-------
-------
EPA-600/9-79-01!
June 1979
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
PROCEEDINGS OF THE NATIONAL SYMPOSIUM
CINCINNATI, OHIO
SEPTEMBER 18-20, 1978
Edited by
Albert D. Venosa
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Re-
search Laboratory, U.S. Environmental Protection Agency, and approved
for publication. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
-------
FOREWORD
The Environmental Protection Agency was created because of increas-
ing public and government concern about the dangers of pollution to the
health and welfare of the American people. Noxious air, foul water, and
spoiled land are tragic testimony to the deterioration of our natural en-
vironment. The complexity of that environment and the interplay between
its components require a concentrated and integrated attack on the
problem.
Research and development is that necessary first step in problem solu-
tion and it involves defining the problem, measuring its impact, and
searching for solutions. The Municipal Environmental Research Labor-
atory develops new and improved technology and systems for the pre-
vention, treatment, and management of wastewater and solid and
hazardous waste pollutant discharges from municipal and community
sources, for the preservation and treatment of public drinking water
supplies, and to minimize the adverse economic, social, health, and
aesthetic efforts of pollution. This publication is one of the products of
that research; a most vital communications link between the researcher
and the user community.
The most important aspect of any successful research endeavor is the
proper transfer of the new or developed technology to the user community.
A research symposium is one mechanism for achieving that objective
quickly and accurately. The proceedings reported herein is a concise
presentation of the latest results from research in wastewater disinfec-
tion technology.
Francis T. Mayo
Director
Municipal Environmental Research Laboratory
111
-------
ABSTRACT
This symposium brought together scientists, consulting engineers,
municipal design engineers, anil various technical, state, and local gov-
ernment officials to listen to presentations on the latest developments
in the field of waste-water disinfection and to actively participate in floor
discussions on the data and interpretations presented therefrom. The
symposium was divided into several main sessions, each session being
limited to a specific topical area. Following each formal presentation,
a 5 minute informal question and answer period was allotted. Following
each session, a 30 to 75 minute panel or round table discussion was held
in order to permit maximum audience participation and complete
thought development. This turned out to be a great success, as many
thoughts flowed forth from the minds of many people who may not have
participated without the proper lime allotment. Some questions were
answered, but many more were raised, which is what should happen
at a gathering of research scientists and engineers actively involved in
public health and environmental research.
Rapid progress is being made in the field of wastewater disinfection,
but much more work is needed he/ore a design manual can beformulated.
It appears that considerable savings in chlorine usage is possible with a
well designed, optimized mixing and contacting system. Dechlorination
with sulfur dioxide is cost-effective, but not with activated carbon or
holding lagoons. Disinfection of well oxidized, filtered secondary effluent
is best achieved with a bubble cliff user contactor and is independent of
contact time. Total costs of ozone disinfection appear to be twice the cost
of chlorine. Ultraviolet light is finally being recognized as an extremely
effective alternative disinfection process, with costs ultimately promising
to be competitive with chlorine. Chlorine dioxide is somewhat disappoint-
ing compared with chlorine on bench-scale analysis, but pilot testing
should reveal the true effectiveness. Very interesting preliminary results
were presented on nonvolatile organic by-product formation by chlorine,
ozone, and ultraviolet light.
A brief comment concerning organization of the proceeding's contents
is in order. The papers are printed in exactly the same order they were
presented, except for the banquet speech. However, most of the printed
material appears in much greater detail than was presented orally. All
extemporaneous discussions were reproduced with as much fidelity as
possible from the original tapes. In transferring the spoken word to the
written word, I exercised a certain amount of editorial prerogative in the
interest of maintaining conciseness and clarity without changing the
thought or intent of the statement.
IV
-------
CONTENTS
Page
FOREWORD iii
ABSTRACT iv
ACKNOWLEDGEMENTS viii
SECTION 1. OPENING REMARKS
1. OPENING SESSION WELCOME,
Albert D. Venosa 1
2. INTRODUCTION AND OBJECTIVES OF SYMPOSIUM,
John J. Convery 2
3. PERSPECTIVES ON WASTEWATER DISINFECTION:
A VIEW FROM HEADQUARTERS,
Courtney Riordan 5
SECTION 2, CHLORINATION/DECHLORINATION
4. FIELD STUDIES IN OPTIMIZATION OF
WASTEWATER CHLORINATION,
Endel Sepp, P. Bao, and A. Custodio 7
5. COMPARISON OF ACUTE TOXICITY OF
CHLORINATED EFFLUENTS FROM OPTIMIZED
AND EXISTING FACILITIES,
B. J. Finlayson and R.D. Hansen 12
6. DESIGN OF A CHLORINATION SYSTEM FOR
LAGOON DISINFECTION,
B.A. Johnson, J.H. Reynolds, J.L, Wight,
and E.J. Middlebrooks 24
7. DECHLORINATION OF WASTEWATER:
STATE-OF-THE-ART FIELD SURVEY AND
PILOT PLANT STUDIES,
H. Gan and C.L. Chen 36
8. ROUNDTABLE DISCUSSION OF
CHLORINA TION/DECHLORINA TION 49
SECTION 3. CHLORINE DIOXIDE
9. EFFECTIVENESS OF CHLORINE DIOXIDE AS
A WASTEWATER DISINFECTANT,
J.D. Berg, E.M. Aietta, and P. V. Roberts 61
-------
Page
10. CHLORINE DIOXIDE: ANALYTICAL MEASUREMENT
AND PILOT PLANT EVALUATION,
E.M. Aietta, B. Chow, and P. V. Roberts 72
11. EFFECT OF PARTICULATES ON INACTIVATION
OF ENTEROVIRUSES IN WATER BY
CHLORINE DIOXIDE,
F.A.O. Brigano, P.V. S carpi no,
S. Cronier, and M.L. Zink 86
12. PANEL DISCUSSION OF
CHLORINE DIOXIDE DISINFECTION 95
SECTION 4. ULTRAVIOLET LIGHT
13. UTILITY OF UV "DISINFECTION" OF
SECONDARY EFFLUENT,
H.W. Wolf, A.C. Petrasek, Jr., and S.E. Esmond 100
14. UV DISINFECTION OF SECONDARY EFFLUENT,
J.D. Johnson, K. Aldrich, D.E. Francisco, T. Wolff,
and M. Elliott 108
15. FIELD SCALE EVALUATION OF ULTRAVIOLET
DISINFECTION,
O.K. Scheible, T. Mulligan, and G. Binkowski 117
16. ROUND TABLE DISCUSSION OF
ULTRAVIOLET DISINFECTION 126
SECTION 5. OZONE
17. COMPARISON OF MPN AND MF TECHNIQUES
OF ENUMERATING COLIFORM BACTERIA IN
OZONATED WASTEWATER EFFLUENT,
M. C. Meckes and A. D. Venosa 136
18. COMPARATIVE EFFICIENCIES OF OZONE
UTILIZATION AND MICROORGANISM REDUCTION
IN DIFFERENT OZONE CONTACTORS,
A.D. Venosa, M.C. Meckes, E.J. Opatken,
and J. W. Evans 144
19. ECONOMIC EVALUATION OF OZONE CONTACTORS,
E.J. Opatken 162
20. FIELD SCALE EVALUATION OF WASTEWATER
DISINFECTION BY OZONE GENERATED FROM AIR,
K.L. Rakness and B.A. Hegg 174
21. FIELD SCALE EVALUATION OF WASTEWATER
DISINFECTION BY OZONE GENERATED
FROM OXYGEN,
J.S. Jain, N.L. Presecan, and M. Fitas 198
22. ROUND TABLE DISCUSSION OF
OZONE DISINFECTION 210
VI
-------
Page
SECTION 6. VIRUSES AND ORGANICS
23. VIRUS INACTIVATION IN WASTEWATER EFFLUENTS
BY CHLORINE, OZONE, AND ULTRAVIOLET LIGHT,
R.A. Fluegge, T. G. Metcalf, and C. Wallis 223
24. EFFECTS OF CHLORINE, OZONE, AND
ULTRAVIOLET LIGHT ON NONVOLATILE ORGANICS
IN WASTEWATER EFFLUENTS,
R.L. Jo/ley, N.L. Lee, W.W. Pitt, M.S. Denton,
J.E. Thompson, S.J. Hartmann, and C.I. Mashni 233
25. MUTAGENIC ACTIVITY OF NONVOLATILE Of3ANICS:
DERIVED FROM TREATED AND UNTREATED
WASTEWATER EFFLUENTS,
R.B. Gumming, L.R. Lewis, R.L. Jolley,
and C. I. Mashni 246
26. PANEL DISCUSSION OF VIRUSES AND ORGANICS 253
SECTION 7. PLANNING AND IMPLEMENTATION
27. PLANNING DECISIONS IN SELECTING WASTEWATER
EFFLUENT DISINFECTION FOR THE OLENTANGY
ENVIRONMENTAL CONTROL CENTER,
C.F. Grissom 258
28. THE CONSULTANT'S VIEWPOINT ON ALTERNATE
DISINFECTANTS,
J.L. Pavoni, M.W. Tenney, M.E. Tittlebaum,
B.F. Maloy, and J.L. Kochert 263
APPENDICES
A. BANQUET PRESENTATION: THE LAW'S RESPONSE
TO PUBLIC HEALTH HAZARDS,
J. V. Karaganis 269
B. SYMPOSIUM RAPPORTEUR
E.J. Middlebrooks
Dean, College of Engineering
Utah State University
Logan, Utah 284
Vll
-------
ACKNOWLEDGEMENTS
The editor acknowledges the perseverance and dedicated efforts of
Mr. Larry Dempsey of the Environmental Research Information Center,
who arranged for the hotel and banquet accommodations and coordinated
registration and other administrative activities. Special thanks are offered
to Mrs. Janice Bader and Mrs. Rita Bender, who donated almost three
days of their time at the registration desk. 1 also wish to thank the session
moderators, Messrs. J. J. Convery, E. F. Barth, E. J. Opatken, M.'C.
Meckes, and Dr. R. L. Bunch, for having maintained all presentations
within the scheduled time frames. I wish to express particular thanks to
Dr. Jessica Barren at the Environmental Research Information Center
for her able assistance in arranging for the editing and typesetting of the
manuscript.
Special appreciation is given to the speakers and authors of the papers
for their many hours of labor and preparation, and to the geneiai regis-
trants whose lively participation in the panel and round table discussions
contributed greatly to the success of the symposium.
Vlll
-------
SECTION 1. OPENING REMARKS
1.
OPENING SESSION WELCOME
Albert D. Venosa
Research Microbiologist
Wastewater Research Division
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
It is a pleasure to welcome all of you to the U.S.
Environmental Protection Agency's National Sym-
posium on Wastewater Disinfection. This symposium
is being sponsored by USEPA's Municipal Environ-
mental Research Laboratory (MERL) of the Office
of Research and Development, and the Environr
mental Research Information Center, Cincinnati,
Ohio. The objectives and goals of the symposium.
will be presented in a few moments by Mr. John J.
Convery, Director of MERL's Wastewater Research'
Division.
I would like to highlight the symposium's program,
of which all of you should have a copy. We have or-
ganized the program into seven major sessions, each
of which deals with a specific topic or aspect of waste-
water disinfection research. In the first session you
will be introduced to the Acting Deputy Assistant
Administrator of the Office of Air, Land, and Water
Use, Dr. Courtney Riordan, and MERL's Wastewater
Research Division Director, Mr. John J. Convery.
The subsequent sessions will be technical in scope and
will involve investigators who have been granted EPA
funds to conduct wastewater disinfection research.
Each of the first four sessions will deal exclusively
with a specific disinfectant. Thus, Session 2 concerns
chlorination-dechlorination, Session 3 chlorine
dioxide. Session 4 ultraviolet light, and Session 5
ozone. Session 6 involves two studies which support
all the disinfection projects, i.e., indigenous virus
inactivation and nonvolatile organic compound
formation by chlorine, ozone, and UV light. The final
session was included to provide insight into how plan-
ning decisions on implementation of new disinfection
technology are made at both the EPA Regional level
and the municipal consultant's level. In order to per-
mit the maximum amount of audience participation,
each session will be concluded by a separate panel
or round table discussion. You are encouraged to air
your views to the fullest during these discussions.
Following the last panel discussion, Dr. E. J.
Middlebrooks, Dean of the College of Engineering,
Utah State University, an able environmental re-
searcher himself, will summarize the findings of the
symposium and provide perspective into what lies
ahead for EPA's Wastewater Disinfection Program.
On Tuesday evening, there will be a social hour
and banquet, beginning at 6:00 P.M. I am certain
you will enjoy listening to the guest speaker,
Mr. Joseph V. Karaganis, the Special Assistant
Attorney General of the State of Illinois, who will
discuss the microbial aspects of the case in which
he won a landmark decision for the State of
Illinois against The City of Milwaukee. The title of
his paper is "The Law's Response to Public
Health Hazards."
All presentations and panel discussions will be
recorded and published in a proceedings. I hope
you enjoy the symposium and encourage you all to
participate actively.
-------
2.
INTRODUCTION AND OBJECTIVES OF SYMPOSIUM
John J. Con very
Director, Wastewater Research Division
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
On behalf of the Municipal Environmental Re-
search Laboratory of the U.S. Environmental
Protection Agency, I welcome you to this National
Symposium on Wastewater Disinfection. The basic
purpose of this symposium is to share with you the
most up-to-date research findings of our disinfection
research program and thereby accelerate the practical
application of these findings to the construction grants
program. A corollary objective is to obtain feedback
from you on your ideas or problems which can be
incorporated into our future research program. The
Office of Research and Development of EPA wants
to be responsive to the needs of you, the practitioners
of pollution control technology.
During the next three days there will be opportun-
ities for audience participation. I would like to second
Al Venosa's invitation to participate in these discus-
sions. If additional thoughts on disinfection research
needs occur to you after the symposium is finished,
please share them by writing to me.
This is the second effort of the disinfection research
staff to share timely research findings through the
medium of a public meeting. In October 1974, we
held a workshop in Wyoming, Michigan, where the
results of continuous fish bioassays on chlorinated,
chlorinated/dechlorinated, ozonated, and bromine
chloride treated wastewaters were presented. The
primary interests at that time were the relative degrees
of disinfectant-induced fish toxicity and the methods
of reducing or eliminating the toxicity. The prevention
of fish toxicity is still an important design objective
and a fundamental reason for our interest in improved
chlorination and dechlorination. The current EPA
criteria for total residual chlorine to prevent fish
toxicity is 2.0 pg/1 for salmonid fish and lO.OMg/1
for other freshwater and marine organisms (3).
Several events have occurred since 1974 which make
the task of developing acceptable disinfection tech-
nology more difficult for all of us. Widespread occur-
rence of the formation of trihalomethanes in chlo-
rinated waters and recognition of their potential
health effects together with passage of the Toxic Sub-
stances Act (PL94-469) and the signing by EPA of
the consent degree (2) to control 129 priority pol-
lutants within 3 years, have added analytical chemistry
and toxicological screening requirements to our evalu-
ation of disinfection alternatives. This is particularly
true for potential reuse situations. A significantly
expanded analytical methods development program is
underway to permit surveying the occurrence of pri-
ority pollutants in municipal raw wastewaters, process
influents and effluents, and sludges. Recommended
analytical methods should be available by the first
of November, 1978. The Health Effects Research
Laboratory, EPA, Cincinnati, is currently evaluating
a variety of toxicological screening tests which may
be useful in evaluating the efficacy of alternative dis-
infectants. A recent paper by Bull, Kopfler and
McCabe (1) describes these tests, which 1 have listed
for your information.
ACUTE TOXICITY -MEDIAN LETHAL
DOSE (LD50)
MUTAGENICITY -AMES TEST:
Salmonella typhimurium
strains (bacteria)
SRI TEST;
Saccharomvces
cerevisiae (yeast)
CARC1NOGENICITY -IN VITRO
-------
OPENING REMARKS
TERATOGENICITY
MAMMALIAN CELL
CULTURE WITH
TRANSFORMED
CELLS INJECTED
INTO MICE TO NOTE
TUMOR
PRODUCTION
-SKIN TUMOR
RESPONSE IN
SENSITIZED MICE
(SEN-CAR MOUSE)
-IN VIVO ASSAY
USING NEONATAL
RATS
-FEEDING OF
PREGNANT RATS
AND NOTING
MORPHOLOGIC
CHANGES IN THE
NEW-BORN
Since 1974, the disinfection program has spent
approximately $3 million on projects to develop and
test alternative disinfection approaches. Many of the
facilities involved are "among the first of their kind"
treatment plants. In this category I would include:
the Estes Park, Colorado, ozonation with air facility;
the Meander, Ohio, ozone with oxygen facility; and,
the Northwest Bergen County ultraviolet light treat-
ment facility. Obviously, most of our limited resources
are used to monitor and evaluate performance rather
than pay for capital facilities.
The amendments to the Clean Water Act (PL95-
217) which passed last Fall, include a provision for
innovative and alternative process or system designs
which permits, after October 1, 1978, 85% construc-
tion grant financing- of the capital requirements and
provides for a 100% replacement or modification
guarantee. To qualify as innovative technology, a
process must save 15% of the total life costs or
20% of the energy requirements compared to con-
ventional technology. This provision illustrates two
other significant elements of concern in comparing
disinfection alternatives-energy utilization and total
treatment costs. Comparisons cannot be made accu-
rately unless the same endpoint or treatment objec-
tive is stated. This brings me to my final point. In
reading through disinfection literature I am always
impressed with the number of variables that need
to be recognized, and the necessity of their meas-
urement or identification to permit meaningful com-
parison of results. The following list may not be all
inclusive but it serves to illustrate my point.
DISINFECTANT type (chlorine), form (mono-
chloramine)
DOSAGE APPLIED - DISINFECTANT DE-
MAND = EFFECTIVE DOSAGE
QUALITY OF THE FEED WATER (AND COST
IF A NECESSARY PRETREATMENT RE-
QUIREMENT FOR A PARTICULAR DIS-
INFECTANT)— Suspended solids, color, COD,
pH, NO2, NH3, H2S.
BACTERIOLOGICAL QUALITY MEASURE-
MENT — indicator (coliform, MPN, MF),
pathogen (bacteria or virus [wild or cell cul-
tured]), density objective, point of measuring
effluent quality, opportunity for regrowth or
photoreactivation, influent density effect on %
removal vs. number remaining.
CONTACTOR — mixing conditions (GT), type
(pressure or open), geometry (length to width
ratio), real vs. theoretical detention time.
My purpose in raising the issue of variables and
comparable treatment objectives is to encourage more
complete descriptions of your evaluations and thereby
permit more meaningful comparisons and the rational
assignment of pre-and post-disinfection process costs
when comparing disinfectants. Examples might be the
cost of chemical clarification and filtration to meet
a 2.2 MPN/ 100 ml coliform disinfectant objective
or the reaeration costs to meet a dissolved oxygen
requirement for a dechlorination facility.
1 hope you experience an enjoyable and worthwhile
symposium.
REFERENCES
I. hull. R. J.,ctal. 1978. Toxicity and Mutagenic Effects ol Organ-
ics. Presented I97X Annual AWWA Conference, Atlantic
City. N.J.
2. Natural Re^out'ct'*
(D.D.C. 1976).
Council vs. li'iiin. H ERC 2120
3. Quality Criteria lor Water. 1976. U.S. Environmental Protection
Agency. Washington. IXC. 20460.
DISCUSSION
DR. HARVEY ROSEN, Union Carbide Corpor-
ation: I would just like to reiterate your last point.
I think it is very important.
Recently, in a National Academy of Sciences study.
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
investigators had been trying to collect information the reporting is evident. Certain specific information,
to do meaningful comparisons among different disin- needed to determine the validity of the results, is miss-
fectants as well as comparisons using a single disin- ing. This is becoming very important in terms of all the
fectant in different quality waters. They are discovering money that is being spent in this area since results that
that, although there are thousands and thousands of can be used and understood by everybody on a corn-
reports dealing with these studies, lack of quality in mon basis are not available.
-------
3.
PERSPECTIVES ON WASTEWATER DISINFECTION
- A VIEW FROM HEADQUARTERS
Dr. Courtney Riordan
Acting Deputy Assistant Administrator
Office of Air, Land, and Water Use
U.S. Environmental Protection Agency
Washington, D.C. 20460
Before beginning my prepared remarks, 1 would like
to second the welcome to the Symposium.
It really warms my heart to see all these people here
because one of the most sensitive areas that we deal
with in headquarters in terms of our research program
in EPA is our delivery system. That is the one that is
constantly questioned. By that I mean, how do we
interact with our user groups? How do we get feedback'
from our user groups and ultimately influence the
character of our research program?
Looking at the group we have convened today and
studying the program, I think it is going to be a very
productive conference. It gives me the kind of first-
hand experience that greatly facilitates returning to
Washington and pointing out that we do have some
fairly effective mechanisms for communicating our
programs to the public.
On July 26, 1976, the U.S. Environmental Protec-
tion Agency made the decision to delete the fecal
coliform bacteria limitations from the definition of
secondary treatment (40 CFR 133). The intent was to
permit reliance on site-specific water quality standards
to establish disinfection requirements for municipal
wastewater treatment plants rather than mandate con-
tinuous, year-round disinfection of all effluents. The
detailed rationale behind this decision has been pub-
lished and need not be discussed today. It is important
to consider, however, the impact this decision has had
both on the Municipal Environmental Research
Laboratory's (MERL) Wastewater Disinfection R&D
Program and on the individual statesf reactions.
Public Law 95-217 states that, wherever attainable,
water quality which protects fish, shellfish, and wild-
life and provides for recreation will be achieved by
July 1, 1983. Two important questions relative to the
1983 goals are raised in connection with the disinfec-
tion deletion from the Secondary Treatment Regula-
tion. First, is it possible to comply with the 1983 goal
of swimmable waters without universal, year-round
disinfection? Secondly, how can the goal of providing
for the protection and propagation of fish, shellfish,
and wildlife be satisfied when disinfection is required
to maintain conditions suitable for recreation?
Water quality standards are presently being re-
viewed and revised to define conditions necessary
to meet the 1983 goal. Reliance on state water
quality standards in lieu of fecal coliform bac-
teria limitations in the Secondary Treatment
Regulation is consistent with the achievement of
the 1983 goal, because water quality standards
are established, in part, to allow recreation in
and on the water. Implementation of water
quality standards for recreation in and on the
water can result in the following types of
situations:
(a) Where water quality standards 'are estab-
lished to allow whole body contact recreation,
disinfection of wastewater discharges will gener-
ally be necessary to meet bathing water criteria,
such as 200 fecal conform bacteria per 100 ml
over a 30-day period. In instances where bathing
waters are used seasonally, water quality stan-
dards for whole body contact may not apply
during non-use periods and, consequently, year-
round wastewater disinfection may not be re-
quired.
(b) Where water quality standards are estab-
lished for secondary contact recreation purposes
such as boating, in-stream criteria may be as
high as 1000 to 5000 fecal coliform bacteria per
100 ml. In these instances, wastewater disinfec-
tion may not always be needed to meet the re-
quired in-stream standards.
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
(c) In accordance with regulations pertaining to
the establishment of water quality standards
(40 CFR 130), the states may, under certain pre-
scribed conditions, set standards at levels less
stringent than the national water quality goal.
Wastewater disinfection will generally not be
necessary in these instances.
Similarly, water quality standards for total residual
chlorine can be established to address the potentially
adverse effect of chlorination on aquatic plants and
organisms. In situations where the needs for disinfec-
tion and protection of aquatic organisms co-exist,
several alternatives are available. First, sufficient
mixing and dilution may be available to meet in-
strcam chlorine criteria, particularly where chlorina-
tion processes are properly designed and operated.
Dechlorination is also available to eliminate the resid-
ual chlorine from disinfected effluents prior to dis-
charge. Finally, alternative disinfectants, such as
o/one or ultraviolet light, may prove to be cost-effective
where stringent effluent standards are specified.
As achievement of Public Law 95-217 by the 1983
goal becomes imminent, the Agency will be in a better
position to re-evaluate the disinfection requirements
for municipal wastewatcr discharges in consideration
of the improved water quality at that time. In the in-
terim, time will be available for further investigation
of cost-effective disinfection processes and analysis of
more conclusive data on the relative benefits and risks
resulting from wastewater disinfection.
Individual Sidles Reactions
When the decision had been made to delete the fecal
coliform limitation, it was intended that the states'
water quality criteria would control disinfection policy
on a case-by-case basis. In some instances the criteria
established for some steams have been too lax, while
in others too strict, relative to the probable transmis-
sion of communicable disease through direct human
contact with contaminated waters.
One such state, the State of Mew York, has estab-
lished 14 different stream classifications based on
water quality relative to use and proximity of human
contact or shellfish contact with contaminated dis-
charges. There are five separate classifications for
fresh surface waters, four classifications for marine
waters, four special classifications related to specific
locales, and one classification established for enjoy-
ment of water in its natural conditions. For each of
these classifications a separate coliform standard is
established with standards ranging from 0 discharge
to a high of 10.000 total coliforms per 100 ml. It is
hoped that other states will follow suit in this regard.
MERl. Wasii'watei' Disinfection R&D Program
The MERL Wastewater Disinfection R&D Pro-
gram is essentially a process development effort in
which the major responsibility is to develop alternative
technology useful to design engineers and municipali-
ties to decide how best to achieve a stated microbio-
logical discharge limitation. Elimination of the Fed-
eral requirement for disinfection of secondary effluent
has led to the additional responsibility for the MERL
Wastewater Disinfection Program to develop the
capability of the alternative technology to achieve
different levels of effluent bacterial numbers rather
than a fixed national goal of 200 fecal coliforms 100
ml. Depending on the water quality standard of the
receiving stream, effluent limitations may range from
as high as 5000 fecal coliforms 100 ml to as low as 2.2
total coliforms 100 ml. Thus, decisions as to which
alternative technology to implement in a particular
receiving stream could be significantly affected by the
microbial limitations established for it.
Research efforts of the MERL Wastewater Disin-
fection Program have demonstrated the process re-
quirements necessary to achieve flexible coliform
limitations. Some of the data will be presented at this
symposium. Other data will be available in the upcom-
ing months. The Municipal Environmental Research
Laboratory has implemented a combination of in-
house and extramural expertise for research and
development to meet these objectives. This symposium
is a hallmark to that effort.
-------
SECTION 2. CHLORTNATION/DECHLORINATION
4.
FIELD STUDIES IN OPTIMIZATION OF WASTEWATER CHLORINATION
E. Sepp, P. Bao and A. Custodio
California Department of Health Services
Berkeley, California
INTRODUCTION
California Department of Health Services is
conducting field studies on optimization of waste-
water chlorination. The project is partially fi-
nanced by the U.S. Environmental Protection
Agency (EPA) and is a joint venture with the
Department of Fish and Game and State Water
Resources Control Board. Three mobile units par-
ticipate in the study; a chlorination pilot plant, a
fish bioassay laboratory and a public health lab-
oratory. Results from the pilot plant are com-
pared to results obtained from parallel studies on
the full scale plants. Performance is measured by
means of coliform bacteria and chlorine residual
tests, as well as various chemical analyses.
Objectives of the project are: reduction of
chlorine induced toxicity; savings in chlorine use
by improved design; and, writing of a design
manual.
MATERIALS AND METHODS
The chlorination pilot plant incorporates opti-
mized features including rapid initial mixing,
automatic control of chlorine dosage by residual
control, and plug flow contact chambers provid-
ing a variety of contact times. Samples are ana-
lyzed for total coliform bacteria using 5 tubes per
dilution. All analyses are done according to Stan-
dard Methods for the Examination of Water and
Wastewater (1). Bacteriological and chemical
sampling is confined to the 8-hr, daytime period.
Flow-through bioassays, done by Department of
Fish and Game, are run continuously for 96 hours.
RESULTS AND DISCUSSION
The first study was done at an activated sludge
plant at San Leandro. The plant flow consists of
almost 50% industrial wastes and the plant is
overloaded. The chlorination system consists of a
manually controlled chlorinator and a spirovortex
chlorine contact tank. Chlorine is injected
through a diffuser into a pipe about 30m (100 feet)
ahead of the contact tank. The pipe is under
pressure; therefore, initial mixing probably is
good. The BOD of the effluent varied from 12
to 29 mg/1, and COD from 69 to 109 mg/1. The
coliform levels were extremely high.
The first study was beset with problems. Dur-
ing maximum flow periods there was a heavy
carryover of sludge and grease from the final
clarifier which clogged up the pilot plant. Conse-
quently, the pilot plant could be operated only
part-time. Bacterial results from the pilot plant
were not remarkably better than those from the
full scale plant. This was inexplicable and disap-
pointing because the contact tank had a t* less
than 10 minutes and suspended solids content was
high (32-102 mg/1). Approximately the same coli-
form bacteria levels were reached with chlorine
dosages of 10-11 mg/1 in the pilot plant and
10-12 mg/1 in the full scale plant. Chlorine resi-
duals were 5-6.5 mg/1 and 6-11 mg/1, respective-
ly, in the two plants after the same theoretical
contact times.
ban Pablo has an activated sludge plant which
also uses a roughing trickling filter and complete
*t. is the time (minutes) for the initial appearance of dye in a dye study.
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
nitrification. The plant receives appreciable
amounts of industrial waste. Effluent quality is
good, with a BOD of 4-5 mg/1 and suspended
solids < 10 mg/1 most of the time. Furthermore,
the quality is fairly constant and the coliform
bacteria levels are low. The chlorination system
consists of a compound loop chlorine control sys-
tem, a hydraulic jump for initial mixing and two
parallel rectangular baffled chlorine contact tanks
with a L/W ratio of 15:1. The contact tank has
a tj of about 30 minutes. Effluent is dechlorina-
ted with S02.
Bacterial results indicated that the same level of
disinfection was obtained with a slightly lower
chlorine dosage in the pilot plant than in the full
scale plant. The chlorine dosages were 8-9 mg/1
for the pilot plant and 9.5-10.5 mg/1 for the full
scale plant. Chlorine residuals at the end of con-
tact were about 1.5-2.0 mg/1 for both plants. The
detention time in the pilot plant was slightly
smaller than in the full scale plant.
The Pinole plant uses activated sludge for
treatment. Chlorination system consists of com-
pound loop chlorine control system, two turbine
mixers and two parallel chlorine contact tanks
with a L/W ratio of 40:1. The waste is entirely
domestic. However, the chlorine flow to each
tank is poorly balanced and during low flows at
night the chlorine residual drops to zero. Further-
more, there is leakage through the baffles in the
contact tank resulting in a t; of only about 15
minutes. Effluent is dechlorinated with S02. The
effluent BOD varied from 8 to 29 mg/1, and sus-
pended solids from 2 to 60 mg/1. There was con-
siderable amount of nitrification, and nitrite con-
tent was fairly high (1.0-2.9 mg/1). The chlorine
demand usually was very high in the morning
and dropped to a low level in the afternoon. The
coliform bacteria content varied from day to day
and week to week.
uue 10 tne varying effluent quality, the results
were difficult to analyze; however the chlorine
dosages to achieve the same bacteria levels were
1.5-3 times higher for the full scale plant than
for the pilot plant. The chlorine residuals were
3.4-3.7 mg/1 for the pilot plant and 5.5-13.6 mg/1
for the full scale plant after detention. The cor-
responding dosages were 7.5-27.7 mg/1 for the
pilot plant and 27-47 mg/1 for the full scale plant.
Pilot plant studies were also done at each plant
using different chlorine dosages to obtain differ-
ent effluent bacteria levels. An MPN level of 2
per 100 ml after 60 minutes contact was achieved
with the following chlorine dosages and control
residuals, respectively: 22 mg/1 and 11.5 mg/1 at
San Leandro; 10-13 mg/1 and 7-8 mg/1 at San
Pablo; and 16.2-22.8 mg/1 and 4.6-5.5 mg/1 at
Pinole (one week only). An MPN level of 23 per
100 ml after 30 minutes contact was obtained
with the following dosages and residuals: 22 mg/1
and 11 mg/1 at San Leandro; 10 mg/1 and 7
mg/1 at San Pablo; and 5.7-8 mg/1 and 2-3.7
mg/1 at Pinole.
Various authors (2,3) have indicated that bac-
terial survival ratio is depende'nt on the product
of chlorine residual and detention time. Collins
(2) developed the following empirical formula:
N N -F^lfl
1 °"L 3 J
where NJ - bacteria number in chlorinated effluent
NQ = bacteria number in unchlorinated
effluent
RQ = chlorine residual, mg 1, after mixing
(the "control" residual)
T = detention time, minutes
b = slope of the curve
a = a coefficient
This formula piots as a straight line on log-log
paper. Collins found that the slope b is -3 for
Lplug flow and -1.5 for completely mixed flow.
The ratio Nj/NQ is the bacterial survival ratio. In
this study the residual after contact was desig-
nated as R, without subscript.
Plots of bacterial survival ratio against RT for
pilot plant and for full scale plant for the three
studies are shown in Figures 1 through 6. The
solid line represents the theoretical plug flow
curve. The data points are daily median values.
The pilot plant data include variation in detention
times from 15 to 60 minutes. Computer regres-
sion analysis yielded the following values:
San Leandro plant 30.1 -1.53 0.68
Pilot plant at San Leandro 5.7 -3.3 0.77
San Pablo plant 0.1 -1.1 0.43
Pilot plant at San Pablo 14.6 -4.1 0.93
Pinole plant 0.2 -1.35 0.40
Pilot plant at Pinole
0.16 -1.71 0.66
r = correlation coefficient.
-------
CHLORINA TION/DECHLORINA TION
10
-3
- 10
.-4
cc
I
cc
CC -(
P 10
o
o
10
-7
10 100 1,000
RoT(MG/LxMIN.)
10
10'
-2
<
CC
< -4
1 10
CC
cc .5
P 10
o
o
10
.-6
10 100 1,000
RT (MG/LxMIN.)
Figure 1. Coliform Survival in Pilot Plant (San Leandro) Figure 3. Coliform Survival in Pilot Plant (San Pablo)
10
-3
- 10
.-4
<
CC
I
CC
en
O
O
10
-7
10 100 1,000
RT(MG/LxMIN.)
10
-2
- 10
-3
<
CC
< -4
> 10
CC
o
o
IO
.-6
IO IOO 1,000
RxT(MG/LxMIN.)
Figure 2. Coliform Survival in San Leandro Plant
Figure 4. Coliform Survival in San Pablo Plant
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
o
Z
*.
g
rr
|
rr
CO
5
rr
e
_j
o
o
IU
Id3
.d4
.d5
,n6
\
V,
V|
V
\l
•
«\ .
V*
l\
\°
o \
&)
, \
10
-2
I 10 100 1,000
RT (MG/LxMIN.)
Figure 5. Coliform Survival in Pilot Plant (Pinole)
The regression curves for the pilot plant differ
considerably from the -3 slope for theoretical
plug flow. One reason for this may be the rapid
variations in effluent quality at some of the
plants. The regression curves for the full scale
plants have small correlation coefficients (r)
reflecting the fact that only small variations in
detention times are available.
o
Z
10
v-3
<
rr
< -4
> 10
rr
CO
rr -5
£ 10
_
o
o
1 10
-6
00
10 100 1,000
RT (MG/LxMIN.)
Figure 6. Coliform Survival in Pinole Plant
REFERENCES
1. American Public Health Association. 1975. Standard Meth-
ods for the Examination of Water and Wastewater,
14th ed. American Public Health Association, Inc.,
New York.
2. Collins, H.F., and Selleck, R.E., "Process Kinetics of
Wastewater Chlorination." SERL Report No. 72-5,
Univ. of California, Berkeley, (Nov., 1972).
3. Sanitation Districts of Los Angeles County, "Pomona
Virus Study," California State Water Resources Control
Board, Sacramento, (June, 1977).
SUMMARY AND CONCLUSION
Results of three Chlorination studies have been
discussed. At two plants, the optimized pilot
plant was able to disinfect the effluents with
lesser chlorine dosage required than the full scale
plants. Dosages and residuals to disinfect to a
particular MPN level appear to depend on ef-
fluent quality, including the initial coliform bac-
teria densities. Bacterial survival ratio appears to
be a function of the product of chlorine residual
and contact time, but is further influenced by
changes in effluent characteristics.
DISCUSSION
MR. GEORGE WHITE: It would be a big help
if we could get you scientists, when you are
making those graphs, to put a ball park range on
N( because N} coliform count is probably the
primary criterion which demonstrates the quality
of the effluent and how difficult it will be to dis-
infect, whether you are using chlorine or some
other disinfectant. If you had a ball park range
on each one of those graphs, you could apply the
formula. This is the first thing you have to know
when you are trying to get down into the
10
-------
CHLORINA TION/DECHLORINA TION
240/100 range or the 23 or the 2.2 range . . .
You have to know what the initial number of the
coliforms is before you hit it with the disinfec-
tant. If you could just put those ball park figures
up in the lefthand corner of those graphs it
would be very helpful.
MR. SEPP: I think this is a very good point.
Coliform bacteria content seems to be the main
criterion.
MR. OPATKEN: What is your definition of
low levels of coliforms? Somewhere during pre-
sentation of your paper you referred to a low
level of coliforms.
MR. SEPP: Low level was 100,000 to 500,000
per 100 ml.
MR. VENOSA: Were the three treatment
plants that you studied capable of meeting the
coliform standard that was established for those
particular plants, and was the pilot plant able to
do better?
MR. SEPP: Yes, but they over chlorinated be-
cause usually their effluent levels were very low.
All their requirements were 240 MPN/100 ml.
MR. VENOSA: All three plants?
MR. SEPP: Their actual levels were more like
10/100 ml.
MR. VENOSA: If I remember correctly, it was
at the Pinole plant that only about a 4 log reduc-
tion was achieved, and I was wondering how
many organisms that represented?
MR. SEPP: It varied . . . sometimes it was
around 50,000, other times there were around
10,000,000. It varied through that range and eacn
week there was some change. It was partly
caused by use' of the community. When school
was out, bacteria levels dropped. When school
was in, bacterial counts were higher.
DR. HARVEY ROSEN: I would like to say
that we reached similar conclusions with ozone. I
am pretty sure you will find that with most disin-
fectants the initial coliform count is extremely
important in determining the amount of disinfec-
tant, assuming that you have a well mixed reac-
tor and other things. The exception is when you
are designing a system; then, it is very difficult
to know what the expected level of coliforms is
going to be. From a practical point of view you
then have, to relate coliform levels to effluent
quality. The better the effluent quality, generally,
the lower the level of coliforms. From a point of
view of design, you then have to generally know
something about or make certain assumptions
about the treatment processes that, preceding dis-
infection, will produce a specific BOD, suspen-
sion solids, and COD in order to estimate and
evaluate a specific dose of disinfectant. It
would be nice if we were able to pilot all of
these things and know what kind of coliform
count we expected from these various treatment
processes. That is the best index, as we have
found, although you do need some other things
as tools.
MR. WHITE: This goes back to the same thing
that I said about putting the range of the coli-
forms on the graphs. This is one of the things
that I tried to do when I was writing the second
book which, incidentally, will be out in October.
I have it all in there so it will answer Harvey
Rosen's question. I collected as much data as I
could about that Nj from all the different types
of effluents. I just put it down in a range for
well oxidized secondary, for primary, which I do
not think you can disinfect, and for tertiary with
and without sedimentation and filtration. It is all
in there and, of course, the more data we collect,
the better it will be. Good data is very hard to
come by.
I might also add that one thing people over-
look when they start talking about disinfection of
wastewater, is that the operator has to have all
of his unit processes working before the disinfec-
tion system will really take over. This is why dis-
infection, if nothing else, is a good monitor of
how well the operator is operating his plant.
11
-------
5.
COMPARISON OF ACUTE TOXICITY OF CHLORINATED EFFLUENTS
FROM OPTIMIZED AND EXISTING FACILITIES
Brian J. Finlayson and Richard J. Hansen
California Department of Fish and Game
Fish and Wildlife Pollution Control Laboratory
Rancho Cordova, CA 95670
ABSTRACT
The California Department of Health in cooperation with the
California Department of Fish and Game developed and imple-
mented a chlorine optimization study which investigated several de-
sign criteria that may improve, the efficiency of wastewater chlori-
nation systems, and hence, provide adequate disinfection without
excessive chlorination. The study has been conducted on site at
several wastewater treatment plants in northern California. Two
mobile units were constructed for the project; a pilot chlorination
plant and a mobile water quality and bioassay laboratory were
operated by the Department of Health and the Department of Fish
and Game, respectively. The pilot chlorination plant tested several
optimized chlorination design criteria against existing wastewater
treatment plant chlorination systems. The bioassay laboratory tested
the toxicities of the optimized and existing chlorinated effluents.
The results obtained thus far indicate that no toxicity was asso-
ciated with either the existing unchlorinated or dechlorinated efflu-
ents. The optimized chlorinated effluents had lower and more
stable chlorine residuals than did the existing chlorinated effluents
and hence, were less toxic. The toxicity of all effluents investigated
increased proportionately with increased chlorine residual. The toxi-
city of chlorine in wastewater was lowered when denitrification was
practiced prior to chlorination.
INTRODUCTION
The toxicities of chlorine and chlorinated threespined stickleback (Gasterosteus aculeatus).
wastewater effluents to fish and other aquatic life The U.S. EPA (18) has recommended that
are well documented. Many investigators [Zillich "safe" chlorine residuals in the environment
(20); Arthur et al (2); Brungs (4,5); Ward et al should not exceed 10 pg/1 (0.01 mg/1) for the
(19); Mattice and Zittel (12); and Finlayson and protection of aquatic life. Even lower levels of
Hinkelman (10)] have clearly demonstrated that chlorine than those recommended have been re-
chlorine is highly toxic to most aquatic life, ported to impair growth and development of
Lethal concentrations of chlorine to fish falls some aquatic biota [Roberts et al. (15)].
between 23 yg/1 (0.023 mg/1) for rainbow trout The disposal of chlorinated wastewater effluents
(Salmo gairdneri) and 500 p g/1 (0.5 mg/1) for in the environment has created a long-standing
12
-------
CHLORINA TION/DECHLORINA T1ON
conflict between public health and conservation
agencies. Health agencies are concerned with the
adequate disinfection of wastewater effluents for
the protection of public health; whereas! conser-
vation agencies are responsible for protecting the
aquatic resources in the environment. Section 5650
of the State of California Fish and Game Code
(1977) prohibits the discharge of any substance
deleterious to fish, plant or bird life into the
waters of the State; consequently, confrontations
between public health agencies and conservation
agencies over the discharge of chlorinated waste-
waters are common.
Our primary concern with the chlorination of
wastewater has been that chlorine is not always
carefully and efficiently applied at wastewater
treatment plants. We have observed the discharge
of effluents into the environment with total resi-
dual chlorine (TRC) concentrations in excess of
10 mg/1. This level is 1000 times the recom-
mended "safe" level. Usually, excessive chlorine
residuals in a discharged waste are the result of
improper chlorine application, an ineffective resi-
dual chlorine control system, or both.
Because of this need to optimize wastewater
chlorination systems, we welcomed the oppor-
tunity to participate in a Chlorination Optimiza-
tion Study which may help to minimize chlorine
residuals discharged into the environment. We
point out, however, successful chlorine optimiza-
tion only solves a portion of the problem.
Chlorinated effluents usually must be subjected to
some form of dechlorination prior to discharge so
that adequate protection is afforded to the biota
using the receiving water.
The purpose of the Chlorination Optimization
Study is to evaluate the influence and importance
of three chlorine application optimization criteria.
These design criteria originally developed by
Collins and Deaner (6) are:
1. rapid and complete initial mixing between
chlorine and waste;
2. a 30-min minimum contact time between the
chlorine and waste with no further
mixing; and
3. a sound and workable chlorine residual
control system.
The Department of Health (DOH) has built a
mobile chlorination pilot plant for the study and
has used the plant to test the chlorine application
optimization criteria against existing wastewater
chlorination facilities. The Department of Fish
and Game's objective in the study is to docu-
ment the comparative toxicities between the opti-
mized and existing chlorinated effluents from
several different wastewater treatment facilities
under study by DOH. We accomplished this by
means of a specially designed mobile laboratory
capable of simultaneously monitoring the water
quality and toxicity of three separate effluents:
1. existing WTP unchlorinated (UnCl)
effluent;
2. existing WTP chlorinated (EC1) effluent; and,
3. DOH optimized pilot plant chlorinated
(DOHCL) effluent.
During periods when the DOH pilot plant was
not in operation, toxicity and water quality data
were also collected from existing dechlorinated
(DeCl) effluents.
To date, both units have participated in com-
parative studies at several wastewater treatment
plants (WTP) in California. Since the project is
on-going, this paper is only a progress report.
The final report will not be available until after
June 1979. Our present analyses of the data
demonstrate the differences between:
1. magnitude and variability of chlorine resi-
duals in optimized and existing chlorinated
effluents;
2. toxicity of optimized and existing chlori-
nated effluents;
3. toxicity of chlorine in ammoniated and
ammonia-stripped (denitrified) effluents; and,
4. toxicity of chlorine to the two test fish,
fathead minnow (Pimephales promelas)
and golden shiner (Notemigonus crysoleucas).
MATERIALS AND METHODS
Project Schedule
To date, both units have been in operation at
four wastewater treatment plants in northern
California (Table 1). We have collected compara-
tive toxicity data between optimized and existing
chlorinated effluents for only two plants, San
Pablo and Pinole. The pilot plant was not in
operation at San Leandro and although recent
comparative toxicity data are now available from
the South San Francisco facility, they have not
been analyzed yet.
Mobile Laboratory
The DFG mobile water quality and toxicity
13
-------
PROGRESS [N WASTEWATER DISINFECTION TECHNOLOGY
TABLE 1. SCHEDULE FOR BIOASSAY TESTING AT VARIOUS WASTEWATER TREATEMENT PLANTS IN
CALIFORNIA DURING THE PERIOD OF FEBRUARY THROUGH AUGUST 1978.
Plant Location
South San
Francisco
Date Begin
Elluent Type
Test Species
(b)
Aug. 14, 78
ECI
FH, GS
Series No.
San Leandro
San Pablo
Pinole
Feb.
Feb.
Apr.
Apr.
Apr.
Jun.
Jun.
Jun.
Jul.
13,
27,
10,
17,
24,
12,
19,
26,
8,
78
78
78
78
78
78
78
78
78
UnCI,
UnCI,
UnCI,
UnCI,
UnCI,
UnCI,
UnCI,
UnCI,
ECI
ECI
ECI
ECI,
ECI,
ECI,
ECI,
ECI,
ECI,
DOHCI
DeCI
DOHCI
DOHCI
DeCI
DOHCI
FH
FH
FH
FH
FH
FH
GS
GS
FH, GS
SL-1&2
SL-3&4
SP-1,2,&3
SP-4,5,&6
SP-7,8,&9
P-1,2,&3
P-4.5&6
P-7,8&9
P-10&11
SSF-1&2
(a) ECI = Existing chlorinated effluent
UnCI = Existing unchlorinated effluent
DOHCI = Department of Health Pilot Plant Chlorinated effluent
DeCI = Existing dechlorinated effluent
(b) FH = Fathead minnows {Pimephales promelas)
GS = Golden Shiners INotemigonus crysoleucas)
FIGURE 1. The California Department of Fish and
Game's Field Laboratory.
laboratory is an 8 by 4 by 2.7 m trailer (Figure
1). The unit is transported on site by a 1-ton
stakeside truck. The total cost of the mobile
laboratory, including all analytical instruments,
is approximately $30,000.
The mobile laboratory is functionally segre-
gated into three separate rooms (Figure 2):
1. Water Control Room; 2. Bioassay Room;
and, 3". Laboratory Room.
The Water Control Room receives up to three
wastes and the dilution water for the bioassays.
Equipment in this room includes: a chiller, heat
exchangers, and an air pump. There are four
heat exchangers, one for each of the wastes and
Object Key:
M/C = Monitor Control T/F = Transformer
PIC = Proportional Diluter Control A/C = Air Conditioning
R/F = Refrigerator W/M = Water Quality Monitor Probe
P/D = Proportional Diluter A/P
E/H = Electrical Connector M/R
H/E = Heat Exchanger
H/W = Hot Water Heater
= Air Pump
- Monitor Reservoir
FIGURE 2. Floor plan of mobile laboratory
one for the dilution water. The heat exchangers,
which have the capacity of lowering the tempera-
tures of incoming water and wastes, are construc-
ted of 10.2 cm and 15.2 cm diameter PVC pipe
with stainless steel tubing contained inside. The
exchangers work on the principal of heat trans-
fer from the water and waste streams to chilled
ethylene glycol in the steel tubing as the streams
flow through the PVC pipes. The chilled ethylene
glycol is supplied from a 1-ton, air-cooled,
water chiller.
The Bioassay Room contains three proportional
diluters [Mount and Brungs (13)] and 36 aquaria
for three continual-flow bioassay tests. An air
14
-------
CHLORINA TION/DECHLORINA TION
conditioner, a waste and dilution water delivery
system, and an automatic water quality monitor
are also present (Figure 2).
The automatic water quality monitor has four
overflowing reservoirs; one for the multiparameter
probe (measuring unit) and one for each of the
three wastes. The monitor continually records
pH, temperature, dissolved oxygen, and conduc-
tivity of each undiluted waste stream on a re-
cycling basis. The recycling interval can be ad-
justed from 30-min to 12-h.
The Laboratory Room houses the electronic
controls for the proportional diluters and the
water quality monitor, and the various instru-
ments needed for the chemical and physical
monitoring of the bioassay aquaria during the
continual-flow bioassays (Figure 2). The control
unit of the water quality monitor, in addition to
recording the four water quality parameters,
records a code for each waste and the time-of-
day on paper and cassette tapes.
The laboratory operates on standard 120/240
VAC ^'ectricitv which ran hp snnnlied from two
separate sources. Generally, a 3P 480 VAC
source is obtained from the treatment plant and
connected to the trailer. The 3P supply is con-
verted to IP 240/120 VAC through a IP 480/240
VAC transformer located underneath the trailer.
A standby IP 240 VAC diesel-powered portable
alternator can also be used.
portlancil OilyUr [ | Pi-nporllin.nl U""''' | i Prg^.tUnol PI
37
*
IB
V
10
c
32
•4
11
%.
to
c
+
31 | IB
I-
~^~
•
C
FIGURE 3. Schematic diayiam of waste and water flow
in mobile laboratory
The flow of water and waste through the
mobile laboratory is schematically diagrammed in
Figure 3. The water and waste delivery system of
the mobile laboratory is constructed out of PVC
SCH 40 and 80 pipe, fittings, and valves. Nylon
garden hoses connected to submersible pumps
supply the wastes to the trailer. Dechlorinated tap
water is used as the dilution water for the
bioassays. Dechlorination is accomplished by ac-
tivated charcoal contained inside of the large
(dilution water) heat exchanger. The flows of the
wastes and dilution water in the mobile
laboratory are controlled by a series of PVC ball
valves (which function as shunts) located down-
stream of the heat exchangers. Proportions of the
wastes are drained off and supplied to the water
quality monitor before they enter the proportional
diluters.
tfiuussuy Methods
Standard 96-h continual-flow bioassay methods
[APHA (1); Peltier (14)] were used in the toxicity
testing. Both fathead minnows and golden shiners
were used as the test organisms. The proportional
diluters supplied continual waste concentrations in
a geometric series (100, 56, 32, 18, 100%) and a
control (100% dilution water) to the bioassay
aquaria during the 96-h tests. All waste concen-
trations and the control were done in replicate.
Fifteen fish were tested in each 10 1 overflowing
plexiglass aquarium. The volume of each aquarium
was replenished every 2.5-h. Fish mortality was
determined every 24-h. Adequate dissolved oxygen
was supplied to the test aquaria by aeration.
Several water quality parameters using Standard
Methods [APHA (1)] were manually measured in
the bioassay aquaria every six hours during the
bioassays. Dissolved oxygen and temperature were
measured with a daily calibrated dissolved oxygen
meter and probe, and pH was determined using
an expanded-scale pH meter and combination
electrode. Total ammonia was determined with a
specific-ion meter and NH3 gas-sensing electrode.
Total residual chlorine was measured with an
amperometric titrator.
During the bioassay tests, total residual chlorine
and total ammonia were manually determined at
2-h intervals in each undiluted waste stream along
with the automatically measured levels of dis-
solved oxygen, pH, temperature, and conductivity.
Data Analysis
We estimated mortality for each waste concen-
tration during the 96-h test from m = (1 - SX/SC)
100, where Sx is equal to the survival in waste
concentration X and Sc is equal to the survival in
the controls. If greater than 15% mortality oc-
curred in tne controls during the test, the test
was considered invalid and no mortalities in the
waste concentrations were calculated. To evaluate
15
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 2. SUMMARY OF MEAN TOTAL RESIDUAL CHLORINE AND TOTAL AMMONIA LEVELS IN THE UNDILUTED
EFFLUENTS DURING THE PERIOD OF FEBRUARY THROUGH AUGUST 1978.
Plant
San Leandro
San Pablo
Pinole
Series
SL-1
SL-2
SL-3
SL-4
SP-1
SP-2
SP-3
SP-4
SP-5
SP-6
SP-7
SP-8
SP-9
P-1
P-2
P-3
P-4
P-5
P-6
P-7
P-8
P-9
P-10&11
South SanFranciso. SSF-1&2
(a) ECI =
UnCI =
DOHCI =
DeCI =
(b) Mean ±
Existing chlorinated effluent
Existing unchlorinated effluent
Department of Health Pilot Plant
Existing dechi^Hnated effluent
standard deviation.
Effluent
Type (a)
UnCI
ECI
UnCI
ECI
UnCI
ECI
DOHCI
UnCI
ECI
DeCI
UnCI
ECI
DOHCI
UnCI
ECI
DOHCI
UnCI
ECI
DeCI
UnCI
ECI
DOHCI
ECI
ECI
Chlorinated effluent
Mean
TRC Residual
(mg/l) (b)
0.00 ±
9.47 ±
0.00 ±
5.93 ±
0.00 ±
2.23 ±
2.15 ±
0.00 ±
2.71 ±
0.00 ±
0.00 ±
2.44 ±
2.14 ±
0.00 ±
5.03 ±
3.01 ±
0.00 ±
4.42 ±
0.14 ±
0.00 ±
8.00 ±
2.62 ±
4.39 ±
4.07 ±
0.00
2.42
0.00
1.99
0.00
0.74
0.52
0.00
0.47
0.00
0.00
0.31
0.31
0.00
2.32
0.44
0.00
2.16
0.52
0.00
4.16
0.36
2.19
2.02
Mean
Total Ammonia
(mg/l (b)
13.9 ± 4.3
12.6 ± 4.5
18.0 ± 3.6
17.2 ± 2.4
<0.1 ± 0.0
<0.1 ± 0.0
<0.1 ± 0.0
<1.0 ± 0.0
<1.0 ± 0.0
<1.0 ± 0.0
<1.0 ± 0.0
<1.0 ± 0.0
<1.0 ± 0.0
15.9 ± 4.1
14.9 ± 4.0
15.1 ± 4.4
23.5 ± 7.8
22.8 ± 6.5
23.9 ± 7.3
19.2 ± 8.1
19.2 ± 8.3
19.4 ± 1.7
24.2 ± 11.7
64.5 ± 28.0
the toxicities of all effluents, we calculated a 96-h
LC50 (the waste [%] or Chlorine [mg/l TRC]
concentration which produced 50% mortality to
the test organisms within 96-h) or noted the mor-
tality at the highest concentration tested for the
test organisms.
Statistically significant (P<.05) differences
among respective mean test results for the various
optimized and existing chlorinated effluents were
determined by subjecting the data to two-tailed t-
tests [Sokal and Rohlf (17)]. Significant
correlations (P<.05) between the toxicity of the
chlorinated effluents and their chlorine residuals
were calculated using linear regressions by the
method of least squares (17).
RESULTS
Mean test chlorine residuals in the existing ef-
fluents varied considerably from a low at San
Pablo of 2.23 mg/l TRC to a high of 9.57 mg/l
TRC at San Leandro (Table 2). Mean chlorine
residuals in the existing effluents at Pinole and
South San Francisco were intermediate within
this range.
In all cases, the optimized chlorination process
lowered the mean chlorine residuals below those
of the existing effluents (Table 2). Hence, it ap-
pears that the DOH Pilot plant optimized
system uses chlorine more efficiently. During the
comparative studies at San Pablo, both mean test
chlorine residuals were lower in the optimized ef-
fluent, but only the second comparative study
demonstrated a significant difference (Figure 4).
Both comparative studies at Pinole demonstrated
the optimized effluents to have significantly lower
chlorine residuals than the existing effluents.
16
-------
CHLORINA TION/DECHLORINA TION
8.
6"
£
~ 4«
2-
0-
r~|Eii*tir>9 WTP CMorinotion
KEY: ^^ DOH Pilot Plant Chlorination
T Standard Deviation -
I -.862 1 = 7.14
.-37 tf = 48
ft fl
i
P<.001
1-598
>.48
.|
X*
SAN PABLO SAN PABLO P|NDLE
T
-£
P<.00t
I 8.70
»-48
P-9
f
*^
PINOLE
FIGURE 4. Differences in chlorine residuals between op
timized and existing effluents
3 3i
\
CD
5 2,
o
ce
*~ 1
0.
DOH PILOT PLANT
SERIES SP-3
7*2
TIME (HOURSI
SAN PABLO WTP
SERIES SP-2
TIME (HOURSI
FIGURE 5. Chlorine residuals in optimized and existing
effluents at San Pablo
DOH PILOT PLANT
Series P-3
7«2
TIME (HOURSI
In addition to having comparatively lower
chlorine residuals, the optimized effluents had
typically less variable chlorine residuals (Table 2).
This occurrence was apparent at both San Pablo
(Figure 5) and at Pinole (Figures 6 and 7). The
variability of chlorine residuals in both the opti-
mized and existing chlorinated effluents during
the second comparative test at San Pablo were
similar (Figure 8). The existing effluents at San
Leandro and South San Francisco also had vari-
able chlorine residuals (Table 2).
DON PILOT PLANT
Strict P-9
TIMt (HOURS)
FIGURE 7. Chlorine residuals in optimized and existing
effluents at Pinole.
S.ri,, SP-9
DOH PILOT PLANT
7*2
TIME {HOURS)
Series SP-8
SAN PABLO WTP
2*1 4'S 7*2 »*
TIME {HOURS!
FIGURE 8. Chlorine residuals in optimized and existing
effluents at San Pablo.
7*1
^ Acute Toxicity
There was no acute toxicity associated with
FIGURE 6. Chlorine residuals in optimized and existing either the existing unchlorinated or dechlorinated
effluents at Pinole effluents at any of the plants investigated (Table 3).
,un,,Dc,
(HOURS)
17
-------
PROGRESS fN WASTEWATER DISINFECTION TECHNOLOGY
Generally speaking, the toxicities of the opti-
mized chlorinated effluents were less than those
for the existing chlorinated effluents. One excep-
tion to this rule was during the first comparative
study at San Pablo; the 96-h LCSO's of the
existing chlorinated effluents were greater than
50% while those of the optimized chlorinated
effluents were 48%. During the second compara-
tive study at San Pablo, the toxicity of the exist-
ing chlorinated effluent (96-h LC50 x = 38%)
was more than the optimized chlorinated effluent
(96-h LC50 x = 58%), but the difference between
the two effluents was not significant. No com-
parison could be made between the two compara-
tive effluents from the first study at Pinole be-
cause of excessive disease-related mortality
which occurred in our test fish. During the
second series of comparative testing at Pinole,
the optimized chlorinated effluent (96-h LC50 x
= 15.5%) was significantly less toxic than the exist-
ing chlorinated effluent (96-h LC50 x = 4.3%/
Both fathead minnows and golden shiners were
used as test organisms in the toxicity tests (Table
1). Golden shiners had to be substituted for fat-
TABLE 3. CONTINUAL-FLOW BIOASSAY RESULTS FOR UNCHLORINATED, CHLORINATED, AND DECHLORINATED
EFFLUENTS AT WASTEWATER TREATMENT PLANTS IN CALIFORNIA FROM FEBRUARY THROUGH AUGUST 1978.
(a) (b)
Plant Test Species Test Series
San Leandro FH
FH
FH
FH
FH
FH
FY
San Pablo FH
FH
FH
FH
FH
FH
FH
FH
FH
FH
FH
FH
FH
FH
FH
FH
FH
FH
Pinole FH
FH
FH
FH
FH
FH
GS
GS
GS
GS
GS
GS
US
GS
GS
GS
GS
GS
SL-1 (1)
SL-1 (2)
SL-2 (1 + 2) ««
SL-3 (1)
SL-3 (2)
SL-4 (1)
SL-4 (2)
SP-1 1)
SP-1 (2)
SP-2 (1)
SP-2 (2)
SP-3 (1
SP-3 (2)
SP-4 1
SP-4 (2)
SP-5 (1)
SP-5 (2)
SP-6 (1
SP-6 (2)
SP-7 1
SP-7 (2)
SP-8 1
SP-8 (2)
SP-9 (1)
SP-9 (2)
P-1 (1)
P-1 (2)
P-2 (1)
P-2 (2)
P-3 (1)
P-3 (2)
P-4 (1)
P-4 (2)
P-5 (1)
P-5 (2)
P-6 (1)
P-6 (2)
P-7 (1)
P-7 (2)
P-8 (1)
P-8 (2)
P-9 (1)
P-9 (2)
(c)
Effluent Type
UnCI
UnCI
ECI
UnCI
UnCI
ECI
ECI
UnCI
UnCI
ECI
ECI
DOHCI
DOHCI
UnCI
UnCI
ECI
ECI
DeCI
DeCI
UnCI
UnCI
ECI
ECI
DOHCI
DOHCI
UnCI
UnCI
ECI
ECI
DOHCI
DOHCI
UnCI
UnCI
ECI
ECI
DeCI
DeCI
UnCI
UnCI
ECI
ECI
DOHCI
DOHCI
96-h
LC50
(% Cone)
No toxicity (e)
No toxicity
2.1
No toxicity
No toxicity
3.5
3.3
No toxicity
No toxicity
>50.0
>50.0
48.0
48.0
No Toxicity
No toxicity
43.0
49.0
No toxicity
No toxicity
No toxicity
No toxicity
39.0
37.0
71.0
45.0
ND (f)
ND (f)
ND (f)
ND (f)
7.2
6.9
No toxicity
No toxicity
6.2
6.3
No toxicity
No toxicity
No toxicity
No toxicity
4.4
4.2
16.0
15.0
96-h
LCSO
(mg/l TRC)
0.10
0.13
0.16
>0.46
>0.52
0.48
0.65
0.53
0.55
0.59
0.57
0.90
0.35
0.12
0.12
0.10
0.13
0.15
0.16
0.18
0.15
18
-------
CHLORINA TION/DECHLORINA TION
TABLE 3. (Continued).
Plant
South San
Francisco
(a) (b)
Test Species Test Series
GS
GS
FH
FH
FH
FH
GS
GS
p-10 (1)
P-10 (2)
P-11 (1)
P-11 (2)
SSF-1 (1)
SSF-1 (2)
SSF-2 (1)
SSF-2 (2)
(c)
Effluent Type
ECI
ECI
ECI
ECI
ECI
ECI
ECI
ECI
96-h
LCSO
(% Cone)
7.4
7.4
6.9
5.4
9.8
9.6
9.2
11.0
96-h
LCSO
(mg/l TRC)
0.26
0.21
0.14
0.14
0.14
0.14
0.21
0.22
(a) FH = Fathead minnows (Pimephales pmmelasl
GS = Golden Shiners (Notemigonus crysoleucasi
(D) See Table 1 tor lest series dates; numoers in parentheses correspond to replicate tests.
(c) boi = Existing chlorinated effluent
UnCI = Existing unchlorinated effluent
DOHCI = Department of Health Pilot Plant chlorinated effluent
DeCI = Existing dechlorinated effluent
(a) MBSUITS Trum the two replicates were averaged to estimate LCSO.
(e) 100% survival in undiluted effluent.
(f) No LCSO determined due to excess mortality in controls.
head minnows when disease was present in our
fathead minnow stock. We found the fathead
minnow (96-h LCSO x = 0.13 mg/l TRC) to be
significantly more sensitive to chlorine than the
golden shiner (96-h LCSO x = 0.18 mg/l TRC)
by a factor of 140%.
We noticed a significant difference between the
comparative toxicities of chlorine to fathead min-
nows at San Pablo and all of the other plants
investigated (San Leandro, Pinole, and South San
Francisco). Chlorine was found to be less toxic at
San Pablo (96-h LCSO x = 0.58 mg/l TRC) than
at the other locations (96-h LCSO x = 0.13 mg/l
TRC) by a factor of 450%. The only. measurable
difference in the effluent quality between the two
groups of plants was in their total ammonia con-
tent. San Pablo had ammonia levels consistently
below 1.0 mg/l while all of the other locations
had ammonia levels in excess of 10.0 mg/l
(Table 2).
In effluents with total ammonia concentrations
greater than 10.0 mg/l, we found significant cor-
relations between the 96-h LCSO levels and cor-
responding chlorine residuals of the undiluted ef-
fluents for both fathead minnows (r = 0.84) and
golden shiners (r = 0.86)(Figure 9). The toxic re-
actions of chlorine to fathead minnows, and
golden shiners in those effluents appear to be si-
milar since the slopes of the regression equations
(-1.41 and -1.86, respectively) are similar. A
nonsignificant correlation (r = 0.40) was found be-
tween the two variables for fathead minnows at
- 20i
30
20-
10-
5-
2.
60'
50i
40.
30«
Arcsinvp (y)= 21.12-1.41(i)
r = .84
P<.01
(2)
Fathead Minnows and Total Ammonia 10.0 mg I
2 ' 4 5 8 15
TRC (MG/L) UNDILUTED EFFLUENT
Arcsinvp(y)=25.51-1.86(*j
r= 86
PC01
(2)
Golden Shiners and Total AmmoniaMOOmfl/l
3 I 5 8 IS
TRC IMG/LI UNDILUTED EFFLUENT
Arcslni/p(y)=66,70-9.79fx)
' V 'r--M
P>.05
ft
Fathead Minnows and
Total Ammonla
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
San Pablo where the undiluted effluents had less
than 1.0 mg/1 total ammonia.
DISCUSSION AND SUMMARY
No acute toxicity was found in any of the un-
chlorinated effluents. Dechlorination of the toxic
chlorinated effluents removed all chlorine-induced
toxicity. Hence, chlorine can be viewed as the
single, most toxic constituent of most wastewater
effluents [Martens and Servizi (11); Beeton,
Kovacic, and Brooks (3)].
The optimized effluents were characteristically
lower in residual chlorine than the existing ef-
fluents at both San Pablo and Pinole. Even
though the optimized effluents at San Pablo had
slightly lower mean chlorine residuals, the first
comparative study showed the existing effluent to
be lower in toxicity than the optimized effluent
while the second comp_arative study showed the
optimized effluent to be lower in toxicity than
the existing effluent. Hence, chlorine optimization
of the existing unchlorinated effluent at San
Pablo did not produce toxicity less than that
present in the existing -chlorinated effluent. The
San Pablo WTP may have been employing op-
timization design criteria into the chlorination of
their effluents similar to that which was employed
by the DOH pilot plant.
The optimized chlorinated effluents at Pinole
were significantly lower in chlorine residual than
the existing chlorinated effluents. During the
second comparative study, the optimized effluent
was significantly lower in chlorine and toxicity
than the existing effluent by factors of 300 and
360%, respectively. In this case, chlorine op-
timization of the existing unchlorinated effluent
produced a toxicity less than that present in the
existing chlorinated effluent.
We found fathead minnows to be significantly
more sensitive to chlorine than golden shiners by
a factor of 140%. Our mean 96-h LC50 values
for both test species agree remarkably well with
those values in the published literature [Arthur
et al (2); Esvelt, Kaufman, and Selleck (8)] for
ammoniated wastewater treatment plant effluents.
We found a significant difference in the sensi-
tivities of fathead minnows to chlorine between
nitrified and ammoniated effluents by a factor
of 450%. The nitrified effluents at San Pablo
had total ammonia concentrations less than 1.0
mg/1 and a mean 96-h LC50 of 0,58 mg/1
TRC. Effluents from all other plants had total
ammonia concentrations greater, than 10.0 mg/1
and a mean 96-h LC50 of 0.13 mg/1 TRC.
The influence of ammonia on chlorine toxicity
has been documented elsewhere. For example,
Finlayson (9) found chlorine in ammoniated
wastewater effluents to be more toxic to rainbow
trout fry by a factor of 410% than in nitrified
effluents containing less than 0.5 mg/1 total
ammonia.
A possible explanation for the observed differ-
ence in chlorine toxicity between these two types
of effluents may in part be due to the relation-
ships in the chemistry of chlorine and ammonia.
At San Pablo, break-point chlorination was
apparently accomplished because the chlorine
dosages, which were in excess of twice the total
ammonia concentrations, were above break-point
[Collins and Selleck (7)]. Break-point chlorina-
tion was not apparently accomplished at any of
the other plants investigated because the total
ammonia concentrations were always higher than
the chlorine residuals. At break-point chlorination
ammonia is lacking; consequently, trichloramines,
nitrous oxide, chlorine dioxide, and complex
organochloramines may be produced from side-
reactions between the chlorine and the waste.
Conversely, when ammonia is present and abun-
dant, mono- and dichloramines are primarily
produced between the chlorine and waste.
Rosenberger (16) suspected mono- and dich-
loramines to be more toxic than the side-reaction
products formed at break-point chlorination. The
toxicities of trichloramines, nitrous oxide, chlorine
dioxide, and complex organochloramines are not
well known. However, all of the aforementioned
chlorine residual species will partially titrate out
as total residual chlorine in the amperometric ti-
tration [APHA (1)]. Hence, if one or more of
these'break-point chlorination products were
present at San Pablo, and were either non-toxic
or less toxic than mono- or dichloramine, and
were included in the total chlorine residual, the
result would be that the test fish would be able
to withstand apparently higher chlorine residuals
than the test fish at the other locations.
In summary, the most significant findings of
this study are:
1. the optimized design criteria investigated in
this study provided lower and more stabte
20
-------
CHLORINA TION/DECHLORINA TION
chlorine residuals in wastewater than were
provided by the existing facilities;
2. because the optimized effluents had lower
chlorine residuals, they were typically
lower in toxicity than existing effluents; and
3. nitrification of wastewaters prior to chlori-
nation can reduce the toxicity of the resul-
ting chlorine residuals.
In view of the latter finding, perhaps nitrifica-
tion should be considered an important design
criterion in chlorination systems.
ACKNOWLEDGEMENTS
Steve Flannery, John Nelson, and Don
Yoshikawa assisted with the collection of the data
analyzed in this paper. Debra Langdon typed the
manuscript and Tracy Watson drew the graphs
and charts.
Investigation supported by funds from the Cali-
fornia Department of Health under Interagency
Agreement No. 75-54104 A-l.
REFERENCES
1. American Public Health Association. 1975. Standard
Methods for the Examination of Water and Wastewater.
14th ed. APHA, New York, NY. pp. 1193.
2. Arthur, J.W., R.W. Andrew, V.R. Mattson, D.T. Olson,
G.E. Glass, B.J. Halligan, and C.T. Walbridge. J975.
Comparative toxicity of sewage effluent disinfection to
freshwater aquatic life. Environ. Prot. Agency, Ecologi-
cal Res. Ser. EPA-600/3-75-012. pp. 53 + Appendix.
3. Beeton, A.M., P.K. Kovacic, and A.S. Brooks. 1976.
Effects of chlorine and sulfite reduction Lake
Michigan invertebrates. Environ. Prot. Agency, Ecologi-
cal Res. Ser. EPA-600/3-76-036. pp. 122.
4. Brungs, W.A. 1973. Effects of residual chlorine on aquatic
life. J. Water Poll. Contr. Fed. 45:2180-2193.
5. Brungs, W.A. 1976. Effects of wastewater and cooling
water chlorination on aquatic life. Environ. Prot.
Agency, Ecological Res. Ser. EPA-600/3-76-098.
pp. 46.
6. Collins, H.F. and D.G. Deaner. 1973. Sewage chlorina-
tion versus toxicity - A dilemma. J. Environ. Eng.
Biol. 99(6):761-772.
7. Collins, H.F. and R.E. Selleck. 1972. Process kinetics of
wastewater chlorination. Univ. of Calif., Sanitary
Eng. Res. Lab., Berkeley, CA. SERL Report No.
72-5. pp. 91.
8. Esvelt, L.A., W.J. Kaufman and R.E. Selleck. 1971.
Toxicity removal from municipal wastewaters. Volume
IV of A Study of Toxicity and Biostimulation in San
Francisco Bay-Delta Waters. Univ. of Calif., Sanitary
Eng. Res. Lab., Berkeley, CA. SERL Report No.
71-7. pp. 153.
9. Finlayson, B.J. 1977. Evaluation of four wastewater treat-
ment facilities in Sacramento County, CA. Calif. Dept.
Fish and Game, Fish and Wildlife Water Pollution
Control Laboratory, Rancho Cordova, CA. Laboratory
Memorandum Report No. 77-3. pp. 22 + Appendices.
10. Finlayson, B.J. and L.A. Hinkelman. 1977. Effects of
chlorinated power plant cooling water on aquatic life.
Calif. Dept. Fish and Game, Environ. Services Branch,
Sacramento, CA. Administrative Report No. 77-5.
pp. 53.
11. Martens, D.W. and J.A. Servizi. 1975. Dechlorination of
municipal sewage using sulfur dioxide. Internat'l Pacific
Salmon Fish. Commission, New Westminister, B.C.,
Canada. Progress Report No. 32. pp. 24.
12. Mattice, J.S. and H.E. Zittel. 1976. Site-specific evalua-
tion of power plant chlorination. A proposal. J. Water
Poll. Contr. Fed. 48:2284-2308.
13. Mount, D.I. and W.A. Brungs. 1967. A simplified dosing
apparatus for fish toxicological studies. Water Res.
1:21-29.
14. Peltier, W. 1978. Methods for measuring the acute toxi-
city of effluents to aquatic organisms. Environ. Prot.
Agency, Environ. Monitoring Ser. EPA-600/4-78-012.
pp. 52.
15. Roberts, M., R. Diaz, M. Bender, and R. Huggett. 1975.
Acute toxicity of chlorine to selected marine species.
J. Fish. Res. Bd. Can. 32:2525-2528.
16. Rosenberger, D.R, 1971. The calculation of acute toxicity
of free chlorine and chloramines to coho salmon by
multiple regression analysis. M.S. Thesis, Mich. State
Univ., East Lansing, Michigan, pp. 33 + Appendix.
17. Sokal, R.R. and F.J. Rohlf. 1969. Biometry. W.H.
Freeman and Company, San Francisco, California.
pp. 776.
18. U.S. Environmental Protection Agency. 1976. Quality
Criteria for Water, pp. 256.
19. Ward, R.N., R.D. Giffin, G.M. DeGraeve, and R.A.
Stone. 1976. Disinfection efficiency and residual toxicity
of several wastewater disinfectants. Volume I -
Grandville, Michigan. Environ. Prot. Agency, Environ.
Prot. Technol. Serv. EPA-600/2-76-156. pp. 131.
20. Zillich, J.A. 1972. Toxicity of combined chlorine residuals
to freshwater fish. J. Water Poll. Contr. Fed. 44(2):
212-220.
DISCUSSION
DR. DONALD JOHNSON: I am interested in
the difference in the toxicity of the chloramines
and particularly whether or not you had organic
chloramines in the San Pablo effluent, which I
understood had the lower toxicity . . . lower
ammonia. Did you have any Kjeldahl nitrogen
values for your wastewater?
21
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
MR. HANSEN: No. We had no Kjeldahl nitro-
gen measurements.
DR. JOHNSON: Were you really up at break-
point? Normally the breakpoint ratio is 10 mg of
chlorine per mg of ammonia. You said the
ammonia concentrations were around 1 mg.
MR. HANSEN: I cannot recall the exact values
but they were less than 1 mg/1, if we take all of
the test into consideration. I think they got
down to around 0.2 mg/1 in some cases.
DR. JOHNSON: The values were 0.2 to 1
mg/1. What was your chlorine dosage?
MR. HANSEN: I do not know what the do-
sages were. We took the effluent that came from
the optimized pilot plant, so the results I am re-
porting are what we measured in our laboratory.
DR. JOHNSON: Did you do any free chlorine
titrations and compare with total chlorine?
MR. HANSEN: No, we just took the total.
DR. RIP RICE, Jacobs Engineering: Did I
understand that all of your dechlorination was
done with activated carbon?
MR. HANSEN: Yes, our dilution water was
tap water which was dechlorinated with charcoal.
DR. RICE: Right. Was this powdered activated
charcoal or was it granular?
MR. HANSEN: It was granular.
DR. RICE: Granular activated carbon?
MR. HANSEN: That is correct. We replaced
that on each site.
DR. RICE: That is what I was going to ask.
Did you have any data as to how long the de-
chlorination efficiencies . . .
MR. HANSEN: Two weeks. That is all we
used it for and we replaced it 100 pounds at a
time.
DR. RICE: And it was good for that length of
time?
MR. HANSEN: Yes. No toxicity at all. We ob-
served no mortality whatsoever in our controls
and that is our guide.
MR. GREG SEEGERT, WAPORA, Inc: I had
a question on the differences which you ascribe
to the toxicity between the various fractions. You
said in one case you were essentially dealing with
solutions which you felt were primarily mono-
and dichloramine and in the other case, since you
said breakpoint chlorination was occurring, you
were getting the various products such as tri-
chloramine and some of the others. Is that correct?
MR. HANSEN: Yes, and that is mainly taken
from the literature. That is just our attempt to
explain why these differences occurred. We have
not measured them ourselves.
MR. SEEGERT: It would seem to me that the
more logical explanation would be the difference
between the mono- and dichloramine versus the
free chlorine fraction which, I think, is one of
the things Don Johnson was essentially leading
up to. In one case, if you have breakpoint chlori-
nation occurring, your total residual in the break-
point situation is going to be either hypochlorous
acid or hypochlorite ion.
MR. HANSEN: Right.
MR. SEEGERT: In the other case you are
going to be having the combination of primarily
monochloramine with maybe a little dichloramine.
It would be more logical to attribute any differ-
ences to that.
In support of that conclusion or the fact that
there were differences in toxicity, you referred to
a paper by Rosenberger. You said that he had re-
ported that there were differences between the
effects of monochloramine, dichloramine and
trichloramine. I do not recall in that paper any
tests on trichloramine. It was only free chlorine
and monochloramine and dichloramine. He did
not identify any trichloramine fraction.
MR. HANSEN: I believe that came out when
he was discussing the various forms.
DR. JOHNSON: I might just say, I was think-
ing of organic chlorine rather free chlorine. I
wonder if the free chlorine residual . . .
MR. HANSEN: I do not think we saw free
chlorine. It disappeared pretty fast.
DR. JOHNSON: There is very little data on
organic chloramine. That is the reason I asked
about critical mass.
MR. SEEGERT: Well that is certainly true.
There could be a sizeable fraction of more com-
plex forms of chlorine, but there is likely
a sizeable fraction of either hypochlorous acid or
hypochlorite ion. Certainly in terms of killing
bacteria and viruses, free chlorine is widely re-
garded as being more toxic.
DR. JOHNSON: Certainly it is the most toxic
form, but I do not think that is what we are
faced with, at least in the wastewater treatment
plants. It is the chloramines which last longer, of
course.
MR. VENOSA: Perhaps Endel could shed some
light on the dosage of chlorine that was used in
22
-------
CHLORINA TION/DECHLOR1NA TION
the pilot plant at that particular location.
MR SEPP: The chlorine dose at the San Pablo
plant was 10-12 mg/1 for the full scale plant and
the dose in the pilot plant was 10-11 mg/1
slightly lower.
Now these were the daytime dosages. We do
not have any data on the night time dosages.
MR WHltE: I want to point out that the or-
ganic chloramines titrate in the forward titration
with the dichloramines, so in wastewater treat-
ment I would say it is impossible to be able to
identify that something is dichloramine because
there is probably no pure dichloramine.
Now, to learn about the effect of organic ni-
trogen in a breakpoint situation in both waste-
water and tap water, Saunier did some research
at the University of California that is in his
thesis. I believe it was printed and excerpted in
an AWWA paper that he gave. I do not know
whether it has been in the journal or not but
he tells about that.
23
-------
6.
DESIGN OF A CHLORINATION SYSTEM FOR LAGOON DISINFECTION
Bruce A. Johnson, J. H. Reynolds, J. L. Wight and E. J. Middlebrooks
Utah Water Research Laboratory
Utah State University
Logan, Utah
ABSTRACT
Chlorine disinfection of waste stabilization lagoon effluents has been
anil is being considered as a solution to bacterial removal prior to dis-
charge to receiving waters. To evaluate the amenability of algae-laden
lagoon effluent to chlorine disinfection, chlorination lest facilities were
constructed at the Logan, L'la/i, wastewater lagoons. An investigation was
conducted at these facilities on primary and secondary, as well as filtered
and unfiliered, lagoon effluents between August 1, 1975, and August 24,
1976. The filtered effluent was obtained by passing lagoon effluent through
an intermittent sand filter prior to chlorination.
The results of this study indicate that, in all cases, adequate disinfection
was obtained with combined chlorine residual within a contact period of
60 minutes or less. Filtered effluent was found to exert less chlorine de-
mand than unfiltered effluent. It was also determined that temperature,
sulfide, and total chemical oxygen demand influence the chlorine dose
necessary to achieve a specified level of disinfection. Suspended solids and
soluble chemical oxygen demand were found to be slightly altered as a
result of chlorination.
A mathematical model was developed to represent the effects of chlo-
rination of lagoon effluents. This model irc/.v used to predict the chlorine
dosages necessary to achieve adequate disinfection for varying effluent
characteristics. A series of design curves was constructed from the model
for use in selecting the optimal chlorine dosages necessary for achieving
prescribed levels of disinfection.
INTRODUCTION standards, many proposed and existing lagoon systems
may not produce satisfactory effluent coliform con-
Wastewater lagoon systems provide simple econ- centrations. Therefore, disinfection of lagoon effluents
omic wastewater treatment for approximately 5,000 is becoming common practice.
communities throughout the United States. Histor- Although various disinfection processes are avail-
ically. hydraulic detention times within many of these able for application to lagoons, chlorination is prac-
lagoon systems have been sufficient to provide ade- ticed in most installations. However, there is evidence
quate total and fecal coliform reduction. However, as [Burkhead and O'Brien, 1973 (3); Echelberger. el a/.,
a result of more stringent state and federal discharge 1971 (8); and Horn, 1972 (II)] that chlorination of
24
-------
CHLORINA TION/DECHLORINA TION
lagoon effluents increases the biochemical oxygen
demand (BOD5) and soluble chemical oxygen demand
of the effluent. Thus, there is serious concern about
the proper chlorine dose and contact time required
for satisfactory chlorination of lagoon effluents.
The paper presents a series of design curves based
on a mathematical model developed by Johnson, et
a/., 1978 (12). The model was developed with data
obtained from operating a field scale waste stabiliza-
tion lagoon facility for twelve months to evaluate
lagoon chlorination practices under varying seasonal
conditions. This evaluation included determination of
the chlorine residual, concentration necessary to re-
duce bacterial population to acceptable levels, and the
effects of temperature, suspended solids, ammonia,
chemical oxygen demand, and sulfide on chlorination
practices.
PROCEDURES
The Logan, Utah, wastewater stabilization lagoons
were selected as the site for this study. Because of the
relatively high bacteriological quality of the final
lagoon effluent, the facilities were constructed with
capabilities of treating either primary or secondary
lagoon effluent. Four systems of identically designed
chlorine mixing and contact tanks, each capable of
treating 50,000 gallons per day, were constructed.
Using recommendations presented by Collins, et at.,
1971 (5), Kothandaraman and Evans, 1972 (13) and
1974(14) and Marske and Boyle, 1973 (16), the chlo-
rination systems were constructed to provide rapid
initial mixing followed by chlorine contact in plug
flow reactors. A serpentine flow configuration, having
a length to width ratio of 25:1, coupled with inlet and
outlet baffles, was used to enhance plug flow hydraulic
performance. The chlorine mixing and contact tanks
are illustrated in Figure 1. Dye studies similar to those
conducted by Deaner [undated] (6) were used to deter-
mine average detention time's for each contact tank.
The theoretical detention time for each tank was 60
minutes, while the actual detention time for the four
tanks averaged 49.6 minutes.
Three of the four chlorination systems were used for
directly treating primary and secondary lagoon efflu-
ent. The effluent treated in the fourth system was
filtered through an intermittent sand filter prior to
chlorination. Filtered lagoon effluent was also used
as the solution water for all four chlorination systems.
Appropriate quantities of chlorine gas were mixed
with solution water by use of vacuum operated dif-
fusers prior to introducing solution lines into chlorine
mixing tanks and exposing the main flow of lagoon
effluent to chlorine. Schematic diagrams to illustrate
the chlorination operation are presented in Figures 2
and 3.
INFLUENT
1
r %
CHLORINE
SOLUTION
^
r
J
~\
V.
r
j
^
INFLUENT
CHLO
SOLU
,
-Hr
RINE
TION
Si_?l
t
rrujtr
v°°
Figure 1. Chlorine mixing and contact tanks.
TREATED
EFFLUENT
TREATED
EFFLUENT
Figure 2. Schematic flow diagram of experimental
chlorination facility.
25
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
STORAGE
_LL
STORAGE
CONTACT
CHAMBER
3
CONTACT
CHAMBER
t
FILTERED
SOLUTION LINES
UNFILTERED
EFFLUENT
EX
.._X
MIX
MIX
CONTACT
CHAMBER
CONTACT
CHAMBER
I
INJECTORS
LORINE GAS
Figure 3. Chlorination facilities.
Chlorination of lagoon effluent began on August 6,
1975, and continued until August 24, 1976. Samples
were collected at least twice a week throughout the
study period except between December, 1975, and
February, 1976. No samples were collected during
that period because of pipeline freeze-up. Samples
were collected just prior to Chlorination as well as at
points corresponding to residence times of 17.5, 35.0,
and 49.6 minutes in each contact tank. Chlorine doses
were varied between 0.25 and 30.0 mg/1. In addition
to chlorinated samples, other samples were collected
from the influent and effluent of the lagoon system
and from the effluent from each cell in the system. This
was done to characterize the performance of the la-
goon system and to assist in determining how to adjust
Chlorination practices to compensate for seasonal
fluctuations in lagoon performance.
The chlorinated samples were analyzed bacterio-
logically for MPN total and fecal coliforms (TC and
FC). Five tubes were used for each dilution. Mem-
brane filter total and fecal coliform determinations
were also conducted on all unchlorinated samples.
Additional water quality analyses included ammonia
(NH3-N), biochemical oxygen demand (BODs), dis-
solved oxygen (DO), total and soluble chemical oxy-
gen demand (TCOD and SCOD), sulfide (S=), sus-
pended solids (SS), volatile suspended solids (VSS),
pH, temperature, and turbidity. Free and combined
chlorine residuals (FC1 and CC1) were also measured
for all chlorinated samples using the amperometric
titration method. With the exception of the sulfide,
all samples were collected and analy/ed using recom-
mended procedures outlined in APHA Standard
Methods, 1971 (1). Sulfide was analyzed using a
method described by Orion [undated] (17).
In addition to the field study as described, labora-
tory studies were also conducted. These studies were
performed to assist in describing relationships be-
tween chlorine and other wastewater constituents
Johnson, el a/., 1978 (12).
26
-------
CHLORINA TION/DECHLORINA TION
RESULTS AND DISCUSSION
Lagoon Disinfection
General
To determine disinfection efficiency, with respect
to total and fecal coliforms, as a function of total
chlorine residual, the data were fitted with a linear
regression equation, which expresses the logarithm
of the fraction of coliform remaining as a function of
total chlorine residual. The intercept values for these
regression equations range from-0.3 to-1.3, indicating
that a 7,ero total chlorine residual concentration will
produce a coliform reduction of from 50 percent to 95
percent. A possible explanation for this anomaly is
that chlorine combines with ammonia, organic mate-
rials, sulfides, and other compounds and thus dissi-
pates leaving no measurable chlorine residual. However,
at some point in time (perhaps only an instant after
addition) the chlorine is also in contact with the bac-
teria in the water and available for disinfection, result-
ing in decreases in coliform concentrations indicated
by the intercept values. The intercept values may also
be due to the statistical confidence intervals associated
with the MPN values.
In an attempt to relate the results of this study with
those reported in the literature, a similar regression
analysis as discussed above was performed using a
forced zero intercept. The results of this analysis are
illustrated in Figures 4, 5, 6, and 1. The correlation
coefficients were significant at the 5 percent level for
all reported regression equations. Because of this high
degree of statistical significance, the forced zero inter-
cept regression analysis was used in discussing the
results. These results are expressed as the logioN/N0
(in which N = the number of organisms per 100 ml after
chlorination, and N0 = the original number of organ-
isms per 100 ml), or the logarithm of the fraction re-
maining after chlorination versus varying concentra-
tions of total chlorine residual.
Figure 5. Summary of fecal coliform removal efficiency in
filtered lagoon effluent as a function of total
chlorine residual at various chlorine contact times.
— IB MIN. CONTACT TIME, fls.922
— 35 " » " , R=939
SO '• " " , B = .908
IB MIN CONTACT TIME, R:.B76
TOTAL CHLORINE RESIDUAL (m5/l) TOTAL CHLORINE RESIDUAL (mg/l)
Figure 6. Summary of total coliform removal efficiency,
Figure 4. Summary of total coliform removal efficiency in using unfiltered lagoon effluent, as a function of
total chlorine residual at various chlorine
filtered lagoon effluent as a function of total
chlorine residual at various chlorine contact times.
contact times.
27
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TOTAL CHLORINE RESIDUAL (mj/l)
Figure 7. Summary of total coliform removal efficiency,
using unfiltered lagoon effluent, as a function of
total chlorine residual at various chlorine
contact times.
Filtered Lagoon Effluent
Total coliforms. The effect of total chlorine
residual on total coliform numbers after chlorine
contact times of 18, 35, and 50 minutes using
filtered lagoon effluent is reported in Figure 4.
Analysis of Figure 4 indicates that the rate of
total coliform removal increases with increasing
total chlorine residual. This result is in agreement
with reports in the literature, [Butterfield, 1948
(2); Chambers, 1971 (4); Green and Stumpf, 1944
(9); Horn, 1970 (10); Kott, 1971 (15); and White,
1972 (18)]. Results from Figure 4 indicate that a
total coliform organism reduction of 99.9 percent
in filtered lagoon effluent can be expected with a
total chlorine residual concentration of 2.7 mg/1
after 18 minutes chlorine contact time. The 99.9
percent level of reduction was chosen for
discussion because at this removal efficiency the
data and subsequent regression lines are well
developed and interpretation of results is accom-
plished with less inference.
By using a statistical test which employs sums of
squares, sums of products, degrees of freedom, sample
regression coefficients, and regression equation slopes,
it was determined that the slopes of the regression
equations for the 35 and 50 minute chlorine contact
times reported in Figure 4 were not significantly dif-
ferent from one another and could be regarded as
having the same slope. This statistical procedure is
used in determining whether confidence intervals for
two regression lines overlap. If an overlap does occur,
it is an indication that the regression equations (regres-
sion lines) are not statistically different. Statistical
results of this kind further suggest that the two regres-
sion lines describe a range in which a single regression
line would be found. Therefore, if the data for both the
35 and 50 minute chlorine contact times wereanaly/ed
together and fitted with a regression equation, the line
described by this equation would fall somewhere be-
tween 35 and 50 minute chlorine contact time regres-
sion lines shown in Figure 4. However, this approach
was not used because it was apparent that to group two
different operational time period data would be statis-
tically incorrect. Because these two regression lines are
not significantly different and because grouping data
is invalid, interpolation between the 35 and 50 minute
chlorine contact time regression lipes in Figure 4 indi-
cate that a total chlorine residual of 1.5 mg/1 is re-
quired to produce a 99.9 percent total coliform reduc-
tion at chlorine contact times between 35 and 50 minutes.
This concentration is contrasted with the 2.7 mg/1
total chlorine residual needed for the same level of
reduction at the 18 minute chlorine contact times,
which is consistent with earlier reports [Butterfield,
1948 (2); Chamber, 1971 (4); and White, 1972 (18)].
The above experimental and statistical results imply
that an increase in chlorine contact time from 18 to
35 minutes will require 1.2 mg/1 or44 percent less total
chlorine residual to obtain the same level of total coli-
form destruction. However, at chlorine contact times
between 35 and 50 minutes, there is no statistically
significant difference in total chlorine residual re-
quired to produce a 99.9 percent reduction in total
coliform concentration. A possible explanation for
this effect occurring consistently at the 35 and 50 min-
ute chlorine contact time is that coliform concentra-
tions are reduced to such low levels within the 35
minute chlorine contact time that further reductions
with increasing time are not statistically measurable.
Fecal coliforms. The effects of total chlorine resi-
dual on fecal coliform bacteria in filtered lagoon efflu-
ent are illustrated in Figure 5 at 18, 35, and 50
minutes of chlorine contact time. This figure
depicts a trend in reduction of fecal coliform levels
with increasing total chlorine residual concentration
and chlorine contact time similar to the trend indicated
by Figure 4 for total coliform reduction. The 35 and 50
minute chlorine contact time regression lines for
Figure 5 were also found to have statistically similar
slopes. An average of 1.7 mg/1 total chlorine residual
after 35 to 50 minutes contact time is needed to effect
a three log fecal coliform reduction. At the 18 minute
chlorine contact time, 2.7 mg/1 total chlorine residual
were required to reduce fecal coliform bacteria to this
same level. This is a difference of 1.0 mg/1 or 37 percent
28
-------
CHLORINA TION/DECHLORINA TION
less total chlorine residual. This suggests longer chlo-
rine contact times will produce the same level of fecal
coliform removal as higher total chlorine residual
concentrations. Again, these results are similar to
earlier published reports: Chambers, 1971 (4); Horn,
1970 (10); and White. 1972 (18).
Unfiltered Lagoon Effluent
Total coliform. Total coliform reduction versus
residua] chlorine at three different contact times (18,
35, and 50 minutes) is illustrated in Figure 6. Figure 6
indicates, again, that increasing amounts of total
chlorine residual and chlorine contact time will in-
crease the rate of reduction of total coliform numbers.
This indicates that a 99.9 percent reduction of total
coliform bacteria can be achieved with 18 minutes of
chlorine contact at total chlorine residual concentra-
tions of 4.2 mg/1. The regression lines in this figure
illustrating the 35 and 50 minute chlorine contact times
are statistically the same, for reasons discussed earlier
in connection with filtered lagoon effluent, and suggest
that an average of 3.0 mg/1 total chlorine residual is
required to reduce total coliform concentrations by
99.9 percent at these chlorine contact times. This is a
29 percent, or 1.2 mg/1, reduction of total chlorine'
residual for the 35 to 50 minute chlorine contact times
over that for the 18 minute chlorine contact time.
Fecal coliform. The effect of total chlorine residual
on fecal coliform reduction in unfiltered lagoon efflu-
ent is shown by Figure 7 for chlorine contact times of
18, 35, and 50 minutes, respectively. Figure 7 indicates
reduced fecal coliform concentrations with increased
chlorine contact time and total chlorine residual. A
3.4 mg/1 total chlorine residual with an 18 minute
chlorine contact time and an average of 2.3 mg/1 of
total chlorine residual for 35 and 50 minutes chlorine
contact times are suggested by this figure to reduce
fecal coliform levels by 99.9 percent. This means 32
percent less total chlorine residual is required at the
longer contact times than fo'r that at 18 minutes chlo-
rine contact to produce the same level of reduction.
Results further indicate that the 99.9 percent fecal
coliform reduction level is achieved at lower concen-
trations of total chlorine residual than for total coli-
form reduction in unfiltered lagoon effluent (an average
of 5.7 mg/1 total chlorine residual for fecal coliform
reduction compared to 7.2 mg, 1 total chlorine residual
for total coliform reduction, or 21 percent less). Fecal
coliform bacteria may be less resistant to chlorine in
unfiltered lagoon effluent. The difference between
wastewater characteristics found in unfiltered lagoon
effluent and those in filtered effluent in combination
with chlorine may cause the fecal coliform to die off at
a greater rate in the unfiltered lagoon effluent. Tr.is
may help to explain why total chlorine residual con-
centrations vary in effectiveness between total and
fecal coliform reduction with unfiltered lagoon efflu-
ent and not with filtered lagoon effluent.
Summary
Results indicate that increasing total chlorine re-
sidual will produce increased total and fecal coliform
reduction for both filtered and unfiltered lagoon ef-
fluent. Results also suggest that statistically significant
reductions in coliform concentration can be accom-
plished at the same residual chlorine level with in-
creasing chlorine contact times.
Indications are that less total chlorine residual is
required for disinfection of filtered lagoon effluent
over that of unfiltered lagoon effluent. An average of
3.6 mg/1 total chlorine residual is required for a 99.9
percent reduction of total coliform numbers in unfil-
tered lagoon effluent, a 42 percent lower chlorine
residual requirement. In addition, 23 percent less total
chlorine residual (2.2 mg/1) is required to achieve a
99.9 percent reduction of fecal coliform numbers in
filtered lagoon effluent, as compared with unfiltered
lagoon effluent (2.85 mg/1).
There is also evidence that fecal coliform bacteria
in unfiltered lagoon effluent are reduced to the 99.9
percent level with smaller concentrations of total
chlorine residual than are the total coliform bacteria.
MODEL DEVELOPMENT
General
The mathematical model developed by Johnson, el
at.. 1978 (12) predicts the appropriate combination of
chlorine dose, contact time, and chlorine residual for
adequate disinfection of a specific waste stabilization
lagoon effluent. The model accounts for the inter-
actions and temperature dependence of chlorine and
sulfide, nitrogen compounds, chemical oxygen
demand, suspended solids, chlorine demand, and dis-
infection (i.e., coliform reduction). Field and labo-
ratory data were evaluated to develop these rela-
tionships and the coefficients employed in the model.
Mode] development
The model incorporates the equations shown in
Table 1. In developing the mathematical model, steady
state representations of Equations 1 to 6 were used to
determine the chlorine consumed by sulfide, the sulfide
remaining after chlorination, and reactions with
ammonia and organic nitrogen. The dynamic portion
29
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
of the model, as represented by Equations 7 to
12 to describe changes in soluble COD, SS, free
and combined chlorine residual, and total and
performed to determine those coefficients most sensi-
tive in producing changes in the model results. It was
found that those stoichiometric constants associated
TABLE 1. SUMMARY OF EQUATIONS EMPLOYED TO DEVELOP DESIGN CURVES.
Function
Equation
Number
Equation*
Comments
To predict the amount
of sulfide remaining
after chlorination
S = SL+ (S0-SL)
To describe effects
of chlorine on
suspended solids
+ (-3.0 x 10-")(SS)(FCI + CCI)
form = exponential decay
r = 0.674
significance level = 95%
To describe the
interreactions
between chlorine and
nitrogen forms
To describe the
effects of chlorine
on suspended solids
and COD
2
3
4
5
6
7
CI2 + H2O ^ HOCI + H+ + Cl
HOCI ^ OCI~ + H+
HOCI + NH3 + NH2CI + H2O
HOCI + NH2CI + NHCI2 + H2O
HOCI + NHCI2:+:NCl3 + H20
d(^fC D) = 7.24 x 10-" (FCI)(TCOD - SCOD)
Assumption is that
reactions occur
rapidly enough to be
instantaneous. Thus,
steady state is assumed.
Rate constant
determined from
regression analysis
(l/mg-min)
Rate constant
determined from
regression analysis
(l/mg-min)
To determine changes
in free chlorine (FCI)
and combined
chlorine (CCI)
dt
10 d(CCI)
FCI "
FCI0 (TCOD)
CCI "
CCI0 (TCOD)
Values for rate and
stoichiometric
constants determined
in model calibration
process.
To determine changes
in total coliform
concentrations (TC
and fecal coliform
concentrations (FC)
1 1 dCTC)
12 rf(FC)
= (-.055)(TC)1-1(CCI)1-35+(-.20)(TC)1-1(FCI)'-3°
= (-.085)(FC)1-OS(CCI)'-35 + (-.35)(FC)1-08(FCI)1-30
Values for rate and
stoichiometric
constants determined
from regression
analysis and model
calibration process.
To account for
temperature effects
13
kT=k20
P =1.15 for chlorine
demand
P = 1.03 for disinfection
'See LIST OF SYMBOLS.
fecal coliforms, was solved using a second order
Runge-Kutta solution technique. Temperature ef-
fects were accounted for by using equation 13.
The model was calibrated by first selecting approxi-
mately 5 percent of the total data (1804 total data
points) representing extremes in effluent characteris-
tics as the calibration data. The model was then applied
to these data and coefficients adjusted by trial and
error until a maximum value of the correlation coef-
ficient, r, was achieved. A sensitivity analysis was
with total and fecal coliform in the disinfection equa-
tions were the most sensitive.
The model was verified by comparing the predicted
values produced by the model with all the observed
values of the field data. Predicted combined chlorine
residuals were found to be correlated to 65 percent of
the observed data (1804 total data points) at the 95
percent confidence level, while 81 percent of predicted
total and fecal coliform values compared with ob-
served values at the same confidence level.
30
-------
CHLORINA TION/DECHLORINA TION
DESIGN CURVES
The model was then used to construct a series of
design curves for selecting chlorine doses and contact
times for achieving desired levels of disinfection. An
example may best illustrate how these design curves
are applied. Assume that a particular lagoon effluent
is characterized as having a fecal coliform concentra-
tion of 10,000 100 ml. 0 mg 1 sulfide. 20 mg 1 TCOD,
and a temperature of 5°C. If it is necessary to reduce
the fecal coliform counts to 100 100 ml, a combined
chlorine residual sufficient to produce a 99 percent
bacterial reduction must be obtained. If an existing
chlorine contact chamber has an average residence
time of 30 minutes, the required chlorine residual is
obtained from Figure 8. A 99 percent bacterial reduc-
tion corresponds to log (N0 N) equal to 2.0. For a
contact period of 30 minutes, a combined chlorine
residual of between 1.0 and 1.5 mg 1 is required to
produce that level of fecal coliform reduction. Upon
interpolation, the actual chlorine residual is deter-
mined to be 1.3 mg, 1. This is indicated by point © in
Figure 8.
INITIAL FECAL COLIFORM MPN : I04/I00m
INITIAL TOTAL COLIFORM MPN = I04/I00m
• FECAL COLIFORM
A TOTAL COLIFORM
30
TIME (Minutes)
Going to Figure 9, it is determined that if a chlorine
dose produces a residual of 1.30 mg 1 at 5°C, the same
dose would produce a residual of 0.95 mg, 1 at 20°C.
This is because of the faster rate of reaction between
I COD and chlorine at the higher temperature. This
is indicated by point® in Figure 9. For an equivalent
chlorine residual of 0.95 mg 1 at 20°C and 20 mg/1
TCOD. it is determined from Figure 10 that the same
chlorine dose would produce a residual of 0.80 mg/1
if the TCOD were 60 mg 1. This is because higher
concentrations of TCOD increase the rate of exertion
of chlorine demand. Point (3) in Figure 10 corresponds
to this residual. The chlorine dose required to produce
an equivalent residual of 0.80 mg 1 at 20°C and 60
mg/1 TCOD is determined from Figure 11. For a chlo-
rine contact period of 30 minutes, a chlorine dose of
2.15 mg I is necessary to produce the desired com-
COM9INED CHLORINE RESIP'.^L AT TEMP I [mg/1
Figure 9. Conversion of combined chlorine residual at
Temp 1 to equivalent residual at 20°C.
Figure 8. Combined chlorine residual at 5°C for coliform
= 10V100 ml.
COMBINED CHLORINE RESIDUAL AT TCOD I AND TEMP=2O° C [mg/D
Figure 10. Conversion of combined chlorine residual at
TCOD 1 and 20°C to equivalent residual at 20°C
and TCOD = 60 mg/l.
31
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
hincd residual as indicated by point (4) on Figure 1 1.
This dose will produee a reduction of fecal coliform ?
from 10,000 100 ml to 100 100 ml within 30 minutes |
at 5°C and with 20 mg 1 TCOD.
g 30
in
UJ
(£
*
I
o
S
5
Chlorine Dose = 10.0 mo/I-1
20 30 40
TIME (Minutes)
Figure 11. Determination of chlorine dose required for
equivalent combined residuals at TCOD - 60 mg/l
and Temp = 20° C.
COMBINED CHLORINE RESIDUAL AT TCOD I AND TEMP - 5-C (mg/l)
Figure 12. Conversion of combined residual chlorine at 5°C
and TCOD 1 to equivalent residual at 5°C and
TCOD = 60 mg/l.
o:
3
If, in the previous example, the initial sulfude con-
centration was l.O mg I instead ofO mg/l, it would be
necessary to go directly from Figure 8 to Figure 12.
Here, chlorine residual of 1.30 mg/l at a TCOD of 20
mg/l and a temperature of 5°C is converted to an
equivalent chlorine residual of 1.10 mg, 1 for a TCOD
of 60 mg/l. This is represented by point (5)in Figure 12.
Going to Figure 13, which corresponds to an initial
sulfide concentration of 1.0 mg/1, it is determined that
a chlorine dose of 6.65 mg/l is necessary to produce
an equivalent chlorine residual of 1.1 mg 1 after a con-
tact period of 30 minutes. Point (S) on Figure 13 cor-
responds to this dose. The sulfide remaining after
chlorination is determined to be 0.44 mg I from Figure Figure 13. Determination of chlorine dose required when
TIME (Minutot)
14 as indicated by point (7) .
S~ = 1.0 mg/l, TCOD = 60 mg/l, and Temp = 5°C.
32
-------
CHLORINA TION/DECHLORINA TION
CHLORINE OOSt (mg/l)
The disinfection model was used to develop a series
of design curves (seven graphs) which may be used to
determine the chlorine doses, contact time, and chlo-
rine residual required for a given level of disinfection.
The use of the curves is illustrated by two separate
examples.
ACKNOWLEDGEMENTS
The work presented in this paper was funded by the
U.S. Environmental Protection Agency, Biological
Treatment Section, Municipal Environmental Re-
search Laboratory, Cincinnati, Ohio, under Contract
No. 68-03-2151, A. D. Venosa. Project Officer.
Figure 14. Sulfide reduction as a function of chlorine dose. LIST OE SYMBOLS
SUMMARY
A disinfection model was developed, verified, and
used to produce design curves which are applicable to
most lagoon systems. The model reflects the follow-
ing variations and trends.
It was determined that disinfection of waste stabili-
zation lagoon effluent can generally be achieved with
relatively low doses of chlorine and in contact times
of less than 50 minutes. The chlorine demand was
found to be less than reported in other literature dur-
ing most of the year. Generally it was found that the
chlorine demand was about 50 percent of the applied
dose during all times of the year except when hydrogen
sulfide was produced. During that period, the chlorine
demand was found to be as high as 85 percent. Com-
bined chlorine residuals of between 0.5 and I.O mg/l
were found to be adequate in reducing fecal conforms
below the discharge standard of 200 100 ml. This
residual is produced by a chlorine dose of between
2-3 mg/l, except during periods of hydrogen sulfide
production when a dose of 7-8 mg/l is required.
Chlorination of these algae laden waters was ac-
companied by very few and minor adverse effects.
Soluble COD was observed to increase in the presence
of free chlorine residual. Increases in turbidity and
reductions of SS were also observed for high chlorine
doses. However, it was rarely necessary to chlorinate
at high enough doses for these responses to have any
major repercussions. Breakpoint chlorination was
observed to be of minimal importance in providing
adequate disinfection. Filtering of lagoon effluent
through intermittent sand filters prior to chlorination
was found to reduce chlorine demand and enhance
disinfection efficiency.
P = Temperature coefficient
CCl = Combined chlorine residual, mg I
CCl0 = Initial combined chlorine residual, mg I
FC = Fecal coliform. N0 100 ml
FCl - Free chlorine residual, mg I
FCIo = Initial free chlorine, mg/l
S = Sulfide
SL = Lower limit of sulfide detection (O.I mg I)
S0 = Initial sulfide concentration, mg I
SCOD = Soluble COD, mg I
SS = Suspended solids, mg I
TCOD = Total COD, mg/l
TC = Total coliform, N0/ 100 ml
X = Chlorine dose, mg I
REFERENCES
I. APHA. I97I. Standard methods for examination of water and
wasteuater. 13 ed. Ameriean Public Health Association.
Inc. pp. 874.
2. Butterfield. C. T. I948. Bacterial properties of chloramines and
free chlorine in water. U.S. Public Health Reports. 63(7):
934-940.
3. Burkhead. C. F... and W. J. O'Brien. 1973. Lagoons and oxida-
tion ponds. .IWPCF 45(10): 1054-1059.
4. Chambers, C. W. 1971. Chlorination for control of bacteria and
virus in treatment plant effluents. .IWPC'F 43(2):228-241.
5. Collins, Harvey K.. Robert E. Sellech. and George C. White.
1971. Problems in obtaining adequate sewagedisinfection.
Journal of the Sanitary Engineering Division of ASCE
97(SA5):349-362.
6. Deaner. David G. Undated. A procedure for conducting dye
tracer studies in chlorine contact chambers and to deter-
mine detention limes and How characteristics. Technical
Paper. Bureau of Sanitary Engineering. California Stale
Dept. of Health.
7. Dinges. Ray, and Alfred Rusl. 1969. Experimental ehlorinalion
ol stabili/alion pond effluent. Public Works I()()(3):9K-10I.
33
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
X. Echelberger. Wayne F.. Joseph I.. Pavoni. Philip C. Singer, and
Mark W. Tenney. 1971. Disinfection ol algal laden waters.
Journal of the Sanitary Engineering Division of ASC'F.
97(SA5):721-730.
9. Green. D. F.. and P. K. Stunipf. 1944. I he mode of action of
ehlorine. Journal of the American Water Works Associa-
tion 38:1301.
10. Horn, [.. W. 1970. Chlorination of waste pond effluent. In:
Second International Symposium for Waste Treatment
Lagoons, (ed. Ross McKinnev). University of Kansas.
Laurence. Kansas.
11. Horn. Leonard W. 1972. Kinetics of chlorine disinfection in an
ecosystem. Journal of the Sanitary Engineering Division
of ASCE 9X(SA11:1X3-194.
12. Johnson. B. A...I. L. Wight. D. S. Bowles..I. H. Reynolds, and
E. .1. Middlehrooks. I97X. Waste stahili/ation lagoon
microorganism removal efficiency and effluent disinfec-
tion with chlorine. Final Report U.S. FPA Contract No.
6X-03-2I5I. Utah Water Research Laboratory. Utah Slate
University. Logan. Utah.
13. Kothandaraman. V.. and R. L. Evans. 1972. Hydraulic model
studies ol chlorine contact tanks. .IWPCF 44(4):625-633.
14. Kothandaraman. V.. and R. L. Evans. 1974. Design and per-
formance ol chlorine contact tanks. Circular 1 19. Illinois
State Water Survey. Urbana. Illinois.
15. Kott. Yehuda. 1971. Chlorination dynamics in wastewater
effluents. Journal of Sanitary Engineering Division ol
ASCE 97(SA5):647-659.
16. Marske. Donald M.. and Jerry D. Boyle. 1973. Chlorine con-
tact chamber design - a lield evaluation. Watei and Sewage
Works 120(11:70-77.
17.. Orion Research Incorporated. Undated. Determination of
total sullide content in water. Applications Bulletin No. I 2.
Orion Research Inc.. Cambridge. Mass.
IX. White, (i. C. 1972. Chlorination of wastewater. In: Handbook
ol chlorination. Van N'orstrand Reinhold Co.. New York.
pp. 420-465.
19. White, (i. Clifford. 1973. Disinfection practices in the San
Francisco Bav area. JWPCF 46( I ):XO-I(II.
DISCUSSION
DR. KARL LONGLEY, U.S. Army:\ found it in-
teresting that you based your nomographs on residual
as a function of dose for 60 mg.'l TOC. In work that I
did with the University of Texas at the Rilling Road
Plant at San Antonio, we were not able to find good
correlation. We were not able to accurately predict
the residual as a function of the dose based on TOC. It
was interesting to see that you were able to do it for
your particular system.
I would like to reemphasi/e what you said at the
end, and that is the fact that you did a tremendous
amount of work here to arrive at your model. The
danger will be of course, particularly when this report
becomes available to consulting engineering firms
and whomever, that there will be a tendency to take
these numbers without checking them on other sys-
tems and seeing if the factors hold up. It is a very dan-
gerous practice but. unfortunately, it is followed by
too many consultants.
DR. REYNOLDS: In response to your first com-
ment, it had been reported in the literature that chlo-
rination of lagoon effluent seriously affected soluble
COD particularly or soluble BOD. thereby releasing
dissolved organics into the effluent. We found that
only occurred in our practice, when excessive amounts
of chlorine were added to the lagoon effluent, specifi-
cally, when breakpoint was approached. As longas the
chlorine residual was less than 5 mg 1, and it was not
above breakpoint conditions, we did not see the dis-
ruption of the algal cell wall and the release of dis-
solved organics. That may explain the discrepancy
between what you have observed in Texas and what
we observed.
We did not observe release of dissolved organics in
the effluent due to chlorination, unless massive doses
of chlorine were used. We studied doses from 0.2 mg I
up to as high as 30 mg 1.
I agree with your comment about application of the
model.
MR. RON SOLTIS, Washington Suburban San-
itary Commission: Was the floating material on top
of the lagoon allowed to be discharged with the efflu-
ent into your filters? I noticed in the pictures what
looked like green material on the top.
DR. REYNOLDS: Yes. The first picture we showed
illustrated that there was some floating algal mass on
top. The discharge at that point was submerged. There
was no attempt to prevent that from going over if it did
occur. So. sometimes it did and sometimes it did not
depending on how things went. Generally, we did not
have mats of algae floating in the chlorine contact
chambers.
MR. SOLTIS: Did the floating material ever cover
the entire lagoon?
DR. REYNOLDS: No
MR. WHITE: You flashed one of those slides too
fast for me. A couple of questions. Was this primary
effluent discharging to lagoons or was it raw sewage
discharging to lagoons?
DR. REYNOLDS: The lagoon system itself treats
raw wastewater without any prior treatment. There
is no comminution. There is a bar screen at the head
34
-------
CHLOR1NA T1ON/DECHLORINA TION
of the lagoon system but no primary treatment and no
preliminary treatment.
MR. WHITE: What was the calculated length of
the residence time in the lagoon?
DR. REYNOLDS: The lagoon system was de-
signed for a wintertime capacity of 180 days and a
summertime capacity of 120 days. There have been
no concrete studies to indicate what the actual hy-
draulic residence time is. We did calculate on one of
the ponds, which had a theoretical residence time
of 9 days, that it was actually 9 hours, so we
would assume that in the primary cell it was sub-
stantially less than the 90 days that it was
designed for. How much less we do not know.
MR. WHITE: How deep were they?
DR. REYNOLDS: They averaged in depth from
about 5 feet in the summer up to about 8 feet during
wintertime storage.
MR. WHITE: What were your initial coliforms just
before disinfection?
DR. REYNOLDS: As 1 indicated, initial coliform
concentrations ranged somewhere between a mini-
mum of 103 up to about I04 or 10s, an average of
around 104.
MR. ZALEIKO, Technical Associates: Did you
get a chance to run any TOC-COD relationships
on both the normal chlorine dosage as well as
the extra heavy dosage?
DR. REYNOLDS: Not that 1 recall. We ran a few
TOC's, but not enough to establish a ratio between
TOC and COD in the chlorinated effluent.
MR. ZALEIKO: So there were no parallel runs of
TOC as well as COD?
DR. REYNOLDS: I do not recall any.
MR. ZALEIKO: Just spot checks on the TOC? 1
was wondering, since the Army gentleman in men-
tioning some of his problems, discussed TOC rather
than COD relationships on his work rather than on
yours. It can make a difference.
DR. HARVEY ROSEN: 1 may be a little embar-
rassed to ask this question, being a chemist, but have
there been any studies done about the potential reac-
tions with respect to toxicity when sulfur is present?
It seems like SOC1 and other things may be formed in
lagoon effluents that would not be formed in most of
the effluents that have been reported relative to the
toxicity studies that tend to be aerobic when chlorin-
ated.
DR. REYNOLDS: I am not aware of any. Now,
some of the other gentlemen who are doing more on
toxicity may know. By the way, the effluent was not
anaerobic when it was chlorinated, it just contained
a high degree of sulfide because of anaerobic condi-
tions in the lagoon, but the total lagoon dissolved
oxygen was still above 6.2 or 6.3 mg/1, even in the
wintertime.
MR. CONVERY: I do know that the Health Effects
Research Laboratory has tried to partition the organic
concentrates from their RO concentrates of drinking
water supplies, and I think they narrowed it down to
468 compounds. So, I am not sure that they can iden-
tify whether it is sulfonated organics or not.
35
-------
7.
DECHLORINATION OF WASTEWATER: STATE-OF-THE-ART FIELD SURVEY
AND PILOT PLANT STUDIES
Henry B. Can,* Ching-lin Chen,* and Albert D. Venosa*
*Los Angeles County Sanitation Districts
**U.S. EPA, Cincinnati, Ohio
INTRODUCTION
Dechlorination is required in six of the County
Sanitation Districts of Los Angeles County (CSDLAC)
water reclamation plants (Figure 1). These plants are
located upstream in the sewerage system to provide
hydraulic relief for downstream sewers. Each day
2.5 x 105m-'(70 million gallons) of filtered chlorinated
secondary effluent are discharged to the Rio Hondo
and San Gabriel River flood control channels for reuse
or overland flow to the ocean. Because most of the
effluent is discharged to navigable waters, the up-
stream WRP's must, therefore, conform to National
Pollutant Discharge Elimination System (NPDES)
requirements.
Discharge requirements for CSDLAC facilities are
set by the Los Angeles Region of the California Water
Quality Control Board. Among other requirements,
the total chlorine residual allowed in effluent dis-
charged to navigable water is generally less than 0.1
NUMBERS IN PARENTHESES DENOTE
PLANT OPERATING CAPACITY
Figure 1. Location of dechlorination facilities sanita-
tion districts of Los Angeles County
mg, I. This requirement has been imposed primarily
for the protection of fish life and other aquatic organ-
isms in the receiving waters. Dechlorination with
sulfur dioxide is employed in CSDLAC facilities to
meet the chlorine standard.
Asa result of the need for dechlorination in CSDLAC
facilities, this study was, therefore, initiated with funds
from the U.S. Environmental Protection Agency's
Municipal Environmental Research Laboratory in
Cincinnati, Ohio. The study consisted of a field survey
of dechlorination facilities in the State and pilot plant
studies.
Objectives of the field survey were to assess the
effectiveness and reliability of full-scale dechlorination
installations in California, which leads the nation in
number of dechlorination facilities. Prime considera-
tions of the survey were on reliability and methods of
control for the sulfonation system and bacterial after-
growth in the dechlorinated effluent. Information for
the survey was gathered through questionnaires,
which were mailed to superintendents of the dechlo-
rination installations. Site visit follow-ups were made
to some of the facilities.
In the pilot plant studies, the effluent was closely
monitored for bacteriological, physical and chemical
degradation effects after dechlorination. The three
methods of dechlorination evaluated were sulfur
dioxide (SO2), granular activated carbon and holding
pond. Emphasis in the study was on SO2, because it
is the most commonly employed method of dechlorin-
ation. Cost effectiveness on all three methods of de-
chlorination was examined.
FIELD SURVEY
General Information
Disinfection of wastewater with chlorine can reli-
ably meet the bacteriological standards for secondary
36
-------
CHLORINA TION/DECHLORINA TION
treatment in California. However, it can result in a
residual chlorine level that is toxic to fish. Available
data indicate that a chlorine concentration of less than
0.1 mg/l is acutely toxic to fish. Dechlorination is
therefore necessary for wastewater discharges that
threaten the ecology of the receiving waters.
The dechlorination requirement for wastewaters in
the State varies from one region to another. There are
nine Regional Water Quality Control Boards in the
State. Each Board sets and enforces the requirements
for the region under its jurisdiction. For dechlorina-
tion, the chlorine discharge limits are set based on the
existing qualities and beneficial uses of the receiving
waters.
The nine California Regional Water Quality Con-
trol Boards (CRWQCB's) were contacted in May 1977
to determine the number of dechlorination installa-
tions in the State. From the list of treatment plants
NOTtS:
Numbers in parentheses denote
dechlorination facilities In
operation.
Other numbers denote regions
of the State Water Quality
Control Board.
Figure 2. Map showing number of operating dechlori-
nation facilities in.California (1977)
provided by CRWQCB, superintendents of individual
plants were contacted by phone in June 1977, to in-
quire if dechlorination was practiced at their plants.
Dechlorination was employed in 31 of the 60 facilities
contacted. The remaining facilities were under design/
construction. Questionnaires were then mailed to the
31 facilities employing dechlorination. Their total
number in each of the regions is shown in parentheses
of Figure 2.
Results from the field survey indicate that sulfur
dioxide (SC>2) is the most widely used dechlorinating
agent in California. It is more popular because of cost
and ease of application. Equipment used for SC>2 is
the same as that used for chlorine, which makes SC>2
application easily adaptable in most wastewater treat-
ment facilities.
The chemical characteristics of the gas have also
added to the attraction for the process. Reaction time
of SC>2 and free chlorine or chloramines is very short.
By-products of SC»2, such as sulfites and chlorides,
have not been shown to be toxic to fish at normal levels
encountered in dechlorination.
Dechlorination is employed mostly in the urban
areas such as San Francisco (Region 2), Sacramento
(Region 5), and Los Angeles (Region 4). The threat to
fish and wildlife is most prevalent in these areas be-
cause of the voluminous amount of wastewater dis-
charges. Receiving streams and rivers are incapable of
assimilating the chlorine discharge from the treatment
plants without endangering fish lives and other aquatic
organisms.
Region 2 dechlorinates approximately 3.0 x lO^m1
(800 million gallons) of wastewater daily to protect
fish and wildlife habitat in the San Francisco Bay
estuaries. The Bay is one of the most important coastal
estuaries in the State. Myriads of fish and wildlife
species utilize the Bay habitats for feeding and nursery
ground. The Bay also functions as the only drainage
outlet for wastewaters in the region. Dechlorination
is, therefore, necessary for wastewater discharges in
the Bay.
Region 5 dechlorinates its wastewater to protect the
fish and waterfowl habitat in the rivers. Region 4 de-
chlorinates the wastewater only in the inland plants to
protect the fish in the low flowing rivers.
The North Coast (Region 1) has three facilities cur-
rently dechlorinating their effluent. Rapid growth has
generated sufficient chlorinated wastewater to threaten
the fish life in the area.
The Central Coast (Region 3) and San Diego (Re-
gion 9) utilize submarine ocean outfall to disperse off
their wastewater discharges. No dechlorination is
practiced in these areas.
The Colorado River Basin (Region 7) and Santa
Ana (Region 8) are water scarce regions. Wastewater
effluents are normally discharged to ephemeral streams
which contained virtually no fish life. Dechlorination
of wastewater is, therefore, not required in these
regions.
The Lahontan Region (Region 6) disposes its waste-
water via land disposal and, therefore, has no dechlo-
rination facilities.
37
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
Questionnaire Responses
A summary of responses to the questionnaires is
shown in Tables 1, 2, and 3. The questions are divided
into three categories: general information, engineer-
ing design information, and operational information
for the SC>2 dechlorination system. Responses to the
questions are tabulated with respect to percent of the
total dechlorination facilities surveyed.
TABLE 1. PROFILE OF DECHLORINATION FACILITIES
IN SURVEY (1977)
Percent of Total
Description Responses*
(a)
(b)
(c)
(d)
(e)
(f)
Startup date of dechlorination facilities
- Before January, 1976
- After January, 1976
Type of treatment preceding dechlorination
- Primary
- Secondary
- Tertiary
Average daily plant flow
- Less than 2.3x1 04m3/d (6 mgd)
- 6 to 10 mgd 2.3x10" to 3.8x104m3/d
(6-10 mgd)
- Greater than 3.8x104mVd (10 mgd)
Sulfur dioxide capacity
- 0 to 45.4 kg/d (0 to 100 Ibs/day)
- 45.8 to 227 kg/d (101 to 500 Ibs/day)
- Greater than 227 kg/d (500 Ibs/day)
Total coliform discharge standard
- Less than or equal to 2.2/100 ml
- Less than or equal to 23/100 ml
- Less than or equal to 100/100 ml
- Less than or equal to 240/100 ml
- Others
Total residual chlorine discharge standard
- 0
- Less than or equal to 0.1 mg/l
- Greater than 2 mg/l
38.7
61.3
9.7
83.9
6.4
68.0
16.0
16.0
12.9
35.5
51.6
22.6
16.1
9.7
41.9
9.7
58.1
29.0
12.9
'Based on 31 respondents
Referring to Table I, dechlorination facilities in the
State are mostly installed after January 1976, thus re-
flecting the recent stringent requirement imposed on
wastewater treatment facilities in California. They are
mostly in secondary treatment plants with an average
daily plant flow of 3.8 to 23 x KPmVdO to6 mgd)and
SO2 capacity greater than 230 kg/d (500 Ibs/d). Sec-
ondary treatment is the minimum requirement for
most wastewater discharges in the State. Consequently,
most plants are in this category, which may account for
the largest percentage of dechlorination found in
secondary treatment.
Before dechlorination, the chlorine residuals in the
treatment plants range from 2 to 10 mg/l. This is to
meet the various disinfection goals shown in Table I.
The disinfection requirement of most wastewater treat-
ment facilities is to meet either a 2.2 or less total coli-
form per 100 ml or 240 or less total coliform per 100
ml. The former limit is for facilities which discharge to
waters for non-restricted recreation usage. The latter
limit is for facilities which discharge to shellfish har-
vesting areas.
The total residual chlorine goal in most wastewater
discharges is generally O.I mg/l or less. Several have
reported total residua! chlorine discharges ranging
from 2 to 6 mg/l. In these plants, it was found that
TABLE 2. ENGINEERING DESIGN INFORMATION OF
DECHLORINATION FACILITIES IN SURVEY (1977)
Percent of Total
Description Responses*
(a) Type of feed control system
- Feedforward
- Feedback
- Feedforward and feedback
- Flow paced
- Residual control
- Flow and residual controls
- Pneumatic flow signal
- Electronic flow signal
- Pneumatic dosage signal
- Electric dosage signal
- Gap residual controller
- Proportional and reset controller
- None
- With multiplier
- Without multiplier
- With adjustable slope factor
- Without adjustable slope factor
(b) Contacting method
- SC>2 injected in mixing chamber
- SC>2 injected in outfall pipe
87.1
9.7
3.2
27.4
27.4
45.2
6.5
93.5
9.7
90.3
16.1
25.8
58.1
35.5
64.5
9.6
90.4
32.3
67.7
- Reaeration provided after 3.2
dechlorination
- Reaeration not necessary after
dechlorination 96.8
- pH adjustment provided after
dechlorination
- Others
3.2
12.9
"Based on 31 respondents
38
-------
CHLORINA TION/DECHLORINA TION
the discharge limitation for chlorine is not required
until several miles downstream. Consequently, the
plant can discharge a higher level of chlorine. Due to
normal chlorine dissipation along the channel. Virtu-
ally all the chlorine is removed by the time it reaches
the required sampling point.
Table 2 summarizes the engineering design informa-
tion gathered in the survey. On the method of SO2
feed, most responses showed a feed forward type
system with primary control based on flow and sec-
ondary control based on residual levels at the chlorine
contact chamber effluent. The secondary signal must
be feed forward rather than a closed loop, since
currently available residual chlorine analy/crs are
not capable of monitoring and controlling on the basis
of a dechlorinated effluent. Both the flow and dose
signals are electronic rather than pneumatic, thus
reflecting the newness of the equipment. Controllers,
multipliers, and adjustable slope factors are not used
in most installations. These devices have been used to
fine-tune the SO2 requirement for dechlorination.
Responses on the SC>2 contacting methods showed
injection of the gas largely occurs through diffusers
TABLE 3. OPERATIONAL INFORMATION OF
DECHLORINATION FACILITIES IN SURVEY (1977)
Description
Percent of Total
Responses*
(a) Is dechlorination system operated
24-hrs daily?
- Yes
- No
93.5
6.5
(b) What is the desirable SO2:CI2 ratio
employed?
- 1 or less 74.2
- greater than 1 25.8
(c) Is overdosing necessary to meet standard?
- Yes 87.1
- No 12.9
(d) Is SO2 feed control system reliable?
- Yes 58.1
- No 41.9
(e) Will system handle drastic fluctuation of
residual chlorine?
- Yes 50.0
- No 50.0
(f) Is biological aftergrowth observed after
dechlorination?
- Yes 6.5
- No 93.5
'Based on 31 respondents
with no mixing chamber provided. Reaction of SO2
and chlorine have been shown to be rapid from rate
equation analyses. Neither reaeration nor pH adjust-
ment has been found necessary in most installations,
which is contrary to what has been anticipated by
many people due to the reducing reactions of SO2.
Table 3 shows a summary on the operational infor-
mation of the plant survey. Approximately 94 percent
of the plants reported operating the dechlorination
system continuously for 24 hours daily. The remain-
ing 6 percent operate the system intermittently, turn-
ing on the SC>2 only at high chlorine flows. Approxi-
mately 74 percent attempt to operate the SO2 close to
the theoretical SO2: chlorine residual ratio of 0.9;
however, most (87 percent) find overdosing necessary
to achieve the low or /ero chlorine residual standard.
Approximately 97 percent reported reaeration to be
unnecessary after SO2 dechlorination, and 97 percent
found no pH adjustment required of the dechlorinated
effluent (Table 2). Biological aftergrowth was not ob-
served in 94 percent of the installations. Approxi-
mately 58 percent felt the SO2 feed control system
used in the plant was reliable.
Site Visits
Table 4 lists the dechlorination facilities visited on
the trip. The facilities are located in the San Francisco,
Sacramento, San Jose, Santa Rosa, and Los Angeles
areas. The purpose of the trip was to study various
SO2 feed control methods and interview operators
regarding operation of their system. A total of 17 in-
stallations were visited. All the wastewater treatment
plants visited, have secondary treatment. The plant
How capacities ranged from 1.1 x 104to3.0 x 10sm-'/d
(3 to 80 mgd). The SO2 capacities ranged from 200 to
6900 kg/d (450 to 15,200 Ibs/d).
Figure 3 shows the most common method of SO2
dechlorination observed in the site visits. In the sys-
tem, a feed forward residual signal and a feed forward
flow signal are fed to the sulfonator. These two sig-
nals are sometimes combined into a product signal
through an electronic multiplier before feeding to the
sulfonator. This is done to avoid having to excessively
overdose the chlorinated effluent with SO2.
To improve the system further, an electronic ratio
controller is adopted after the multiplier in some
facilities. The ratio station maintains the desired ratio
between the flow and residual signals combined in the
multiplier. This provides a more precise control on the
SO2 dosage requirement for dechlorination.
Alternate methods have been devised to overcome
inability of the chlorine analy/er in providing a feed-
39
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 4. DECHLORINATION FACILITIES VISITED
IN SURVEY (1978)
Dechlorination
Facilities
City Main WTP
San Jose/Santa
Clara WPCP
Northeast WRP
Arden WTP
San Pablo
San. WPCP
Main Treatment
Plant
Hayward WTP
Pomona WRP
Harold May WQCP
Laguna WTP
Valencia WRP
Cordova WTP
Richmond WPCP
Irvington WTP
Newark WTP
Alvarado WTP
West College WTP
Operated
by
City of Sacramento
City of San Jose
County of
Sacramento
County of
Sacramento
San Pablo San. Dist.
San Rafael San. Dist.
City of Hayward
L.A. Co. San.
Districts
City of Palo Alto
L.A. Co. San.
Districts
L.A. Co. San.
Districts
County of
Sacramento
City of Richmond
Union Sanitary Dist.
Union Sanitary Dist.
Union Sanitary Dist.
City of Santa Rosa
SO, Daily Avg.
Capacity, Flow,
kg/day m'/d
(Ibs/day) (MGD)
6,900 1.8X105
(15,200) (48)
3,450 3.0x1 05
( 7,600) (80)
2,724 5.7x10"
( 6,000) (15)
1,725 1.9x10"
( 3,800) ( 5)
908 3.0x10"
( 2,000) ( 8)
908 1.1x10"
( 2,000) ( 3)
908 4.5x10"
( 2,000) (12)
908 2.3x10"
( 2,000) ( 6)
863 1.1x105
( 1,900) (30)
454 1.1x10"
( 1,000) ( 3)
454 1.1x10"
( 1,000) ( 3)
431 7.6x103
( 950) ( 2)
409 2.7x10"
( 900) ( 7)
227 2.3x10"
( 500) ( 6)
227 1.9x10"
( 500) ( 5)
227 1.1x10"
( 500) ( 3)
204 1.5x10"
( 450) ( 4)
back signal to the sulfonator. Figure 4 shows a sche-
matic of such a system.
In alternate No. 1, a two-stage method of dechlo-
rination is used. Analyzer No. 1 is used to instruct
Sulfonator No. 1 to dechlorinate to a 10:1 ratio of the
discharge limit. The analyzer performs best within
a 10 to 1 setting. Calibration is maintained because of
the continuous presence of chlorine residual in the
effluent. Sulfonator No. 2 is then used to remove the
remaining residual chlorine. Because the residual
chlorine has been reduced to 1 mg/1 or less in the first
stage, excessive overdose of the SO2 with sulfonator
No. 2 is avoided.
Figure 3. Feed control system most commonly em-
ployed in sulfur dioxide dechlorination
facilities in California.
In alternate No. 2. a biased residual chlorine signal
is sent through the analyzer to keep it in calibration.
A feedback residual signal from the dechlorinated
effluent greater than the biased signal signifies in-
complete dechlorination. The S"O2 is paced to dose
proportional to any signal greater than the biased
signal.
The simple feed forward SC>2 feed control system is
adequate for most dechlorination installations. It re-
quires a small capital investment and offers simplicity
of controls. Disadvantage of the system is that SC>2
overdosing is necessary to accomplish disinfection.
Overdosing (SO2) cost is a significant factor in large
dechlorination installations. Alternate SO2 feed con-
trol systems are beneficial for the large installation. It
reduces the SC>2 overdose requirement and hence the
operating chemical cost.
The weakest link in the SO2 feed control system is
the chlorine residual analyzer. The measuring electrode
of the analyzer loses its sensitivity rapidly in a de-
chlorinated effluent. The presence of chlorine residual
helps prevent oxidation on the electrode. Abrasive
grits in the measuring cell block are incapable of pre-
venting oxides from forming on the electrode in the
absence of chlorine.
ALTERNATE NO. I
CHLORINAT
SULFONATOR
1
ED j)
SET POINT
ANALYZER TO
10- 1 RATIO OF
DISCHARGE
LIMIT
L !
+ ALTERNATE NO. 2
FEED BACK
RESIDUA
I
FEED FORWARD
FLOW SIGNAL
NFLOWMETER
EFFLUENT "V^
Figure 4. Feed
. SIGNAL CHLO
ANAL
SOj SOLUTION
1
I BIASED
RINE _| RESIDUAL
rzER *• I CHLORINE
| SIGNAL
I . _ _ .J
SAMPLE SOLUTION
DECHLORINATED
EFFLUENT
control systems used in dechlorination
facilities to avoid excessive SO2 overdose.
40
-------
CHLORINA TION/DECHLORINA TION
Figure 5 shows a schematic of a residual chlorine
sampling cell. Measuring of the chlorine takes place
in the measuring cell block. The platinum is the ref-
erence electrode and the copper is the measuring elec-
trode. A small D.C. current is produced in the presence
of free chlorine or iodine proportional to its concen-
tration. This current is measured by a recording
ammeter in terms of mg/1 of chlorine. The efficient
operation of the analyzer is dependent upon several
factors, all of which are equally important. The cell
block and electrodes have to be free from biological
or chemical fouling of the electrodes. Abrasive grit
and installation of filters have been successful to some
extent in preventing contamination of the electrodes.
The buffer solution must be maintained since the
analyzer is calibrated to read accurately at a specific
pH. The flow and pressure of the sample water must be
fairly constant, since they affect cell current produc-
tion. A malfunction in any of the above conditions
will give a false reading and result in underdosing or
overdosing of the
BUFFER
SOLUTION, pH
WATER
REGULATOR
t
MEASURIN
CELL
BLOCK
IP. Cu
•<
1 \
G
WASTED
SAMPLE WATER,
FLOW 8 PRESSURE
Figure 5. Critical components of a chlorine residual
analyzer
Poor analyzer performance has been found in SC>2
dechlorination installation where there are high in-
dustrial waste discharges. Plugging of the measuring
cell block frequently occurs. Installation of filters
preceding the cell block did not improve the analyzer
reliability.
Operators' comments on the reliability of the sul-
fonation system have been generally satisfactory.
Most felt a slight overdose of the effluent'with SO2
is necessary to meet the discharge requirement of 0.1
mg/1 or less chlorine residual consistently. Biological
growth in the dechlorinated effluent was not observ-
able in most plants since the effluents were usually
discharged in a submerged outfall pipe or intermingled
with the receiving waters containing natural fresh
water slime growth. However, laboratory reports
from these plants indicated an increase in total coli-
form after dechlorination. As a result, coliform dis-
charge limitations were usually exceeded at the point
of discharge. The Regional Water Quality Control
Board has now allowed the treatment plants to sample
before dechlorination to meet the disinfection stan-
dard.
A continuous monitoring of the dechlorinated
effluent is required in some cases for compliance with
the discharge standard of CRWQCB. However, since
a continuous recording of zero chlorine on the chart
is not possible because of the limitation of the analyzer,
the CRWQCB has allowed monitoring of the dechlo-
rinated effluent for residual compliance once every
hour. The analyzer monitoring the chlorinated efflu-
ent is used to monitor the dechlorinated effluent for
residual chlorine compliance. Because this is done
only a few minutes every hour, the analyzer is kept in
calibration.
PILOT PLANT STUDIES
General Information
Figure 6 shows a schematic of the three dechlorina-
tion systems evaluated at the Pomona Research Facil-
ity. Unchlorinated secondary effluent from the Pomona
Water Reclamation Plant was pumped to the Pomona
Research Facility. The effluent was then chlorinated
in a pilot system for two hours and then fed to the
dechlorination systems. The sulfur dioxide and hold-
SULFUR DIOXIDE
ACTIVATED CARBON
CHLORINATED
EFFLUENT
CARBON'
•FILTER
DECHLORINATED
EFFLUENT
HOLDING POND
.CHLORINATED r
EFFLUENT ***
HOLDING POND
T DECHLORINATED
EFFLUENT *
Figure 6. Flow diagram of dechlorination pilot plant
systems at Pomona, California.
41
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
ing pond studies were conducted recently. The acti-
vated carbon study was done several years ago as part
of another study.
DESCRIPTION OF PILOT PLANT
Sulfur Dioxide
The sulfur dioxide (SO2> feed control system used
in the pilot studies was obtained from a previous
chlorination study. It employed only a feed forward
chlorine residual signal to dechlorinate the effluent.
The SO2 was paced in proportion to the chlorine
measured in the effluent. Because the flow was held
constant, no flow signal was required. A multiplier
unit and ratio station was not considered for the pilot
systems because the objective here was to study the
effects on water quality after dechlorination. Sim-
ilarly, the other alternate schemes found in the
field survey were not considered.
The chlorine residual of the effluent feeding through
the SO2 dechlorinations system averaged 7 mg 1.
This was continuously monitored by a chlorine ana-
lyzer and a recorder. An electronic signal was sent
from the analyzer to the sulfonator to control SO2
dosage based on chlorine residual. The sulfonator was
readjusted to feed SO2 in direct proportion to the
chlorine in the effluent.
Chlorinated effluent flow through the SO2 dechlo-
rination system averaged 95 1/min (25 gpm). A steel
tank measuring 13.7m by 0.9m (45 feet by 3 feet) deep
was used as a mixing and contact chamber. SC>2 was
added to the chlorinated effluent in the mixing cham-
ber at ratios ranging from 1-2 mg/1 of S(>2 to 1 mg/l
of chlorine residual. The ratio employed in each exper-
iment was dependent on the SC>2 required to remove
the chlorine in the pilot studies. Theoretical contact
time in the mixing chamber was 10 minutes. From the
mixing chamber, the dechlorinated effluent was fed
through the contact chamber to study bacterial after-
growth and physical-chemical degradation of the
effluent after dechlorination. The theoretical contact
time in the chamber was two hours.
TABLE 5. AVERAGE* WATER QUALITY
CHARACTERISTICS OF CHLORINATED ACTIVATED
SLUDGE EFFLUENT IN PILOT PLANT STUDIES
Parameters
PH
Dissolved Oxygen (mg/l)
Sulfite (mg/l SO3)
Alkalinity (mg/l)
Total Dissolved Solids (mg/l)
Total Residual Chlorine (mg/l)
Before
S02 DeCI2
7.2
4.8
0.0
172.0
499.0
5.0
After
SO2 DeCI2
7.2
6.1
2.6
164.0
471.0
0.0
'Average of 56 samples
Activated Carbon
Key operating parameters of the activated carbon
dechlorination system may be summarized as follows.
The activated carbon was a Calgon Filtrasorb 400
with 12 X 40 U.S. standard sieve size. It was contained
in a steel tank measuring 1.8m (6 feet) in diameter
and 4.9m (16 feet) high. Carbon depth in the column
was (3.1 m) 10 feet. Chlorinated secondary effluent
flowed downwards through the carbon at a loading
rate of 0.28m'min m: (7 gpm ft:). The empty bed de-
tention time at this rate was 10 minutes. Sampling
parts were provided at various depths in the column.
Holding Tank
An open steel tank of the same size as the SC>2 de-
chlorination tank was used. Chlorinated secondary
effluent at 5 mg 1 chlorine residual was fed to the hold-
ing tank and held until the chlorine had dissipated.
Samples in all three systems were collected before
and after dechlorination. In addition, for the sulfur
dioxide and holding pond experiments, samples were
also collected at several periods after dechlorination.
Both 24-hour composite and grab samples were col-
lected during the pilot plant operation. The composite
samples were analyzed for water quality. The grab
samples were analyzed for bacterial aftergrowth.
Pilot Plant Results
The pilot plant studies addressed the question of
water quality and bacteria regrowth after dechlorina-
tion. All three systems were evaluated with emphasis
on the sulfur dioxide dechlorination system.
TABLE 6. AVERAGE* WATER QUALITY
CHARACTERISTICS OF CHLORINATED ACTIVATED
SLUDGE EFFLUENT IN PILOT PLANT STUDIES
Parameters
Total COD (mg/l)
Dissolved COD (mg/l)
Color (units)
Turbidity (Jtu)
Suspended Solids (mg/l)
Total Residual Chlorine (mg/l)
Before
.Carbon DeClj
19.0
16.0
14.0
1.8
2.4
8.0
After
.Carbon DeClj.
13.0
10.0
4.0
2.4
2.1
0.0
"Average of 52 samples
Water Qiialitv Effects
Tables 5 and 6 summarize the water quality charac-
teristics before and after dechlorination for the sulfur
dioxide and activated carbon dechlorination systems,
respectively. No data are shown for the holding pond
experiments since no change other than the chlorine
content was observed in the effluent.
Referring to Table 5, no adverse effects on the water
quality were found after dechlorination with SO2. The
42
-------
CHLORINA TION/DECHLORINA TION
pH remained relatively unchanged at 7.2. This may
have been due to the high alkalinity (172 mg/1) in the
Pomono effluent. Dissolved oxygen was increased
from 4.8 mg; 1 to 6.1 mg 1. The increase could be at-
tributed to turbulent mixing in the dechlorination
chamber. The average sulfite level in the effluent was
2.6 mg/1 due to a slight overdose of SO2. Alkalinity
was reduced from 172 mg/1 to 164 mg/1 possibly due
to reaction with the SC^. This decrease may have re-
sulted in the reduction of total dissolved solids
from 499 mg/1 to 471 mg/1.
Dissolved oxygen depletion and pH reduction after
SC>2 addition have been anticipated by many people
because of the reducing quality of SO2- These were
not observed either in the pilot p'ant results or in the
data collected in the plant full scale survey. A brief
study during the pilot plant runs shows an SC>2 over-
dose as high as 50 mg/1 is necessary to incur any
significant change in the above parameters.
For the activated carbon system (Table 6) no ad-
verse effects on the water quality were observed in the
dechlorinated effluent. The effluent quality after
chlorine by the carbon did not seem to alter the
ability of the carbon for removing COD (total
and soluble) and color.
TABLE 7. JAR TEST RESULTS IN BACTERIA
AFTERGROWTH AFTER DECHLORINATION
Experiment
No.
1
2
3
4
5
6
7
Chlorinated
Effluent
(2 hrs)
2
<2
5
<2
2
2
2
Dechlorinated in
Sample Bottle
(10 min)
2
<2
2
2
2
<2
<2
Note: All results expressed at total conform MPN
Dechlorinated In
Pilot Plant Tank
(10 min)
49
79
33
49
33
49
490
per 100 ml.
Bacteriological Aftergrowth
Bacteria measured for this phase of work included
total and fecal coliforms, total aerobic plate count at
35 °C, fecal streptococci, and Salmonella. All tests
were performed in accordance with Standard Methods
(1) except for the Salmonella spp. which were isolated
using methods recommended by the District's micro-
biology laboratory and the EPA project officer.
Figure 7 shows typical total coliform results for the
dechlorination systems. An increase in total coliforms
shortly after dechlorination was consistently observed
in both the filtered and unfiltered dechlorinated sec-
ondary effluents. The origin of the increase was not
known. It was presumed that the increase might be
due either to a reactivation of the bacteria in-.
jured during chlorination or to contamination
from the tank. Contamination from the air was
not considered likely since the sulfonation cham-
ber was covered throughout the tests.
Jar tests were performed in the laboratory to deter-
mine if the total coliform increase could be attributed
to repair of injured bacterial cells. A chlorinated sec-
ondary effluent sample from the pilot plant was
collected in a sterilized beaker and then dechlorinated
with SO2 solution (sulfurous acid). Table 7 is a sum-
mary of the results. No increase in total coliform
density was observed in the beaker after dechlorina-
tion.
The dechlorination systems were subsequently in-
vestigated for possible sources of contamination. For
the SO2 system, the mixing and contact chambers
o
-. 2.0
LJ
CO
^
UJ
o:
0 1 5
?
^
COLIFORI
b
_i
< 0.5
H
O
r-
n
LOG AVERAGE OF 19 SAMPLES
—
-
_
LTERED
U.
Q
MFILTERE
^
LTERED
u_
Q
UJ
rr
UJ
I-
iZ
~z.
—
—
-
—
Figure 7.
10 MINUTES 2 HOURS
PERIOD AFTER DECHLORINATION
Total coliform responses in dechlorinated
secondary effluents of pilot plant studies
were thoroughly cleaned and disinfected. For the car-
bon system, freshly regenerated carbon was used. The
holding tank was also thoroughly cleaned and dis-
infected prior to the experiment. The systems were
then put in operation and the dechlorinated effluent
samples collected at 1,2, 3, and 4 days after startup.
For the holding tank, the samples were collected as
soon as the chlorine had dissipated.
Figure 8 shows a summary of typical results
for all three systems. Total coliforms were in-
creased by about two log units in all three
dechlorination systems within the first four days.
It is not known why the increase leveled off after
43
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
four days. The gradual total coliform increases
suggest that the source of contamination may be
due to a buildup of bacterial growth in the
dechlorination systems. Slime and scum growth
were consistently observed in the sulfonation
chamber a few days after operation. It is
presumed that the same buildup may have oc-
curred in the activated carbon column, although
this was not observable because of the enclosed
steel tank used. Growth in the holding pond was
also observed after the chlorine residual had
dissipated.
0.(
01
o
co _
< 2
LJ
cc
o
cc
o
8 i
<
o
HOLDING POND-
SULFUR
DIOXIDEr
10 minute
sample
CARBON, 10 minute
sample
( /' TOTAL COLIFORM MPN PRIOR
TO
I
I
I
I
01 2345
PERIOD AFTER STARTUP, days
Figure 8. Rate of contamination after initial startup in
clean dechlorination pilot plant systems.
Besides the total coliform analyses, other bacteri-
ological analyses were performed. These tests were
conducted only in the SOa experimental phase. Figure
9 is a summary of the results. Increases were found in
the fecal coliform and total plate count populations.
These increases (less than 0.5 log unit) were signifi-
cantly lower than those found for the total coliform
populations. No change in the fecal streptococci count
was found after dechlorination.
To pursue the increase of the total and fecal coli-
forms further, photomicrographs of slime, scum, and
foam samples collected at the head end of the dechlo-
rination chamber were taken. The slime is the slippery
substance that attaches to the insides of the tank; the
scum is the deposit at the bottom of the tank; and the
foam is the flocculant material that floats on the sur-
0.5
- 0.4
to
UJ
cc
0.3
cc
UJ
o
m
0.2
O.I
LOG AVERAGE OF 19 SAMPLES
-
—
—
-
OJ
O
f*v
CC
UJ
1—
1 1
^
CO
MINUTE
O
OJ
O
UJ
Q
CC
1 1 1
) —
u.
<
co
cc
o
OJ
PJ
0
o
cr
UJ
r-
U-
-------
CHLORINA TION/DECHLORINA TION
The foam had MPN's of 6.0 x lO'Vgm material and
6.0 x I05/gm material for the total and fecal coliform,
respectively. The slime had MPN's of 39; 100 ml and
23 /100 ml for the total and fecal coliform, respectively.
In addition to the above bacteriological tests, Sal-
monella analyses were also performed. Two enrich-
ment media were employed to isolate the Salmonella;
namely, dulcitol selenite broth (DSE) and tetrathi-
onate broth (TB). Both have been shown to be effective
media for enriching Salmonella. Following enrich-
ment, isolation and confirmation of Salmonella col-
onies were carried out. Isolation media used were
Xylose Lysine Desoxycholate (XLD) and Bismuth
Sulfite (BS). Confirmation of typical colonies was
performed biochemically using Lysine Iron Agar,
Triple Sugar Iron Agar, and Urea Broth, and serologi-
cally using Salmonella polyvalent-O antiserum. A
total of 26 tests were conducted. All were qualitative
tests since low numbers of Salmonella were antici-
pated due to chlorine disinfection in the effluent. The
results provided an estimate of sampling volume size
for the quantitative tests performed later.
TABLE 9. SALMONELLA ISOLATIONS IN EFFLUENTS OF
PILOT PLANT STUDIES WITH KNOWN
SALMONELLA INPUT
Chlorinated
Sample Activated Sludge
Date Effluent
3/29/78 +
4/05/78 +
5/08/78 +
5/31/78 +
6/19/78 +
6/28/78 +
7/06/78 +
7/10/78 +
7/12/78 +
7/17/78 +
7/24/78 +
7/28/78 +
8/02/78 +
8/23/78 +
8/25/78 +
Dechlorinated
Activated Sludge
Effluent
-
-
-
-
-
-
+
-
+
-
+
-
-
-
-
Table 9 summarizes the qualitative results of the
Salmonella tests. Only results where Salmonella were
isolated in the unchlorinated effluent are presented.
Of the 26 unchlorinated samples. Salmonella were
isolated in 15. Of these samples, only three, or 20 per-
cent of the samples, showed Salmonella presence after
dechlorination. It is not known if the Salmonella pres-
ence was due to regrowth after dechlorination.
Table 10 lists the quantitative results and the sample
volume filtered for each corresponding test. The pro-
cedure used was the same as the qualitative method
except that 16-24 liter samples were filtered; the filters
were placed in 10 ml enrichment broth and homogen-
ized in a Waring blender; and the resulting homogen-
ate tested by the MPN procedure, using tetrathionate
broth at 37°C. No Salmonella were found in the de-
chlorinated effluent even though positives were found
in the unchlorinated effluent. The Salmonella counts
for the unchlorinated secondary effluent ranged from
8 to 40 per liter sample. The Salmonella counts after
dechlorination were less than 0.06 per liter sample.
The lowest detection limit for these tests was 0.05
per liter sample.
TABLE 10. QUANTITATIVE ANALYSES FOR
SALMONELLA IN EFFLUENTS OF PILOT PLANT STUDIES
Sample
Date
8/15/78
8/16/78
8/17/78
Chlorinated Activated
Sludge Effluent
(Sa/mone«a/liter)
10 .
8
40
Dechlorinated Activated
Sludge Effluent
(Sa/mone//a/liter)
<0.05
<0.05
<0.06
The results showed that prevention of slime growth
in the dechlorination test chambers is virtually im-
possible. The growth appears to be similar to that
found in natural streams. These slime growth in de-
chlorinated effluent may be attributed to available
nutrients in wastewater, the moderate ambient tem-
perature and the lack of a toxic chlorine residual. It
is felt that the presence of chlorine is an important
factor in deterring this growth. This was learned from
the site visits to dechlorination installations in North-
ern California.
COST ESTIMATES
Cost estimates for three dechlorination systems
were derived using the information gathered in the
study. These were the SOj, activated carbon, and
holding pond dechlorination systems. In deriving the
costs, it was assumed that the wastewater treatment
plant capacity is 38000m'/d (10 mgd) with 5 mg/l
residual chlorine to be dechlorinated. The residual
chlorine concentration is based on a requirement to
meet the State's disinfection standard of 2.2 or less
total coliform MPN per 100 ml.
Table 11 summarizes results of the cost analyses.
The amortization period for the capital costs was 15
years at an annual interest rate of 8 percent.
Sulfur Dioxide
Total capital cost for the system was $100,000. The
cost includes two 390 kg/d (850 Ibs/d) sulfonators
with one used as standby, control building, piping and
45
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
instrumentation. The sulfonator was sized based on a
SO2:C12 ratio requirement of 2:1. In the pilot studies
at Pomona this ratio was found necessary when high
alkalinity of 200 mg/ 1 was found in the chlorinated
secondary effluent. The theoretical SC^C^ ratio re-
quirement is 0.9:1.0. No provision was made for a
contact chamber since the SOj, Cl2 reactions are rapid
as found in the pilot plant studies and treatment plant
survey.
TABLE 11. COST ESTIMATES FOR A 38000mVd (10 mgd)
ACTIVATED SLUDGE DECHLORINATION FACILITY
Dechlorination
Process
Sulfur Dioxide
Activated Carbon
Holding Pond
Capital
Cost •
«/3.8m'
(C/1000 gal)
0.4
2.3
5.4
Operating
Cost
C/3.8m>
(C/1000 gal)
1.3
10.5
0.4
Total
Cost
e/3.8m>
(C/1000 gal)
1.7
12.8
5.8
Total operating cost for the SO2 dechlorination
system was $50.000 per year. The cost was based on a
chemical cost of $220 per ton of SO2 and an annual
salary of $15,000 for one operator.
Total cost for the SC>2 dechlorination system was 1.7
cents to treat 3.8m1 (1000 gallons) of chlorinated sec-
ondary effluent. Of this total, 76.5 percent was for
operation and maintenance and 23.5 percent was for
capital recovery.
Carbon Dechlorination
Total capital cost for the system was $400,000. Capi-
tal cost for the carbon includes the dechlorination
column and holding tanks, regeneration furnace,
instrumentation, piping, valves, pumps and initial
carbon load. The carbon filters were si/ed based on a
chlorine adsorption capacity of 0.068 kg (0.15 Ibs) Clg
per 0.45 kg (1 Ib) carbon for 1.8m (6 feet) of carbon
depth. Loading rate for the carbon was 0.28m1/ min/ m:
(7.1 gpm/ft2) for an empty bed detention time of 10
minutes. Calgon Filtrasorb 400 carbon with U.S. stan-
dard mesh si/e of 12 x 40 was used. Cost of this carbon
was 58 cents per 0.45 kg (1 Ib) carbon.
Total operating cost for the carbon dechlorination
was $400,000 per year. Operating cost for the carbon
was based on 45 days regeneration frequency. Re-
generated carbon cost was 35 cents per 0.45 kg (1 Ib)
C, based on information generated at the Pomona Car-
bon Filtration Plant. Carbon loss during regeneration
was assumed at 6 percent. One full time operator was
considered for the system. Operators for the regenera-
tion system were not considered since the regenerated
carbon cost includes the operators' cost.
Total cost for the carbon dechlorination system was
12.8 cents to treat 3.8m1 (1000 gallons) of chlorinated
secondary effluent. Of this total, 80 percent was for
operation and maintenance and 20 percent was for
capital recovery.
Holding Pond
Total capital cost for the holding ponds was
$1.700,000. This includes excavation and lining the
pond with 10 cm (4 inch) gunite. The pond was sized
to allow a detention time of five days for the 5 mg; 1
chlorine residual to dissipate. This dissipation time
was derived in the pilot studies and is based on abun-
dant sunlight during most part of the day as in South-
ern California.
Total operating cost for the holding pond was
$15,000 per year. This was to pay the salary for one
operator.
Total cost for the holding pond dechlorination
system was 5.8 cents per 3.8m1 (1000 gallons) of de-
chlorinated effluent. Of this total, 93 percent was for
capital and 7 percent was for operation and main-
tenance.
The cost data confirm the economic justification
of the municipalities in California for using SOg over
the other dechlorination processes. The cost presented
is fora 3.8 x I04m-Vd(10 mgd) plant. In an EPA cost
analysis (2) for SO2 dechlorination, the cost for a 3.8 x
I01m1 d (1 mgd) dechlorination facility was shown to
be many times higher than for a 3.8 x 104m1 d (10
mgd) plant. The same report showed the cost for a
3.8 x lO'm-Vd (100 mgd) plant to be less than for the
3.8 x l04mVd(IO mgd) plant.
Dechlorination by the other methods such as carbon
or holding pond does not seem to be economically
feasible. Water quality effects and bacteriological
aftergrowth after dechlorination were observed to be
similar to SOj dechlorination.
The use of carbon for dechlorination may be attrac-
tive when removal of trace organics is required. In the
pilot carbon studies, it was found that the adsorption
capacity of carbon for the organic material is only
slightly reduced by the presence of chlorine. The total
dissolved COD removal was reduced by only 5 percent
in the carbon treating chlorinated effluent after 90
days onstream.
The use of a holding pond for dechlorination may
be attractive in a region where land is plentiful and
inexpensive. The presence of sunlight is an important
criterion for this process installation. An advantage of
the system is that virtually no mechanical equipment
needs to be used. Considerable savings in operating
cost and many years of services are obtainable with
this system.
46
-------
CHLORINA TION/DECHLORINA TION
CONCLUSIONS
I. Sulfur dioxide (SOa) is the most preferred method
for dechlorination of wastewater in California.
2. The feedforward method of control with flow as
the primary signal and residual chlorine as the
secondary signal is most commonly used in SO?
dechlorination facilities.
3. Overdosing the chlorinated wastewaters with SO2
is essential to accomplish consistent dechlorina-
tion.
4. Excessive overdose of SO2 can be avoided by
using discrete instruments and alternate methods
of feed.
5. Except for the residual chlorine analyzer, the
equipment in an SO2 feed control system is reli-
able.
6. The analyzer is the weakest link in a SO2 feed
system. Most analyzers manufactured today are
incapable of maintaining calibration in the ab-
sence of chlorine.
7. No significant physical-chemical degradation of
the effluent was found after dechlorination with
SO2- Depletion of dissolved oxygen or change in
pH was not observed in the pilot studies at SO2
dosage to residual chlorine ratio of 2:1.
8. Bacteriological aftergrowth in some microorgan-
ism populations was found after dechlorination.
This was observed predominantly for the total
coliform group. Some increases in fecal coliforms
and other bacteria (as detected in the total plate
count) were also found to increase in the dechlo-
rinated secondary effluent. Salmonella was not
detected in most of the samples. Fecal streptococci
in the effluent remained relatively unchanged
after dechlorination.
9. The bacterial increase in the dechlorinated efflu-
ent is attributed to contamination from bacteri-
ological aftergrowth in the dechlorination
chamber rather than reactivation of injured bac-
terial cells.
10. Sulfur dioxide dechlorination is a more cost-
effective process than either activated carbon or
holding pond.
11. Similar effects on water quality and bacteriological
contamination of the dechlorinated effluent were
found for the activated carbon and the holding
pond as with the
ACKNOWLEDGEMENTS
The dechlorination study has been jointly sponsored
by the Sanitation Districts of Los Angeles County and
the U.S. Environmental Protection Agency.
The cooperation from respondents of the question-
naires in the survey and plant personnel in the site
visits is greatly appreciated.
Cieorge C. White gave invaluable assistance in the
field survey.
Credits for this research are extended to staff of the
Pomona Research Laboratory and the San Jose Creek
M icrobiology Laboratory.
REFERENCES
I. "Standard Methods lor the Examination ol Water and Waste-
vvater." 14th ed.. Amer. Pub. Health Assn.. New York (1975)
2. U.S. EPA. "Disinfection ol Water- Task Korce Report." MCD-
21. No. FPA-430 9-75-012. Centrali/ed Mailing Lists
Services (X FSS). Denver Federal Center. Denver. Colorado
(1975)
DISCUSSION
MR. RON SOLTIS, Washington Suburban San-
itary Commission: How long have the carbon col-
umns been in service?
MR. GAN: You mean the carbon that we used for
the dechlorination study?
MR. SOLTIS: Yes, for the dechlorination studies.
MR. GAN: Those are freshly regenerated carbon.
MR. SOLTIS: Do you have any facilities in Cal-
ifornia that use this as a routine method of dechlorina-
tion?
MR. GAN: Not that I know of.
MR. VENOSA: You established a cost of dechlor-
ination. What did you assume to be the chlorine resid-
ual to establish your costs?
MR. GAN: The chlorine residual is 5 mg/l. This is
the condition for the Pomona Water Reclamation
Plant.
MR. BUONOMO, Envirotech, NationalSonics,
Delta Scientific: I want to comment on your state-
ment that at close to zero residual chlorine you had
difficult problems monitoring and therefore maybe
that explains why closed loop control was only 58%
effective. Your statement does hold for naked elec-
trode amperometric type technique for monitoring
residual chlorine. We have been using that for the last
10-15 years. With the polarographic membrane tech-
nique, you can go to zero or close to zero residuals and
47
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
have excellent recovery capabilities. That is more than
a statement of theory. On the West Coast, six utilities,
funded at Trojan Nuclear, Portland General Electric.
proved that they could achieve closed loop control and
approach /ero residual using the polarographic mem-
brane. Incidentally, the maximum discharge limita-
tions there were 0.05 parts per million.
MR. ED JONES, District of Columbia Govern-
ment, Blue Plains: Do you think that 3 parts per
billion is reasonable treatment for chlorine residual?
MR. GAN: You want my opinion on it?
MR. JONES: Yes, that has been the initial decision
Region III imposed on Blue Plains.
MR. GAN: No, 1 do not think it is a reasonable
level.
48
-------
8.
ROUNDTABLE DISCUSSION OF CHLORINATION/DECHLORINATION
Participants in Chlorination/Dechlorination
Roundtable
John J. Convery, Moderator
U.S. EPA, MERL-Cincinnati
I. Endel Sepp
California State Dept. of Health
2. Richard J. Hansen
California Dept. of Fish & Game
3. James H. Reynolds
Utah State University
4. Henry Can
Los Angeles County Sanitation District
5. George White
Consultant, San Francisco
6. Karl Longley
U.S. Army "
MR. SAULYS, EPA, Chicago, Illinois, Region
Five: In the beginning, there was some mention about
the design factors for chlorination systems. I was
particularly interested in the data that were presented
on the lagoon systems. I wonder if anybody has in-
vestigated more fully the use of alternative forms of
chlorine besides gaseous, as a means of optimizing or
controlling chlorine feed, particularly, for small sys-
tems, say a hundred thousand gallons per day.
MR. WHITE: A hundred thousand gallons per day
. . . you could use hypochlorite, but, do not forget, if
you use hypochlorite, which would be versus, say 150
pound cylinders, I believe you are talking about three
and half to four times the cost of chlorine gas. The
installation itself is a standoff. Diaphragm pumps are
legion. You can use the same automatic controls for
diaphragm pumps. You can have a residual signal plus
a flow signal, the same as you can for chlorination
equipment. The capital cost is about the same, but
your chemical cost would be about three to four times
the quantity that you would buy for an 0.1 MGD plant.
MR. SAULYS: We did not find, by the way, any ref-
erence in the literature, other than for gaseous chlo-
rine. The only advantage I could see to have the hypo-
chlorite, would be from an operator safety point of
view, especially if you have a very inexperienced
operator and you want to protect him, and if you are
not too worried about cost. Otherwise, I would agree
with what has been said.
MR. WHITE: First of all, when using a hypochlo-
rinator, you have to have power, but you do not have
to have a booster pump to operate the injector, as you
do in the gaseous chlorine. There are some little minor
differences in that, but the biggest difference comes
when you are treating small water supplies. This is a
different ball game. Hypochlorite is very effective and
very interesting. The other thing is that in any chlo-
rinator installation that small, where you are using 150
pound cylinders, and because the amount of gas with-
drawn is at such a slow rate, the ambient temperature
change . . . diurnal temperature change, affects the
way you build the house for a gas chlorinator. It has
to be better insulated. You do not have to worry about
anything except freezing, insofar as a hypochlorinator
is concerned. I think that about covers it.
MR. OPATKEN: This one is for Endel Sepp. I
thought that, in your Paper about the first plant you
tested, you said your effluent had a BOD of about 30,
and, in the second plant, you had BOD's of approxi-
mately five. Then, I thought you said the applied dose
was the same in both cases. And then, I thought I heard
you say that the residual for the second case, or the
good quality, was about 1.5 to 2 mg/1. And, for the
poor quality, or the thirty BOD, it was about 5 or 6
mg/1. Is that correct?
MR. ENDEL SEPP: I do not know. Let me look it
up. For the poor quality plant which was the pilot
plant, the residuals were 5 to 6.5 mg/1.
MR. OPATKEN: And the dose is still 10 to 11?
MR. SEPP: Yes.
MR. OPATKEN: Then, for the good quality waste-
water, your residual is down, but your dose is still the
same?
MR. SEPP: They were slightly less. I think the re-
49
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
siduals were lower than that. I do not have the data
here.
MR. OPATKEN: Can somebody explain that one
from the board there?
MR. SEPP: One reason was that at San Pablo, the
effluent was nitrified, and there probably was break-
point chlorination taking place. So, we had a good
effect, but a lower residual.
MR. WHITE: I might just add something, or rather
an anomaly. We have found, in some plants, where
they inadvertently lose the ammonia nitrogen, so that
the ammonia nitrogen may drop off to /ero, that we
do not find any free chlorine, and the disinfection
system goes to hell in a hurry. And, that by artificially
feeding ammonia nitrogen, we can bring back the dis-
infection efficiency. The only thing that we think
happens is that there are some organic nitrogen com-
pounds in these one or two particular sewages that
react very rapidly with free chlorine. Now, this is
something that is not generally talked about, because
the broad picture has always been that it takes a much
longer time for free chlorine to react with organic ni-
trogen. And, this is shown up in the tail away of the
germicidal efficiency of chloramine residuals in sew-
ages. So, every once in a while, we get one of these
anomalies that is just absolutely impossible to explain.
DR. LONGLEY:Extending on George White's
remarks just a little bit, it has generally been thought
that in sewage, when we have good mixing, the initial
high rate of disinfection we see is due to the free re-
sidual. And, Dr. Walter, who is here, did some work,
about ten or twelve years ago, at Hopkins, which was
attributed to the free residual. I have noticed in work
by Stinquist and Kaufman, and I think by Mr. White
here, that they claim that the initial high rate of dis-
infection that you get in a wastewater is not due to a
free residual, because they say, it goes so quickly, but
is due to what they call a breaking residual. Now, my
chemist friends tell me that these equations we see
showing the chloramine reaction do not really describe
the pathway very well from a mechanistic standpoint.
and probably may have a number of intermediates
there. And, what the virucidal species is, as the case
may be. is something which we may not even have our
finger on, at this point in time. Now there are some
who, like myself, still believe that probably what we
are seeing is a free residual kill, even though it is very
transitory in the case where you have good mixing. I
do not know if you will subscribe to that or not, George.
MR. WHITE: Well, I give up on that subject, be-
cause 1 am still waiting for Robert Jolley to use that
wonderful mechanism that he has down there, that will
tell us, precisely in milliseconds, how long it takes free
chlorine to combine with the ammonia nitrogen in the
sewage. That will settle that part of the argument.
DR. REYNOLDS:! might throw something else
into the mix. The curves that we had, showing coliform
removal versus total chlorine residual all went through
/ero. and that is because they had a forced /ero inter-
cept. We originally ran those equations without a
forced /ero intercept, and we got negative intercepts
ranging from a -0.1 to a -1.3, which says at zero chlo-
rine residual, we were getting as high as maybe fifty
percent kill, if you look at those correctly. Well, obvi-
ously, at least from a theoretical point of view, that
cannot happen, unless you consider what has just been
said. There is an instantaneous kill before that chlo-
rine is taken up, as chlorine demand, or there is a
species of disinfectant in there that we are not measur-
ing.
MR. JIM HAGEN, EPA, Region Three: Dr.
Reynolds, when you ran your tests, what was the sus-
pended solids levels that you encountered in your
unfiltered effluent?
DR. REYNOLDS: The suspended solids level?
They ranged all the way from a low of 10 mg clear up
to a high of 60. 70, and 80 mg 1.
MR. HAGEN: As a result of some regulations that
were issued last October, the state of Maryland is
issuing permits with levels at ninety mg 1 suspended
solids. Do you see any difficulties with the averages
that high, in terms of adequate disinfection?
DR. REYNOLDS: On lagoon effluent?
MR. HAGEN: Yes, lagoon effluent.
DR. REYNOLDS: 1 do not see any problems of
achieving a given level of disinfection, with suspended
solids that high, in a lagoon effluent, because we are
talking, primarily, of suspended solids being algae
or possibly Daphnia.
MR. HAGEN: What about subsequent release
of soluble BOD? You stated that you did not
find it at your levels.
DR. REYNOLDS: Ourexperience was that as long
as we did not overchlorinate, and especially, get any-
where near breakpoint, we did not observe disruption
of the algal cell wall, or in other words, an increase in
soluble COD in our work. And, as long as we main-
tained fairly low residuals, with reasonable contact
times, we did not see it incur work. Now, there is some
work out, and 1 will hasten to add, by Echclberger. in
particular, which shows that you do have that happen.
We were able to duplicate that, by the way, to a certain
extent, in our laboratory, under laboratory controlled
conditions where we used heaw chlorine concentra-
50
-------
CHLORINA TION/DECHLORINA TION
tions, but we were not able to duplicate it in the field
under normal conditions, unless we went to break-
point.
MR. HAGEN: Well, he had a paper back in 1971,1
thought, that demonstrated that phenomenon. But,
you were unable to duplicate that, or you think it was
just because his high chlorine levels were . . .
DR. REYNOLDS: Yes, if you will look atthechlo-
rine concentrations he had in that paper, you will see
that they are pretty substantial. And, in normal prac-
tice, our feeling is, you do not need to be that high, and
as a result, you do not have that problem.
MR. HAGEN: There were masking effects and you
felt your indicators were sufficient and you did not go
beyond that?
DR. REYNOLDS: Yes, we felt that it was not
masking the soluble COD, if that was what you
were suggesting.
MR. HAGEN: The next question is directed to Mr.
Can. Would you see any problems in dechlorinating
an effluent with high suspended solids levels?
MR. GAN: I cannot answer that, because with the
effluent we worked on, the suspended solids were usu-
ally about 5 or 10 mg/1. None of our experiments had
high suspended solids, so I would not know what the
response from dechlorination would be.
MR. HAGEN: Would anybody have a feeling on
that? See, the problem now is that we have been re-
quested to run power plant work and I just wondered if
it would be necessary to do that.
MR. WHITE: Well, personally, 1 do not think you
are going to have any problems. But, 1 would like to
point out the original work that Collins did in which he
developed the mathematical model, which I find very
helpful in designing chlorination systems providing I
know that initial coliform count. He found no varia-
tion to the model, up to suspended solids as high as, I
think, 120 or 140 mg/1. So, I personally would not be
worried about 90 mg/1 in a well oxidized effluent,
which pond effluent is, or a well oxidized secondary
effluent. When you start talking about primary efflu-
ent, I have backed off of that, and I claim that primary
effluents cannot be disinfected as we think of disinfec-
tion in California.
DR. REYNOLDS: I would probably take a little
more conservative response to that. I have had some
experience, over the last year, with using sulfur dioxide
as a disinfectant, as opposed to dechlorination. And,
we are trying to work out the specific mechanism of the
disinfectant relationship. But, it looks like it could
possibly be an oxidation-reduction reaction, which
would be very similar in the reaction, chemically, of
chlorine in water. If that is the case, and depending on
the relative strength of that oxidation-reduction re-
action, it is conceivable to me that you might disrupt
those algal cell walls. I say, conceivable. I am not sure
that I would really buy it in practice, because 1 think
you would have to look at the oxidation-reduction
potential between the two to really say whether or not
it happened. In gene'ral, the amount of SO2 you add
for dechlorination is substantially less than what you
would add for, say, disinfection. But, I think it bears
some looking at. I would not just throw it out the
window because it does not sound right on the surface.
MR. HAGEN: Well, let me ask you a follow-up,
perhaps. I guess what 1 am fishing for is a recommen-
dation as to how we should proceed with the situation.
Are you saying that, perhaps, given the dechlorination
situation, that we should investigate that further on
site before allowing them to construct?
DR. REYNOLDS: Yes, I think it bears some look-
ing at. That is, to completely dismiss it at this point, I
think is a little iffy. Now, whether you actually have
to run pilot plant work to verify that, I am not certain.
1 think you could look at the oxidation-reduction po-
tentials between the two and probably do it on a lab-
scale basis, as opposed to a pilot-scale basis, to come
up with the answer. My first reaction is the same as
Mr. White's down here; there is probably not a prob-
lem. But, by the same token, I can see where there is a
potential there.
MR. HAGEN: Fine, thank you very much.
MR. WHITE: Before I forget it, one of the inter-
esting characteristics of sulfur dioxide, which is differ-
ent from chlorine, and which simplifies the design, is
that you can put sulfur dioxide ahead of the lip of a
weir and get wonderful mixing right over in the nap of
the weir. You cannot do that with chlorine, because
you will start getting aerating. The chloramines will
aerate out and get some fuming. But, in that way, chlo-
rine dioxide is easier to handle than chlorine.
MR. VENOSA: 1 have a question for Henry Gan.
In the state of California, when you are required to
meet 2.2 total coliform, 23 total coliform, and so forth,
are these samples taken at end of pipe? Based upon
what Henry has presented this morning that ten min-
utes after dechlorination, a one- to two-log increase
in total coliforms occurs, where must the State of
California's compliance standards be met? At end of
pipe or in stream? Before or after dechlorination?
MR. GAN: To meet the disinfection standard, most
of the treatment facilities have to collect the bacteri-
ological sample before dechlorination. And, in most
cases, this is done with the state's approval. Now, for
51
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
Los Angeles, the 2.2 standard is for some part in the
treatment scheme, so the state really does not require
that we sample after dechlorination, but in the treat-
ment scheme. So, we were able to meet that standard,
in spite of the coliform aftergrowth. Does that answer
your question?
MR. WHITE:It also is very critical where you
sample it. You learn these things the hard way. When
you have an effluent that is discharging into a tidal
stream, you do not put the sample point where it can
get the back flow of the tidal stream, because then you
get the aftergrowth, after dechlorination. So, you want
to be in an appropriate place upstream from dechlor-
ination, which does not get that back diffusion of the
dechlorinated effluent.
MR. VENOSA: It seems to me that the purpose of
a coliform standard is to control discharge of micro-
organisms into a stream. And, you have an after-
growth problem based strictly upon where you take
your sample. It seems to me somewhat of a cop-out to
allow a sample to be taken prior to the dechlorina-
tion, just to enable compliance with the standard.
DR. REYNOLDS: Coliforms, as everybody knows,
are only an indicator. Do we know for certain that, just
because coliforms are regrowing, the actual patho-
genic organisms are occurring in the regrowth? And
also, 1 was just going to mention, with some of Henry's
data, especially taken along the canal where it is un-
lincd, some of those coliforms arc probably coming
from leaching out of the soil, as opposed to actual
regrowth in the canal. But, I really do not know. Has
there been any work done to indicate that when coli-
forms regrow the pathogens are regrowing?
DR. LONGLEY: I think work done by Chambers
and a number of others, over the years, has shown that
the pathogens, for the most part, do not regrow. And,
if. in fact, we can consider coliforms a suitable indi-
cator. I know that George has some comments on that
too . . . but. if they are, in fact, a suitable indicator and
we can show that at some point in the system we have
suitable indicator kill, we, therefore, assume that we
have killed the pathogens and they are not regrowing.
If the coliforms regrow downstream this is maybe in-
dicative of other problems, but it is certainly not a
problem of having pathogens in your downstream.
MR. PETER DeSTEFANO, Thomas M. Riddick
and Associates: I think you may have answered the
question already. But, I was a bit confused with this
regrowth problem. If the effluent had a residual of five
parts per million, there probably was not anything still
alive. Are we talking about coliform regrowth or re-
contamination from some sort of infusion from the
plant or aerosol or some other way? Mr. Can?
MR. GAN: In my presentation of data, I said it is
an aftergrowth, which means that there is a possibility
that the coliform is in the chlorinated effluent, but only
to a small amount. But, then after dechlorination,
you are creating an enviroment where you have rapid
aftergrowth of the coliform, in the floe material.
. MR. WHITE:Maybe I could summarize it by read-
ing what 1 wrote about it in my new book.
It is under the heading of The Regrowth Phenom-
enon. It is easy to find because it is in the index. "Num-
erous studies have demonstrated the regrowth phe-
nomenon of coliform and fecal coliform organisms
after disinfection. This may be due to the destruction
of bacterial predators and may depend upon the pres-
ence of certain nutrients in the wastewater or the re-
ceiving waters. It has been observed both in waste-
waters and in receiving waters downstream from a
disinfepted sewage discharge. It is the judgment of
authorities, such as Geldreich, that pathogenic bac-
teria, like Salmonella and Shigella. of the same family
as the coliform group, also regrow. If the disinfection
process has a comparable effectiveness against such
pathogens as with coliforms, a disinfection require-
ment, resulting in reduction of numerous coliform
bacteria, down to a level of 23. or 230 per 100 ml,
would virtually assure the absence of pathogenic bac-
teria in sewage in the smallest numbers, etc." But, this
is the view of the California State Department of
Health, that the pathogens are knocked out.
MR. VENOSA: I would like to make one comment
on that, George. I would have to respectfully disagree
on one thing. You are talking about nutrients and
absence of predators and so forth, and Henry pre-
sented evidence that regrowth occurs within ten min-
utes. There is no known bacterium on earth that is
capable of dividing in ten minutes. It cannot possibly
be "growth". That is my opinion. 1 do not think it can
be growth. I like the term "recontamination" that was
just suggested from the audience.
MR. WHITE:! understand what you are saying.
But, do not forget, when you have been dechlorinating
for a long time, you have an entirely different biomass
downstream from the dechlorination point that you
do upstream. This makes a big difference.
MR. BOB MASTERSON, Tufts University: Not
to add another term to it, but at Tufts we have exper-
ienced maybe what you would call reactivation. And,
in a tertiary effluent we had what we thought was re-
growth, but, on the same sample, we had a die-off of
the unchlorinated one. Therefore, the substrate was
insufficient to support reproduction. But, the injured
52
-------
CHLORINA TION/DECHLORINA TION
organisms, we felt, were coming back and surviving
the actual chlorination. The levels were not nearly as
high as five, but there were around one.
MR. READ WARRINER, CH2M Hill: I would
like to ask about optimization, in connection with
the first two papers this morning. I gather it is
defined as rapid initial mixing, thirty minute con-
tact, in a residual control system. Could you tell
me the basis for each or any of these three and
possibly also, the name of the proprietry equip-
ment that was employed in the pilot plant?
MR. SEPP: The pilot plant has a static mixer... a
tubular mixer. It is simply a T in a three-inch pipe. It
has a Reynolds number of 34,000. And, we have just
installed a 1.5 inch T. We wanted to see whether that
makes any difference. I do not know if it does or not.
And, mixing is accomplished within two seconds,
which is pretty fast. The control system consists of a
residual analyzer and a chlorinator. The residual an-
alyzer has an automatic closed loop control. It is based
on direct residual control. The flow is constant, so
therefore, we do not need any flow proportional con-
trol. The contact tanks are long narrow baffled tanks
with a length-to-width ratio of 130 to 1 and 1 hour
contact. Design was based on what was thought to be
good practice. I understand many plants have the
same kind of a mixer.. . they have a diffuser in a pipe.
But, you have to provide turbulent flow in that pipe,
otherwise it will not do much good. Does this answer
your questions?
DR. LONGLEY: I would like to make a comment.
In a very general sort of way, Reynold's number is, of
course, as far as mixing is concerned for chlorine, an
indicator of whether or not you have a good or poor
mixing. But, from a design standpoint, we have found
in some work we presented at ACS last March, and in
some work we will be presenting in the future, that
Reynold's number, in fact, cannot be correlated to the
amount of inactivation you get, given a certain sewage
and a certain amount of chlorine dose to that sewage.
In truth, the Reynold's number does not describe such
factors as the chemical reaction rate. And, most of
these systems are either reaction rate limited or diffu-
sion limited. I tend to believe that in practically every
system we have, we still are diffusion limited, particu-
larly in the dirty systems. The microorganisms are
going to have a certain amount of organics associated
with them that are going to compete with the chlorine
even if you remove the solids. The problem is trying to
get the chlorine to the active sight on the organism,
whatever that may be, whatever type of mechanism
there is, prior to the conversion of the chlorine to a less
active biocidal species. So, to be able to describe this,
we need some other descriptor than Reynold's number
. . . something which can express time, concentration
gradients, and things such as this. Now, I appreciated
some of the work that was presented on the ponds this
morning, simply because it attempted to get away
from the black box type of design. Hopefully, some
day, when we are talking about wastewater disinfec-
tion, we can describe it in some sort of a formalistic
form that will give us a handle that we can use when
we start to design these systems, other than just black
boxing them.
MR. WHITE: I have long ago abandoned the idea
of trying to use the Reynold's number described by
Karl. What I have been using lately is based on some
work of the County Sanitation Districts of Los Angeles,
with whom Henry Gan is associated, and I also had the
opportunity of working on a couple of jobs with some
consultants. We tried to pick an arbitrary number of
the mean velocity gradient G. LA County used 500. I
do not know whether they picked that number, but
that is the way it came out . . . 500. Obviously, work-
ing down to 2.2 MPN, they got good results in their
work in the plant. I would raise it to 1000. If you
needed 23.2 or 2.2 MPN/100 ml, if ppssible, I would
design my mixer so that the mean velocity gradient is
1000.
DR. JOHNSON: On the other side of the coin con-
cerning the residual control, of course, we know the
detention time, residual concentration, turbulence, all
are important factors in the disinfection process. The
literature is replete with data that show that waste-
water disinfection control through residual is a little
iffy. I wonder whether or not this might be due to a
couple of different things. One is, the residual, of
course, changes quite a bit with time. You put in ten
parts and maybe it comes down to five, in the first
second of rapid demand. The residual concentration
is decreasing all the way through the contact tank.
Usually, we just measure the residual at the end . . .
thirty minutes or so, and there is a lot of change in
concentration. You can measure the residual all the
way through, or you can measure the residual after
two minutes, in order to do closed loop control. But,
most plants are manual control plants, and they con-
trol on the basis of a thirty minute residual. An oper-
ator must crank the control every thirty minutes.
MR. WHITE: That is known as a poor control set
up.
DR. JOHNSON: I am talking about standard
practice today.
MR. WHITE: It is not in California.
53
-------
PROGRESS fN WASTEWATER DISINFECTION TECHNOLOGY
DR. JOHNSON: California is an unusual place.
Anybody that disinfects down to 2.2 MPN/100 ml has
to be unusual, right? So, the residual changes as a
function of time, not just with respect to the concen-
tration, but also with respect to type of residual. I have
seen, in my data, a change from primarily a mono-
chloramine type residual, in the initial part of the chlo-
rination process, to primarily a dichloramine residual,
at the end of the disinfection process, after thirty
minutes. I have forgotten whose data it was, but I
found it interesting that they get any additional dis-
infection from thirty minutes to sixty minutes. Whose
data was that?
MR. VENOSA: Jim Reynolds'.
DR. JOHNSON: Yes. I wonder if that might be
due to the changing character of the residual. 1 won-
der whether or not we should be controlling more
wastewater plants on the basis of the monochloramine
residual which tends to be more inorganic chloramine,
instead of the total residual which has the dichlora-
mine and probably a lot of organics in it. I just throw
that out as a kind of point of discussion to see what you
all think of that.
DR. REYNOLDS: I think that when you start
comparing coliform removal efficiency, as a function
of chlorine residual, after whatever contact time you
speak about, whether it be eighteen min., thirty-five
min., an hour, or whatever the case may be, you have
to look upstream at how the chlorine is injected, the
degree of mixing, and the actual contact time. Whether
you really are getting that kind of contact time that is
presented with the data. 1 think that has a great bear-
ing on this iffyness that you talked about in coliform
removal efficiency as being controlled by chlorine
residual downstream, because all of those factors go
into that residual. I think in ourcasethereare probably
two things which contributed to the fact that there
was no more removal, or no different removal between
thirty-five and fifty minutes. We had very intense mix-
ing and good contact with the disinfectant early in the
process, and most of the organisms were affected by
the thirty minute contact time. We just did not see any
additional kill after thirty-five minutes. We were also
getting down to some very low coliform concentra-
tions by that time.
MR. WHITE: I think the important thing, in work-
ing on contact times, is tj, which Endel Sepp referred
to, and that is the initial appearance of dye. So, what I
have been telling my clients, when they design a con-
tact chamber, is that they have in the contract that
there has to be a dye test made for the contact tank at
three different flows, provided they have a reasonable
amount of wastewater to put through. Usually, in new
plants, you cannot vary the flows too much. So. you
have three flows, and you do this two or three times
and you get some kind of a mean value. You do not
worry about the peak of the curve, where the dye is,
you worry about the tj. That is the first appearance of
the dye, and that is the thing that you plug into Collins'
formula for your kill.
Now, going back to my shaking my head at Don
Johnson while he was telling about the residual. The
only good control system is one that takes the two or
hree minute chlorine contact time for the control of
the chlorination equipment..Then, you can plot the
die-away. You are going to have a residual analyzer at
the end of the system anyway for dechlorination, so
you can use that other analy/er to come up with the
residual concentration time, at the end of the contact
chamber, which is tj, to plug into the Collins'formula.
Collins' formula is close enough to design the capacity
of the equipment. This is the whole point. That is what
we ought to do with contact chambers. We do know
that the germicidal efficiency tails away in thirty min-
utes. We think that is because the monochloramine
and the dichloramine start to hydroly/e and form
organic chloramines which are less germicidal. No-
body has really proved this, but they have come pretty
close to it. I agree with Dr. Reynolds that beyond
thirty or thirty-five minutes you are not improving
coliform kill. But, if you want virus inactivation, Los
Angeles County has proved that at the end of about
fifty or sixty minutes, chloramines catch up with free
chlorine and with o/one. So, there are two different
ball games. You are going to kill viruses and you are
going to kill coliforms.
DR. JOHNSON: What about the difference in
controlling the monochloramine compared to control-
ling the total chlorine?
MR. WHITE: We have enough trouble with con-
trolling total chlorine to get into that one. But, 1 am
working on a couple of things with cells where we do
not go through the classical way of measuring iodine
release in the total residual, which comes closer to
what you are saying. It looks pretty good.
DR. LONGLEY: Just a little more on developing
one point that George made. Some work by Stinquist
and Kaufman at the University of California a number
of years ago. which we more or less duplicated, showed
that, if you had just fair initial mixing, the initial coli-
form kill was only moderate. They were injecting their
chlorine in a pipe with one outlet and they were using
a grid for their good mixing. With very good mixing,
they achieved a high initial kill of coliform, much more
54
-------
CHLORINA TION/DECHLORINA TION
so than they did with just the center point injection.
However, after thirty minutes both were the same. In
the second system a very high initial rate of kill was
achieved and then very little decay thereafter. In the
other system, a more or less uniform kill occurred
throughout the thirty minutes, thereby achieving the
same final product. We have seen this same phenome-
non in a pretty well mixed system and in an extremely
well mixed system. In an extremely well mixed system,
you are just shoving energy into it to get the initial
mixing, really to no added benefit at thirty minutes.
MR. WHITE: But, was your contact chamber com-
pletely plug flow?
DR. LONGLEY: Yes, the contact chamber was
completely plug flow.
MR. WHITE: Okay, that is the difference.
MR. SAULYS:In the midwest, one of the battles
that the biologists have been fighting is trying to insti-
gate total residual control. As far as I am aware, in
most of the counties using free chlorine residual con-
trol, where the state engineers are progressive, they
limit chlorine residua! at the end of the discharge from
the contact chamber to between 0.2 and 0.75 parts per
million. Total residuals could be as high as 10 or more
parts per million.
MR. WHITE: I am confused. Did I hear you say
something about 0.2 mg/1 free chlorine residual?
MR. SAULYS: That is correct.
MR. WHITE: There isn't any 0.2 mg/1 free chlorine
residual in wastewater effluent, unless you are nitri-
fying.
MR. SAULYS:In the NPDES permits issued by
the state of Illinois, many require sewage treatment
plants to discharge a free chlorine residual from the
contact chamber into the river or receiving stream be-
tween 0.2 and 0.75 mg/1 free chlorine residual, after a
half hour of contact time as designed.
MR. WHITE: You would have to breakpoint it and
then some.
MR. SAULYS: That is why we end up with pretty
high totals.
MR. WHITE: Even if you breakpoint it, all your
residual would not be free chlorine. You would have
some dichloramine and maybe some monochloramine
present. That kills the fish, too. What are you going
to do with that?
MR. SAULYS: I agree, but this is the problem that
the biologists are facing out in the midwest, where the
engineers, in the interest of public health, are requir-
ing high chlorine residuals, particularly, free chlorine
residuals, to insure a good disinfection, to insure kill,
with the resultant effect that some of our streams that
are being cleaned up are pretty damn sterile, in terms
of aquatic life.
MR. WHITE: It is absolute folly to try to specify
any kind of a residual and say that this is a disinfected
sewage. The only way to be sure is with coliform test-
ing. Are any of these engineers, from these places that
you speak of, here? They are going down the wrong
track, in my mind.
MR. SAULYS: Maybe next time there is a stan-
dards hearing in Region 5, I will take some efforts to
invite you down and speak to the boards.
MR. WHITE: Okay. I will start a riot, maybe.
MR. VENOSA: I thought I was recently informed
that the State of Illinois had doneaway with theircoli-
form standards altogether. Isn't that true?
MR. SAULYS:It may be another year or so before
they decide to issue a standard. I am not aware that
they have changed it. They have held a number of
hearings. They have tried to do away, at least, with
year around chlorination. They received a mixed re-
ception from USEPA, where some people came out in
support of them, others warned them of public health
menaces. The state is sitting on the whole question
right now, because, potentially, it is a hot political
issue. The issue had come up. for example, in the state
of Wisconsin. It came up before the Department of
Natural Resources Board. One of the members stood
up and said that one of his relatives had contracted
polio in the past and he did not want it on his con-
science that anybody would ever get polio again be-
cause we are not chlorinating year around, and voted
down the proposal to seasonal chlorination. This is the
type of situation, at least, that the states in the mid-
west are faced with right now. Some four or five years
ago there was seasonal chlorination. There was judi-
cious use of chlorine. The federal EPA came in and
imposed year around chlorination through the grants
system . . . through the enforcement system. Now,
when the USEPA is trying to back down, trying to
balance some of their aquatic concerns against public
health concerns, we are again receiving resistance
because everybody is used to chlorinating and they
would rather be safe than sorry.
MR. WHITE: Well. 0.2 mg/"l free chlorine in half
an hour is not going to kill any polio.
MR. CAL SAWYER, Dewberry, Nealon and
Davis in northern Virginia:! have a couple of ques-
tions. The first to Mr. Hansen. When you talk about
lethal concentrations and mixing 53% dilution of your
chlorinated wastewater with your control water, did
you work with your unchlorinated wastewater, same
samples unchlorinated, mixing with 53% and exposing
55
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
the fish to this?
MR. HANSEN: We had a hundred percent of the
unchlorinated water and all test fish have survived.
MR. SAWYER: So, that the fifty-three percent was
the ...
MR. HANSEN: Fifty-three percent is the LC50.
That was that point where half the population was
calculated to have died in the test.
MR. SAWYER: They all survived in a hundred
percent effluent?
MR. HANSEN: Yes
MR. SAWYER: Next question to Mr. Can. You
indicated in your results that dissolved oxygen adjust-
ment and pH adjustment were not necessary on most
of these plants because of the high alkalinity content.
What is the lower level of alkalinity where you would
need pH adjustment, if you were using SO2dechlorin-
ation?
MR. GAN: 1 cannot answer that question because
what we did is just monitor the plant effluent, and all
our treatment plants have an alkalinity of 200 mg/1.
So, I do not know what the response would be down at
lower alkalinity levels.
MR. SAWYER: Once you get about 200 mg/1 of
alkalinity, does your SO2 to chlorine residual ratio go
up above one-to-one?
MR. GAN: Yes. It will go up about two-to-one.
MR. SAWYER: Maybe if you are below a hundred
mg/1 of alkalinity in your effluent, you might be think-
ing of pH adjustment?
MR. GAN: It is a good possibility, yes.
MR. SAWYER: How about with dissolved oxygen?
MR. GAN: There was natural aeration going on
through the treatment systems, and right in our SO2
mixing chamber there was turbulence, mixing the SO2
in the solution. For that reason, we never had any
problem with having to reaerate our water.
MR. SAWYER: If you are using aeration for mix-
ing the contact chamber itself, probably there would
be no need for post aeration?
MR. GAN: Yes. It depends on how much SO2. In
our Pomona plant we do not evqn have mixing. We
jrst add the SO2 right at the out-fall. It is just the
turbulence at the out-fall that mixes the SO2 hydrauli-
cally with the solution.
MR. SAWYER: Do you think in most cases that
you need a rapid mixer to make SO2 effective?
MR. GAN: Definitely, yes.
MR. RON SOLTIS, Washington Suburban San-
itary Commission: I am mainly concerned with
operation of existing wastewater facilities. 1 have
heard a lot today on monitoring and taking chlorine
residuals, and I have seen some of the graphs up here
that have zero chlorine residual as points on the
graphs. At the plants that we have to operate under
current discharge permits, we are running into a prob-
lem with chlorine residuals approaching 0.02 total
chlorine residual in the effluent. There was reference
earlier to Blue Plains which has gone down to 3 parts
per billion free, in the effluent. The question that 1
have (and I am wondering what EPA is doing or not
doing), is where do you all stand on research into how
are we, the people in the field, going to monitor these
wonderful discarge limitations so that we can say, yes,
EPA and water resources administration, we have
noncomplied with the permit, although we did not kill
anything, but we do not know what the heck we did. 1
really have a problem with it, and with operators who
use amperometric titrators. The amperometric titrator
is only accurate down to what the limit is, so essentially
you do not know whether you have ever gone Below
the limit. You are either at noncompliance level or you
do not know where you are. Is any of the research that
EPA is funding going toward operation, rather than
whether we are going to kill anything or not? How are
we going to find out what we are doing?
MR. BEN BUONOMO, Delta Scientific Enviro-
tech: I was pleased to hear that question because no
matter what we do in a laboratory, we are using port-
able instruments. We are measuring discrete samples.
In the real world, all of our fine measurements and the
theory and the models that we develop from these
portable instrumentation, measuring discrete samples
in a clean environment, does not hold true when you
are on continuous on-line monitoring, twenty-four
hours a day, seven days a week. We recogni/e this
problem of measuring very, very low chlorine residuals.
With utilities discharging into the river waters, in the
regions we have worked. 0.25 mg/1 is the maximum.
When they are discharging, as with condensor cooling
water into an ocean, it is 0.5 parts per million maxi-
mum. EPA now, in Cincinnati, under Tony Mentink's
quality assurance laboratory, is evaluating a polaro-
graphic membrane technique for monitoring chlorine
residuals. We have one now working in a laboratory,
0.25 parts per million full-scale. Currently, we are
working on one that was the zero point off the end
of the scale, not on it directly, so we can even measure
negative when we are doing sulfur dioxide, to see how
much have we dechlorinated. There is some work
being done by the EPA with us.
MR. VENOSA: Excuse me, how well does your
56
-------
CHLORINA TION/DECHLORINA TION
membrane electrode work in the wastewater environ-
ment?
MR. BUONOMO: It is an iodine release from
iodide, so we measure total release.
MR. VENOSA: Yes, but 1 mean as far as gumming
up with solids and so forth?
MR. BUONOMO: Thank you for that question.
Instead of exposing naked electrodes as convention-
ally has been done for the last ten or fifteen years in
an amperometric cell where your electrodes are ex-
posed completely to the effluent, we isolate the elec-
trodes from the process liquid by a membrane. It is
constantly enclosed with a clean electrolyte and the
membrane is blocking the effluent. Through the mem-
brane, the specific chlorine species are coming through
that we monitor.
MR. SOLTIS: 1 have one other question, again,
concerning operation. We have several . . . actually
two plants now, that have upflow carbon columns that
are supposed to have been designed fordechlorination
They were designed with a free residual of three going
into the carbon column, and the effluent coming out
was supposed to be 0.05 or less total chlorine. These
were supposed to last for the life of these plants which
were interim. We found out that after approximate!)
three to four months, in order to stay within the per-
mit limitation, we could no longer put in three mg/ 1
free chlorine. We had to go down to between 0.5 and
1.0 free chlorine in order to stay at the 0.05 total resid
ual out in the effluent. I previously asked ifCaliforni;
had any systems that were actually on-line, and tht
answer, I believe, was no, at that time. I ask the pane
or anyone in the audience, if there are any full-scale
facilities that are actually using either up-flow carbon
columns or any other type of carbon column, as a
means of dechlorination?
MR. WHITE: I believe there is one that is either in
the process of design or construction, and that is at
Vallejo, California. But, it sounds to me like you have
a whole lot of problems. The easiest way to do dechlo-
rination with SC>2 is either by the biased residual that
Mr. Gan had a little schematic on, or to just put in a
slight excess of SO2 on the feed forward basis, and
intermittently sample the dechlorinated sample on
your control analyzer, through a timer, five minutes
out of the hour. Those two work very well.
MR. SOLTIS: What we have gone to is an SOa
system as a backup. Because of the investment in the
carbon and the facilities for storing the carbon, or
actually running the flow through them, we utilize the
carbon to remove the maximum amount of chlorine
possible and then we trim it with sulfur dioxide. What
I am looking for is help in trying to find out (which we
have not been able to get an answer for yet) what has
happened that the carbon has dechlorinated in both
cases, satisfactorily for three to four months, and then
over almost a one to two week period, it was almost a
straight line drop from 3 mg 1 free down to 1.0 or 0.5
mg/1.
MR. WHITE: Well, is this regenerated carbon or
is this just . . .
MR. SOLTIS: No, this was designed to last the
life of the plants, which was . . .
MR. WHITE: Eighteen months?
MR. SOLTIS: No, it was two years. And, in both
cases . . . one is a 200,000 gal day flow and the other
one is 2 million. And. in both cases, it just did not last
longer than the three . . .
MR. WHITE: All right, if you have a backup de-
chlorination for SO2, then the easiest thing to do is to
come down to a half a milligram per liter total, which
is easy to do with conventional chlorination equip-
ment, and use that as a feedback for your control, and
then dechlorinate the other half. But, do you do this?
Is this what you are doing, and you mean that the
carbon is petering out?
MR. SOLTIS: It has petered out. We have a /inger
in here, in that we have to maintain'3 mg 1 free chlo-
rine residual prior to dechlorination. At the end of one
hour contact time we have to have a 3 mg; 1 free chlo-
rine residual for virus inactivation.
MR. WHITE: 3 mg/1 free chlorine''
MR. SOLTIS: It is one of the requirements, not of
EPA or the state of Maryland, but it is a resolution of
the county that we have.
MR. WHITE: I see.
MR. SOLTIS: They are trying to put a virus limita-
tion on us, and one of the requirements, like I say, is a
3 mg/1 free chlorine residual, at the end of one hour
contact time, prior to dechlorination.
MR. WHITE: What kind of coliform counts are
you getting before dechlorination, by the carbon?
MR. SOLTIS: In most cases, the coliform is less
than three.
MR. WHITE: What do you do to take care of the
carbon and keep it from fouling?
MR. SOLTIS: The plants are all designed with
either a chemical addition, for tertiary treatment, and
filtration prior to chlorination ... we have to go
through filters, so the effluents, in most cases, have
turbidities, for instance, in the 0.6 range and sus-
pended solids less than 5.
MR. WHITE: Yes, but don't you get solids fouling
in the carbon?
57
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
MR. SOLTIS: So far, in these two particular facil-
ities, we have not noticed any.
DR. ARCHIBALD HILL, Louisiana Tech Uni-
versity: 1 have a question for Henry Can. I was inter-
ested in your recontamination after dechlorination.
You had ten minute and two hour samples. I was won-
dering if these were estimated from points on your
contact chamber, and if so, could part of the recon-
tamination have been due to back mixing effects?
MR. GAN: The time is based on points in the con-
tact chamber. The recontamination, we are sure, is
not due to back mixing, as I showed in one of the
figures. Initially, we do not get any recontamination,
or did not get any recontamination right after cleaning
of the tank. The recontamination occurred after a few
days operation. Another thing I want to add is, we are
certain there is recontamination because we have col-
lected grab samples in dechlorination experiments in
the lab, and we found no coliform increase several
hours later. But, we did find it in the dechlorination
chamber.
DR. GILBERT GORDON, Miami University:
Later today and tomorrow, we will hear about chlo-
rine dioxide and ozone, primarily, because of some
concern over removing haloforms in general. Are any
studies being carried out on the chlorination test to
see what kind of reduction or what kind of chloroform
and any haloform residuals are formed as a function
of the various ways in which mixing and chlorination
are carried out?
MR. WHITE: It is well documented that the chlo-
ramines contribute very little to the THM's. They
contribute other organics. The biggest concern is in
potable water because that is where the precursors are.
DR. GORDON: But, if you remove the precursors.
perhaps, at one of the . . .
MR. WHITE: Yes, but now you are getting into
water treatment.
MR. MIKE WITT, Wisconsin Department of
Natural Resources: 1 would like to ask Mr. Venosa.
what, if any, research EPA has going, as far as the need
for disinfection? The program here is geared a lot
toward different types of disinfection. As the gentle-
men from Region 5 said before, one of the questions
we have in the midwest is whether or not we need it at
all in the winter time, since nobody swims in ice. I
would like to know what your opinion is of that, and
if there is anything that we can get ahold of that might
give us more information?
MR. VENOSA: We have a project that we have
planned for next year to try to determine, or assess,
the significance of antibiotic resistant coliforms in
sewage effluents because the increase in the number of
antibiotic resistant coliforms could someday, possibly.
result in an increase in the number of antibiotic re-
sistant pathogens as well. Maybe that would be one
argument in favor of disinfection at the wastewater
treatment plant, to control the proliferation of such
types of organisms. So, we have a project to address
that type of problem. The Health Effects Research
Laboratory is conducting some research to establish
correlations between coliform numbers and epidemio-
logical transmission of disease in bathing beaches in
the salt water areas of New York. 1 am hoping that they
do that in fresh water, too.
MR. WITT: In our work, about a year ago, we tried
to make that administrative change in Wisconsin, and
go to seasonal disinfection. It was clear at that time
that no one could make any correlation either way. We
tried to show one side, and other people tried to show
the other side. Some of the research, at the University
of Wisconsin, has dealt with the probabilities of differ-
ent disease transmission routes, showing that waste-
water disinfection, whether you have it or not. makes
little difference. People get diseases from other sources.
We also have a person with the Department of Medi-
cine, at the University, doing some research on disease
transmission in children swimming in bathing beaches.
swimming in pools, or not swimming at all. The chil-
dren that swim in the pools get sick the most. The
thought is that, perhaps, it is because there are so many
so close together, and not because they jump in the
water.
DR. LONGLEY: 1 think the problem though is
when you try to do any of this type of epidemiology,
the numbers that you are incurring are so low that they
are lost in the total numbers, and to try to attribute
any signifieanee to them becomes a burden, which we
really have not found out how to deal with yet. The
same problem, of course, holds with cancer caused by
THM's. The other point is, if you are going to accept
that we might have, in fact, a virus problem. I think
you have to reali/e that viruses can survive for some
long periods of time in the environment. If we put them
out in the winter time, they might still be there next
summer.
DR. REYNOLDS: Another approach to that is
looking into actual coliform die-aways in various
treatment trains and systems. As I indicated earlier,
a well designed lagoon system, with sufficient contact
time, will achieve verv low coliform numbers in the
58
-------
CHLORINA TION/DECHLORINA TION
effluent, with no disinfection. EPA has funded (and
those reports, I think, are available) about nine sys-
tems nationwide, where the quality of the lagoon efflu-
ent was monitored. You can check those systems to
determine the need of chlorinating lagoon effluents,
if they were properly designed. A lagoon effluent is not
an activated sludge or a trickling filter effluent.
MS. BARBARA BROOMFIELD, Argonne Na-
tional Lab: Now, aside from the already stated health
effects of chlorinated and nonchlorinated effluents, 1
was wondering if anyone bothered to assess the effects
of nonchlorinated or nondisinfected wastewaters on
test organisms or other aquatic organisms? Mr.
Hansen mentioned that he observed no death rate
from organisms exposed to nondisinfected waste-
waters, but did he bother to examine them to notice
if there were other physiological changes that might
have taken place in these organisms?
MR. HANSEN: No, we did not examine them any
further. What we are looking for, and I thought I made
it clear, is for acute toxicity. That is all we are looking
for in this particular study.
MS. BROOMFIELD: Would there be any studies
being done to determine if there were any long-term
effects from discharging large amounts of nondisin-
fected wastewaters into the streams and rivers?
MR. WHITE1 Yes, there was a study, a very com-
prehensive study in San Francisco Bay. To answer
your question, very briefly, I think the man's name was
Brock, but it was done under the Sanitary Engineering
Research Laboratory, University of California, about
1969-70. They found out that chlorinated and decho-
rinated effluent were less toxic than an unchlorinated
effluent. This was on the biomass . . . attached organ-
isms, if you will, that were growing. Before 1 forget.
I want to tell you about this dechlorination problem,
and the contact time. I have seem somewhere in print,
that the EPA or some other regulatory agency, did not
consider an out-fall as a contact chamber. On the con-
trary, a long out-fall is absolutely the best type of
contact chamber that you can make. If you are smart
enough when you are designing it, you can wind the
out-fall around the plant long enough to get your thirty
minutes, and dechlorinate to 0.5 mg/1 and then the rest
of the out-fall will accumulate enough biomass in there
to destroy the remaining residual. This has been done
in a couple of places and it is very effective. Of course.
Fish and Game might not like that idea, but it works.
MR. VENOSA: I would like to provide some more
information for Miss Broomfield. We funded a study
with Wyoming, Michigan, (using two study sites.
Grandville and Wyoming) where we evaluated the
effects of chlorinated, dechlorinated. brominated, and
o/onated effluents on fish, both acute and chronic
toxicities. If I remember correctly, at least in the
Grandville portion of the study, the dechlorinated.
o/onated and unchlorinated effluents were nontoxic.
The o/onated effluent was not only nontoxic. but
actually stimulatory, so that in the chronic exposures
the fish actually grew bigger. They laid more eggs and
the hatchabilities were higher than the controls. Pre-
sumably, this was due to the higher dissolved oxygen
content. At Wyoming, Michigan, we repeated the ex-
periments, but the unchlorinated effluent (control)
was toxic because half the raw wastewater was of in-
dustrial origin, containing cyanides and other toxic
compounds, so we really could not tell too much from
that one.
MR. GREG SEEGERT, WAPORA, Inc.: Before
asking my question, I would also like to respond.
briefly, to the lady's inquiry. The problem that you
typically get into is, there is no such thing as a typical
effluent. There are many effluents which arc just plain
bad news, whether they are chlorinated or not. So, as
Al was just mentioning, in Wyoming there was a high
industrial input with lots of heavy metal contamina-
tion. The effluent was highly toxic whether it was
chlorinated or not. In many cases, if you are talking
about relatively high quality water systems, there is
not much of a problem. It depends almost entirely on
the overall quality of the effluent. Chlorine only com-
prises a portion of the potential toxicity. My question
is for Mr. Hansen. You had characteri/ed all your
LC50 values, in terms of percent effluent. I was won-
dering if you had calculated what the total residual
chlorine concentrations were for those same LC50
values?
MR. HANSEN: Yes, I have. As I recall, the higher
one was 0.60 total residual chlorine. The lower one
was, I believe, 0.13, at San Pablo.
MR. SEEGERT: The higher value was for the
golden shiner? You said that the golden shiner was
considerably more resistant?
MR. HANSEN: No, this was on fathead minnows.
At San Pablo, the total ammonia was less than one,
and the 96-hour LC50 was 0.60 TRC. Effluents from
all the other plants had ammonia concentrations
greater than 10 and mean 96-hour LC50 of 0.13 TRC.
MR. SEEGERT: And that was for the fathead
minnow?
MR. HANSEN: That was for the fathead minnows,
ves.
59
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
MR. SEEGERT: What was it for the golden shiner, 96-hour LC50 values were about 40 parts per billion.
it you have that information? and the fathead minnow was approximately two to
MR. HANSEN: 1 do not recall if I have that. three times more resistant, which would place the
MR. SEEGERT: As I recall, you did make the overall toxicity almost an order of magnitude lower
point that the fathead minnow was considerably more than your LC50 values for 96 hours, and also would
sensitive. reverse the relative sensitivity of the two species.
MR. HANSEN: Yes. that is correct. Would you respond to that, please?
MR. SEEGERT: Al probably also has the infor- MR. HANSEN: Well, in the ones that we calcu-
mation, but as I recall, in the study that was done at lated, we found that they did agree very closely with
Grandville and Wyoming, I think, the golden shiner what we found in the literature.
was the most sensitive species that they tested. The
60
-------
SECTION 3. CHLORINE DIOXIDE
9.
EFFECTIVENESS OF CHLORINE DIOXIDE AS A WASTEWATER DISINFECTANT
James D. Berg,* E. Marco Aieta* and Paul V. Roberts,* and Robert C. Cooper**
*Civil Engineering Department
Stanford University
Stanford, California
**Sanitary Engineering Research Laboratory
University of California
Berkeley, California
ABSTRACT
Chlorine dioxide (C/O2) /.v evaluated as an alternative to chlorine (CI2)
for use in the disinfection of municipal wastewater. Bench-scale experi-
ments in three topic areas are discussed. First, two methods for counting
bacteria, Membrane Filter and Most Probable Number, are compared.
Second, factors affecting the relatively inadequate performance of CIO2,
based on kill of coliform bacteria, are investigated. Third, the two dis-
infectants are compared according to inactivation of viruses in secondary
effluent. The information from these experiments will ultimately be used
to develop a concept for a continuous reactor system that will be tested
on a pilot scale.
INTRODUCTION
Alternatives to chlorine for the disinfection of
wastewater are being sought for several reasons, such
as developing a process that is more efficient and that
yields an effluent that is environmentally safe. A vi-
able candidate to replace chlorine should meet the
following set of criteria. The disinfectant should:
1) be cost effective;
2) provide a measurable residual;
3) be easy and safe to handle; and,
4) produce a minimum of undesirable by-products.
Chlorine dioxide (C1O2) is being investigated as an
alternative under EPA Research Grant R-805426,
"Feasibility of Using Chlorine Dioxide in the Dis-
infection of Municipal Wastewater." Chlorine dioxide
has been used successfully in Europe to disinfect water
supplies and is used extensively as a bleaching agent
in the pulp and paper industry. With respect to the
disinfection of wastewater, chlorine dioxide has sev-
eral exceptional characteristics; it
1) is a strong oxidant over a broad pH range;
2) provides a measurable residual;
3) does not react with ammonia to form less effec-
tive chloramines; and,
4) does not react to yield trihalomethanes.
The principal objectives of the feasibility study are to:
1) use these characteristics to develop a concept for a
continuous reactor system that will be tested on a pilot
scale and used to prepare a preliminary design for full-
scale treatment facilities and 2) compare chlorine
61
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
dioxide and chlorine as wastewater disinfectants on
the basis of effectiveness and cost.
This paper will address three topics that describe
the behavior of chlorine dioxide in secondary effluent
and begin to define those characteristics that will lead
to a process design. The three topics, from the first
nine months of the study, are:
I. Comparison of MPN and MF Methods of
Counting Bacteria.
II. Factors Influencing the Bactericidal Efficiency
of Chlorine Dioxide.
III. Virucidal Effectiveness of Chlorine Dioxide.
GENERAL METHODS
MPN and MF tests were done according to Stan-
dard Methods (1975)(5). MPN media used were lac-
tose broth as a presumptive medium and brilliant
green bile broth and EC medium for confirmation.
Gelman GN-6 filters were used for the MF tests and
M-Endo Agar for total coliforms using the single-
step direct technique and M-FC medium for fecal
coliforms.
Virus samples were collected, placed on ice, and
transported to Dr. Robert Cooper, Sanitary Engineering
Research laboratory, University of California at Berkeley,
for analyses. Transit time was less than 2 hours and
the analyses were begun immediately upon receipt of
samples. The animal virus used in the study was Polio-
virus I LCS strain that had been grown in Buffalo Green
Monkey Kidney cells (BGM). The viruses were har-
vested from the cells by freezing, thawing, and centrifu-
gation to a liter of approximately 107 plaque-forming
units per milliliter (PFU/ml). The virus was recovered
after experimentation using a standard plaque assay.
A cofiphage that had been isolated from Palo Alto secon-
dary effluent (the wastewater used throughout the study)
was grown on host Escherichia coli B to a titer of ap-
proximately 109 MPN/ml and used in the experiments.
The phage was recovered and assayed using a modifi-
cation of the MPN described by Kott (1966)(2).
Secondary effluent from the Palo Alto Water
Pollution Control Plant was used in our ex-
periments. Samples were collected at about 8:00
A.M. each experiment day to minimize problems
of variability due to hydraulic surges and to
minimize NH3-N concentrations. The effuent was
characterized by measuring the following
parameters:
1) total COD
2) alkalinity (by an amperometric titration)
3) pH
4) temperature
5) total filterable residue (dried at 103°C)
6) ammonia as nitrogen
All parameters were measured singly except for the
total filterable residue (total suspended solids) which
was done in duplicate. All tests except for ammonia
nitrogen were done according to methods prescribed
in Standard Methods. Ammonia nitrogen was mea-
sured using an ammonia electrode. Previously, results
using the ammonia electrode and those of nessleriza-
tion (a procedure which is prescribed in Standard
Methods) had been compared. A t-test based on 13
paired values in the range of 0.5 to 30 mg NH3-N per
liter showed no significant difference between the
ammonia electrode and the determination by distil-
lation and nesslerization.
The oxidant chemistry consisted primarily of chlo-
rine dioxide and chlorine generation and standardi/a-
tion, and residuals measurement. Chlorine dioxide
was generated using the sodium chlorite acid activa-
tion technique described in Standard Methods. A
suggested chemical reaction for this procedure is:
4H* + 4C1O2 -*-CP + 2C1O2 + ClOg + 2H+ + H2O.
The schematic drawing of the reactor is shown in
Figure I. Chlorine solutions were made by bubbling
chlorine gas through distilled water. Both chlorine
dioxide and chlorine solutions were standardi/ed at
CHLORINE DIOXIDE GENERATOR
VENT
TO
HOOD
REACTION
VESSEL
NaCl02
SALT
TOWER
CI02 STOCK SOLN
Figure 1. Chlorine dioxide generator using acid acti-
vation of a sodium chlorite solution to prepare
an approximately 500 mg/l CIO. stock solu-
tion.
the beginning of every experiment using the iodo-
metric method prescribed in Standard Methods. So-
dium thiosulfatc was placed in the sampling flasks to
reduce chlorine dioxide or chlorine residuals immedi-
ately upon sampling. Chlorine dioxide and chlorine
62
-------
CHLORINE DIOXIDE
residuals were measured by the iodometrie back-
titration using an amperometrie titralor (Fisher-
Porter).
RESULTS AND DISCUSSION
I. Comparison of MPN and MF Methods of
Counting Bacteria
This series of experiments had two objectives:
1) a preliminary comparison of the disinfection
effectiveness of chlorine dioxide and chlorine ac-
cording to the kill of natural populations of
coliform organisms in wastewater and 2) a com-
parison of the Most Probable Number Test
.(MPN) and the Membrane Filter Test (MF) for
counting coliform bacteria in wastewater. The
disinfectant comparison will be addressed in
greater detail in another paper.
Each disinfectant experiment had bacteria coun-
ted by both the MPN and MF methods so that
both goals were achieved. Duplicate samples were
taken for both counting methods. All samples in
a given experiment for the MPN test were ran-
domized to eliminate any bias associated with or-
der or time of analysis. The same was done for
the MF test.
The comparisons of disinfectants and counting
methods were conducted at doses of 5 mg/1
chlorine dioxide as mg/1 C1O2 and 5 mg/1 C12 as
mg/1 Cl, and a contact time of 5 minutes. This
dose and contact time were chosen to achieve
substantial bacterial kills, yet avoid overkill to the
extent of no measurable bacterial counts.
The results of these experiments are presented
in Table 1. Each "experiment number" indicates
a different grab sample of secondary effluent'
from the Palo Alto Water Pollution Control
Plant. The associated values in each cell (Table
1A) are log reduction values of the disinfected
sample after a five-minute contact time. The
coliform values in Table IB are logarithms of the
replicate bacterial counts in the grab sample
before disinfection along with chemical charac-
TABLE 1A: SUMMARY OF LOG REDUCTION A (LOG N)
Experiment
No.
6
7
10
16
18
20
1
mg/l
cio2
MPN
TC
0.10
0.76
0.17
-0.02
-0.17
0.39
FC
-0.22
-0.66
0.07
0.07
0.45
0.08
MF
TC
-0.17
-0.29
-0.01
0.31
0.15
-0.04
FC
0.24
0.21
0.16
0.19
0.08
0.14
Experiment
No.
8
9
11
12
14
15
5 mg/l, 5 min contact time
cio2
MPN
TC
2.13
3.23
2.86
1.89
2.51
2.52
FC
1.91
3.71
3.31
2.34
2.05
2.27
MF
TC FC
2.79 2.36
4.39 1.55
2.18 1.94
3.60 2.52
2.48 2.10
2.39 2.10
CI2
MPN
TC FC
2.10 2.93
2.88 3.10
3.17 3.29
1.86 2.09
2.67 1.94
3.32 3.13
MF
TC
3.78
4.76
2.63
3.60
3.05
2.78
FC
3.00
2.07
2.86
2.28
2.44
2.51
Log reduction = log average of number of organisms (N/100 ml) in undisinfected sample minus log average number of organisms in disinfected sample.
TABLE 1B: SECONDARY EFFLUENT CHARACTERISTICS
Parameter
Temp., °C
PH
Alkalinity, mg/l as CaCO,
NH3-N, mg/l
Total COD, mg/l
Total Suspended
Solids, mg/l
Total Coliforms
MPN, LOG N/100 ml
MF, LOG N/100 ml
Fecal Coliforms
MPN, LOG N/100 ml
MF, LOG N/100 ml
Exp.
6
22.6
7.15
221.
42.
45.6
15.5
5.64
5.93
4.97
4.83
Exp.
7
21.2
6.9
210.
36.
34.3
14.
4.81
5.10
3.95
4.92
Exp.
8
22.5
7.1
249.
30.5
35.9
8.5
4.73
4.95
4.47
4.04
Exp.
9
21.2
7.3
249.
42.5
37.5
7.5
4.73
6.27
5.27
3.24
Exp.
10
21.8
7.5
232.
40.
41.3
3.0
3.70
3.81
3.43
3.49
Exp.
11
21.5
7.5
260.
21.6
49.0
24.7
4.48
3.88
4.23
3.46
Exp.
12
21.8
7.45
240.
18.0
52.0
9.5
3.69
5.51
3.64
4.35
Exp.
14
21.5
7.55
296.0
19.6
46.2
16.0
4.97
4.70
4.24
4.26
Exp.
15
22.0
7.35
285.
19.5
55.3
7.5
4.62
4.87
4.43
4.59
Exp.
16
22.8
7.8
290.
21.5
50.0
5.5
4.05
4.69
4.33
4.89
Exp.
18
23.4
7.5
268.
22.
39.1
10.0
3.95
4.58
4.29
4.65
Exp.
20
22.3
7.7
363.
24.5
39.0
21.5
5.57
5.62
5.24
5.66
63
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
teristics of the secondary effluent used for each
experiment. The difference between averages of
influent and disinfected sample is the reduction,
A (log N),caused by each disinfectant. The data
from these two tables were then subjected to the
statistical analysis which follows.
STATISTICAL INTERPRETATION
Two statistical computer packages were used in
analysis of the data. The first of these is the Biomedical
Computer Programs Package, P-series (BMDP).
From this package the BMDP-2V analysis of variance
program was used to generate the analysis of variance
(ANOVA) tables that follow. The second package
used was Minitab II. This package provided the results
of the paired t-tests and the correlation tests that also
follow.
TABLE 2. CONDENSED RESULTS OF ANOVA TC
LOG-KILL DEPENDENT VARIABLE
Source of Varialiont
Experiment
Method
Experiment-Method
Interaction
F-Value
12.37131*
33.87109*
15.16024*
Prob. F Exceeded
0.008
0.002
0.005
tDisinfectant main effect and disinfectant-experiment interaction omitted.
"Significant at 1% level.
To examine the differences between the method of
counting organisms on log-kill, an ANOVA was run
with Total Coliform (TC) log-kill as the dependent
variable, and method of counting and experiment
(since experiments were performed on different waste-
waters and on different days) as grouping variables.
This same ANOVA was also run with Fecal Coliform
(FC) log-kill as the dependent variable. The results
are shown on Tables 2 and 3.
TABLE 3. CONDENSED RESULTS OF ANOVA FC
LOG-KILL DEPENDENT VARIABLE
Source of Variationf
Experiment
Method
Experiment-Method
Interaction
F-Value
2.21334
7.04393*
5.2105 *
Prob. F Exceeded
0.202
0.045
0.047
tDisinfectant main effect and disinfectant-experiment interaction omitted.
'Significant at 5% level
It is apparent from Tables 2 and 3 that there is a sig-
nificant experiment-method interaction which com-
plicated interpretation of the ANOVA results. This
implies that the relationship between log-kill and
method of bacterial enumeration is not consistent
from experiment to experiment. One suggested hy-
pothesis to explain this inconsistency concerns the
effect on the experiments by a variation in Palo Alto
wastewater. Correlations of the parameters temper-
ature, pH, alkalinity, ammonia concentration, total
COD, and total suspended solids with TC and FC log-
kills, however, showed no consistent relationships. It
is possible, but unlikely, that an unmeasured param-
eter or trace industrial constituent may be responsible
for the experiment-method interaction.
To exclude the effect of experiment in the compari-
son of counting methods, a pooled variance paired
t-test was run. For this test, a mean of the log number
of organisms from each cell was calculated. The differ-
ences between method for each experiment were calcu-
lated and the mean of these values was subjected to the
paired t-test. This comparison was performed over all
combinations of parameters. In addition, the log-kill
values for fecal and total coliforms were tested. The
results, tabulated in Table 4 indicate that, indeed,
there is no difference between the MPN and MF
Methods of counting organisms in determining the
quantity of organisms present.
TABLE 4. COMPARISON OF MPN AND MF METHODS
USING A POOLED VARIANCE t-TEST
Analysis
TC Influent
FC Influent
TC Effluent
CIO;,
CI2
FC Effluent
CIO2
CI2
TC Log-Kill
CI02
CI2
FC Log-Kill
CIO2
ci-2
t-Value
-1.408
1.002
-0.322
0.662
-0.744
0.541
-1.231
-1 .636
1.199
0.944
Probability t Exceeded
0.2182
0.3622
0.7603
0.5373
0.4904
0.6117
0.2731
0.1627
0.2843
0.3887
Standard Deviation
of Difference
1.010
0.912
0.393
0.654
0.750
0.641
0.892
1.150
1.030
0.571
There is, however, an important difference between
these two methods. To see this, the standard deviation
of each cell is calculated and these standard deviations
are weighted and summed over all experiments for
both MPN and MF. A comparison of the standard
deviation (i.e., precision) between the two methods
can be made. 1 he results of such an analysis appear in
Table 5. In every case the MPN method shows greater
64
-------
CHLORINE DIOXIDE
variation than the MF method. This result seems rea-
sonable when one considers that the MPN method
itself is based on a statistical probability and not on a
direct organism count as is the MF method. The MPN
value is the modal value of a 95% confidence interval.
Another point to be made from Table 5 is that the
magnitude of the difference between two samples as
measured by MPN must be greater to be statistically
significant than the magnitude as measured by MF
[(e.g., (Influent-log) - (Effluent-log) = log-kill)].
TABLE 5. COMPARISON OF METHOD PRECISION
Parameter
TC-lnfluent
FC-lnfluent
TC-Effluent
CIO2
TC-Effluent
ci2
FC-Effluent
CIO2 -
FC-Effluent
CI2
Method
MPN
MF
MPN
MF
MPN
MF
MPN
MF
MPN
MF
MPN
MF
5
.61
.11
.69
.14
.38
.23
.42
.21
.46
.13
.58
.13
s2
.37
.01
.48
.02
.14
.05
.18
.04
.21
.02
.34
.02
The mechanisms of disinfection are not completely
understood, but most information on the subject indi-
cates that powerful chemical oxidants such as C1O2
and C^ act destructively upon cell protein, either by
inactivation of critical enzyme systems or disruption
of protein synthesis directly. Benarde et al (1967)(1)
have shown that cell wall lysis is not the cause of cell
destruction. The disinfection mechanism must then in-
volve at least two steps: 1) chemisorption of the dis-
infectant at selective active centers on the cell surface,
and 2) surface and intrasurface diffusion of the active
chemical complex across the cell wall with resultant
chemical attack on cellular components. It should be
possible, then, for an organism to become "stressed"
without being destroyed. If an organism in this stressed
state were placed into an environment in which the
optimum conditions for growth were present, its
chance of survival would be greater than if left in a less
favorable environment. This is indeed the case for
MPN vs MF methods of organism counting. It has
been shown by Mowat (1976) (4) that the M PN method
is in fact a more favorable environment for bacterial
growth. Therefore, the survival rate of the stressed
organisms would be greater in the MPN method than
in the M F method, giving results as seen in Table 6. In
an actual disinfection situation, the surviving and/or
stressed organisms are not subjected to an optimal
growth environment, suggesting that the MF method
is a more realistic indicator of the number of surviving
organisms than is the MPN method.
TABLE 6. COMPARISON OF TWO DISINFECTANTS
BY MPN AND MF USING A POOLED VARIANCE t-TEST
Analysis
TC Log-Kill
MPN
MF
FC Log-Kill
MPN
MF
l-Valuet
-0.899
-3.514
-0.588
-2.731
Probability I Exceeded
0.4099
0.0170*
0.5819
0.0412*
Standard Deviation
0.391
0.322
0.645
0.387
"Significant at the 5% level.
tt-Value of t < 0 indicates that log kill with CIO2 < log-kill with
SUMMARY
I. The MPN and MF methods of counting total
and fecal coliform organisms are equivalent in
accuracy of measurement.
2. The MF method is a more precise method of
measurement than the MPN method.
3. The MF method is a more appropriate estimator
of realistic situations.
II. Factors Influencing the Bactericidal Efficiency
of Chlorine Dioxide
Chlorine dioxide had been observed to have an
unexpectedly low efficiency when compared with
chlorine in disinfecting Palo Alto secondary ef-
fluent (Table 1A). Two experiments were designed
to investigate factors that may influence the per-
formance of chlorine dioxide as a disinfectant: 1)
suspended solids and disolved constituents present
in the wastewater as well as 2) the nature of the
indicator organism.
EFFECT OF SUSPENDED SOLIDS
Experiment "A" was designed to study the effect of
suspended solids on disinfection. Two parallel reactors
were dosed with 5 mg/1 chlorine dioxide. One reactor
contained secondary effluent with 12 mg/1 suspended
solids. The second reactor contained no measurable
suspended solids. The reactor containing the sus-
pended solids showed less bacterial kills than the
65
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
reactor containing settled secondary effluent (Table 7).
These results indicate that the suspended solids offered
some physical protection to the bacteria, and possibly
exerted an additional chlorine dioxide demand. The
difference between these two systems, however, is not
dramatic.
TABLE 7. EXPERIMENT A: EFFECT OF SUSPENDED
SOLIDS ON DISINFECTION WITH 5 mg/l CIO2,
5-MIN CONTACT TIME
Secondary
Effluent
Settled
Secondary
"Affluent
NINF
NEFF
AN
^INF
^EFF
AN
MPN
TC
4.45
1.30
3.15
4.88
1.54
3.34
FC
4.45
1.30
3.15
4.88
1.45
3.43
MF
TC
4.65
2.21
2.44
4.67
1.28
3.39
FC
4.88
2.30
2.58
4.79
1.09
3.70
Secondary
Effluent
Settled
Secondary
Effluent
Chemical Characteristics
Temp. Alk.
°C pH mg/l
22.3 7.3 290.4
22.3 7.4 290.4
NH3-N
mg/l
18.5
18.5
CODf
mg/l
46.4
33.0
TSS
mg/l
12.0
ND*
"Not detectable
EFFECTIVENESS AS A BACTERICIDE
Experiment "B" was designed with two purposes in
mind: I) to provide a link with the literature that
reports superior bacterial kills with C1O2, and 2) to
indicate whether the dissolved constituents in the Palo
Alto secondary effluent exert a demand for ClOj. Ster-
ile buffered deionized water (D.I.) and sterile filtered
(0.45p) secondary effluent were inoculated with 10®
fecal coliforms/ml. The bacteria had been obtained
from the completed MPN procedure and purified to
yield a strain of Escherichia coli. This culture was then
grown overnight in nutrient broth, centrifuged, and
resuspended in sterile buffer water.
The data are summarized in Table 8, along with
similar data from the work of Walters (1976)(6). A
comparison of the secondary effluent results with Walters'
data indicates that our method of bench-scale dis-
infection is comparable with reported studies that use
a "clean system" with a pure inoculum. Also, these
two systems are identical with the possible exception
of the type of pure inoculum used: a "fecal strain of
coliform" which we used versus a culture of Escherichia
coli B used by Walters. While one experiment is incon-
clusive, the variable resistance of strains of one bac-
terial type seems to be important. Hence, the validity
of predicting the disinfection performance on a mixed,
native population of bacteria in wastewater based on
results of pure culture studies is questioned.
TABLE 8. EXPERIMENT B: EFFECT OF BACTERIAL
CULTURE AND DISSOLVED SUBSTANCES ON
DISINFECTION WITH 5 mg/l
Contact
Time
0
5 min
10 min
Log Number Bacteria/100
Data of D.I.
.' Walters (1976X6) H,O
7.91 8.49
3.78 1
3.58 0
ml
Secondary
Effluent
8.54
1.43
0.3
Chemical Characteristics
Alk. NH,-N
Sample pH mg/l mg/l
D.I.
H O 6.5 10.3 0
Second.
Effl. 8.2 289.2 45
CODT
mg/l
8
39
A comparison also can be made between the D.I.
water and secondary effluent. After a 5-min. contact
time, which should ensure complete mixing, there is
no substantial difference in bacterial kills. Therefore,
in this sample of wastewater, there seems not to have
been any inhibition of disinfection caused by dissolved
substances in the secondary effluent.
Based on the results of these experiments, a study
was designed to verify the comparison between a pure
culture and a native population, to compare two
wastewaters and their respective coliform populations,
and to compare chlorine and chlorine dioxide on pure
cultures of Escherichia co/i.
The first comparison, a pure culture of E. coli with a
native coliform population, verified the results ob-
tained earlier. Namely, pure batch-grown cultures of
E. coli seem to be more susceptible to disinfection than
the native populations from which they were isolated.
Samples of secondary effluent, settled and containing
66
-------
CHLORINE DIOXIDE
non-settling floe particles, were blended in a Waring
blender for 30 seconds to minimize any physical pro-
tection afforded by aggregation (Moffa and Smith,
1975X3). These samples were then dosed with chlorine
dioxide and chlorine, as were samples of filter-sterilized
secondary effluent and buffered D.I. water that con-
tained an inoculum of E. coli. The results are summar-
ized in Table 9 and are similar to the results obtained
earlier and indicated in Table 8. Again, pure batch-
grown cultures of E. coli seem to be more susceptible
to disinfectants than the native populations from
which they were isolated. Also, the data indicate that
there is not a significant demand exerted by dissolved
substances in the effluent relative to demand exerted
by the disinfection process, since both the filter-sterilized
effluent and buffer water yield approximately the
same log survival. This had also been observed earlier
and is noted in Table 9. Finally, a comparison of the
influent ("0-time") fecal coliform counts between the
blended secondary and settled, blended secondary
effluents reveals that most coliform organisms are not
associated with settleable floe-type particles. Instead,
these bacteria seem to be present as more dispersed
growth, or associated with smaller suspended particles.
Therefore, these indicator organisms are generally
less protected than other bacteria which are enriched
selectively in the activated-sludge process by virtue of
their ability to flocculate.
Another comparison between two different
wastewaters was made to verify the data relating
to the relative inefficiency of chlorine dioxide as
a bactericide in secondary effluent. Effluent from
San Jose Wastewater Treatment plant, a 3.4 x
105mVd (90 mgd) activated sludge plant, was
compared with Palo Alto secondary effluent
which has been used throughout this feasibility
study. The data are summarized in Table 10. The
coliform survival data (Table 10A) are similar to
the results that have been obtained with Palo
Alto secondary effluent.
The performance of chlorine dioxide and
chlorine are nearly identical when compared using
pure cultures in secondary effluent (Tabje 10B).
This result has been observed before (Table 9),
but is contrary to other information (Bernarde et
al., 1967; Walters, 1976) indicating that chlorine
dioxide is a more efficient bactericide than
chlorine. There was a greater demand for
chlorine dioxide in San Jose effluent than in
Palo Alto effluent. However, this seemed to have
a minimal effect on the bactericidal effectiveness.
The equal disinfection ability of chlorine dioxide
and chlorine 'is still unexplainable; the only ap-
parent difference between the compared ex-
periments is the type of organism and possibly,
TABLE 9. EFFECT OF BACTERIAL CULTURE ON DISINFECTION WITH 5 mg/l CIO2 AND 5 mg/l CI2
Contact
Time
0
2
5
10
LOG SURVIVAL
Native Population of Fecal Conforms
Blended Secondary
Effluent
CIO 2
4.6
1.9
1.3
1.0
CI2
4.6
3.8
2.4
2.1
Settled, Blended
Secondary Effluent
CIO 2
4.5
2.0
0
0.3
CI2
4.8
2.5
2.0
1.3
Pure Cultures of E coif
Filter-Sterilized
Secondary Effluent
CIO 2
8.3
6.4
1.0
0
CI2
8.4
7.2
—
1.2
Sterile
D.I.
Buffered
Water
CIO2 CI2
8.0
6.8
0
0.5
8.6
6.7
0
0
TABLE 10. PALO ALTO vs SAN JOSE SECONDARY EFFLUENT 5 mg/l CIO2 AND
A. Secondary Effluent, Native Populations
SAN JOSE
Contact
Time,
minutes
0
2
5
10
CIO:
Fecal Conforms
N/100 ml
4.8
2.2
2.0
1.8
Residual
mg/l
1.64
1.70
1.58
Cf2
Fecal Conforms
N/100 ml
4.8
1.9
1.7
1.6
Residual
mg/l
3.37
3.31
3.32
67
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 10. PALO ALTO VS SAN JOSE SECONDARY EFFLUENT 5 mg/l CIO2 AND CI2 continued
B. Filter-Sterilized Secondary Effluent, Pure Cultures of E. coll
Contact
Time,
minutes
0
2
5
10
SAN JOSE
CIO,
Fecal
Conforms
N/100 ml
8.5
1.9
1.1
0
Resid-
ual
mg/l
—
1.64
1.35
1.56
Cl,
Fecal
Conforms
N/100 ml
8.4
1.4
0
0
Resid-
ual
mg/l
—
3.34
3.45
3.21
PALO ALTO
CIO 2
Fecal
Coliforms
N/100 ml
8.5
0
0
0
Resid-
ual
mg/l
—
3.48
2.76
2.33
CI2
Fecal
Coliforms
N/100 ml
8.6
2.8
0
0
Resid-
ual
mg/l
—
3.71
3.79
3.74
C. Sterile Buffered Water, Pure Cultures of E. coll
Contact
Time,
minutes
0
2
5
10
SAN JOSE
CIO2'
Fecal
Coliforms
N/100 ml
8.2
0
0
0
Resid-
ual
mg/l
—
4.81
4.60
4.32
CI2
Fecal
Coliforms
N/100 ml
8.1
0
0
0
Resid-
ual
mg/l
—
4.23
4.12
4.06
PALO ALTO
CIO 2
Fecal
Coliforms
N/100 ml
8.6
3.5
0
0
Resid-
ual
mg/l
—
4.64
4.32
4.13
Clj
Fecal
Coliforms
N/100 ml
8.6
2.8
0
0
Resid-
ual
mg/l
—
4.29
4.09
3.94
Note: All bacterial counts are log number surviving at the given contact time; residual Cl2 is as total available Clg; residual
is as CIO2
the conditions used to culture the populations.
These variables and mechanisms of disinfection
will have to be studied further to explain the
discrepancies.
Finally, two different populations of E. coli,
isolated from San Jose and Palo Alto secondary
effluent, were compared. No significant difference
was observed in log-kills of the two different
cultures. Therefore, it seems that the method of
culturing pure populations of bacteria used in
disinfection studies as opposed to the source of
the population may be important regarding
mechanisms of disinfection.
SUMMARY
I. Suspended solids offer some physical protection
to the bacteria from the action of chlorine diox-
ide.
2. Pure batch-grown cultures of E. coli are more
suseptible to ClOj and Clj than the native popu-
lations from which they were isolated, and the
method of culturing pure populations of bacteria
may be important regarding responses to disin-
fectants.
3. Chlorine dioxide and chlorine are nearly identi-
cal when comparing kills of pure cultures of E.
coli or native populations of fecal coliforms in
either Palo Alto or San Jose secondary effluent.
III. Virucidal Effectivensss of Chlorine Dioxide
Selected experiments throughout the study were
conducted to assess the virucidal effectiveness of chlo-
rine dioxide as compared with chlorine. Coliphage
and Poliovirus I were inoculated into secondary efflu-
ent, then dosed with 5 mg/l chlorine dioxide and
chlorine. Samples were taken at 2-, 5-, and 10-minute
contact times and analyzed for virus and fecal coli-
forms by Dr. Robert Cooper, University of California
at Berkeley. The results are shown in Figure 2. In both
trials, chlorine dioxide was a much more effective viru-
cide than was chlorine, although the coliform survivals
would indicate that both disinfectants were performing
equally well.
The log reductions of the coliphage and poliovirus
by each of the disinfectants was similar, a response
that had been observed in preliminary experiments.
Both phage and virus experienced significant log re-
ductions by chlorine dioxide, and both were affected
less by the same dose of chlorine. Since the test pro-
68
-------
CHLORINE DIOXIDE
cedure for poliovirus is much more expensive and time
consuming than the recovery procedure for coliphage,
the coliphage was chosen as an indicator of virus re-
sponse to both disinfectants. The coliphage had been
isolated from this secondary effluent, as was described
earlier. Therefore, in situ phage, which may be more
realistic indicators of native populations of other path-
ogens, was used as opposed to inocula of laboratory-
grown strains of organisms.
o
Z
O
o
-6
\
5mg/i
C102 •
LEGEND
FC = FECAL COLIFORM
4> = COLIPHAGE
VIRUS = POLIOVIRUS
\\\
\\
\ ^d
\
Q
\
D-
FC
. Q
VIRUS
*
02 5
CONTACT TIME (min)
Figure 2. Comparison of In situ coliphage and an
inoculum of Poliovirus I in secondary effluent.
The remaining virus experiments were conducted in
parallel with a large grid of experiments designed for
modeling dose-contact time relationships for chlorine
dioxide and chlorine. A full factorial experiment in
randomized block design was selected. Three dosage
levels (2, 5, and 10 mg/1) of each disinfectant, and
three contact times (5-, 15-, and 30-minutes) were
tested in eight replicate experiments. Reductions of
total coliform organisms were measured in all experi-
ments, and coliphage reductions were measured as
well in Experiments 1 and 8. The order of choice of
disinfectant and dose and the sequence of analyses for
bacterial and viral counts were all randomized.
The reactor for these experiments was a 4-liter aspir-
ator bottle which had been modified with 4 identations
BENCH SCALE DISINFECTION REACTOR
INDENTATIONS
FOR HIGH SHEAR
MIXING
STIR BAR
CONSTANT TEMP.
H2O BATH
Figure 3. A rapid-mix, rapid-sampling, 4 liter reactor.
Dye studies indicate that complete mixing
takes place in less than 5 sec.
2 mg/1
5 mg/l
(10 mg/l
CONTACT TIME (min)
Figure 4. Total coliform survival comparing CIO2 at
3 doses and 3 contact times in secondary
effluent.
69
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
similar to those of a trypsinizing flask (Fig. 3). The
reactor was also equipped with an injector and sampler
that were located in the most active mixing zone of the
reactor. All connections were glass or teflon. The
vessel was pressurized (12 psig) and both disinfectant
and sample delivery were pressure driven. A dye study
indicated that complete mixing of injected disinfectant
liquid and the wastewater occured in less than five sec-
onds. The reactor was placed in a constant-temperature
bath and all experiments were conducted at 24±0.5°C.
o
oc
E'o
o -«-
in —,
O
<
x
Q.
_J
O
o
CONTACT TIME (min)
Figure 5. In situ coliphage survival in secondary efflu-
ent at 3 doses and 3 contact times.
The results of the virus analyses that were run paral-
lel with the total coliform analysis are shown in Figures
4 and 5. Wastewater characteristics are summarized in
Table 11. Once again, the total coliform bacteria re-
spond similarly to the stress of a given disinfectant
dose (Fig. 4), whereas all three doses of chlorine diox-
ide (2, 5, and 10 mg/l) are much more effective at in-
TABLE 11. WASTEWATER CHARACTERISTICS
Parameter
Temperature, °C
Total Filterable Residue, mg/l
Chemical Oxygen Demand, mg/l
Alkalinity, Total mg/l as Ca CO,
PH
Nitrogen: Ammonia-N, mg/l
Experiment Number
1
24
28.0
30.8
222.1
7.4
29
8
24
23.5
34.0
247.9
7.0
34
activating the coliphage than even the highest dose (10
mg/I) of chlorine (Fig. 5). This suggests that accepted
indicator organisms (total coliform bacteria) may
yield conservative, possibly inaccurate, performance
data for disinfection with chlorine dioxide in secondary
effluent
A final set ot experiments was run to assess
he biocidal effectiveness of both disinfectants at
v'ery short contact times, since the shortest con-
tact time studied had been 2 minutes (Fig. 2).
These experiments were run at 5 mg/1 doses C1O2
and C12 and sampled at 15, 45, 90, and 120
seconds. The results (Fig. 6) again indicate that
C1O2 is a more effective virucide than a bac-
tericide, and that C1O2 performs better than CI2
in a short contact time (<120 seconds) in secon-
dary effluent.
I "'
-M
o
o
^ -3
O
1 -4
-5
TOTAL COLIFORM
FECAL COLIFORM
L^-^_ nTOTAL COLIFORM
-=>-~ «=r— b D
' D
FECAL COLIFORM
. COLIPHAGE
D — D
OC
3
V) -6
O
0
-7
LEGEND
D--- O C102
• • C12
i i
5 mg/l
5 mg/l
i i
15 45 90
CONTACT TIME (sec)
120
Figure 6. A comparison of the responses of several
organisms to ClC^and Cljat very short con-
tact times in secondary effluent.
The differential behavior of chlorine dioxide com-
pared with chlorine and of bacteria compared with
virus in response to chlorine dioxide in secondary
effluent merit further investigation. A study of the
factors that influence mechanisms of inactivation by
C1O2 in a complex matrix such as secondary effluent
may have important implications regarding the design
of processes for disinfection.
70
-------
CHLORINE DIOXIDE
SUMMARY
1. In situ coliphage and inocula of Poliovirus I ex-
hibited the same response to both chlorine diox-
ide and chlorine dosed at 5 mg/1 in secondary
effluent.
2. Coliphage inactivation at all doses of chlorine
dioxide (2, 5, and 10 mg/1) is much greater than
the highest dose of chlorine (10 mg/1) in secon-
dary effluent.
3. Chlorine dioxide is a more effective virucide
than a bactericide, and chlorine dioxide performs
equal to or better than chlorine at all contact
times (15 sec. to 30 min.) and doses (2 to 10 mg/1)
tested.
CONCLUSIONS
I. The M PN and M F methods of counting TC and
FC are equal in accuracy of measurement; how-
ever, the MF method is more precise.
2. The variable resistance to disinfection of types
of indicator organisms and types of populations
chosen for a disinfection study is critical.
3. Chlorine dioxide is a significantly more effective
virucide than is chlorine in secondary effluent,
although disinfection efficiency as measured by
coliform survival indicates that C1O2 is not dif-
ferent from Clg.
REFERENCES
I. Benarde. M.A.. W.B.S. Snow, V.P. Olivieri and B. Davidson.
Applied Microbiology. 1967. I5(2):257.
2. Kott, Y'., Applied Microbiology. 1965. 14(2):141.
3. Moffa, P.E., and .I.E. Smith, "Bench-Scale High-Rate Disinfec-
tion of Combined Sewer Overflows with Chlorine and Chlo-
rine Dioxide." EPA-670/2-75-021, U.S. Environmental
Protection Agency. Cincinnati, Ohio, April 1975.
4. Mowat, A. 1976. JWPCF. 4X(4):724.
5. Standard Methods for the Examination of Water and Waste-
water, 14th edition. 1975, APHA, Washington, D.C.
6. Walters, Gary E., "Chlorine Dioxide and Chlorine: Comparative
Disinfection." M.S. Thesis. Johns Hopkins University,
1976.
71
-------
10.
CHLORINE DIOXIDE: ANALYTICAL MEASUREMENT
AND PILOT PLANT EVALUATION
E. Marco Aieta, * Bruce Chow* and Paul V. Roberts1
Department of Environmental Engineering
Stanford University
Stanford, California
"Graduate Research Assistant, Stanford University and
"Adjunct Professor of Environmental Engineering, Stanford University
INTRODUCTION
Chlorine has long been favored as a disinfectant in
both water and wastewater treatment by virtue of its
bactericidal effectiveness, low cost and convenience,
and relatively long-lived residual. However, chlorina-
tion as normally practiced in water and wastewater
treatment has been found to result in the formation
of trihalomethanes and other chlorinated organics
that are undesirable from the viewpoint of water
pollution control in general and human health specifi-
cally (11).
The EPA is actively seeking substitutes for chlorine
as a disinfectant. Chlorine dioxide has received con-
siderable attention in this respect. Chlorine dioxide
has been used most extensively in this country by the
pulp and paper industry as a bleaching agent and to a
small extent for the control of taste and odor problems
in a few water treatment systems (14). Chlorine dioxide
has been found in a number of investigations to be
equal or superior to chlorine as a disinfectant (1, 2,9,
10, 13). However, most of the studies were conducted
in clean systems or synthetic sewages inoculated with
lab-grown strains of bacteria. These results may not
be directly applicable to typical municipal waste-
waters.
The topics from this research that will be presented
here include:
1. Analytical methods for the measurement of both
stock and residual concentrations of chlorine
and chlorine dioxide.
2. The design and results of a dose-time matrix
experiment in which chlorine and chlorine di-
oxide are compared in the disinfection of a mu-
nicipal secondary effluent.
3. The development from the results of the dose-
time matrix experiment of a simple disinfection
model for chlorine or chlorine dioxide.
4. Plans for upcoming verification of the laboratory-
derived model on a pilot plant scale.
ANALYTICAL DETERMINATION
OF CHLORINE AND CHLORINE DIOXIDE
Measurement of chlorine dioxide in the mg/1 range
is a relatively new field. The experience gained over
many years in the study of chlorine and its residual
forms provides a starting point for the development
of a chlorine dioxide residual measurement technique.
The following is a brief description of the stock genera-
tion techniques, stock measurement techniques, and
residual measurement techniques for both chlorine
and chlorine dioxide as used in this study.
Chlorine stock solutions were generated by bubbl-
ing chlorine gas through chilled distilled water. Chlo-
rine dioxide solutions were generated by the sulfuric
acid activation of sodium chlorite as described in
Standard Methods (12). In order to insure a pure chlo-
rine dioxide solution, the generated chlorine dioxide
gas was passed through a tower of sodium chlorite
flakes and collected in chilled distilled water, thereby
removing any chlorine impurity. Both chlorine and
chlorine dioxide stock solutions were standardized
(in triplicate) before each experiment using the iodo-
metric technique prescribed in Standard Methods.
The concentration of chlorine dioxide stock varied
from 300 mg/1 to 600 mg/1. The concentration of chlo-
rine stock varied from 3000 mg/1 to 8000 mg/1.
72
-------
CHLORINE DIOXIDE
Several methods are available for the measurement
of chlorine residuals in wastewater. These methods
have been extensively developed over many years and
the strengths and weaknesses of each are reasonably
understood. Many of these same methods are also
applicable to the measurement of chlorine dioxide
residuals, with some modifications. Two methods for
residual measurement of chlorine dioxide were chosen
for investigation in this study: 1. DPD—Ferrous
Titrimetric Method, and 2. Amperometric Method.
DPD—Ferrous Titrimetric Determination of Chlo-
rine Dioxide Residuals
In attempting to measure residual chlorine dioxide
by the DPD-FAS technique (6, 7, 8), some difficulties
were encountered. The residual as measured by DPD
was consistently higher than the chlorine dioxide dose.-
The stock solution was standardized immediately
before use by the iodometric method, and the dose
calculated on this basis. Not only was this discrepancy
seen in wastewater, but also in deionized water, dis-
tilled water and tap water. This discrepancy was not
seen when chlorine was the oxidant of interest.
In the iodometric method, chlorine dioxide is
reduced to chloride (Cl~). However, several
researchers (6, 9, 10) have reported that the pH
used in the iodometric analysis (pH*1.8) is not
low enough to effect complete reduction of the
chlorine dioxide to Cl~ (see Fig. 1). If indeed,
this was the case, the iodometric results would
give a low value for the concentration of chlorine
dioxide in the stock solution. The DPD method,
which is performed at pH of 6.2-6.4 is assumed
to measure the reduction of chlorine dioxide to
chlorite, and would therefore indicate a higher
value of chlorine dioxide than that given by the
iodometric analysis..
UlUo ' 6
r*in i RO — .-
* Pin ~
— , 9 o
pH 7
pH<2
FIGURE 1. Reduction of CIO2 According to pH.
To test this hypothesis, two experiments were per-
formed. In the first experiment, the iodometric anal-
ysis was performed at various pH's ranging from 2.0
to 0.8. No difference in titration value was found. In
the second experiment, the iodometric analysis per-
formed at pH 1.8 was compared with the amperometric
analysis at pH 7. The amperometric method is analo-
gous to the iodometric method with these exceptions:
I. In the amperometric analysis the end-point is
detected by observing the change in current flow
in the sample while adding a strong reducing
agent as titrant.
2. In the amperometric analysis phenylarsine oxide
is the titrant instead of sodium thiosulfate as used
in the iodometric method.
3. In the amperometric analysis the determination
is performed at pH 7. At this pH only one elec-
tron transfer is possible for chlorine dioxide.
Since at a pH of 7, chlorine dioxide is only reduced
to chlorite, a higher concentration of chlorine dioxide
as measured by the amperometric analysis would in-
dicate that chlorine dioxide was not being fully re-
duced to chloride as assumed in the iodometric method.
The analyses were run in distilled water. The results
appear in Table I, and a statistical analysis appears in
Table 2.
TABLE 1. COMPARISON OF AMPEROMETRIC AND
IODOMETRIC METHODS FOR DETERMINING
CONCENTRATION OF CIO2
Trial
1
2
3
4
5
Mean
S.D.
S2
Iodometric*
563.7
563.7
566.4
561.0
560.3
563.0
2.18
4.77
Amperometric*
547.2
554.8
571.9
560.5
558.6
558.6
8.06
64.98
*mg/l AS CIO2.
TABLE 2. COMPARISON OF AMPEROMETRIC AND
IODOMETRIC METHODS BY ANALYSIS OF VARIANCE
Source of Variance
Between Methods
Error
Degrees
of
Freedom
1
8
Sum
of
Squares
48.8
348.7
Mean
Square
48.8
43.6
F-Ralio
1.12
TOTAL.
397.5
As can be seen from Table 2, the difference between
the values obtained by the two methods is not signifi-
cant. It can be inferred from these data that the iodo-
metric method accurately measures the concentration
of chlorine dioxide in stock solutions. The inconsis-
tency must then be in the assumed reactions of the
DPD-FAS method. It is hypothesized (although un-
tested at this time) that the chlorite from the reduction
73
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
of chlorine dioxide is reacting with the DPD reagent to
a slight extent. Due to the continuing problems with
the DPD-FAS method, further investigations into
utilizing the amperometric method for measurement
of residual chlorine dioxide were made.
Amperometric Determination of Chlorine Dioxide
Residuals
To investigate the possible interference of chlorite
in the determination of chlorine dioxide by the amper-
ometric method, a solution of sodium chlorite was
added to a sample containing chlorine dioxide in dis-
tilled water. Since the sodium chlorite used was only
~ 95 percent pure, glycine (aminoacetic acid) was
added to the sample to react any chlorine or bromine
compounds present which would give an apparent
positive result from the chlorite. Results of this exper-
4. The experiment involved dosing 2 liters of Palo
Alto secondary effluent to 4.50 mg/1 with pure chlo-
rine dioxide. The sample was then contacted for 2
minutes. After the contact period, sufficient KI and
pH 7 buffer were added to the sample to react with all
the chlorine dioxide present. In the case of the I2 back-
titration, sufficient PAO was added to swamp the re-
leased l^- Eight (8) replicates were run at approxi-
mately one-minute intervals. The experiment was
repeated for filtered Palo Alto secondary effluent. The
values reported have been corrected for blank titra-
tions. In the case of the forward titration, no correction
was necessary, but the I2 back-titration more I2 solu-
tion was required in the blank than was indicated by
TABLE 4. AMPEROMETRIC TITRATION EVALUATION*
iment are reported in Table 3.
As seen in Table 3, chlorite does not interfere with
the chlorine dioxide measurement.
TABLE 3. THE EFFECT OF CHLORITE ON THE
DETERMINATION OF CHLORINE DIOXIDE
BY AMPEROMETRIC TITRATION
CIO, + 4.33 mg/l CIO,
Sample CIO2 + Glycine-
1 5.47 5.55
2 5.55 5.51
3 5.72 5.55
4 5.61 5.61
5 5.59 5.61
Mean 5.59 5.57
Sample
1
2
3
4
5
6
7
8
Mean
S.D.
S2
' Results are
Secondary Filtered
Forward
Titration
1.90
1.86
1.90
1.84
1.81
1.73
1.67
1.91
1.80
±0.083
0.0069
Back
Titration
2.12
2.31
2.13
2.22
2.22
2.12
2.12
2.12
2.17
±0.067
0.0045
Secondary Unflltered
Forward
Titration
1.62
1.62
1.52
1.52
1.52
1.48
1.39
1.27
1.49
±0.109
0.0119
Back
Titration
1.88
1.97
2.07
1.88
1.88
1.88
1.88
1.79
1.90
±0.077
0.0060
3iven as mg/l CIC^ at 2-minute contact time; dose = 4.50 mg/l CIC^.
' Two ml of 10% W/V solution of Glycine added to 200 ml of sample.
The amperometric titration measures iodine (I2)
released into solution by the oxidation of iodide by
any powerful oxidant such as chlorine dioxide or
chlorine. If this released I2 comes into contact with
organic constituents in the sample (as in the case of
wastewater), the I2 will react with the organics. To
prevent this, a back-titration procedure is recom-
mended, in which the liberated I2 is immediately re-
acted with phenylarsine oxide (PAO), which has been
added to the sample prior to the addition of iodide (as
KI). The PAO is added in excess and the unreacted
PAO is back-titrated with an I2 solution of known
concentration. The concentration of chlorine dioxide
(or chlorine) is then determined by difference. To eval-
uate the need for this back-titration procedure, chlo-
rine dioxide residual measurements were made in
Palo Alto secondary effluent.
The results of this experiment are reported in Table
TABLE 5. ANALYSES OF VARIANCE FOR
AMPEROMETRIC TITRATION
Source of Variation
Degrees
of
Freedom
Sum
of
Squares
Mean
Square
F-Ratio
ANOVA for filtered
secondary effluent:
Between forward
and back titrations
Error
TOTAL
1
14
15
.54023
.09135
.63158
.54023 82.79*
.006525
ANOVA for unfiltered
secondary effluent:
Between forward
and back titrations
Error
TOTAL
'P< .001 that F is exceeded.
1
14
15
.67651
.14274
.81924
.67651
.01020
66.35*
74
-------
CHLORINE DIOXIDE
the amount of PAO added. (The \z solution was stan-
dardized immediately prior to use). Although not
shown in Table 4, the blank titrations (done in tripli-
cate) showed the same variation as the forward or
backward titration. The standard deviations shown in
Table 4 are not statistically different from one another
as measured by the F-test.
Table 5 gives two analyses of variance tables compar-
ing the forward titration with the ^ back-titration. As
indicated, there is a significant difference between the
values measured by the two titrations. Based on these
results, the amperometric back-titration method was
chosen for the measurement of chlorine dioxide resid-
uals in wastewater.
Amperometric Determination of Chlorine Residuals
The amperometric back-titration method for mea-
surement of chlorine residuals in wastewater is a widely
used and well-accepted method. Since this method was
found acceptable for chlorine dioxide residual mea-
surements, it was also chosen for the measurement of
chlorine residuals. The analyses for chlorine residuals
(as Total Available Chlorine) are performed at pH 4 to
insure quantitative response from all forms of com-
bined chlorine. At this pH there is no need for a blank
correction. The reaction of PAO with \% must be some-
what slower at pH 7 than at pH 4, so that some of the
added \2 titrant would be available for reaction with
other sample constituents; hence the need for blank
correction in determining chlorine dioxide residuals
at pH 7.
Future Requirements
The amperometric back-titration procedure reported
here has proved adequate for the needs of this study.
There is, however, a continuing need for a residual
measurement technique capable of differentiating
between chlorine dioxide and chlorite in the same
wastewater sample. Chlorite is suspected of being a
human health hazard, specifically of being a cause of
methemoglobinemia. In addition, the use of chlorine
dioxide in conjunction with chlorine may be the most
„economically attractive means of pathogen destruction
in wastewater, so that an analytical method to differ-
entiate chlorine dioxide from chlorine would be re-
quired to control the disinfection process. An ideal
method, then, would be able to differentiate chlorine
(as chloramines), chlorine dioxide, and chlorite in a
wastewater matrix. Work is continuing in this area
with the amperometric technique as the primary ana-
lytical tool.
COMPARATIVE BACTERICIDAL EFFECTIVE-
NESS OF CHLORINE AND CHLORINE DIOXIDE
Reactor Design
To compare the relative effectiveness of chlorine
dioxide with that of chlorine, secondary effluent from
the Palo Alto Water Pollution Control Plant was
dosed with various concentrations of either chlorine or
chlorine dioxide. The reactor for these experiments
was a specially modified four-liter aspirator bottle.
The modifications include four indentations similar to
those of a trypsinizing flask, oxidant injector and
sample ports located in the most active mixing zone
of the reactor, and all glass or teflon connections. The
reactor was maintained at 15 psig and both disin-
fectant injection and sample delivery were pressure
driven. Stirring was provided by a 21/:" magnetic
stir bar (cylindrical) with ratational speed of
~ 600 rpm. The reactor was maintained at
24°C for all experiments.
To evaluate the dispersion efficiency of added
chemicals within this batch reactor, a dye-train study
was performed. The dye-train study involved dosing
three liters of deionized water with a methylene blue
stock solution of 500 mg/1. Sufficient stock was added
to make a final concentration of 5 mg/1 in the reactor.
These conditions mimicked those in which actual dis-
infection experiments were performed. Samples were
taken at 2, 5, 10, 20, 40, 60 and 120 seconds and ana-
lyzed spectrophotometrically for methylene blue. The
results of this experiment indicate that complete mix-
ing is effected in less than five seconds.
Experimental Design
The experiments for the comparison of bactericidal
effectiveness of chlorine and chlorine dioxide were
performed on wastewaters from the Palo Alto Water
Pollution Control Plant. Secondary effluent was col-
lected at about 8:00 AM on the day of each experiment.
Eight separate experiments were performed on five
different days over a two-week period. On three days,
two experiments were run on the same wastewater. On
two days, only one experiment was run. An experi-
ment is a complete dose-time matrix in which either
chlorine or chlorine dioxide is added to wastewater
at a dosed concentration of 2, 5 or 10 mg/1. All of the
six runs within an experiment (two disinfectants, three
doses) were done in randomized order. Samples were
taken at 5-, 15- and 30-minute intervals for both bac-
terial analysis and residual disinfectant determination.
75
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
Samples for bacteria analysis were also taken at time 0.
Table 6 graphically illustrates the dose-time matrix of
this study.
TABLE 6. EXPERIMENTAL DOSE-TIME MATRIX FOR
CHLORINE CHLORINE DIOXIDE COMPARISON
g 5 MIN
1-
u 15 MIN
K
2
8 30 MIN
CHLORINE
DOSE
mg/l
2 5 10
CHLORINE
DIOXIDE
DOSE
mg/l
2 5 10
Experimental Results
All samples for bacterial analysis were analyzed by
the membrane filter technique for total coliform bac-
teria (12). Four dilutions of each sample were analyzed
in order to obtain an incubated sample plate with a
significant number of colonies to count but a suffi-
ciently small number, so that the colonies do not over-
lap (usually 20-80 colonies are adequate). It is possible,
then, to have from one to four measurements of coli-
form bacteria for each sample. The logarithms of the
numbers of bacteria for each sample are given in Table
7 for samples disinfected with chlorine and in Table 8
for samples disinfected with chlorine dioxide. Where
more than one value appears for each sample, the in-
dividual values represent bacterial counts from differ-
ent dilutions but from the same sample.
Statistical Analysis of Experimental Results
Preliminary statistical analyses were done on the
data from Tables 7 and 8. To investigate any biases in
the data due to the order of experiments or sampling
technique, the data from time 0 were subjected to an
analysis of variance (ANOVA). The grouping vari-
ables used were wastewater, disinfectant, and disin-
fectant dose. In setting up the ANOVA in this way, all
experiments are considered but there is no difference
between experiments except the wastewater on which
the experiment was run. The res-ults of this ANOVA
are shown in Table 9.
The only effect that is significant is that of waste-
water, indicating that the variance in bacterial num-
bers at time 0 for each experiment is due only to the
difference inherent in the wastewater used for that
particular experiment. No significant correlation was
found between measured wastewater parameters
(Total Filtrable Residue, COD, Alkalinity, pH, am-
monia-nitrogen) and bacterial counts.
The next step in the statistical analysis was to inves-
tigate any differences between experiments performed
on the same wastewater; that is, the reproducibility of
experiments. Again, analysis of variance was used
with grouping variables of wastewater, experiment,
disinfectant, disinfectant dose, and contact time. The
ANOVA is reported in Table 10. As seen in this ANOVA,
all variables except experiment have a significant
TABLE 7. LOG [SURVIVING BACTERIA] IN EIGHT REPLICATE EXPERIMENTS USING CHLORINE AS DISINFECTANT
Dose:
Experiment:
2 mg/l
4 5
6
1
5 mg/l
4 5
1
10 mg/l
4 5
5.85 6.09 6.00 6.43 6.81 6.63 6.62 5.89 5.78 6.14 6.10 6.79 6.83 6.67 6.63 5.83 5.92 6.30 6.08 6.70 6.84 6.67 6.60 5.73
OMIN 5.906.286.186.74 5.90 5.666.326.36 6.30 6.00 5.84 5.75
5.78 5.30 5.60 5.30
5.48
4.78 5.75 5.95 6.16 6.13 6.30 6.34 5.37
5 MIN 5.00 5.30 5.70 6.48 6.15 6.28 5.40
5.60
4.09 3.83 3.96 4.15 4.15 4.53 3.90 3.86
4.08 3.90 4.08 4.06 4.15 4.26 4.06
3.60 4.08 3.78 4.41 4.30 4.08
3.95 3.70 4.14 4.49 3.95
2.91 2.51 2.34 3.77 2.81 3.24 3.24 3.00
2.81 2.26 2.40 2.69 3.08 2.67
2.16 2.42 2.42
2.07 2.09
4.48 4.70 4.92 5.34 4.72 5.45 5.41 4.60 2.84 2.28 2.36 2.78 3.49 4.26 3.85 3.01
15 MIN 4.62 4.79 5.25 4.46 5.62 5.55 4.58 3.15 2.00 3.23 3.34 3.70 4.79 3.75 3.40
4.30 4.90 5.72 4.78 5.68 3.48 3.30 3.48 3.30 3.34
4.30 4.84 5.53 4.48 4.30 3.60
2.27 1.00 1.74 1.34 2.04 2.24 2.22 2.04
1.81 1.41 1.76 1.30 1.91 1.79 1.66
1.76 1.20 1.50 1.75 1.91 1.41
1.57 1.04 1.48
30 MIN
4.34 3.30 3.30 4.70 4.00 4.78 4.41 4.00
3.30 4.58 3,70 4.70 4.80 3.95
4.00 4.60 5.00 4.78 4.00
4.00 4.85 5.08
2.71 1.88 1.76 2.91 3.23 2.86 3.30 2.34 1.04 0.60 1.15 1.53 0.30 0.95 1.23 0.48
2.48 1.90 1.30 2.87 3.21 2.89 3.11 2.73 0.60 0.60 0.34 1.30 0.70 1.26 0.34
1.30 2.15 3.38 3.20 2.84 3.00 2.86 0.48 0.54 0.48
2.00 2.60
76
-------
CHLORINE DIOXIDE
TABLE 8. LOG [SURVIVING BACTERIA] IN EIGHT REPLICATE EXPERIMENTS
USING CHLORINE DIOXIDE AS DISINFECTANT
DOSE :
EXPERIMENT:
2 mg/l
5 mg/l
10 mg/l
0 MIN
o
o
5 MIN
15 MIN
30 MIN
1 I 2
5.91 6.05
5.90 6.18
5.08 4.41
5.95 4.36
4.30
5.00
5.26 4.38
5.00 4.38
4.48
4.00 4.20
4.78 4.32
4.48
| 3 | 4
6.09 6.73
6.45
4.34 5.00
4.50 4.97
508
5.00
438 506
4.26 4.95
4.30 4.78
5.04
3.60 4 78
3.04 4.81
4.30 4.90
4.30 5:00
| 5 | 6
6.79 6.41
6.67
4.99 5.21
4.99 5.02
5.11 5.34
500 5.00
5.25 5.17
5.44 5.00
5.15 5.48
5.00 5.15
4.81 5.09
4.90 4.92
5.26 5.08
5.08 5.00
| 7 | 8
6.48 585
591
5.90
570
5.08 5.03
494 504
5.41 5.30
5.00
5.20 4.93
5.09 461
5.38 460
526 4.78
5.15 4.91
5.05 5.08
545 411
5.36
1 I 2 |
5.83 6.02
5.60 628
5.00
3.99 2.00
391
323 200
320
300
3.33 1.78
3.37 1.90
3.11 190
3 I 4
6.26 683
278 3.76
3.30 3.90
360
3.30
1.70 238
227
2.30
3.00
1 20 2.89
1.30 306
1 30 3.11
| 5 | 6
676 6.52
3.45 3.41
3.75 3.30
4 15 3.00
404
2.89 303
315 382
2.90 3.72
370
3.15 3.12
315 3.15
350 3.83
7 1 8
667 577
5.88
5.00
3.98 3.20
408 3.30
3.30 3.70
3.30
2 28 268
228 3.04
3.00 2.30
3.00
2.99 3.46
3.24 3.56
3.00 3.32
1 I 2
5.87 6.12
5.60 6.38
1.90 1 30
1.95 1.00
2.18 0.30
2.04
1.98 0.60
1.75 0.48
1 49
031 0.34
0.60
040
i 3 | 4
6.15 6.83
620
1 78 2.62
1.78 2.46
1.56
0.30 2.11
0.34
0.34 0.34
0.30
5
6.70
2.15
2.15
1.70
1.40
1.68
1.66
0.34
0.30
6 | 7 | 8
6.73 6.93 5.93
5.75
2.26 2.45 1.60
2.56 2.43 2.08
2.32 1.95 1.65
198 1.53
1.60 1.18 0.70
1.56 1.30 1.00
1.20 1.08 1.15
1.11 0.90 048
1.15
TABLE 9. ANOVA FOR BACTERIAL ANALYSIS BEFORE DISINFECTION
Grouping Variable Information
Variable
Number
In Group
Values
Wastewater
Disinfectant
Disinfectant Dose
1,2,3,4,5
CIO2, CI2
2, 5, 10 mg/l
ANOVA - Log [N (I)] Dependent Variable*
Wastewater
Disinfectant
Disinfectant Dose
Degrees
of Freedom
4
1
2
Sum of Squares
13.2259
0.0002
0.1625
F-Value
80.64
0.00
1.98
Probability F
F-Exceeded
0.0001
0.9516
0.1483
• Interactions omitted; N (t) = Number of bacteria at time t.
effect on the number of surviving bacteria. (These
effects will be examined in subsequent ANOVA's).
This implies that the duplicate experiments run on the
same wastewater are not significantly different from
each other. For this reason, the experiments were no
longer grouped according to wastewater, so that varia-
tions seen between experiments in subsequent statis-
tical analyses are due to variations in the wastewater.
In other words, the wastewater variable is redundant
in that the experiments differ only due to the effects
of wastewater composition.
The previous statistical analyses have been run with
the logarithms of bacterial numbers from each sample
(see Tables 7 and 8) as the dependent variable. In order
to compare the bactericidal efficiency of chlorine and
chlorine dioxide and to examine the relationships of
disinfectant doses and contact times to bactericidal
efficiency, a survival ratio for each experiment was
calculated and the logarithms of these survival ratios
were substituted as the dependent variable.
The survival ratio is defined as:
N(t)/N(0)
Where:
N(t) = number of bacteria measured at time t,
N(O) = number of bacteria measured at time O.
The logarithm of survival ratio is then:
log[N(t)/N(0)] = log[N(t)] - log[N(0)]
77
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
log[N(T)] was calculated as the arithmetic mean of the
logarithms of bacterial numbers from each sample
(see Tables 7 and 8) for each time t. The log [N(O)J was
similarly calculated at time O. This calculation yields
the geometric mean of the replicate analyses for bac-
terial numbers.
Experiment, disinfectant, disinfectant dose, and
time were used as independent variables. The resulting
ANOVA is shown in Table 11. This ANOVA identifies
several sources of variation:
1. There is a significant difference between experi-
ments. This has been discussed above and is due
to variations in the wastewater used for the ex-
periments.
2. There is a significant difference between disin-
fectant doses; i.e., as the dose increases, the
survival ratio decreases.
3. There is a significant difference between contact
TABLE 10. ANOVA FOR EXAMINING EXPERIMENTAL REPRODUCIBILITY
Grouping Variable Information
Variable
Wastewater
Experiment
Disinfectant
Disinfectant Dose
Contact Time
Source
Wastewater
Experiment
Disinfectant
Disinfectant Dose
Contact Time
Number
In Group
5
8
2
3
3
ANOVA —
Degrees
of Freedom
4
3
1
2
2
Values
1,2,3,4,5
1,2,3,4,5,6,7,8
CIO2, CI2
2,5, 10 mg/l
5, 15, 30 min
Log [N (t)] Dependent Variable*
Sum ol Squares
40.3664
0.5532
7.5686
781.5436
48.2474
F-Value
90.86
1.66
68.15
3518.42
217.20
Probability
F-Exceeded
0.0001
0.1735
0.0001
0.0001
0.0001
Interactions omitted; N (t) = Number of bacteria at time t.
TABLE 11. ANOVA FOR COMPARISON OF CHLORINE AND CHLORINE DIOXIDE AS BACTERICIDES
Variable
Experiment
Disinfectant
Disinfectant Dose
Contact Time
Number
In Group
8
2
3
3
Grouping Variable Information
Values
1,2,3,4,5,6,7,
CIO2, CI2
2,5, 10 mg/l
5, 15, 30 min
8
ANOVA — Log [N (t)/N (0)] Dependent Variable*
Source
Experiment
Disinfectant
Disinfectant Dose
Contact Time
Disinfectant -
Disinfectant Dose Interaction
Disinfectant -
Contact Time Interaction
Disinfectant Dose -
Contact Time Interaction
Degrees
ol Freedom
7
1
2
2
2
2
4
Sum of Squares
21.8691
4.8584
279.2971
37.3407
1 .3392
6.4239
3.4030
F-Value
27.83
43.28
1244.08
166.33
5.97
28.61
7.58
Probability
F-Exceeded
0.0001
0.0001
0.0001
0.0001
0.0037
0.0001
0.0001
Other interactions omitted; Log [N (t)/N (0)] = Log (survival ratio).
78
-------
CHLORINE DIOXIDE
TABLE 12. EXPERIMENTAL DOSE-TIME MATRIX FOR CHLORINE—CHLORINE DIOXIDE COMPARISON
WITH MEAN LOG [SURVIVAL RATIO]* AND RESIDUALS
Contact Time
5 Minutes
10 Minutes
30 Minutes
2
-0.49
(1.36)
-1.38
(1.28)
-2.18
(1.18)
Chlorine
Dose mg/l
5
-2.23
(4.03)
-2.98
(3.88)
-3.72
(3.78)
10
-3.44
(8.64)
-4.62
(8.47)
-5.53
(8.08)
Chlorine Dioxide
Dose mg/l
2
-1.34
(.66)
-1.43
(.58)
-1.61
(.49)
5
-3.01
(2.05)
-3.69
(1.44)
-3.50
(1.23)
10
-4.40
(6.33)
-5.09
(5.48)
-5.81
(4.87)
"Calculated means of 8 experiment replicates.
times; i.e., as the contact time increases, the
survival ratio decreases.
4. There is a significant difference between the dis-
infectants, chlorine dioxide and chlorine. The
survival ratio overall islowerforchlorinedioxide
than for chlorine.
5. There is a significant interaction between dis-
infectant and disinfectant dose; i.e., the relation-
ship between log [survival ratio] and dose is not
the same for both disinfectants.
6. There is a significant interaction between dis-
infectant and contact time; i.e., the relationship
between log [survival ratio] and contact time is
not the same for both disinfectants.
0
-1
ii -2
<
IT
< ~3
I ~4
C/5
C3
O
-6
-7
(.49)
CHLORINE
DIOXIDE
CHLORINE
RESIDUALS
DISINFECTANT DOSE (mg/l)
Figure 2. Comparison of Chlorine and Chlorine Dioxide
Bactericidal Effectiveness at 30 Minute Contact Time
7. There is a significant interaction between dis-
infectant dose and contact time; i.e., the relation-
ship between log [survival ratio] and contact
time is not the same for all doses.
In Table 12, the means of log [survival ratio] for
chlorine dioxide are compared with that of chlorine by
dose and contact time. These means were calculated
over all eight experiments. Mean residual concentra-
tions are shown in parentheses. The means of Table
12 show the differences between chlorine and chlorine
dioxide observed in the analysis of variance main
effects of Table 11. (Observation 1 from above has
14
10 15 20
TIME IN MINUTES
25
30
INITIAL DOSE OF 12.5 mg/l FOR BOTH
CHLORINE AND CHLORINE DIOXIDE
Figure 3. Chlorine and Chlorine Dioxide Die-Away
in Palo Alto Secondary Effluent
79
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
already been discussed.) From Table 12, the difference
between disinfectant doses (observation 2) is obvious,
as in the difference between contact times. The differ-
ence between disinfectants seen in the ANOVA of
Table 11 (observation 4) is modified by the ANOVA
interactions (observations 5-7). To examine the differ-
ence between disinfectants and their relationships to
dose and contact time, Figures 2,3 and 4are presented.
-3
EC
cc
D
C/D
U
o
-4
-5
-6
(864)
CHLORINE
DIOXIDE
CHLORINE
) = RESIDUALS
(8.08)
30
CONTACT TIME (MINUTES)
Figure 4. Comparison of Chlorine and Chlorine Dioxide
Bactericidal Effectiveness at 10 mg/l Dose
Chlorine and chlorine dioxide are compared by
dose at 30-minute contact times in Figure 2. These
values are means over eight experiments. The values
in parentheses are the means of the residual over the
eight experiments. Chlorine gives a lower log [survival
ratio] at the 2 and 5 mg/1 doses, while chlorine dioxide
gives a lower log [survival ratio] at 10 mg/1 dose. This
is the disinfectant dose interaction mentioned above
(observation 5). The reason for this dose dependent
disinfectant difference can be found in the die-away
curves of chlorine and chlorine dioxide. Figure 3
shows typical chlorine and chlorine dioxide die-away
curves in secondary effluent. The initial oxidant
demand (within the first three minutes of contact
time) is much greater for chlorine dioxide than for
chlorine. This initial oxidant demand has been shown
by others not to contribute to the destruction of bac-
teria but is a function of waste composition. The dis-
infection process proceeds after an initial time lag and
is dependent on the concentration of bactericidal agent
present (4, 14). If, as in the case of a 2 mg/l dose of
chlorine dioxide, the initial demand reduces the dis-
infectant concentration significantly, the rate of dis-
infection will decrease.
When disinfectants are compared with respect to
contact times, the disinfectant-contact time interaction
(observation 6) can be evaluated. In Figure 4 chlorine
and chlorine dioxide are compared by contact time at
10 mg/l dose. This figure indicates a different time-
dependent mode of action for chlorine versus chlorine
dioxide. Chlorine dioxide affects a major portion of
bacteria destruction at shorter times, while chlorine
requires longer contact times to effect the same degree
of disinfection.
The disinfectant dose-contact time interaction
(observation 7) is a combination of the disinfectant-
dose interaction and the disinfectant-contact time
interaction and indicates that the rate of bacterial
destruction is dependent on both the concentration of
the disinfectant present and contact time.
A final analysis of variance was computed to assess
the effect of residual on log [survival ratio]. In this
ANOVA, disinfectant dose grouping variable was
replaced by residual concentration as a co-variate.
Also, experiment grouping variable was replaced by
total filterable residue, pH and COD, as co-variates
in an attempt to isolate the experimental variation.
The results of this ANOVA are reported in Table 13.
As seen from this ANOVA, residual concentration
has a significant effect on log [survival ratio]. Also,
total filterable residue has a significant effect on log
[survival ratio]. This relationship is such that as total
filterable residue increases, log [survival ratio] in-
creases. This observation supports the finding that
solids protect bacteria from attack by bactericidal
agents.
Conclusions
The conclusions concerning the comparative
bactericidal action of chlorine and chlorine dioxide
derived from the the statistical analysis are as
follows:
1. Both chlorine and chlorine dioxide give de-
creased survival ratios when dose or contact
time is increased.
2. Although some variations exist, chlorine and
chlorine dioxide give essentially the same
survival ratios when compared on a mass
dose basis at 30 minute contact time.
80
-------
CHLORINE DIOXIDE
TABLE 13. COMPARISON OF CHLORINE AND CHLORINE DIOXIDE AS BACTERICIDES
WITH RESIDUAL CONCENTRATION AS CO-VARIATE
Variable
Disinfectant
Contact Time
'Log [N (t)/N (0)] = Log [survival ratio].
Grouping Variable Information
Number
In Group
Values
CIO2, CI2
5, 15, 30 min
ANOVA — Log [N(I)/N (0)] Dependent Variable
Disinfectant
Contact Time
Residual Concentration
Disinfectant-Contact Time
Interaction
Disinfectant-Residual
Interaction
Total Filtrable Solids
PH
COD
Degrees
of Freedom
1
2
1
2
1
1
1
1
Sum of Squares
4.8584
37.3407
242.8966
6.4239
8.8487
15.2176
0.8702
0.0156
F-Value
12.02
46.19
600.88
7.95
21.89
37.65
2.15
0.04
Probability
F-Exceeded
0.0007
0.0001
0.0001
0.0005
0.0001
0.0001
0.1447
0.8448
3. Chlorine dioxide is a more rapid disinfecting
agent, effecting greater bacterial destruction
than chlorine at the shorter contact times.
4. Comparing chlorine and chlorine dioxide on
a residual basis, chlorine dioxide effects the
same survival ratio as chlorine but with a
much lower residual concentration.
DISINFECTION MODEL FOR CHLORINE AND
CHLORINE DIOXIDE
To assimilate the information of the previous sec-
tions into a useful tool for the design of chlorine
dioxide disinfection facilities for wastewater, a simple
model was devised which is applicable to both chlorine
and chlorine dioxide disinfection. The original form
of this model was proposed by Gard (5) to describe the
inactivation of viruses by chemical agents. This model
has been modified by other researchers, and used to
describe disinfection by chlorine (14). Gard found that
the relationship between time and virus inactivation
describes a linear function on a log-log plot. He also
argued that toxicant concentration and time had the
same relation to inactivation, and therefore could be
used interchangeably as the product ct, where c is
concentration at time, t. Gard also recognized that
this model was not limited to the polio virus-formalde-
hyde system which he had studied.
The model which follows is a simplification of
Gard's original model. When survival ratio is plotted
versus the residual-time product on a log-log plot, a
straight line reiationsnip results. Survival ratio was
calculated as previously, and residual-time is the
product of the amperometric back-titration residual
measured at time t, and the sampling time, 5, 15, or 30
minutes. A linear regression of log [survival ratio] on
log [residual-time] yields an equation of the form:
N(t)/N(0) =
where
N(t)/N(0)
RT
b
k
survival ratio
residual-time product, (mg-min/1)"1
lag coefficient (mg-min/1)
velocity coefficient
The coefficient b is a relative measure of the lag period
between dosing and the onset of bacterial destruction;
the larger the value of b, the shorter the lag time. The
coefficient k, is a measure of the rate of kill; the larger
the absolute value of k, the faster the kill. Figures 5
and 6 show the relationship between survival ratio and
residual-time for chlorine and chlorine dioxide. The
equations given are the least-squares line of best fit,
and r is the correlation coefficient.
\
Both curves are plotted in Figure 7 to allow a com-
parison of the disinfectants. The rates of coliform kill
are essentially the same for both disinfectants (k values).
The major difference between chlorine and chlorine
dioxide is in the lag coefficient, b. This implies that for
81
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
I
-6 F
10 10
RESIDUAL -- TIME IN MG -MIN, L
Figure 5. Reduction of Coliform Bacteria by Chlorine
10' 10'
RESIDUAL -- TIME IN MG-MIN/L
Figure 6. Reduction of Coliform Bacteria by Chlorine Dioxide
r
i r
95% CONFIDENCE INTERVAL
FOR EXPONENT
. CHLORINE DIOXIDE -2.70 TO 3 10
\CHLORINE -2.99 TO -3.31
CHLORINE DIOXIDE
N(T)/N(0) = [.64(RT)]"
r = 0 86
CHLORINE
N(T)/N(0)-[ 17(RT)r315
r^O 92
10' 10"
RESIDUAL--TIME IN MG-MIN/L
Figure 7. Comparison of Coliform Reduction
equal coliform reduction, a higher residual-time
product is required by chlorine than by chlorine di-
oxide.
Although these results are from only one waste-
water, Palo Alto Water Pollution Control Plant sec-
ondary effluent," they agree very well with chlorine
disinfection results reported by other researchers on a
variety of wastewaters.
White (14) reports a b value of 0.23 and a k value
of -3 when primary effluent is disinfected with chlo-
rine. Selleck and Collins (4) report b values of 0.25 for
a stirred batch reactor and 4.2 for a plug flow reactor
and k values of -3 for both in the disinfection of pri-
mary effluent with chlorine. In both White, and Selleck
and Collins the chlorine residual concentration is not
the residual as measured at time t, but the residual
measured after the immediate chlorine demand has
been satisfied. Using the residual concentration
measured in this way, as opposed to residual concen-
tration measured at time t, will increase the lag coef-
ficient b. Also, both residual concentration and lag
coefficient are functions of wastewater composition
and reactor design and comparisons should be made
with this in mind.
The most interesting comparison of the work re-
ported here with the work of others is the agreement
of the velocity coefficient k. The work reported here
measured only total coliform bacteria while in the
work of Selleck and Collins, both total and fecal coli-
form organisms were measured. This comparison indi-
cates that the rate of kill will not be a function of the
composition of natural populations of coliform bac-
teria.
Application of Disinfection Model for Design
To be a useful design tool, a laboratory model de-
rived from batch disinfection studies must have verifi-
cation from continuous disinfection of wastewater in
situations analogous to those found in actual use. In
evaluating the model presented here, the basic prin-
ciples of good disinfection practice must be adhered
to: namely, effective rapid mixing of the disinfectant
at the point of dose and close approximation to plug
flow in the contact zone. If these conditions are satis-
fied, then the results of batch disinfection studies can
be used with confidence in designing disinfection
facilities for full-scale operations.
The attractiveness of the proposed model as a design
tool rests in its simplicity. It is only necessary to esti-
mate the coliform concentration of the waste stream
and calculate the required survival ratio to meet the
effluent requirements. Then if the equations given in
Figure 7 are solved for RT, it only remains to choose a ;
contact time and corresponding residual concentra-
tion. There is, however, an important aspect of the
residual disinfectant concentration that must be con-
sidered when applying this model to full-scale opera-
tions. The laboratory residual concentrations were
82
-------
CHLORINE DIOXIDE
determined at the specified contact time, 5, 15 or 30
minutes. Most installations, however, monitor and
control the disinfectant concentration in the waste
stream some short distance downstream of the injec-
tion and mixing point but before the stream enters
the contact chamber. The measurement of disinfectant
at this point in the process does not correspond to the
residual measurement upon which the model is based.
It does, however, measure disinfectant concentration
after the initial rapid oxidant demand of the waste-
water discussed previously (see Fig. 3). In the case of
chlorine, the die-away curve after this initial demand
has a relatively flat slope, so that severe deviation from
the chlorine model is not expected. In the case of chlo-
rine dioxide, however, the slope is not nearly so flat
(see Fig. 3) so that measurement of the chlorine di-
oxide residual at the entrance to the contact chamber
may substantially overestimate the effective residual
concentration. The method of residual measurement
must also be taken under consideration when usinj
this model. Most commercial units available utilize
continuous flow amperometric determination. As
long as these units are calibrated using the amper-
ometric back-titration procedure at pH 7 for chlorine
dioxide, design assumpiions based on the proposed
model will be met. If other than amperometric residual
measurement techniques are used, the design pro-
cedure must be modified accordingly.
Another aspect of the proposed model that requires
verification is its applicability to other wastewater.
Although chlorine results reported here correlate
very well with the chlorine results of others, it re-
mains to be seen if the chlorine dioxide model will
adequately predict the disinfection of other secon-
dary effluents and other effluents of varying quality.
In addition, the economic implications of the pro-
posed model need to be considered. It is obvious from
Figure 7 that a shorter contact time is required for dis-
infection when chlorine dioxide is used as opposed to
chlorine at the same residual concentration. There-
fore, a smaller contact chamber, or possibly no contact
chamber at all, would be required and a savings in
capital costs could be realized. However, this savings
may be offset by the higher chemical costs of chlorine
dioxide (on a mass basis) and a higher chlorine dioxide
dose required than that required for chlorine to pro-
duce the same residual concentration.
Verification of the proposed model as a design tool
entails pilot plant studies (presently in the planning
stage) in which the disinfectant is generated and dosed
continuously over a period sufficiently long to appr^-
imate quasi-steady-state behavior. In addition to
confirming the validity of the model as a design ap-
proach, the pilot-plant study will provide independent
data on the yield and purity of the chlorine dioxide
generated by alternative processes.
The pilot plant will be assembled at the Palo Alto
Water Pollution Control Plant. This facility is a non-
nitrifying, activated-sludge plant with a capacity of 35
mgd. The laboratory-scale studies from which the
model was formulated were conducted with this sec-
ondary effluent. The pilot plant will be a 10 gpm facility
with both chlorine dioxide and chlorine feed systems.
This will allow a direct comparison of chlorine dioxide
and chlorine.
The feed systems for both the chlorine dioxide and
chlorine will be commercially available units or scaled-
down versions of these units. Ideally, the chlorine
dioxide generator will provide chlorine dioxide free
of chlorine and chlorite (i.e., the acid activation of
chlorite method of generation). It remains to be seen
if units utilizing this process are available for treat-
ment of only 10 gpm, or if design and manufacture of a
special unit for pilot-plant studies will be required.
Manufacturers of chlorine dioxide generators have in-
dicated that yiejd and purity may suffer when attempt-
ing to scale down units designed for much higher flows.
Within the constraints imposed by small-scale opera-
tion, information on yields and purity of the chlorine
dioxide will be obtained that will be useful in evaluat-
ing alternative generation processes.
The contact chambers for pilot-plant studies will be
plug flow units. Two types of reactors approximating
plug flow conditions can be used: 1) a baffled rectang-
ular tank; 2) a closed conduit. If possible, both types
of reactors will be constructed and compared. The
rectangular tank contactor will have removable baffles
so that the length-to-width ratio may be varied to
determine optimum contactor configuration. The
contactors will provide 30-minute contact time, and
have sample ports placed at 15-minute detention inter-
vals along the reactor. Because initial mixing is an
important factor in rate of aisimection, a hign turbu-
lence in-line mixer or batch mixer will be used immedi-
ately following chemical addition.
Influent concentration of chlorine dioxide to the
contactors will be analyzed on a continuous basis. The
feed rate will be controlled by the output signal of the
continuous monitor. Chlorine concentration will also
be monitored and controlled by this method, when
chlorine is being fed. Commercial units are available
for this type of control for chlorine only. However,
83
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
these units are based on amperometric measurements
and should be readily adaptable to chlorine dioxide
measurements.
The pilot plant work described above will provide
the information needed for any modification of the
proposed model. In addition, data will be gathered
on the yield of chlorine dioxide from several genera-
tion processes in order to provide guidelines for the
design and economic analysis of full-scale disinfection
facilities utilizing chlorine dioxide.
SUMMARY
What has been reported in this paper are the results
of comparative chlorine and chlorine dioxide disinfec-
tion studies. These studies were carried out in munici-
pal secondary effluent and the organisms used as a
measure of disinfection were the naturally present
population of coliform bacteria. Every effort was
made to simulate an actual disinfection process on a
laboratory scale. The results of these experiments were
developed into a simple disinfection model. Subject
to full-scale verification this model can serve as a
guideline in the design of disinfection facilities utiliz-
ing chlorine dioxide. The rapid destruction of bac-
teria and virus (14) by chlorine dioxide indicate that
chlorine dioxide is a very promising candidate to re-
place chlorine in wastewater disinfection. The eco-
nomic evaluation of chlorine dioxide in the disinfec-
tion of municipal wastewater is continuing under this
EPA research grant.
ACKNOWLEDGEMENT
This paper reports on findings from the EPA Re-
search Grant R-805426, "Feasibility of Using Chlorine
Dioxide in the Disinfection of Municipal Wastewater."
REFERENCES
1. Benarde, M.A., B.M. Israel, V.P. Olivieri, and M.L.
Granstrom. 1965. Appl. Microbiol. 13(5):776.
2. Benarde, M.A., B.M. Israel, V.P. Olivieri, and M.L.
Granstrom. 1967. Appl. Microbiol. 15(2):257.
3. Berg, J., E.M. Aieta, P.V. Roberts, unpublished data.
4. Collins. H. F., and R. E. Selleck. Process Kinetics of Waste-
water Chlorination, Sanitary Engineering Research Lab-
oratory Report. University of California, Berkeley, August
1973.
5. Card. S. 1957. Chemical Inactivation of Viruses. Ciba Founda-
tion Symposium. Nature of Viruses, pp. 123-146.
6. Miltner, R. J. "Measurement of Chlorine Dioxide and Related
Products," Paper given at AWWA Workshop, San Diego.
December 1976.
7. Palin. A. 1967. Jour. In.tt. Water Engr. 21:537.
8. Palin, A. 1974. Jour. Insi. Water Enf>r. 28:139.
9. Ridenour, G. M., and R. S. Ingols. 1947. JAWWA. 39:561.
10. Ridenour, G. M..and E. H. Armburster. 1949. JA WWA. 41:537.
II. Rook. J. J. May 1977. Envir. Sci. & Tech. 11:478-482.
12. Standard Methods of Water and Wastewater Examination.
14th edition, 1975.
13. Walters. Gary E. 1976. "Chlorine Dioxide and Chlorine: Com-
parative Disinfection," M. S. Thesis, Johns Hopkins
University.
14. White, G. C. 1972. Handbook of Chlorination. Van Nostrand-
Reinhold.
DISCUSSION
MR. MECKES: Mr. Aieta and Mr. Berg will now
entertain questions from the floor.
DR. LONGLEY: I would like to ask a question of
the first speaker . .. in fact, I have a couple of questions.
You mentioned the membrane filter technique as being
preferable to the MPN. 1 am curious whether you are
using the one-step procedure or the two-step enrich-
ment procedure.
MR. BERG: For the membrane filter technique,
we used a direct one-step technique. We did not enrich.
DR. LONGLEY: At the University of Texas, San
Antonio, we are doing very similar work, and it is on-
going, comparing chlorine and chlorine dioxide. In
general, the results that you discussed here and pre-
sented here today correspond very closely to what we
are seeing also, particularly in relation to the coli-
phage. The big difference is, though, that on the fecal
coliform, we are finding on a mass basis of applied
chlorine or chlorine dioxide a much more rapid fecal
coliform kill, and a much greater fecal coliform kill
with chlorine dioxide than with chlorine. We are using
fecal coliform ... we are not looking at total coliform
and we are using the same dose ui disinfectant as you
are, that is, 5 mg/1. These results will be presented in
about a week and a half at WPCF in Anaheim.
QUESTION: Are we making the ClOj the same
wav. that is. sodium chlorite with hydrochloric acid?
DR. OLIVltMl: I teel like my thunder has been
stolen. Unfortunately, Dr. Longley has reported on
his studies, and I would like to just indicate that we
agree with the virus data that you have. Apparently, the
bacteriophage are very nice little systems with which.
to evaluate disinfectants. In any event, we find the
same thing. In pilot plant studies at 10 gpm with sec-
ondary trickling filter effluent, we find that the chlo-
84
-------
CHLORINE DIOXIDE
rine dioxide performs much, much better than chlo-
rine, with several orders of magnitude difference
between the inactivation of total coliform done by the
most probable number procedure, compared with
chlorine. Do you have any response or any reasoning
why the coliform kill in your studies is so much lower
than many of the other studies reported in the litera-
ture? This, by the way, is an extension of Walter's work
that was done at Hopkins.
MR. BERG: By using both counting methods, we
achieved comparable results, typically two to three log-
reductions at a 5 mg/1 dose, and five minute contact
time.
DR. OLIVIER!: I am talking about differences be-
tween chlorine and chlorine dioxide of 2 to 3 mg/1.
The overall kill is about five log reduction, bringing
the level of total coliforms well below a hundred.
MR. BERG: Okay. I am substantiating our results
in that, by both counting methods, we did get two to
three log reductions, and that was consistent through-
out the study. Why we never exceeded that inefficient
log kill, compared with what we see in the literature, I
tried to address in the second part of the Paper. The
only response I have to that is, we compared pure
cultures in native populations.
DR. OLIVIER): These are natural populations,
and 1 think in Karl's study they are also natural popu-
lations. In the studies of O'Brien and Gier, using a
combined wastewater overflow, they are also natural
populations of coliforms.
MR. AIETA: And you are showing a two or three
log reduction . . .
DR. OLIVIER): Difference between chlorine and
chlorine dioxide ... an overall four and a half to fivf
log reduction, with comparable dosages.
MR. AIETA: Is this on a lab scale?
DR. OLIVIER): This is pilot plant, 10 gpm.
MR. AIETA: I do not have any explanation at all.
We have seen other reports in the literature. All we can
say is, on Palo Alto's secondary effluent in our labor-
atory reactor, we are showing chlorine and chlorine
dioxide doing about the same job, as far as coliforms
go. We have not gotten to our pilot plant work yet, so
the situation may change there. 1 really do not have
any insight into the difference.
DR. RICE, Jacobs Engineering: You had said
that you are going to be looking at the effects of C1O2
synthesized other ways. Have you decided which ways
you are going to be synthesizing C1O2?
MR. AIETA: The units that are available to us uti-
lize, essentially, the same reaction that we generate
our chlorine dioxide with here. The only difference is
that we strip the chlorine dioxide out of solution, so
that we are sure that we have a pure chlorine dioxide
solution to deal with. I know in the chlorine/chlorite
procedure, unless it is controlled very carefully, you
can have your stream contaminated with chlorine. I
call it contaminated, because what you are looking
for is pure chlorine dioxide. The kind of things we
would be looking at are what effects the generation
procedure, in a natural disinfection scheme, might
have on disinfection.
MR. PETER DeSTEFANO, Riddick and Associ-
ates: I understand that chlorine dioxide is available
commercially in a stabilized form in a five percent
solution, as active C1O5. I was wondering if you looked
into that at all?
MR. AIETA: No, we have not.
MR. DeSTEFANO:Could you give us just a hint
of what the costs of ClOj are, as compared to chlorine?
MR. AIETA: I really do not know, but 1 think there
may be some people in this audience who can give you
a hand on that.
DR. SUSSMAN, Olln Corporation: On the ques-
tion of stabilized chlorine dioxide, be careful. The
solution to the best of our knowledge is a solution of
sodium chlorite. If you have worked at all with chlo-
rine dioxide, you will know that it is colored . . . if you
get even a very dilute solution, you have a yellow solu-
tion. If you look at the stabilized chlorites that are
available, they are colorless.
MR. FLUEGGE, Carborundum Company:Mr.
Berg, in looking at your virus work, you, evidently,
were using poliovirus 1 seed as your virus, as opposed
to naturally occurring virus. Do you have any indica-
tion of the background virus levels in the water?
MR. BERG: Zero time samples were taken before
inoculation of the poliovirus and no recoveries were
made. In fact, that is one of the reasons we went with
the phage as an in situ virus, because poliovirus was
very hard to recover.
MR. FLUEGGE: What were the levels of the
poliovirus 1 seed . . . what concentrations?
MR. BERG: At about 106/100 ml.
85
-------
11.
EFFECT OF PARTICULATES ON INACTIVATION OF ENTEROVIRUSES
IN WATER BY CHLORINE DIOXIDE
F. A. O. Brigano, P. V. Scarp/no, S. Cronier and M. L. Zink
Department of Civil & Environmental Engineering
The University of Cincinnati
Cincinnati, Ohio 45221
ABSTRACT
The ability of suspended matter and viral aggregation to affect disin-
fection efficiency assumes importance in wastewater treatment. Reduced
reactivity of chlorine dioxide (CIO^) to form carcinogenic compounds is
known, but information is needed about the disinfecting abilitv of ClO-^
as affected by particulates and viral aggregates. In laboratory studies at
5° C and pH 7, poliovirus I preparations containing mostly viral aggre-
gates took 2.7 times longer to inactivate with €10% than single state virus
preparations containing 93% single and 7% clumped viruses, as determined
by electron microscopy. The disinfection efficiency ofC/O2 with unassoci-
ated poliovirus 1 singles increased as the temperature increased from 5
to 15 to 25° C at pH 7. However, the disinfection efficiency of CIO2 with
bentonite-adsorbedpoliovirus I singles decreased with increasing temper-
ature relative to the efficiencies obtained with unassociated poliovirus I
singles. Protection from inactivation due to the bentonite was found to
be 11.4% for turbidities of ^ SNTU's (Nephelometric Turbidity Units)
and 24.8% for > 5 S 17 NTU's. Poliovirus in association with BGM
(Buffalo Green Monkey) tissue culture cells show no trend towards cell-
ular protection at pH 7.0 in a turbidity range of 1.10 to 2.00 NTU and5° C,
or in 1.12 to 3.10 NTU at 25° C. Thus, temperature and the amount of
turbidity affect the rate of inactivation of bentonite-adsorbed poliovirus,
while there is no affect seen with the turbidity levels of cellular-associated
virus examined.
INTRODUCTION . . , , . . ... , ...
water without loss of infectivity has been well docu-
It has been amply demonstrated in recent years that mented (Schaub et al., 1974 (21); Schaub and Sagik,
the majority of viruses in the natural environment are 1975 (20)). It has been also demonstrated that animal
associated with solids and are not in a "free" state viruses (e.g., poliovirus 1 and coxsackievirus A9) are
(Berg, 1973) (2). Wastewater influent, effluent, and more resistant to inactivation by free and combined
chlorinated effluent samples have been found to have forms of chlorine and chlorine dioxide than the path-
16.1 to 100% of their total virus content associated with ogenic indicator of fecal pollution, Escherichia coli
solids (Wellingspf al., 1975) (32). The ability of viruses (Scarpino et al., 1972 (18); Scarpino et al.. 1974(19);
to associate with clays and other suspended solids in Cronier et al., 1978 (4)). This lact demonstrates the
86
-------
CHLORINE DIOXIDE
need to increase our understanding of the effectiveness
of the viral disinfection process in wastewater. The
capability of coliforms and fecal coliforms to regrow
in chlorinated wastewater effluents (Shuval et a!.,
1973) (24) may necessitate the presence of a disinfec-
tant residual in these effluents. Such residuals should
be non-toxic to aquatic life. The importance of under-
standing the disinfection kinetics of viruses in effluents
from wastewater treatment is clear, since these efflu-
ents discharge into receiving waters which may serve
either as a source of potable water, an area in which
shellfish may be harvested, or as a recreational re-
source (Murphy, 1977) (13). Emphasis must also be
placed on the products of the chemical and physical
interactions and reactions between the disinfectant
and wastewater effluent, since these products may
ultimately affect the environment and its life forms.
Such products of disinfection are chlorinated hydro-
carbons. These compounds are formed when disinfec-
tants such as chlorine, bromine, and iodine (Mills,
1978) (11) react with precursor organics in water. Tri-
halomethanes, which are carcinogens and suspect
:arcinogens are, however, produced to a lesser extent
when wastewater effluent is chlorinated because the
high ammonia content of wastewater leads to the
formation of the less reactive chloramine species
(Mills, 1978). The precursor organics necessary fortri-
halomethane formation have been demonstrated to be
the natural humic material present in virtually all
source waters (Stevens et a/., 1976 (30), Rook, 1976
(17)). Therefore, disinfectant residuals present in dis-
charged, treated wastewater effluent are of concern
since they may react with these precursors in receiving
water environment to form trihalomethanes and other
substances potentially adverse to humans and aquatic
life. Disinfectants such as ozone (Falk and Moyer,
1978) (7) and chlorine dioxide (Miltner, 1976) (12) are
not known to form trihalomethanes. Their oxidation
products created by reaction with organics in water
and wastewater are in low amounts and of still largely
unknown toxicological significance. Although disin-
fectant residuals in wastewater effluents are not avail-
able with ozone, they are available with chlorine di-
oxide. As seen in Figure 1, chlorine dioxide (ClOg). is
just as efficient as chlorine when comparison is made
between 99% poliovirus 1 inactivation at 15°C by
hypochlorous acid at pH 6 to that of chlorine dioxide
at pH 7 (Siders et ai, 1973 (25); Scarpino et ai, 1974
(19); Esposito et at., 1974 (6); Cronier etai, 1978(4)).
The present study was undertaken to evaluate the
effects of particulates on inactivation of enteroviruses
by C1O2.
1000
KX>
10
I
t 1.0
0.1
— NHCI2
(pH4.5)
.Oil 1_L
.01
I I I I II IN I I I I I lilt I I I I Mill 1 1 I I MM
0.1 1.0 10 100 1000
Time in Minutes for 99% Inactivation
FIGURE 1. Comparison of the Relative Inactivation of
Poliovirus 1 by Hypochlorous Acid, Hypochlorite Ion,
Monochloramine, Dichloramine, and Chlorine Dioxide at
15°C at Different pH Values.
MATERIALS AND METHODS
Chlorine Dioxide Generation: A stock solution of
chlorine dioxide was prepared by the generation of
chlorine dioxide gas which occurs according to the
following reaction:
2C1OJ+ K2S2O8-*-2ClO2 \ + 2K++2SO4.= •
The chlorine dioxide gas was the result of the reaction
of sodium chlorite (NaClO2) and potassium persulfate
(K2S2O8) solutions. The evolved gas was swept from
solution by purified nitrogen gas and passed through a
column of sodium chlorite to absorb any chlorite gas
or volatilized hypochlorous acid that might also be
present. Any sodium chlorite dust was retained in an
empty vessel prior to collection of the gas in deionized
distilled water held at 5°C. The stock solution was
prepared prior to experimentation and had a concen-
tration of 500 to 1,000 mg/1 after 15 to 20 minutes of
generation.
Chlorine Dioxide Measurement: The concentra-
tion of chlorine dioxide was determined by the DPD
(diethyl-p-phenylene diamine) method of Palin (1974)
(14) and as also set forth in the 14th Edition of Stan-
dard Methods for the Examination of Water and
Wastewater (Wlb) (27).
-------
PROGRESS fN WASTEWATER DISINFECTION TECHNOLOGY
Virus Quantitation and Characterization:^^
virus inocula were quantitated and characterized by
electron microscopy using the kinetic attachment pro-
cedure as described by Sharp (1974) (22), using a JEM
100 B electron microscope (JEOL, Ltd., Tokyo).
Poliovirus Preparation: The poliovirus 1 used in
these studies were prepared by two different methods.
Poliovirus for both methods were first grown in mono-
layers of Buffalo Green Monkey kidney cells (BGM).
The chlorine dioxide-demand-free poliovirus prep-
aration, which contained numerous viral aggregates
and cellular debris similar to the presumed state of the
virus in the natural environment, was prepared by
repetitive freezing (to -70°C) and thawing with differ-
ential and high speed centrifugation. The 10K plaque
forming units (pfu)/ml virus was stored at-70°C until
used.
Purified chlorine dioxide-demand-free poliovirus
with the virions virtually in a "single" state was pre-
pared by freon extraction of poliovirus from intact
BGM cells followed by density gradient centrifugation
in sucrose. Fractions were collected and examined by
electron microscopy. All relevant fractions were
pooled and stored at 4°C without any attempt to re-
move the sucrose (Floyd et ai, 1976) (8). The resultant
10'° pfu/ ml virus preparation contained no cellular
debris and consisted of greater than 93% single virions
as determined by electron microscopy.
Bentonite Preparation: The chlorine dioxide-
demand-free bentonite suspensions with particulates
of approximately 2 urn or less used in the turbidity
studies was prepared using the method of Stagg et a/.,
(1977) (26). Turbidity was measured in Nephelometric
Turbidity Units (NTU) using a Hach 2100 A Turbidi-
meter.
Poliovirus-Bentonite Complex Preparation: The
virus-bentonite suspensions were prepared by allow-
ing the "singles" poliovirus preparation to associate
with the bentonite for 1 hour at constant mixing in
demand-free phosphate buffer followed by low speed
centrifugation to prepare only bentonite adsorbed-
virus for use in experiments.
Cell Associated-Poliovirus 1 Preparation: Cell
asscciated-poliovirus 1 was prepared from BGM
Monolayers which had been infected 11 hours earlier.
The poliovirus-containing cells were harvested and
washed repeatedly in demand-free phosphate buffer
prior to experimentation. Despite these repeated
washings this preparation still maintained a consider-
able chlorine dioxide demand. Analysis of the cell
associated-poliovirus aggregation state by the Multi-
Poisson Distribution Model (Wei and Chang, 1975)
(31) indicates that the preparation contained a high
population of single virions.
Experimental Procedures: The chlorine dioxide
disinfection studies were performed using the kinetic
(stirred beaker) apparatus (Scarpino, 1972 (18) and
1974 (19)) and the dynamic (flowing stream-rapid mix)
apparatus (Sharp et ai., 1976) (23).
The kinetic apparatus (Figure 2) consisted of C1O2
test and control solutions in stainless steel beakers
held at the desired temperature in a water bath, and
stirred throughout by glass stirring rods connected to
an overhead stirring device. Virus was then inoculated
into the solutions, and at specific contact time inver-
vals removed and placed in a chlorine dioxide neutral-
izing thiosulfate solution.
FIGURE 2. Kinetic (Stirred Beaker) Apparatus.
The dynamic apparatus (Figure 3) is ideally suited
for yielding kinetic data of short reaction times of less
than one minute. The apparatus provides rapid injec-
tion and mixing of the virus inoculum into the flowing
stream of buffer-containing chlorine dioxide solution.
After injection of the inoculum, samples were rapidly
withdrawn at the syringe sampling points at appropri-
ate time intervals and immediately mixed with sodium
thiosulfate, which was contained within each syringe
to stop the reactions. The time of transit of the moving
stream of water from the injection point to a sampling
CONTINUOUS ROW APPARATUS
for
VIRUS DISINFECTION STUDIES
FIGURE 3. Dynamic (Flowing Stream-Rapid Mix) Apparatus.
88
-------
CHLORINE DIOXIDE
point was determined by the flow rate and the distance
traversed. Turbulent mixing of the inoculum within
the stream was assured by maintaining a Reynolds
number greater than 3000.
Poliovirus Assay Procedure: The surviving polio-
virus from these studies were assayed by the plaque
forming system in BGM cells (Dulbecco and Vogt,
1954(5), Hsuingand Melnick, 1951 (10)). The associ-
ation of bentonite with poliovirus has previously been
shown not to affect the plaque forming ability of the
virus (Schaub and Sagik, 1975) (20).
RESULTS AND CONCLUSIONS
A typical disinfection survival curve using the dy-
namic apparatus and poliovirus 1 singles in contact
with 12.1 mg/1 chlorine dioxide at 5°C and pH 7
(Figure 4) is obtained by plotting the log of the percent
survival against the time of exposure to chlorine diox-
ide. One hundred percent survival times are derived
from the controls. The 99% inactivation points or the
1% survival points were then extrapolated from the
survival curves to give the time necessary for 99% in-
activation of the viruses. These 1% survival points for
each of the C1C>2 levels used were then replotted on"
100
10
1.0
0.1
0.01
10 15
Seconds
20
FIGURE 4. Inactivation of Poliovirus 1 Singles
Using the Dynamic Apparatus with 12.1 mg/l Chlorine
Dioxide at 5°C at pH 7.
log-log paper to show C1O2 concentration versus the
previously determined 99% inactivation times. These
new concentration-time plots (Figure 5) were used to
compare the rates of inactivation, with the curves
closer to the left hand corner of the graph representing
the faster reaction. Here we see the effects of temper-
ature on poliovirus inactivation with C1O2 at pH 7.
The 99% inactivation rates of the poliovirus singles
are increasing from 5 to 15 to 25° C.
1001-
o
u
10
pH7
25T
10
Seconds
100
FIGURE 5. Concentration-Time Relationship for 99% In-
activation of Poliovirus 1 Singles Using the Dynamic
Apparatus at 5, 15, and 25°C.
The effect of viral aggregation, which is similar to
the state of the virus found in the natural environment,
was evaluated at 5° C and pH 7 using the kinetic appar-
atus (Figure 6). The 93% singles poliovirus prepara-
tion reached 99% inactivation at a rate of 2.7 times
faster than the aggregated virus preparation.
Studies on the effect of inorganic turbidity on the
disinfection of bentonite adsorbed-poliovirus 1 singles
were conducted at pH 7 using the dynamic apparatus.
The turbidity levels used in these studies fell within the
range of expected turbidity levels of final secondary
effluent from a well run wastewater treatment plant.
The results (Figure 7) for 99% inactivation of free-
unassociated poliovirus is represented in the left-hand
portion of each graph and the 99% inactivation points
for the poliovirus-bentonite complex on the right side
for each temperature; the solid line represents the 99%
inactivation curve for the free poliovirus. At 5° C there
is no trend towards protection from chlorine dioxide
inactivation offered by the bentonite to the attached
poliovirus with turbidities ranging from 1.14 to 16.5
89
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
"NTU. At 15°C (Figure 8) a slight trend towards pro-
tection from C1O2 inactivation by the bentonite devel-
ops, with turbidity levels that vary from 0.64 to 12.24
NTU. At 25° C (Figure 9) a definite trend toward pro-
tection by bentonite is evident where the 99% inactiva-
tion points of adsorbed poliovirus at turbidities of
from 1.35 to 2.29 NTU are found clustered around the
free poliovirus inactivation curve. As the turbidity
increases from 3.22 to 14.10 NTU, the 99% inactiva-
tion points are found further from the free poliovirus
inactivation curve, thereby indicating a definite pro-
tective effect at higher levels of turbidity at 25°C.
These data indicate that the amount of protection
from C1O2 inactivation by the bentonite to the ad-
sorbed poliovirus increases with increasing temper-
ature and turbidity.
10r
1.0
o.i
5'C
pH7
Poliovirus 1
« Aggregates
• Singles
25
100
Seconds
1000
FIGURE 6. Concentration-Time Relationship for 99% In-
activation of Poliovirus 1 Using the Kinetic Apparatus
Comparing Single Virions to Aggregated Virions.
When these data are placed into turbidity groupings
of < 5 NTU's and > 5 < 17 NTU and graphed along
with the unassociated "singles" poliovirus data onto a
plot of rate of inactivation versus the product of con-
centration times temperature a linear relationship re-
sults (Figure 10). From this relationship it is found that
bentonite-adsorbed virus of the ^ 5 NTU group is
protected to the extent of 11.4% (or 88.6% unpro-
tected) when comparison is made to the unassociated
poliovirus. The > 5 < 17 NTU group is protected to
24.8% (or 75.2% unprotected).
Thermodynamic analysis of the data (Table 1) yields
mean values for the QIQ> Energy of Activation (Ea),
Enthalpy of Activation (AH), and Entropy of Activa-
tion (AS) for the unassociated poliovirus, the ^ 5
NTU group, and the > 5 < 17 NTU group. These
values are consistant with those obtained in protein
denaturation (Adams, 1959 (1); Cliver el a/., 1978(3);
Ginoza el a/., 1964 (9); Pollard, 1953 (15); Prokop el
a/., 1970 (16); Steam, 1949 (28);and Woese, 1960(33)).
Thus, the information suggests that the mechanism of
inactivation for the poliovirus by the chlorine dioxide
is by means of protein denaturation. The values for the
>5 £ 17 NTU group indicate that the bentonite is
interacting with the chlorine dioxide, thus, inhibiting
the chlorine dioxide's ability to react with the virus and
cause its inactivation.
TABLE 1. Q10, ENERGY OF ACTIVATION (Ea), ENTHALPY
OF ACTIVATION (AH), AND ENTROPY OF ACTIVATION
(AS) FOR THE UNASSOCIATED POLIOVIRUS, AND THE
<5 AND > 5<17 NTU GROUPS OF THE POLIO-
BENTONITE COMPLEXES.
GROUP Q
• Unassociated
Poliovirus
S5NTU 2.35
Polio- Bentonite
>5<17NTU ]56
Polio- Bentonite
cal/mole
12353
14020
7289
cal/mole
11781
13448
6717
* S
cal/mole-deg
98
104
80
E - Energy of Activation
AH = Enthalpy of Activation
*S : Entropy of Activation
Inactivation studies of BGM cell-associated polio-
virus by chlorine dioxide were conducted at pH 7 using
the kinetic apparatus. The results (Figure 11) for 99%
Fre« Poliovirus
SC pH7
Polio-B*ntonit«
MTU
• 1.14
• 2.20
• 3.09
° 4.15
* 6,25
o 6.70
• 12.38
o 16.50
100 1
Seconds
100
FIGURE 7. Concentration-Time Relationship for 99% In-
activation of Poliovirus 1 Singles and Bentonite Adsorbed-
Poliovirus 1 Singles Using the Dynamic Apparatus at
5°C at pH 7.
90
-------
CHLORINE DIOXIDE
inactivation of the cell-associated virus are represented
here at 5 and 25° C. The solid line at each temperature
represents the 99% inactivation curve without data
points for the free-unassociated poliovirus (0.15 to
0.20 NTU). Cell-associated turbidities at 5° C of 1.10
to 2.00 NTU and at 25° C of 1.13 to 3.10 NTU show no
trend towards protection at these turbidity levels.
100
$
u
15'C pH7
Fr«e Poliovirus
Polio- Bcnlonit*
NTU
• 0.64
• 1.65
o 103
o Z48
• 193
• £18
»1Z25
10
100 1
S.condt
100
FIGURE 8. Concentration-Time Relationship for 99% In-
activation of Poliovirus 1 Singles and Bentonite Adsorbed-
Poliovirus 1 Singles Using the Dynamic Apparatus at
15°C at pH 7.
g
u
25'C PH7
Fr«* Poliovirut
Polk>-B5 £17 NTU Polio-Bentonite
10 20 30 40
10"* "C moles/I
50
FIGURE 10. Relationship of the Product of Concentratior.
of CIO2 times the Temperature and the Rate of Inactiva-
tion of the Unassociated-Poliovirus, and the <5 and
5-< 17 NTU Groups of the Poliovirus-Bentonite Complexes.
1.0
o.i
M'C
NTU
•1.13
•133
• 3.10
100
WOO
no
woo
FIGURE 11. Concentration-Time Relationship for 99% In-
activation of BGM Cell-Associated Poliovirus at Various
Turbidities Compared to the 99% Inactivation Curve for'
Unassociated Poliovirus Using the Kinetic Apparatus at
5° and 25°C at pH 7.
91
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
that the cellular material surrounding the viruses is
readily oxidized by the chlorine dioxide, thus exposing
them to the disinfectant as if single virions alone were
present. Finally, a correlation exists between benton-
ite protection of poliovirus during disinfection at
increasing temperatures and increasing turbidities,
i.e., as the temperature and bentonite turbidity in-
creases, the disinfection efficiency decreases for the
bentonite adsorbed poliovirus.
ACKNOWLEDGEMENTS
This study was supported by the U.S. Environ-
mental Protection Agency, Grant R-804418. Our
thanks to Mrs. June Schuck for typing the manuscript.
REFERENCES
I. Adams, M. H. 1959. Bacteriophages. Interscience Publishers,
New York.
2. Berg, G. 1973. "Reassessment of the Virus Problem in Sewage
and in Surface and Renovated Waters." in Progress in
Water Technology, S. H. Jenkins (ed.), Pergamon Press,
New York.
3. Cliver, D. O. and R. J. Salo. 1978. "Indicators of Viruses in
Foods Preserved by Heat" in Indicators of Viruses in Water
and Food, G. Berg (ed.), Ann Arbor Science, Ann Arbor,
Michigan, pp 329-354.
4. Cronier, S., P. V. Scarpino, and M. L. Zink. 1978. "Chlorine
Dioxide Destruction of Viruses and Bacteria in Water."
in Water Chlorination Environmental Impact and Health
Effects Volume 2, R. L. Jolley, H. Gorchev and D. H.
Hamilton (eds.), Ann Arbor Science Publishers, Ann
Arbor, Michigan, p. 651.
5. Dulbecco, R and M. Vogt. 1954. Plaque Formation and Isola-
tion of Pure Lines with Poliomyelitis Virus. J. Exp. Med.
99: 167-182.
6. Esposito, P., P. V. Scarpino, S. L. Chang and G. Berg. 1974.
"Destruction by Dichloramine of Viruses and Bacteria in
Water," Abstracts of the Annual Meeting -1974, American
Society for Microbiology, G-99.
7. Falk, H. L. and J. E. Moyer. 1978. "Ozone as a Disinfectant of
Water" in Ozone-Chlorine Dioxide Oxidation Products
of Organic Materials. R. G. Rice and J. A. Cotruvo (eds.),
Ozone Press International, Cleveland, Ohio, p. 38.
8. Floyd, R., J. D. Johnson and D. G. Sharp. 1976. The Inactiva-
tion by Bromine of Single Poliovirus Particles in Water.
Applied Microbiology, 31: 298.
9. Ginoza, W. C, C. J. Hoelle, K. B. Vessey, and C. Carmack.
1964. Mechanism of Inactivation of Single-Stranded Virus
Nucleic Acids by Heat. Nature (London) 203: 606.
10. Hsuing. G. D. and J. L. Melnick. 1955. Plaque Formation with
Poliomyelitis, Coxsackie, and Orphan (Echo) Virus in
Bottle Cultures of Monkey Epithelial Cells. Virology,
7:533-535.
11. Mills, J. F. 1978. "Competitive Oxidation and Halogenation
Reactions in the Disinfection of Wastewater" in Ozone-
Chlorine Dioxide Oxidation Products of Organic Mate-
rials, R G. Rice and J. A. Cotruvo (eds.), Ozone Press
International, Cleveland, Ohio, p. 400.
12. Miltner, R. J. 1976. The Effect of Chlorine Dioxide on Trihalo-
methanes in Drinking Water. Master of Science Thesis,
University of Cincinnati.
13. Murphy, K. L. 1977. The Current Role of Wastewater Disin-
fection. Water and Pollution Control, //J:13-16.
14. Palin, A. T. 1974. Analytical Control of Water Disinfection with
Special Reference to Differential DPD Methods for Chlo-
rine, Chlorine Dioxide, Bromine, Iodine and Ozone.
Journal Institute of Water Engineers, 2#: 139-154.
15. Pollard, E. C. 1953. The Physics of Viruses. Academic Press,
Inc. New York, New York.
16. Prokop, A. and A. E. Humphrey. 1970. "Kinetics of Disin-
fection" in Disinfection. M. A. Bernarde (ed.), Marcel
Dekker, Inc. New York.
17. Rook, J. J. 1976. Haloforms in Drinking Water. Journal
AWWA, 6#:168.
18. Scarpino, P. V.,G. Berg, S. L. Chang, D. Dahlingand M. Lucas.
1972. A Comparative Study of the Inactivation of Viruses
in Water by Chlorine. Water Research 6:959-965.
19. Scarpino, P. V., M. Lucas, D. R. Dahling, G. Berg, and S. L.
Chang. 1974. "Effectiveness of Hypochlorous Acid and
Hypochlorite Ion in Destruction of Viruses and Bacteria"
in Chemistry of Water Supply, Treatment, and Distribu-
tion, A. J. Rubin (ed.), Ann Arbor Science Publishers, Ann
Arbor, Michigan, p. 359.
20. Schaub, S. A. and B. P. Sagik. 1975. Association of Entero-
viruses with Natural and Artificially Introduced Colloidal
Solids in Water and Infectivity of Solids-Associated
Virions. Applied Microbiology, .30:212-222.
21. Schaub, S. A., C. A. Sorber and G. W. Taylor. 1974. "The
Association of Enteric Viruses with Natural Turbidity in
the Aquatic Environment." in Virus Survival in Water
and Wastewater Systems. J. F. Malina and B. P. Sagik
(eds.). Center for Research in Water Resources, The Uni-
versity of Texas at Austin.
22. Sharp, D. G. 1974. in Proc. 32nd. Annual Meeting of the
Electron Microscopy Society of America (C. J. Arceneaux,
ed.) Claitor's Publishing Division, Baton Rouge, La.
23. Sharp, D. G., R. Floyd, and J. D. Johnson. 1976. Initial Fast
Reaction of Bromine on Reovirus in Turbulent Flowing
Water. Applied and Environmental Microbiology, 31:
173-181.
24. Shuval, H. L, J. Cohen and R. Kolodney. 1973. Regrowth of
Coliforms and Fecal Coliforms in Chlorinated Waste-
water Effluent. Water Research, 7:537-546.
25. Siders, D. L., P. V. Scarpino, M. Lucas, G. Berg and S. L.
Chang. 1973. "Destruction of Viruses and Bacteria in
Water by Monochloramine," Abstracts of the Annual
Meeting - 1973, American Society for Microbiology, E-27.
26. Stagg, C. H., C. Wallis and C. H. Ward. 1977. Inactivation of
Clay-Associated Bacteriophage MS-2 by Chlorine. Ap-
plied and Environmental Microbiology, J3:385-391.
27. Standard Methods for the Examination of Water and Waste-
water- 14th Ed. 1976. APHA, AWWA, WPCF, Washington,
D. C.
92
-------
CHLORINE DIOXIDE
28. Steam, A. E. 1949. Kinetics of Biological Reactions with
Special Reference to Enzymic Processes. Advances in
Enzymology. IX:25-14.
29. Stevens, A. A., D. R. Seeger, and C. J. Slocum. 1978. "Prod-
ucts of Chlorine Dioxide Treatment of Organic Materials
in Water" in Ozone-Chlorine Dioxide Oxidation Products
of Organic Materials, R. G. Rice and J. A. Cotruvo (eds.),
Ozone Press International, Cleveland, Ohio, p. 383.
30. Stevens, A. A., C. J. Slocum, D. R. Seeger, and G. G. Robeck.
1976. Chlorination of Organics in Drinking Water. Journal
AWWA, <5S:615.
31. Wei, J. H. and S. L. Chang. 1975. "A Multi-Poisson Distribu-
tion Model for Treating Disinfection Data" in Disinfec-
tion Water and Wastewater, J. D. Johnson (ed.), Ann
Arbor Science Publishers, Ann Arbor, Michigan, p. 11.
32. Wellings, F. M., A. L. Lewis and C. W. Mountain. 1965. Dem-
onstration of Solids-Associated Virus in Wastewater and
Sludge. Applied & Environmental Microbiology, 31:354-
358.
33. Woese, C. 1960. Thermal Inactivation oj Animal Viruses. Ann.
N. Y. Acad. Sci. &J:741.
DISCUSSION
DR. JOHNSON: I was interested in the effect of
your picking the 99 percent point for your analysis of
data. I noticed the one curve that you showed on the
log survival versus time was very linear. Was this typi-
cal of your log survival versus time curves?
MR. BRIGANO: Yes.
DR. JOHNSON: They were all linear, is that right?
MR. BRIGANO: Right. These are fitted to a first
order reaction, as under the model equation of van't
Hoff - Arrhenius.
DR. JOHNSON: Was this true of not only the free
unassociated virus, but also the virus which was cell
associated and the virus which was associated with the
bentonite? You got linear log survival time?
MR. BRIGANO: Yes.
DR. JOHNSON: Surprising. That does not agree
with data that we have gotten on aggregation of virus
to virus, where we were looking at aggregated virus,
rather than virus which was ... as you are, in your
study, associated with cells or bentonite. Our virus-
virus aggregation data showed nonlinear log survival
versus time plots.
MR. BRIGANO: Log/log plots?
DR. JOHNSON: No, log survival versus time.
MR. BRIGANO: These are log/log plots. The log
of the concentration versus the log of the time for 99
percent inactivation.
DR. JOHNSON: Yes, okay. Now, that is my point.
If you took, say the 99.9 percent level, or the 99.99
percent level, ... if your plots of log survival versus
time . . . your basic disinfection kinetic data . . . if they
are not linear ... if it does not obey Chick's Law, then
where you pick your survival level very much affects
the ...
MR. BRIGANO: Right. I thought you were talk-
ing about the van't Hoff plots.
DR. JOHNSON: No, I am talking about the log
survival versus time . . . the kinetic plots. Were they
linear?
MR. BRIGANO: For the dynamic apparatus, they
were. For the kinetic apparatus, they were not. The
dynamic apparatus is the second apparatus I showed
with the tubing.
DR. JOHNSON: Okay. That is the Sharp, John-
son, Floyd apparatus. With that kinetic data, log sur-
vival versus time, if you pick a two log survival level,
you get a very different kind of effect, if the survival
kinetics are not linear, than if you pick a four or five
log survival. My question is, were all of your kinetic
data linear?
MR. BRIGANO: Yes, for that apparatus, they
were all linear, as shown in that survival curve that I
showed in the beginning.
DR. JOHNSON: That is in contrast to our data
where they were nonlinear if they were aggregated
virus to virus.
MR. BRIGANO: Those were not done using that
apparatus for the aggregated virus. The aggregated
virus experiments were done in the kinetic apparatus
. . . the stirred beaker apparatus.
DR. JOHNSON:Oh, and those log survivals were
not linear?
MR. BRIGANO: Right.
DR. JOHNSON: Now, that does make a point,
because, if you look at, say your two log survival, that
is more associated, because you are taking the initial
early part of the log survival curve.
MR. BRIGANO: No, the singles virus was also
done. For that graph, where I showed the difference
between aggregated virus and the singles virus, they
both were done using the kinetic, stirred beaker appar-
atus. They were not done using the dynamic apparatus.
They both were done with all the conditions the same.
DR. KARL LONGLEY: I want to commend you,
first of all, on a very interesting paper. Secondly, I
want to suggest that it departs quite a bit from what we
see at the end of a wastewater treatment plant, where
a good part of the viruses are actually occluded in the
solids and are not adsorbed, with possibly some ex-
posure externally to the solution containing the waste-
water disinfectant. I think, in a real world situation,
93
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
you may find that the controlling rate parameter is
really the rate of diffusion of the disinfectant through
the occluding organic material, and the competition of
that organic material for the disinfectant, rather than
other parameters that might be considered.
MR. BRIGANO: We understand that. Thank you.
94
-------
12.
PANEL DISCUSSION OF CHLORINE DIOXIDE DISINFECTION
Participants of Chlorine Dioxide Panel Discussion
Mark C. Mcckes, Moderator
U.S. EPA, MERL-Cincinnati
I. James Berg
Stanford University
2. Marco Aieta
Stanford University
3. Paul Roberts
• Stanford University
4. HA. Brigano
University of Cincinnati
5. P. V. Scarpino
University of Cincinnati
DR. ROSEN: I would like to make these com-
ments, with respect to the Stanford work. Paul Roberts
and I were recently at a conference where Dr. Kott was
talking about some work that was done at Israel. He
found very similar results in an ozone/chlorine com-
parison to the work that you have reported with chlo-
rine dioxide and chlorine, with respect to the greater
relative efficiency of ozone with respect to viruses, in
secondary effluents. Now, he did not present data. His
comment was from the floor, with respect to someone
else's work, so that I cannot comment on the absolute
magnitudes. In relative terms, he found a similar rela-
tionship to that which you have reported. The other
thing that he reported, that I found of interest, that
Clifford White is always banging the drum about, is
that, really, these problems tend to be sometimes
solved by, not one or the other, but by combinations.
He reported that, in this case, chlorine and ozone were
more efficient (this is, again, relative efficiency, and
this was not defined because it was a comment from
the floor), in terms of finding some sort of synergism,
with respect to both bacteria and viruses on secondary
effluents, independent of the method of addition, or
the order of addition . . . that is, together, chlorine
first, ozone first. Also, with respect to whether it be
bacteria or viruses, he found they could get an effect,
which was, apparently, synergistic. Perhaps we ought
to be talking, in some of these cases, about combined
disinfection, maybe something for viruses and some-
thing for bacteria, and ending up with something that
overall is the most cost effective.
MR. AIETA: Just a comment to go along with that.
Hopefully, in our pilot study, we will be able to take a
look at the combination of chlorine and chlorine di-
oxide. Our pilot studies are still pretty much in the
planning stage, so we do not know exactly how far we
will go with that.
DR. PAUL ROBERTS: .I think that may be a good
practical suggestion, especially, in that it has been
found that combinations of chlorine and chlorine di-
oxide have the effect of inhibiting, at least haloform
formation, and it may be that a good economic com-
bination of chlorine and chlorine dioxide avoids the
formation of some of the organics that we would like
not to see, and at the same time achieves economic kill
of both bacteria and virus.
MR. WHITE: Isn't it because ozone has a different
killing mechanism than chlorine?
DR. ROBERTS: I cannot answer that definitively,
but it is a good . . .
MR. WHITE: Insofar as bacteria versus virus?
DR. ROBERTS: That is a reasonable hypothesis,
I think. Would someone like to comment on that?
DR. ROSEN: I did ask if the work that was done in
Israel was with seeded viruses or natural coliphages,
and it was with both, and results were similar. That
question always arises when you talk about seeding
and protection, and some of the work that is coming
out of Pat Scarpino's group. Combining disinfectants
is something that I am for, and something that I do not
think we ought to jump into, maybe, until we under-
stand some of the basics about the individual disin-
fectants. With respect to the mechanism, I do not
know if the mechanisms are the same, but certainly,
I think that one indication that the mechanisms are
different (there are a couple, in the literature) is, of
course, the relative rates of inactivation. The second
has to do with the MPN/MH work that has been done,
and the fact that, apparently, when that comparison
is made with chlorine, there is a different result than
when that comparison is made with ozone. That is,
95
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
ozone seems to make no statistical difference with
respect to survivors that may regrow, whereas chlo-
rine does.
DR. RICE, Jacobs Engineering: This is a differ-
ent point. Three points, actually, one is a comment,
one is a partial answer, and one is a question. The com-
ment is, so that the audience will not think that the
only person hawking a new book here is George White:
Those of you who have read ClOj literature have run
up against a little monograph on the oxides of Cl,
published in 1959 by Dunod, written by Dr. Willy J.
Masschelein, who is the technical director of the Tailfer
Belgium Drinking Water Works. About a year and a
half ago, 1 met Dr. Masschelein, impressed him with
a growing amount of interest in the U.S.A. on C1O2,
and talked him into updating his monograph and do-
ing it in English. I have been the editor on it. It is in
proof now, and it will be published by Ann Arbor
Science Publishers before the end of the year. So, if
anybody is interested in that, good. The partial answer,
which I have, to the question that was raised a while
ago about cost . . . what does it cost to use ClOjover
Cl2? That depends pretty much on the way you make
ClOa, and I am not going to get into numbers, but that
is why my earlier questions, of how is C1O2 made, are
very important. Again, hawking a book, this is an EPA
final report. Some of you may know, I have been in-
volved with a nonprofit organization in Washington,
which has been granted by drinking water people here
to survey the use of ClOjand ozone in drinking water
treatment plants. That final report is in print, and in
fact, 1 think it has been shipped from the printer, but
it does have a fairly good state of the art review of C1O2
in drinking water.
Now, back to the cost situation. The people in Euro-
pean drinking water plants that use ClOj normally use
it as a terminal disinfectant. They normally make it
using excess Ci2 over NaClOj. In some circumstances,
it is made by using HC1 and the sodium chlorite. These
people say that it costs them about three times the
amount to use C1O2, than it does Cl2. They use very
low levels in drinking water. I think, in Zurich it is
0.15 ppm. So, I do not think that any of this reads
directly on the wastewater end, but those two com-
ments.
Now, my question is, in the drinking water side,
there is some concern by EPA for the toxicity of C1O2
ion for people. And, I have not heard any of this yet, in
the wastewater discussion. I wonder if there is any
testing going on, concerning the toxicity to organisms
of the C1C>2 ion, in the wastewater area?
MR. AIETA: It is my understanding that there is
work happening now in the EPA. 1 think they are feed-
ing chlorite to rats. I have not seen anything come out
of that work as of yet. As a corollary to that question,
there will be a need to be able to measure chlorite in
the presence of chlorine dioxide, at some point in time.
That is no easy task.
MR. RICK TRAVER, EPA, R & D Storm and
Combined Sewer Section: Mr. Berg and Mr. Aieta,
in your generation of chlorine dioxide, in your labora-
tory, how was that accomplished, and to what quanti-
ties did you generate it? Was it on a continual basis?
Did you store it?
MR. BERG: We generated batches of chlorine di-
oxide, on the day that we did an experiment, so they
were fresh on that day. We took a sodium chlorite 750
ml solution and, I believe it was a ten percent solution
of that salt, to which we added a ten percent solution of
hydrochloric acid. The CIO 2 that would be evolved
from that reaction was then passed through a sodium
chlorite salt tower. We have been told that that should
remove any chlorine present, although it has also been
shown that it is most likely that no chlorine is gener-
ated in that process. Then, the gas that was passed
through the salt tower was trapped and iced down in
deionized water. We then measured the concentrations
of that, and, typically, we could get around 500 mg/1.
MR. TRAVER: Was there any storage of the chlo-
rine dioxide, over a period of time, between day to
day?
MR. BERG: We stored it, and if we used it later for
other studies, we noticed a drop in concentration. We
could probably lose up to a 100 mg/1 over a few days.
MR. TRAVER: O'Brien and Gier's research was
referenced earlier on the utilization of chlorine and
chlorine dioxide singly and also combining them,
which I have not heard here, either, in tandem in com-
bined sewer overflows, research funded by my particu-
lar program. This was one of the problems that was
typified in bench scale study, that of storage of the
chlorine dioxide solution and the decay of the concen-
tration over a period of time. One of the suggestions
was that better results could be found on a continuous
full-scale manufactured solution. That was my ques-
tion ... a couple of points, though. At the present time,
in addition to the research of O'Brien and Gier, a full-
scale evaluation is being conducted at the East Chicago
Sanitary District in East Chicago, Indiana, coupling a
160 million gallon lagoon, which is receiving secondary
plant effluent, and during wet weather 75,000 GPM of
combined sewage. A full-scale 16 million gallon chlo-
96
-------
CHLORINE DIOXIDE
rine/chlorine dioxide disinfection system is being in-
stalled and will be dosing at 8 mg/1 of chlorine, two
minutes detention time, followed by 2 mg/1 of chlorine
dioxide, fifteen seconds detention time, attempting to
achieve a four to five log decrease, as a result of the
O'Brien and Gier study. We will be optimizing that
situation, in Chicago, once that goes on line, we hope,
later this winter and early spring.
MR. WHITE: Is that the greater sanitary district
of Chicago?
MR. TRAVER: No, sir, East Chicago Sanitary
District.
MR. WHITE: The answer to the question on cost
will be found in the Culp-Wesner-Culp Report, and it
is EPA #600/2-78-182, August, 1978. It compares
chlorine, onsight generation of hypochlorite, onsight
generation of chlorine/chlorine dioxide, chlorine
cylinder storage, and ozone. They split it up into three:
process energy KW per year, dollars construction
cost, and total cost per year, in dollars (thanks to Bob
Baker, who took all the tables and put them on log/ log
paper; I just received them from him, in case this might
be asked at the previous panel). This is on potable
water, but anybody that wants to get a comparison of
costs, can use this very recent report. Now, my ques-
tion. Mr. Aieta, I am slightly' confused about youi
analytical technique, and I wonder if you would jus
explain a few things. You talked about iodometric; yoi
talked about amperometric; you talked about differ-
ent pH's, and you talked about DPD. Now, how do
you characterize iodometric, because amperometric
is really iodometric back titration, but it is an ampero-
metric end point. So, what is iodometric? And, what
were the different pH's? This is what confused me.
MR. AIETA: What I call the iodometric technique
is, essentially, the starch iodide. And, that technique
for chlorine dioxide is done at a pH below two. The pH
characteristically runs about 1.8, 1.85. The ampero-
metric back titration that I spoke of is done at a pH of
7, and at that pH, the chlorine dioxide only goes
through a one electron change. The starch iodide
method at pH 1.8 should take the chlorine dioxide all
the way to chloride. So, that by comparing those two,
what I attempted to show was that, if the results did
, not agree, I would not be anywhere. But, since the
results agreed, it showed that, yes, the iodometric is
measuring chlorine dioxide, five electron . . . five
valance changes. And, the amperometric is measuring
one valance change for chlorine dioxide.
MR. WHITE: Okay, so in effect, the amperometric
is really the same as the forward titration for free chlo-
rine, right?
MR. AIETA: No, it ...
MR. WHITE: I am sorry, monochloramine . . .
MR. AIETA: Right, exactly.
DR. OLIVIER!: I would like to direct a question to
Mr. Berg and Mr. Aieta. Since we brought up the
stock solutions, did you people ever compare the levels
of chlorine dioxide, using a direct spectrophotometric
measurement at 357 nm with your starch iodide or
amperometric measurements? In addition, is thereany
idea of the percent yield that you got in the prepara-
tion of your stock solution?
MR. AIETA: I do not think we ever calculated per-
cent yield on our stock solutions. We never verified
our chlorine dioxide spectrophotometrically.
DR. OLIVIERI: I realize that it is not going to
work very well in the sewage, but it would certainly
work well on vour stock solutions.
MR. AIETA: We felt that the starch iodide method
was sufficient, and gave us good enough results, so we
really did not go any further than that.
DR. OLIVIERI: But, the starch iodide at the low
pH does not distinguish between chlorite and chlorine
dioxide.
MR. AIETA: Well, in our generation procedure, we
feel we have a pure chlorine dioxide solution.
DR. OLIVIERI: The point is, can you demonstrate
that?
MR. AIETA: Not with what we have done now, no.
DR. GILBERT GORDON, Miami University:
We have done work on chlorine dioxide for many
years. The direct spectrophotometric analysis is, in-
deed, a challenge to almost anybody doing it. We
developed a shrinking bottle to do that ... a bottle
which has no air volume above the liquid. It is really
an inverted syringe, so that as you withdraw solution
to analyze it, you are pumping solution out. You are
not changing the relative volume above it. Thus, you
do not have to worry about the Cl 2 dispersed both in
the gas phase and the liquid. Under these conditions,
we can reproducibly do several parts per thousand
spectrophotometric, and what you are calling iodo-
metric titration. However, we do two added titrations.
If you do that iodometric at pH about 8 or 7.8, there is
a single electron change. The ClO2goes to ClOj, a
single electron change. You can then acidify that solu-
tion even more and do either a separate or a continued
titration, getting the five electron change. And, if you
compare those, one can find internal consistency, and
check out this point of the purity of the solution . . .
the amount of ClOj, or, conversely, if one has stored
these at high pH, the problem of chlorate and chlorite
formation. In answer to the general question of stabil-
97
-------
PROGRESS //V WASTEWATER DISINFECTION TECHNOLOGY
ity, we have stored solutions for weeks in dark bottles
in refrigerators at zero degrees, and find no decompo-
sition. However, the analytical techniques are some-
what trying, and it does take one quite a long time to
develop a method where you do not lose gas upon
analysis, and are not convinced that you are getting
something else. However, on the iodide analysis do
you find all your results showed amperometric titra-
tions were higher and iodide were lower as you drew
them on the slides, if I recall. I think the numbers you
had showed the iodometric values were slightly lower.
The amperometric were slightly higher tilers, in the
titration. I would suggest that may be due to iodate
formation, in that, chlorine dioxide reacts with iodide
and forms iodate. That iodate does not react very
rapidly with iodide to reform iodine. That, quantita-
tively, can account for the low tiler values. We can
show those Iwo come out exactly Ihe same, if you do
very slow addition, very dilule solution.
MR. BERG: I have a question for you. Have you
found, in generaling chlorine dioxide by the method
we described, thai there was any level of chlorite ion
in the stock, or was thai produced over a lime?
DR. GORDON: I Ihink your point is well laken,
lhat if you go through a chlorite purificalion lower,
you do end up wilh extremely pure chlorine dioxide
solulions. We have had solulions which showed no
lesls for any other foreign material, other than ihe
buffers, in which we may have collected Ihem. There
have been several queslions aboul synergislic effecls.
1 Ihink lhal in Ihe chlorine/chlorine dioxide sludies,
I would like lo suggest a series of what I think are very
important polential synergistic tests. If one is adding
chlorine dioxide after Ihe chlorine, there is one kind
of information. If, however, one injects sodium chlo-
rite, or a solution conlaining a chlorile ion, I suggesl
lhal one mighl, indeed, observe considerably differenl
resulls. I am a kineticisl, basically, and we have sludied
ihese solulions for years. If we look at pure phenol,
and look at Ihe oxidation wilh chlorine, or Ihe oxida-
lion of phenol wilh chlorine dioxide, or mixlures of
Ihose, where we are now injecling phenol wilh a mix-
ture of ihe Iwo, one sees considerably differenl resulls.
We have been able to demonslrale, I Ihink unambigu-
ously on pure solulions initially containing whatever
organic malerial we want lo oxidize, and sodium chlo-
rile, lo which we add chlorine, lhal inlermediales are
formed, and Ihose inlermediales are differenl in Ihe
way in which Ihey read wilh phenols or whalever Ihe
organic malerials are. If one is looking for synergistic
effects, I think one can easily demonstrale il in lhat
case. In the olher case, it may be more difficult, be-
cause only the chlorite product, after chlorine dioxide
reaction, is available for reaction. That synergistic
effecl may be much further downstream. But, the
other one is very easy to demonstrale.
MR. WHITE: The French tell me thai ihe ozone,
following ihe applicalion of chlorine dioxide, puls
some of the chlorite form back lo chlorine dioxide . . .
if you can believe Ihem.
DR. GORDON: The polenlials are righl for oxi-
dalion.
MR. WHITE: They also say, lhal ozone can reverse
the chloride that comes out from, maybe, the prechlo-
rinalion, back to HOC1.
DR. WAITE, Northwestern: I would like to ask
if any of Ihe chlorine dioxide people have looked al Ihe
reaclion of chlorine dioxide wilh any differenl lypes
of organics? Realizing Ihe demand of chlorine dioxide
is prelly high, in highly conlaminaled walers, am-
monia probably is nol going lo affecl il. Working wilh
a different type of disinfeclanl, that I am not allowed
lo menlion here, we see that, depending on the type of
organic in the water, we see a lot of different reactions,
even though the TOC or COD may be the same. So, I
would like to ask if any of you have looked at different
types of wastewaters, or waters containing a basic
TOC bul differenl makeup?
DR. ROBERTS: We have just begun to do so, and
I really do not have any results to present. However, I '
might comment on anolher finding al the European
conference on drinking water treatment, thai some of
us attended. In Germany and Holland, the use of total
organic chlorine, as a collective parameter, is becom-
ing more widespread. Some measuremenls have been
made, particularly, comparisons of formation of total
chlorinated organics by chlorine, compared to chlo-
rine dioxide. And, the few data lhal were shown indi-
caled aboul one-fourlh as much total organic chlorine
formed by chlorine dioxide, as for chlorine. In other
words, it seems as if Ihose organic chlorine com-
pounds, measured by this method, are formed lo a
lesser extent using chlorine dioxide than chlorine,
which ought lo be favorable for chlorine dioxide.
DR. SCARPING: I just wanted lo commenl on
some other aspect of Ihe research by Mr. Berg and
Mr. Aiela. Nowhere in the papers have you mentioned
the pH of thesyslem, or Ihe chemical paramelersof Ihe
syslem. I may have missed lhal. Whal were Ihe pH's in
your system?
MR. AIETA: Our pH's were lypical of pH's of sec-
ondary effluent, i.e., 7 to about 6.5. We measured
COD, ammonia nitrogen, alkalinity, and there were
a couple of others . . . total suspended solids . . .
98
-------
CHLORINE DIOXIDE
DR. SCARPINO: I was especially interested in
pH, because the work of Cronier in our lab establishes
the fact that as you raise your pH from 4.5 to 9, you in-
crease the efficiency of your chlorine dioxide. This is
consistent with other work in the field, but not con-
sistent with the work of O'Brien and Gier, I believe,
or Smith from Syracuse.
MR. AIETA: Is that relative to chlorine?
DR. SCARPINO: Chlorine dioxide, yes. I was
wondering if you noticed the same thing, as you in-
creased your pH. Did you have a more efficient system?
MR. AIETA: We did run a couple of statistical anal-
yses on our wastewater parameters. The only correla-
tion we could find between log-kill and our parameters
was with the solids. As the solids increased, the kill
decreased. We found no correlation with pH or any of
the other parameters.
DR. SCARPINO: Your pH range was quite nar-'
row.
MR. VENOSA: I would just like to make one com-
ment before adjourning. I am getting the impression
here that there is just as much disagreement in the
chlorine dioxide field now as there was with ozone
several years ago. Everybody is getting controversial.
They are getting results that are different from each
other. So, it suggests that we need to do a lot more
research in this field on chlorine dioxide in both mea-
surement as well as disinfection efficiency.
DR. SCARPINO: I can agree with you on the mea-
surement of chlorine dioxide. 1 think one of the prob-
lems we had, initially, was what method to select for
chlorine dioxide. We finally ended up calling Dr. Palen
in England to find out exactly how he did it.
99
-------
SECTION 4. ULTRAVIOLET LIGHT
13.
UTILITY OF UV "DISINFECTION" OF SECONDARY EFFLUENT
Harold W. Wolf, * Albert C. Petrasek, Jr., ** and Steven E. Esmond***
"Head, Environmental Engineering, Texas ABM University, College Station, Texas 77843
*Director, Environmental Engineering & Science, Albert H. Halff Associates, Inc., Dallas, Texas 75219
***Esmond-Haner, Inc., Odessa, Texas 79761
INTRODUCTION
The criterion utilized in these studies to determine
the adequacy of the "disinfection" level achieved was
a fecal coliform content of 200/ 100ml. Fecal coliforms
comprise only a part of the total coliform group, and
since some members of the coliform group are ac-
knowledged pathogens, we believe that some caution
and understanding should accompany the use of the
terms "disinfection" or "disinfected". There is also
the matter of the Glossary (2) definition of disinfection
which excludes viruses, thus reinforcing the need to
exercise caution and understanding.
At the onset of this project, the factors that we
thought would be among the most important of the
many confounding variables (other than the absolute
number of fecal coliforms and the ultraviolet dose)
were total suspended solids, turbidity, and probably
the transmissability at 254 nm. Our subsequent exper-
ience taught us that over the range of suspended solids
(5-50 mg/1) and turbidity (0.5-12 Ntu) that we en-
countered, these two quality parameters had relatively
little influence on the results. Transmissability, how-
ever, emerged as quite important, and furthermore
was observed to be a factor of the operation of the
biological treatment facility — a complete-mix acti-
vated sludge system.
This project was conducted over a 16-month period
at the Dallas Water Reclamation Research Center by
personnel from Dallas' Water Utilities Department,
the Civil Engineering Department of Texas A&M
University, and the U.S. Environmental Protection
Agency. Detailed descriptions of the Demonstration
Plant at this Facility have been published (1). Two
ultraviolet (UV) generating units were used (3). First,
a Kelly-Purdy unit originally designed for use in shell-
fish depuration studies (3), and later, a proprietary
unit loaned by Ultraviolet Purification Systems, Inc.,
located in Scarsdale, N.Y. The UV units were applied
directly to final effluent, to mixed-media filtered efflu-
ent, to dual-media filtered effluent, to tertiary clarified
effluent, to chemically clarified effluent, and to ter-
tiary clarified plus dual-media filtered effluent.
Additionally, following the protocol used in previous
work (5), and in full recognition of its limitations,
three virus seeding experiments were conducted
using poliovirus type 1 and F2 coliphage.
An attempt was made by the plant operational staff
to utilize as a baseline of biological operation the
production of a highly nitrified effluent, i.e.Ammonia-
nitrogen (NHs-N) effluent concentrations of <(l mg/1
as N. We failed all too often to achieve this level be-
cause of fragile oxygen transfer systems and a gross
operator failure. Actually, three different oxygen-
transfer systems were used during the study, two pro-
prietary types and one inadequate home-made type.
The operator failure occurred when he left a waste
sludge valve wide open all-night long — in the cold
month of December.
A stringent sampling program was followed in these
studies. The unit processes prior to the UV unit were
sampled on a flow-composited basis and analyzed by
Standard Methods (7) procedures. The UV units were
grab sampled at 11 AM and 4PM — the times when
the diurnal curve for organic content was maximum at
the Dallas facility. Separate samples were taken for
chemical and microbiological analyses. The mem-
brane filter procedure was used with the Kelly-Purdy
unit and the MPN procedure with the Ultraviolet Pur-
ification Systems unit. Eight paired irradiated samples
were run for comparison using both methods. The
100
-------
UL TRA VIOLET LIGHT
mean log results were 2.07 by MF and 2.01 by MPN
showing essentially no difference between the two.
Kelly- Purdy Unit Results
The Kelly-Purdy (K-P) ultraviolet light disinfection
unit (Figure 1) consists of a shallow-tray exposure
chamber 185 cm (6 ft.) long, 92 cm (3 ft.) wide, and 9.2
cm (3-5/8 in.) deep. Flow passes through the tray
underneath 13 30-watt UV lamps (G30T8). The influ-
ent flow was measured with an undulating disc meter,
and water depth was controlled by changing the eleva-
tion of the effluent weir.
10°
10'
101-
Figure 1. Kelly-Purdy UV Unit (lid open) and Ultraviolet
Purification Systems Model EP-50 (on blocks
on the floor)
Five runs were made with the K-P unit ranging from
2-1/2 hrs. long to 24 days. The first three runs were
applied directly to activated sludge effluent, the last
two to a mixed-media filtered effluent. In the first
three runs, flow rate was varied from 0.32 to 1.58 1 / sec
(5-25 gpm) and depth from 2.54 to 5.08 cm (1 to 2 in.).
The microbiological results of the samplings showed
no association with flow and little with depth, but
more importantly, the effluent quality did not achieve
the fecal coliform level sought, 200 fecal coliforms per
100 ml. However, the results were definitely promis-
ing, giving over-all reductions in fecal coliform con-
centrations of 2-3 logs. Figure 2, for example, illus-
trates the results obtained during Run K3. The the-
oretical detention time for this run was 43 seconds and
the theoretical dose 32,000 Mwatt-sec/cm2. Table 1
shows for each run the types of effluent treated, the
time and duration of each run, and the average COD
and NH3-N content of the composited samples. For
run K3, the effluent composited samples averaged a
COD of 43 mg/1 and NH3-N of 1.1 mg/1 which indi-
cates good biological treatment — although not quite
as good as was sought (NHs-N <(1.0 mg/1). COD:
BOD5 ratios for this effluent generally average 2.7:1,
hence, a BODs of about 16 mg/1 is suggested.
10
INFLUENT
EFFLUENT
10
15
20
25
30
October 1974
Figure 2. Fecal coliform data for Run No. K3
One of the difficulties encountered with the K-P unit
was the accumulation of solids in the tray. Hence, for
the last two runs made with that unit, the final effluent
was filtered through a mixed-media filter (anthracite-
sand-garnet, Neptune-Microfloc Co.) prior to UV
irradiation. The microbiological results of Run K4
(Figure 3) show that all samples met the 200/100 ml
fecal coliform limit. For Run K5 the flow rate was in-
creased from 0.96 I/sec to 1.4 I/sec (15.25 to 22.2
gpm) thus decreasing the theoretical detention time to
29 seconds. More importantly, however, it was during
this run that the operation failure occurred. The result
was that only a couple of effluent samples met the fecal
coliform limit.
Ultraviolet Purification Systems Unit Results
It was at this point that the Ultraviolet Purification
Systems (UPS) firm loaned their Model EP-50 unit
(Figure 1). This unit consists of a 53.6 liter (14.2 gal.)
stainless steel cylindrical chamber that houses nine
longitudinally-mounted 40-watt U V lamps. Each lamp
is enclosed by a quartz sleeve and has an individual
ammeter mounted on a control panel. The unit is
101
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
10'
10-
10'
10
10'
10'
INFLUENT
5
10
25
15 20
November 1974
Figure 3. Fecal coliform data for Run No. K4.
equipped with a water quality meter which measures
UV light intensity on a unitless scale. Hence, all UV
intensity measures in this study were made with an IL
• 500 radiometer manufactured by International Light,
Inc., Newburyport, Massachusetts, through a quartz
window on.the side of the tank.
The location of the nine UV lamps with respect to
the system geometry is shown in Figure 4 which also
shows lines of equal intensity calculated for the unit.
Flow enters from above at a right angle to the lamps at
one end, moves parallel to the lamps toward the other
end, and exits at the top at the far end.
The quartz sleeves were cleaned with a solution sup-
plied by the manufacturer. Each run was commenced
with a freshly cleaned unit. The cleaning frequency
required to keep the system operating at peak efficiency
can be expected to vary with the quality of the effluent,
but intervals of two to three weeks seem reasonably
consistent with the data and temperatures observed.
A total of eight runs were made with the UPS unit
varying from 2 days duration to 127 days. All effluent
samples of Run Ul of 13 days duration applied to
straight secondary effluent at a flow of 1.8 I/sec (29
gpm), resulting in a detention time of 29.8 seconds,
met the 200/100 ml fecal coliform criterion (Figure 5).
Runs U2 and U3 applied to tertiary settled effluent and
to tertiary settled dual-media filtered effluent were
even more effective — the latter is shown in Figure 6.
Run U4 was the 2-day run and was short because we
only had a 2-day supply of chemicals. Ferric chloride
was applied in the Densator (Infilco Co.) prior to UV
Run
No.
K1
K2
K3
K4
K5
U1
U2
U3
U4
U5
U6
U7
U8
V1
V2
V3
Type of Effluent
Treated
Secondary effluent
Secondary effluent
Secondary effluent
Mixed-media filtered effluent
Mixed-media filtered effluent
Secondary effluent
Tertiary clarified
Clarified and dual-media filtered
Chemically clarified
Secondary effluent
Dual-media filtered effluent
Dual-media filtered effluent
Secondary effluent
Secondary effluent
Secondary effluent
Secondary effluent
Date
Run
7/26
9/14-10/7
10/8-10/30
11/1-11/25
12/7-12/27
1/23-2/5
2/7-2/17
2/19-3/3
3/6-3/7
3/8-7/14
7/15-8/3
8/8-10/5
10/7-11/30
4/22
5/13
6/26
Length
of Run,
days
2'/2 hrs.
23
23
24
20
13
10
12
2
127
19
58
54
—
—
~
Composite
COO
mg/l
42
42
43
14
92
33
35
42
—
54
107
62
63
46
37
75
Sample
NH3-N
mg/l
1.1
3.2
1.1
0.2
10.1
2.2
1.8
1.5
0.2
7.1
12.8
4.6
5.4
1.8
4.4
10.3
TABLE 1. TYPES OF EFFLUENTS TREATED, TIME AND LENGTH OF RUNS, AND COMPOSITE SAMPLE COD
AND NHj-N MEAN CONCENTRATIONS.
102
-------
ULTRA VIOLET LIGHT
Figure 4. Isointensity patterns for UPS exposure
chamber.
irradiation. Geometric mean fecal coliform values of
the effluent were less than 4.3/100 ml with a maxi-
mum of 33/100 ml.
The shortest UPS run was followed by the longest.
EFFLUENT
KM
25
ITT
Figure
January February
5. Fecal coliform data
for Run No. Ul.
25
February
4
March
Figure 6. Fecal coliform
data for Run No. U3.
Run U5 lasted for 127 days. It was during this run that
the oxygen transfer system commenced to fail and
many heartaches were encountered in trying to limp
along. The fecal coliform limit during this period was
not often achieved.
The proprietary oxygen transfer equipment was re-
moved and a home-made diffused air system substi-
tuted. It was inadequate to achieve the nitrification
desired, but Run U6 using dual-media filtered effluent
was nevertheless conducted. The fecal coliform limit
was achieved about half the time (Figure 7).
i
i
1 15 20 25 | 3
July August
Figure 7. Fecal coliform data for Run No. U6.
In Run U7, new Penberthy oxygen transfer equip-
ment had been installed and the U V system was applied
once again to a dual-media filtered effluent. Most of
the effluent samples met the fecal coliform limit but,
as compared to the earlier runs, the UPS unit during
Run U7 was operating near its design hydraulic capac-
ity of 3.1 I/sec (49 gpm) giving a theoretical contact
time of 17.3 seconds.
• - • ,
The last run, U8, was made at the maximum flow
rate attainable, 3.2 I/sec (51 gpm) which gave a the-
oretical contact time of 16.8 seconds, and using straight
final effluent. Most of the fecal coliform samples ex-
ceeded the limit desired during this run (Figure 8).
Nitrification was still incomplete — NHs-N averaging
5.4 mg/1. The resulting effluent COD of 63 mg/1, al-
though higher than when better nitrification is occur-
103
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
io5
10 15 20
November
Figure 8. Fecal coliform data for Run No. US.
ring, is still at a reasonable level for the usual activated
sludge effluent. For example, applying the COD:
BOD5 ratio of 2.7:1 results in an effluent BOD5 of
about 23 mg/1.
Virus-Seeding Experiments
The three virus runs were each made at four flow
rates which provided exposure periods of 11.4 to 85
seconds (theoretical). The runs were made during the
period of equipment difficulties when effluent COD
values ranged from 37 to 75 mg/1. The highest COD
waters gave the poorest microbiological results, but
little difference was observed between phages and
poliovirus. The phages (y = 1.59x - 4.68, x = log cal-
culated U V dose in u watt-sec/ cm2, y = log reduction)
were possibly a little more resistant than the poliovirus
(y = 1.62x - 4.48), and both viruses, in turn, were more
resistant than the fecal coliforms (y = 1.48x - 3.21).
Figure 9 shows fecal coliform log reductions as a func-
tion of the log of the calculated UV dose. Figure 10
shows the same curve for coliphages. The correlation
coefficients for these curves were fairly low (0.70-0.72);
hence, additional research is warranted. The no effect
dose (determined by setting y = O and solving for x)
ranges from 148 (fecal coliforms) to 871 (coliphages)
u watt-sec/ cm2.
CONCLUSIONS
During the course of these experiments, we once
again observed that high quality effluents (low COD's)
were associated with nitrified operation, a relationship
that is even more marked after an adsorption process
application (6). Our plots of turbidity vs. transmit-
tance, and suspended solids vs. transmittance yielded
shotgun patterns (4). On the other hand, transmittance
and COD correlated at 0.76 and transmittance and
TOC at 0.95.
Both TOC and COD correlated well with effluent
NH3-N, COD at 0.97 and TOC at 0.88 (Figure 11).
The result is the seemingly unlikely correlation of
effluent NH^-N content with transmittance at r = 0.81.
LOG UV DOSE
Figure 9. Fecal coliform reduction vs. UV dose
Additionally, highly nitrified operation results in
lower coliform concentrations. Figure 12 reflects this
observation. The data in Figure 12 compare the aver-
age activated sludge effluent NHs-N concentrations
for each K and U run of Table 1 with the mean log
fecal coliform content entering the U V unit. The latter
value is impacted by the additional unit processes em-
ployed such as filtration or tertiary settling. For ex-
ample, the lowest fecal coliform value on Figure 12
(6.3 x 10-1) was obtained during Run U4 when chemical
clarification was utilized. In spite of these additional
treatments, a relationship is apparent. A correlation
coefficient was not calculated because two points (^)
are less-than values and one point (/^) is a greater-
than value, and also because of the ameliorating effect
104
-------
ULTRA VIOLET LIGHT
of the additional treatments on the coliform values.
Light-dark experiments were performed during the
course of these studies (6). UV irradiated effluent was
directed into two parallel and baffled chlorine-contact
basins (without chlorine addition) providing a deten-
tion period of about an hour. One basin was covered
with black plastic sheet, the other was open. Statistical
evaluations of the data showed significant (at a 95%
'level) regrowth of total coliforms but not of fecal coli-
forms — although the latter also showed higher efflu-
ent concentrations.
LOG UV DOSE
Figure 10. Coliphage reduction vs. UV dose
Clearly, ultraviolet irradiation of secondary efflu-
ents is a viable alternative "disinfection" procedure
for achieving a 200 fecal coliforms/100 ml effluent
criterion. Like ozone, however, UV is quite sensitive
to effluent quality variations and will have to be de-
signed to accommodate the expected variations.
18
16
14
12
10
8
2
0
20
18
16
14
12
10
8
6
4
2
0
NH3-N = .27 COD - 3.
r = 0.97
0 10 20 30 40 50 60
COD, nig/1
NH3-N = 0.96 TOC - 5.3
r = 0.88
I I
12
16
20
TOC, mg/1
Figure 11. Correlations between ammonia-N and COD
and TOC.
E
8
2
oc
O
NH3-N, mg/l
Figure 12. Relationship of Fecal Coliform Concentration
to NH.-N Content of Activated Sludge
Effluent
105
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
REFERENCES
1. Esmond, E. S., and H. W. Wolf. 1973. The status of organics -
the Dallas Water Reclamation Pilot Plant. Proceedings,
Organic Matter in Water Supplies: Occurrence, Significance,
and Control. Univ. of Illinois, Urbana, 111.
2. Glossary Water and Wastewater Control Engineering. 1969.
APHA, ASCE, AWWA and WPCF.
3. Kelly, C. B. Nov. 1961. Disinfection of sea water by ultraviolet
radiation. Am. J. Public Health. 51:1670.
4. McCarthy, J. J. The Influence of Particle Si:e in Oxidation of
Total, Soluahle, and Paniculate Municipal Wastewater
Components by Ozone. Dissertation, Southern Methodist
University. December 5, 1974.
5. Petrasek, A. C., Jr. November 1977. Wastewater Characteriza-
tion and Process Reliability for Potable Wastewater Rec-
lamation. EPA-600/ 2-77-210.
6. Petrasek, A. C., Jr., H. W. Wolf, S. E. Esmond,and D. C. Andrews.
1977. Ultravioret Disinfection of Municipal Wastewater
Effluents. EPA-R-803292.
7. Standard Methods for the Examination of Water and Waste-
water. 13th ed. 1971. American Public Health Association,
Inc., Washington, D.C.
DISCUSSION
MR. DeSTEFANO, Riddick and Associates:
Did you do any correlations between UV transmit-
tance and coliform counts?
DR. WOLF: I will have to check that.
MR. DeSTEFANO: - It seems to me the correlation
between the ammonia and the coliform counts are
more correlation between the transmittance and the
counts rather than the ammonia itself.
DR. WOLF: I will check that. I have the report
with me.
MR. SEVERIN, FMC Corporation, Chicago:
Did you change your flow rate on this unit during any
fecal coliform runs?
DR. WOLF: Yes.
MR. SEVERIN: Did you find a straight line rela-
tion with time or did your curves taper off?
DR. WOLF: I had one transparency I was going to
show in which we maintained a constant flow rate with
the Kelly-Purdy unit and doubled the depth. There
was an increased effectiveness of UV due to the in-
creased exposure time which resulted, on the order of
about one log. We did vary flow, and flow enters into
the U V dose very much by affecting the exposure per-
iod. Does that answer your question?
MR. SEVERIN: The reason I am asking is that I
have done tests with the EP 50 unit, and found that in
primary and filtered secondary effluents there wa.s a
linear relationship between survival and t'/3 and I
am trying to explain this to myself.
DR. WOLF: What kind of factor?
MR. SEVERIN: Time to the one-third.
DR. WOLF: Time to the one-third. I will have to
check that out.
MR. SEVERIN: Also, I would like to agree with
your suspended solids information. I was able to put
all these on a single line, and the indication that you
can use a single line for a log reduction with primary
effluent as well as with dual media filtered effluent
would indicate that suspended solids are not really
interfering in the range up to say 100 mg/1.
DR. WOLF: Good. That one has been worrying
me right along.
MR. FLUEGGE, Carborundum Company: I
have tried to go through some calculations in deter-
mining what dosages of ultraviolet light are to the
wastewater. Looking at your numbers it looks like you
have assumed for one thing only one pass of the pho-
ton, that the photon is not reflected off the walls or
not absorbed by the bulb and then readmitted. Is that
what you have done, or did you try to take that into
account through measurements?
DR. WOLF: When you get into photons, you are
getting out of my field, but we went through two basic
types of calculations. This gave us problems all the
way through. With the Kelly-Purdy unit we could put
the sensing meter on the bottom of the tray, move it
along different locations, integrate it and so forth.
With the U PS unit we could not get inside to do these
things. So we had to do it on a calculated basis, and if
you do it by a calculated basis you have some problems
because you have a slime buildup on the lamps. It is
very slow initially, but once it starts it develops very
rapidly, and you also have the problem on the inside of
the quartz window.
MR. FLUEGGE: You mean between the bulb and
the outer casing.
DR. WOLF: Yes, and so we are relatively insecure
with respect to our measurements. When we did the
Kelly-Purdy test, for example, we did not even have
the UV measuring device at the time, but we went
ahead and did it anyway thinking we could come back
and measure it, that perhaps there would not be that
much deterioration in the bulbs, but when we got into
this study it is not as simple as we thought originally.
The extinction coefficient becomes terribly important,
106
-------
ULTRA VIOLET LIGHT
and so we could not run the extinction coefficient on
the early examples any more, so we assumed an extinc-
tion coefficient for those particular studies.
MR. FLUEGGE: All of your data then would indi-
cate a constant light source intensity. Is that correct?
DR. WOLF: That is correct.
DR. JOHNSON: Your iso-intensity lines on your
diagram there show very low values along the wall.
You did consider reflections off the wall.
DR. WOLF: That is a good question. I believe that
is correct. I will have to check that out.
QUESTION: Question on maintenance. These
units look like they would be maintenance nightmares
on a large scale, cleaning them off, especially if you
could not easily replace the bulbs or get inside these
units. Is that true?
DR. WOLF: There is a lot of proprietary people
here and I will let them answer that question, but one
of the things that we wanted to do in this study was
some cost work with respect to an application in
Dallas, and we came up with a pretty large unit for
application to Dallas. Furthermore, we just did not
have enough information to make a reasonable esti-
mate of cost. So we did not even bother to do it. I also
thought we could store the water and get some off-
peak power rates and then disinfect during the off-
peak power periods, but as I found out the Dallas
Light and Power gives power so cheaply to the Dallas
Water Utility that they would not even entertain off-
peak rates.
QUESTION: Was the smaller unit rated at the
same capacity as the large one?
DR. WOLF: It was higher capacity. Almost double.
MR. ELLNER, Ultraviolet Purification Systems:
Just one point here: the gentleman mentioned the fact
earlier that the two units were rated similarly. I think
Harold will point out that there is a considerable dif-
ference in the UV generators even though you had
almost a similar number of lamps in the Kelly-Purdy.
There are differences in UV sources, and the ratings of
the 30 watt or 40 watt did not refer to U V output but to
electrical input, so you cannot make any correlation.
The two quick points that I wanted to make were:
there is a tendency to try and describe the acceptability
of an effluent based upon visible light determinations,
terms such as color and turbidity which are measure-
ments of visible light. I think Harold mentioned earlier
in his statement that those factors just do not apply to
ultraviolet. You can have situations with very high
visible light transmittance but very low UV transmit-
tance, and you can have the opposite under certain
circumstances. The point that I would like to make is
the combination of all of the factors can be measured
with just one measurement, and that is the UV trans-
mittance at 254 nm on a spectrophotometer. That
would take into consideration all of these variables,
and I was curious, Harold, whether you had some of
those ratings such as UV transmission to correlate.
DR. WOLF: Yes, we do, Sid. They are in this report.
MR. ELLNER: And the only other comment at
this time is that you will agree that the dosages are
theoretical. Would you hang your hat on any of those
dosage numbers that appear in the paper?
DR. WOLF: No, I am rather reluctant.
107
-------
14.
UV DISINFECTION OF SECONDARY EFFLUENT
J. D. Johnson, K. Aldrich, D. E. Francisco, T. Wolff, and M. Elliott
Department of Environmental Sciences and Engineering
School of Public Health
University of North Carolina
Chapel Hill, North Carolina 27514
ABSTRACT
Ultraviolet disinfection studies are being conducted on a pilot scale
using two commercial 230-2851 /min (60-75 gpm) units. Potassium ferri-
oxalate actinometry is used to measure U K dosages. Log reductions of
Escherichia coli in buffered tap water were compared to dosages measured
both by the actinometry method and by multiplying the ultraviolet in-
tensity times the retention time. Actinometry dosages were found to have
the advantage of being directly related to the average depth of fluid
through which the light penetrates and the average intensity within these
multiple lamp units. Photoreactivation in UV treated samples exposed
to 45 minutes of sunlight caused an average 1.4 log recoverv in total coil-
forms at 25° C. Lower photoreactivation was found at a lower temperature
and at a higher UV dose in one of the units. A study of the effects of water
quality parameters and unit design on UV disinfection of filtered and un-
filtered 2° effluent is still under investigation. Log reductions of total and
fecal coliforms are being influenced primarily by the different flow pat-
terns within the units rather than bv variations in the parameters or the
UV dosages applied. Short-circuiting of fluid through one of the chambers
appears to be responsible for this observation.
INTRODUCTION stances. Also, its germicidal effect is not limited pri-
Chlorination has to date been the most widely util- marily to bacteria. UV absorption by nucleic acids
ized disinfection process because of its low cost, sim- causes disruption of the DNA or RNA molecules
plicity of operation, and ability to provide residual which are vital to all organisms including viruses. This
protection. However, in light of the suspected carcino- suggests that UV radiation should be as effective a dis-
genic properties of chlorinated hydrocarbons and the infectant of the viral component of wastewater as it
undue stress of chlorine residuals on stream ecology, is of the bacterial component.
the use of ultraviolet disinfection is being seriously In the present study, some aspects of the UV disin-
considered as an alternative. Further justification fection process which are still poorly understood are
arises from the fact that chlorination does not reduce under investigation. These are the use of chemical
wastewater virus concentrations effectively. actinometry to measure UV dosages, the occurence of
Ultraviolet radiation, being a physical agent, is not photoreactivation in UV treated samples exposed to
believed to cause the formation of toxic chemical sub- sunlight and the effects of water quality parameters on
108
-------
ULTRAVIOLET LIGHT
the UV disinfection of filtered and unfiltered secon-
dary effluent.
EXPERIMENTAL SYSTEM
UV disinfection of filtered and unfiltered secondary
effluent is being conducted on a pilot scale in Durham,
North Carolina at the city's 3.8 x lO-'m-Vd (1.0 mgd)
Sandy Creek contact stabili/ation plant. The plant
produces an effluent of a quality comparable to a typi-
cal activated sludge plant. For the unfiltered experi-
ments, water is drawn directly from the plant effluent.
For the filtered runs, effluent is diverted into one of
three downflow multimedia filters before entering the
UV units. Water leaving the units is pumped into a
large storage tank where it is subsequently used as
backwash water for the filters.
The project employs two commercially available
UV sterilization units. Unit No. 1 included an 8
liter cylindrical tank containing 14 25-watt, 38 cm
(15 in.) mercury low pressure lamps arranged
around the upper two-thirds of the chamber. The
lamps are packed closely together to reduce the
width of the fluid through which the UV
radiation must penetrate. This is referred to as a
thin-film design. Water enters the bottom of the
unit through a distribution tube and flows up
and over a central shaft in somewhat spiral pat-
tern before exiting at the bottom again. Dye
studies, shown in Figure 1, indicate that a
significant amount of fluid short-circuits the
chamber, while other portions are retained for
excessively long periods. The former can be seen
by noting that the actual retention times are shor-
ter than V/Q, the theoretical retention time.
Unit No. 2 features an 11 liter cylindrical tank hous-
ing six 40-watt, 91 cm (36 in.) low pressure mercury
ACTUAL RETENTION
TIME
10 15 20 25 30
TIME FROM INJECTION,sec
35
lamps. The lamps are not as closely spaced as those in
unit No. 1, and thus the reactor operates in a "thick-
film" mode. Flow through the chamber is essentially
linear with water entering and exiting each end of the
unit from the top. Four disk shaped baffles are in-
cluded to insure adequate mixing. The results of dye
studies, shown in Figure 2, indicate a moderate amount
of dispersion in the chamber but no short-circuiting
or tailing.
For both units, the voltage input to the lamps is
adjusted via a Variac. Intensity readings are taken at
254 nm with an International Light IL-500 research
radiometer calibrated to standards traceable to the
National Bureau of Standards. Measurements are
made approximately 5 cm (2 in.) above quart/ win-
dows situated at the wall of both reactors.
1.0
0.8
§0.4
0.2
ACTUAL RETENTION
TIME
- THEORETICAL
RETENTION TIME
15 20 25 30 35
TIME FROM INJECTION,sec
40 45
Figure 1. Unit No. 1 dye studies
Figure 2. Unit No. 2 dye studies
ACTINOMETRY STUDIES
INTRODUCTION
UV light exhibits a maximum germicidal effect on
microorganisms at approximately 265 nm. In most
theoretical investigations, the log of the ratio of in-
fluent to effluent microbial counts is essentially linear
with UV dosage. In wastewater studies, these dose-
response curves are usually linear at first, but as dos-
ages increase to more or less saturation levels, the
curves begin to tail off before eventually assuming
a slope of xero.
A lack of consistency in dose-response data, how-
ever, is a problem which is caused by the method used
currently to determine UV dosages. Before UV light
can be considered a viable alternative to chlorine, a
means of determining a dose which more accurately
reflects the actual amount of radiation to which the
microorganisms are exposed must be found.
Presently, dosages are usually determined by multi-
plying an intensity measured at the wall of a reactor
109
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
times the retention time inside the unit. Unfortunately.
intensity readings are more dependent upon the par-
ticular location of the detector in relation to the lamps
than on the actual amount of radiation supplied to the
unit. Secondly, the retention time does not describe
the actual flow patterns within the unit such as the
degree of mixing and the amount of short-circuiting.
The geometry of the reactor is a very important vari-
able in UV disinfection, which is almost completely
ignored by this method of determining dosages. There
is no way in which the relative effectiveness of two dif-
ferent units can be accurately predicted.
There are a number of factors that need to be con-
sidered in attempting to obtain a true measure of the
average amount of UV radiation that reaches the
microorganisms. Ideally, one would prefer a dose
which takes into account variables such as the trans-
mission of the fluid, the amount of reflections off the
wall, the degree of scattering, the average distance
between the lamps, and so forth. Various methods for
calculating average intensities have been presented in
the literature which deal with some of these factors.
These procedures have eliminated the bias due to the
position of the UV detector in relation to the lamps
and have also taken into consideration the average
depth of fluid.
Another way in which a more meaningful UV dose
may be obtained is through the use of potassium fer-
rioxalate actinometry. Solutions containing 0.006N
ferrioxalate ion in 0.1 N H2SO4 undergo a photo-
chemical decomposition of known quantum yield,
$ . upon exposure to UV light. In the process, ferric
iron is reduced to the ferrous form. Because the solu-
tion absorbs essentially 100%. of the UV emitted by the
lamps, and because
-_ ft molecules of product formed
# photons absorbed
METHODS
The potassium ferrioxalate solutions were pre-
pared in a 1.3 m' (350 gal.) polyethylene tank. From
there they were pumped through either of the two units
at flow rates ranging from 76-2851 min. (20-75 gpm).
Samples were collected in brown bottles at three lamp
voltage settings for each flow rate tested. In addition,
two blanks were collected ahead of the units. To in-
sure that control over the lamps' output .was main-
tained from one experiment to the next, intensity
readings were obtained in tap water prior to each run.
Intensity measurements could not be made in the ac-
tinometry solution itself, as it absorbs all UV light.
Samples were analy/ed for ferrous iron spectro-
photometrically using 1:10 phenanthroline mono-
hydrate as the colorometric indicator. The exact pro-
cedure employed was described by Hatchard and
Parker (2). Throughout the analysis, all samples were
protected from exposure to light to eliminate any
additional photodecomposition of the ferrioxalate ions.
The following equation was then used to calculate
dosages in pw-sec cm1.
dose -
1.196 X 10" pW sue
254 nm
$ X 1000 ml/1
where $ = 1.25 (5> 254 nm
This calculation is based on the assumptions that 1)
the actinometry solution absorbs lOO'/r of the emitted
UV radiation and 2) all of the radiation emitted is at
254 nm.
Disinfection studies were carried out by irradiating
pure cultures of E. co/i in buffered tap water over the
same range of voltage and flow rates used in the actin-
ometry studies. Organisms were enumerated before
and after treatment using the 5 tube multiple-tube
fermentation technique as described in Standard
Methods (1).
a measurement of the quantity of ferrous iron pro-
duced can be converted to an average number of pho-
tons emitted per volume of solution. Thus, dosages
are measured in units of energy per unit volume rather
than in the customary units of energy per unit area. In
the present study, dosages measured in this way and in
the conventional fashion were compared to log reduc-
tion ot Esi'herichia co/i in the two units to determine
what advantages actinometry might hold over the
latter method.
RESULTS AND DISCUSSION
Typical results from the actinometry studies arc
shown in Figures 3 and 4. All curves exhibited some
degree of non-linearity, particularly those obtained
from unit No. I. All plots extrapolate to the origin as
would be expected except for the dose vs. retention
time plots from the first unit. The reason for the excep-
tion is probably that actual retention times are sig-
nificantly shorter than theoretical retention times in
that unit.
110
-------
ULTRA VIOLET LIGHT
LJ
CO
O
Qro
rr \
8x10"
6xl04
4xl0
p 2x10'
o
0
ret. time =6.24 sec
0 250 500 750 1000 1250
UV INTENSITY, uW/cm2'
e
o
« 8xl04
LJ- 6xl04
CO
o
a .
>- 4xl04
2x10^
o
<
0
UV intensity =1190 uW/cm?-
0 2.0 4.0 6.0 8.0 10.0
THEORETICAL RETENTION
TIME,sec
Figure 3. Unit No. 2 actinometry dosages vs. a) UV in-
intensity at reactor wall, and b) retention time
The results of the tap water disinfection studies with
pure cultures of E. co/i are plotted in Figures 5 and 6
vs. the conventional, or intensity times retention time,
dose, and the actinometry dose, respectively. Unfor-
tunately, only the tail of the dose-response curve
could be obtained because of operational constraints
on the system which prevented lower dosage levels
from being achieved.
If the actinometry had provided a true measure of
the UV dose supplied to the microorganisms, i.e., one
that takes into account all of the geometrically related
variables such as reflection, short-circuiting and the
average depth of fluid, then the two curves in Figure 6
would have been identical. Although they are quite
different, there is an even greater lack of similarity
between the two curves in Figure 5. Dosages obtained
using actinometry appear to control at least some of
the factors that are neglected by the conventional dose.
lxlOE
ji> 8xl04
I
jj 6xl°4
8
Q
>- 4xl04
rr
2x10'
O
<
0
ret. time = 6.02 sec
0 1000 3000 5000
UV INTENSITY, uW/cm2
8xl04
ul
co • .
§ro 6XI04
tn«4x|o
li
p 2x10'
o
0
b.
UV INTENSITY
4930 uW/cm2
0 2.0 4.0 6.0 8.0
THEORETICAL RETENTION
TIME,sec
Figure 4. Unit No. 1 actinometry dosages vs. a) UV in-
tensity at reactor wall, and b) retention time
The major difference in the relative position of the
curves in Figures 5 and 6 is probably due to the acti-
nometry 's elimination of the bias caused by the posi-
tion of the UV detector. The quart/ window through
which intensity readings were made was situated much
closer to a lamp in unit No. I than in unit No. 2. This
caused UV intensities and hence conventional dosage
measurements on the former unit to be grossly inflated
relative to the latter. At a given voltage setting and
retention time, the first unit was found to supply a UV
dose five times that of the second unit if calculated
111
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
using 1 x t. Actinometry dosages, on the other hand,
showed only an increase of approximately 1.6 times.
Actinometry dosages provide a measure of the total
number of photons emitted by the lamps per unit
volume and time. This dose is directly related to the
average depth of fluid through which the light passes
if the latter is defined as the1
total volume of unit
surface area of lamps
This is because the total number of photons a lamp
emits is proportional to its surface area. The average
depth of fluid is an important treatment parameter
which actinometry dosage measurements inherently
take into account.
u
0 2.0
h-
0
-)
Q 4.0
£
8 6.0
;j
an
\
J \
\ V UNIT NO. 1
\ ^*^- ± 4 A J A
-\
\
- V&A A A UNIT NO. 2
^l **— — .
A ^
1 1
10,000 20,000
CONVENTIONAL DOSE (IxT),
uW-sec/cm 2
30,000
Figure 5. Log reductions of E. Coli vs. conventional
dose
P 2.0
1-
0
u
S <*'(")
cc
8 6.0
Q n
iVx
\ N* UNIT NO. 1
\ ^-^*- 4f.A A» ,
\ » * . «. i
' \
\
\ A ^ UNIT NO- 2
V., A _ . ^
A A
1 1 1 1
ACTINOMETRY DOSE, uW-sec/cm3
Figure 6. Log reductions of E. Coli vs. actinometry dose
The differences between the two curves in Figure 6
emphasize the failure of the actinometry method to
control all of the important variables mentioned
earlier; in particular, the degree of short-circuiting in
the two units. For example, the first unit was unable
to achieve log reductions greater than 3.6 probably
because a small percentage of the fluid short-circuited
the chamber. This short-circuiting, however, would
not be expected to have as significant an effect on
actinometry dosages as it would on disinfection effi-
ciencies, which are based on a logarithmic scale.
Actinometry would also not be expected to control
variables such as the degree of reflection of UV light
off the reactor walls and the configuration of the lamps
in relation to the overall unit. This is because the solu-
tion absorbs almost 1009r of the radiation within a
depth of only 3-4 millimeters. Finally, the effects of
UV absorbing substances present in the waters to be>
treated are not considered. For this reason, attempts
to correlate actinometry dosages with log reductions
of microorganisms in wastewaters of widely varying
characteristics are meaningless.
UV dosages measured using chemical actinometry
are more useful than those obtained by the conven-
tional method because they are related directly to the
average depth of fluid through which the light passes.
Although not all of the treatment variables involved
are controlled by this method, it is still a potentially
useful technique for comparing and calibrating UV
disinfection units.
PHOTOREACTIVATION STUDIES
INTRODUCTION
An additional problem may be associated with the
use of UV light to disinfect wastewater effluents. A
situation may arise where organisms supposedly in-
activated by UV radiation may restore their repro-
ductive abilities upon exposure to visible light. Such
an event is termed photoreactivation (PR). It has been
defined by Jagger (3) as "the restoration of ultraviolet
lesions in a biological system with light of wavelength
longer than that of the damaging radiation". The re-
pair mechanism of PR is activated upon exposure to
light of wavelength between 300 and 500 nm; a major
portion is stimulated in the region of 355-385 nm for
most microorganisms. A phenomenon such as photo-
reactivation must be taken into consideration when
evaluating the overall efficiency of an ultraviolet dis-
infection process.
METHODS
Samples of secondary effluent treated and untreated
by both units were collected in sterile bottles wrapped
in aluminum foil. Flow rates and voltage levels used
were comparable to those employed in the actinometry
studies. After being placed immediately on ice, the
samples were transported back to the laboratory while
being kept completely unexposed to light. Sample
aliquots were pipetted into petri dishes labelled as
follows: 1) influent dark control, 2) effluent dark con-
112
-------
ULTRA VIOLET LIGHT
trol (wrapped in aluminum foil), 3) influent sunlight
exposed and 4) effluent sunlight exposed. All dishes,
control and exposed, were kept in a constant temper-
ature bath for forty-five minutes in direct sunlight.
The intensity of PR light was measured at 15 minute
invervals with a General Electric light meter equipped
with a glass 320-390 nm filter. After completing the
exposure, the samples were immediately placed in a
dark, 4°C refrigerator until further use within the
hour. Total coliforms were enumerated using the
membrane filter technique as recommended in Stan-
'(/ur(/ Methods (1).
RESULTS AND DISCUSSION
The initial point of interest concerning the photo-
reactivation of organisms in a wastewaterenvironment
dealt with the effects of temperature. Influent and
effluent samples collected from unit No. 2 were ex-
posed to sunlight as described earlier. Temperature
baths at 4°C and 25°C were used to compare PR
simultaneously. As shown in Table I, the disinfection
in the dark controls at each temperature was consis-
tantly higher than in their counterparts exposed to
sunlight. The decrease in coliform kills may be attrib-
uted to the organisms ability to photoreactivate and
resume replication. Also of interest is the fact that sun-
light exposed log reductions at 25° C are approxi-
mately 40% less than those at 4°C. This decrease,
although small, is consistant and suggests that PR is
more probable in a 25°C environment than at 4°C.
TABLE 1. EFFECTS OF TEMPERATURE
ON PHOTOREACTIVATION
4«c
25°C
Mean Log Reduction
Std. Deviation
95% Confidence Interval
Mean Log Reduction
Std. Deviation
95% C.I.
Dark
Control
2.07
0.14
1.73-2.41
2.07
0.14
1.94-2.20
Sunlight
Exposure
1.59
0.58
0.61-3.02
0.93
0.21
0.40-1.46
An interesting result which arose during the temper-
ature study was the decrease in influent total coliform
counts from the dark control to the sunlight exposed
sample. At both 4° and 25°C, a small but consistant
decrease was observed between all dark and light
samples, especially with the first run, where greater
than one log kills were found. UV light disinfects coli-
forms most effectively in the region of 265 nm, yet very
little energy of wavelengths shorter than 290 nm are
found at the earth's surface. The results thus suggest
that disinfection due to the action of longer wave-
lengths is occuring simultaneously with photoreac-
tivation. During sunlight exposure, these two phe-
nomena compete with one another, with photoreac-
tivation usually predominating.
Samples from unit No. 2 exposed to high UV dos-
ages behaved differently upon exposure to sunlight
from those exposed to a lower dose. A finding con-
sistant with previous data was the second unit's super-
ior performance relative to unit No. 1 in the dark
controls. The result was further manifested at high
dosages in the relative degree of photoreactivation
observed in samples treated by the two units. The
relatively ineffective unit No. I inflicted only minor
damage to most of the inactivated organisms, thus
allowing them to make use of the PR light. See Table 2.
The second unit, on the other hand, operated with
such a high efficiency that no organisms were capable
of repairing themselves. In fact, additional disinfection
was observed when these samples were exposed to sun-
light.
TABLE 2. PHOTOREACTIVATION USING HIGH UV DOSE
Unit No. 1
Unit No. 2
Mean Log Reduction
Std. Deviation
95% C.I.
Mean Log Reduction
Std. Deviation
95% C.I.
Dark
Control
3.24
0.14
3.14-3.38
4.01
0.08
3.93-4.09
Sunlight
Exposure
2.08
0.45
1.60-2.55
5.13
0.54
4.56-5.70
This study's approach to photoreactivation was
unusual in that PR has always been studied with pure
cultures of microorganisms rather than the hetero-
geneous populations characteristic of municipal efflu-
ents. As such, control over variables such as the chem-
ical and biological characteristics of the wastewater
and the intensity of sunlight could not be exercised
in these experiments. Nevertheless, the results show
quite clearly that the occurence of photoreactivation
may significantly reduce the efficiency of wastewater
UV disinfection.
WASTEWATER STUDIES
INTRODUCTION
A pilot plant study is being conducted to define the
water quality parameters useful for predicting and
113
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
monitoring the effectiveness of a UV treatment opera-
tion in wastewater. Many parameters are therefore
being measured in this study in an attempt to isolate
those most directly related to disinfection efficiencies.
Two independent variables are also being investigated:
1) filtered vs. unfiltered effluent and 2) lamp spacing
in the UV units. The thin-film desisin characteristic of
the first unit has potential advantages over the
thick-film arrangement particularly in water with
a high UV absorbance.
METHODS
The experimental design includes four variables,
each of which can assume two conditions: 1) U V unit
(No. I or No. 2), 2) flow rate (approximately 130 or
2851 min), 3) voltage input to lamps (60 V or 127 V).
and 4) quality of the influent to the units (filtered or
unfiltered). Batch experiments are performed on each
unit using randomly determined combinations of the
six possible conditions. A total of nine repeats of each
set of conditions will eventually be performed.
Samples used for chemical analysis are collected
in 3.8 liter plastic containers, while those for bacter-
iological examination are obtained using sterile brown
glass bottles. Lamps are allowed to warm up to 35-
40°C before the samples are taken.
In an effort to correlate parameters of water quality
with the efficiencies of the U V units, a battery of chem-
ical tests are performed on all samples using pro-
cedures outlined in Standard Methods (1). Influent
wastewater to the units is analyzed for the following:
1. Temperature
2. UV absorbance (Beckman DB spectrophotom-
eter)
3. Alkalinity
4. Conductivity (Radiometer model CDM 2e)
5. Suspended solids
6. Nitrogen forms (NH3, NO2. NOs, Auto Ana-
ly/er)
7. Total Iron (Atomic Absorption Spectrophotom-
eter)
8. Total Organic Carbon (Beckman combustion
IR carbon analyzer)
9. Turbidity (Hach 2424 nephelometer)
10. pH
II. Chemical Oxygen Demand
Effluent samples from the units are examined for
temperature, dissolved oxygen (YS1 Model 51B),
nitrogen forms, conductivity, TOC, COD, and pH.
The 5 tube multiple-tube fermentation technique
as outlined in Standard Methods (1) is employed for
the enumeration of total and fecal coliforms in all
samples. Results arc reported as log reductions in coli-
form counts. The procedure is performed to the con-
firmed level with lauryl tryptose utili/ed as the pre-
sumptive medium.
RESULTS AND DISCUSSION
Approximately 30% of the experiments have been
completed thus far. The data have been correlated and
preliminary efforts to analyze the results for signifi-
cant trends have been made.
Two findings are particularly noteworthy at this
time: 1) unit No. 2 performs much more efficiently
than unit No. 1 under all conditions, and 2) flow rate
and lamp voltage have almost no effect on log reduc-
tions of coliforms at the levels tested. Both of these
observations are consistant with the pure culture tap
water results discussed earlier. Apparently, at even the
lowest possible UV dosage level available, the units
are operating where there is little increase in inactiva-
tion with increased dose. In this region, the major
source of variation in the coliform log reductions is
the differences between the units themselves. The sec-
ond unit gave a mean log reduction of 3.56 in total
coliform while the first unit gave only 1.92 logs inacti-
vation.
In this investigation, conclusions regarding the
advantages of the thin-film vs. thick-film concept in
reactor design will not be possible. As shown in Fig-
ures 1 and 2, differences in the flow patterns within the
units are obscuring any possible effect of film thick-
ness. Although unit No. 1, with its 14 closely spaced
lamps, yields a UV actinometry dose 1.6 times unit
No. 2, short-circuiting in the reactor appears to be
preventing it from achieving high log reductions of
coliforms.
In attempting to correlate the various water quality
parameters and process control variables with UV dis-
infection, problems have been encountered so far due
to the overpowering effect of the unit on log reduc-
tions. For example, a strong negative correlation be-
tween log reductions and UV intensity was obtained
simply because the less effective unit No. 1 actually
yields a higher UV intensity reading than does the
second unit. Plots of UV intensity vs. log reductions
for each unit separately show no correlation. Most of
the water quality parameters exhibit little or no cor-
relation with log reductions at this time. However,
once the statistical analysis is performed with respect
to each unit, the importance of these variables will be-
come clearer.
114
-------
ULTRA VIOLET LIGHT
SUMMARY AND CONCLUSIONS
Ultraviolet disinfection studies are being conducted
at a pilot plant using two commercially available 230
to 285 1/min (60-75 gpm) UV units. The use of potas-
sium ferrioxalate actinometry to measure UV dosages
has been found to have advantages over the conven-
tional method. Actinometry dosages are directly re-
lated to the average intensity of all lamps and the aver-
age depth of fluid in the units, while conventional
dosages are biased by the position of the UV detector
in relation to the lamps.
Results from photoreactivation studies of UV in-
activated coliform organisms exposed to sunlight
show a 1 to 2 log recovery in total coliforms at 25° C.
Lower recoveries were found at a lower temperature.
Under high dosage conditions, photoreactivation was
found to be a function of the unit, with samples from
unit No. 2 actually showing additional disinfection
upon exposure to sunlight.
Ultraviolet disinfection studies on filtered and un-
filtered secondary effluent are still in progress. Only
small effects of water quality parameters and UV dos-
ages on log reductions of total and fecal coliforms
have been found so far. The primary variable affect-
ing inactivation is the non-ideal flow patterns of the
two units. This result is due probably to the effect
. that short-circuiting is having on the effectiveness of
unit No. 1. The importance of the other variables will
become clearer following a more complete analysis
of the data.
REFERENCES
I. American Public Health Association. 1975. Standard Methods
for the Examination of Water and Wastewater. 14th edition.
APHA, AWWA and WPCF, Washington, D.C.
2. Hatchard, C. G. and C. A. Parker. 1956. "A New Sensitive
Chemical Actinometer II. Potassium Ferrioxalate as a
Standard Chemical Actinometer." Pruc. Roval Soc. Lon-
don. A235:518.
3. Jagger, J. 1958. "Photoreactivation". Bat: Rev. 22:99.
DISCUSSION
MR. HEINSOHN, Pollster Ind., Inc.: Clarify
for me unit 1 and 2. Which was the thin film?
MR. ALDRICH: The thin film was unit num-
ber 1, and the thick film is unit number 2.
MR. HEINSOHN: Unit number 2 was inter-
nally baffled. Is that correct?
MR. ALDRICH: Yes. It has four internal baf-
fles.
M R. TON ELLI,Onfar;o Ministry of the Environ-
ment: Apart from being a descriptive term,
what does the term "thin film" versus "thick
film" actually mean as regards to fundamentals
of design?
MR. ALDRICH: It refers to the lamp spacing
within the units. I am sure there is an arbitrary
cutoff where you define something as thick film
and where you define it as thin film so it is more
of a qualitive kind of comparison. It relates to
how closely spaced the lamps are together, and
particularly what the average depth of fluid is
through which the light must penetrate.
MR. TONELLI: From your actinometry results,
you do not feel there is anything fundamentally
different. Is it just a matter of definition or
semantics?
MR. ALDRICH: Are you talking about dif-
ferences between measuring the average depth and
the actinometry doses?
MR. TONELLI: No, I am still talking about
the thin film and thick film concept, because on
the face of it there seems to be an infinity of in-
termediate thin films and micro-thin films. Was
there any theoretical basis for selection of depth
or separation of lamps for the study, or were
there two units that were available?
MR. ALDRICH: No. We originally were in-
terested in comparing the two units on the basis
of the thin film versus the thick film. That was
one of the original objectives of the study. As
far as the actinometry is concerned, the spacing
of the lamps did not really have that much to do
with the actinometry. The spacing is related to
the average depth of fluid and so is the ac-
tinometry, but other than that there really was no
particular basis for comparison.
MR. DeSTEFANO: I have a question about the
photoreactivation study. Did you do any counts
immediately after collecting from the effluent,
because there is also a phenomenon of dark.
repair and I would think that your control says
that photoreactivation is a more powerful repair
mechanism than dark repair is.
MISS ELLIOTT: I did not do any immediately
afterwards. There was a matter of transport time
in getting everything set up, and before we ex-
posed these samples they were kept in a dark
refrigerator at about 4°C. So I did not do any
immediate ones afterwards.
115
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
MR. DeSTEFANO: But dark repair could
have occurred in forty-five minutes?
MISS ELLIOTT: There was a possibility, but I
would think that the cool environment of the
refrigerator would possibly prevent that.
MR. VENOSA: I would like to make one par-
ticular comment. Our original idea in this project
was that we wanted to compare thin film versus
thick film, and there is no clear distinction bet-
ween thin film and thick film. There is one com-
pany that makes a thin film type design, and this
refers to approximately 0.25 inch of water wall.
That is the distance the light traverses through
the water. The distance between any two quartz
sleeves is approximately 0.5 inch. Therefore, the
water wall is 0.25 inch. The other unit is ap-
proximately 1 to 1.5 inches. The concept is what
we were interested in, i.e., thin film, high inten-
sity, short detention time, versus thicker water
wall, long detention time, lower intensity.
MR. TRAVER, EPA'- Any type indication we
have had as far as UV disinfection deals with
cleansing of the ultraviolet lamp sources in a
submerged condition, be it either some type of a
cleansing solution or mechanical light system. I
do not remember or recall in the presentation any
indication of this situation. Did you have to go
on a down time basis, or how was this attempted
in units 1 and 2?
MR. ALDRICH: We are not running any con-
tinuous experiments at this time. All our experi-
ments are short term, and the two units are dif-
ferent with respect to the cleaning mechanism.
Unit number 1 has a mechanical cleaning system
powered by compressed air, and unit number 2 is
cleaned with the use of a cleaning solution which
is circulated through the chamber. We have not
run into any problems with down time since we
are not running continuously, and both of the
systems seem to work quite well at this time.
MR. TRAVER: Can unit number 1, using a
mechanical wiper system, be cleaned while in
operation, or does it have to go down?
MR. ALDRICH: No, it is cleaned while operating.
116
-------
15.
FULL SCALE EVALUATION OF ULTRAVIOLET DISINFECTION
OF A SECONDARY EFFLUENT
O. Karl Scheible, Gloria Binkowski and Thomas J. Mulligan
Hydroscience, Inc.
363 Old Hook Road
Westwood, New Jersey 07675
INTRODUCTION
This presentation is a progress report on a full scale
evaluation of U.V. Disinfection of an effluent from a
secondary wastewater treatment plant. The study is
funded by the U.S.E.P.A. (MERL)and the Northwest
Bergen County Sewer Authority. The principal
investigator for the study is Hydroscience, Inc.,
Westwood, New Jersey. The ultraviolet equipment
was developed and is manufactured by Pure Water
Systems, who operate and maintain the on-site
equipment.
The tasks involved in the overall program included
the installation and shakedown of the U.V.
equipment. Experimental work was then directed to
an evaluation of the system under various operational
conditions to determine dosage requirements relative
to disinfection efficiency. Earlier in the program the
system was continuously monitored during a viral
sampling program conducted by the Carborundum
Company, the results of which are to be reported in a
separate presentation. The phenomenon of
photoreactivation was evaluated concurrently with
the primary sampling program, and the system was
continually monitored for operation and maintenance
efficiency. The final report to the EPA is expected in
the spring of 1979.
SITE LOCATION
The site of the installation is the Northwest
Bergen County Water Pollution Control Plant,
located in Waldwick, New Jersey. Figure 1 presents a
site schematic of the plant. It is a conventional air
activated sludge plant with a design capacity of 30,000
m-Vday (8 mgd), and an average yearly flow, at
present, of approximately 18,900 mVday (5 mgd).
Completed in 1968, the plant is a modern, efficient
facility discharging a well treated secondary effluent to
Ho-Ho-Kus Brook, a water quality stream.
The plant has dual chlorine contact chambers.
Under present flow conditions, one has remained
inactive. This provided an ideal location for the
installation of the gravity feed U.V. disinfection
system.
EQUIPMENT INSTALLATION AND SPECIFI-
CATIONS
A plan view of the chlorine contact chambers
presented in Figure 2 shows the overall U.V.
installation. To the left is the active chlorination
system, and on the right the inactive chlorine contact
chamber which provided the site location. The unit
itself was installed at the head end of the chamber, the
influent flow rate being controlled by weir gates on the
influent channel. The U.V. unit was set into the
channel and a platform set above it to support the
ballast and provide a work area. Flow was measured
by 4 V-notch weirs located at the effluent end of the
contact chamber with a level sensor and recorder. A
C12 diffuser was located approximately 6 to 9
meters downstream of the U.V. unit in compliance
with New Jersey law.
A cutout view of the C12 contact chamber on Fi-
gure 3 shows the actual installation of the U.V. lamp
battery. Two concrete webs provide the support. A
pump is placed between the webs to keep the space
dry. The lamp battery itself is supported by two steel
bulkheads set into the concrete, webs, with an air seal
similar to a rubber inner tube along the perimeter
between the bulkhead and concrete webs. Figure 3
also shows the mechanical wiper and the pneumatic
cylinder which drives the wiper mechanism.
117
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
SERVICE BUILDING
THICKENERS
A~
CHLORINE STORAGE AREAx
INLET
BUILDING
PRIMARY
SETTLING
TANKS
AERATION
TANKS
I ^ INSTALLATION
FINAL
SETTLING
TANKS
,*^-f
\
\
r
i
CHLORINE //
CONTACT CHAMBERS — '
s'
WATER
SUPPLY
BUILDING
EFFLUENT TO
HO-HO-KUS BROOK
FIGURE 1. SITE SCHEMATIC
FROM
SECON
CLARIFIER
XRY •• INFLUENT CHANNEL
ER
|
-—~—.
-GATE
i EFFLUENT CHANNEL
. r-*- : : — cfc.
r
UV UNIT--
BALLAST
/
PLATFORM^
1
' '\j
CHLORINE CONTACT
CHAMBER
___
-a
/
/
°7=
r
c
•••
7^
MV-NOTC
WEIRS
^FLOW
RECORD
^LEVEL
H
ER
SENSOR
ZCI2
DIFFUSER
^-^
TO
) (
„
range. Each lamp is jacketed in quartz. The total
effective arc length of the unit is 610 meters (2,000 ft).
Total power consumption by the unit is 45 KVA at
an operating voltage of 480 V. The unit is capable of
lamp battery shutoff in 1/6 increments and power
variability from 40 to 100% full power.
The overall dimensions of the unit are 76 \76 \ 142
cm (3x3\6 ft) with a void volume of 0.63m' (22.2 ft3).
Head loss is estimated at 15 cm (6 in) at a flow rate of
21,000 m-Vday (5.5 mgd). The mechanical wiper
mechanism is comprised of replaceable elastomeric
glands fitted over each of the quartz tubes. The wipers
are cable driven at a variable stroke rate by a
pneumatic cylinder. A unique feature of the U.V. unit
it the utilization of the "thin film" concept, which is
induced by the spacing of the lamps. The nominal
liquid film thickness is 0.6 cm (0.25 in).
OUTFALL EXPERIMENTAL PROGRAM
FIGURE 2. PLAN VIEW OF ULTRAVIOLET UNIT
INSTALLATION
The specifications of the unit used at Northwest
Bergen County are presented in Figure 4. It has a total
of 400 germicidal lamps. They are 142 cm (6 ft long) 85
W lamps with an output of 30 Win the germicidal U.V.
The data reported herein represent the analysis
period of June through August, 1978. During this
time, two randomized sampling series were conducted.
The first, conducted in June and July, evaluated the
system at two specific flow rates and four applied
voltage settings. The second series evaluated the
118
-------
ULTRA VIOLET LIGHT
STAINLESS STEEL
BULKHEAD
WIPER DRIVE
. PISTON
INFLUENT
EXISTING CI2
CONTACT
-REINFORCED
CONCRETE
SUPPORT WALLS
FIGURE 3. ULTRAVIOLET DISINFECTION UNIT
INSTALLATION
system at a single flow rate with variable lamp
operation and variable applied voltage settings.
Sampling was conducted two days per week at eight
operational modes per day. Thus, sixteen samplings
were conducted per week. The sequence of operational
conditions was randomly selected to minimize bias
due to variations in water quality. The randomized
sampling series conducted during the June
through August period were part of an intensive
sampling program, statistically designed to
provide an evaluation of the system relative to
bactericidal efficiency under a wide range of
ultraviolet energy dosage applications.
RESULTS
Wastewater Characterization
Figure 5 is a summary of the analyses performed on
the influent to the U.V. unit. Sampling was by grab
only, in sterile opaque glass bottles. Total and fecal
coliform densities were measured by the 5-tube
multiple tube MPN procedure. All influent grabs were
analyzed for the specific wastewater quality
parameters shown on Figure 5. The mean densities of
total and fecal coliform were 3.6xl05 and 9.5xl04,
respectively. The influent wastewater was relatively
stable, and highly indicative of quality treatment; the
COD averaged 26 mg/1, while the turbidity and SS
were relatively low at 4 FTU and 6 mg/1, respectively.
The water temperature averaged 22° C. As shown by
SPECIFIC A TIONS
MODEL : PWS SE-7.5
LAMPS : NUMBER 400
TYPE : 85 W PWS-L60; 152 CM. LENGTH
30 W AT 2537 A
2.3 CM. 0 QUARTZ JACKETING
TOTAL EFFECTIVE ARC LENGTH: 610 METERS
TOTAL POWER : 45 KVA
OPERATING VOLTAGE: 480 VAC/60 HZ/3 PH
LAMP BATTERY SHUTOFF IN 1/6 INCREMENTS
POWER VARIABILITY: 40 TO 100%
DIMENSIONS : 76 x 76 x 152 CM.
VOID VOLUME : 0.63 M3
HEAD LOSS ~ 15 CM. AT 21,000 M3/DAY
WIPER: REPLACEABLE ELASTOMERIC GLANDS,
CABLE DRIVE OFF PNEUMATIC CYLINDER
FIGURE 4. U.V. DISINFECTION UNIT
JUNE, JULY, AUGUST, 1978
(II9SAMPLES)
MEAN
MIN. MAX.
TOTAL COLIFORM
FECAL COLIFORM
COD ( F ) 2 6.0 _
S S 6.0
3.63xl05 3.2xl04l.6xl07
9.5 xlO4 IxlO4 3.2xl06
14.0 - 47.0
1.0 - 18.0
TURRIDITY
pH
TKN (F)
NOo -N
N03 -N
NH, -N
\J
COLOR
U.V. ABSORBANCE (T) '
4.0
II O
I.I
8,9
9.0
34.0
0.171
1.0 -
6.5 -
3.0 -
0,3 -
1.8 -
1.9 -
18.0 -
0.277
25.0
7.6
24.0
3.7
20.0
20.0
75.0
'-0.137
TRANSMITTANCE
(I)
. 67.4
52.8 - 72.9
(I) I CM. CELL AT 253.7 nm
FIGURE 5. SUMMARY • INFLUENT ANALYSIS
119
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
the distribution of the nitrogen species, the plant was
operating in a nitrification mode during the three
month period. U.V. transmittance was measured in 1
cm quartz cells. The average transmittance was 67%,
with a range of 52 to 73%. This transmittance level falls
in the range indicative of a high quality secondary
effluent.
Approximately one half of the effluent samples
collected were analyzed for these same parameters, the
results of which were tested against the influent
analyses to determine if there was any change upon
irradiation. A Student's t analysis at the 95%
confidence level indicated no significant alteration
in the wastewater parameters measured. The
results of this analysis are summarized on Figure 6.
Disinfection Efficiency
Before presenting the disinfection results, the
manner in which dosage is reported should be
explained. It is computed simply as the applied
germicidal power (KW) divided by the flow rate
(m3/sec):
D
where D
AGP
Q
AGP/Q
dosage (KW-sec/m3)
applied germicidal power (KW)
flow (m3/sec)
This dosage unit, presented for use in
conjunction with, or as an alternate to the
standard /uw-sec/cm2 term, is reportable in both'
total and germicidal power applications, and may
present a practical procedure for sizing and
comparing alternative U.V. equipment. The
reader is cautioned, however, that this type of
dosage unit used alone represents a "black box"
approach. Consideration must be given to
intensity levels and water quality. It is anticipated
that this study will provide a greater input to a
more effective U.V. sizing parameter.
Figure 7 presents a composite of all coliform
data collected during the three month period,
representing a total of 119 samplings. The data
are presented as a log-log relationship of
surviving fraction to dosage. Least squares
analysis was performed to compute the
regression lines as shown for both TC and FC. As
the figures indicate, the range of dosage levels
investigated was approximately 4 to 90 KW-
sec/m3, representing exposure times in the order
of 0.3 to 5 seconds. In the situation of the
Northwest Bergen plant, a 99.9% removal would
Ho: MEAN A = 0.0
INF. TO EFF.
(56 SAMPLES)
MEAN A
CALC. t'
COD m
COD (F)
COLOR
U.V. ABSORBANCE (T)
U.V ABSORBANCE (F)
TSS
TURBIDITY
TKN (F)
NOo -N
NO, -N
NH, -N
-0.48
-1.0
-O.8
-O.OOPR
O.OOO7
O.I 9
0.18
-2.45
Q.OI 1
O.O4I
-0.085
O.87
1.71
O,9
1.39
0.53
1.37
0,84
1,37
0 73
O.55
079
(I) SIGNIFICANT AT t >2.005
FIGURE 6. TEST OF DIFFERENCES
require a dosage of 35 KW-sec/m3 under average
flow conditions and an exposure time of
approximately 2-2.5 seconds.
Design Nomograph
Assuming a linear relationship between log
•surviving fraction and a log dosage (Figure 7), a
design nomograph was developed relating
influent flow and expected influent coliform
densities to germicidal power requirements. The
nomograph presented in Figure 8 is based on a
desired effluent average fecal coliform density of
<200 MPN/100 ml.
Equipment sizing would be based on peak flow,
which is assumed to be twice the average design
flow of the plant. As an example, if a plant is to be
designed at an average flow capacity of 30,000
mVday (8 mgd), the peak flow condition would
be 60,000 mVday (16 mgd). If the expected
influent fecal coliform density is 105 MPN/100 ml,
Figure 8 indicates an estimated germicidal power
requirement of 18 KW. Utilizing lamps with a
germicidal output of 30 W/lamp,'the implied
lamp requirement would be 600. Similarly,
120
-------
ULTRAVIOLET LIGHT
,0*
Id3
z .4
O 10
Id5
10
-3
10
Id'
TOTAL
COLIFORM
0.5
SECONDS
1.0 IQ.O
. . FECAL
*. COLIFORM
90.
99.
99.9
99.99 <
o
99.999 £
I-
z
UJ
o
a:
UJ
90. °"
99.
99.99
j.i.LuJ 99.999
1.0 10.0 100.0 1000.
DOSAGE (KW-SEC/M3)
FIGURE 7. SURVIVING FRACTION
assuming a total power consumption of 110 W per
lamp, the total power application becomes 66
KW.
As indicated by Figure 8, the assumption of a
linear relationship in log surviving fraction with,
log dosage induces a sensitivity of the system
design to influent coliform densities. Single log
increments in influent density levels will affect
system design requirements by a factor between
3 and 3.5.
Preliminary Cost Estimates
Costs were developed on a preliminary basis
using equipment purchase figures provided by
the manufacturer and the design nomograph
shown on Figure 8. The costs included in the
capital cc«st estimates are the equipment
purchase (frame, unit and control panel, and
installation), and excludes the support structure
and any ancillary equipment requirements.
Design and cost estimates for these have not
been developed as yet. Design of the U.V.
equipment is based .on peak flow conditions and-
amortization is over a twenty year period at an
interest rate of 6%%.
Operation and maintenance costs are reported
on an average flow basis. Service is estimated at
15% of annualized capital costs. It is presently
assumed that the lamps would need to be
replaced at an annual rate. Power costs were
assumed to be 3.5
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
10
o
o
o
10'
10'
z
UJ
Q.
O
io'
NUMBER LAMPS(30 W/LAMP)
100 _ 1,000 10.000 3
ESTIMATE ANNUAL
OSM PLUS EQUIPMENT-
ANNUAL
08M
EQUIPMENT
PURCHASE
PRELIMINARY ESTIMATE
1 OF COSTS
10
10'
O
o
Z
Z
I 10 100 1,000
PEAK GERMICIDAL POWER REQ'D (KW)
10 100 1,000
TOTAL POWER REQUIREMENT (KW)
FIGURE 9. TOTAL POWER REQUIREMENT (KW)
DESIGN PEAK 0 S M TOTAL
AVG. GERMICIDAL (t/IOOOgal.)(t/IOOOgal.)
(mgd) (KW)
10
40
100
1.0
5.2
10.0
43.0
100.0
0.6
0.6
0.5
0.5
0.5
1.4
1.2
0.9
0.9
0.9
FIGURE 10. PRELIMINARY ESTIMATE OF COSTS
the 19,000 mVday (5 mgd) are estimated at 1.2C/
1,000 gal. The cost for plants greater than 38,000
mVday (10 mgd) are estimated at 0.9
-------
ULTRA VIOLET LIGHT
determined. Approximately 25% of the samples
analyzed by this procedure were also held for one hour
in dark bottles as controls. There was no measurable
difference beteen the coliform densities in the dark
bottles after one hour, and the densities observed
immediately after irradiation.
Figure 11 presents the results of the photo-
reactivation analysis as a correlation of log reduction
and log dosage. Regression lines were computed by the
least squares procedure and are shown on the Figure.
Regression lines computed for samples analyzed
immedately upon irradiation are superimposed upon
the photoreactivation results. The obvious shift is due
to the repair of injured cells.
o
o
a.
5
z
o
CPO
LU
a:
cc
o
o
o
TOTAL
COLIFORM
0 HOUR
O.I
1.0 IO.Q,
SECONDS
FECAL
COLIFORM
0 HOUR
1.0
10.0
100.0
1000
DOSAGE (KW-SEC/M0)
FIGURE 11. REDUCTION AT ONE HOUR
Similarly, the regression lines computed for log
effluent coliform versus log dosage at 0 hour
(immediately after irradiation) and I hour (exposure
to visible light) suitably demonstrate the impact of the
photoreactivation phenomenon as shown on Figure
12. Repair accounts for fairly consistent one log
increase in effluent total coliform densities. The
increase in fecal coliform levels varied from
approximately a one log increase at the lower dosage
level to 0.6 log increase at the higher levels.
O
o
z
CL
tr
£
o
o
U-
UJ
10°
icr
10
10
10
10
10
10
10
TOTAL
COLIFORM
AFTER
HOUR
(r = 0.69)
FECAL
COLIFORM
AFTER
I HOUR
(r = 0.68)
EFF. FECAL COLI.
< 200 MPN/IOO ML I
1.0
10.0
100.0
1000
DOSAGE (KW-SEC/M0)
FIGURE 12. PHOTOREACTIVATION
SUMMARY
The results of the U.V. disinfection demonstra-
tion project to date have suggested the following
preliminary findings:
The thin film, gravity flow disinfection unit has been
found to provide effective treatment with low
maintenance over a four to five month period. It is
flexible in its operation and mechanically simple. The
wiper mechanism has had approximately 3,000 hours
continuous operation with no apparent degradation in
cleaning efficiency. The lamps have typically operated
an average of 2,000 hours.
123
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
At fecal coliform influent densities of 105 volume over the flow.
MPN/100 ml, a dosage requirement of 35 KW- MR. WHITE: Very unscientific.
sec/m3 is suggested to meet effluent criteria of MR. DeSTEFANO, Riddick and Associates:
<200 MPN/100 ml. This, of course, relates to I noticed that you had to chlorinate. I visited the
'the water quality conditions at the Northwest plant, and I noticed that the injection point was
Bergen County plant. Costs, without probably about ten feet beyond the unit. Now,
consideration of support structures and ancillary did you check to make sure there was no back
equipment, are estimated to be in the order of diffusion of chlorine? Did you take chlorine
0.9 to 1.4c/1,000 gal. Both cost and disinfection residuals in your samples?
unit sizing are sensitive to the influent density MR. SCHEIBLE: Yes, we did.
levels by factors between- 3 and 3.5 per log MR. DeSTEFANO: There was no chlorine re-
increment in coliform density. sidual?
Photoreactivation has been evaluated, and sig- MR. SCHEIBLE: No. That was moved, inci-
nificant increases (0.6 to 1.0 log increments in dentally, down to 30 feet just for safety purposes.
fecal coliform levels) have been noted after MR. DeSTEFANO: Also, this idea of calcula-
exposure of irradiated samples to visible light, ting dosage is very convenient but you seem not
The implication of repair (or aftergrowth noted to be taking into account UV absorbance, and it
in other disinfection procedures) may be does not matter how much power you are
significant when related to unit sizing and costs. applying to the liquid if the liquid is going to be
absorbing some of the power. All you are
AUTHORS concerned about is how much power is hitting
..„„.„,.,,. „ . r . . the cell, not the liquid, because you are not dis-
Mr. O. Karl Scheible is a Senior Engineer at . - ... ., \, , -„ u i,
... . ... ^,- r,-i i • infecting the liquid. You want to kill the cell.
Hydroscience; Ms. Gloria Binkowski was a
Microbiologist with Hydroscience at the time of MR- SCHEIBLE: I agree. I do not know if I
this study and is now affiliated with Rutgers Uni- can cal1 il a drawback to this study, but we have
versity; Mr. Thomas Mulligan is Technical Direc- dealt with a very consistent quality of water, and
tor at Hydroscience ^ue to l^at we have not been able to really
correlate anything. We have tried multiple
regressions bringing in the idea of water quality
DISCUSSION but due to the consistency of the water over a
three month period we are not really able to
MR. WHITE: Did you run a dye study curve correlate UV absorbance.
on your contact in the unit? MR- DeSTEFANO: A comment about the
MR. SCHEIBLE: No, we did not. quality of the water. The effluent quality was
MR. WHITE: Are you going to? verY g°od, but there were also some very large
MR. SCHEIBLE: It is not in the program. I floe particles when I saw it. I do not think the
find some difficulty there in trying to get any P'ant was down. I think that is a normal condi-
accurate measures even on a dye study in the tion, very light though. You probably do not
order of fractional seconds. That scares me. have a high suspended solids, but in terms of
MR. WHITE: Don't you think it is better than volume you do, and I am not sure that when
nothing? that thing goes zipping through your unit at 2.9
MR. SCHEIBLE: Probably. We did some seconds that a little bug inside one of those floe
velocity profiles across the units both horizontally particles is going to get hit.
and vertically to see if there was any dramatic MR- SCHEIBLE: We did have a condition of
changes in the velocity on the effluent side of the floating material, large clumps. This could be
unit, and we did not see any. No evidence of 'brought' on by two operating conditions. During
real short-circuiting through the unit. a portion of this study at night we dropped the
MR. WHITE: In other words, you are assuming flow back down and dropped the lamp power
there is no short-circuiting so you simply do a V down just to conserve. If you recall the slide
over Q to arrive at your seconds. Is that right? showing the installation of that unit, there is ef-
MR. SCHEIBLE: That is correct. It is a void fectively a dead space in the influent channel to
124
-------
ULTRA VIOLET LIGHT
the time when it hits the UV unit. The UV unit
sits about six feet up. If you drop the flow, there
is material that tends to settle in that chamber. If
you come in in the morning and crank up the
flow, it would scour the bottom and push the
particles through, and we could only exercise the
caution that sampling would not be performed
during that flush period.
MR. DeSTEFANO: When I saw the plant it was
about two in the afternoon, and someone did take
a sample as we were there, and you could actually
see the floe going through the unit. They are not
really floating. They seem to have about the same
density as water in that they stay .somewhat sub-
merged on the bottom, not on the top, and you
could see it just ripping through.
MR. SCHEIBLE: One other point I was just going
to make is that plant is nitrified. Thus, they do have
problems at times with floating sludge in their
secondary clarifiers. They also pull off a lot of algae
from the secondary clarifier.
MR. DeSTEFANO: One last question. Did you
correlate UV light intensity with voltage?
MR. SCHEIBLE: It is linear.
MR. DeSTEFANO: But you cannot go all the
way down. I understand that once you get down
about 40% the lamps will go out.
MR. SCHEIBLE: The lamps start shutting off on
you at about 40 or 45% of full power.
MR. VENOSA: Don't forget that some floe will
go through a chlorine contactor too.
125
-------
16.
ROUNDTABLE DISCUSSION OF ULTRAVIOLET LIGHT DISINFECTION
Participants in Ultraviolet Roundtable
Albert D. Venosa, Moderator
U S. EPA, MERL-Cincinnati
I. Harold Wolf
Texas A&M University
2. Kent Aldrich
University of North Carolina at Chapel Hill
3. M. Elliott
University of North Carolina at Chapel Hill
4. O. Karl Scheible
Hydroscience Associates, Inc.
MR. GOLDRING,Or/on Research Incorporated:
I have what are in essence, I think, one question
and a comment on Mr. Aldrich's work. His results,
I believe, are much easier to understand if one as-
sumes that he is killing all of the micoorganisms
in .99% of the solution and 1% of the solution
never sees the UV at all. As he mentioned he had
difficulties with the bypassing in his equipment, so
essentially he is measuring the bypass rather than
any effect of UV, I believe.
The other thing is that in the actinometry he did
not mention the transmittance of the solution that
he used for his actinometry. If it is low, that is, if he
has a high absorption coefficient, he is measuring
the output of the lamps and not the exposure of
the solution. In other words, to measure the ex-
posure of the solution and compare the micro-
biological data he should be working at an ab-
sorption coefficient that is similar to that of the
untreated waste water. In that case he would at
least be seeing the way the UV is absorbed by the
solution when its absorption coefficient is ap-
proximately the same as it is when he has micro-
organisms present.
MR. ALDRICH: 'hat is correct, basically. What
we started with was the high concentration actinom-
etry and with the concentrations that high the solution
absorbed the UV within a depth'of only 3 to 4 mm.
What we are interested in doing now is lowering the
concentration of the actinometry solution, as you sug-
gested, and we hope to be able to obtain a dosage
which is even more meaningful.
DR. JOHNSON: I think you have to put this
dosage business in context. The way that it has
been classically done and that the standards are
assigned today is so many microwatt seconds per
square centimeter, and as we look at dosage, of
course, the amount of UV added to the water can
be measured just simply on the basis of inches of
lamp and the way the manufacturer rates his
thirty watt bulb, which is essentially what was
done with the practical study in Northwest
Bergen County. The next step is to measure the
real output of the lamp with actinometry, and that'
is what you are measuring. You are not measur-
ing the germicidal dose to any one micro-
organism, because he sees all of the UV absorb-
ing fluid draining. He sees the wall over there that
may be stainless steel; it may be aluminum that is
reflecting the UV off the wall and back to him, and
they are multiple lamps. These units are six and
nine or so lamps apiece, not one lamp, so that you
just cannot take the intensity or distance and
measure the decrease in intensity with distance,
because there are several lamps influencing one
spot within the unit. The classical work that Hill
did was with a two-lamp unit, and that is what the
16,000 microwatt seconds per square centimeter
standard of UV disinfection in drinking water is
based upon. So we are taking this a bit at a time.
Our eventual goal, of course, is to measure how
much UV energy comes to the microorganism,
but we are a long way away from that as you point
out.
DR. ROSEN: I have a question about the
possibility of free radical formation and molec-
ular rearrangements. After all, we are apparently
absorbing UV and disrupting bonds, especially
with high intensity UV. I wonder if anybody has
any comment on this, or did anybody look at this,
or what kind of theoretical probability or possibil-
ity is assigned to it?
MR. VENOSA: Bob Jolley tomorrow will be
126
-------
ULTRA VIOLET LIGHT
giving a presentation where he has evaluated the effect
of chlorine, ozone and ultraviolet light on secondary
and primary effluents. So he will answer your
questions tomorrow.
DR. JOHNSON: The energy is fairly low, 254
nm, not a lot of energy. It is not down to the 180
nm region where you get significant bond disrup-
tion. It would take a fairly weak bond to come
apart at 254 nanometers.
DR. ROSEN: Yes, but you know you
have a spectrum. The 254 nm line is the major
line.
DR. JOHNSON: Well, the mercury lamp puts
out about 90% of its energy or more af the 254 nm
region. There is some energy at 186, and then
there are a few lines up in the visible region, but
they are very small. These lamps are tuned to the
mercury line at 254 nm.
MR. DeSTEFANO, Riddick and Associates:
I would just like to explain my dual nature
here. I am representing a consulting engineering
firm that is also, in turn, representing the Village
of Suffern in New York, which has now a discharge
permit for class A waters. They do not want any
chlorine residual. We looked at various alternatives
and in the most recent facility plan we are recommend-
ing a UV system. But the problem we are facing, as I
see it as a consulting engineer (I am not myself an
engineer; 1 am an undergraduate student at Princeton
helping out for the summer) is when the engineer goes
out and talks to the manufacturers, who seem to be the
only ones up until now who had information, he gets
very conflicting ideas, very conflicting dose
recommendations, flow recommendations, and
geometry recommendations. We talked about it. We
got a little bit into it, thick film, thin film, batch, flow
through, and I am really at a loss on how to handle all
this. I know there are a lot of manufacturers here, and I
won't be afraid to say it but . . . each one has a bad
word for the other, and everyone thinks that they have
the best system, and 1 Was wondering if someone up
there could give us a hint on what sort of things you
want to look for when you are designing the plant. Do
you just want to look at performance, or do you want
to have a margin of safety, or do you want to just look
at your counts, or do you want to look at the dose? I
will leave it at that, and you can play with it from there.
DR. JOHNSON: Well, I am not sure what your
question is, but . . .
MR. DeSTEFA'NO: I will clarify it. Could you
just generally give something that an engineer
could look at and say "These are my design pa-
rameters. This is my effluent quality. I want to have
a certain type of ultraviolet disinfection system
with a certain amount of wattage or a certain
amount of dose or whatever method you want to
use to measure the disinfecting power of the
ultraviolet taking into account geometry",
because obviously there is some difference in
proprietary units available.
DR. JOHNSON: Well, current design is focused
upon dosage, that is, microwatt seconds per square
centimeter. It is conventionally measured in the
literature, or as we have talked about today in terms of
watts per cubic meter, which is essentially the same
sort of thing just on a different basis in terms of power
output from the lamp in germicidal power. In our'
studies we found that there was way too much lack of
concern for flow characteristics. Most of the UV has
been applied to drinking water. Drinking water does
not usually have 106 microorganisms in it, hopefully.
So they are not looking for 104 kinds of orders of kill
like they are in waste waters, and of course if you have a
\% short circuit in your contactor then you are never
going to get much passed the 102 kill. Thus retention
time is very important, and the flow characteristics of
your contactor are very important. There has not
really been enough considered in wastewater design in
disinfection units.
MR. DeSTEFANO: What about the issue of ultra-
violet absorption of the effluent?
DR. JOHNSON: That is, of course, the other side
of the coin. What is the tradeoff between thickness?
One person says you need a quarter of an inch and the
next person says you need three inches. Basically what
you are looking for is UV energy to the micro-
organism. Now, there is a lot of theorizing you
can do, and I think that is where we are at this point. It
is obvious that UV transmission at 254 nanometers is
important, and if you have a high quality of waste
which does not have much absorbanceas they seem to
in Northwest Bergen County, then I think they are
kind of foolish to use a quarter of an inch of lamp
depth, because they are throwing a lot of energy away
by reflections off of the quartz tubes next to each
other. Just heating the unit is where most of their
power is going to go. They are doing a rather
inefficient job when they work with a quarter of an
inch film thickness, and anabsorbanceof .17, but it is
theory you know. We do not really have a design
parameter as yet to do a good job, but that is obviously
very important. Harold said ammonia is a good
127
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
parameter, but of course when you look at sewage
treatment, microorganisms first go after carbon and
then go after nitrogen, and so naturally the amount of
organics is important in terms of the 254 nanometer
absorption, and that follows that ammonia might be a
good way of correlating on a control basis, but a much
more direct basis rather than measuring TOC or
ammonia or COD. It is just a measure of 254
nm absorbance
MR. DeSTEFANO: But what about a plant that is
not running very-well. I know the Northwest Bergen
County plant is running under capacity, and it is
producing a very good effluent, but what about just
standard secondary effluent or bad secondary effluent
such as the other plant in Bergen County, the Bergen
County Sewer Authority plant, which, as I understand
it, is rated at 40 mgd and is running at 70 mgd,
something like that. I think that is where this unit was
initially installed and they were not getting
disinfection at all because it was essentially a primary
effluent. What I want to emphasize or try to get from
you is what kind of effluent characteristics do you need
to make a valid application of ultraviolet light, and
what kind of effluent characteristics coupled with
what kind of geometry? Obviously, I know you do not
want to recommend any particular manufacturer or
any particular geometry, but outside of just ultraviolet
transmission there has to be some other things that
you can look at to try to verify that you are getting the
energy to the organism.
DR. JOHNSON: We had a fairly low dosage at
the beginning of our curve, 10,000 microwatt seconds
or less which is generally rather small, and we were
putting it through wastewater which had suspended
solids in the moderately good range. It was not the
best. This is a contact stabilization plant that runs
pretty much up to capacity, and we were doing a good
job with one inch depth of water through a
conventional kind of unit. I guess what I am trying to
convey is my skepticism that UV would work to begin
at all on wastewater, and surprised that is has worked
as well as it has.
MR. DeSTEFANO: What about something that
you might not be able to measure except by
ultraviolet-absorbance some sort of dissolved iron. I
know in our village of Suffern there are a couple of
industries that might be giving us something funny.
Avon is one of them, and who knows what they are
dumping into the line that might just absorb incredibly
at 254 nm but the effluent characteristics by standard
methods by turbidity would be looking good.
DR. JOHNSON: Yes, you cannot use the TOC,
COD, suspended solids kind of measurements. You
have to measure the 254 nm absorbance because iron is
one of the best UV absorbers, and iron is not measured
by any of the standard procedures. So if you have
some iron coming down from an industrial waste you
could completely destroy UV disinfection.
.MR. VENOSA: Are you ready to hang your hat
on the statement that COD and TOC do not interfere
with UV?
DR. JOHNSON: I did not say that. I said you
cannot just use those alone. You have to use the
parameter that is important interacting with UV. Why
measure TOC when you can correctly measure 254 nm
absorbance. It just does not make any sense.
MR. ELLNER: To change the subject a little bit,
there are two points I would like to make. One is, I do
not want anybody here to get the impression that
photoreactivation is a phenomenon that is limited to
ultraviolet disinfection of organisms, and I think the
point should be made that this repair has been studied
and I think should be studied, and has been
demonstrated with more conventional types of
disinfection as well, and you can apparently establish
increase of organisms through photoreactiviation.
MR. SCHEIBLE: ' Just want to qualify your state-
ment. Photoreactivation is unique to UV. It is due to
the presence of photoreactivating enzymes. I agree
with you that other disinfection procedures should be
looked at for "aftergrowth" or any other mechanism
that increases the effluent quality.
MR. ELLNER: While we are on that subject too, I
think some review should be made of the actual.
practicality of the conditions of photoreactivation
and actual practice of a discharge into a receiving
stream as clear as the Hudson River and things of that
nature which offer tremendous amounts of protection
from visible light available.
MR. SCHEIBLE: I will take issue with that too.
If you look at a number of the streams that the U V or
any disinfected effluent discharges into, those are
small and shallow, and they are normally of high
quality.
MR. ELLNER: The point I was trying to make is
that it has to be related to the practical aspect. This
concept of dosage probably highlights the major
problem with UV from both the theoretical and
practical standpoint. The researchers have no sure fire
way of coming up with a dosage to quantify their
results, and the engineering and specifying people have
no way of describing a U V unit to meet a given design
128
-------
ULTRA VIOLET LIGHT
parameter. Now, there have been attempts to come
close to that, and I would like to offer this because
researchers are starting to report it, instead of
something that is questionable in the measurement.
There is no conventional UV dosage determination, at
least you are not going to get a consensus on people in
the field. There has been an attempt though to relate
results to how many gallons per minute per unit length
of ultraviolet source, and how many seconds retention
time, defining, of course, the output of the U V source,
because that is something that can be duplicated by the
next researcher. I propose that as some thought that
certain people have started to specify ultraviolet on
that basis. I am going to add just one more thing. I do
not want to monopolize this, but I represent an
ultraviolet firm, and I make the statement now that we
do not utilize this conventional estimate measurement
of ultraviolet dosage technique. We have developed
proprietary dye tests for determining ultraviolet, but
we also have used some follow up approaches which I
think we can reveal here, and I think should be under
consideration, and that is a very simple scientific
technique known as the bioassay, and essentially what
is done is that you develop pure strains of organisms,
plot the log reduction in relationship to the U V dosage
which can be determined on a static basis, and then
introduce your organism to the test unit, measure the
log reduction coming out on the other end, and have a
good cross check as to what the dosage delivered was.
MR. WHITE: I have a couple of quick questions I
want to sneak in here. Number one is: you speak of
slime on the lamp, and that is what I heard you say, Dr.
Wolf and Mr. Scheible, and this is where the UV
intensity is highest, right on the lamp. What is the
nature of the slime? Is this bacteria? If it is, why isn't it
killing the slime? That is question number one.
MR. SCHEIBLE: I did not speak of slime on the
quartz tubes themselves.
MR. WHITE: You talked about the wipers. What is
this stuff that you have to wipe off? Is it a zoogleal
slime, or what is it? Does it have organisms in it?
MR. SCHEIBLE: There was an occasion very early
in the study when we were still in the procedure of
equipment shakedown and getting everything working
properly, and it happened that the wipers were turned
off for a period of days so we had no mechanical
wiping system in operation, and what seemed to
develop on it was a thin, white film coating, that
essentially became baked on.
MR. WHITE: Just like a calcium scale deposit?
MR. SCHEIBLE: I do not know. It is bakfd on
simply from the heat of the lamp. When we pulled the
unit up out of the water to take a look at this, we
experimented with several cleaning solutions and
came on one and just sprayed it, turned the wiper on,
the wiper scrubbed it off, put it back in the water. I
think I stated at the end that the wiper mechanism
itself operated over a period of 3,000 continuous
hours, and we have not to date seen any deterioration
in the intensity off the quartz tubes.
MR. WHITE: Okay, I was just curious whether or
not it was some organism. The second question is: do
you generate any ozone in the operation of UV
systems?
MR. SCHEIBLE: Yes, it does in fact. As the study
continues on into the winter, we may have to blow
warm air between the lamps and the quartz sleeves to
maintain the lamps at optimum temperature. We can
then take that exit air, which becomes enriched to a
degree in ozone, make use of it, and inject it into the
front end of the system, and get whatever synergistic
effects may occur. We are not sure what will occur.
DR. JOHNSON: We were interested in the ozone
question, and thought that if there is oxygen in the
water, why don't you get ozone in the water. The
trouble is the water is very good at quenching the UV.
The wavelength that produces ozone is 186 nm. Even if
you can get 186 nm through your solution, which is
doubtful in wastewater, the oxygen that is dissolved is
not capable of producing much ozone. Most of the
ozone comes from the air space that is around the
lamps. These lamps have to be cooled. The operate
very much at an optimum temperature. At slightly
below or slightly above that temperature their
efficiency falls off, so that it may not always produce
the .optimum 40 watt germicidal output. And what it
puts out in the 186 nm region is mainly producing
ozone in the air space.
MR. WHITE: An efficiency of 200 fecal coliforms
per 100 ml, which you spoke about as the goal, is not
considered disinfection in California. It is a long story,
but I just wanted to mention that. The other thing is
that some of the people I talked to in France, who have
been looking into this, suspect that UV will cause
mutagenesis in surviving viruses.
MISS ELLIOTT: I have found in my reading that
there are mutagenic effects of UV light, and as far as
the photoreactive ability of these organisms, it varies
among the organisms of course, but mutagenesis has
been observed.
MR. HEINSOHN: I was an old Princeton hat. Let
me answer this young Princeton hat-back here. First
129
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
his Suffern plant is about fifteen minutes trom where I
live, and that plant is an operational disaster. The first
thing you have to do over there is re-engineer that
plant, because what comes out of it is just about what
goes into it. I was called over to have a look at this
place. I could show you some others in the country too
where the lawns look like putting greens and a few
other things, but what comes out of the plant is
terrible. I just want to answer him this way, and say
that from a manufacturing position of UV equipment
for wastewater, we do have production systems on the
line. Let me tell you all that the target in our business is
to reduce any coliform count down tp a count of less
than 3. That is the target, not 200 per 100 ml but less
than 3. Some states have that spec right now.
MR. DeSTEFANO: Can 1 just respond to save the
face of Suffern? Admittedly the plant is now not
operating at optimum. It was a WPA plant. It is old,
and it is running at, I think, twice it design capacity.
There is also a problem with inflow. There are a lot of
other problems, but that is why we wrote a facility
plan. The plan will be expanded, and we will be
working with what we assume will be a good
secondary effluent, and there is going to be some
tertiary treatment, not filtering, but tertiary treatment
for nitrate removal. I would just like to respond to my
fellow Princetonian and maybe I will get some of these
UV manufacturers again. The reason we got involved
in this is the project engineer told me to get on the
phone and start calling people, and it is a very strange
way to try to get involved in a subject you do not know
much about. I called Mr. Heinsohn. We got his name
from some people up at New York EPA, and we talked
about the plant over the phone. He took it upon
himself to visit the plant and then make his design
recommendations. Really at that point we were just
feeling out manufacturers there. We are a bit further
now, and the facility plan is written, but we are still not
at a point where we have written a spec or anything.
The design still is in the future, and that is why I am
here to find out what sort of design considerations we
have to make before we go into the construction phase.
DR. GREENBERG, Department of Nuclear En-
gineering at The University of Cincinnati: I would
like to offer a suggestion to those who are working
with these photo-chemical reactors. We in chemical
engineering characterize chemical reactors by
residence distribution studies, and I am rather
suprised that I have not heard or seen any reports of,
such studies among the presenters-today. I suggest that
that would be an easy way to characterize the
circulation and distribution around the photo-
chemical lamps. I would also IIKC to commeni uu some
parallel work that I have been involved in using
coherent light for disinfection of various
microorganisms. In scanning from the near UV, 265
nm being the peak absorption for DNA-RNA,
individual species of organisms absorb in a very
characteristic pattern. At least this is our preliminary
analysis, and this extends, I guess, through the visible
and into the near IR. You can promote growth by
irradiating at different wavelengths. You can inhibit
growth at the same wavelength by looking at
fingerprints of different organisms and overlaying
these fingerprints, and irradiating at a selected
wavelength, using again coherent light or laser light,
very selectively. We have had some success in doing
this.
I would also like to call attention to some work that
was done by a colleague at the Applied Physics
Laboratory, Johns Hopkins in Maryland about two
years ago. Dr. John Parker irradiated a sample of
sea water or wastewater in a simple batch experiment
with infrared light from a CC>2 laser, and discovered
that by pressuring the sample with oxygen he was able
to generate the oxygen singlet, which has the same
germicidal activity as ozone. He was very effectively
able to reduce the microorganism count. I suggest
that this is something that perhaps ought to be
considered.
DR. JOHNSON: Is this in water?
DR. GREENBERG: In water, yes. Also the
penetration as 1 understand it at these high
frequencies, that is, in the near U V, is on the order of a
few mm, at most, and I am surprised that we are
talking in terms of inches. I think the absorption is
greatest . . .
DR. JOHNSON: We measure directly the UV
absorbance at 254 nm, the wavelength of interest, and
the absorbance values come out to be relatively small.
In operating an actual plant, your values were up to
three-tenths of an absorbance unit. We did not find
anything like one absorbance unit, which would be
10% of the material going through a one centimeter
thickness. These wavelengths are not all that far down
in the ultraviolet.
DR. GREENBERG: Are you suggesting then that
there is very little absorbance as a function of distance
out to several inches?
DR. JOHNSON: Well, we measured. I am not
suggesting anything. I am telling you what the
measurements are. The measurement said in an actual
130
-------
ULTRA VIOLET LIGHT
'wastewate that the iabsorbance was less than one
absorbance unit, and if you look at Beer's Law
relationships, that tells you that 10% of the material
got through one centimeter of fluid, 90% got absorbed,
and the absorbances were well below one. Most of
them were around 0.17. That is in one cm of the fluid; if
you convert that, say 80% goes through.
DR. GREENBERG: 80% goes through, and this is
wastewater?
DR. JOHNSON: In wastewater,
DR. GREENBERG: Is this information available?
Is it published?
DR. JOHNSON: There is published data. It goes
back to some works that were done by General
Electric. Lukeisch did work back in the fifties or earlier
than that. He.looked at UVabsorbance of wastewater,
drinking water, and a number of different types of
waters in the country. This was typical.
DR. GREENBERG: But as one goes down in
frequency towards the IR, the absorbance, the
penetration let's say is much much greater.
DR. JOHNSON: Yes, if you go down to 200 now
everything absorbs.
DR. GREENBERG: But in the near IR you can
penetrate through several feet of dirty water which is
what Parker did in his study.
DR. JOHNSON: Let me just reinforce something
that was said by the last commentor. There has been
far too little concern about the detention time. George
showed some data on detention time, and talked about
a tj, the initial breakthrough of material inadye study.
Even the better of the two units that we looked at
where the mean detention time was not significantly
different from the volume divided by the flow rate,
the tj value was something like six seconds when the
mean detention time was twenty seconds. Now, that is
•at a level that you can read off this curve of ours which
is maybe at 1%. If you are thinking about wanting to
get four or five log reduction, then you ought to call tj
where you get the first 0.01% or four logs, and that
would be a rather short contact time even for the better
of the two units. So there is far too much concern
about dosage, microwatt seconds per square
centimeter in ultraviolet disinfection of wastewater
today. There has not been enough concern about
detention time, the main point of our work.
MR. WOOD: On that" subject, the detention time
can be two ways. It can be long or it can be short.
We are paying attention to the short retention times,
because that marries to a piece of equipment that
practically fits a sewer plant operation. You do not
want a monstrous piece of equipment that takes tw*o'
acres of land to put it on. What you end up with if you
are talking about long retention time, is hydraulic loss
problems. You get into flow conditions. You get into a
wide gradient of exposure as you go away from the
source. These are some of the things that are of
concern. We promote thin film, short retention time.
So we are paying attention to retention time on the
short side.
One other thing I would like to clear up. It was
brought up here that this particular unit at Northwest
Bergen Sewer Authority was installed at Bergen
County. That is an error. The unit that was installed in
Bergen County was the first prototype unit that we
built. It had around a 6 mgd flow capacity. We did
get performance on that unit at a flow condition where
the plant was twice capacity, and effluent quality did
not meet standards, had suspended solids up in the two
hundred range, color and turbidity up in the fifty to
hundred range. We met the 200 per 100 ml standard in
that installation at a flow rate of about 5 mgd under
.those conditions. At that point, we made the decision
that we just did not want to have that particular system
associated with that type of plant. At the time we were
conducting talks with EPA, they felt that if ultraviolet
was going to be a successful disinfection alternative
you did not really want to evaluate it at a very poorly
run plant. It would mean that you could run your plant
anyway you want, just place a UV unit at the end of it,
and it overcomes anything you might do in the
operation of your plant. That was our reason for going
to Northwest Bergen where we had a more model
plant, and we felt it gave U V a more representative test
also. I would subscribe to the fact that the plant should
be run well and if it is run well, ultraviolet is a
tremendous disinfection means.
The other point I would like to make is that I think
we have the tendency in statistical analysis to look at
log reductions and maybe not relate them to actual
conditions, and I would just like to point out a
layman's approach to the testing we are doing at
Northwest Bergen. We have had ultraviolet
disinfection system in there that has been running
around four months, full plant flow, pretty much
somewhere between 80 and 100% power. In four
months of day-to-day testing, twenty-four hour
operation, we have never on a single test reached a 100
per 100 ml count. Now your curves say that that unit
cannot do it. I want to point out that the operator's
standard is 200 per 100 ml. That is the performance
standard he has to meet, and if the unit does it, in my
131
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
opinion that unit performs. Statistical analysis be
damned!
DR. SCHWARTZ: I have some questions. First of
all, U V is known to induce moleclar rearrangements in
DNA. Ultraviolet is known to induce transformation,
mutagenesis, and carcinogenesis when organisms are
irradiated, and molecular rearrangements do occur in
the DNA, not directly due to the ultraviolet but due to
the repair processes within the cells and viruses. The
thing that I want to bring up that is an important
point is not the photoreactivation, which is a rather
benign sort of repair, but the dark repairs which are
error prone. Has anyone who has run these UV dis-
infection facilities done any transformation studies
of the viruses that they inactivated?
MISS ELLIOTT: As I mentioned earlier, we were
working with just total coliform evaluations, so I have
not.
DR. SCHWARTZ, Deltech: I have some~ques-
tions. First of all, UV is known to induce molecular
rearrangements in DNA. Ultraviolet is known to in-
duce transformation, mutagenesis, and carcinogenesis
when organisms are irradiated, and molecular re-
arrangements do occur in the DNA, not directly due
to the ultraviolet but due to the repair processes within
the ceils and viruses. The thing that I want to bring up
that is an important point is not the photoreactivation,
which is a rather benign sort of repair, but the dark
repairs which are error prone. Has anyone who has
run these UV disinfection facilities done any trans-
formation studies of the viruses that they inactivated?
DR. JOHNSON: I think I would like to comment
about this "super oxygen". We went through a nascent
oxygen hulabaloo with chlorine years ago. These
compounds like singlet oxygen and super oxygen are
very unstable. Their lifetimes are very short. One of the
major problems with disinfection from a
microbiological point of view is to get the disinfectant
to the vital site within the microorganism. In fact,
many disinfection processes, and ultraviolet not being
one of which, are transport limited. The rates of the
disinfection process is more a matter of getting the
disinfectant to the organism than it is getting the
molecular reaction to occur. Ultrav^'pf is not like
that. It is definitely going right to the active site, which
makes ultraviolet kind of unique among disinfectants.
But the tremendous amount of energy required to get a
singlet oxygen or super oxygen is one thing that
mitigates against that approach, and also the other is
getting that form into the microorganism.
DR. OLIVIERI, Jo/ins Hopkins Univ.:\ wonder if
both of the gentlemen from N W Bergen County might
comment on the frequency in which they met the 200
fecal coliform level, taking into account the factor of
ten that they observed for photoreactivation, be it light
or dark reaction. What percentage of the time did you
meet that taking into account reactivation?
-MR. WOOD: I can answer that. We ran our own
photoreactivation studies. We disinfected the
downstream chlorination tank with 5 ppm overnight
dosage of that downstream tank. We then drained that
tank down with the UV in full operation, full power,
and ran until we got zero chlorination in the entire
downstream contact tank. We had a time basis to go
by. We then continued to run 100% power on the UV
and then took ten minute up to seventy minute
samples in the channel, downstream from the UV and
zero chlorination. We found, I think, good correlation
between what Dr. Johnson got and what Karl got. We
got a one log increase basically, 1 to 1.2 logs. Again I
have to relate this back to the real world. At the time
we had 20 fecal coliforms/ 100ml coming out of the
unit with roughly 105 coming in, and we had a 25 count
seventy minutes downstream. Now that, too, is a one^
log increase, but it is not anywhere near 200.
DR. OLIVIERI: A factor of ten increase is going to
bring you from 25 up to 200.
MR. WOOD: We had less than that. We went from
20 to 30.
DR. JOHNSON: Well, of course, we saw some
dechlorination data yesterday that had two logs of
regrowth. There was a question about whether it was
really growing or'whether it was just coming up from
the sludge at the bottom of the tank, but UV is not
unique I guess.
MR. SCHEIBLE: With the data that we are re-
porting I could not make the simple calculation
of percent time that we met the effluent criteria
of 200. I was more concerned about relating that
unit . . .
DR. OLIVIERI: That is going to be the
ultimate . . .
MR. SCHEIBLE: I agree, but most of our
program to date has been to evaluate dosage
levels relative to disinfection efficiency, and in
my mind . . . correct me if I am wrong . . .
when an engineer comes in and he wants to
design a system, he is going to know what the
relative characteristics of his influent are, and he
can then relate a dosage to a log reduction scale.
132
-------
ULTRA VIOLET LIGHT
He can get an idea of what the expected effluent
will be. That has been the thrust of our work to
date. Now, I can go through that calculation, but
it would be a difficult one in that we varied the
operating conditions of that unit considerably.
DR. OLIVIERI: Once you put in the lamps
how much flexibility do you have to overcome a
poor design?
MR. SCHEIBLE: Considerable flexibility. The
unit itself has the capability of shutting off
banks of lamps. Now in my mind the ultimate
design of this would be simply that you rate the
power input to the flow. So you vary your power
with your variation in flow through the day. You
can tone down during low flow conditions at
night or whatever. It also has the ability to turn
off banks of lamps so you can vary . . .
DR. OLIVIER!' Well, the situation that I am
concerned about is if you have to add banks of lamps.
MR. SCHEIBLE: If you design your system
properly that should not happen.
MR. SCHEIBLE: 1 do not know how else to
answer your question. It all comes down basically to a
correct design of the system, and you should be able to
handle what your expected conditions are in the near
future. That is all I can really say. And photo-'
reactivation, to answer that question, from the
data that we showed over a variety of dosage levels, we
got a fairly reasonable correlation which showed an
average of one log increase over effluent conditions.
DR. OLIVIERI: This is precisely what I want to get
at. That is an increase, and you are showing me that
you are reducing to say a fecal coliform level of one
hundred. If that is going to go through a one log
increase, I am going to have, coming out of the plant or
at some point downstream, a fecal coliform level
equivalent to a thousand.
MR. SCHEIBLE: That is correct.
DR. OLIVIERI: You have not met the requirement
then. That is the point I am making.
MR. SCHEIBLE: Let me make two statements on
that. First of all, the UV equipment can achieve a level
which, if you even take into consideration
photoreactivation, will achieve effluent criteria. I
would love to throw it open to the floor here to
anybody who has any comments, and I wanted to
bring it up yesterday and Al did. What is the
implication of this after-growth and what is the
implications of photoreactivation?
DR. OLIVIERI: They are two different processes.
They are two distinctly different things.
MR. VENOSA: It does not make any difference.
There is still the same effect. If you look at Kent
Aid rich's data, he showed that when he achieved three
log reduction he got about a one log
photoreactivation, but when he achieved a four log
reduction he did not get any photoreactivation. So I
think possibly the answer might be simply stricter
•coliform standards.
MR. DeSTEFANO: I think the answer might be
that it is not when you are getting a high ultraviolet
dosage like you were in that second unit that maybe
the mechanism and the site of action is different, and
that it is not the DNA dimeration because if you are
getting photoreactivation obviously it is not a
resurrection happening. The cell never died. It was
possibly altered in some point.
So my concern, somewhat similar to the gentleman
from M IT, is that (and this is sort of my other nature as
a Princeton student interested in public health rather
than an engineer because I am going to throw a pipe
wrench in the works) it is possible, I think
theoretically, that what you are doing with the
ultraviolet, especially at low radiation similar to the
dosage used by biologists to study mutagenesis is you
are somehow just stopping the cell from its ability to
reproduce. It can catch up with that in
photoreactivation. So it starts to reproduce again in
your lactose. Also I know that (I have seen some
references in the literature) one of the first uses of UV
was as a mutagen, and it changed resistance of a
bacterium to an antibiotic. I think it was streptomycin,
and I think it was the same wavelength that was used. I
just saw the reference to the article or the abstract. I
have not read the whole article, so I am not going to
hang my hat on that, but I would suspect (and I do not
think any of the work done in Bergen County or the
work done in North Carolina has taken this into
account) that possibly at a certain dosage of U V, what
you are merely doing or could merely be doing is
inhibiting the cell's ability to ferment lactose. If that is
done it is not going to show up on your test unless you
are doing total counts too, because (I have worked
with the MPN method) if it does not ferment lactose'
you are never going to see it.
. MR. VENOSA: The cell's ability to ferment
lactose is controlled genetically anyway.
MR. DeSTEFANO: But aren'1 Y°u altering the
genetic structure?
MR. VENOSA: Yes
MR. DeSTEFANO: ^ut cou'd vou a'so a'ter the
genetic structure in a way that would inhibit its ability
133
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
to ferment lactose?
MR. VENOSA: Yes you could, and you could
also produce other types of mutations such as
antibiotic resistance too.
MR. DeSTEFANO: I do not know if any of you
are familiar with some other things that I- have come
across in my search of the literature of actions of U V
light, one of them being its action on the membranes of
'the lysozomes and inducing lysis. It also induces lysis
,in another way, and I am not that versed in this, by
releasing the coliphagesin the E. coliand they all come
out from wherever they are hiding and go with the cell.
'Again I just see those references. I think it was
^discussed in Strickberger's Genetics.
MR. VENOSA: I hardly doubt that that is going
on in this system. Microbiology literature is replete
with that type of information. It has been known for
many years, but I do not think that is what is going on
here'.
MR. DeSTEFANO: How would you define its
action?
MR. VENOSA: The mechanism of action for UV
has been known for twenty-five years and it is the
absorption by DNA and RNA. That is the primary
effect. Now, certainly there are other secondary
effects, but the primary effect is the absorption by the
nucleic acid.
MR. DeSTEFANO: Now in doing my homework,
just this section on UV says that you can get thymine
dimeration, the DNA is altered, or you can get also I
think uracil dimeration in the RNA. Now if the
messenger RNA gets through and uses it as a decoder
to produce the protein before the cell actually dies, you
are getting a different code than the one that was
originally in the cell.
MR. VENOSA: You are not going to get
translation if the messenger RNA is blocked.
MR. DeSTEFANO: If you have only produced
one dimer, why can't it go through. What is stopping
it?
MR. VENOSA: By the very fact that the dimer
forms it blocks further replication. You cannot get it.
It is genetically impossible.
MR. DeSTEFANO That is not wnat Dyson
implied. I am sure you know much more about it than
I do.
MR. VENOSA: I think we are getting beyond the
subject of this conference.
Why don't we call for one or more questions from
the audience.
DR. ROSEN:- I would like to ask a question of Mr.
Scheible. I am sorry, it is practical.
You indicated that based on this particular design in
this wastewater that the cost of a system for one
additional log increase reduction would be on the
order of a factor of four from the number that you
calculate. Is that correct? Total cost?
MR. SCHEIBLE: Three and a half. I would also
state that these are preliminary cost estimates.
DR. ROSEN: Okay. This gets back to some of
these questions about where you sample, and whether
you are looking at a sample after photoreactiviat ion or
not in terms of what you count and what vou do not
count. If you are talking about a difference between a
200 fecal coliform, assuming you are starting at the
same level, and a 2.2 total coliform in California, you
are talking about a factor sixteen in the cost or 9, 10 or
12, somewhere in that range, based on these
preliminary estimates, and I think that is significant
and has to be considered in the whole evaluation here.
DR. JOHNSON: I might say that in myjudgement
with that particular wastewater this unit is killing the
gnats with a sledge hammer. They are putting in a very
large dosage of U V into the sewage of the type that was
being treated in Northwest Bergen County. In other
words, I think we are at the stage in ultraviolet
disinfection we were with ozone five years ago. UV is
just getting started, and I think that ultraviolet
disinfection can do a much better job as we learn more
about the process. It is going to get better, and I think
we are going to have trouble if we continue to talk
about ultraviolet disinfection in terms of suspended
solids and conventional kinds of wastewater
measurement parameters. It does not make any
difference whether your suspended solids are 5 or 500.
What is important is how much UV transmission you
have.
DR. HILL, Louisiana Tech. Univ. A couple of
the speakers have mentioned that the presence of iron
inhibits UV disinfection. I am wondering if someone
can put a number on this. I am thinking of a relatively
small community that uses ground water for drinking
water with just chlorination, and it comes out of the
ground close to the public health service maximum of
0.3 mg/1, and there is some contribution in the
distribution system from corrosion and iron bacteria
in some places.
MR. ELLNER: Just as Dr. Johnson mentionu
even with iron you cannot take a number and say 0.3
or 3 ppm of iron is going to inhibit UV performance.
You have to relate it to UV transmission. From a
practical standpoint and not taking up everybody's
134
-------
ULTRA VIOLET LIGHT
time, I will give you a copy of a study done here in
Cincinnati at the Taft Sanitary Engineering Center
examining the effects of iron, color, turbidity, but
relating it to UV transmission. In their studies for your
specific question, three parts per million of iron did not
significantly affect the performance of a specific UV
design parameter. One other point I would like to
leave you with. When you review this discussion, and
you take figures, they are not absolutes. For example,
M r. Scheible's economic figures, his cost figures, relate
to a particular configuration that he studied. Now it is
.possible that there might be certain UV approaches
that can cost three times as much, and there might be
other UV approaches that can cost qne-third. The
suggestion I make is: keep an open mind and realize
what we are discussing here is the broad brush stroke.
Don't either dismiss or accept something based upon
specific statements that are made today. I think we are
creating the background for the picture, but a lot of
things still have to be filled in.
DR. ROSEN: Getting back to using this UV
absorbance as a measure of design or acceptability or
efficiency of use of UV in disinfection, we are still
talking about things like color and turbidity and things
that we see in visible light. There must be a lot
of UV studies done on wastewater historically in
terms of trying to identify compounds, etc. My
question is: Does anybody know if we do have
some sort of range that we might expect based
on lots of different kinds of effluents to try to
get a feeling of where this fits? The things that
have all been studied so far are pretty high
quality relative to nitrification and some of the
other things.
MR. REYNOLDS: Aquafine, Inc.: Most of the
work done, and you .can find this in Photochemistry
and Spectroscopy by Simmons, shows general waste-
water at 254 nm to have an absorption coefficient
between 0.13 and 0.2. Typically that is going to take
you down to about 60% transmission in. a one centi-
meter cell. Now, there is also some recent data in
Photochemistry and Photobiology where they are
doing specific work on absorbance with anilins and
other compounds .which do have a tremendous
absorptivity of UV at 254 nm. To elaborate on the
iron, if you go to 0.3 mg/1 iron, you are going to find
tremendous fouling in the UV system very quickly
because of the thermal differential on the quartz in the
wastewater. Typically it tends to play down on the
quartz jackets, thereby having to have some type of
either mechanical or chemical cleaning in the system.
MR. WARRINER: CH^M Hill: This may not
be a response to Harvey Rosen's comment. I ran
.across proprietary equipment a couple of years ago in
Britain for continuous measurement of TOC, which
involved an absorption cell at 254 nm, and 1 wonder if
it is used at the othersite from the other point of view if
that is not a possible source of this kind of data, and 1
puzzled about that because I continued to hear
suspended solids, other visible light parameters used. I
guess it is a question for Dr. Wolf.
DR. WOLF: I am familiar with the fact that in
German drinking water practice there is some
application made to measuring UV absorbance on a
continuous twenty-four hour basis. 1 am unaware of
this type of application in wastewater. Certainly from
what 1 have found myself this would be a terribly
fruitful path to follow with respect to getting some
kind of handle on the quality of effluent with respect to
organic properties. 1 really think so.
135
-------
SECTION 5. OZONE
17.
COMPARISON OF MPN AND MF TECHNIQUES OF ENUMERATING
COLIFORM BACTERIA IN OZONATED WASTEWATER EFFLUENT
Mark C. Meckes and Albert D. Venosa
U, S. Environmental Protection Agency
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
ABSTRACT
In 1976, an amendment to the Code of Federal Regulations (40 CFR
Part 136) (12) was adopted establishing use of the Most Probable Number
(MPN) Technique as the method of choice for enumerating fecal coliforms
in chlorinated effluents in enforcement situations. From this it would
follow that the MPN test would be the method of choice for all disinfected
effluents regardless of the nature of the disinfectant. This study was
conducted to determine if the more rapid Membrane Filter (MF) Tech-
nique could he used to enumerate eoliform populations in ozonated efflu-
ent. A comparison was made between the nro methods at different applied
ozone dose levels. No significant difference between the two methods
was found regardless of the magnitude of the applied ozone dose. Thus,
either the .MPN or the MF technique mar he used to quantify eoliform
populations in ozonated secondary effluent. However, since the Code of
Federal Regulations specifies the MPN as the method of choice in en-
forcement situations, compliance with the USEPA Environmental Mon-
itoring and Support Laboratory's (EMSL) equivalency testing program
would still be required to get approval to use another method such as
the MF.
INTRODUCTION reach the organisms on the membrane filter by
Several researchers have reported that the diffusion, are presumably less readily available to the
membrane filter (MF) technique significantly organisms to allow them to survive this stressed state
underestimates the true population of coliforms in (7). The MPN presumptive broth medium provides a
chlorinated water and wastewater samples when more favorable environment for sublethally injured
compared with the multiple tube fermentation (MPN) cells since it is non-selective and does not expose cells
procedure (3,4,6,8,11). The reasons for the to the additional temperature stress. (5,9)
discrepancy between the two methods are unknown Ozone is rapidly gaining ground in the U.S. as a
for certain, but one possibility is that the surviving viable wastewater effluent disinfection process
organisms, which have already undergone one alternative to chlorination. Several ozone disinfection
environmental stress (i.e. exposure to chlorine) are systems are on stream with many others under
rapidly subjected to a severe temperature stress when construction or in the design stages. Since the
transferred to the 44.5°C water bath. This, plus the amendment to the Code of Federal Regulations (40
fact that nutrients in the MF medium, which must CFR, Part 136) (12) was adopted establishing use of
136
-------
OZONE
the MPN procedure as the method of choice for
enumerating fecal coliforms in chlorinated effluents in
enforcement situations, it follows that the MPN test
would be the method of choice for all disinfected
effluents regardless of the nature of the disinfectant.
Since the MF technique is rapid, simple, and more
economical than the MPN method, it is desirable to
use it rather than the MPN method if comparability
can be demonstrated. To this end, this study was
performed.
MATERIALS AND METHODS
Media
The MPN and the MF procedures were performed
according to Standard Methods (10). In the MPN
tests, lactose broth (Difco) served as the presumptive
medium and brilliant green bile broth (Difco) and EC
broth (Difco) as the confirmatory media for total and
fecal coliforms, respectively. The media used in the
MF tests were M-Endo agar (Difco) for total
coliforms and M-FC agar (Difco) for fecal coliforms.
Membrane filters used were Gelman GN-6, lot #81347.
Incubation of total coliform MPN's and MF's was
carried out at 35.0±0.5°C, while fecal coliform
MPN's and MF's were incubated at 44.5±0.2°C. All
liquid samples were neutralized with sodium
thiosulfate prior to inoculation to destroy residual
ozone.
Ozone Generator and Contactor
The ozone generator was obtained from Com-
puterized Pollution Abatement Corporation (CPAC).
It was an air-cooled, plate-type ozonator capable of
producing 4 kg of ozone per day from air. Ozone was
produced from air after the air had been dried to a dew
point below -40° C.
The contactor (Figure 1) consisted of three
aluminum columns, each 3.7 m high and 30 cm in
diameter, connected in series and staggered vertically.
Wastewater effluent* was pumped to the top of the
first column at a flow rate of 75 I/min. The liquid
emerged at the bottom of the column and flowed by
gravity through a pipe to the top of the next column.
This flow configuration was repeated to the third
column.
The o/one-air gas mixture was injected at a rate of
75 I, min (25 1 min per column) through ceramic
diffusers at the bottom of each column. The concen-
tration of ozone in the gas was 9.0 mg/1 gas. Since the
gas stream was split equally among three columns,
'From the Robert A I,ill I abimitoi) C'omenlional Acmalcd Sludge I'ilol Plant.
the liquid in each column received an applied ozone
dosage of 3 mg 1. Thus, by the time the liquid emerged
from the third column, a total, cumulative dosage of
9.0 mg 1 had been applied. Residence time in each
column was approximately 3 minutes.
With the above arrangement and the system
operating continuously, we were able to sample any of
the three columns at any desired time without the need
to make adjustments in flow rates or ozone
concentrations.
Analytical Method
Ozone was measured in the gas streams before and
after each column by passing a sample gas stream
through gas-washing bottles containing a 2% potas-
sium iodide (KI) solution and then through a wet
test meter to measure the volume. The liberated iodine
was then titrated with a standardized solution of
0.0500 N sodium-thiosulfate. (2)
Experimental Design
A factorial experiment in randomized block design
was performed to compare the two coliform enumer-
ation methods as a function of applied ozone dosage.
The three dosages chosen (i.e., 3, 6, and 9 mg 1) were
considered fixed variables, since they were chosen
to represent low, intermediate, and high dosage levels.
corresponding to high, intermediate, and low coli-
form numbers, respectively, in the treated effluent.
The order in which samples were collected at given
dosage levels was randomized each day. Samples
were collected at all three dosage levels on any given
day, and thus a day was considered a replicate block
experiment. A total of six replicates were performed.
When a liquid sample was collected (according to
the predetermined randomized dosage order) it was
immediately split into 8 equal fractions, labeled 1
through 8. Each label corresponded to either an M PN
or an MF, as shown in Table I. Two duplicate MPN's
and two duplicate MF's were performed for each
coliform type. Two technicians were utilized to inocu-
late the samples. One was assigned exclusively to
MPN inoculations, while the other was assigned to
MF's. Since it was obviously impossible to inoculate
all eight sample fractions simultaneously, the order in
which the sample fractions were inoculated was
randomized to eliminate error due to standing. All
inoculations were completed within one hour follow-
ing sample collection.
Following completion of the six replicate exper-
iments, separate analyses of variance (ANOVA) of the
data for total coliforms and fecal coliforms were
137
-------
PROGRESS IN WASTE WATER DISINFECTION TECHNOLOGY
GAS SAMPLE
INFLUENT
WASTEWATER (
•GAS I ^A0 T
OUT
1
1
1
1
1
A
1
i
^
SAMPLE f-CXJ-»- SAMPLE | f
1
i f
£ ww
^ SAMPL
E
i
i
1
T
1
f 1
i
1
r i
^_l
f
i
i
4
T
1
1
1
i
i
i
i
( i
i
i
, i
i
V >sl^
s
GAS
SAMPLE
SAMPLE
O3
--- 03
- WW
WW DRAIN
SAMPLE
FIGURE 1. BUBBLE DIFFUSER OZONE CONTACTOR
performed in an IBM 370 computer using Biomedical
Computer Programs BMD02V (Analysis of Variance
for Factorial Design)(l). In each ANOVA there were
three "main effects" under consideration and four
"interaction effects" (Table 2). The Methods Main
Effect (M) tested the overall difference in coliform
numbers between methods. The Dosage Main Effect
(D) tested the overall effect of ozone dosage on
coliform numbers as determined by both methods
combined. The Replication Main Effect (R) tested the
overall effect of days(i.e., experimental replication) on
coliform numbers. The Method-Dose Interaction
Effect tested whether the difference between methods
was consistent over all dosage levels or changed when
different dosage levels were applied. The Method-
Replication Interaction Effect tested whether the
difference between methods was consistent or changed
from day to day. The Replication-Dosage Interaction
effect tested whether the difference between coliform
numbers with respect to dosages was consistent c r
changed from day to day. The Replication-Method-
Dose Interaction effect tested whether the difference
between the method-dose interaction was consis-
tent or changed from day to day.
TABLE 1. LABELING OF SAMPLE FRACTIONS
ACCORDING TO COLIFORM ENUMERATION METHOD
Label Number
1
2
3
4
5
6
7
8
Analysis'
MPN
MPN
MF
MF
MPN
MPN
MF
MF
-TC
-TC
-TC
-TC
-FC
- FC
- FC
- FC
* TC = total coliforms; FC = fecal coliforms.
138
-------
OZONE
TABLE 2. EFFECTS TO BE TESTED
BY ANALYSIS OF VARIANCE
1. Overall difference between methods (M)
2. Effect of dosage (D) on coliform numbers
3. Effect of experimental replication (R) on coliform
numbers
4. Interaction between methods and doses (MD)
5. Interaction between methods and replicates (MR)
6. Interaction between replicates and doses (RD)
7. Interaction between replicates, methods, and
doses (RMD)
RESULTS
Total Coliforms
Table 3 summarizes the differences between the MF
and MPN methods of enumerating total coliform
bacteria in ozonated effluent as a function of applied
ozone dosage. Each number in the MF and MPN
columns is the mean of 12 determinations (averaged
over the 6 days). In Table 4, the same data are
presented to illustrate the differences between the
methods as a function of experiment days. Each
number in the M F and M PN columns is the mean of 6
determinations (averaged over the 3 dosage levels).
The ANOVA is presented in Table 5.
The analysis of variance indicated that the
replication main effect (R) was highly significant
(p<0.01). Thus, mean total coliform numbers,
determined by both enumeration methods over all
dosage levels, were significantly different from day to
day (note row means in Table 4). This finding is not
unexpected, since effluent quality varies widely from
one day to another.
The methods main effect (M) was not significant
(Table 5). Thus, the average number of total coliforms
enumerated by each method over all dosage levels was
similar. The method-replication interaction (MR) was
also not significant. Thus, although total coliform
numbers varied significantly between replicate runs(R
effect), the average difference between the MF and
MPN was not significant from one run to another.
The dosage main effect (D) was highly significant
(p<0.01), indicating, as expected, that applied ozone
dosage affected the final number of total coliforms as
measured by both methods (note row means in Table
3). The difference between methods with respect to
individual dosage levels (method-dosage interaction)
was significant (p<0.05). Thus, although the two
methods gave equivalent results when averaged over
all dosage levels, differences did exist at one or more
individual dosage levels. To determine where the.
differences occurred, the data were re-examined
statistically at each dosage level. It was found that the
MF gave significantly higher results than the MPN at
the 6.0 mg/1 dosage level, but the two methods
produced similar results at the 3.0 and 9.0 mg/1 dosage
levels. From a practical microbiological viewpoint, the
difference noted at the 6.0 mg/1 dosage level was not of
sufficient magnitude to warrant a strict interpretation
of the data. Therefore, it is concluded that either the
MPN or the MF technique can be used to enumerate
total coliforms in ozonated secondary effluent.
The replication-dosage interaction (RD) was also
highly significant, indicating that different effluent
total coliform numbers resulted from the same given
ozone dosage level on different days. This was prob-
ably due to the variable effluent quality on different
days causing differences in ozone utilization efficiency,
TABLE 3. DIFFERENCE BETWEEN TOTAL COLIFORM
METHODS AT EACH OZONE DOSAGE LEVEL
Dosage, mg/l
3.0
6.0
9.0
All Dosages
Log^Q Total Coliforms/100 ml
(Standard Deviation)
MF
5.13
(0.63)
3.97
(0.27)
3.33
(0.22)
4.14
(0.86)
MPN Both Methods
5.31
(0.93)
3.58
(0.32)
3.15
(0.20)
4.01
(1.10)
5.22
(0.78)
3.78
(0.35)
3.24
(0.22)
TABLE 4. DIFFERENCE BETWEEN TOTAL COLIFORM
METHODS IN EACH REPLICATE RUN
Replicate
1
2
3
4
5
6
All Replicates
Log
MF
3.62
(0.61)
3.83
(0.65)
4.32
(0.86)
4.45
(0.90)
4.06
(0.66)
4.59
(1.23)
4.14
(0.86)
1Q Total Collforms/100 ml
(standard Deviation)
MPN Both Methods
3.29
(0.32)
3.92
(0.77)
4.26
(1.25)
4.31
(1.39)
3.84
(1.04)
4.47
(1.44)
4.01
(1.10)
3.45
(0.50)
3.88
(0.68)
4.29
(1.03)
4.38
(1.12)
3.95
(0.84)
4.53
(1.28)
139
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
which in turn affected the final effluent coliform
density. The 3-way interaction between replicates,
methods, and dosages was also significant, indicating
that the method-dose interaction was different from
one day to another. Microbiologically, this effect is
not considered important relative to the conclusions
made above and will not be discussed in this paper.
TABLE 5. ANOVA FOR LOG TOTAL COLIFORM MPN,
MF COMPARISON DATA
Source ol Variation
Replicates (R)
Methods (M)
RM
Doses (D)
MD
RD
RMD
Error
Total
Degrees
ol
Freedom
5
1
5
2
2
10
10
36
71
Sum
of
Squares
9.46
0.30
0.31
50.20
0.96
5.11
1.05
0.90
Mean
Square
1.89
0.30
0.06
25.10
0.48
0.51
0.10
0.02
F
76.06"
4.90
2.48
49.10"
4.59*
20.54"
4.23"
• p<0.05
••p<0.01
Fecal Coliforms
Table 6 summari/es the differences between the M F
and M PN methods of enumerating fecal coliforms in
o/onated effluent as a function of applied ozone
dosage. Each number in the MFand MPN columnsis
the mean of 12 determinations (averaged over the 6
days). In Table 7 the same data are presented to
illustrate the differences between the methods as a
function of experiment days. Each number in the M F
and MPN columns is the mean of 6 determinations
(averaged over the 3 dosage levels). The ANOVA is
presented in Table 8.
TABLE 6. DIFFERENCE BETWEEN FECAL COLIFORM
METHODS AT EACH OZONE DOSAGE LEVEL
Log 1Q Fecal Colltorms/100 ml
(Standard Deviation)
Dosage, mg/l
3.0
6.0
9.0
All Doses
MF
4.12
(0.79)
2.87
(0.26)
2.27
(0.13)
3.09
(0.91)
MPN
4.51
(0.97)
2.77
(0.56)
2.27
(0.29)
3.18
(1.17)
Both Methods
4.32
(0.89)
2.82
(0.43)
2.27
(0.22)
The ANOVA indicated that R was highly significant
(p-O.Ol). Thus, mean fecal coliform numbers
determined by both methods over all dosage levels
were significantly different from day to day (note row
means in Table 7). This again reflects a variabk
effluent quality.
TABLE 7. DIFFERENCE BETWEEN FECAL COLIFORM
METHODS IN EACH REPLICATE RUN
Log 10 Fecal Coliforms/100 ml
(Standard Deviation)
Replicate
1
2
3
4
5
6
All Replicates
MF
2.49
(0.42)
2.84
(0.55)
3.35
(0.96)
3.20
(1.13)
3.09
(0.75)
3.55
(1.32)
3.09
(0.91)
MPN
2.48
(0.74)
2.81
(0.70)
3.56
(1.28)
3.38
(1.49)
3.05
(1.08)
3.83
(1.42)
3.18
(1.17)
Both Methods
2.48
(0.57)
2.82
(0.60)
3.45
(1.08)
3.29
(1.27)
3.07
(0.89)
3.68
(1.32)
TABLE 8. ANOVA FOR LOG FECAL COLIFORM MPN,
MF COMPARISON DATA
Source of Variation
Replicates (R)
Methods (M)
RM
Doses (D)
MD
RD
RMD
Error
Total
Degrees
of
Freedom
5
1
5
2
2
10
10
36
71
Sum
of
Squares
11.44
0.18
0.30
53.90
0.80
7.85
0.42
2.43
Mean
Square
2.29
0.18
0.06
26.95
0.40
0.78
0.04
0.07
F
32.71"
3.00
0.86
34.55"
5.71*
11.14**
0.57
• p<0.05
"p<0.01
The methods main effect (M) was not significant
(Table 8). Thus, the average number of fecal coliforms
enumerated by each method over all dosage levels was
similar. The MR interaction was also not significant.
Thus, although the fecal coliform numbers varied
significantly between replicate runs, the average
difference between the MF and MPN was not
significant from one run to another.
140
-------
OZONE
The dosage main effect (D) was highly significant
(p-cO.Ol), indicating again that applied ozone dosage
affected the final numbers of fecal coliforms as
measured by both methods (note row means in Table
6). The M D interaction was significant, indicating that
differences between methods did exist at one or more
individual dosage levels. To determine where the
differences occurred, the data were re-examined
satistically at each dosage level. The only significant
difference found was at the 3.0 mg/1 dosage level. At
this level the MPN gave the higher result. From Table
6 it is clear that at a dosage of 3.0 mg/1 the difference
between the two methods wasO.39 log units. Although
statistically significant, the difference was too minor to
warrant a choice of one method over another.
Therefore, it is concluded that either method will give
equivalent results when quantifying fecal coliforms in
ozonated secondary effluent.
The RD interaction was also highly significant
(p<0.01), indicating that different effluent fecal
coliform numbers resulted from the same given ozone
dosage level on different days. Finally, the 3-way
interaction between replicates, methods, and dosages
was not significant.
CONCLUSIONS
Results from this experiment demonstrated that no
significant difference exists between the MFand MPN
techniques for enumerating total and fecal coliform
bacteria in secondary effluent disinfected with ozone.
Ozone dosage significantly affected total and fecal
coliform numbers in the treated secondary effluent.
Total and fecal coliform densities following ozonation
varied significantly between replicate experiments,
indicating effluent quality may have affected the
variation in the final effluent numbers.
The difference between methods was not consistent
from one dosage level to another, indicating that
significant differences did exist between methods at
one or more dose levels. However, the magnitude of
the difference was not considered substantial enough
to warrant a choice of one method over another.
Finally, the differences between the MPN and MF
techniques were not affected by day-to-day changes in
effluent quality. From the foregoing analysis, it is
reasonable to conclude that either the M PN or the M F
technique can be used to enumerate total and fecal
coliform bacteria in secondary effluent that has been
disinfected with ozone. However, since the Code of
Federal Regulations specifies the MPN as the method
of choice in enforcement situations, compliance with
the US EPA Environmental Monitoring and Support
Laboratory's (EMSL) equivalency testing program
would still be required to get approval to use another
method such as the MF.
ACKNOWLEDGEMENTS
The able efforts of Messrs. Glen Gruber, Leo
Fichter, and Burney Jackson in constructing the ozone
disinfection system are gratefully acknowledged.
Mr. Richard Butler and Ms. Karen Hoskins
conducted the sampling and assisted in the operation
of the ozone disinfection system.
We are indebted to Mr. Joseph Santner for his
invaluable guidance in planning the factorial
experiment. The able assistance of Mr. E. J. Madison
in computer data handling is gratefully acknowledged.
REFERENCES
I. Biomedical Computer Programs 1973. W. .1. Dixon (Ed.).
Uni\ersit\ of California Press, Berkeley. California.
2. Birdsall, C. M.. A. C. Jenkins, and E. Spadinger 1952. lodo-
metric Determination of O/one. Anal. Chem., 24: 662-664.
3. Bissonnette. G. K.. J. J. Je/.eski, G. A. McFeters, and D. G.
Stuart 1977. Evaluation of Recovery Methods to Detect Coli-
forms in Water. Applied Microbiol.. .?.?: 590-595.
4. Braswell, J. R., and A. W. Hoadley 1974. Recovery of £K-/KT-
ichia coli from Chlorinated Secondary Sewage, Applied
Microbiol.. 28: 328-329.
5. Green, B. L. E. M. Clausen, and W. Litsky 1977. Two-
Temperature Membrane Filter Method for Enumeration Fecal
Coliform Bacteria from Chlorinated Effluents, Applied
Microbiol.. 33: 1259-1264.
6. I.in, S. D. 1973. Evaluation of Coliform Tests for Chlorinated
Secondary Effluents, J. Water Pollut. Control Fed., 45:498-506.
7. Maxcy, R. B. 1970. Non-Lethal Injury and Limitations of
Recovery of Coliform Organisms on Selective Media, J. Milk
and Food Technol., 33: 44544X.
8. McKee. J. E.. R. T. McLaughlin and P. Les Gourgues 1958.
Application of Molecular Filter Techniques to the Bacterial
Assay of Sewage 111. Effects of Physical and Chemical Disinfec-
tion Sewage Ind. Wastes, 30: 245-252.
9. Rose, R. E.. E. E. Geldreich. and W. Litsky 1975. Improved
Membrane Filter Method for Fecal Coliform Analysis.
Applied Microbiol., 29: 532-536.
10. Standard Methods for the Examination of Water and
Wastewater. (14th Ed.) 1975. Amer. Public Health Assn.,
New York. N.Y.
II. Stuart. D. G., G. A. McFeters, and .1. E. Schillinger 1977.
Membrane Filter Technique for the Quantification of Stressed
Fecal Coliforms in the Aquatic Environment. Applied
Microbiol.. 34: 42-46.
12. U.S. Environmental Protection Agency. December I. 1976.
Guidelines Establishing Test Procedures for Analysis of
Pollutants. Ammendment of Regulation 40 CFR, Part 136,
Federal Rentier 40. No. 232: 52780-52786.
141
-------
PROGRESS fN WASTEWATER DISINFECTION TECHNOLOGY
DISCUSSION
DR. RICE. Jacobs Engineering:Mark, I did not
look at your diagram hard enough at the beginning.
Did you have any ozone in the off gases, and, if you
did, then should we be talking ozone dosage, or should
we be talking ozone utilized?
MR. MECKES: We did do off gas analysis on these
experiments, I did correlate them, and at this point in
time I would rather defer the results from that to Al's
talk who will be discussing the relationship between
absorbed ozone and coliform reduction.
DR. RICE: I am not being a nitpicker. I am only
concerned with the final engineering design of the
plant. It probably should be made on the basis of the
amount of ozone you are going to use rather than the
amount of ozone you are actually dosing, so that was
the point.
MR. AIETA, Stanford University: Mark, in your
statistical analysis did you take a look at the relative
precisions of the MPN and the MF methods?
MR. MECKES: I certainly did, Marco, and we
found that for this wastewater there was no difference
between the precisions.
MR. AIETA: As measured by the F test?
MR. MECKES: Measured by the T-test.
MR. WHITE: I have a practical question. I cannot
for the life of me understand why we are worrying
about the membrane filter technique versus the MPN
when we do not know anything about what the effects
are of the coliform organisms in the effluent insofaras
disease is concerned, and I would just like to implant
that idea here, because I think we are getting far away
from the practical application of disinfection. You do
not have to answer that question. I do not think there
is an answer, but I would just like to point it out.
MR. VENOSA: How do you measure disinfection,
George?
MR. WHITE I have no way of measuring disin-
'fection except going through all the rigamarole that
the State of California did beginning in 1920 to try to
find out what did represent a threat to public health
and what did not represent threat to public health. If
you read what I write, if you read what other people
write, they claim that in surface water, if the total
coliform count was in excess of 1,000 per 100 mililiter,
this was the factor they used to establish that it was
pollution and there should be something done about
it. But the only work I know is Stevenson's, who did
some work on the threat of disease and how it affected
public health. So I cannot understand why we are
wasting time on the membrane filter technique versus
the MPN because it is so imprecise.
MR. VENOSA: The membrane filter technique is
much more precise that the MPN (except in the
ozonated effluent study we found that they were
similar). The only reason we compared the two
methods was basically because of our ozone contactor
study. We did not want to use MPN's because it takes
four days to get a result, while it only takes twenty-four
hours to get a result with the MF. We wanted, first of
all, to see if there were any differences between the two
methods so that then we could use the M F method and
do more experiments in a shorteramount of time,and
spend fewer research dollars to determine the
differences between ozone contactors. The only reason
we are doing that is to establish a methodology for us
to use. We all agree that there are limitations to
coliforms, and what their significance is in relation to
disease transmission. We are merely using them as we
always have in sanitary engineering to indicate the
possible presence of pathogens, and we cannot
measure pathogens. There are no precise methods of
measuring pathogens in the environment. We have to
rely on the fecal coliforms and the total coliforms to
indicate the presence of those pathogens, and that is
why we are doing it. The only way that we can reliably
measure disinfection efficiency today is by measuring
the coliform numbers.
MR. WHITE: Well, you have answered the
question. The membrane filter technique gets the
answer fast since it is imprecise insofar as coliform'
numbers versus public health. Then your membrane
filter technique is the answer. It does not take so long.
DR. REYNOLDS,Utah State Univ: In re-
sponse to George's question, as Al knows, we ran
comparisons of MPN and MF tests in the lagoon
study that I reported on yesterday morning, and we
did find a substantial difference between coliforms
enumerated by MPN and MF. It was not as sound a
statistical test as was just described here, that is, it was
not a randomized block statistical test, but it was
splitting samples, identical samples, and running them
with various technicians who happened to be there in
connection with the project (who were competent by
the way), and we did see a significant difference
between those two methods. It may be characteristic
of the lagoon effluent we were working with as com-
pared to an ozonated effluent, but we found differ-
ences. 1 think it is important that this work has been
done to substantiate the validity of the research that is
going to follow in the next paper.
142
-------
OZONE
MISS ELLIOTT: I was just curious if you were
using an enrichment broth prior to the full
incubation. If you did, how significant was that
effect?
MR. MECKES: We did not bother to consider
that, especially since the results from this study
showed that we did not have any significant difference
by just using the one step technique, so therefore we
found no enhancement through enrichment would be
necessary anyway. Why go through the additional
effort when one does not have to?
143
-------
18.
COMPARATIVE EFFICIENCIES OF OZONE UTILIZATION AND
MICROORGANISM REDUCTION IN DIFFERENT OZONE CONTACTORS
A. D. Venosa, M. C. Meckes, E. J. Optaken, and J. W. Evans*
Wastewater Research Division
U. S. Environmental Protection Agency
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
'Dept. of Statistics, University of Kentucky, Lexington, Kentucky 40506
INTRODUCTION
This project was initiated to evaluate various
types of ozone contacting devices operating in
parallel on the same source of secondary effluent,
so that the disinfection efficiency of ozone can be
optimized and operating costs defined and quanti-
fied. The types of contactors being studied in this
project were selected to be representative of vari-
ous generic devices commericially available or in
widespead use in the chemical engineering field.
It was the intent of the project to compare dif-
ferent contactor types on the basis of log total
and fecal coliform reduction and ozone utilization
under controlled conditions.
At the International Ozone Institute's Sympo-
sium on Advanced Ozone Technology held in
Toronto, Ontario, Canada on November 16-18,
1977, Scaccia and Rosen stated: ". . . properly
designed ozone contactors show similar perfor-
mance such that there is no 'best' state-of-the-art
ozone contactor." (7) They went on to say that
any generic type contactor can be optimally de-
signed to achieve the same degree of mass trans-
fer, ozone disinfection, and ozone utilization at
approximately the same applied ozone dose. The
selection of a contactor then becomes one pri-
marily of equipment economics, which are
affected by contact time, materials, and energy
trade-offs.
This paper will present evidence that there can
be a "best state-of-the-art ozone contactor" when
one considers and understands all factors respon-
sible for the operating characteristics of the
specific contactors. For purposes of understanding
this and subsequent discussions, it is necessary to
start off with sorne fundamental definitions
of terms.
1. Applied Ozone Dose (D)
D = Y1(QG/QL)
(D
where Y, = concentration of ozone in the carrier
gas, mg 03/lgas
QG = carrier gas flow rate, lgas/min
QL = liquid flow rate, l]ici/min
By inspection of equation 1, it is clear that
the applied dose can be varied either by changing
the ozone concentration in the inlet carrier gas
(Y,) or by changing the QG/QL ratio.
2. Percent Ozone Utilization (%U)
%U =
Y,-Y,
(100)
(2)
where Y2 = concentration of ozone in the gas
leaving the contactor, mg O3/lgas.
3. Actual Ozone Utilization (U) or Absorbed Dose
U = D x Fraction Utilized
= QG/QL (Y,-Y2) (3)
In this phase of our investigation, three contac-
tors were studied: 1. a packed column (PC);
2. a positive pressure injector (PPI); and 3. a
144
-------
OZONE
bubble diffuser (BD). The above contactors were
all designed and fabricated in-house, with the
mixing head for the PPI provided by the Union
Carbide Corporation.
MATERIALS AND METHODS
Source of Wastewater
Secondary effluent was obtained from the con-
ventional activated sludge pilot plant at the
Robert A. Taft Laboratory, Cincinnati, Ohio.
Wastewater entering the plant was of municipal
origin. Effluent from the final clarifer was fil-
tered through a multi-media pressure filter
(Baker Filtration Company, Model HRC-3OD)
and then split equally between two of the contac-
tors for parallel evaluation. Total flow rate was
150 1/min (75 1/min per contactor).
It was found in an earlier phase of the project
(8) that unusually high pressure drops occurred
over the length of the packed column contactor
when effluent was pumped through it. The pres-
sure drops were high enough (greater than 75 cm
H2O) to cause flooding of the column. To rectify
this situation, a small amount of antifoam mate-
rial (Dow-Corning Antifoam emulsion H-10) was
added to the filtered secondary effluent ahead of
the packed column. Only this contactor received
antifoam. This eliminated foaming, which was
causing the liquid hold-up in the packed column,
and the pressure drops returned to normal (less
than 20 cm H2O).
Ozone Concentration
Ozone was generated from air in a plate-type,
corona discharge ozonator (Computerized Pollu-
tion Abatement Corporation Model OZ-180-G).
Gas from the generator was split into two lines,
each with a rotameter and valve for flow control.
These lines served as ozone feed to the contactors.
The carrier air was filtered through a prefilter
(Pall Trinity Micro Corporation Model No.
MCC1001SV160) and then dried to a dew point
of less than -40°C by an activated alumina
dryer (Pall Trinity Micro Corporation 25 HA1).
The dew point was monitored by a Shaw Mini-
Hygrometer, which was calibrated against an
Alnor Dew Pointer (Model 7300, Alnor Instru-
ment Co., Niles, 111.). Carrier air leaving the
drying unit was filtered through an afterfilter
(Pall Trinity Micro Corporation Model No. MC-
C1101EC12) prior to entering the ozonator.
Ozone Contactors
Figure 1 is a schematic diagram of the packed
column contactor. This was a 230 mm diameter
glass column packed with 3.1m of 12mm ceramic
intalox saddles. A teflon redistributor plate was
located midway in the column. Filtered secondary
PACKED COLUMN OZONE CONTACTOR
ww •
INFLUENT WASTEWATER
EXHAUST GAS SAMPLE
•oo-L
EFFLUENT SAMPLE
TO DRAIN
FIGURE 1. PACKED COLUMN OZONE CONTACTOR
effluent entered the top of the column and exited
at the bottom. Mean residence time of the secon-
dary effluent was approximately 35 seconds (mea-
sured by a dye study) at a liquid flow rate of
75 1/min and a gas flow rate of 40 1/min.
Ozone, injected at the bottom of the column,
flowed upward countercurrent to the secondary
effluent and exited at the top. A gas sample tap
was located on the exhaust line. Effluent from
the PC was directed to two 200-liter covered hold
tanks connected in series to allow an additional
five minutes contact time with the ozone residual.
Liquid sample taps were located as shown in
Figure 1.
Figure 2 is a schematic diagram of the PPI
contactor. Ozone enriched gas was injected into
the influent liquid flow in a specially designed
"tee-mixing chamber" under a slight positive
145
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
INFLUENT
>
— *- WW !
— 03
'
1
1
t
t
t
i
U
r
r
i
t
t
i
I)
EXHAUST
GAS
SAMPLE
1
'EFFLUENT
TO DRAIN
FIGURE 2. PRESSURE INJECTOR CONTACTOR
pressure of 50 to 60 kPa. The gas-liquid mixture
flowed concurrently downward through an inner
cylinder. Upon reaching the bottom of the inner
cylinder, .the mixture reversed direction and
flowed upwards through a larger cylinder. At the
top, the gas and liquid separated, the gas flowing
to the exhaust gas line and the liquid flowing
downwards through a second concentric cylinder,
finally emerging at the bottom to a drain line.
Mean liquid residence time in the system was
approximately 15 seconds (measured by a dye
study) at a liquid'flow rate of 75 1/min and a
gas flow rate of 40 1/min.
Figure 3 is a schematic diagram of the bubble
diffuser (BD) contactor. It consisted of three
aluminum columns, each 3.7m high and 300mm
in diameter, connected in series by PVC piping.
The three columns were staggered vertically so
that secondary effluent, which has been pumped
to the top of the first column, could flow by
gravity to each of the two successive columns.
The ozone enriched gas was injected through
domed ceramic diffusers (Norton Chemical
Process Products Division) located at the bottom
of each column and connected in parallel to a
common header pipe. The first column received
50% of the total gas flow, while the other two
columns received equal fractions of the remaining
50%. Gas and liquid sample taps were located as
shown in Figure 3. Since the BD was operated as
a system, treated effluent samples were collected
at the sample tap of the last column only, while
exhaust gas samples were collected after the
exhaust gas from each column had been com-
bined into a common line (see Figure 3). Mean
liquid residence time in the BD contactor, mea-
sured by a dye study was 9.4 minutes at a liquid
flow rate of 75 1/min and a gas flow rate of
40 1/min.
Sampling
All effluent samples were grab samples and
were analyzed for total and soluble chemical oxy-
gen demand (TCOD and SCOD, respectively),
total organic carbon (TOC), total Kjeldahl nitro-
gen (TKN), ammonia-nitrogen (NH+-N), nitrate-
nitrogen (NO^-N), nitrite-nitrogen (NO^-N),
total suspended solids (TSS), turbidity (turb),
temperature, and pH by Standard Methods (1) or
U.S. EPA approved methods (6). Ozone concen-
tration in the inlet and exhaust gas was deter-
GAS SAMPLE
INFLUENT
WASTEWATER
£HX}-»- GAS ~~l
^-^^SAMPLE
GAS
te- Oj OUT
I
I M^ GAS
• ^ SAMPLE
WW
SAMPLE
*
i
r 1
1 1
1
i 1
V
T
i
!
WW
SAMPLE
O3 IN)
Oj
-WW
FIGURE 3. BUBBLE DIFFUSER OZONE CONTACTOR
mined iodometrically by the method of Birdsall,
Jenkins, and Spadinger (3), and a wet test
meter was used to measure gas volumes.
Samples collected for bacteriological analysis
were assayed for total and fecal coliforms by the
standard Membrane Filtration (MF) method (1),
using 0.45 pm GN-6 membrane filters (Gelman
Instrument Company.). Media for total and fecal
coliforms were M-Endo Agar (Difco) and M-FC
146
-------
OZONE
Agar (Difco), respectively. Plates were incubated
in water baths adjusted to 35±0.5°C for total
coliforms and 44.5±0.2°C for fecal coliforms.
Colonies of total coliforms were randomly
picked and inoculated into Lauryl Sulfate Tryp-
tose broth (LST) (Difco) in fermentation tubes
and incubated for 24 to 48 hours at 35±0.5°C
for confirmation. Positive confirmation was indi-
cated by the production of gas in the fermenta-
tion tubes. Fecal coliform colonies were randomly
picked, inoculated into LST medium, and incu-
bated for 24 to 48 hours at 35 °C. A loopful of
suspension from gas-positive tubes was transferred
to EC broth (Difco) in fermentation tubes and
incubated for 24 hours at 44.5±0.2°C. Positive
confirmation was indicated by production of gas
in the EC medium at the end of the incubation
period.
During this study a total of 561 total coliform
colonies were picked from the membrane filters
used on effluent and ozonated effluent. Of these,
488 were confirmed to be total coliforms, result-
ing in a verification percentage of 87%. The
percent verification of fecal coliforms was 95%
(538 confirmed out of 566 isolates picked).
Experimental Design
Testing of the different contactor types was
performed using a split plot design where the
whole plots were arranged in a balanced incom-
plete block design. A discussion of this type of
design can be found in Federer (4). According to
the design, two contactors per day were set up in
parallel. Secondary effluent was pumped to each
ozone contactor at a rate of 75 1/min. The
ozone-air mixture was fed to the contactors at
various gas flow rates and ozone concentrations.
Three different pre-selected dose levels were used.
The three contactors formed the whole plot
portion of the design. The whole plots were
arranged in a balanced incomplete block design
due to the physical restriction that only two of
the contactors could be set up in parallel at any
given time. Thus, two contactors per day were
tested with both contactors receiving the same
applied dose at any given time. All three dose
levels were used each day. The order of the dose
levels was balanced so that each dose level was
used first, second, and third in a day the same
number of times for each contactor. This was
done to minimize the effect of any trend in the
secondary effluent over the day. Randomization
consisted of randomizing the order in which the
pairs of contactors were compared.
Thus, an entire experiment (block), encom-
passing two contactors and all three dose levels,
was conducted in one day. The dependent vari-
bles (i.e., performance criteria) forming the basis
of comparison of the contactors were total and
fecal coliform log reduction (Iog10 N0/N, where
N0 = initial coliform number, and N = final
coliform number), ozone utilization, and percent
ozone utilization. Separate analyses of variance
(ANOVA) were performed to compare the con-
tactors on the basis of each of the four perfor-
mance criteria. The typical form of each ANOVA
was as follows:
Source of
Variation
Degrees of Sum of Mean
Freedom Squares Square F
Whole Plot Analysis
Blocks (Bj)
(adj. for contactors)
Contactors (Cj)
(adj. for blocks)
Whole plot error
Split Plot Analysis
Doses (Dk)
C.D.,
J *
BiDk
BiCjDk
2
4
16
14
Total
53
The contactor main effect tested the overall dif-
ference in performance in the contactors. Its F-
value was computed using the mean square for
contactors ratioed against the mean square for
the whole plot error. A significant contactor
effect led to a comparison of the contactor
means adjusted for blocks using Tukey's HSD
multiple comparison procedure (5). The dose
main effect tested the effect of different ozone
dose levels on the combined performance of the
contactors. The dose F-value was the ratio of the
dose mean square to the block-by-dose mean
square. The contactor-by-dose interaction tested
whether the effect of the contactors was the same
or different for each dose level. The contactor-
by-dose interacton F-test was the ratio of the
147
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
CjDk mean square to the B|CjDk mean square. A
significant contactor-by-dose interaction meant
that the effect of the contactors had to be evalu-
ated separately for each dose level. Finally the
block-by-dose interaction was tested using the
BjDk mean square divided by the BjCjDk mean
square. This tested whether the effect of dose
levels was consistent or changed from day to day.
Data analyses were facilitated by an IBM com-
puter, using the Statistical Analysis System (SAS)
General Linear Model (GLM) procedure (2).
RESULTS AND DISCUSSION
Effect of Increasing QG/QL at a Constant \l
The objective of the first balanced incomplete
block experiment was to evaluate the effect of in-
creasing the applied ozone dose to each contactor
by increasing the QG/QL ratio while maintaining
a constant inlet ozone concentration (Y]). Secon-
dary effluent flow rate (QiJ to each contactor
was 75 1/min and the gas flow rates (Q^) were
25, 50, and 75 1/min, corresponding to QG/QL
ratios of 0.33, 0.67, and 1.00, respectively. The
concentration of ozone in the inlet gas was
maintained at 10.0 mg/lgas (approximately 0.8
weight percent). Thus, the three applied ozone
dose levels studied were 3.3, 6.7, and 10.0 mg/1.
This experiment was conducted during the period
of June 22 to July 21, 1978.
Quality of Wastewater
In Table 1, the minimum, mean, and maximum
values of the physical-chemical and bacteriological
characteristics of the filtered pilot plant effluent
prior to ozonation are presented. Quality was
typical of a well-treated secondary effluent. Note
that the total coliform density was approximately
four times the fecal coliform density.
Percent Ozone Utilization
In Table 2, the percent ozone utilization data
for the first experiment are summarized. In Table
3, the analysis of variance (ANOVA) for those
data is presented. In the whole plot analysis, the
'ANOVA revealed that the Block Main Effect
(Bj) was not significant, indicating that very little
day to day variation in percent ozone utilization
occurred in the three contactors. The Contactor
Main Effect (Cj) was highly significant (p<0.01).
This indicated that the mean overall percent
ozone utilization (i.e., averaged over all 3 dose
levels) was significantly different in at least two
TABLE 1. SUMMARY OF FILTERED PILOT PLANT
EFFLUENT CHARACTERISTICS PRIOR TO OZONATION
(JUNE 22 to JULY 21, 1978)
Parameter
TCOD, mg/l
SCOD, mg/l
TOG, mg/l
TSS, mg/l
NHJ'-N, mg/l
TKN, mg/l
Turbidity, JTU
pH
Temperature, °C
Total Conforms/
100 mi-
Fecal Coliforms/
100 ml"
Mean
35
31
11.5
3.8
13.4
17.1
2.2
—
22
2.6 x 106
(geometric
mean)
6.5 x 105
(geometric
mean)
Minimum
22
22
7.7
1.4
7.7
10.0
0.6
7.2
20
2.4 x 105
1.4x 105
Maximum
46
38
16.5
7.2
19.2
28.7
3.8
8.3
23
7.9 x 106
2.0 x 106
" Adjusted for 87% colony verification.
"Adjusted for 95% colony verification.
TABLE 2. MEAN PERCENT OZONE UTILIZATION AS A
FUNCTION OF APPLIED OZONE DOSAGE
(INCREASING QG/QL, CONSTANT Y,)
Mean Percent Ozone Utilization
(Standard Deviation)
Dose, mg/l
3.3
6.7
10.0
Mean % U,
All Doses
Positive
Pressure
Injector
77
(2)
53
(3)
41
(3)
57
(16)
Packed
Column
85
(4)
58
(2)
46
(3)
63
(17)
Bubble
Difluser
90
(2)
86
(2)
81
(3)
86
(5)
Mean % U,
All
Contactors
84
(6)
66
(15)
56
(18)
of the three contactors (note column means in
Table 2). To determine where the significant
differences occurred, Tukey's HSD multiple com-
parison test (5) was used. Results indicated that
overall percent ozone utilization was highest in
the bubble diffuser, followed by the packed
column, and then the positive pressure injector.
The differences were significant between all three
contactors (p<0.05).
In the split plot analysis, the ANOVA revealed
that the Dose Main Effect (Dk) was highly signi-
ficant (p <0.01). This indicated that the percent
148
-------
OZONE
ozone utilization in all contactors combined was
significantly different at each dose level .studied
(note row means in Table 2). As dose was in-
creased, percent ozone utilization in all contactors
decreased. The Contactor-Dose Interaction (CjDk)
was also highly significant (p< 0.01, Table 3),
indicating that the differences in percent ozone
utilization between contactors were not the same
at each dose level. To facilitate visualization of
these effects, the data from Table 2 are depicted
graphically in Figure 4. It is clear that the rate
of decrease in percent ozone utilization, as a
function of applied dose, was substantially less in
the bubble diffuser than in either the packed
column or the positive pressure injector. When
the effects of the contactors were re-examined
statistically at each individual dose level, it was
found that the contactor main effect was signifi-
cant even at the lowest dose level, and that the
bubble diffuser significantly outperformed the
packed column, which in turn outperformed the
positive pressure injector at each dose level studied.
Finally, the Block-Dose Interaction (BjDk) was
not significant (Table 3), indicating that the effect
of dose levels on percent ozone utilization in all
contactors was consistent from day to day. The
paired percent ozone utilization data on each
experiment day (Block) are plotted in Figure 5, to
aid visualization of the performance differences
between contactors and the performance consis-
TABLE 3. ANOVA FOR PERCENT OZONE UTILIZATION
DATA
(INCREASING QG/QL, CONSTANT Y,)
Source of Variation
Degrees of Sum of Mean
Freedom Squares Square F
Whole Plot Analysis
Blocks (B|) 8
(adj. for contactors)
Contactors (Cj) 2
(adj. for blocks)
Whole Plot Error 7
63.12 7.89 2.20
1911.44 955.72 267.01*
25.06 3.58
Split Plot Analysis
Doses (Dk)
CjDk
BiDk
B|CjDk
TOTAL
2
4
16
14
53
7433.33
1694.56
73.19
34.93
3716.67
423.64
4.57
2.49
812.55*
169.81*
1.83
O BUBBLE DIFFUSER
A PACKED COLUMN
O POSITIVE PRESSURE INJECTOR
FIGURE 4. MEAN PERCENT OZONE UTILIZATION AS A
FUNCTION OF APPLIED OZONE DOSE
g
~~ra
O
1 1 1 1 1 T
00°
'Significant at the p<0.01 level.
EXPERIMENT DAYS
FIGURE 5. OVERALL MEAN PERCENT OZONE
UTILIZATION PER REPLICATE RUN
tency of each contactor from day to day. The
consistency in the percent ozone utilization data
was likely due to the relative lack of change in
wastewater quality over the duration of the study
period.
Actual Ozone Utilization (Absorbed Dose)
In Table 4 the actual ozone utilization data
(i.e., the amount of ozone actually transferred to,
but not necessarily consumed by, the wastewater)
from the first experiment are summarized. The
ANOVA for those data are presented in Table 5.
In the whole plot analysis, the ANOVA revealed
that Bj was not significant, again indicating very
little change in overall ozone utilization in all
contactors from day to day. Cj was highly signi-
ficant (p< 0.01). Thus, the mean overall ozone
utilization was significantly different in at least
two of the three contactors (note column means
in Table 4). Using Tukey's HSD multiple com-
parison test, actual overall ozone utilization was
highest in the bubble diffuser, then the packed
column, and finally the positive pressure injector.
149
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 4. MEAN OZONE UTILIZATION AS A FUNCTION
OF APPLIED OZONE DOSE
(INCREASING QG/QL, CONSTANT Y,)
Mean Ozone Utilization, mg/l
(Standard Deviation)
Positive
Pressure Packed
Dose, mg/l Injector Column
3.3
6.7
10.0
2.5
(0.1)
3.5
(0.2)
4.0
(0.2)
Mean U, mg/l, 3.3
All Doses (0.6)
TABLE
2.
(0.
3.
(0.
4.
(0.
3.
(0.
8
1)
8
1)
5
3)
7
8)
Mean U, mg/l
Bubble All
Ditfuser Contactors
3.0
(0.1)
5.7
(0.2)
7.9
(0.3)
5.5
(2.1)
5. ANOVA FOR OZONE UTILIZATION
(INCREASING QG/QL, CONSTANT Y,)
Source of Variation
Degrees
of
Freedom
Sum
of
Squares
Mean
Square
2.8
(0.2)
4.4
(1.0)
5.2
(1.7)
DATA
F
Whole Plot Analysis
Blocks (Bj)
(adj. for contactors)
Contactors (Cj)
(adj. for blocks)
Whole Plot Error
-------
OZONE
Total Coliform Log Reduction (TCLR)
In Table 6, the total coliform log reduction
(TCLR) data from the first experiment are sum-
marized. The ANOVA for those data are pre-
sented in Table 7. In the whole plot analysis, the
only significant effect noted was Cj, indicating
that mean TCLR was significantly different in at
least two of the three contactors (note column
means in Table 6). Tukey's HSD multiple com-
parison test of the data revealed that overall
mean TCLR was higher in the bubble diffuser
than in either the packed column or the positive
pressure injector (p < 0.05). However, the overall
mean TCLR in the latter two contactors was not
significantly different. Thus, the significant Cj
effect in Table 7 was due entirely to the higher
TCLR in the bubble diffuser contactor. Further-
more, even though the actual ozone utilized in
the packed column was significantly higher than
in the pressure injector (Table 5), the increased
utilization was not sufficient to cause a noticeable
difference in total coliform reduction between the
two contactors.
TABLE 6. MEAN TOTAL COLIFORM LOG REDUCTION
(TCLR) AS A FUNCTION OF APPLIED OZONE DOSE
(INCREASING QG/QL, CONSTANT Y^
TCLR
(Standard Deviation)
Dose, mg/1
3.3
6.7
10.0
Mean TCLR,
All Doses
Positive
Pressure
Injector
2.90
(0.36)
2.94
(0.44)
3.04
(0.34)
2.96
(0.36)
Packed
Column
2.85
(0.36)
3.13
(0.10)
3.31
(0.11)
3.10
(0.29)
Bubble
Diffuser
3.12
(0.10)
3.56
(0.25)
4.52
(0.55)
3.73
(0.69)
Mean TCLR
All
Contactors
2.96
(0.31)
3.21
(0.38)
3.62
(0.75)
In the split plot analysis, the ANOVA revealed
that both the Dk and the CjDk interaction were
highly significant (p< 0.01). Thus, the applied
ozone dose significantly affected the magnitude of
the total coliform reduction in all contactors, and
the differences in TCLR between contactors were
not the same at each dose level. These effects are
easily visualized by examining Figure 8. Note the
apparent differences in slopes of the three TCLR
curves. When the data were re-evaluated at each
individual dose level, it was found that: 1. there
was no significant difference in TCLR between
O BUBBLE DIFFUSER
A PACKED COLUMN
D POSITIVE PRESSURE INJECTOR
FIGURE 8. MEAN LOG TOTAL COLIFORM REDUCTION
AS A FUNCTION OF APPLIED OZONE DOSE
the bubble diffuser, the packed column, or the
positive pressure injector at applied ozone doses
of 3.3 and 6.7 mg/1; 2. at an applied ozone dose
of 10.0 mg/1, TCLR was significantly higher in
the bubble diffuser than in the packed column,
and higher in the packed column than in the
positive pressure injector (p < 0.05). The geometric
mean total coliform densities in the three contac-
tors after 10.0 mg/1 applied ozone dose were:
160/100 ml in the bubble diffuser, 1300/100 ml
in the packed column, and 2900/100 ml in the
positive pressure injector.
TABLE 7. ANOVA FOR TOTAL COLIFORM LOG
REDUCTION DATA (INCREASING QG/QL, CONSTANT Y,)
Source of Variation
Degrees Sum
of of
Freedom Squares
Mean
Square
Whole Plot Analysis
Blocks (B,) 8
(adj. for contactors)
Contactors (Cj) 2
(adj. for blocks)
Whole Plot Error 7
Split Plot Analysis
0.68 0.086 1.81
1.13 0.57 12.00*
0.33 0.047
Doses (Dk)
C: Dk
BiDk
BicjDk
TOTAL
2
4
16
14
53
4.08
2.76
0.86
0.89
2.04
0.69
0.054
0.063
37.75*
10.87*
0.85
'significant at the P<0.01 level.
Fecal Coliform Log Reduction (FCLR)
In Table 8, the fecal coliform log reduction
(FCLR) data from the first experiment are sum-
marized. The ANOVA for those data are pre-
sented in Table 9. In the whole plot analysis, the
block main effect (Bj) was highly significant
151
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 8. MEAN FECAL COLIFORM LOG REDUCTION
(FCLR) AS A FUNCTION OF APPLIED OZONE DOSE
(INCREASING QG/QL, CONSTANT Y^
FCLR
(Standard Deviation)
Dose, mg/l
3.3
6.7
10.0
Mean FCLR,
All Doses
Positive
Pressure
Injector
2.99
(0.38)
3.08
(0.59)
3.26
(0.13)
3.11
(0.40)
Packed
Column
3.08
(0.40)
3.16
(0.42)
3.34
(0.46)
3.19
(0.42)
Bubble
Diffuser
3.22
(0.16)
3.71
(0.24)
4.29
(0.36)
3.74
(0.51)
Mean FCLR,
All
Contactors
3.10
(0.33)
3.32
(0.50)
3.63
(0.58)
TABLE 9. ANOVA FOR FECAL COLIFORM LOG
REDUCTION DATA (INCREASING QG/QL> CONSTANT
Source of Variation
Degrees Sum
of of
Freedom Squares
Mean
Square
Whole Plot Analysis
Blocks (B:)
(adj. for contactors)
Contactors (Cj )
(adj. for blocks)
Whole Plot Error
8 1.47 0.18 13.23*
2 0.60 0.30 21.76*
7 0.097 0.014
form reduction was different in at least two of
the three contactors (note column means in Table
8). Application of Tukey's HSD multiple com-
parison test to the data revealed that, just as
with the TCLR data, FCLR in the bubble dif-
fuser was significantly higher than in either the
packed column or the positive pressure injector.
Moreover, there was no difference in overall
FCLR between the latter two contactors.
In the split plot analysis, the ANOVA revealed
that both the Dk and the CjDk interaction were
highly significant (p < 0.01). Again the applied
ozone dose significantly affected the magnitude of
the fecal coliform reduction in all contactors, and
the differences in FCLR between contactors were
not the same at each dose level. These effects are
graphically depicted in Figure 9. The slopes of
O BUBBLE DIFFUSER
A PACKED COLUMN
Q POSITIVE PRESSURE INJECTOR
Split Plot Analysis
Doses (Dk)
Cj°k
B:D,
TOTAL
2
4
16
14
53
2.56
1.29
0.85
0.77
1.28 24.04*
0.32 5.85*
0.053 0.97
0.055
'significant at p<0.01 level.
(p < 0.01). This -is not consistent with the
ANOVA's for the other three dependent variables
(Tables 3, 5, and 7). On the first experiment day,
when the packed column and positive pressure
injector were run in parallel, the FCLR's in both
contactors were much lower than on any other
day. This observation was apparently of sufficient
magnitude to cause the significant B; effect.
.Noteworthy among the effluent quality charac-
teristics on this particular day is the fact that the
fecal coliform density prior to ozonation was
approximately 10-fold lower than on any other
day.
The Cj effect was also highly significant
(p
-------
OZONE
this, we plotted the total and fecal coliform log
reduction data in all contactors against the ozone
utilization data in all contactors. Results are
shown in Figures 10 and 11. Data from the three
individual contactors are differentiated by the
symbols, as defined in the legends. It is clear that
our data are consistent with the conclusion
reached by Scaccia and Rosen, namely, that
equivalent total or fecal coliform reductions are
achieveable in a given wastewater independent of.
contactor type as long as the absorbed dose is
O BUBBLE D1FFUSER
A PACKED COLUMN
O POSITIVE PRESSURE INJECTOR
Q./Q, IV, -Y,
FIGURE 10. MEAN LOG TOTAL COLIFORM REDUCTION
AS A FUNCTION OF OZONE UTILIZATION
O BUBBLE DIFFUSER
A PACKED COLUMN
Qg/Q, [Y, -Y,|
FIGURE 11. MEAN LOG FECAL COLIFORM REDUCTION
AS A FUNCTION OF OZONE UTILIZATION
the same. However, we have shown in the
foregoing data analyses and discussion that
equivalent absorbed doses (i.e., utilization)
necessary to achieve a given bacteriological stan-
dard may not be possible in some contactors.
The dashed horizontal lines depicted on Figures
10 and 11 represent calculated total and fecal
coliform log reductions, respectively, necessary to
achieve 1000 total coliforms/100 ml and 200 fecal
coliforms/100 ml in our filtered pilot plant ef-
fluent. The corresponding ozone utilization
needed to achieve those bacteriological levels is
approximately 4.8 mg/1. Referring back to Figure
6 wherein applied ozone dose is plotted against
absorbed ozone dose (utilization) in each contac-
tor, it is clear that only the bubble diffuser con-
tactor was able to achieve utilization of 4.8 mg/1.
In order for the packed column and positive
pressure injector to achieve comparable levels of
ozone utilization, the concentration of ozone in
the inlet gas (Yj) must be raised.
Effect of Increasing Y, at a Constant QG/QL
Ratio
The objective of this experiment was to deter-
mine the change in percent ozone utilization, ac-
tual ozone utilization, and microorganism reduc-
tion in the three contactors when the applied
dose was varied by changing the ozone concen-
tration (Y]) and maintaining a constant Qo/QL
ratio. The Qo/Qt ratio chosen for this ex-
periment was 0.5 (37.5 1/min gas flow rate and
75 1/min liquid flow rate per contactor). The
maximum Yj attainable with our ozone generator
at this gas flow rate was 15 mg/1 (ap-
proximately 1.25 weight percent). Thus, the three
concentration levels selected for this portion of
the study were 3.0, 9.0, and 15.0 mg/lgas,
corresponding to applied dose levels of 1.5, 4.5
and 7.5 mg/1, respectively. The period during
which this experiment was conducted was July 25
to August 16, 1978.
Quality of Wastewater
In Table 10, the minimum, mean, and maxi-
mum values of the filtered pilot plant effluent
characteristics prior to ozonation are summarized
for the second study period. There were several
rainy days during this time period, which
probably accounts for the lower readings relative
to those of the first study period. Again, the
filtered secondary effluent was of excellent quality.
Percent Ozone Utilization
In Table 11, the mean percent ozone utilization
data per contactor at each dose level are sum-
marized. The ANOVA for those data are presen-
ted in Table 12. In the whole plot analysis, the
Block effect (B;) was not significant, indicating
very little day to day variation in percent
utilization occurred in the contactors. However,
Cj was highly significant (p< 0.01), indicating
that overall percent ozone utilization was signifi-
cantly different in at least two of the contactors
(note column means in Table 11). To determine
where the significant differences occurred,
153
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 10. SUMMARY OF FILTERED PILOT PLANT
EFFLUENT CHARACTERISTICS PRIOR TO OZONATION
(JULY 25 to AUGUST 16, 1978)
Parameter
TCOD, mg/l
SCOD, mg/l
TOC, mg/l
TSS, mg/l
NHj-N, mg/l
TKN, mg/l
Turbidity, mg/l
PH
Temperature, °C
Total Coliforms
per 100 ml*
Fecal Coliforms
per 100 ml"
Mean
26
25
8.1
2.9
11.4
12.5
2.0
—
23
9.1 x 105
(geometric
mean)
2.7 x 105
(geometric
mean)
Minimum
18
17
4.4
1.2
5.8
6.2
0.8
7.4
22
8.5 x 10"
2.0 x 10"
Maximum
44
42
13.0
5.4
17.0
18.9
4.3
8.1
23
4.8 x 106
1.7 x 106
' Adjusted for 87% colony verification.
"Adjusted for 95% colony verification-
TABLE 11. MEAN PERCENT OZONE UTILIZATION AS A
FUNCTION OF APPLIED OZONE DOSE
(INCREASING Yv CONSTANT QG/QL)
Mean Percent Ozone Utilization
(Standard Deviation)
Dose mg/l
1.5
4.5
7.5
Mean % U,
All Doses
Positive
Pressure
Injector
84
(4)
68
(5)
62
(4)
71
(10)
Packed
Column
79
(3)
68
(4)
62
(3)
69
(8)
Bubble
Ditfuser
89
(2)
80
(4)
77
(3)
82
(6)
Mean % U,
All
Contactors
84
(5)
72
(7)
67
(8)
not the same at each dose level. The data from
Table 11 are depicted graphically in Figure 12.
TABLE 12. ANOVA FOR PERCENT OZONE UTILIZATION
DATA (INCREASING Y,, CONSTANT QG/QL)
Tukey's HSD multiple comparison test was used.
i Results indicated that overall percent ozone
utilization was higher in the bubble diffuser than
in either the packed column or the positive
pressure injector, while percent utilization in the
latter two contactors was similar.
In the split plot analysis the Dose (Dj) effect
was highly significant (p< 0.01) indicating that
percent ozone utilization in all contactors varied
with applied dose (note row means in Table 11).
The CjDk interaction was also highly significant
[(p<0.01), Table 12]. Thus, differences in per-
cent ozone utilization between contactors were
Source of Variation
Whole Plot Analysis
Blocks (Bj)
(adj. for contactors)
Contactors (Cj )
(adj. for blocks)
Whole Plot Error
Split Plot Analysis
Doses (Dt)
Cj Dk
B;Dk
BiCjD,
TOTAL
•significant at the p<0.01
Degrees
of
Freedom
8
2
7
2
4
16
14
53
level.
Sum
of
Squares
101.67
532.00
28.00
2786.78
173.78
87.67
81.78
Mean
Square F
12.71 3.18
266.00 66.50*
4.00
1393.39 254.31*
43.44 7.44*
5.48 0.94
5.84
Note that the trend is similar to that which was
observed when dose was increased by changing
the QG/QL at a constant Yj (Figure 4). The rate
of decrease in percent ozone utilization at higher
doses was substantially less in the bubble diffuser
than in the other two contactors. When Cj was
re-evaluated statistically at each individual dose
level, it was found that percent utilization in the
bubble diffuser was significantly higher than in
the other two contactors at each dose level
studied. Percent utilization in the positive
pressure injector was higher than in the packed
• BUBBLE DIFFUSER
A PACKED COLUMN
• POSITIVE PRESSURE INJECTOR
I—I-
FIGURE 12. MEAN PERCENT OZONE UTILIZATION AS
A FUNCTION OF APPLIED OZONE DOSE (INCREASING
Yv CONSTANT QG/QL)
154
-------
OZONE
column at the lowest dose level, but the two con-
tactors were equivalent at the higher dose levels.
The BjDk interaction (Table 12) was not
significant, indicating that the effect of dose
levels on percent ozone utilization in all contac-
tors was consistent from day to day. In Figure
13, the daily paired overall percent ozone
utilization data are shown to illustrate the per-
formance differences between contactors during
the study period.
SO
O 60
i
H
20
1
I
~
1
1
I
I
|
! I
. I
i i
i i i
I : -
I ! I
I I
I I -
-
I I
EXPERIMENT DAYS
FIGURE 13. OVERALL MEAN PERCENT OZONE
UTILIZATION PER REPLICATE RUN (EXPERIMENT 2)
Actual Ozone Utilization
In Table 13 the actual ozone utilization data
(i.e., the amount of ozone transferred to the
secondary effluent) are summarized. The ANOVA
for those data is presented in Table 14. Results
were similar to the percent ozone utilization data.
In the whole plot analysis the Bj effect was not
significant and the Cj effect was highly significant
(p< 0.01). Tukey's HSD test indicated that
overall ozone utilization was higher in the bubble
diffuser than in the other two contactors. Overall
ozone utilization in the packed column and
positive pressure injector was similar.
In the split plot analysis, both Dk and CjDk
were highly significant (p
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
analysis, neither the B; effect nor the Cj' effect
was significant. Thus, overall TCLR in all con-
tactors was similar (note column means in Table
15). This effect, which at first may appear
somewhat surprising, is not unexpected when one
considers the magnitude of the overall differences
in ozone utilization between contactors (column
means, Table 13). Only 0.5 mg/1 more ozone was
absorbed in the bubble diffuser than in the other
two contactors, and, although this was
statistically significant, the increase in absorbed
ozone was not sufficient to cause a marked in-
crease in TCLR.
_ 4
> 3
8"
• BUBBLE DIFFUSER
A PACKED COLUMN
• POSITIVE PRESSURE INJECTOR
I
FIGURE 14. MEAN OZONE UTILIZATION AS A
FUNCTION OF APPLIED OZONE DOSE (INCREASING
Y1? CONSTANT Qg/Q,)
In the split plot analysis (Table 16), the dose
main effect (Dk), the contactor-dose interaction
(CjDk), and the Block-Dose interaction (B;Dk)'
were highly significant (p<0.01). When the
results were re-analyzed statistically at each dose
level, it was found that all contactors elicited
similar TCLR responses at the 1.5 and 4.5 mg/1
applied dose levels. However, at the 7.5 mg/1
level, TCLR in the bubble diffuser was significan-
tly higher than in the packed column, while there
was no difference between the packed column
and the positive pressure injector. These effects
are displayed graphically in Figure 15.
In this experiment, more day-to-day variation
in the coliform reduction data occurred because
of the variable effluent quality. This would most
likely explain the significant BjDk interaction.
Fecal Coliform Reduction
Results from this experiment were almost exactly
equivalent to the TCLR data. They are plotted in
Figure 16. The fecal coliform log reduction data,
from this experiment are summarized in Table 17
and the ANOVA in Table 18.
TABLE 15. MEAN TOTAL COLIFORM LOG REDUCTION
(TCLR) AS A FUNCTION OF APPLIED OZONE DOSE
(INCREASING Yv CONSTANT QQ/QL)
' Mean TCLR
(Standard Deviation)
Dose mg/l
1.5
4.5
7.5
Mean TCLR,
All Doses
Positive
Pressure
Injector
0.81
(0.68)
3.07
(0.43)
3.28
(0.40)
2.39
(1.25)
Packed
Column
0.73
(0.59)
2.90
(0.39)
3.37
(0.30)
2.34
(1.25)
Bubble
Diffuser
0.56
(0.40)
3.24
(0.49)
4.00
(0.49)
2.60
(1.58)
Mean TCLR,
All
Contactors
0.70
(0.54)
3.07
(0.44)
3.55
(0.50)
TABLE 16. ANOVA FOR TOTAL COLIFORM LOG
REDUCTION DATA (INCREASING Yv CONSTANT QG/QL))
Source of Variation
Degrees Sum
of of
Freedom Squares
Mean
Square
Whole Plot analysis
Blocks (Bj) 8
(adj. for contactors)
Contactors (Cj) 2
(adj. for blocks)
Whole Plot Error 7
Split Plot Analysis
Doses (Dk) 2
CjDk 4
B]Dk 16
TOTAL 53
0.98 0.12 2.06
0.062 0.031 0.52
0.42 0.059
83.94 41.97 132.82*
1.69 0.42 6.14*
5.06 0.32 4.59*
0.96 0.069
"significant at the p<0.01 level.
Coliform Reduction vs. Ozone Utilization
Previously, we showed that, by plotting TCLR
and FCLR vs. ozone utilization (Figures 10 and
11), equivalent coliform reductions were
achieveable in a given wastewater, independent of
contactor type and contact time. The increase in
log reduction in Figures 10 and 11 was relatively
flat, following the first 2 to 3 mg/1 ozone
utilized. This apparent "all-or-none" phenomenon
suggested that perhaps log coliform reduction was
a function of log ozone utilization. When all the
156
-------
OZONE
_, 3 LOG, (No/N) (FECAL COLIFORMS) "" LOG. (No/N| (TOTAL COLIFORMS)
•=> 3>£ 0
,5 rjj m C -.roufeoi ^ ^ -• N w * o
I III
/ 0 BUBBLE DIFFUSER
/ A PACKED COLUMN
_ / m POSITIVE PRESSURE INJECTOR_
II I I I I
1 2345678
Y. (Qg/Q,)
RE 15. MEAN LOG TOTAL COLIFORM REDUCTION
AS A FUNCTION OF APPLIED OZONE DOSE
(INCREASING Y.,, CONSTANT Q /Q,)
I I I I I I I
~~ / • BUBBLE DIFFUSER ~~
/ A PACKED COLUMN
/ • POSITIVE PRESSURE INJECTOR
I I I I I I
1 2345671
Y, (Og/0.,1
RE 16. MEAN LOG FECAL COLIFORM REDUCTION
DUCTION AS A FUNCTION OF APPLIED OZONE
DOSE (INCREASING YV CONSTANT QO/Q,)
LE 17. MEAN FECAL COLIFORM LOG REDUCTION
:LR) AS A FUNCTION OF APPLIED OZONE DOSE
(INCREASING Yv CONSTANT QG/QL)
Mean FCLR
(Standard Deviation)
Positive Mean FCLR,
Pressure Packed Bubble All
Dose mg/l Injector Column Diffuser Contactors
1.5 0.58 0.58 0.54 0.57
(0.57) (0.60) (0.50) (0.52)
4.5 3.17 2.94 - 3.27 3.13
(0.45) (0.40) (0.37) (0.41)
7.5 3.37 3.48 4.00 3.62
(0.38) (0.32) (0.43) (0.45)
Mean FCLR, 2.37 2.34 2.60
All Doses (1.38) (1.37) (1.59)
TABLE 18. ANOVA FOR FECAL COLIFORM LOG
REDUCTION DATA (INCREASING Yv CONSTANT QG/QL)
Degrees Sum
of of Mean
Source of Variation Freedom Squares Square F
Whole Plot Analysis
Blocks (B; ) 8 0.60 0.075 1.68
(adj. for contactors)
Contactors (Cj ) 2 0.058 0.029 0.65
(adj. for blocks)
Whole Plot Error 7 0.31 0.045
: (B|Cj)
Split Plot Analysis
< Doses (Dk) 2 96.75 48.37 130.83*
CjDk 4 0.95 0.24 4.73*
B;Dk 16 5.92 0.37 7.40*
B|CjDk 14 0.70 0.050
TOTAL 53
"Significant at the p<0.01 level.
i
j
ozone utilization data from both experiments
were transformed to Iog10 values and plotted on
the abscissa against the corresponding log
coliform reduction data, the resulting graphs con-
firmed this hypothesis. The total coliform reduc-
tion data are depicted in Figure 17 and the fecal
coliform data in Figure 18. A highly significant
log-log correlation between coliform reduction
and ozone utilization was found. The correlation
coefficients for the total and fecal coliform data
were 0.985 and 0.987, respectively (the intercepts
were forced through the origin). Each figure in-
cludes 108 data points, gathered from both ex-
periments, plus an additional 12 data points (in-
dicated by triangular symbols) collected later for
verification of the regression equations. Although
it is possible that higher order relationships
(cubic, quadratic, etc.) may exist in the data, we
concluded that for all practical purposes the
linear log-log relationship is sufficient as a predic-
tive tool. For example, to achieve a final effluent
total or fecal coliform density of 1000/100 ml or
200/100 ml, respectively, starting with an initial
number of 1.55 X 106 total coliforms/100 ml or
4.17 x 105 fecal coliforms/100 ml (the respec-
tive geometric means of our data), 4.0 mg/l
ozone must be absorbed by the effluent (dotted
line labeled "a" in Figures 17 and 18). To
achieve 70 total coliforms/100 ml or 14 fecal
157
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
5
"5
o
o>
o
o
u
s
£
6.0
5.0
4.0
3.0
2.0
1.0
LOG1Q No/N = 5.4 LOG10 QG/QL
r=0 985
j h
O 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
FIGURE 17. Mean Log Total Coliform Reduction as a function of Log Ozone Utilization
6.0
5.0
4.0
3.0
2.0
1.0
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
FIGURE 18. Mean Log Fecal Coliform Reduction as a function of Log Ozone Utilization
ozone must be absorbed (dotted line labeled "b" is clear from the foregoing data presentation that
in Figures 17 and 18). Finally, to achieve the only one of the three contactors studied would be
strict State of California standard of 2.2 total capable of meeting all the above effluent
coliforms/100 ml, 12.1 mg/1 ozone must be ab- limitations reliably, with ozone generated from
sorbed (dotted line labeled "c" in Figure 17). It air. That contactor is the bubble diffuser.
158
-------
OZONE
ACKNOWLEDGMENTS
The able efforts of Messrs. Glenn Gruber, Leo
Fichter, and Burney Jackson in constructing the ozone
disinfection pilot system are gratefully acknowledged.
Ms. Catherine Yeats and Messrs. Harold L. Sparks
and Charles Palopoli assisted in the sampling,
performance of ozone mass balances, and membrane
filtration assays. Mr. Harold P. Clark identified,
enumerated, and confirmed all coliform colonies.
Chemical analyses were conducted by the Waste
Identification and Analysis Section, Wastewater
Research Division, Municipal Environmental
Research Laboratory, U.S.E.P.A., Cincinnati, Ohio.
REFERENCES
1. American Public Health Association. 1975. Standard Methods
for the Examination of Water and Wastewater, 14th ed.
American Public Health Association, Inc. New York.
2. Barr. A.J.. J.H. Goodnight, J.P. Sail, and J.T. Helwig. 1976. A
User's Guide to SAS 76. SAS Institute, Raleigh. North
Carolina, 329 pp.
3. Birdsall, CM, A.C. Jenkins, and E. Spadinger. 1952. "lodo-
metric Determination of Ozone." Anal. Chem., 24: 662.
4. Federer, W.T. 1975. "The Misunderstood Split Plot."In: Applied
Statistics, R.P. Gupta, ed. North Holland Publishing
Company, Amsterdam.
5. Kirk, R.E. 1968. Experimental Design: Procedures for the
Behavioral Sciences. Brooks-Cole Publishing Company,
Monterey, California.
6. Methods Development and Quality Assurance Research Labor-
atory. 1975. Methods for Chemical Analysis of Wastes.
EPA-625/6-7-003, U.S. Environmental Protection Agency,
Cincinnati, Ohio.
7. Scaccia, C. and H.M. Rosen. "O/one Contacting: What is the
Answer?" Presented at the International Ozone Institute's
Symposium on Advanced Ozone Technology, Toronto,
Ontario, Canada, November 16-18, 1977.
8. Venosa, A.D., M.C. Me.ckes, and E.J. Opatken. "Comparative
Disinfection Efficiency in Two Ozone Contactors: Packed
Column and Pressure Injector." Presented before the Div-
ision of Environmental Chemistry, American Chemical
Society's Annual Meeting, Anaheim, California, March
12-17, 1978.
DISCUSSION
DR. LONGLEY: I assume we can draw one
more conclusion from what you said, talking
about the bubble diffuser, and that is if you do
not kill them right away you do not kill them at
all. In other words, within thirty-five seconds.
MR. VENOSA: No. In the bubble diffuser, the
residence time was 9.4 minutes.
DR. LONGLEY: I got it turned around. That
was the positive pressure injector.
MR. VENOSA: Thirty-five seconds was the
residence time in the packed column.
DR. LONGLEY: Okay, well then let's take the
packed column at thirty-five seconds. All the kill
then occurs in thirty-five seconds.
MR. VENOSA: Correct, but the kill was inde-
pendent of detention time.
DR. LONGLEY: I would like to ask you
another question and you may have covered this
and I may have missed it. How did you do your
ozone residual analysis?
MR. VENOSA: I did not mention anything
about the residual, but we measured the residuals
by the backward amperometric titration procedure.
DR. LONGLEY: Did you scrub the gas out
before . . .
MR. VENOSA: We have done that. We have
looked at the forward titration, the backward
titration, and the ozone scrubbing method, and
each one gives a different result. That is why I
did not mention it.
DR. LONGLEY: If you are measuring directly
in the liquid rather than scrubbing the gas, are
you not really measuring total oxidants, at least
the materials that oxidize the iodide? You are not
really measuring ozone, are you?
MR. VENOSA: Presumably so, yes.
DR. ROSEN: You did say that for this parti-
cular work that these numbers you were com-
paring were at the end of the contact time, at
which time you added thiosulphate or something
to kill the reaction. You do not know that the
residuals would not show a continued kill with
time.
MR. VENOSA: We found that with the packed
column we were capable of evaluating that effect.
Remember those two 200 liter hold tanks that we
had? When we enumerated coliforms at the end
of those hold tanks versus the beginning of the
hold tanks, there wasn't really much difference.
It is when the water is actually in contact with
the ozone in the contactors when you get your
kill.
DR. ROSEN: Bollyky in the New York studies
found that there was some holding time effect.
Part of what .1 am trying to get at is to see if
this regrowth or reactivation occurs with ozone as
opposed to some of the other disinfectants.
MR. VENOSA: We have not looked at
159
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
regrowth. We have not looked at that with
ozone.
QUESTION: I have a question on the residual.
If you measured the residual, why didn't you in-
clude it in your mass balance when you were
figuring ozone utilization in the columns?
MR. VENOSA: Residual is included in that.
When you measure the off-gas concentration
relative to the in-gas concentration, the difference
is that which has been added to the wastewater,
not necessarily consumed by the wastewater. Now
if you look at the consumption of the ozone by
the wastewater whatever is left is the residual.
QUESTION: Right, but then you are not get-
ting a true kill per mg of ozone, if you do not
subtract out the residual. You are saying that all
the ozone is being utilized if you are only sub-
tracting the off-gases.
MR. OPATKEN: That is the reason that the
two hold tanks were placed after the packed
column, which provides a true countercurrent
operation. Normally if you are going to have a
substantial residual, you would get it in a packed
column. So, we put the two hold tanks in to
provide additional residence time to consume
the ozone and it did. So, there was six minutes
of residence time there which consumed the
ozone, and that also was then utilized. For the
other contactors, the PPI operated cocurrently.
You do not have the residual that you do in the
countercurrent operation. In the diffuser the
ozone is consumed as it is sitting in the tank it-
self for three minutes in each column. So you do
not have a high residual coming off the bubble
diffuser.
QUESTION: But even if you have 0.3 ppm at
a low dosage of 3 ppm, that a 10% effect.
MR. OPATKEN: We did measure the residuals
coming off the columns, and they are lower than
that.
QUESTION: But if you used standard
methods, you said you used the standard back
titration method, and they say right in the stan-
dard method that they are not accurate down
past ranges in that area.
MR. OPATKEN: True.
MR. VENOSA: What area?
QUESTION: Down past three parts per
million, but if you use Schechter's spec-
trophotometric method that was published in
1973 it goes down past those ranges.
DR. HILL: I have a question about the method
you use to measure the ozone concentration in
your air stream. I think you mentioned that you
used BirdsalPs method. Is this one based on buf-
fered solution of the iodide, or what method is
this?
MR. VENOSA: The pH of the titration is per-
formed at 2.
DR. -HILL: When you are bubbling your ozone
through the iodide solution, is the iodide solution
neutral?
MR. VENOSA: Unbuffered.
MR. OPATKEN: There was a Dasibi monitor
that corresponded exceedingly well with the wet
test method. It was within plus or minus 2%.
DR. ROSEN: With respect to the Schechter
method, that was done on very clean water as
opposed to these kinds of waters. Then you
would expect certainly the residuals to hang
around long enough to get a reading in the spec-
trophotometer cell. Here we are talking about
wastewater, and it does have some interferences
and other things. So I am not sure how valuable,
especially when you consider the low concen-
trations you are talking about, that Schechter's
method is in wastewater.
DR. JOHNSON: I find it disturbing when we
talk about ozone concentration in wastewater. As
Longley pointed out at the beginning of the
discussion, there is not any ozone in that
wastewater. It is oxidant. One of the reasons you
do not see the classical kind of time dependence
of disinfection is that there is ozone there but
just for a very short period of time. When you
measure with the KI right in the sewage you are
not measuring ozone even though Schechter says
it is ozone.
MR. OPATKEN: We have stripped it out as it
was coming into our sample, using a glass
distillation flask. Now we cannot do it
stoichiometrically, but we do know that we have
ozone residual, because we strip it out and then
we bubble it into potassium iodide. So we do
have an ozone residual.
DR. JOHNSON: What are your comparisons?
You are talking about some new data we have
not seen. What is the comparison between the
ozone that you measure in the solution and the
ozone that you measure in the stripped out
ozone?
MR. OPATKEN: The ozone stripped out may
160
-------
OZONE
be 50 to 60% of that which we meausre by the
back titration method. That 50 to-60% is not
sacred either, but that is the reason we just have
not been able to define ozone residuals as such,
but we do have ozone residuals by the stripped
method with wastewater.
DR. JOHNSON: After how much time?
MR. OPATKEN: No time. As we take our
liquid sample we are stripping. In other words,
we have gas flowing and stripping out the ozone
when we start our liquid sampling.
DR. JOHNSON: You are pumping in ozone
gas.
MR. OPATKEN: No. We are putting liquid in
and pumping in dry air, and stripping that ozone
that is in the wastewater from the wastewater.
161
-------
19.
ECONOMIC EVALUATION OF OZONE CONTACTORS
Edward J. Opatken
Municipal Environmental Research Laboratory, U.S. EPA, Cincinnati, Ohio
INTRODUCTION
This paper discusses the economics of disinfec-
tion in three different contactors. The evaluation
is based heavily on pilot plant results (1). De-
tailed cost sheets and summary cost data have
been developed on the three contactors. The cost
data are summarized in the accompanying
18 tables.
Conditions and Assumptions
The ground rules that were employed during
this evaluation are:
1. This paper compares the economics of disin-
fecting wastewater with ozone using the
following contactors:
a. Packed Column (PC)
b. Positive Pressure Injector (PPI)
c. Bubble Diffuser operating at:
(1) 5 mg/1 O3 concentration.
.(2) 10 mg/1 O3 concentration.
(3) 15 mg/1 O3 concentration.
2. The evaluation is based on pilot plant data
and extrapolated to one treatment plant size.
That size is 5000 mVd.
3. The data are in metric units. English units
are displayed in the paper within paren-
theses. In the tabular material the roman
numeral "M" refers to 1000 throughout
the paper.
4. The process selected for disinfection may
not be the most economical, but the costs
are used to compare contactors, not opti-
mize ozonation. An illustration might be
helpful. To produce ozone from air, the air
should be dried to below -40°C dew point.
The process employed at the pilot plant con-
sisted of compressing ambient air to 700
kPa and cooling the compressed air to am-
bient temperature. This pressurization and
cooling lowered the dew point to -7°C
.which removed 87% of the moisture. The
compressed air was then dried to below
-40°C by adsorbing water vapor on an
alumina bed. Approximately 25% of the dry
air stream was passed through a second bed
of alumina to regenerate depleted alumina
beads and thus maintain a continuous flow
system. This probably was not the most
efficient method for drying air. However,
since it was the method used at the pilot
plant, it will serve as the basis for the en-
suing cost analysis. The individual power
cost of compressing and regenerating the air
can be compared with other methods of
drying, but the prime purpose of this paper
is to compare costs associated with contact-
ing, not drying. Since the air was dried by
the same technique for all contactors, the
comparative evaluation is valid and differen-
tial costs would only be slightly affected.
The same is true for the ozonator. Ozone
generation power costs were based on the
pilot plant ozonator. If there are ozonators
that require less power at these concentra-
tions, and capital costs are about the same
as the pilot plant unit, then there could be
a reduction in the cost of disinfection, but
it probably will not significantly affect the
differential costs between the contactors.
5. In reporting operating and overhead costs,
the assumptions that were used in calculat-
ing these unit costs are given. These values
are estimates and as such they are not
sacred. However, the. values on power costs
or material costs, such as antifoam, are
based on actual costs, and for this particu-
lar wastewater they are firm.'
6. The unit cost for producing ozone was
162
-------
OZONE
TABLE 1. SUMMARY OF CONTACTOR EVALUATION
' Concentration (mg/l) O3
Capital Investment ($)
Disinfection Cost (S/Mm3)
Disinfection Cost («/Mgal)
DIFFUSER
5
406,000
39.4
15
10
266,000
29.2
11
15
'336,000
34.4
13
PACKED
COLUMN
15
445,000
47.3
18
PRESSURE
INJECTOR
21
665,000
60.00
23
developed and then carried over to the dis-
infection cost.
Results and Discussion
Table 1 is a summary of the contactor evalua-
tion. The diffusers were studied at three different
ozone gas concentrations: 5, 10, and 15 mg/l.
The packed column and pressure injector were
evaluated in parallel with the diffuser.
The diffuser at 10 mg/l had the lowest capital
investment, i.e., $266,000. Capital costs range
from this minimum to $665,000. The diffuser at
10 mg/l also had the lowest disinfection cost,
which was $29.20/Mm3 or HC/Mgal. These
values include operating and overhead costs.
Since this evaluation is scaled to a treatment plant
size of 5000m3/d, it can be expected that operating
labor and amortization costs contribute a dispro-
portionate share of the total disinfection cost.
OZONE COST SHEETS
Listed on Table 2 are the bases for determining
the quantity and concentration of ozone. These
data were developed at the pilot plant.
With the information from Table 2 the indivi-
dual cost sheets can be developed. The ozone
costs for the packed column are given on Table 3.
The direct out-of-pocket cost for ozone is 99
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 3. OZONE COST SHEET - PACKED COLUMN
O3 Rate = 42 kg/d @ 15 mg/l
FCI = $370,000
I. Operating Cost
A. Utility Cost
1. Electric Power
Ratio
kWh/kg
5
1
26
Mgal/kg
.3
a. Compression
b. Regeneration
c. Generation
2. Water
a. Cooling
Total Utility Cost (c/lb)
B. Operating Labor (1/2 MY)
C. Repair Labor (1% FCI)
D. Supervision [15% (O.L. + R.L)]
E. Repair Materials (1% FCI)
F. Supplies [10% (O.L. + R.L. + Supv)]
Total Operating Cost
Overhead Cost
A. Ins. (1% FCI)
B. Amortization (7% FCI)
Total Overhead Cost
TOTAL OZONE COST (C/lb)
Price
c/kWh
3
3
3
c/Mgal
10
Cost
e/kg
15
3
78
41
24
10
24
8
24
170
99 (45)
206
194
400 (182)
attributed to the high fixed capital investment. In
the case of the PPI, 83 ozone units are required
to obtain a production of 50 kg/d @ 21 mg/l,
whereas 39 units were required for the PC, since
only 42 kg/d @ 15 mg/l O3 concentration were
required.
Tables 5, 6, and 7 cover the ozone costs for
the bubble diffuser, at the three ozone concentra-
tions. The quantity of ozone required for the
bubble diffuser is considerably less than it is for
the PC or PPI. This is the result of a signifi-
cantly higher utilization efficiency. The utilization
is 80, 85 and 90°/o for the diffuser at ozone con-
centrations of 5, 10, and 15 mg/l, respectively,
and only 60% for the PC and 50% for the PPI.
The three different bubble diffuser cases show
the effect of concentration on ozone generation
power: as the concentration increases so does the
power to generate ozone. For 5 mg/l the genera-
tion power is 15 kwh/kg, at 10 mg/l the genera-
tion power increases to 21 kwh/kg, and at 15
mg/l the power is 26 kwh/kg. However, as the
ozone concentration decreases, the volume of air
required for compression and drying increases to
obtain the necessary dosage. The advantage of
lower power generation is lost in air preparation,
in that 35 kwh/kg is the total power required at
5 mg/l and 32 kwh/kg is the total power at
15 mg/l. The major differences in cost for the
three operating conditions on the diffuser are
related to the overhead differences caused by the
FCI for the three cases. The numbers of ozone
units required to achieve the production level of
31 kg/d are 37 for 5 mg/l; 18 to produce 29
kg/d at 10 mg/l; and 26 units to produce 28
kg/d at 15 mg/l.
G = GAS FLOW (I/mm)
250 500
750 1000 1250 1500 1750 2000
POWER (WATTSI
Figure 1. Ozone Concentration vs. Power at
Specified Gas Flow Rates
164
-------
OZONE
TABLE 4. OZONE COST SHEET - PPI
O3 Rate = 50 kg/d @ 21 mg/l
FCI = $630,000
I. Operating Cost
A. Utility Cost
1. Electric Power
a. Compression
b. Regeneration
c. Generation
2. Water
a. Cooling
Total Utility Cost (e/lb)
B. Operating Labor (1/2 MY)
C. Repair Labor (1% FCI)
D. Supervision [15% (O.L + R.
Repair Materials (1% FCI)
Ratio
kWh/kg
4
1
35
Mgal/kg
.2
L.)]
F. Supplies [10% (O.L. + R.L. + Supv)]
Total Operating Cost
II. Overhead Cost
A. Ins. (1% FCI)
B. Amortization (7% FCI)
Total Overhead Cost
TOTAL OZONE COST («/lb)
Price Cost
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 6. OZONE COST SHEET - DIFFUSER
O3 Rate = 29 kg/d @ 10 mg/l
FCI = $210,000
I. Operating Cost
A. Utility Cost
B.
C.
D.
E.
F.
1. Electric Power
a. Compression
b. Regeneration
c. Generation
2. Water
a. Cooling
Total Utility Cost (e/lb)
Operating Labor (1/2 MY)
Repair Labor (1% FCI)
Supervision [15% (O.L + R.
Repair Materials (1% FCI)
Supplies [10% (O.L + R.L.
Total Operating Cost
Ratio
kWh/kg
8
2
21
Mgal/kg
.4
.L)]
+ Supv)]
Price
c/kWh
3
3
3
C/Mgal
10
Cost
«/kg
24
6
63
4
97 (44)
59
18
12
18
9
213
II. Overhead Cost
A.
B.
Ins. (1% FCI)
Amortization (7% FCI)
Total Overhead Cost (c/lb)
TOTAL OZONE COST
18
130
148
361 (164)
TABLE 7. OZONE COST SHEET - DIFFUSER
O3 Rate = 28 kg/d @ 15 mg/l
FCI - $280,000
I. Operating Cost
A. Utility Cost
1. Electric Power Ratio
kWh/kg
a. Compression 5
b. Regeneration 1
c. Generation 26
2. Water Mgal/kg
a. Cooling .3
Total Utility Cost (c/lb)
B. Operating Labor (1/2 MY)
C. Repair Labor (1% FCI)
D. Supervision [15% (O.L. + R.L.)]
E. Repair Materials (1% FCI)
F. Supplies [10% (O.L. + R.L. + Supv)]
Total Operating Cost
II. Overhead Cost
A. Ins. (1% FCI)
B. Amortization (7% FCI)
Total Overhead Cost
TOTAL OZONE COST (
-------
OZONE
TABLE 8. SUMMARY OF OZONE COST
Contactor
Concentration (mg/l)
Fixed Capital Inv. ($)
Ozone Gen. Units
Ozone Production kg/d
O Utility Cost ($/kg)
($/lb)
O Total Cost ($/kg)
($/lb)
Gas Flow Rate (m3/d)
Diffuser
5
350,000
37
31
1.13
0.51
5.07
2.30
6,200
10
210,000
18
29
0.97
0.44
3.61
1.64
2,950
15
280,000
26
28
0.99
0.45
4.74
2.15
1,870
Packed
Column
15
370,000
39
42
0.99
0.45
4.00
1.82
2,780
Pressure
Injector
21
630,000
83
50
1.22
0.55
5.23
2.38
2,380
Table 8 summarizes some of the significant
differences:
1. The diffuser at an ozone concentration of
10 mg/l has a decided edge over the other
conditions and contactors evaluated. The
FCI and the ozone disinfection cost are at a
minimum for both conditions.
2. The second best contactor appears to be the
PC because of an advantage from an ozone
cost standpoint. However, the advantage
may be tied in with low unit operating costs
due to the higher production levels of
ozone. When disinfection comes into play,
the advantage for the PC will be offset by a
higher dose and therefore higher disinfection
costs.
3. The ozone cost between $1.64 and $2.38 per
pound is expensive because of the low pro-
duction rate. To keep this chemical, ozone,
in perspective, remember that 30,000 Ib/yr
is considerably different from chlorine plants
that have capacities in the tons/day cate-
gory. And this is really why the cost dif-
ference is significant.
DISINFECTION COST SHEETS
Packed Column
With the ozone costs established, these values
were then applied to the various contactors to
determine the cost of disinfection. Table 9 shows
the costs for disinfecting with the packed column.
The English equivalents in
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 9. DISINFECTION COST SHEET - PC AT 15 mg/l
(1) Treatment Plant Size = 5000 m3/d
(2) Dose at 8.4 mg/l
I. Materials Ratio Price Cost
kg/Mm3 $/kg $/Mm3
A. Ozone 8.4 4.00 33.6
B. Antifoam (c/Mgal) 2.0 2.05 4.10(1.6)
Total Materials cost 37.7
II. Operating Cost kW/Mm3
-------
OZONE
TABLE 11. DISINFECTION COST SHEET -DIFFUSER AT 5 mg/l
(1) Treatment Plant Size =
(2) Dose at 6.2 mg/l
I. Materials
A. Ozone
5000 m3/d
Ratio
kg/M3
6.2
Price
$/kg
5.07
Cost
$/Mm3
31.43
II. Operating Cost
A. Power kW/Mm3 C/kW
1. Electric .146 3 .44
B. Operating Labor (1/2 MY) 3.42
C. Maintenance (1% FCI) .31
D. Supv. [15% (O.L. + M.L)] .56
E. Maintenance Materials (1% FCI) .31
F. Supplies [10% (O.L. + M.L + Supv.)] .43
Total Operating Cost 5.47
III. Overhead Cost
A. Ins. & Taxes (1% FCI) .31
B. Amortization (7% FCI) 2.T5
Total Overhead Cost 2.46
TOTAL DISINFECTION COST (C/Mgal) 39.36 (15)
SUMMARY OF DISINFECTION COST
yielded reduced unit cost of ozone is significantly Tat,ie 14 shows the applied dose as the major
increased by the increased dose of 8.4 mg/l. The cause for the cost differences. The doses shown
material costs jumped because of the higher ratio were established at the pilot plant. There are dif-
of kg/Mm3 which negated the effect of a lower ferences in contactor FCI, but these are relatively
unit cost of ozone. The disinfection costs showed small, especially when compared with the FCI for
only a slight effect by the increased FCI of the ozonation. Finally, the diffuser at 10 mg/l is the
PC and no difference in the operating labor cost most economical process for achieving disinfec-
since the base unit is identical in terms of Mm3. tion with ozone.
TABLE 12. DISINFECTION COST SHEET - DIFFUSER AT 10 mg/l
(1) Treatment Plant Size = 5000 m3/d
(2) Dose at 5.9 mg/l
I. Materials Ratio Price Cost
kg/Mm3 $/kg $/Mm3
A. Ozone 5.9 3.61 21.3
II. Operating Cost
A. Power kW/Mm3 C/kW
1. Electric -146 3 .44
B. Operating Labor (1/2 MY) 3.42
C. Maintenance (1% FCI) .31
D. Supv. [15% (O.L. + M.L)] .56
E. Maintenance Materials (1% FCI) .31
F. Supplies [10% (O.L. + M.L. + Supv.)] .43
Total Operating Cost 5.47
III. Overhead Cost
A. Ins. & Taxes (1% FCI) .31
B. Amortization (7% FCI) 2.15
Total Overhead Cost 2.46
TOTAL DISINFECTION COST (C/Mgal) 29.23 (11)
169
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 13. DISINFECTION COST SHEET - DIFFUSER AT 15 mg/l
(1) Treatment Plant Size = 5000 m /d
(2) Dose at 5.6 mg/l
Materials
A. Ozone
Ratio
kg/Mm3
5.6
Price
$/kg
4.74
Operating Cost
A. Rower kW/Mm3 C/kW
1. Electric .146 3
B. Operating Labor (1/2 MY)
C. Maintenance (1% FCI)
D. Supv. [15% (O.L. + M.L)]
E. Maintenance Materials (1% FCI)
F. Supplies [10% (O.L. + M.L. + Supv.)]
Total Operating Cost
Overhead Cost
A. Ins. & Taxes (1% FCI)
B. Amortization (7% FCI)
Total Overhead Cost
TOTAL DISINFECTION COST (C/Mgal)
Cost
$/Mm3
26.5
.44
3.42
.31
.56
.31
.43
.31
2.15
5.47
2.46
34.43 (13)
TABLE 14. SUMMARY OF DISINFECTION COSTS
Concentration (mg/l)
Applied Dose (mg/l)
Utilized Dose (mg/l)
Capital Investment ($)
Treatment Plant Size (m3/d)
(mad)
Disinfection Cost ($/Mirn
(it/Mgal)
Disinfection Cost-Direct ($/Mm3)
(c/Mgal)
Dilfuser
5
6.2
5
56,000
5,000
1.3
39.4
15
7.45
2.8
10
5.9
5
56,000
5,000
1.3
29.2
11
6.16
2.3
15
5.6
5
56,00
5,000
1.3
34.4
13
5.98
2.3
Packed
Column
15
8.4
5
75,000
5,000
1.3
47.3
18
13.5
5.1
Pressure
Injector
21
10
5
35,000
5,000
1.3
60.00
23
13.7
5.2
The last line of the summary indicates the out-
of-the-pocket costs. These values range between
2.3 and 5.2C/Mgal, which is more typical of pub-
lished ozone costs. However, the problem arises
that with this type of evaluation we could select
the diffuser at 15 mg/l as an economical proces-
sing condition.
To place these costs in proper perspective a
yearly budget should be considered. Tables 15
and 16 show the proposed yearly budget for the
diffuser at 10 mg/l as an ozone budget and a
disinfection budget, respectively. These two are
combined .on Table 17 to show the overall bud-
geted items. Amortization accounts for 33% of
the budget and operating labor at 25%. The out-
of-the-pocket cost or in this case, electrical
power, constituted approximately 20% of the
total budget. It is important to single out these
individual cost items to determine where futu/e
research efforts should concentrate in order to
reduce disinfection costs. In this case a program
on reducing the FCI would prove more beneficial
than reducing the power cost associated with
generating ozone.
This type evaluation can be used to determine
the cost of chlorine disinfection. Such an evalua-
tion appears in Table 18. The cost is 5.1
-------
OZONE
TABLE 15. OZONE BUDGET FOR DIFFUSER @ 10 mg/l
O3 Required = 29 kg/d * 10,600 kg/yr
Operating Budget $/yr
A. Electric Power 9,800
B. Cooling Water 400
C. Operating Labor 6,200
D. Repair Labor 1,900
E. Supervision 1,200
F. Repair Materials 1,900
G. Supplies 900
Total Operating Budget 22,300
Overhead Budget
A. Insurance 1,900
B. Amortization 13,800
Total Overhead Budget 15,700
TOTAL OZONE BUDGET 38,000
TABLE 16. DISINFECTION BUDGET
FOR DIFFUSER @ 10 mg/l
WW Rate - 5 Mm3/d * 1,800 Mm3/yr
I. Material Budget
A. Ozone
$/yr
38,000
Operating Budget
A. Electrical Power
B. Operating Labor
C. Repair Labor
D. Supervision
E. Repair Materials
F. Supplies
Total Operating Budget
Overhead Budget
A. Insurance
B. Amortization
Total Overhead Budget
TOTAL DISINFECTION
800
6,200
600
1,000
600
800
10,000
600
3,900
4,500
BUDGET 52,500
TABLE 17. COMBINED OZONE AND DISINFECTION BUDGET FOR DIFFUSER @ 10 mg/l
WW = 1,800 Mm3/yr
O, = 10,600 kg/yr
o
Operating Budget $/yr
A. Electric Power 10,600
B. Cooling Water 400
C. Operating Labor 12,400
D. Repair Labor 2,500
E. Supervision 2,200
F. Repair Materials 2,500
G. Supplies 1,700
Total Operating Cost
Overhead Budget
A. Insurance 2,500
B. Amortization 17,700
Total Overhead Cost
TOTAL OZONE DISINFECTION COST
32,300
20,200
52,500
171
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 18. DISINFECTION COST SHEET - CHLORINE
(1) Treatment Plant Size = 5000 m3/d
(2) Dose at 10 mg/l
(3) FCI = $80,000
I. Materials
A. Chlorine (c/Mgal)
Ratio
kg/Mm3
10
Operating Cost
A. Power
B. Operating Labor (1/2 MY)
C. Repair Labor (1% FCI)
D. Repair Materials (1% FCI)
E. Supv. [15% (O.L + R.L)]
F. Supplies [10% (O.L. + R.L. + Supv.)]
Total Operating Costs
Overhead Cost
A. Ins. & Taxes (1% FCI)
B. Amortization (7% FCI)
Total Overhead Costs
TOTAL DISINFECTION COST (l/Mgal)
Price
$/kg
.41
.44
3.42
.44
.44
.58
.44
.44
3.07
Cost
$/Mm3
4.10
(1.6)
5.76
3.51
13.37 (5.1)
REFERENCE
1. Venosa, A.D., M.C. Meckes, E.J. Opatken, and J.W. Evans,
"Comparative Efficiencies of Ozone Utilization and
Microorganism Reduction in Different Ozone Contac-
tors." This Symposium.
DISCUSSION
MR. GIAIMO: Pure Water Systems: It was
pointed out this morning that we reached a cer-
tain disinfection level at a certain cost. What
disinfection level was achieved at these costs?
MR. OPATKEN: 200 fecal coliforms/100 ml.
We said 200 was achieved at 4 mg/l, and our
safety factor was applied on to that to make it
5 mg/l.
MR. GIAIMO: Did you have a corresponding
increase to achieve the 2.2 MPN level?
MR. OPATKEN: No.
MR. GIAIMO: In other words you are saying
that it cost the same to get 2.2 as it cost to
get ...
MR. OPATKEN: No. You want the cost for
2.2 MPN. I do not know. Al said he has one
point that he got for you this morning. He said
it took 15 mg/l, but I think he wants to check
that point a lot more thoroughly than that.
MR. GIAIMO: That would be around 3 times
the cost, then? In other words, if you have
5 mg/l to get 200 fecal coliforms, and you have
to go to 15 mg/l to get 2.2, you're about 3
times the cost . . .
MR. OPATKEN: No. I haven't calculated that.
MR. GIAIMO: You are saying really you did
not look at it.
MR. OPATKEN: At 2.2, no. I did not look at
2.2.
DR. RICE: Question and a comment. I did not
get the amortization time on the ozonation sys-
tem. It looks like it is about three years.
MR. OPATKEN: No, I checked with Bob
Smith, and we rounded it off at 7% for 20
years. I think it was 6% or 67/8.
DR. RICE: I did not know whether the 7%
meant interest on the money and I wanted to get
the years. The second thing is, the half a man
year. In the European drinking water plants they
are using ozone or C102. These plants effectively
run themselves. You do not have a half a man
operating those things.
MR. OPATKEN: I think Dr. Jain will actually
increase this estimate when he discusses start-up
172
-------
OZONE
at the Meander facility, and he may use up his
ten year contingent of manpower just for start-up.
DR. RICE: The third thing I would like to say
is you have based all of your ozone generation
production rate versus power on minus 40 °C air,
and mostly in European plants it is generally
around minus 60 °C. Minus 40 °C in those genera-
tors over there seems to be just skirting the edge
of getting too little ozone for the power.
MR. MECKES: We did dry our air below
minus 40 °C. However, it did vary. We did have
some days when we were below minus 60 °C as
far as drying.
DR. ROSEN: I am obviously not going to be
able to sit here and take all the numbers that
you took and calculate the same sorts of things
that you did, but I have done enough work in
this area in making cost estimates to know that
you are off by at least a factor of two. First of
all, the particular unit that you are using to
generate ozone is a one-of-a-kind non-production
unit.
MR. OPATKEN: And have you got a curve of
your unit to compare with that one?
DR. ROSEN: Yes. There are certain proprietary
reasons why those things are not published, not
because we do not know them. But you are
talking about a one of a kind unit. Your genera-
tor is very much limited on the cooling capacity
and some other things. Aside from that you re-
ferenced to the paper that Scaccia and I presented
on contactors, and you basically came up with
the same conclusion, that independent of the con-
.tactor you can get the same kind of disinfection
if you get the same kind of utilized dose.
MR. OPATKEN: I think that is where we do
agree.
DR. ROSEN: Now where we disagree, which is
very important, is that in that same paper we
showed that for a positive pressure injector op1
crating somewhere near 6 mg/1 there was some-
thing like 80% utilization, and for a diffuser sys-
tem like you were using something like 98%
utilization.
MR. OPATKEN: What was your gas-to-liquid
ratio?
DR. ROSEN: It was a range, and that was all
reported in that paper.
MR. OPATKEN: What was the concentration?
DR. ROSEN: It was around 1% by weight, 1
to 2% depending whether it was generated from
air or oxygen.
MR. OPATKEN: We received the positive
pressure injector from you.
DR. ROSEN:Yes, but things have changed.
What you are saying is that prototypes are what
ultimately become commercial products, and that
is wrong. You are talking about a field that is
changing very rapidly from day to day.
MR. OPATKEN: They are not changing that
fast, Harvey.
173
-------
20.
FIELD SCALE EVALUATION OF WASTEWATER DISINFECTION
BY OZONE GENERATED FROM AIR
Kerwin L. Rakness and Bob A. Hegg
M & I, Inc., Consulting Engineers
4710 South College Avenue
Fort Collins, Colorado 80525
ABSTRACT
The upper Thompson Sanitation district (UTSD) A WTfacility located
at Estes Park, Colorado, incorporated one of the first full-scale ozone
wastewater disinfection systems in the United States. Plant design flow is
5,680 cu mj day (1.5 mgd) and current operating flows have ranged from
1,140 cu ml day to 3,790 cu mj day (0.3 mgd to 1.0 mgd). The ozone system
has been operational for 2 years and good disinfection performance has
occurred, but several required modifications have resulted in intermittent
operation.
The ozone generation system is air-fed. The air pretreatment unit
performed well, but periodic malfunctions have damaged the ozone
generator. A high dew point alarm is being investigated to prevent this
problem from reoccurring. The dew point of air from the pretreatment unit
increased during an 8-hour drying to wer cycle by approximately 20° C, and
during normal operation all readings were better than the manufacturer's
rated value. The increase in dew point resulted in as much as a 30 percent
decrease in ozone production, which resulted in a decrease in the ozone
dosage to the wastewater. Air pretreatment system design should address
the fluctuation in air dew point.
Ozone power utilization (kwhjlb) was significantly greater at lower
ozone production levels due to the effect of the constant power
consumption of the air pretreatment system. The UTSD plant typically
operates at lower ozone production levels and thus at inefficient power
utilization values, because wastewater flows are lower than design and the
ozone dosage requirements are not as high as anticipated. Because of the
relatively limited information on ozone disinfection, most plants will
probably be designed conservatively and therefore will require less
ozone than for design conditions for most of the plant life. Con-
sideration should be given during plant design to provide a power
efficient unit over the entire system's operating range.
174
-------
OZONE
The UTSD ozone system has been intermittently shut down due to
excessively high ambient ozone concentrations in and around the plant
area. Several modifications were made to prevent ozone leakage including:
contact basin covering and sealing, revisions to basin exhaust fan,
modifications to basin baffles and scum skimmers, complete ozone
repiping from U.P. V.C. to stainless steel piping and installation of new-
ozone diffusers. An additional modification that has been designed
but not yet installed is an off-gas ozone destruct system.
Much conflicting, information still exists concerning ozone srstem
design and operation. Two major areas of disagreement concerning the
UTSD ozone system were ozone transfer efficiency and the need for an off-
gas ozone destruct system. Based on operation experiences at the UTSD
plant, the measured ozone transfer efficiency has conformed to
ozone/liquid gas transfer theory when good disinfection occurred. The
conclusion is that the contact basin is operating as expected, despite the
fact that transfer efficiencies are considerably less than values claimed for
ozone s vsterns. Ozone in the basin off-gas has existed and has caused ozone
exposure to plant operating personnel. To avoid this potentially
hazardous situation an off-gas ozone destruct unit was deemed necessarv
at the UTSD facility and should be strongly considered as part of an ozone
wastewater disinfection system.
INTRODUCTION
Background, Purpose and Scope
The Upper Thompson Sanitation District (UTSD)
was formed to provide sanitary services to the area
surrounding the town of Estes Park, Colorado. Estes
Park is located about 75 miles northwest of Denver,
Colorado, and is adjacent to Rocky Mountain
National Park. The elevation of the community is
2,286 m (7,500 ft), above sea level. The UTSD
wastewater treatment .plant discharges to the Big
Thompson River which is a cold water fishery and
highly visible with respect to the tourist industry. The
plant was placed in operation in April 1976, and has a
design flow rate of 5,680 cu m/day (1.5 mgd).
Significantly large seasonal variations in plant flow
occur due to the high influx of tourists in the summer
months. The plant flow rate has ranged from 1,140 to
3,790 cu m/day (0.3 to 1.0 mgd). Several unit processes
were incorporated into the plant's advanced
wastewater treatment (AWT) design to handle the
flow variations and discharge a high quality plant
effluent. Among these unit processes were flow
equalization, activated sludge, attached growth
nitrification, tri-media filtration and ozone
disinfection.
The incorporation of several unique unit processes
and the combination of these processes at one facility
aided the UTSD in obtaining funds from the U.S.
Environmental Protection Agency (EPA) for a two
and one-half year research project entitled, "An
Evaluation of Pollution Control Processes — Upper
Thompson Sanitation District". Data collected for the
research effort included all unit processes. The scope
of this paper includes the ozone disinfection process
only.
The UTSD ozone system was one of the first
full-scale ozone wastewater disinfection processes in
the United States. Several state-of-the-art problems
were encountered, and continuous ozonation has
not yet occurred. As such, limited cost and perfor-
mance information was developed. This paper dis-
cusses the problems encountered with the UTSD
ozone system and presents conclusions on various
aspects of operation, maintenance and design.
UTSD Plant Description
The UTSD advanced wastewater treatment facility
and 33 miles of interceptor and collection lines were
constructed concurrently. The majority of flow
received at the plant is from domestic and commercial
sources. Since the plant discharges to the Big
175
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
Thompson River, stringent effluent requirements were
established to protect the high quality of the river. (See
Table 1).
TABLE 1. PLANT EFFLUENT DESIGN CRITERIA
FOR THE UTSD
Parameter 30-Day Average
5 Day BOD (mg/l)
Total Suspended Solids (mg/l)
Fecal Coliform (#/100 ml)
Ammonia (as N - mg/l)
Chlorine Residual (mg/l)
Residual D.O. (mg/l)
5
5
200
2.0
0.1
2.0
7-Day Average
10
10
400
—
0.4
2.0
Figure I shows a flow schematic of the processes
that were selected to achieve the desired effluent
quality. The process design flow is 5,680 cu m/ day (1.5
mgd), with a hydraulic capacity of 13,250 cu m/day
(3.5 mgd). Parallel treatment systems are provided for
the activated sludge and nitrification, processes. At
flows of 2,840 cu m/day (0.75 mgd) or less, only one-
half of these processes are operated.
All flow entering the plant is pumped to an
equalization basin. A controlled rate of flow is
directed through an aerated grit chamber to the
activiated sludge process. Secondary clarifier effluent
is pumped to a fixed-film nitrification tower. Tower
effluent is directed through multimedia filters to the
plant disinfection process. All sludge generated within
the plant was wasted to aerobic digesters,
concentrated on a pressure roller filter and hauled to a
sanitary landfill.
Stringent disinfection and stringent chlorine
residual effluent requirements necessitated that
alternative disinfection methods be considered.
Chlorination plus dechlorination and ozonation were
evaluated. It was determined that chlorination plus
dechlorination would have required continuing
chemical costs and necessitated complicated control
equipment to feed and maintain proper dosages.
Additionally, dissolved oxygen uptake during
dechlorination chemical addition would have required
reaeration capabilities prior to effluent discharge.
Finally, the hazards and costs of transporting chlorine
to the mountain plant site made the chlorination
alternative unattractive. With ozone, it was felt that
disinfection standards could be maintained, dissolved
oxygen would be added to the effluent and the ozone
could be generated on-site. Studies had also shown
that ozone was effective in color and turbidity
reduction, which would be beneficial because of the
plant discharge going to the high quality waters of the
Big Thompson River. The ozone system was finally
selected as the disinfection process.
HOW t
Xlank
qualization
Grit
l|J Chamber 1
^Rate Controller .
1
Aeration
^
To Sludge
Handling
Effluent
Wastewater Flow •
Ozone
Contact
Basin
Multi-Media
Pressure Filters
Sludge Flow
Figure 1. Plant flow schematic diagram for the UTSD.
176
-------
OZONE
OZONE SYSTEM DESCRIPTION
The UTSD ozone disinfection process consists of
two air-fed ozone generation systems and one ozone
contact basin. Each ozone generation system is
designed to provide adequate disinfection at the plant
design flow rate of 5,680 cu m/ day (1.5 mgd) The other
generation system was provided for stand-by capacity.
The two ozone generation systems were identically
constructed and were labeled No. 1 and No. 2. The
ozone research data presented in this report is for
Generation System No. 2.
Ozone Generation System
At the USTD plant, ozone is generated with
Welsbach air-fed units, Model CLP-68F20L. A
schematic diagram of the ozone generation equipment
is shown in Figure 2. Three components make up each
ozone generation system: generator, power supply,
and air pretreatment. Each generator was designed to
produce 1.43 kg/hr (76 Ib/day) of ozone at a
minimum concentration of 1% by weight. This
provided for a maximum ozone/liquid dosage of 6
mg/1 at the plant design flow of 5,680 cu m/ day (1.5
mgd). The generators are "iron lung" tube-type units
containing 68 discharge chambers. The generators are
water cooled using potable well-water at a rate of
approximately 1.26 l/sec(20gpm). Power is supplied
to the generator through variable voltage
transformers. A controller assembly is used to adjust
the voltage from the transformer to the generator,
which controls the ozone generator output. The
controller was designed for manual or automatic
adjustment. During the research project, manual
adjustments based on generator amperage readings
were used.
TO AIR
PRETREATMENT
DRYING TOWERS
OZONE GENERATOR
Figure 2. Ozone generation schematic diagram for UTSD.
The air pretreatment components of the ozone
generation system are an air compressor, refrigerant
drier and air drying towers. Air pretreatment was
designed to provide particle-free dry air with a dew
point of -51°C at a pressure between 0.42 and 1.05
kg/sq m (6 and 15 psig). During the research project,
the air pretreatment pressure was maintained at 0.53
kg/sq m (7.5 psig).
Ambient air is compressed to 0.53 kg/ sq m (7.5 psig)
by a Nash Model L3 water ring compressor. The
compressor operates at a constant speed and has an
output of about 160 cu m/hr (94 scfm). It should be
noted that standard conditions used throughout this
report are one atmosphere pressure and 25° C
temperature. A baffle separator is provided to
separate the water and air. A bleed-off air valve is
provided downstream of the baffle separator. The
bleed-off valve at this point was determined to be
necessary during the research project to enable the air
flow rate to the drying towers to be controlled to
prevent overloading of these units.
Compressed air is cooled to between 3.3°C and
5.6° C in a Zeks, Model 9 J refrigerant drier, in order to
remove excess moisture in the air. The unit was
designed so that the air dew point does not exceed
8.9°C.
Refrigerant dried air is further dewatered to a dew
point less than -51°C in Kemp Model 100 UEA-1
absorptive drier. The drier uses molecular sieves and
activa-ted alumina as absorptive material. Dual towers
are provided for continuous operation. Tower
operation is cycled at 8-hour intervals to provide
regeneration of one tower while the other is in use.
Each tower was rated by the manufacturer at 131 cu
m/hr (78 scfm) air flow.
Several design modifacations and operation and
maintenance changes were implemented based on
problems encountered with the air pretreatment and
ozone generator units. These problems and the
approach to their solutions are further described later.
Ozone Contact Basin
The ozone contact basin is located adjacent to the
tri-media filters in the main control building. The
ozone generation systems are located directly above
the contact basin. The ozone contact basin is 1.30 m
(4.25 ft) wide, 11.6 m (38.0 ft) long and 3.66m (12 ft)
deep, which would give an ozone contact time of 14
min at 5,680 cu m/day (1.5 mgd) design flow. The first
8.22 m (27 ft) of the ozone basin is divided into nine
equal sized compartments with U. P. V. C. baffles. The
177
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
baffles are placed to allow vertical, serpentine
flow of wastewater through the basin. A schematic
of the contact basin is shown in Figure 3.
Treated water from the ozone contact basin passes
over a weir and into a backwash water storage basin
where it is used for backwashing the tri-media filters.
The backwash water storage basin is located beneath
the main offices for the UTSD facility. Effluent
discharges from the backwash water storage basin
through a morning glory overflow weir to the Big
Thompson River. The ozone contact basin and the
backwash water storage basin share a common air
space above their water levels due to the described
overflow weir arrangement. Ozone off-gas problems
within the main offices occurred because of this
arrangement, and modifications were necessary to
correct these problems.
The contact basin is covered with aluminum plates
which are bolted in place. Hypalon gasket material
was placed beneath all joints. Additionally, silicone
sealant was used to cover all exposed joints. Ozone
laden off-gas from both the contact and backwash
water storage basins is vented to the roof of the main
control building. The exhaust fan is located in the roof
and provides a negative pressure (about 1/4 inch
water) in each basin so as to prevent ozone leakage
into the main control building and/ or offices. A water
spray nozzle is located in the vent duct above the ozone
basin tank cover to prevent foam from blocking the
exhaust air flow.
Ozone is injected into the wastewater in the contact
basin through porous stone diffusers. The diffusers are
Kullendite, Model FAO 50 as manufactured by Ferro
Corporation. The diffusers are located in each of the
nine baffled areas. Each diffuser is 6.4 cm (2-1/2 in) in
diameter and 61 cm (24 in) long, and has an air
permeability between 12and 15 scfm/ ft2/ in. thickness
and a maximum pore diameter of 140 microns.
Distribution piping consists of type 304 schedule 40
stainless steel piping with welded or threaded joints.
Distribution of ozone to each compartment is
controlled by nine individual valves. Both the diffusers
and distribution piping represent modifications of the
original equipment.
Four adjustable height weir scum skimmers are
located along the length of the ozone contact basin to
facilitate removal of any scum that may be generated
as a by-product of ozonation. Scum removed by the
skimmers can be pumped to the head of the plant for
recycle or to the secondary clarifiers. To date, only
small amounts of scum have formed which has been
easily handled by the existing scum skimmers.
A proposed ozone off-gas destruct unit is shown in
Figure 3. A heat/ catalyst ozone destruct unit has been
designed and is currently being constructed. Tie
ozone destruct unit represents a major design
modification which will be further discussed later.
BACKWASH,
BASIN VENT~~
OVERFLOW , .. M [J-H-. OZONE
WEIRX ci«iiiiipn LJ F - -- DESTRUCT UNIT
'COVER (SKIMMER 1 "[PROPOSED!
/ PLATES) \ WATER SPRAY-^.f- ., ,„_._ _„.„
J ^
0
b
. y, _j
6
a
6
6
b
3
O:
-
i.
- INFLUENT
- UPVC BAFFLE
OZONE DIFFUSER
_,/'
Figure 3. Ozone contact basin schematic diagram
for the UTSD.
OZONE SYSTEM DATA COLLECTION
Data collection for the ozone system comprised
several different areas: chemical and microbiological
analyses, ozone in air concentrations and mass
measurements, ozone in water-concentration
measurements, electrical power consumption
measurements and miscellaneous measurements. In
this section of the pa per the equipment and procedures
used to collect these data are described.
Chemical and Microbiological Analyses
Most chemical analyses of the ozone contact basin
wastewater influent and effluent were conducted on
samples which were collected at 2-hour intervals, 24-
hours per day, 5-days per week (Sunday through
Thursday). For some chemical and for all
microbiological analyses, grab samples were taken
because of the nature of the test. The various chemical
and microbiological analyses conducted and the type
of sample collected is shown in Table 2. All analyses
were conducted in accordance with procedures
outlined in Standard Methods for the Examination
of Water and Wastewater, 14th Ed. (2).
Ozone in Air-Concentration and Mass Measurement
The ozone concentration in the ozone/ air flow from
the generators was measured by a wet chemistry
procedure and by a high concentration continuous
reading ozone meter (Dasibi High Concentration
Ozone Meter, Model 1003-AH). The wet chemistry
method involved a sodium thiosulfate titration of 3
prepared solution of potassium iodide which had bee i
178
-------
OZONE
TABLE 2. LIST OF CHEMICAL AND MICROBIOLOGICAL
ANALYSES CONDUCTED ON THE UTSD OZONE
DISINFECTION SYSTEM
Parameter
Total Suspended Solids (TSS)
Chemical Oxygen Demand
(COD)
Total Kjeldahl Nitrogen (TKN)
Ammonia Nitrogen (IMHg-N)
Nitrate/Nitrite Nitrogen
(NO3NO2 - N)
5 Day Biochemical Oxygen
Demand (BODg)
Total Phosphorus (Total P)
Alkalinity (Alk)
Total Coliform (T.Coli)
Fecal Coliform (F.Coli)
Dissolved Oxygen (D.O.)
PH
Frequency
5/wk
5/wk
5/wk
5/wk
5/wk
2/wk
5/wk
5/wk
5/wk
5/wk
7/wk
5/wk
Type of Sample
24 hr. Comp.
24 hr. Comp.
24 hr. Comp.
24 hr. Comp.
24 hr. Comp.
24 hr. Comp.
24 hr. Comp.
24 hr. Comp.
Grab
Grab
Grab
Grab
exposed to a known volume of the ozone/air flow
stream.
An alternate acceptable method of monitoring
ozone concentration in air was available after the
Dasibi continuous reading ozone meter was properly
set up and calibrated. Originally, meter readings did
not correlate with wet chemistry results. The problem
was isolated to the ozone/air and purge air flow rate to
the meter. Lowering and controlling the ozone/ air and
purge air flow rates to about 2 1/min resulted in
consistent meter readings that correlated well with wet
chemistry results. A comparison of meter results with
wet chemistry results is shown in Table 3 and Figure 4.
A very good correlation of results exists. Due to the
Q
5 6000
Q.
Q.
o5000
z
§ 4000
UJ
oc
DC 3000
UI
1-
Ul
Z 2000
m
W 1000
o
/
^^— DIRECT COR
o ACTUAL
J
/
/
RELAT
/
ON
\/
0 1000 2000 3000 4000 5000 6000 7000
WET CHEMISTRY RESULT- PPM IVOLI
Figure 4. Comparison of Dasibi meter and wet chemistry
ozone/air concentration measurements.
good correlation between wet chemistry and Dasibi
meter results, it was concluded that the Dasibi meter
could be used to determine the ozone in air
concentrations so that data points could be more
readily obtained. However, the Dasibi meter was used
to obtain only about 25 percent of the ozone/air
concentration measurements used in this report,
because the problem with the meter was not corrected
until the later part of the research evaluation.
The ozone/air concentration was combined with
the generator air flow to obtain the mass of ozone
produced. The generator air flow was measured with a
Fischer and Porter Series 10A3500 "Flowrator"
meter. The recorded flow was corrected to standard
pressure and temperature conditions of one
atmosphere and 25° C.
TABLE 3. SUMMARY OF COMPARISON OF CONTINUOUS
MEASUREMENT DASIBI OZONE METER RESULTS
WITH WET CHEMISTRY RESULTS
Date
Dasibi Meter
Span Setting
Number of Tests
Average Wet
Chemistry* Result
[ppm(vol)]
Average Actual
Dasibi* Reading
[ppm(vol)]
Average Difference
(%)**
Range of Difference
(%) -0.17
7/19/78
80570
4
5598
5532
0.12
to 2.0
7/21/78
80570
9
5372
5351
0.39
-0.72 to 2.65
7/25/78
80570
5
3222
3204
0.56
-0.71 to 1.66
* Corrected to standard conditions of 1 atmosphere pressure and 25°C temp.
(Wet Chemistry - Actual Dasibi) (100)
Percent Difference =
(Wet Chemistry)
The ozone/air concentration of the contact
basin off-gas was measured using a wet chemistry
procedure. The flow rate of off-gas through the
vent duct was measured using a pitot tube. The
pitot tube was not ozone resistant so the off-gas
flow rate determinations were made with the
ozone generator shut down and only the air pre-
treatment system running. Off-gas flow measure-
ments were taken at different air flow rates from
the air pretreatment unit. The data points were
very reproducible and a curve was developed re-
lating off-gas flow rate to air flow rate from the
ozone generation system. The curve was used to deter-
mine off-gas flow rate during testing of the ozone
179
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
disinfection system. The off-gas flow rate, corrected to
standard conditions, was coupled with the off-gas
concentration and was used to obtain the mass of ozone
contained in the contact basin off-gas.
Ozone In Water - Concentration Measurements
Ozone residual concentrations in the effluent from
the ozone contact basin were initially made using a
volumetric titration procedure. Using this procedure it
became apparent that the color change at the end of
the titration was nearly impossible to detect. It was
decided that an amperometric titration method would
be employed. An amperometric titrator was pur-
chased and was used to obtain ozone/water concen-
trations. With the present operating procedures,
good disinfection results were often obtained with no
detectable or negligible ozone residuals.
Electrical Power Consumption Measurements
Ozone generator and air pretreatment power
consumption measurements were made with a
Sangamo type S3DS watt-hour meter, which was tied
into the electrical feed lines to both the ozone
generator and air pretreatment units. The meter
provides the capability to determine totalized kilo-
watt-hour readings, maximum kilowatt demand
readings and instantaneous kilowatt demand
readings, which are the basis of the power cost
determination at the UTSD plant.
Miscellaneous Measurements
Other measurements and gauge readings were taken
in conjunction with generator production
determinations. The air pretreatment dew point was
measured with a Shaw Model "S" Mini Hygrometer
which had a Red Spot probe. Using the meter, the
changes in air dew point from the air pretreatment
system were recorded throughout the day. The air dew
point typically ranged from -70° C to -54° C.
The air compressor seal water pressure reading was
recorded daily and also every time ozone generator
production testing was conducted. The seal water
pressure was the line pressure of the water that entered
the water ring air compressor, and was located before
a filter screen in the line. The seal water pressure
indicated the relative flow of water to the water ring
compressor, in that a higher pressure forced more
water to the compressor when the filter screen was
clean. Varying seal water pressure resulted in
operational problems with the ozone generator, which
are described later. A greater or lesser water flow to the
compressor was important in that it influenced the
compressed air temperature, which in turn affected
power consumption of the refrigerant drier and will be
discussed later.
RESULTS AND DISCUSSION
The discussion of results from the research
evaluation of the UTSD ozone disinfection system are
separated into four general categories: ozone
generation, ozone system power requirements, ozone
contacting system, and disinfection performance.
During the UTSD research evaluation the ozone
disinfection system was intermittently operated.
Intermittant operation resulted from numerous
problems requiring design or operations changes.
Several of the problems encountered were associated
with the state-of-the-art design of one of the first full-
scale ozone wastewater disinfection systems in the
United States. Much conflicting and confusing
information was provided by various "ozone experts"
about the UTSD ozone disinfection system. The
conflicting information provided often delayed the
correction of the design and operational problems
encountered. The major disadvantage of the
intermittent operation of the ozone disinfection
system is that a thorough evaluation of disinfection
performance was not achieved.
Ozone Generation
The three components of the ozone generation
system are air pretreatment, ozone generator and
power supply. These three areas were evaluated both
separately and in combination during the course of the
research effort. The ozone generation system
performance is dependent upon good quality dry air.
As such, the operation and performance of the air
pretreatment system is just as important as the
. operation and performance of the generator itself.
Air Pretreatment. Several problems were
encountered with the air pretreatment system, most of
which have been corrected. However, the potential for
some problems to reoccur exists and studies are in
progress to eliminate this potential.
The air compressor for each air pretreatment system
is a constant speed unit and continuously discharges
an air flow of 160 cu m/ hr (94 scfm). By bleeding-off
excess air, the air pretreatment system can operate at
variable air flow rates. Originally, excess air from the
compressor was bled-off after passing through the air
drying tower. Each tower is rated at a maximum air
flow of only 130 cu m/hr (78 scfm), and thus was
constantly overloaded. This caused an .excessively
high dew point of the "dried" air. Subsequently, a
180
-------
OZONE
minor modification was made by installing a bleed-off
valve after the air compressor and before the air drying
tower so that excess air was bled-off before the air
drying tower. This modification solved the
operational problems of excess air flow to the drying
tower. Variable air flow not to exceed the air
pretreatment system's ability to provide dry air is
highly desirable, but more economical means of
providing variable air flow are required and will
be discussed later.
Compressed air is directed through a refrigerant
drier that has an input voltage of 440 volts. This
voltage is compatible with the voltage to the ozone
generator, but when operational problems were
encountered with the refrigerant drier it was quickly
learned that all parts locally available were for 220 volt
refrigeration units. These parts were not suitable for
the installed 440 volt refrigerant drier, and parts had to
be special ordered which delayed the unit's repair. The
time delay for repair of the refrigerant drier was the
main reason Generation System No. 1 was not
operable during most of the time the production data
used for this paper was developed. To maintain both
ozone generation systems fully operational, the UTSD
will have to purchase additional spare parts for the
refrigerant driers. Also, the UTSD staff was not
trained nor had the equipment to repair the refrigerant
unit. The corrective maintenance and spare parts
problems encountered with the UTSD 440 volt
refrigeration system should be considered in the design
of air pretreatment capability for other ozone systems.
Each of the two air drying towers contains activated
alumina, molecular sieves and alumina balls desicant
material to absorb the water in the air and lower the
dew point to less than -51°C. One tower "dries"air
from the compressor while the other tower is
regenerated by a combination of heating the desicant
material to release the bound water and puring the
tower contents with dry air to remove the excess
moisture. The towers are cycled for drying and
regeneration on 8-hour intervals.
Generally, the air drying towers worked well after
the air bleed-off valve was moved to a location in the
air flow scheme which was prior to air entering the
towers. However, a malfunction of the tower
switching mechanism used to alternate the towers
from the drying to regeneration cycles occurred. This
problem occurred due to "sticking" of the linkage of
the pneumatically operated switching mechanisms. The
arm was lubricated and this problem was corrected,
but the potential for this problem to reoccur still
exists. The signal for the tower change is electrical
while the tower switching mechanism is pneumatic. If
problems are encountered with the pneumatic
switching mechanism, the electrical system will still
indicate that the towers are functioning normally even
though "wet" air could be passing through the tower
that is regenerating. If this occurred, excessive
moisture could be directed to the ozone generator.
Under this condition the ozone generator could be
"flooded". The term flooded is used to describe
moisture build-up in the ozone generator which causes
short-circuiting and can cause electrode tube and or
fuse failure.
On several occasions flooding of the ozone
generator did occur. On one occasion flooding
occurred due to a problem with the refrigerant drier.
The refrigerant drier motor overheated and burned
out, for an as yet unknown reason. A new refrigerant
drier has been ordered. When received, careful
electrical checks will be made during its installation to
try to isolate the problem.
Flooding of the ozone generator on three other
occasions has been tentatively associated with the seal
water pressure of the water ring air compressor. These
three generator floodings were expensive because
several electrode tubes and fuses blew out. As
described earlier, the greater the seal water pressure
the greater the water flow rate through the compressor
and the lower the temperature of compressor air. The
temperature of the compressed air is important
because if the temperature is too high the refrigerant
drier cannot cool the air to reduce the dew point and
the air drying tower will be overloaded. The flooding
problems were believed to have occurred due to
plugging of an in-line filter screen which was used to
remove any particulate matter from the water that was
directed to the compressor. Plugging caused the flow
of water to the compressor to decrease and the
temperature of the compressed air to increase, which
eventually led to overloading of the drying tower. The
entire seal water pressure system is being re-evaluated
to determine what design modification or preventive
maintenance checks can be instituted to reduce the
frequency and/or effects of the screen plugging
problem.
Although numerous sources of ozone generator
flooding problems have been isolated, the possibility
of flooding still exists. Based on this experience, it
appears that a high dew point alarm and an associated
automatic generator shut-off is necessary to prevent
the expense and loss of production associated with
181
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
generator flooding. Originally, the relative air dew
point was monitored with cobaltous chloride pale blue
to pink color changing indicator. This indicator is
inadequate as an alarm or protection device for the
ozone generator for two reasons: 1) someone has to see
the color change and shut down the system before
flooding occurs and 2) the color change is not sensitive
to gradual changes in dew point so potential problems
cannot be detected until they are quite far along.
A dew point indicator was purchased as part of the
research project to obtain more exact information on
air dew point versus generator production. This
indicator has greatly aided in the detection of changes
in dew point and correction of problems before
flooding of the generator occurred, but has not solved
the generator flooding problem. Observations of the
indicator on a continuous basis is still required. As a
better solution, a high level dew point alarm and
associated automatic generator shut-down is being
considered. To date, this would have saved the UTSD
at least $1,600 in electrode tubes and fuses during the
past two years of operation.
Ozone Generator Production. Many parameters
can influence the rate of ozone production including:
power supply, air dew point, ozonated air temperature
which is influenced by the volume of air directed
through the generator and the ozonated air heat
removal capability of the cooling water jacket. These
items are not all inclusive, but were selected for
analysis because of their potential application to a
variety of ozone generation systems.
The UTSD ozone generators were designed to
supply ozone at a rate of 1.43 kg/ hr (76 lb/day) at an
air flow of 118 cu m/hr (70 scfm). The ozone
production for Generation System No. 2 only is
discussed, because Unit No. 1 was not operational as
described earlier. The production rate for different air
dew point levels will be presented because the ozone
production rate decreased as the dew point of the air
increased. This change in production may be a
significant operational consideration with respect to
continuous, satisfactory wastewater disinfection. The
production rate will also be compared to the relative
power setting of the generator for two different air
flow rates, namely the design air flow rate of 118 cu
m/ hr (70 scfm) and a lower air flow rate of 79 cu m/ hr
(47 scfm).
Ozone Production Versus Air Dew Point. The air
dew point readings increased proportionately to the
drying tower operating time. Typically, soon after a
regenerated tower came online and began drying, the
air dew point reached its lowest level. As the tower
dried more air, the dew point increased. Apparently,
as the desicant absorbed and contained more and
more moisture, less moisture was absorbed as
indicated by the dew point readings. The rate of
increase of the dew point was greater at the design air
flow rate of 118 cu m/hr(70scfm)thanatthelowerair
flow rate of 79 cu m/hr (47 scfm). The changes in air
dew point for the two air flow rates versus drying
tower operating time is shown in Figure 5.
012345678
DRYING TOWER OPERATING TIME-MRS.
Figure 5. Change in air pretreatment dew point with drying
tower operating time (scfm x 1.70 = cu m/hr).
The drying time per operating cycle for each tower
was eight (8) hours. The lowest dew point reading
shown in Figure 5 is -12° C, although readings as low
as -74°C were achieved. The highest dew point
recorded was -54° C. All dew point levels recorded,
except when obvious air pretreatment problems were
noted to cause generator floodings, were better than
the manufacturer's rated minimum dew point level of
-51°C. In this regard, the UTSD ozone air
pretreatment system functioned very satisfactorily.
The change of ozone production per degree of
change in dew point was evaluated at different
generator power settings. Two power settings were
evaluated and results are shown in Figure 6 and in
Table 4. In general, the ozone production level
decreased as the air dew point increased. It is noted
that all the changes in ozone production occurred at
better than the manufacturer's rated dew point level of
182
-------
OZONE
-74 -70 -66 -62 -58 -54
AIR DEW POINT -°C
Figure 6. Change in ozone production with dew point
at two generator power settings (Ib/day x 0.0189 = kg/hr).
As shown in Table 4, the mass decrease in ozone
production per degree change in dew point was about
three times greater for the 130 amp power setting than
for the 40 amp setting. However, the percentage
change was slightly less for the 130 amp setting at
l.70%/°C than for the 40 amp setting of l.93%/°C.
From the data presented it was concluded that the
overall decrease in ozone production varied
considerably with a decrease in dew point, which is
important with respect to ozone dosage to the
TABLE 4. EFFECT OF DEW POINT AND POWER SETTING
ON OZONE PRODUCTION
Maximum
Production
Minimum
Production
Dew Point
72° C
,70°C
56.5° C
59.5° C
130 Amps
57 Ib/day
42 Ib/day
40 Amps
17.1 Ib/day
13.6 Ib/day
Total Dew Point
Increase
Ozone Production
Decrease
Specific Ozone
Production
Decrease
Mass Decrease
Percent Decrease
15.5°C 10.5°C
15 Ib/day 3.5 Ib/day
0.97 Ib/day 0.33 Ib/day
1.70%/°C 1.93%/°C
wastewater. These findings would be expected to
apply to all ozone generation systems. However, the
magnitude of the decrease in ozone production per
degree increase in dew point may vary from generator
to generator. The implications of these findings on
design and operation are significant.
At the UTSD plant the wastewater flow rate was
controlled through a flow equalization basin and a
negligible variation in daily plant flow occurred. The
ozone dosage to the wastewater was manually
controlled by adjusting the generator power setting.
However, at a given power setting the ozone dosage
decreased as the dew point increased. The potential
magnitude of the decrease in ozone dosage for
observed changes in air dew point is summarized in
Table 5.
As shown in Table 5, the ozone dosage could vary
from 5.0 to 3.7 mg/1 if the air dew point changed from
-72 to -56.5°C. A change in dosage because of dew
point is important from a design and operation basis
because disinfection performance is heavily influenced
by ozone dosage. Therefore, ozone production
information at different power settings and at variable
dew point levels are required in order to properly
design and operate ozone disinfection systems. In the
final analysis it may be required that ozone systems be
designed with multiple units which have the air
TABLE 5. POTENTIAL DECREASE IN OZONE DOSAGE
FOR OBSERVED CHANGES IN AIR DEW POINT
Wastewater Flow (mgd)
Ozone Dosage
(mg/l)
(Ib/day)
Air Dew Point*
-72° C
1.37
5.0
57"
Air Dew Point *
-56.5" C
1.37
3.7
42"
Ib/day x 0.0189 = kg/hr
mgd x 3785 = cu m/day; Ib/day x 0.0189 = kg/hr.
" It should be noted that the manufacturer's minimum rated dew point level
is -51°C.
"For generator power setting at 130 amps.
pretreatment drying towers changed sequentially so as
to reduce the overall effect of an increase in dew point
on ozone production.
It should be noted that the sensitivity of ozone
dosage to system disinfection capability was not
evaluated due to the intermittent operation of the
UTSD ozone system. It may be that an ozone dosage
between 3.7 and 5.0 mg/l yields the same general'
disinfection level, especially when other system
variables like influent wastewater COD, TSS and fecal
183
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
coliform concentrations are considered. If the
disinfection capability is not sensitive, the ozone
dosage variation because of dew point would not be as
critical as earlier described. This aspect should be
further evaluated. However, it is still concluded that
more information should be developed by ozone
manufacturers on ozone production versus dew point
levels and ozone production versus generator power
settings in order to provide design engineers and plant
operators with a better basis for ozone system design
and operation.
Overall Ozone Production. The major factors
affecting ozone production are generator power
setting, air dew point and ozonated air temperature
which is influenced by the air flow rate to the ozone
generator and the heat removal capability of the
cooling water jacket. The cooler the temperature of the
ozonated air, the less ozone will be decomposed after it
is generated. The UTSD ozone generator begins
producing ozone at a consistant, reproducible level at
a power setting of 40 amps. The maximum power
setting tested was 150 amps, when the generator
voltage was 450 volts. For the production evaluation,
the mass of ozone produced was determined for power
settings at 10 amp intervals between 40 and 150 amps.
The ozone production values shown in this paper were
taken when the air dew point was between -70°C
and -74°C. This dew point level represented the
best condition for ozone production within the
limits of the air pretreatment unit.
To insure that the cooling water system was
performing at optimum conditions, the cooling water
jacket was inspected for possible scaling which could
have reduced its heat removal effectiveness. No scaling
was noted at cooling water jacket sites that were
inspected. This was as expected because no scaling
problems were encountered on other equipment in the
plant that used the same water supply. The
temperature of the cooling water ranged between 10° C
and 12°C, which was within the ozone manufacturer's
specifications.
The actual temperature of the ozonated air was not
recorded. However, the relative temperature of the
ozonated air was investigated within the limits of the
UTSD system by adjusting the air flow rate to the
generator. The design air flow rate of 118 cu m/ hr (70
scfm) was expected to develop the lowest ozonated air
temperature, and in turn produce the highest mass of
ozone at a given power setting. The lower air flow rate
of 79 cu m/hr (47 scfm) was expected to produce a
lower mass of ozone, especially at the higher power
settings. The ozone production levels for the two air
flow rates is shown in Figure 7. As shown, ozone
production is nearly the same for both air flow rates at
all generator power settings, although the higher flow
rate generally had a slightly higher ozone production
level. Apparently, the ozone air temperature change
and hence ozone production was not significantly
affected within the range of air flow rates tested. The
fact that little production difference was shown for the
UJ
z
o
N
O
10
A = 70 SCFM AIR FLOW
o=47 SCFM AIR FLOW
40 50 60 70 80 90 100 110 120
GENERATOR SETTING - AMPS
Figure 7. Ozone generator production at various generator
power settings (Ib/day x 0.0189 = kg/hr);
(scfm x 1.70 = cum/hr).
130 140 150
184
-------
OZONE
lower air flow rate is significant because a smaller air
compressor apparently could have been used which
would have resulted in an electrical power savings.
The ozone production levels shown in Figure 7
ranged from 0.32 kg/hr( 17 Ib/day) to 1.08 kg/hr(57
Ib/day), and a fairly uniform increase in production
occurred for each 10 amp increase of the power setting.
A slight decrease occurred at the highest amp settings,
probably due to a relatively higher ozonated air
temperature and associated increased destruction of
ozone due to temperature. The maximum ozone
production level was 1.08 kg/hr (57 Ib/day). This
occurred at the air flow rate of 79 cu m/hr (47 scfm).
The maximum amp setting was not tested at the design
air flow rate of 118 cu m/hr (70 scfm), because the
ambient ozone concentration within the building
became too high on that testing day and the ozone
generator was shut down.
The maximum ozone production level of 1.08 kg/ hr
(57 Ib/day) at 79 cu m/hr (47 scfm) air flow was 25
percent less than the manufacturer's rated value of
1.43 kg/hr (76 Ib/day). Based on comparable results
for other amp settings, it is not expected that the ozone
production level at the design airflow of 118 cu m/hr
(70 scfm) would be significantly higher. The reason for
the lower than design ozone production level was not
known and is still being investigated. All known
influences on ozone production were optimized during
the ozone production tests, including lowest possible
air dew point and clean (no scaling) water jacket. Also,
prior to production testing the generator was
thoroughly cleaned, all electrode tubes were removed
and checked for damage, and all tubes were replaced
according to the manufacturer's recommendations.
Ozone System Power Requirements
One of the advantages considered in the selection of
ozone for the UTSD plant was on-site production. On-
site production capabilities were selected over on-
going chlorine and dechlorination chemical costs and
chemical hauling in the canyon roads which led to the
plant. The initial cost of the ozone generation
equipment and the on-going power costs needed.to
generate ozone were considered in this selection.
During the research, an evaluation of the ozone
generation power requirement was made to determine
if the initial cost assumptions were adequate.
The UTSD ozone system has been intermittently
operated, and the typical operating procedure was to
dose at a rate which would insure disinfection. No
attempt was made to optimize ozone dosage. As such,
realistic values' for on-going power consumption to
achieve disinfection were not obtained. However,
power consumption of various units over the
operating range of the ozone generating system were
determined (i.e., ozone generation "mapping"). Each
major unit of the ozone system was separately
evaluated. Presently, power is consumed by the air
pretreatment system, by the cooling and seal water
system, and by the ozone generation process. In the
future, an ozone destruct unit for the contact basin off-
gas will add to the power consumption.
Air Pretreatment and Cooling Water Power Con-
sumption. Power consumption for the air pre-
treatment system included power for the air com-
pressor, refrigerant drier, air drying tower heater,
pneumatic control system air compressor, and
electrical control circuit. The power requirements
for the air compressor, refrigerant drier and air
drying tower heater were most significant. The
power requirement for the pneumatic control sys-
tem air compressor and electrical control circuit
were insignificant in terms of total power usage,
and were not included in the power consumption
evaluation.
The air compressor operated continuously and used
8.35 kw of electrical energy. The air compressor was
designed to provide a nominal 118 cu m/hr (70 scfm)
air flow to the ozone generator. As described, the unit
discharged 160 cu m/hr (94 scfm) of air, and theexcess
air had to be bled-off to avoid overloading the air
drying tower. It was also determined that generator
production did not significantly change from an air
flow rate of 118 cu m/hr (70 scfm) to a lower rate of 79
cu m/hr (47 scfm). It appears that a smaller air
compressor could have supported similar ozone
production with a significant savings in power
consumption, especially at lower than design ozone
production requirements.
The instantaneous power consumption of the
refrigerant drier was 2.0 kw. However, a lower average
daily power consumption was determined because the
drier operating time varied with the inlet air
temperature and air flow rate to the drier. As
described, inlet air temperature increased as the air
compressor seal water pressure decreased. Generally,
the average inlet air temperature was about 33°C. The
relationship between the refrigerant drier operating
time and inlet air temperature for an air flow rate of 79
cu m/ hr (47 scfm) is shown in Figure 8. At an inlet air
temperature of 33°C, the average refrigerant drier
operating time was 11.1 hrs/day. A similar evaluation
185
-------
PROGRESS IN WASTE WATER DISINFECTION TECHNOLOGY
for the higher flow rate of 118 cu m/hr (70 scfm)
indicated an average operating time of 14.4 hr/day.
These operating times were coupled with the 2.0 kw
instantaneous power consumption to determine the
average daily power requirements of the refrigerant
drier.
TABLE 6. SUMMARY OF POWER CONSUMPTION FOR THE:
AIR PRETREATMENT AND COOLING WATER UNITS
28 30 32 34 36
INLET AIR TEMPERATURE -°C
Figure 8. Refrigerant drier operating time at various inlet
air temperatures and an air flow rate of 79 cu m/hr (47 scfm).
The air drying tower used electrical energy in the
tower regeneration cycle. During regeneration the
tower was heated by an electrical heater that had an
instantaneous power requirement of 3.65 kw. The 8-
hour regeneration cycle consisted of tower "heating"
for 4 hours and "cooling" for 4 hours. During the 4-
hour heating cycle, heater operation was controlled by
a high temperature cut-off and a low temperature
start-up system. Therefore, actual heater operating
time was less than the total 4 hours. The average heater
operating time per heating cycle was determined to be
3.25 hours, resulting in a daily average electrical usage
of 1.48 kw.
Air pretreatment power consumption data is shown
in Table 6. The daily average power consumption for
the air compressor and air drying tower heater was not
affected by air flow rate. Power consumption of the
refrigerant drier was affected by air flow rate.
However, the net effect was that the daily average
power consumption was not significantly different for
the two air flow rates.
One other use of power for the ozone system was
supply water for ozone generator cooling and for air
compressor operation. This water was provided by the
Unit
Air Compressor
Refrigerant Drier
Air Drying Tower
Cooling and Seal
Water
TOTAL
Instantaneous
Power Daily Average
Consumption Power Consumption
(kw)
8.35
2.00
3.65
5.00
@ 47 scfm
(kw)
8.35
0.93*
1.48'*
2.10***
12.9
@ 70 scfm
(kw)
8.35
1.20*
1.48**
2.10***
13.1
scfm x 1JQ = cu m/hr
' Refrigerant Drier on-time at an average inlet air temperature of 33°C.
" Average drying tower heater on-time of 3.25 hours per 8-hour cycle.
'" Average potable water pump on-time of 10.5 hours. (95% of potable water
used for cooling and seal water).
plant potable water pumping system. A special power
consumption measurement was taken to determine the
power usage of the potable water system, and the
average daily power consumption was 2.2 kw. The
potable water demand for the ozone generation
operation was about 95 percent of the total plant
potable water usage. Therefore, the power required to
supply water to the ozone generation system was 2.10
kw, and is also shown in Table 6.
Ozone Generator Power Consumption. The power
requirement of the ozone generator increased as the
level of ozone production increased. The most
important consideration was the power required to
produce a given mass of ozone (i.e., power utilization
in terms of kwh/ Ib). Two different air flow rates were
used in determining generator power utilization, and
power consumption measurements were taken at 10
amp intervals starting where reliable and reproducible
ozone production began (40 amps) and were
continued to the generator's maximum setting (150
amps). Only Generator No. 2 was used for the power
consumption determinations.
Power consumption for the ozone generator had to
be carefully determined because power consumption
measurements for the air pretreatment and the
generator were combined in the readings obtained
from the single ozone system watt-hour meter. In
order to attain the power consumption for the ozone
generator, the power consumed by the air
pretreatment units that were operating at that time
was substracted from the total measured ozone system
power that was indicated by the watt-hour meter.
Using this procedure, reproducible ozone generator
power consumption values were obtained.
186
-------
OZONE
TABLE 7. OZONE GENERATOR POWER REQUIREMENTS
Amperage
(amps)
40
50
60
70
80
90
100
110
120
130
140
150
Ozone Production
@
47 scfm
(Ib/day)
16.8
22.8
27.4
31.5
35.1
43.1
45.9
50.0
51.8
55.4
56.3
@
70 scfm
(Ib/day)
17.1
22.9
32.5
36.2
40.0
43.9
47.2
50.3
53.9
56.8
Ozone Generator
Power Consumption
@
47 scfm
(kw)
4.4
6.2
7.8
9.7
11.4
15.1
17.0
19.7
21.4
23.5
25.1
@
70 scfm
(kw)
4.6
6.4
9.9
11.4
13.1
15.1
17.2
18.8
20.8
22.6
Ozone Generator
Power Utilization
@
47 scfm
(kwh/lb)
6.2
6.5
6.9
7.4
7.8
8.4
8.9
9.5
9.9
10.2
10.7
@
70 scfm
(kwh/lb)
6.5
6.7
7.3
7.6
7.9
8.3
8.7
9.0
9.3
9.6
scfm x 1.70 = cu m/hr; Ib/day x 0.0189 = kg/hr; kwh/lb x 2.21 = kwh/kg.
The ozone generator production, power 26
consumption and power utilization values for the two
evaluated air flow rates are shown in Table 7. 24
Generator power consumption varied from a low of
4.4 kw to a high of 25.1 kwas the generator amperage 22
setting increased. Ozone production also increased as
the amperage increased, but at a lesser rate than power m 20
consumption as evidenced by the increase in power ^
utilization. Power utilization increased from a low of z 1 8
13. 7 kwh/kg (6.2 kwh/lb) to a high of 23.6 kwh/kg j*
(10.7 kwh/lb), which is graphically illustrated in , 16
Figure 9. z
O
K 14
The power utilization values shown in Table 7 for ]*
Ozone Generator No. 2 were obtained under — 1 2
conditions that would yield maximum ozone — •
production. These conditions were discussed 3 10
previously and include-: air dew point equal to or less tt
than -70 C, all electrode tubes operational, negligible !JJ 8
scaling of the cooling water jacket and a recently Q
cleaned ozone generator. It should be noted that a a 6
series of power utilization measurements were taken
before Generator No. 2 was cleaned. The uncleaned 4
generator power utilization was an average 15 percent
greater than values that are presented for the cleaned 2
generator.
Total Ozone System Power Utilization (Existing
System). A summary of the total ozone system power °
requirements for each air flow rate evaluated is shown
in Tables 8 and 9. The power utilization values for the _. „
two air flow rates were compared graphically in ozone g
Figure 9. As shown, power utilization for the two air x
GE
\
\
NERA
*
GENERATOR PLUS
AIR PRETREATMENT
I
I AND
\
\
^
TOR C
-^
COO
1
)NLY
>
LING
*szc
4*
f
+ S
f°
W^
£
r
0 = 47 SCFM AIR FLOW
A= 70 SCFM AIR FLOW
10 20 30 40 50 60
OZONE PRODUCTION - LB/ DAY
Measured power utilization for the existing UTSD
eneration system (scfm x 1 .70 = cu m/hr; Ib/day
0.0189 = kg/hr; kwh/lb x 2.21 = kwh/kg).
187
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 8. TOTAL OZONE SYSTEM POWER REQUIREMENT AT AN AIR FLOW OF 79 CU M/HR (47 SCFM)
Generator
Power
Setting
(amps)
40
50
60
70
80
90
100
110
120
130
140
150
Ozone
Production
(Ib/day)
16.8
22.8
27.4
31.5
35.1
43.1
45.9
50.0
51.8
55.4
56.3
Generator
-------
OZONE
efficient electrical energy usage occurs at lower
ozone production requirements. The fact that the
power utilization increases dramatically at low
production requirements has significant impact
when applied to wastewater treatment plant
operation.
At the UTSD plant, variations in wastewater
volume and effluent quality occur. Although flow
equalization is available, plant flows have ranged from
1140 cu m/day (0.3 mgd)to3780cu m/day (1.0 mgd).
To optimize ozone dosages in line with these flow
variations, an adjustable ozone generation system is
required. An" adjustable system is available at the
UTSD, but low ozone production operating
requirements have caused the system to operate in the
least efficient electrical energy usage range.
Similarly, municipal wastewater treatment plants
are typically designed for flow rates greater than
existing rates. Also, because of the limited information
available, most ozone systems will probably be
conservatively designed with respect to ozone dosage.
The combination of these factors will result in lower
ozone production requirements than the design
capability of the system. At a lower ozone production
level, the power utilization value could be much
greater than at design production levels which would
GENERATOR PLUS
AIR PRETREATMENT
AND COOLING
GENERATOR ONLY
A 70 SCFM AIR FLOW
— ACTUAL
! THEORETICAL
10 20 30 40 50 60 70 80
OZONE PRODUCTION - LB/DAY
Figure 10. Comparison of theoretical and actual UTSD
ozone generation system power utilization (scfm x 1.70
cu m/hr; Ib/day x 0.0189 = cu m/hr; kwh/lb x 2.21 = kwh/kg)
result in operation of a proportionately less
economical ozone disinfection system than would
occur at design flows. The need for adjustable ozone
production exists, and in order to have this flexibility
and be more energy efficient a more uniform power
usage efficiency is desireable. Multiple ozone
generation and/or multiple air pretreatment units
should be considered in order to achieve more uniform
power utilization values for the expanded range of
ozone production requirements associated with
municipal wastewater treatment facilities.
Proposed Off-Gas Ozone Destruct Unit Power
Requirement. The UTSD ozone system was
intermittently operated because of periodically high
ambient ozone concentrations in the plant working
environment. Several modifications to the ozone
piping and contact basin were made to reduce the
ambient ozone concentration when the generators
were operated. One remaining modification is an
ozone destruct unit for the off-gases from the ozone
contact basin. The unit has been designed and is being
constructed, and represents another source of power
consumption associated with the ozone system. The
expected power consumption is between 8 and 10 kw.
Ozone Contacting System
Ozone produced in the generators was directed to
the ozone contact basin. Several design modifications
were made to the ozone contact basin and ozone
piping arrangement. Some of the modifications
represent state-of-the-art design changes that evolved
over the 2-year operation period of the UTSD ozone
system. The modifications made include: contact
basin covering, basin exhaust changes, baffle and
scum skimmer changes, and ozone piping and diffuser
replacement. In addition, an ozone destruct system
has been designed and is in the process of being
constructed.
Covering of Contact Basin. An initial obstacle in
operating the ozone disinfection system was the
presence of high ambient ozone concentrations in the
operator's working environment. Present minimum
acceptable standards for human exposure to ozone are
0.1 ppm by volume for a period not to exceed 8 hours.
During initial start-up ambient ozone concentrations
of 3-5 ppm by volume for 2-hour periods, with peaks
of 15-30 ppm by volume were encountered.
It was determined that a portion of these high
ambient ozone concentrations were the result of a
partially covered contact basin. Additionally, the part
of the tank which was covered was not sealed. The
189
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
contact basin was modified by covering the entire
basin with aluminum plates and sealing the joints with
hypalon gasket material. It was anticipated that a
covered and sealed basin would prevent off-gas leaks.
When the revisions were completed, ozonation was
again attempted but high ambient ozone
concentrations still occurred. This problem was
isolated to ineffective sealing of the basin cover. A
silicone caulk was applied in a continuous bead to
obtain a positive seal. High ambient ozone
concentrations above the contact basin were reduced,
but were then detected in the main plant offices. Off-
gas that could no longer escape through the basin
cover was being forced to the air space above the
backwash water storage basin. This basin was adjacent
to the ozone contact basin and below the main plant
offices. Additionally, contact basin off-gases were not
being properly vented to the roof discharge because
foam produced by the addition of air and ozone to the
contact basin was blocking the exhaust air flow to the
vent duct.
Ozone Contact Basin Exhaust System Revision. To
correct the problem of off-gas leakage from the
backwash storage basin, a duct connecting the existing
ozone contact basin with the air space above the
backwash storage tank was installed. The addition of
the duct eliminated high ambient ozone
concentrations in the plant offices, because when foam
blocked the movement of air above the water surface
of the ozone contact basin the ozone laden off-gases
would transfer to the backwash water storage basin
where they would be pulled back into the main exhaust
duct and discharged. This duct provided an alternate
means for ozone laden air to be removed when
foaming occurred. In addition to the duct
modification, a water spray nozzle was installed in the
ozone contact basin duct to depress the foam as it
developed. It should be noted that the excessive
foaming problem occurred during ozone system start-
up and typically lasted for only 2-3 hours.
Baffles and Scum Skimmers Revisions. During
initial operation of the ozone system it was determined
that short-circuiting of flow was occurring across the
top of the compartment baffles. This short-circuiting
resulted when air/ozone from the diffusers"air-lifted"
the water level in the basin. The baffles were raised
slightly, but not to the extent that air movement to the
exhaust vent was blocked.
The air-lift action caused by the diffuser system also
caused flooding of the scum skimmer mechanisms
which were initially set too low. The adjustment range
of the skimmer units was expanded and adjustment
handles were extended through the basin cover. To
facilitate adjustment of the overflow weir plates,
plexiglass sight windows were installed into the basin
cover above each skimming unit. To date, the
continuous use of the scum skimmers has been
unnecessary as little foam or the predicted gelatinous.
type froth has been produced. The lack of appreciable
foam or froth is believed to be due to the high quality
water entering the ozone contact basin. This water has
very little material (i.e., total suspended solids)
available to be coagulated into froth.
Ozone Piping Modifications. The combination of
modifications previously discussed allowed operation
of the ozone system whenever the wind was blowing
sufficiently to remove the ozone laden off-gases from
the area surrounding the discharge stack. The system
had to be shut down when the wind was not blowing
because high ambient ozone concentrations would
develop in and around the plant area. Plans were
developed to install an ozone destruct unit for the
contact basin off-gases to prevent excessive discharge
of ozone to the atmosphere. The UTS D was instructed
by State Department of Health officials to operate the
ozone system to achieve disinfection, but the District
was allowed to shut down the system when necessary
to prevent human exposure to high ambient
concentrations of ozone. Under this arrangement, it is
•estimated that the ozone disinfection system was
operated only about 50 percent of the time.
The UTSD ozone system was operating under these
conditions for about one year when problems were
encountered with leaks in the ozone piping from
the generator to the ozone diffusers. The original
ozone piping from the generator to the ozone diffusers was
Schedule 40, U.P.V.C. pipe with both solvent weld
and threaded joints. The problems with leakage
occurred around the joints and along straight pipe
sections near pipe hangers. This leakage problem may
have been caused by inadequate care during
installation; however, good plumbing practices by the
contractor were felt to exist because very few problems
occurred with other plant piping systems. Although no
definite conclusions could be developed, the leakage in
the ozone piping may have occurred as a result of
ozone exposure over a one-year time period and not
due to poor workmanship.
Because of the excessive ozone leakage of the
U.P.V.C. pipe, all piping was replaced with Schedule
40, Type 304 stainless steel. Some of the stainless steel
pipe connections were threaded and some" were
190
-------
OZONE
welded. Some of the threaded stainless steel
connections were noted to leak and were tightened,
but some could not be sealed and were welded. It was
concluded that welded connections provided the best
assurance for sealing ozone piping.
The U.P.V.C. piping within the ozone contact basin
was also replaced with stainless steel, primarily to
facilitate installation of new ozone diffuscrs that were
concurrently installed. The diffusers were connected to
the stainless steel pipe outside the basin, and the pipe
with diffusers was lowered into the basin aided by the
structural integrity of the stainless steel pipe. During
installation another reason for using stainless steel
pipe became evident. Upon removal of the U.P.V.C.
pipe it was observed that the portion of the pipe that
had been underwater was extremely brittle and
shattered easily when dropped. The U.P.V.C. pipe's
condition was not felt to be acceptable for long-term
operation and replacement with stainless steel pipe
was considered appropriate. These results indicate
that strong consideration should be given to using
suitable grade stainless steel pipe and welded
connections for all ozone/air piping.
Ozone Diffuser Modifications. The UTSD ozone
contact basin achieved ozone transfer efficiencies
ranging from 35 to 70 percent. Transfer' efficiency (TE)
as used in this context was calculated as follows:
(Mass of Ozone Produced - Mass of Ozone
in Off-Gas)(100)
Mass of Ozone Produced
Typically, the transfer efficiencies were between 50
and 60 percent, which were considerably less than the
design TE of 90 percent. One reason for the lower than
expected TE was breakdown of the ozone diffusers.
The original ozone diffusers were tubuler in shape and
were attached to a piping connection nipple with a "2-
part epoxy bond." The epoxy served in a structural as
well as a gas sealing capacity. Because the measured
transfer efficiencies were lower than expected, the
contact basin was drawn down and the diffusers were
inspected. It was noted that the epoxy had become
extremely soft. The diffuser manufacturer claimed
that the epoxy would probably become "a little soft"
when exposed to water. A new diffuser was exposed to
water for 3 months in the laboratory. No softening
effect was noted. During this time the TE had reduced
to only about 35 percent from 50 to 60 percent. The
contact basin was again drained and this time some
diffusers were noted to have completely separated
from the connection nipple and had fallen to the
bottom of the contact basin.
The original o/one diffusers were replaced with new
diffusers at the same time the stainless steel piping
modification was completed. The new ozone diffusers
were of similar material and were also tubular in
shape, but a bolted stainless steel connection and
hypalon gasket material were incorporated in the
diffuser construction. The reason for replacing the
original diffusers was to obtain a diffuser that
hopefully was more ozone resistant. Initial indications
are favorable, but operation with the new diffusers has
been limited and final conclusions concerning the
suitability of the diffusers has not been reached.
Evaluation of Transfer Efficiencies. After
installation of the new ozone diffusers and stainless
steel piping. TE tests were conducted and were found
to be similar to efficiencies that were attained when the
o/.one system was first started up, namely 50 to 60
percent. These values were better than the 35 percent
efficiency attained when some of the original diffusers
were known to have failed, but were still less than the
design efficiency of 90 percent. A re-evaluation of the
o/one contact basin design was made.
The UTSD ozone contact basin design was similar
to a basin that was designed and tested in 1971 at the
Louisville, Kentucky Wastewater Treatment Plant
The ozone basin at Louisville was reported to have
consistently achieved 90% or greater TE. At about thai
same time at the University of Louisville, research
work was conducted to measure the TE of a contact
basin similar to the basin that was tested at the
Louisville plant, and a paper was published
concerning the results (1). The author reported that
the 9091 to 95% TE that was achieved at the Louisville
plant could not be duplicated at the University
laboratory. Laboratory results indicated a TE of
about 50%. The difference in transfer efficiency was
attributed to a "with reaction" consumptive use of the
applied ozone at the Louisville plant, which was
ozonating secondary effluent. The tests completed at
the University of Louisville, which did not correlate
with Louisville plant results, were labeled "without
reaction" test results. These "without reaction" tests
also correlated well with ozone/liquid gas transfer
theories.
Transfer efficiencies achieved at the UTSD facility
have also correlated well with ozone/liquid gas
transfer theories, and are believed to be more in line
with the "without reaction" tests conducted at the
University of Louisville. The wastewater ozonated at
the UTSD plant is tertiary effluent and of considerably
better quality than typically associated with secondary
191
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
treatment.
The belief that the UTSD ozone contact basin was
performing similarly to the "without reaction" tests as
established at the University of Louisville was further
supported. Periodically, the ozone transfer efficiency
has reached a level of about 70 percent. When
this occurred, poor disinfection results were often
achieved even if disinfection previously occurred at
similar ozone dosages. At the same time, the TSS con-
centration through the basin increased. This condition
existed when the ozone generator was started after
being shut down for a day or two. The air pretreatment
system was always operated; thus air was continuously
diffused into the wastewater. When the ozone
generator was shut off or was operated at a very low
ozone dosage level, biological growth developed in the
basin in the form of a slime on the basin walls, the
U.P. V.C. baffles, and other surface media in the basin.
When the ozone generator was started and operated at
a higher ozone dosage, a "with reaction" ozone
consumption probably occurred and increased the
ozone transfer efficiency to near 70%. At the same
time, the biological slime would slough-off and
increase the contact basin effluent TSS concentration
which interferred with the disinfection capability of
the system. However, when the UTSD ozone contact
basin is operated on a continuous basis it is expected
that it will operate according to gas/liquid transfer
theories. It was concluded that the basin is achieving
expected ozone transfer efficiency for the quality of
wastewater treated. Based on these developments,
engineers should design ozone contact basin transfer
efficiencies based on ozone/liquid gas transfer
theories and not on pilot scale tests.
Biological slime build-up occurs in the UTSD ozone
contact basin when the ozone generator is shut down
and/or operated at a very low ozone dosage level.
Because of this problem with intermittent ozone
operation, a good microorganism reduction versus
ozone dosage relationship was not obtained. When
continuous ozonation and continuous good disinfec-
tion is achieved, the ozone dosage will be adjusted
to determine the minimum level necessary to achieve
disinfection. Transfer efficiency tests will be taken to
determine the effective ozone dosage as opposed to
applied ozone dosages (i.e., dosage excluding the
ozone lost in the off-gas) so that a common ground
comparision can be made with other systems that
achieve a better transfer efficiency.
Off-Gas Ozone Destruct System. Intermittent
ozone generator operation occurs because of periodic,
excessive ozone exposure to operating personnel.
Originally, the ozone exposure was due to several
sources: contact basin leaks, backwash storage basin
leaks, ozone piping leaks and excessive ozone in the
contact basin off-gas exhaust. Excessive ozone from
these sources, except contact basin off-gas exhaust,
have been corrected. Ozone operation is now
dependent upon wind direction and velocity to
appropriately disperse the contact basin off-gas
exhaust. To achieve continuous ozonation, several
options were considered to control the off-gas ozone
discharge. Among these were: heat destruct,
heat/catalyst destruct, activated carbon, recycle to
sludge and discharge through a tall stack. The option
selected was heat/catalyst destruct. Heat destruct was
rejected because of an excessively high power
consumption. Activated carbon was rejected because
of its explosive potential when combined with ozone.
Recycle to sludge and discharge through a tall stack
were rejected because they were felt to likely resu-lt in
transferring the problem to another area within the
plant. The heat/catalyst ozone destruct system for the
contact basin off-gases has been designed and is being
constructed. The system is manufactured by Emery
Industries. Cost, performance and operating
information of the heat/catalyst unit should be
developed when installed and operational. An off-gas
ozone destruct unit should be strongly considered for
all newly designed ozone systems.
Disinfection Performance
Operation of the ozone system at the UTSD plant
was sporadic due to a variety of problems that resulted
in high ambient ozone concentrations that represented
a hazard to operating personnel. Shown below is a
synopsis of the problems encountered.
Time Period Comment
June — Off-gas problems due to contact basin cover
December, 1976 problem.
December, 1976 — Good disinfection achieved, but equipment
February, 1977 to measure ozone concentration not
available.
March — Good disinfection achieved.
April, 1977
May — Poor disinfection achieved because of ozone
June. 1977 diffuser problem.
July — System shut down for inspection and repair
October, 1977 of original ozone diffusers.
October — • Relatively good disinfection, but only with
December, 1977 extremely high ozone dosages because of.
further problems with o/one diffusers.
January — Design, construction and installation of
April, 1978 new ozone diffusers completed.
May — Sporadic operation due to excessively high
September, 1978 ozone levels in and around the plant area due
to no off-gas ozone destruct system.
192
-------
OZONE
Routine collection of analytical data for the ozone
system was initiated during the week of December 12,
1976. In general, good disinfection performance could
be achieved when the ozone diffusers were in good
condition and the system was operated for an
extended period of time (several days). Disinfection
performance was poor when problems were occuring
with diffusers and when the ozone system was
operated on an intermittent basis. A summary of
performance data for selected time periods is shown is
Table 10. Periods 1, 2 and 3 represent data collected
when the original ozone diffusers were in operation.
Period 4 represents data collected after the new
diffusers were installed.
During the 8-weeks of Period 1, the original ozone
diffusers were new and were operating satisfactorily.
Very good disinfection occurred at an average ozone
dosage of about 11 mg/1. (Note: Effective dosage was
only 50 to 60 percent of the applied dosage.) The
effluent fecal coliform concentration was reduced to
30 per 100 ml, much better than the design standard of
200 per 100 ml. The COD reduction during Period 1
was 11.7 percent and the TSS reduction was 35.6
percent. Both the influent COD and TSS
concentrations were relatively low at 29.7 and 4.5
mg/1, respectively.
During Period 2, the average applied ozone dosage
was about 9 mg/1, but disinfection performance
deteriorated significantly. The effluent fecal coliform
concentration was 2,080 per 100 ml. The reason for the
poor performance was primarily attributed to
problems with the ozone diffusers. During Period 3
disinfection improved, but only after the applied
dosage was more than doubled to about 19 mg/1.
The performance data for Period 4 represents
information collected after the new ozone diffusers
were installed. The data was collected for two different
time periods because of problems with intermittent
ozone generation and biological slime build-up. As
during Period 1, good disinfection was achieved
during Period 4. The effluent fecal coliform concen-
tration was only 9 per 100 ml at an applied ozone
dosage of about 7 mg/1. (Note: Effective dosage
was still only 50 to 60 percent of the applied dosage.)
These performance data are limited, but indicate
that good disinfection can be achieved with the UTSD
ozone system. More definitive conclusions are
expected when the ozone destruct unit for the contact
basin off-gas is operational and continuous ozonation
can be implemented.
TABLE 10. SUMMARY OF PERFORMANCE DATA FOR SELECTED TIME PERIODS
Period
1
2
3
4
Date
2/27/77
To
4/23/77
4/23/77
To
6/18/77
10/16/77
To
12/10/77
4/16/78
To
4/29/78
Plus
5/21/78
To
6/3/78
Applied Effluent COD in
No. of Ozone Fecal Effluent &
Wks. Dosage* Conform Reduction
(mg/l) (ff/100 ml) (mg/l) (%)
8 11.* 30 26.2 11.7
8 9." 2,080 39.5 7.3
8 19." 91 43.3 7.5
4 7.* 9 35.6 10.2
TSS in
Effluent &
Reduction
(mg/l) (%)
2.9 35.6
6.0 22.1
7.6 8.6
3.1 3.0
During these time periods the system achieved a maximum of 50 to 60 percent ozone transfer efficiency.
"During these time periods the system achieved a maximum of about 35 percent ozone transfer efficiency.
193
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
CONCLUSIONS AND RECOMMENDATIONS
1. The "Dasibi Meter" continuous measure-
ment ozone concentration meter readings
correlated well with wet chemistry results after
the meter was properly set up and calibrated.
a. Ozone and purge air flow to the meter
was controlled at 2 1/min.
b. The meter span setting was adjusted
based on wet chemistry results.
2. The dew point of air from the ozone system
air pretreatment unit varied, but as a whole the
unit performed better than rated by the
manufacturer.
a. The rated minimum air dew point was
-51 °C.
b. The operating dew point ranged from
-74°C to -54°C over an 8-hour operating
cycle.
3. The ozone generator periodically "flooded"
and was damaged due to malfunctions to the air
pretreatment system.
a. The cobaltous chloride color changing
indicator provided to show an increase in air dew
point was not sensitive to gradual changes and
potential problems associated with dew point
could not be detected until they were quite far
along.
b. A dew point measuring device was
sensitive to gradual changes in dew point, but
"flooding" still occurred when the meter was
monitored only once per day.
c. A dew point high level alarm and/or
automatic system shut-down would substantially
reduce generator "flooding" potential and should
be incorporated in all air pretreatment unit
designs.
4. The air pretreatment refrigerated drier
required special maintenance considerations.
a. The drier voltage of 440 volts was
compatible with the voltage of the ozone
generator, but required special order parts
because 220 volt refrigeration units were more
common in the community.
b. Repairs to the refrigerant drier were quite
technical and had to be completed by an
experienced repairman who had special equip-
ment.
5. Ozone production decreased with an increase
in air dew point. (Note: Actual air dew point was
better than rated by the manufacturer.)
a. The ozone production decreased by 1.70
percent per °C change in air dew point at a
higher ozone production level, and by 1.9.3
percent per °C change at a lower ozone
production level.
b. A decrease in ozone production because
of an increase in dew point could decrease the.
ozone dosage to the wastewater from 5.0 mg/1 to
3.7 mg/1.
c. Ozone system air pretreatment units may
have to be designed with multiple components
that are changed in sequence so as to reduce the
overall effect of an increase in air dew point on
ozone production.
d. More information should be developed by
ozone manufacturers on ozone production versus
dew point levels over the entire ozone generator
operation range in order to provide design
engineers and plant operators with a better basis
for ozone system design and operation.
6. Ozone production of the LJTSD ozone
generator was 25 percent less than rated by the
manufacturer, even though production testing was
completed under controlled conditions as follows:
a. The air dew point was equal to or less
than -70°C.
b. The generator cooling water jacket was
free of scaling.
c. The generator cooling water temperature
was 10°C to 12°C, which was within the ozone
manufacturer's specifications.
d. All electrode tubes were removed, in-
spected, cleaned and replaced according to the
manufacturer's recommendations.
7. Ozone production of the UTSD ozone
generators was not significantly lower at air flow
rates of 79 cu m/hr (47 scfm) as opposed to air
flow rates of 118 cu m/hr (70 scfm).
8. Ozone generation system power utilization at
the UTSD plant was greater at lower ozone
production levels due to the relatively constant
power requirements of the air pretreatment unit,
and caused inefficient power usage under existing
operating conditions.
a. Total power utilization varied from about
55 kwh/kg (25 kwh/lb) at production levels of
0.32 kg/hr (17 Ib/day) to 33 kwh/kg (15 kwh/lb)
at production levels of 1.08 kg/hr (57 Ib/day).
b. Lower ozone production levels of 0.32
kg/hr (17 Ib/day) are usually sufficient at the
UTSD plant for current wastewater flow rates
194
-------
OZONE
and ozone dosage requirements. Thus, the least mg/1. (Note: Transfer efficiency was 50 to 60
efficient power utilization values presently exist. percent.)
c. More efficient power utilization values b. When the ozone system is operated for
over the entire range of ozone generation system only a short period of time, poor disinfection is
operation would improve system power experienced and increased effluent TSS concen-
consumption and cost at lower than design trations occur due to biological slime growth
wastewater flow rates and ozone dosage sloughing off.
requirements, and should be considered for all
ozone system designs.
9. Contact basin off-gas, ozone discharge and
other sources of ozone leakage have caused ex-
cessively high ambient ozone concentrations in
and around the plant area and have required that
several system design modifications be made.
a. The contact basin had to be totally
covered and sealed.
b. The off-gas exhaust system was re-
designed.
c. The basin baffles and scum skimmers
were modified.
d. The UPVC ozone piping was replaced
with stainless steel piping.
e. The off-gas ozone must be destroyed. A
heat/catalyst ozone destruct unit has been de-
signed and is being constructed.
10. The original ozone diffusers were not ozone
resistant. They failed and were replaced with new
diffusers.
11. The UTSD ozone contact basin was
designed for 90 percent transfer efficiency and
was based on incomplete information.
a. Ozone transfer efficiency was variable
and was affected by wastewater quality.
1. Typically, the transfer efficiency was
between 50 and 60 percent.
2. The transfer efficiency increased when
the wastewater quality was poor.
3. With good wastewater quality, mea-
sured ozone transfer efficiencies correlated
well with ozone/liquid gas transfer theory.
b. To insure achieving a desired, minimum
transfer efficiency, ozone contact basin design
should be based on ozone/liquid gas transfer
theory.
12. Good disinfection performance has occurred
at the UTSD plant when the ozone diffusers were
in good condition and the system was operated
consecutively for several days.
a. Effluent fecal coliform concentrations
averaged less than the design requirement of 200
per 100 ml at applied ozone dosages of about 7
195
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
REFERENCES
1. Hill, A. G. and H. T. Spencer, "Mass Transfer in Gas
Sparged Ozone Reactor," Proceedings of the First
International Symposium on Ozone for Water and
Wastewater Treatment sponsored by the International
Ozone Institute (1975).
2.. Standard Methods for the Examination of Water and
Wastewater, 14th Ed. 1975. American Health Associa-
tion, New York, New York.
ACKNOWLEDGEMENT
The work upon which this paper is based was
performed pursuant to EPA Research Grant Number
R-803831-01 entitled, "An Evaluation of Pollution
Control Processes — Upper Thompson Sanitation
District." The research effort was conducted under the
direction of Mr. Edwin F. Barth, Municipal
Environmental Research Laboratory, U.S. Environ-
mental Protection Agency, Cincinnati, Ohio.
The prime contractor for the project was the Upper
Thompson Sanitation District. Mr. Giles Gere,
Chairman of the Board, was the Principal Investigator
for the research effort. M & I, Inc. served as a
subcontractor for the District and acted as the
technical advisor for the research effort.
Appreciation is expressed to all of the UTSDand M &
I, Inc. personnel who worked on the research effort.
UTSD Staff:
Plant Superintendent - Mr. Bob Cheney
Lab Chemist - Ms. Barbara Baldwin
Plant Operators - Mr. Rawle Alloway
Mr. Larry Boehme
Mr. Roger Hess
Mr. Tim Hunter
Mr. Bob Tardy
M & I, Inc. Staff:
Project Technician
Lab Chemist
Project Engineer
- Mr. Jan Cranor
- Mrs. Sue Martin
- Mr. Larry Stanton
(Formerly with
M & I, Inc.; now
TST Associates)
Appreciation is also expressed to Mr. Edwin F. Barth,
Mr. Albert D. Venosa and Mr. Edward J. Opatken,
EPA MERL, Cincinnati, Ohio, for their direction
and assistance regarding the ozone system research
effort, and to Dr. Sumner M. Morrison Colorado
State University, Fort Collins, Colorado, for technical
advice regarding microbiological testing.
DISCUSSION:
MR. DALE WOOD: I would like to ask, num-
ber one, what was the flow capacity of that plant?
MR. RAKNESS: Design flow is 1.5 mgd, opera-
ting flow ranges from 0.3 to 1.0 mgd.
MR. WOOD: Does this take into account a
flood peak? In this area we have plants that will
have a flood condition that is above peak flow
during heavy rainfall.
MR. RAKNESS: The maximum hydraulic capa-
city through the plant is 3.5 mgd, but the 1.5 mgd
is the average daily operating design flow for
about twenty years from now.
MR. WOOD: Is the capability of your equip-
ment at that level?
MR. RAKNESS; The capability of the equip-
ment is for 1.5 mgd.
MR. WOOD: And is that a single unit or dual
units?
MR. RAKNESS: Good question. We have dual
units. One unit is operated as a standby. Each
unit is capable of 76 pounds per day, which gives
a design dosage of 6 mg/1 at a flow rate of 1.5 mgd.
DR. RICE: Quick comment in terms of your
remark about the uncovered contactors, which had to
be covered. Any ozone generator manufacturer that
would recommend uncovered contactors should be
run out of the business and never talked to again. That
cannot be considered a proper design.
MR. WHITE: I think this transfer efficiency is very
important. You went by it so fast you lost me; I
understand the 90% transfer, which means from what I
know about it that 10% comes in the off-gas. But, you
were showing 50 and 60% transfer efficiency. Will you
explain it again?
MR. RAKNESS: We measured transfer efficiency.
We measured the mass of ozone supplied to the ozone
contact basin, and the mass of ozone in the off-gas,
subtracted those two and divided by the mass of ozone
supplied, and that is what we call transfer efficiency.
MR. WHITE: Well, when you showed 50 to 60%
transfer efficiency, are you saying that 40 to 50% was
coming in the off-gas?
MR. RAKNESS: That is right.
MR. WHITE: Okay.
QUESTION: Instead of running a destruct set up
on your off gas, you are certainly throwing away some
very valuable material. Why not run a recycle line and
either go into a tower feeding the waste stream before
your initial diffuser contactors, which certainly most
196
-------
OZONE
chemical engineering handbooks would show you how another characteristic of 254 nm UV is that it is
to design that, or go to a second bank of diffusers destructive to ozone.
before you go into any destruct mechanism. That is
really throwing good dollars away.
MR. RAKNESS: We investigated different ways of
handling that off-gas. Among mem were heat-only
destruct, which was not cost effective at all. Another
way was diffusion into the sludge. Another one was
diffusion into the wastewater stream. As a result
of off-gas that has been in and around the plant
area and to which operators were exposed, the
operators, the superintendent, the board members,
and we did not feel that we wanted to chance trans-
ferring the problem to another area. For that rea-
son the last two alternatives were excluded. As far
as redesigning the ozone contact basin or adding
to it with a diffusion system prior to ozone diffu-
sers, when we considered the cost effectiveness of
that, it was just better to destruct. The problem with
Upper Thompson (I should not say it is a problem;
I guess it is a blessing in disguise) is that we are not at
the maximum capacity of the generators. We still have
excess capacity. So we have not had to reach the limits.
DR. NOACK: I have just a brief comment on the
ozone and the off-gases. I think they will require much
attention in days to come in view of the fact that
OS HA under their generic carcinogen policy in one of
the lists that are attached to the regulatory proposal
that classified ozone, I believe it is categorized as a
category two carcinogen. I have reviewed the list and 1
have not found chlorine in there.
MR. ELLNER: I would be curious to know what
you found the half life of ozone to be. Our experience
has shown it to be rather short, and I am curious why
you would be looking for a destruct requirement
for the off-gases.
MR. RAKNESS: Well, 1 should say another
alternative that we thought about was just venting it to
the atmosphere through a thirty or forty foot stack and
let the air take care of it. Right now it is vented on top
of the roof about twenty feet above the plant area with
operators exposed to it, and that has not been
sufficient to disperse the ozone in the atmosphere.
Thus, because of aesthetic reasons for a higher tower
we did not want to go any higher than that. Also,
because it is twenty feet in the air right now, we are not
sure that it would be adequately dispersed if we went
higher. I am sure that at some elevation you could let it
go and it would not affect the ground, but it still
would not be aesthetically pleasing to the area.
MR. ELLNER: If you do need a destruct system,
197
-------
21.
FIELD-SCALE EVALUATION OF WASTEWATER
DISINFECTION BY OZONE GENERATED FROM OXYGEN
Jain S. Jain,* Nicholas L. Presecan* and Michael Fitas**
^Engineering-Science, Ltd., Cleveland, Ohio
**Mahoning County Sanitation Commission, Youngstown, Ohio
INTRODUCTION
This paper pertains to the findings of a full-scale
U.S. Environmental Protection Agency (USEPA)
funded R&D study being conducted on the Meander
Ozonation System (MOS). The MOS provides the
final treatment step of disinfection in the 1.5 x 104
m3/d (4 mgd) regional Meander Advanced
Wastewater Treatment Facility (Meander AWTF).
The facility is owned by Mahoning County and is
located in Trumbull County, Ohio. The entire
Meander AWTF, including the ozonation system, was
designed by Engineering-Science, Ltd., Cleveland,
Ohio.
The effluent discharge standards for the Meander
AWTF require summer/winter nitrification,
phosphate removal, and postaeration as shown in
Table 1. To meet these standards, the following
treatment scheme was selected for the Meander
AWTF. The wastewater from interceptor is pumped to
comminution and aerated grit removal facilities.
Carbonaceous BOD is removed in a first-step pure
oxygen activated sludge reactor and clarifier, and
nitrogenous BOD is removed in a second-step pure
oxygen activated sludge reactor and clarifier. Lime is
mixed with second-step effluent, flocculated, and the
chemical/phosphorus sludge is separated in a solids
contact clarifier. Following recarbonation and dual
media filtration, the effluent is ozonated for
disinfection prior to discharge to Meander Creek. The
process flow diagram of the Meander AWTF, in-
cluding sludge handling facility, is shown in Figure 1.
The USEPA was interested in technical
developments which could be forthcoming from the
TABLE 1. OHIO ENVIRONMENTAL PROTECTION AGENCY EFFLUENT DISCHARGE STANDARDS
FOR MUNICIPAL WASTEWATERS DISCHARGING TO THE MAHONING RIVER AND ITS TRIBUTARIES
Present Effluent Standards
Critical
Constituent
BOD5
Suspended Solids
Free Ammonia
Summer
Winter
Fecal Conforms
Summer
Winter
Oils and Greases
Phosphorus
DO
Meander Report
Design Standard %
1970 Reduction
% Monthly
Reduction Average
90 95
95 95
— 85
- 85
— 99.99
— 99.99
- 90
90 90
Effluent Concentration
(mg/l)
Monthly
Average
10
12
2.5
2.5
200'
1 ,000*
10
1.0
Effluent
Maximum
15
18
5.0
5.0
400*
2,000*
20
1.5
Concentration 5 mg/l
' Counts per 100 ml.
198
-------
OZONE
Figure 1. Process Flow Diagram for Meander Advanced Wastewater Treatment Facility
MOS. This facility was considered ideal for a USEPA
Research and Demonstration Grant because it was to
be the first full-scale plant in the country using ozone
to disinfect the final effluent of a wastewater plant,
coupled with on-site oxygen generation unit and two-
step pure oxygen system. Moreover, the size of the
MOS was considered ideal to develop parameters
which could easily be scaled up for larger treatment
facilities. The USEPA interest in the project finally
resulted in a grant to the Mahoning County in June
1976 to conduct a two-year full-scale study on
the MOS. Mahoning County is presently conduct-
ing this study through its consultant, Engineering-
Science, Ltd., Cleveland, Ohio.
The Meander Uzone K&D study was to be
completed by 31 August 1978. However, this was not
achieved for various reasons resulting from the unique
features of the ozonation system. This is the.first
facility in the country where ozone is produced for
wastewater disinfection from oxygen recycled from
the contactor. At the time this paper was committed
for the symposium, it was expected that the ozonation
system would have been operating for several months
before the presentation and that there would be
sufficient data to report. For several reasons beyond
the control of Mahoning County, the ozonation
system did not become operational until only a few
weeks ago and, therefore, very limited data are
available. Under these circumstances, this paper will
report on the problems encountered in placing this
system into operation, and will report on the limited
data obtained as of 31 August 1978.
MEANDER OZONATION SYSTEM
Process Description
The Meander Ozonation System is designed to be
operated on air or oxygen recycled from the contactor.
The ozonators are capable of producing 180 kg/d (400
Ib/d) ozone providing an ozone dosage of 6 mg/1 for
the peak flow of 3 x 104 m\/d (8 mgd). There are two
trains of ozone generation units which can be operated
separately or in combinations. Both trains combined
can handle flows up to 3 x 104 m-'/d (8 mgd) at an
ozone dosage of 6 mg/1. However, one ozone
decomposer at a time is used since each decomposer is
199
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
designed to take the full load; thus, one decomposer is
at 100 percent standby. A common header between
compressor and dryer allows compressor No. 1 to
work with dryer No. 2. Electrically, however,
'compressor No. 1 operates in conjunction with dryer
No. 1. Similarly, compressor No. 2 operates in
conjunction with dryer No. 2. This system can be
operated in two modes.
The first mode is the open-loop system wherein air is
used as feed gas for ozone generation, as shown in
Figure 2. In this case off-gases from the contacting
tank are discharged directly to the atmosphere after
residual ozone is destroyed through the ozone
decomposing device.
The second mode of operation of the ozonation
system is the closed-loop system which utilizes
oxygen-enriched feed gas as shown in Figure 3. In this
case, the off-gases from the ozone contacting tank are
recycled and serve as feed to the ozone generator.
Make-up oxygen is necessary in this case to
compensate for the oxygen losses in the system.
To handle a broad range of flow variations in the
Meander AWTF, a contacting system consisting of
eight Positive Pressure Injectors (PPI), each capable
of handling 3.8 x 10' m'/d (1 mgd) flow, is also
provided.
System Components
In general, the Meander Ozonation System consists
of five main components: on-site ozone generation
facility, ozone contacting system, ozone decomposing
device, control system, and the indicating and
recording instruments. Particulars of these
.components are summarized in Table 2.
MEANDER OZONE R&D STUDY OBJECTIVES
Field Study
The initial field study work is directed towards
collecting background data on the plant flow and
processes, and to calibrate the instrumentation and
perform the mapping of the ozone generating
TABLE 2. MAIN COMPONENTS OF OZONATION SYSTEM AT MEANDER
No. of Units
Type
Capacity
Compressor
Dryer
Ozone Generator
Ozone Contacting System
Ozone Destruct Unit
Control System
Indicating & Recording
Instruments
Nash
Model L-3
Deltech
Model ESL-4
(Refrigerative/Dessicant Type)
Union Carbide
Model A-128
Lowther Plate
9,000 volts (peak)
2,000 Hertz
Positive Pressure
Injectors (PPI) supplied by
Grace/Union Carbide
Each 142 cfm at 15 psig
Outlet Pressure (0 psig Inlet and 60° F)
Each 140 scfm at 15 psig
Each 100 Ib/day (Air)
200 Ib/day (Oxygen)
Heat-catalyst unit supplied
by Mining Safety Appliances
Company
Supplied by Union Carbide
Most of them supplied as part
of the ozonation system by
Union Carbide
A total of 8 PPI heads each with
a nominal capacity of 1 mgd,
installed in a concrete tank having
two compartments, each containing
four PPI's. Each compartment has
inside dimensions: 7'-6"W x 38'-1"L
x 13'-10" deep with a water depth
of approx. 10'
Two units each designed to handle
up to 248 scfm of gas containing up
to 3000 ppm ozone
Ozone Gas Flow Control, Ozone
Production Control, Makeup Oxygen
Control, Regulation of Contactors
There are several indicating and
recording instruments but of
significant importance are: three
Dasibi Ozone Gas Analyzers, one
Delta Scientific Dissolved Ozone
Analyzer, and one LEL Monitor
200
-------
OZONE
t
COMPRESSOR
NO. 1
Nfl ?
DRYER
NO; i
NO 2
03 GENERATOR
NO. 1
NO 2
-oo-
OFF-GASES RECYCLE
Figure 2. Open-loop Ozonation System Using Air as Feed Gas
02 MAKE-UP ^
°3
DECOMPOSER
NO. 1
i
t—
LU
Csj
NO. 2
i
COMPRESSOR
NO. 1
1
1
DRYER
NO. 1
1
1
1
1
1
|
i
i 4
Q
LU
I —
z
o
r^i
O
0^ GENERATOR
NO. 1
t
h-
LU t—
1 LU
UJ U.
-^
-O^O-
*"""
0
w—
-OgO
z
-ogo-
OFF-GASES RECYCLE ,
LU
CSL
LU
^
U.
\
f
.Figure 3. Closed-loop Ozonation System Using Oxygen Enriched Feed Gas
201
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
equipment. After the collection of background
information, the efforts are directed towards achieving
the objectives of the Meander Ozone R&D Study
which are:
1. Determination of the power and oxygen
consumption.
2. Evaluation of the off-gas recycle system.
3. Determination of the optimum ozone dosage to
achieve a desired disinfection level.
4. Evaluation of the efficiency of the ozone
decomposing device.
5. Development of cost data.
Instrumentation
The Meander Ozonation System was initially
furnished with a variety of instrumentation for
automatic control of the process as well as for data
recording. In addition, several other instruments have
been incorporated into the ozonation system to
. accomplish the objectives of the Ozone R&D Study. A
process and instrumentation diagram (P&ID) is
presented in Figure 4 which shows the total ozonation
system which is being operated during Ozone R&D
Study. Those instruments specifically provided for the
Meander Ozone R&D Study are marked with an
asterisk. Major instruments purchased are: one gas
chromatograph (HP Model 5840A); one Dasibi ozone
analyzer; one BIF venturi tube with Hagen flow
indicating recorder; one sample pump (processor
Model 9-1311); and one dew point analyzer (Alnor
Model 7000L) for monitoring the moisture in the feed
gas to the ozone generator.
Additionally, there are a number of other
instruments/equipment such as pressure gauges,
thermometers, and watthour meters, which have also
been purchased for the Ozone R&D Study.
AEJIc
* ADDITIONAL INSTRUMENTATION
PROVIDED FOR THE. OZONE RID
STUDY
FILTER EFFLUENT
SAMPLE PUMP
Figure 4. Process and Instrumentation Diagram for Meander Ozone R&D Study
20?
-------
OZONE
PROBLEMS FACED IN CONDUCTING R&D
STUDY
The ozonation system was originally supplied by
W.R. Grace Company, who bid on a performance
specification. Prior to completion of the installation,
the ozone division was purchased from W.R. Grace by
the Union Carbide Company. After a review of the
system by its engineers, Union Carbide insisted upon
making several major modifications to the system (at
its expense) before they would assume full
responsibility. These modifications resulted in
delaying startup by almost two years.
Once startup was initiated, the usual number of
other problems were encountered with each
component. However, there were several major
problem areas which resulted in extended delays in
getting the system operational. The main problem
ireas of the ozonation system are: 1) Drying
Equipment; 2) Contacting Tank; 3) Certain Controls
and Instruments; and 4) Materials of Construction.
Drying Equipment
The drying equipment in the Meander Ozonation
System is a refrigerative/desiccant type and is
manufactured by Deltech Engineering Company. The
incoming compressed gas passes through a porous
stone prefilter to remove the finely divided water mist.
The prefiltered gas is first chilled to about 7°C (45°F),
at which temperature most of the contained water
vapor condenses. The cooled gas is then passed
through one of the two desiccant towers containing
activated alumina held at about 7°C (45°F). The gas
leaving the desiccant bed is about 7()C (45°F) with a
minimum dew point of -40°C. Each bed remains in
service for two hours. About 15 percent of this dry gas
is heated to 52°C (125°F) and is fed back into the other
desiccant bed in a reverse direction at atmospheric
pressure. This action regenerates the out-of-service
bed over a two-hour period by purging it of water it
had absorbed during its previous cycle. Every two
hours the towers alternate their functions and the
process continues uninterrupted. The dried gas then
passes through an after-filter. This is a sintered
alumina with Spm openings which is supposed to
trap particles greater than 40pm in size.
Besides some relatively minor problems, three
major problems in the drying equipment have been the
refrigeration system, the drying media and the after-
filter. The ozone generation equipment was installed
almost two years ago but the drying equipment always
gave problems. These problems were associated with
leaks in the refrigeration system, moisture in the
drying bed, and insufficiency of drying media. All of
these components were serviced and replaced when
necessary, but even so the desired dew point of -40°C
has not been achieved on a sustained basis. Another
problem area is the after-filter. The after-filter clogs
often, necessitating frequent cleaning. Old after-filter
elements have been replaced with new ones, but there
is a feeling that the present after-filter openings (5^,)
and length (41 cm) (16-inch) are too small and it would
require coarser and longer after-filter elements to cut
down the frequency of cleaning.
Ozone Contacting System
The ozone contacting system is comprised of eight
Positive Pressure Injectors each with a nominal
capacity of 3.8 x 10' m3/d (1 mgd). They are installed
in a concrete tank having two compartments, each
containing four PPI's. Each compartment is 2.3m (7
feet 6-inch) wide, 11.6m (38-feet 1-inch) longand 4.3m
(13-feet 10-inch) deep (inside dimensions). The water
depth in the tank is normally about 3m (10-feet). The
flow sequence in the contacting tank for the
liquid and the ozone bearing gas is as given below.
The 76 cm (30-inch) concrete pipe carrying filter
effluent to the ozone contacting tank is divided into
two 46cm (18-inch) cast iron pipes at the south end of
the contacting tank. Each of these pipes is reduced to
41cm (16-inch) and then further divided into four
25cm (10-inch) which are finally reduced to 15cm (6-
inch) just before entering PP1. Ozone-bearing gas
from the ozone generators is piped through a 10cm (4-
inch) PVDC pipe to a 20cm (8-inch) central header
which is connected to each individual PPI with a 5cm
(2-inch) pipe. In each PPI, the filter effluent and the
ozone bearing gas are mixed concurrently to create a
selected gas to liquid (G/ L) ratio which can be varied
to optimize performance. This two-phase mixture is
driven down a 25cm (10-inch) PVC pipe and the,
released about 3m (10-feet) below water surface. The
mixture then rises to the top of a 0.9m (36-inch) RCP
concentric riser column where the gas and liquid
separate. The liquid spills out of the columns onto a
platform and drains away, passing under a baffle and
over a weir before discharge. The baffle provides a gas
seal to contain the ozone bearing gas. The gas is
collected and piped back through a 20cm (8-inch) PVC
pipe to the ozone feed gas preparation equipment for
reuse in venting. Plan and section views of the ozone
contacting tank are shown in Figure 5. Figure 6
presents a schematic of the PPI functioning.
203
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
0)
c
o
a
s
o
w
0)
•a
i
c
o
o
0)
o
o
(A
0>
(0
13
C
(0
c
(0
in
o
204
-------
OZONE
With the arrangement of the PPI contactor as
shown in Figure 6, there was a possibility that during I
low filter effluent flow, or no flow condition, o/one
might travel upstream and back up to the filter
building where there was a potential for the equipment
and the personnel to be exposed to the ozone
atmosphere. Low flow and no flow conditions can
exist as the effluent of the filter is used to backwash the
filter cells. To alleviate this possibility, J-traps were
installed in each PPI contactor. A schematic of the J-
.trap installed is shown in Figure 7. Under this
arrangement, under low flow conditions, the ozone
gas will have to overcome higher pressure though the
J-trap (20cm) (8-inch), which will always be filled with
water, as compared to the pressure to be overcome in
the 25cm (10-inch) PVC pipe. The installation of the J-
trap has completely eliminated the potential of ozone
leaks in the filter building.
H20
A
"2°
Figure 6. Positive Pressure Contactor (Original) Meander
Ozonation System
Another problem encountered in the contacting
tank is leaks in the PVC pipe. The 20cm (8-inch) central
header, carrying the ozone bearing gas, is connected to
each individual PPI through a 5cm (2-inch) pipe. The
joints ot tnese individual PPI connections with .the
central header frequently leak and consequently have
to be frequently repaired. Initially, repair was done
using plastic rod and compressed hot air to apply
molten plastic around the joints. Later, plastic cement
was used. Neither of the two techniques has proved
completely successful. Perhaps a permanent solution
resides in providing tee connections at these joints.
Another problem encountered at the ozone
contacting tank is the occasional opening of the
pressure/vacuum relief valves installed at the
contacting tank. There is a 10cm (4-inch) pressure-
vacuum relief valve at the contacting tank designed for
1 m (39-inch) water column (WC) positive pressure and
76cm (30-inch) WC negative pressure. Although the
cause has not been fully identified, it seems that when
the gas to liquid ratio exceeds 0.25, the pressure relief
valve opens. On the other hand, if there is a greater
demand for oxygen at the make-up point than is being
supplied, the vacuum relief valve opens,
contaminating the system. These problems would
indicate that the ozonation system at Meander can
only be operated within a very limited range of
parameters and cannot be subjected to a wide range of
conditions without upsetting the operation of the
system.
The ozone contacting tank has a total of four
manholes and eight hatches for the removal of PPI
heads. Gaskets provided with the original installation
started leaking and had to be temporarily replaced
with BUNA-N rubber. These were finally replaced
with viton gaskets, which have proved extremely
reliable.
Figure 7. Positive Pressure Contactor (Modified)
Meander Ozonation System
205
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
Controls and Instruments
Oxygen Recycle System
A schematic of the basic oxygen recycle loop is
shown in Figure 8. There are three main sources of
oxygen loss in the system which must be replenished in
order to maintain a constant concentration of oxygen
in the feed gas to the ozone generator: 1) Oxygen
consumed in producing ozone; 2) Oxygen lost as
dissolved oxygen in the final effluent; and 3) Oxygen
lost in the nitrogen purge line. From the limited
operating experience of the Meander Ozonation
System, it appears that sufficient oxygen cannot be
supplied for all operating conditions. Also, it is
difficult to measure the amount of gas purged through
the nitrogen vent line for all operating conditions.
Ozone Gas Flow and Production Control
Control of the amount of gas and the production of
ozone is necessary to apply a predetermined ozone
dosage to disinfect the final effluent. In the MOS, a
signal from the final effluent flow transmitter can be
used to control the amount of gas, concentration of
ozone in the gas, amount of nitrogen purge, and the
number of the PPI contactors in operation. Thus, the
fluctuating demand of the varying plant flow is met by
regulating these four parameters, thereby maintaining
Figure 8. Basic Oxygen Recycle Loop Meander
Ozonation System
a predetermined dosage of ozone and predetermined
concentration of oxygen in the feed gas to the ozone
generators; and to open required number of PPI
contactors for an optimum treatment. Such a control
schematic is depicted in Figure 9. In this scheme, valve
46 (Figure 9) which controls the gas flow to the ozone
contactor caused several problems. Initially, the valve
supplied was too big for the system. It took several
months for Union Carbide to realize that in fact this
was the problem. An almost equal amount of time was
taken to replace this valve by a smaller one. Even after
replacement, the valve did not function well. Although
the cause of malfunctioning of V46 could not be
determined, it was made operational after a long
period of trial and error. In the same control loop, the
rotameter supplied to measure the gas flow through
the nitrogen vent controlled by V21 (Figure 9) does not
seem to be of adequate capacity. This poses a problem
in that there is no measurement of the amount of vent
gases when the capacity of the rotameter is exceeded
and, thus, a mass balance of the gases in and out of the
contactor cannot be made.
-o o-
-o o-
-o o-
-o o-
Figure 9. Control System Schematic Meander
Ozonation System
Dasibi Ozone Meters:
The Meander Ozonation System has three Dasibi
ozone meters. This instrument measures gaseous
ozone by means of ultraviolet (UV) light absorption by
the ozone molecule. Two of them are for measuring
high ozone concentration in the exit gases from the
ozone generator and in the off-gases from the ozone
contacting tank. The third meter is for monitoring the
ambient ozone concentration inside the ozone
building. It appears that the electronics of these meters
is good and reliable but the optics is not. The cell in the
meter often fogs up with a yellowish film giving false
reading of ozone concentration in the process gas. This
film is very difficult to remove. Alcohol is not effective
in this respect; however, we have successfully been able
to fully remove it with sulfuric acid. The frequency
with which these cells have to be cleaned is such that it
defeats the purpose of spending money on the
purchase of these expensive monitoring devices.
Although unlikely, there may be some special
contaminant in the gases at Meander which is causing
this problem.
Delta Scientific Dissolved Ozone Meter
A dissolved ozone meter supplied by Delta
Scientific is used at the MOS to monitor the dissolved
ozone in the final effluent. This meter utilizes an
amperometric membrane electrode for aqueous ozone
determination. The Delta Scientific Ozone meter was
206
-------
OZONE
delivered almost a year ago and installed in the final
effluent channel. Since the system could not be made
operational until recently, there was no opportunity to
test the performance of the meter. However, when the
ozonated effluent became available, the meter did not
perform well. The meter has been sent for repairs. It is
difficult to say at this time whether the problem lies in
the design of the meter or its performance was
adversely affected due to its long storage.
LEL Monitor
Although other safety features exist, monitoring of
the hydrocarbons based on their lower explosion limit
(LEL) is provided in the design of MOS to prevent
buildup of hydrocarbons in the contacting tank and in
the closed recycle loop. Should the hydrocarbon level
reach a predetermined set point, the monitoring
system provided will alarm, discontinue introduction
of gas into PPl's, and turn on a purge air system. This
monitoring system has been functional off and on for
different reasons. Presently, the LEL monitor is not
operational. The cause has not been determined.
Union Carbide, supplier of the ozonation system, is
working on it.
MATERIALS OF CONSTRUCTION
Initially, PVC pipe was used for both oxygen and
ozone application in the Meander Ozonation System.
In the present design, the use of PVC pipe is restricted.
PVC can become brittle with age, and use of PVC in
oxygen application is potentially hazardous in that the
PVC ignition point is about 430° to 480"C (800° to
900°F). Although there are some exceptions, the use of
construction materials in the MOS has been according
to the following guidelines:
I. Copper or stainless steel is used in the application
of wet oxygen.
2. PVC or stainless steel is used in the application of
ozone and oxygen.
Use of PVC for ozone and oxygen is restricted to
underground applications. However, some exceptions
had to be made. For economic considerations, the
main header and the individual pipes feeding ozone to
PPl's are made of PVC and are above ground.
However, in these pipes, leaks at the joints have been a
Constant problem.
• Initially, teflon tape was used as a lubricant/sealer
in the joints of stainless steel and steel pipes. However,
it was found to be inadequate for sealing the joints.
Teflon tape with RTV cement gave an effective seal
and it has been working satisfactorily for the last year
without any problems. The original gaskets (material
not known) in manholes and hatches were found very
unsatisfactory; they could not stop leaks in the tank.
They were replaced by gaskets of BUNA-N rubber for
temporary use. Now the gaskets used are of viton
(fluoroelastomer) and have been performing
satisfactorily for the last year.
Although hypalon (chlorosulfonated polyethylene)
is recommended for ozone applications, the hypalon
diaphragm in an ozone meter pump was reported to
have been attacked by the ozone. The manufacturer
considered replacing it with a viton diaphragm.
DATA OBTAINED
As previously discussed, the ozonation system at
Meander AWTF has been operational only a short
time and, therefore, only one week's data are available
for reporting here. Data pertaining to ozone
disinfection efficiency are given in Table 3. Although it
is difficult to draw a definite conclusion from such
little data, it appears that ozone dosage of 4 mg/1
should be sufficient to achieve a disinfection level
below 200 fecal coliform/100 ml. This dosage is 67
percent of the design dosage of 6 mg/1. All the
disinfection data were obtained using a gas to liquid
ratio in the range of 0.18 to 0.27. The plant flow during
this study varied from 6400 to 8300 mVd (1.7 to 2.2
mgd).
Data on the quality of the influent to and effluent
from the ozonation tank for the period of the
disinfection study are presented in Table 4. As can be
seen from the table, the influent to the contacting tank
was of very high quality; BOD ranged between 0.6 and
1.8 mg/1, COD ranged between 5.1 and 28.6 mg/1, and
TSS ranged between 1.2 and 3.4, except in one case
where the TSS concentration was 12.8 mg/1.
There is an indication of a slight increase in BOD of
the ozonated effluent. Concentrations of COD in the
ozonated effluent remained mostly unchanged except
in the last iwo cases where there was an abrupt
increase. These results cannot be explained and this
increase can only be attributed to analytical errors. In
. the case of TSS, there was an increase in five out of
seven cases. This might explain in part the reason for
increased BOD in the ozonated effluent. There was an
increase in dissolved oxygen in the ozonated effluent,
which was to be expected, as the ozone was produced
with gas feed containing oxygen in the range of 45.3 to
61.2 percent. Ozonation, at least for the ozone dosage
applied in this study, does not seem to affect the pH of
the effluent as can be seen from the data presented in
Table 4.
207
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
Data pertaining to power and oxygen consumption
are presented in Table 5. These data do not show any
trend. However, average oxygen consumption was
8.85 kg per kg of ozone produced and power
consumption averaged 18.1 KWH per kg (8.24 KWH
per pound) of ozone produced. These values
correspond to an average oxygen concentration of
51.94 percent in the feed gas to the ozone generator.
They appear to be high and more data will be required
to develop reliable values.
CONCLUSIONS
Based upon a short operating experience with the
Meander Ozonation System and from analysis of the
very limited data available, the following conclusions
might be suggested:
1. The Meander Ozonation System supplied by
Union Carbide (previously .W.R. Grace), even
after making extensive modifications and
changes, has not performed reliably to date.
TABLE 3. OZONE DISINFECTION EFFICIENCY, MEANDER OZONE R&D STUDY
Influent
Sample
Date
August 1978
23
24
25
28
29
30
31
Flow
mgd
1.7
1.7
1.8
2.2
2.1
2.0
2.0
Ozone
dosage
mg/l
3.38
3.12
3.13
4.15
4.79
6.57
8.38
(a)
G/L
0.27
0.24
0.24
0.19
0.20
0.21
0.18
Total
Conform
per 100 ml
480,000
700,000
800,000
140,000
45,000
54,000
>1 ,000
Fecal
Conform
per 100 ml
20,000
220,000
80,000
4,600
3,000
2,300
< 1,000
Effluent
Total
Conform
per 100 ml
3,400
3,700
10
760
150
490
1,000
Fecal
Conform
per 100 ml
260
600
6
8
25
20
15
(a) Gas to Liquid Ratio.
TABLE 4. WASTEWATER QUALITY, MEANDER OZONE R&D STUDY
Sample
Date
August
1978
23
24
25
28
29
30
31
Inf.
0.6
1.4
1.1
1.7
1.2
1.8
0.9
BOD
mg/l
Eff.
3.0
3.0
1.3
2.9
1.3
2.4
—
Inf.
28.6
28.3
56.6
5.1
20.2
20.4
20.0
COD
mg/l
Eff.
28.6
18.9
56.7
5.1
20.2
91.8
60.0
Inf.
2.4
3.4
1.2
12.8
1.4
2.8
2.6
TSS
mg/l
Eff.
5.4
5.4
3.4
18.2
1.6
1.6
0.8
Inf.
7.08
7.59
7.44
8.24
7.88
7.88
10.19
pH«
unit*
Eff.
7.4
7.70
7.40
8.04
7.99
8.10
10.16
Inf.
0.6
0.6
7.3
<0.1
<0.1
<0.1
0.3
NH3
mg/l
Eff.
0.6
0.7
6.5
0.1
0.1
0.1
0.4
Inf.
8.0
7.1
7.1
7.7
8.2
8.2
7.8
DO2
mg/l
Eff.
11.4
14.0
11.8
11.2
11.3
13.5
10.2
Average
TABLE 5. OZONE GENERATING PARAMETERS, MEANDER OZONE R&D STUDY
Sample Date
August 1978
23
24
25
28
29
30
31
% Oxygen
Feed Gas
50.4
54.9
52.0
47.2
45.3
52.6
61.2
Dew Point
Feed Gas,
0 C
—
—
—
0
-59
-59
-50
% Ozone
Produced
0.87
1.01
1.22
1.84
1.97
2.59
3.35
kg Oxygen/kg Ozone
—
12.1
7.6
4.7
14.0
9.0
5.7
KWH/kg (lt>) Ozone
25.5 (11.6)
22.4 (10.2)
11.9 ( 5.4)
17.0 ( 7.7)
16.1 ( 7.3)
11.9 ( 5.4)
22.1 (10.0)
51.9
8.85
18.1 ( 8.2)
208
-------
OZONE
2. Except for some surface discoloration, no visible
material degradation has been noticed so far in
the PVC pipe used in the Meander Ozonation
System. However, use of PVC in ozone
applications should be carefully investigated for
each individual case. Stainless steel, although
expensive, would be a more durable and reliable
material in such applications.
3. Viton (fluoroelastomer) is highly recommended
for gaskets and diaphragms in ozone'
applications. Hypalon (chlorosulfonated
polyethylene) appears amenable to ozone attack.
4. Operation of an ozonation system with recycled
oxygen feed is a fairly complex system for the
disinfection of treated wastewater effluents.
5. Average consumption of oxygen and power per
kg of ozone produced on 52 percent oxygen feed
are 19.47 kg (8.85 Ib) and 18.13 KWH,
respectively. These values are almost twice the
values normally expected from such systems.
6. Initial indications are that disinfection to meet the
200 fecal coliform/100 ml discharge can be met
with an ozone dosage of 4 mg/1.
7. In its present condition, the oxygen concentration
in the recycle feed gases to the ozone generator
cannot be achieved more than 60 percent.
209
-------
22.
ROUNDTABLE DISCUSSION OF OZONE DISINFECTION
Participants in Ozone Roundtable
1. Mark C. Meckes
U.S. EPA-Cincinnati
2. Albert D. Venosa
U.S. EPA-Cincinnati
3. Edward J. Opatken
U.S. EPA-Cincinnati
4. K. Rakness
M&I Consulting Engineers
5. Rip Rice
Jacobs Engineering
6. Jain S. Jain
Engineering Science, Inc.
MR. VENOSA: Rip Rice has asked that he be
given a couple minutes to make a few points. So, we
will start off with Rip.
DR. RICE: Thank you, Al. From the discussions
we were hearing yesterday afternoon, I thought it
would be smart to make some general points which we
found in our survey of drinking water treatment plants
using ozonation systems and using ClO2systems.
Most of the comments that I am going to make are
applicable directly to ozonation but some of them are
also applicable to C1O2 systems, too.
There are something like 1,050 drinking water
treatment plants throughout the world using
ozonation systems and these have been in constant use
since 1906. One of these plants in Nice operated
actually 60 years using the same ozonation hardware
prior to upgrading it.
More than 500 of these plants are in France and
most of the rest are in Europe. The USA has only five,
Canada has 19 or 20. The important point is that in
sewage treatment with ozonation the U.S. is
pioneering this application. The Europeans know all
about how to generate ozone, dry air, contact it for
drinking water, but.they know very little about its use
in sewage treatment. So when you hear the horror
stories, keep in mind that we are on the cutting edge
here and, number two, that Union Carbide is
pioneering a brand new type of ozonator and a brand
new type of contactor.
I have a list of some of the plants in the USA that
are involved with ozone (slide). There are six or seven
of them which are operational. That means they are
either operating or they are in the startup phases.
The Indiantown, Florida plant has been operating
this November for three years, constantly, without
having to maintain either the ozonator or the
contacting system, and this is the longest operational
sewage disinfection plant using ozone in the world. It
is a very small plant. The asterisk over at Chino Basin
means that this is the one plant in this group that is not
using ozone for disinfection. It is using it for
suspended solids removal.
The next overhead shows those plants which are in
construction using ozone for disinfection. All the
other plants other than the one asterisked are using
ozone for disinfection, and here we have six plants, all
of which are being built, about to be started up. The
next slide shows some of the plants which are under
design using ozone for disinfection. There are 16 here.
Six of them have been named. It is not right to name
the other 10 yet because there is, as you know
competition among the manufacturers.
The asterisk against the Cleveland Westerly Plant
means that they are not using ozone for disinfection
here but for pretreatment of water prior to passage
through dual media filters, one medium of which is-
activated carbon which is now biologically active.
One of the things that I think is important is that in
the European drinking water treatment plants the
normal dosage of ozone for any purpose averages
between 1 and 4 mg/1. During the survey which we
made we did cost estimates of several of the larger
European drinking water treatment plants and we
found that the cost for ozonation systems, cost for air
preparation, for generation, for contacting, for
instrumentation, for control, for the analyses, for the
operation, the maintenance, all factors, amortization
and so forth, costs ran between 1.75$ per thousand
210
-------
OZONE
gallons to 4c per thousand gallons.
We found the major variable factors in the cost for
ozonation systems were twofold. These are variables
now. The first one was the housing that you put the
ozonation system in. If you go out and. build a brand
new plant and build a separate housing for the
ozonation system as opposed to taking an existing
plant and jamming the ozonation system into it, that
gave us a broad range in the cents per thousand
gallons.
The other one is the cost of power which over there
varies from 2 to 5c per kilowatt-hour.
Another point that I think is worth making is that in
drinking water treatment practice multiple uses of
ozonation are being made. In other words, at the end
of the process if they are using ozone for disinfection
or bio-inactivation, many times they will take the off-
gases and put them in the front end of the process with
the chemical addition to aid in bringing down
suspended solids. This gives them a much clearer, less
turbid solution to take onto the plant.
In terms of sewage treatment disinfection, it is my
understanding that the recommendation to this point
is that secondary effluents should be filtered prior to
ozonation. In that case, I would think that it would be
wise to look at taking contactor off-gases and not
throwing them up the stack but putting them up in
front of the filter step. This should do several things.
One is to help you knock down suspended solids. Also,
partially oxidize some of that COD into BOD and if
you can adjust the flow rates through the sand filter you
will have some biological degradation and
nitrification. This should give you a better quality
effluent if not a somewhat more constant effluent, but
that is argumentative and I am not saying that there is
any data to show that to be the case.
Finally, in Berlin, .Germany I think there is a new
approach being taken in sewage treatment. The pilot
plant is taking primary sewage effluent, ozonizing
it with 1 to 2 ppm dosage of ozone, not for disinfection
but merely to partially oxidize BOD and COD, make
it more biodegradable, saturate the sewage with
oxygen, and then that effluent goes into the ground,
into a sand filtration system in the ground and then on
out into the rivers. So, effectively what they are trying
to do is "secondary treatment" in the sand prior to the
water going out into the river. That is kind 'of
interesting and so I wanted to make those general
comments.
MR. VENOSA: Thank you. Rip. Do we have any
comments from the audience now? Okay, Harvey.
DR. ROSEN: Well, first I would like to apologize
to the other manufacturers when I said that you should
have called me for information. I meant you should
have called us. We are not the only manufacturers.
There are many of them here and we all provide a lot of
information and do a lot of hard work to provide
information.
With respect to the particular economic analysis
done by Ed Opatken. I called the home office and I got
authorization to sell what I understand to be the
$280,000 system under Table 7 ... that is, equipment
cost for ozone generation, for compression, and
drying .... for $65,000 to anybody in the room. That is
the current market price. It is one unit. It is not 28
units. It is skidded. It goes in with a forklift truck. It is
about 5 feet by 12 feet, takes almost no room as far as
the building space involved; I do not know what to say
beyond that.
Rather than picking on any individual parts of the
analysis, as Ed indicated when you do this analysis,
that capital cost for the ozone equipment is the
overriding factor which comes down to affecting the
bottom line, and what that does, if you change nothing
else except that capital cost, (and, by the way, we will
warranty better operating costs than are shown in this
analysis) comes down to is: look at the capital
investment under diffuser 15. We chose this because
that was the point of 90% utilization. That is another
thing which manufacturers will warranty, at least we
will if we are involved in a design.
Ozone utilization is not a black art. It is mass
transfer. It is gases dissolved in liquids and things like
that and we think we understand it enough to put up or
shut up in terms of penalties and warranties with
respect to that, as well as all the performance that you
would expect with respect to how many pounds a
system will make, how much power that system draws,
and those sorts of things. We even warranty reliability
up to a point more than just the one year normal
manufacturer's warranty of parts and labor, because
they tend to be new things and people are concerned
about the problems with new equipment.
To corroborate that number, I do not know if the
gentleman from the Washington Suburban Sanitary
Commission was involved in our recent negotiations
to sell him a system that is similar to what I am talking
about. It is something that comes off the shelf.. It is
called an MA-75. With a device to decompose ozone
and with diffusers and some other things on a rental
basis, I think the price would be $82,000, if after it was
all rented out over a year or something like that the
Commission decided to keep the equipment.
So, these are the kind of prices there are available on
211
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
the marketplace. There are, of course, as Dr. Rice
showed, a lot of plants being bid and that is where to
go out and find out how much these things cost. There
is competition out there, and you will find on a dollar
per pound of ozone basis that you get a pretty good
price now compared to even a couple of years ago,
because there is some volume in sales and
manufacturers can afford to reduce the cost of
manufacture and sell systems at a cheaper price.
The particular system that I am quoting, the MA-
75, is a 75 rather than a 62 pound a day system.
They come pretty much standard. So, it is even a
little more ozone that we are talking about that
was required for this particular application here.
With respect to the general technology development
and innovative technology, I do not think, as Rip Rice
mentioned, that it is fair to judge what the future costs
will be necessarily on the first one or two start-ups. We
have a plant that was listed up there at 35 mgd at
Springfield, Missouri, which has started up, and had
minor start-up problems relative to most wastewater
treatment plants or any plant that size which includes
over 3,000 pounds a day of ozone. 1 do not want this to
get into a commercial, but 1 do want to put things in
perspective. That is the third plant now as opposed to
the other two that we have been involved in, and the
performance characteristics are embarrassing relative
to the warranties because of the improvements made
between the time we sold the equipment and the time it
was installed and working. These are some of the
things 1 was trying to say yesterday.
There is over 94% utilization of the ozone and these
are warranties or acceptance tests done by Birdsall and
Jenkins, and all the standard methods that are done by
the consulting engineer when he accepts the
equipment, and tests that system.
So, I think we have to take this whole thing,
regardless about whether it is ozone or UV or any of
the new technologies and be a little careful about . . .
and this was mentioned by one of the UV gentlemen
yesterday .... before we start to condemn them too
early. Thank you.
MR. BOB FLEISCHER: I did not have to call
the home office. I did some evaluation in my own
head last night and my evaluation is based on
actual bids. In fact, the numbers I am going to use
are actually .higher than actual bids that took place
less than six months ago.
For instance, there were two 23 kilogram units with
.several metering devices that 1 would say retailed for
approximately $4,000 apiece. I think there were three
on this job, and two ozonators. In other words,
redundancy .... went for $69,000 on the marketplace.
DR. RICE: 23 kilograms per hour or per day?
MR. FLEISCHER: 23 kilograms per day, excuse
me. And tms 100 is an odor control job on Long
Island. I can discuss it with you turtner if you would
iiKe. it was not my company that got the job.
It was | another manufacturer, as a matter of fact.
However, based on Table 6, I went backjust figured
on what capital costs would be, power would be, etc.,
for that size plant. I was extremely conservative. For
instance, I have enough money in here to build a house
with a refrigerator and a freezer and stove around my
ozone generator, and total cost with full redundancy is
$104,000.
MR. OPATKEN: Installed.
MR. FLEISCHER:Installed, piping . . . .
MR. OPATKEN:' Housed and everything ....
MR. FLEISCHER: Housed, diffusers, contactor,
labor, everything. . . .which at first cut. . . .and this is
just the first cut in my head, brought the cost down
from 361 to 209, which is a 42% reduction, and I did
not even breathe hard.
MR. OPATKEN: Good. You also instrumented it
with Dasibi's and so forth?
MR. FLEISCHER: That is correct. There are over
$ 10,000 worth of meters that I threw in. I would like to
comment that I do not really want for anybody to put
the blame anyplace. Nobody ever called me and I am
the man who receives these type of phone calls at PCI.
There is no way that I would ever recommend anybody
scaling up from anybody else's ozone generator,
including my own. On top of that this job looks like,
from the regeneration numbers, oxygen feed, which is
absolutely absurd for 65 pounds a day of ozone. That
is not real world.
We have several plants that are going on for as little
as 3 pounds a day ozone disinfection and in which the
engineers are finding it cost effective. There are two
projects in West Chester County and one is running, it
is called Hunter Highlands, I will take you up there, 3
pounds a day ozone for a condominium complex.
Another one is Oak Ridge which is in West Chester
County. Saratoga Waste Water Treatment Plant is
using ozone at 37 pounds a day. That is even smaller,
yet the engineer has found it cost effective.
I could go on and on, but I would rather not. I would
like to make one other comment and this goes back to
the previous study.
We at PCI feel that the type of diffuser used for the, I
think it was 85% mass transfer, really is not the best
diffuser available today and that higher mass transfer
values can be gotten with a cylindrical type diffuser. If
212
-------
OZONE
that was used, mass transfer would be better and
perhaps even the amount of ozone utili/ed would be
greater versus the amount of ozone applied.
MR. OPATKEN: Where is your data on your
diffuser analysis?
MR. FLEISCHER: If you look at Bollyky's mass
transfer paper. . . .
MR. OPATKEN: All right, I talked to Dr. Bollyky
at the last EPA meeting.
MR. FLEISCHER: Well if you look deeply into his
paper, you will find it.
DR. JOHNSON:. Isn't this a kind of an unending
argument? That is, naturally, as the technology
improves and if you compare government specs, with
all due respect to the government well represented at
the front, we all recognize that when you build
something according to government specs, I mean it is
a different type of deal than if you do it yourself.
So, cost is cost and I think this is an unending
argument. I would like to go back to a more
fundamental question.
Let me first preface this by stating, I think that
ozone is probably the best method for killing viruses
and other very difficult to disinfect microorganisms
such as cysts, of all of the various methods of chemical
disinfection available. At the same time, ozone is very
reactive. It is a very strong oxidant. We all recognize
that. It is rather selective in its reaction with organic
compounds, primarily reacting with unsaturated
compounds as we will hear more about later today, I
think, from Dr. Jolley and Dr. Cumming.
But the basic problem with ozone is control, in my
estimation. We have a very strong oxidant, that is very
reactive, and we are putting it into sewage, which has a
lot of compounds in it that the ozone can react with,
which will take and pull the ozone off to reactions
other than what we would like it to do, which is kill
microorganisms.
The side reactions may actually consume all of the
ozone that we have put in. In fact, in many cases it is
very likely to consume all the ozone. Even if we had
done as the manufacturers would like for us to do, (run
pilot plant studies, determine on site with the actual
waste what kind of demands we have for ozone and
what kind of dosages we have to put in with that
particular contactor and that particular waste and that
site), who is to prevent manufacturer A from dumping
a little phenol that particular day into the waste stream
and suddenly we do not have any disinfection going
because all of the ozone we added in our pilot studies
without the phenol coming down the pipe is suddenly
being consumed by what we got that morning. That is
a real problem for any kind of wastewaterdisinfection
facility.
This is the question then. How do we control ozon-
ation processes so that we can be sure that, on a
continuing basis where we are controlling with dosage
alone, we have a continuing disinfection process going
on? It is a very difficult question for ozone, and I think
the question that we need to focus on in controlling the
ozone disinfection process on a continuing basis so
that we have assurance on a continuing basis that we
have disinfection and not some side reaction
occurring.
MR. OPATKEN: I would like to make one
statement. During my paper I said that what I used
was the pilot plant unit and the production required
was based upon that pilot plant unit being a modular
unit and scaled up in terms of units. I obtained those
units through the mapping curve and I said earlier that
I was not concerned with the optimization of ozone
disinfection costs. I was concerned with ozone
contactor evaluation and economic evaluation. The
differential costs could be adjusted accordingly if you
have an ozone generator that can produce this at a
lower kilowatt-hour per kilogram. That is an
advantage and you can take it right into these costs
and apply it. If you can get more production out
of it, fine. You can then adjust your fixed
capital investment, but for me to be sitting there
at the pilot plant, I have to estimate a fixed
capital investment. I did it on the best possible
method I thought fair. That is the data that I have.
Now, as I said when I started my paper, I think the
consultants are smart enough to be able to specify, and
we have tried to help them. Al gave a paper saying that
so much of this is necessary, at so much concentration
to get this level of disinfection.
I have not heard the ozone manufacturers come out
with that type of data. They always come out with the
fact that we will give you something. Now, the
consultants are also smart enough to take that
mapping curve and specify where they want to be.
They then have the knowledge to them to start up that
plant.
One of the difficulties we have in ozone is that a lot
of this information is with the manufacturer. Then we
get to the field and this manufacturer has that
information. The engineer in the field is blind. He
doesn't know where he's supposed to be.
We were given an ozone system for our pilot plant
without a dew point meter. We were ignorant at the
time, too. Now, I would not recommend any facility
213
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
without a dew point meter. We were given a dryer that
supposedly had a fixed orifice. We could not dry our
air. We had to readjust our regeneration. We had to
add a rotameter. So we had to apply control methods.
Now, I have seen plants out in the field and they also
were lacking in some of these fine operational details,
and so when you say they are skid-mounted and so
forth, that is all well and good, but remember that I
laid everything on the line when I calculated out my
number of units.
Now, these units are based upon that curve. I have
asked one of my fellows to duplicate that'curve so that
each of you can have a copy of it. All I am saying is, let
the consultants do the analyzing instead of the
manufacturers hiding this information among
themselves. I think you ought to have an open
competition here.
DR. ROSEN: 1 think if you were at an ozone bid at
any time you will find that there is a very stiff
competition.
MR. OPATKEN: And whatever you are offering is
hidden.
DR. ROSEN: No, it is not hidden as a matter of
fact. It is very nice to offer that curve, but that curve is
for your unit that you are using in that laboratory and
that is part of the problem with putting out that
information. It will become a gospel. It will become a
standard for all ozone generators, and although you
say no, and I am telling you that is what will happen
and it is happening in the marketplace. I guess in
somebody's file this is an ozone generator. Generically
it does not make any difference what type is used. We
design and provide information relative to a particular
problem at a particular plant.
Now, in our brochures we will be publishing for our
equipment on a relative basis, some of the kinds of
curves you are talking about, because again we are.
talking about covering a large size range of equipment
and all performing the same. It has to be on a
relative basis as opposed to an absolute basis, but
as the equipment gets harder and we learn more
about it, we do publish that information in bro-
chures. You know, it is not even that secret.
Another point is that if you change that price in that
particular diffuser (15 milligrams per liter ozone
concentration) and just make the adjustments on the
basis of that capital price, you come down to 6.3c per
thousand gallons rather than 13. So, when I said a
factor of 2, I was not too far off.
MR. RAKNESS: I would like to speak for a minute
on the gentleman's question about operation of a plant
and what do you do to continue to get disinfection.
At the Upper Thompson plant, since we are sure we
need disinfection, what the operators do if there is any
doubt at all, is to crank the ozone production up
probably greater than what is optimum from a cost
standpoint, and get disinfection levels much less than
what are required. Now, that does not take into
account any industrial dumping and things like that,
but. . . .
DR. JOHNSON: What do you do about the big
demand that comes down the pipe you do not know
about?
MR. RAKNESS: Well, the Upper Thompson
facility does not have a lot of manufacturing plants
that are hooked on the industrial sources. It is a
residential community with commercial tourist-
related type wastewater. jt does not have industries
that would dump things like phenol.
DR. JOHNSON: Unfortunately, most of our
facilities are not that well off.
MR. RAKNESS:Well, I just wanted to make the
comment that we do probably dose at a higher level
than what is needed to just achieve our disinfection
level.
DR. JOHNSON: Of course one answer is to set
your dosage up here when most of your needs are
down here, but that is not very cost-effective, is it? You
hope that your excursions of demand do not ever
reach your dosage level.
DR. ROSEN: Oh, yes. 1 do want to try and answer
your question. Part of the problem with what Ed was
saying is that you can instrument and control and do
all kinds of things with an ozone system or any system
for that matter to any degree to which you are willing
to pay for it, and there are a lot of sophisticated. . . .
DR. JOHNSON: Wait a minute. There are
measurements to make. How are you going to
instrument it and control. . . .
DR. ROSEN: You know how. You are developing
a technique yourself. There are a lot of other people...
DR. JOHNSON: That is why I asked the question.
DR. ROSEN: I know, but there are a lot of
other people also developing techniques. One of the
techniques that we are looking at and Al has told me he
will be looking at is, if you are changing ozone
demand, you are changing the amount of ozone
coming out in the gas phase because more of it is being
reacted or less of it is being reacted as compared to that
which you are applying, and it is easy to measure
ozone in the gas phase.
The question is, can we take that delta, how sensitive
it is, and can we relate that back to demand and
disinfection? That we do not yet know. Now, that
214
-------
OZONE
method of control is used in European plants to
control a constant dissolved amount and has been
used successfully for a number of years.
DR. JOHNSON: Okay, but in that method of
control where we are working with the difference
between the ozone in the gas phase coming in and the
ozone in the gas phase coming out, that only tells you
how much ozone is going in the solution. It does not
tell you whether it goes to the microorganism or
whether it goes to the demand, and if it all goes to
demand, then you are lost as far as the microorganism
is concerned. So, what are you going to do?
DR. ROSEN: Well, isn't that true of chlorine or
anything else?
DR. JOHNSON: No, with chlorine we measure
residuals and we control on the basis of residuals.
When the residual goes down, we up the dosage.
DR. ROSEN: Someone said the other day, that
from a practical point of view, that is more a crutch
than a practical solution.
MR. WHITE: That is general practice.
DR. ROSEN:! know it is standard practice,
George. There have been a number of reports in the
literature where people have studied that standard
practice with the chlorine residual that the state
recommended and Don, I think, will agree that
they have not achieved the disinfection standard.
MR. WHITE: That is how we achieved disinfection
standards in California and I wanted to say that
nobody has defined disinfection. If 200 fecal coli-
forms per 100 milliliters is the standard, you
could do better with a hammer. That is not
disinfection.
Disinfection is a graded thing. There is a certain
water requirement. There is a confined estuary
requirement. There is a negative estuary requirement.
There is fish breeding place requirement. There is
water sports requirement. They are all different. They
imply certain grades of water quality coming out of the
treatment plant and the disinfection is to a
certain number of total coliforms.
DR. JOHNSON: I am a native Californian and I
would like to make one point and that is Californians
are crazy.
MR. OPATKEN: We have a grant with the city of
Marlborough to look at ozone for disinfecting at what
we consider high levels of disinfection. So, that is
down the road.
MR. WHITE: Right, but the thing is when we talk
about disinfection, we have to define what we mean.
That is all I am saving.
DR. ROSEN: As I heard someone once express it,
and I do not think it is a bad way of thinking of it
necessarily, wastewater treatment is the preparation of
wastewater to be disinfected. So. the up-stream
processes and the quality are important, regardless of
disinfectant, because they all have other demands
(UV. chlorine, ozone). If you do not have a relatively
good water to disinfect, whatever your definition of
disinfection is. it creates a big problem and that is part
of the overall problem that I think we are discussing
here.
DR. RICE: Don, I would like to add a little and
maybe clarify. There are actually two control items
here. One is the ozonation system itself which can be
controlled very well to put out any amount of ozone in
any concentration at any rate you want. What you
have asked really is the major critical question, and
this is one of the things I had in mind when I said we
are pioneering the use of ozonation in sewage
treatment.
In the ideal ultimate, you should never think of
pacing disinfection by ozone dosage, ozone demand,
because that implies you know what the demanding
components arc. In the ultimate end, in my opinion,
the ideal way is to have a real quick bacteriological
test, which we do not have.
DR. JOHNSON: Well, you cannot grow bacteria
instantaneously.
DR. RICE:That is correct. Therefore, in my
opinion the best approach is the last comment that
Harvey made which is to make sure that the treatment
process which you are going, to use is good enough to
handle the components which may come into that
sewage in preparation for disinfection. That is a crutch
and is not an ideal situation, but that is where we are
with ozone,
DR. JOHNSON: I will direct this to the panel (and
this is where I would like to see some more data from
Mark and Al and the kind of studies they are doing
now). I was encouraged to find that somebody had
found some ozone residual in wastewater. The first
time I had heard that anybody has been able to strip
out ozone from a wastewater contactor and then
measure some ozone after a gas phase transfer in
another vessel was in what Mark found. We went up
and talked about the data.
That is encouraging and I think that is where we are
going to have to go. I am sorry that AFs comment was
that you could not correlate the ozone residuals that he
had found with his disinfection. That is too bad. Again
relying on our ancient history from chlorine, we know
215
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
that sometimes chlorine residuals do not correlate very
well with disinfection and I think the reason for that is
we are measuring something other than chlorine
residuals. In other words, the analytical method is
getting us in trouble, particularly when we measure
total chlorine. We take a lot of iodide and we take the
pH down and we measure a dichloramine fraction,
which, I agree with George, is probably in large part
not really inorganic ammonia-type chloramine.
Bob Baker did some nice work on the Philadelphia
supply, looking at chloramine residuals and their
various components. The dichloramine fraction really
was not dichloramine. It was organic chloramines,
which are probably not very good disinfectants.
So, I hope that we can start to do some
measurements and I hope that you all will put into
your data, even though it does not correlate very well,
the different ways'you measured ozone residual and
how it did correlate if at all. That is where we are going
to have to go with the ozone field if we are going to
answer this question of what to do with demand as it
goes up and down, other than hitting it with a huge
amount of dosage which we cannot afford to do.
MR. VENOSA: I would like to say one thing
about ozone residual.
The magnitude of the residual has a lot to do with
the type of contactor being used. If you have a counter-
current flow configuration, you can have higher
residuals. With the co-current flow configuration,
residuals hurt you because the higher the residual the
higher must be the off-gas concentration; thus, mass
transfer is inhibited.
DR. JOHNSON: Mark, can you give us some of
the data that you had on residuals and make a few
comments that are appropriate to be made about how
you measure residuals. Somebody back there talked
about the Shuval-Schechter method for measuring
residuals in which you add the KI directly to the
wastewater and measure the iodine generated with
spectrophotometry. There you are measuring all the
oxidants in the solution and that is very different from
what you find in the Standard Methods. . . .the book
that nobody pays attention to but everybody refers to.
It says in Standard Methods that you must, before
you measure ozone residual, strip the ozone out of the
solution and into another container, and that the
Schechter method does not do that.
MR. MECKES: Don, I cannot give you results off
the top of my head. However, we did test residuals by
amperometric forward titration methods at pH4 and
amperometric back titration methods at pH4. We did
do a method that was done several years ago in a
couple of previous studies whereby it is a colorimetric
type measurement, directly taking residual and then
adding postassium iodide to produce the iodine, which
then you titrate. We found this was very erratic.
The backward and forward titrations measured
quite accurately in water. However, when we got to
wastewater, we could correlate the two only
sporadically. Sometimes they would be right on the
money together and other times they would not. We
did do a stripping method as is reported in Standard
Methods. However, we put the unit on-line so that we
had no waiting period from the time we drew our
sample until the time we purged it with air to release
any ozone to the gas phase into our potassium iodide
washing bottle. These gave us extremely low results.
However, we were able to get some measurements
from these and at this point in time I really feel that
there is still a lot of work that needs to be done.
We did not look at Schechter's method.
DR. JOHNSON: It really makes little difference
how you measure the iodine generated. Whether you
use Schechter's method, whether you use
amperometric titrations or what you use. The problem
is getting the ozone measurement separate from the
oxidant measurements and stripping is one method for <
doing that.
Dr. Jain, did you get any data with ozone
electrodes?
DR. J.JAIN: We bought it but it was not
operational. We had to send it back for repairs.
DR. JOHNSON: You got-the very first one, I
think.
DR. JAIN: 1 guess so. I guess we got everything
first.
DR. JOHNSON: That is the trouble with
pioneering.
DR. JAIN: I would like to make a couple of
comments before we divert the subject, not for the
analysis but for the ozone generator as a whole.
1 do not mean to knock down the salesmen, but they
can sell you anything. Don't really believe them what
they say. I am being honest. We have had the
experience. I have salesmen friends who are fine
gentlemen. They are really professional, but yet they are
under pressure to sell you something. Also I have been
in the business for the last six years and I have known
ozone development in this country. All these people
who are the key people from four or five firms, have
been changing with each other like musical chairs.
Now, you go and talk to one person who asks. . .
why did you come here and you say the reason I came
was because the others were not right enough. The
216
-------
OZONE
data they gave was not right, and I could not justify it.
Now, at this point Ed Opatken's data may sound a
little on the high side but for the consulting engineers,
for others who are really wanting to use this data, this
is good data. You will find it is more useful than what
the salesmen are telling you now.
For example, at Meander they state that intially we
were supposed to have four ozonators for the total
capacity. Only two ozonators could be
accommodated. For the next two we had to have more
building. At the time they told us that the ozone
machine would be of this size and you can have four
ozone machines without any difficulty.
Power costs. At this point the indication we have is
that the kilowatt on the air costs us something like 18
kilowatt-hours per pound of ozone. In Mahoning we
are talking something 8 to 10 kilowatt hour per pound •
of ozone. The capacity, rated capacity, the maximum
we could get was something like 150 pound per
machine as opposed to 237 pound, and this I have in
writing. It is not just talking. It can very well be that it
is the first machine, but still they guarantee.
Now, even though they guarantee you, you go to the
specification. We have penalties. How do you really
apply those penalties? Could you go into court? You
are under pressure, you want to make the system work.
Nobody's interested in going to court. So, it does not
really help. Those penalties in the long run are so
minor, for the project to start it takes almost 3 to 5
years and maybe $25,000.. . .they say they do not care
foi that. They made the money.
So, please be sure, before you take a number, you
have to have confidence in yourself, not what they are
telling. They are telling because from their interest
they want to sell you anything.
An example is this: at Meander we have been
struggling three years. We have people with W.R.
Grace, who are the same people with Union Carbide
and they are helpless to help us. For example, coming
back to the Springfield plant. I was told the
Springfield plant is working fine and I suppose it is
working, but it is very different from Meander. They
are not. using recycled oxygen for producing ozone.
Instead they have a once through system. They are
treating their off-gases to the UNOX system. So this is
different.
When we had a meeting with these people they told
us the control system is the same. If the control system
is the same, instrumentation the same, why doesn't
Meander work? They have tried and it has not worked
so far. We are still praying, we are praying from
different mediums, and it has not worked.
MR. RAKNESS: I would add to that by saying that
there is such a thing as innovative technology and I
think that that is where we are now. Six years ago, or
so we first looked at design, and two years ago
we installed this and that. If you are thinking as
design engineers and as communities that are
going to put in ozone systems, you now are
probably looking at some innovative technology.
What I would recommend to you is to understand
the entire system and talk to different companies,
different manufacturers, different people who had
been involved with it, and understand the entire
process and all the in's and out's about it. At that time,
make your own decisions as to the approach to use and
the things that you can rely upon. Some of the major
things I consider are to look at two different
conditions: the design condition and the initial
operating condition. That is what the client sees when
you first start up. As I showed on the graph yesterday,
the initial power utilization was much greater than at
design conditions by a factor of half again as much.
That is what the consumer initially sees when you start
up the system.
So, understand the system, talk to different
people, and make your own decisions as to what
approach to take.
DR. RICE: I have one word . . . amen.
MR. FLEISCHER: I have several comments
and I want to make them very quickly. The first
is that, as I said before, nobody came to me. I
have had power curves available for you for two
years that are free and open to the public.
MR. OPATKEN: It is not necessary to come to me.
Go to the consultants. . . .go to the industry itself.
MR. FLEISCHER:The consultants have these
curves when they ask for them and with all due respect
to Dr. Jain, 1 have been dealing with Dr. Jain on
Cleveland Westerly for two years. My power numbers
have not changed for two years. My capital costs have
not changed for two years, and in fact, I was told that a
building for Cleveland has to be built with a second
floor that PCI will not even utilize because the other
manufacturers need it for space.
I agree that the manufacturers can vary in power
requirements because their generators are different.
Harvey represents Union Carbide. I represent PCI.
These are two state-of-the-art ozone generators that do
not nearly consume as much power, but Dr. Jain is
stuck with the situation, in a bid situation where he
may not have a Union Carbide generator or PCI
generator. He may have one that consumes more
power, and there is a problem. How do you penalize
217
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
somebody who already has a generator on-site?
The second comment I would like to make is that for
at least the two years that I have been with PCI, dew
point indicators or analyzers are standard on our
equipment, and, in fact, I have not seen a specification
come from a consultant that has not required, whether
it be standard or optional, a dew point analyzer. It is
just as plain and simple as that. We have never put out
one without one.
MR. OPATKEN: Did you have one on yours,
Kerwin?
MR. RAKNESS: No
MR. OPATKEN: There is an answer.
MR. FLEISCHER: What size generator was it?
MR. RAKNESS: 76 pounds per day.
MR. FLEISCHER: That is not true. There is a
dew point indicator.
MR. RAKNESS:There is a color changing
indicator that suggests as the dew point reaches the
point where you have moisture. . . .
MR. FLEISCHER: That is the dew point indicator
which is standard on smaller models; on larger models
when you get over 100 pounds a day, then you get into
analyzers.
MR. RAKNESS: That is a good point, but what I
want to add is that when the operator looks at the dew
point indicator and sees a color change, then he knows
something is wrong. He can look for a problem.
The problem with the color changing indicator was
that it was not sensitive to gradual changes in dew
point and you could not sense a problem as quickly as
you could with the dew point analyzer. The problem
with the dew point analyzer was that it still required an
operator to look up and observe a change in dew point
and that was checked once per day, every morning,
and between the morning check and the next morning
check, the generator could flood and did flood.
That is where we got into the idea in the
recommendation that we made in the talk yesterday
that consideration be given to a high dew point alarm
or an automatic shut-off that prevents the generator
from flooding.
MR. FLEISCHER: I agree with you. The indicator
is just as you said, but this is typically what is required
in the smaller type of units. When we are getting to the
point of 100 pounds or more ozone, then the dew point
analyzers are typically supplied instead of the
indicator.
By the way, the consultant is always made aware of
what is available. He is told specifically that that is an
indicator and what it does and that if he wants an
analyzer, it will cost him in the 100 pounds a day
probably less than 1% of capital cost to add an
analyzer and that is his option.
MR. OPATKEN: Yes and I think that is great
because this technology that we are transferring today
is telling this consultant to make sure he does have a
dew point meter that is not a color changer.
DR. RICE: I have a point to make on dew
point meters. In the European ozonation systems
they say there are two reasons for having a dew
point monitor, not an indicator.
First is you want to make sure you have low dew
point to make the most ozone at the lowest power cost.
The second is, which to me is even more important, is to
lower the maintenance requirements. Once the air goes
above -40"C, then you are going to have to maintain
the ozone generator, because you are going to have
moisture getting in there and I am not sure exactly
what all happens. What they say is, as long as we are
monitored at -60°C and lower, we are only going
to have to maintain that ozone generator once a
year if that. Thus, once the indicator says you have
exceeded the dew point, it is really too late in terms
of maintenance. The damage has been done.
DR. HILL: I just wanted to compliment the
EPA's work on the contactors. The comparison
of the three types of contactors was something
that was needed for some time. Regarding the
talk about the family of curves on ozone
production, I would like to make a comment on
that, showing one slide.
What this slide shows is the performance of a
laboratory ozone generator, the Welsbach 223, as a
function of pressure. People who have discussed
ozonator performance so far have mentioned that
power, as it is controlled by voltage and frequency, is
very important and they have also talked about how
the gas flow is very important.
In this experiment the ozonator gas flow was
maintained at 0.1 standard cubic foot per minute of
oxygen and the primary voltage, that is before it is
transformed to high voltage, was maintained at 100.
The only thing that was changed was the pressure
within the ozone discharge reactor, where the ozone is
generated. In the far left you can see that at 1 psig
the concentration of the ozone is 37.5 mg/1, and
then when you come upt to 8 psig, it is down to
20 mg/1. So, for this type of ozone generator there
is a very big difference in pressure. It amounts to
85% increase in ozone generator production in
going from 8 psig to 1 psig. I believe this parti-
cular manufacturer rates its equipment at 8 psig.
This is an important consideration.
218
-------
OZONE
Now. not all ozone generators work better at a low
pressure. Each ozone generator has one optimum
pressure, and in the more traditional designs that have
a long tube and about 3 inch diameter they usually
have a discharge gap of about 3 millimeters. This
discharge gap is one of the most important things in
determining what the optimum pressure is.
For these types of traditional generators they seem
to work a little better at low pressure. Some of the
newer designs such as the PCI and the Lowther plate
that Union Carbide sells, have a smaller discharge gap.
They are made to operate at their optimum pressure at
a higher pressure, closer to what you would need to go
into a tall column of water.
I just wanted to make this caution. Those people
who are using these families of curves should make
sure that the pressure is the same when they attempt to
use that in their equipment. A good reference I would
like to recommend, if someone is interested in finding
out how an ozone generator works, is a dissertation by
a fellow named Anthony Edelman. It came out in 1967
at a university in the Netherlands, and where I
found it was at the Library of Congress, but it
may be available other places.
MR. WHITE: As a special advisor to consultants,
we have a problem and I think we are missing the point
here. When a consultant specifies a piece of
equipment, he is going to expect the manufacturer to
perform, and I have no reason to believe that any of
the ozone manufacturers, in spite of what we heard
today, would not want to perform. Why keep
yammering about disinfection limits when we start
talking about disinfection. I am sure we are not going
to agree. The people in California are not going to
agree with the people in Alabama, or for instance,
Vancouver, British Columbia, where they think 10,000
per 100 milliliter coliforms is disinfecting their sewage.
So, we can just forget all about that and all these
requirements in disinfection.
I am working on two projects where I am
recommending a combination of ozone and chlorine
and what I want to know is: what is the ozone
consumption on a well oxidized domestic sewage? I
need to know this and we just do not have any
performance data that we can hang our hat on. We
have to go back to performance data that is on potable
water which is relatively clean and that is all we have.
So, what we need is more do and less talk.
We need this performance data and I do not know
how we are going to get it except having installations
where we can actually observe the operation. I might
add that one of the ways that the people I talked to in
France operate their ozone equipment is that they
routinely make ozone demands everyday and they
have a historical record of the ozone demand of the
water that they are treating. This goes back to the same
thing that we have done in sewage chlorination. We get
a historical record. We know what the chlorine
demand is, except that we do have the advantage of
being able to control from residual and this hampers
the ozone situation. But I need information, and that is
the kind of information I need. I need cost per pound
of ozone and somebody has to tell me how many
pounds of ozone per day I need to put into an effluent
to achieve a certain disinfection.
MR. VENOSA: I will give you a copy of my
paper, George. It is all there.
MR. WHITE: I have a copy of your paper.
DR. LONGLEY: I want to give credit to Archie
Hill. He is coming up with what appears will be some
very interesting information. His report should be
written in about two months. He is under contract to
the U.S. Army Medical Bio-Engineering Research and
Development Laboratories. So it is not a proprietor-
oriented report. Hopefully, it is a report written with
great objectivity.
I want to develop a little on what Don was saying a
while ago, and I want to speak as a member of the
Standard Methods Committee. Don happens to be the
chairman of the Joint Task Force for Chlorine
Residual Techniques.
I think what we have discussed here a little bit in the
past few days is some of the problems with residual
analysis. At least it keeps rearing its ugly head. As I
understand, one of your objectives of this symposium
might be to identify future research needs. I suggest
that what we need, whether it be an inhouse EPA
effort or whether it be an extramural effort, is to look
at the analysis of oxidants and to start developing
some of the information that we need on accuracy,
precision, the problems that we have with the various
techniques, their applicability, and we can talk forever
about it. I think this is an important area that has been
overlooked, at least in the recent past insofar as the
research dollar and the research effort is concerned.
MR. OPATKEN: We are working on a grant
agreement between Miami University and EPA in
regard to the residual analysis. That is underway at
this time.
DR. RICE: On this subject, extending it a little
further and back into drinking water where the
example is, the French use ozone for many reasons,
one of which is virus inactivation. What they do is
attain an ozone residual level of 0.4 mg/1, and after
219
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
attaining that, maintain it in a plant for 6-10 minutes.
Then that guarantees polio virus 1, 2, and 3 have been
inactivated. This is a performance disinfection
standard which was first adopted in Paris and is now
adopted throughout France. France is the only
country which has formally adopted that.
The other countries that are using ozone for
disinfection of drinking water are doing the same
thing. The extrapolation is that if one would take the
time to pretreat sewage to the point where you can get
residual ozone and would have confidence that there
are no hills, valleys, particles for the viruses to hide,
one could then use that technique for virus
inactivation.
1 am not recommending that in sewage treatment
because I think that would cost a fortune.
DR. JOHNSON: That is something chlorine
cannot do. So, I think that we want to put a little
encouragement into folks particularly after they have
been discouraged as Dr. Jain has by all of his
problems, because ozone has some real potential. It is
an excellent disinfectant.
MR NOVAK: Rip, I would just like to add to what
you are saying. I agree with you that there is moni-
toring equipment that is being used now and, just
to add to what Harvey said before, they do measure
ozone concentrations out of the ozonator going into
the system. They measure the off-gas, but they also
can measure residual ozone in the water. As Rip just
said, they measure 0.4 ppm ozone in the water after
5 minutes and consider that a disinfected water.
DR. RICE: A virally inactivated water, which is
also bacterially disinfected, but they do not have a
bacterial disinfection standard. It is only that viral
thing.
MR. NOVAK: The work that Dr. Rook has done
also indicates that the bacteria kills are no longer
there.
DR. RICE: Well, yes, but how many fecal
coliforms do you have in the drinking water?
MR. NOVAK: If you look at raw water supply and
if you are taking water from the Rhine you are talking
about a sewage treatment plant. They do phys-chem.
They do promoted sedimentation. They have laminar
flow devices, sand filtration, activated carbon. They
treat it just like a sewage treatment plant. So there is
not that much difference. Mr. White, you asked for
data. We have published data on that Indiantown
plant that has been running for about three years now,
which gives a chemical analysis, biological analysis. It
gives residual ozone dose, residual ozone absorbed in
the entire system: So you have some back-up data for
engineering design.
One other thing on contactor systems, when we talk
about killing bacteria, we first built our plant with a
bubble tower type reaction chamber and we ran into
major problems with that type of an injector system. It
would run very well for approximately 1-2 weeks, 3
weeks, and then all of a sudden we would start to get
some types of slime growths on the side of the wall and
we could not get consistent, reliable bacteriological
kills. We would have to shut the whole system down,
scrape the walls, and then start our system up again,
and we would get good bacteria readings for
approximately 2-3-4 weeks.
I admit it was in a tropical climate in Florida. It got
very hot and very humid and that may not be the case
all over the United States. But we switched to a very,
very rapid contacting system and were able to get
consistent bacteriological kill with one minute contact
time of ozone and the water and the wastewater with
very low dosages. We reduced the ozone dose rate
from about 12 to 14 parts per million down to about
4.5 to 5 parts per million. If we use a single stage
reactor we can get transfer ratios of about 80 to 90%. If
you go to a two-stage reactor you can get 100%
utilization of the ozone.
MR. DeSTEFANO: I have a question for Mr.
Venosa that is sort of related to this residual question.
Is ozone a surface active oxidant? What do we know
about the action of ozone versus chlorine on cell surfaces
or, is it surface or is it in through the membrane or what?
MR. VENOSA: I have absolutely no idea.
DR. RICE: If you have a surface which can be
oxidized, yes, ozone is surface active, and so are all of
the other oxidants as well.
MR. DeSTEFANO: Right. I am sure a lot of the
engineers are familiar with the fact, maybe more
through water treatment than with surface charges
and surface-surface interactions, when you are doing
coagulation control and maybe adding alum and a
polymer, you are trying to neutralize the negative
charge on whatever is in there. They are all colloids.
One of the jobs that I have with the firm I am with is
to keep up research on this because we manufacture an
electrophoresis instrument. I will not mention the
name, I am not trying to sell them. It seems to me with
any surface-active oxidant that you have to get past
whatever is absorbed onto the cell before you are going
to kill the cell, and I am not sure if any research has
been done, maybe not in this field but maybe in
microbiology, about how much choice the cell has on
what absorbs onto its membrane on a macro scale,
talking about big proteins that are present in
220
-------
OZONE
wastewater. I have seen research done with albumin on
clays and what you get with albumin absorbing onto a
clay is a high buffering capacity, where say a dosage
1,000 parts per million of salt would salt out or
coagulate a clay. If you absorb a certain amount,
maybe 1,000 ppmof albumin or just put egg white in it,
you can go up to maybe 20,000 without coagulating.
DR. RICE: I am not aware of any research of this
sort in sewage treatment where they are using ozone,
but there is some in the drinking water area, notably
pesticides in the presence of added Kaolin or
suspended solids, something that can adsorb the
pesticide on the surface. Now, without the adsorbing
substrate, the ozone takes the pesticide which 1 think
was Aldrin which is very reactive with ozone, down to
zero in a very short time, but adsorbing it onto the
substrate makes it much more oxidation resistant. It
took more ozone, took more time and so forth.
MR. DeSTEFANO: But what I am more concerned
with is the smaller molecules adsorbing onto the
larger cells. Now, Dr. Johnson was saying that you
cannot get complete correlation between residuals and
disinfection. He thought that that is probably because
you are not measuring the residual correctly. Well
assume you are measuring the residual correctly, it
seems to me that possibly the protein, or unless you
have something that is chlorine resistant and is going
to buffer the cell against the reaction of the
hypochlorous acid or hypochlorite ion or whatever
you consider the disinfecting ion, that is going to be
the limitation to disinfection rather than a strict
chlorine residual.
It is something that 1 admit is hard to measure and I
am not even sure you would be able to use this as a
process control. Maybe as consulting engineers keep
in mind when you are talking about disinfection, you
try to imagine yourself as an E. coli floating through
the plant and what you see and what is happening to
you when you go through. . . .
MR. VENOSA: Thank-you. We will keep it under
advisement!
_DR. LONGLEY: What the gentleman I think is
talking about is mechanisms and we probably need
some more mechanism work. Vincent Olivieri,-who
was here yesterday, was talking about CIC^, has done
mechanism work on phage. It is interesting work. It is
long and hard and tedious and what comes out of it is
ultimately useful. Well, I think it is, but apparently
EPA does not because they do not want to fund that
type of research right now.
QUESTION: I have a question in connection with
analytical procedures. I was wondering whether
anybody has had any experience utilizing
methodology which FMC has published using a
titanium sulphate solution acidified with sulphuric.
and then diluted. It is a colorimetric reaction which
works very well for unreacted hydrogen peroxide. If it
works on that, it should work pretty well on ozone too,
since basically the chemistry is the same. Has anybody
had any experience with this methodology? The
reason I am asking is 1 am the new chairman of ASTM
D-22 for ozone and we do not have a method.
1 am guessing here but the problem is not that you
cannot analyze for strong oxidants, The problem is
that as the strong oxidants react with substrate
materials, you get a lot of weaker oxidants produced
and it is very difficult to distinguish among this
spectrum of oxidants. The problem is to analyze for a
strong oxidant like ozone and the presence may
include peroxide and God knows what other organic
peroxy-compounds of a variety of types of varying
reactivities.
MR. WOOD: I would like to make a comment
here. 1 am somewhat amazed by a reaction that is
described and not giving the end product. We hear
people talking about putting controls on sensing
flooding of a reactor with the answer that you shut that
unit down in case of that malfunction. We have to put
equipment in and one of the stipulations is in no way
can we shut down that equipment. In watching a sewer
plant operation, you cannot shut off disinfection. So,
sensing it and shutting down the piece of equipment is
not the end of the reaction because you just cannot in
my opinion do this.
You have to cut in the standby piece of equipment
which means you do have to have a standby.
MR. RAKNESS: That is right and that is what we
have. So it is shutting down one and turning the other
one on.
DR. JAIN: 1 want to make a couple of very brief
comments. I want to make myself clear that I am for
ozone. Ozone has a big future. In fact, Rip Rice,
Harvey Rosen, and I started at the same time in the
business.
The reason why I am complaining about Meander
and keep on harping on that is because it is hurting the
cause of ozone. It is in the interest of everybody that
Meander should work. If it does not work and we have
to go to chlorine, it is going to hurt ozone and not any
particular company like Union Carbide.
The second thing was regarding the ozone analyzer.
We have ozone analyzers now especially for the gas
phase, Dasibi. This is a very expensive piece, we are
talking $4,000-5,000. In a plant you need several of
221
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
them and you also would like to have a standby. So,
this is one area where we need research and
development of instruments which are cheaper and
perhaps more reliable. The Dasibi ozone analyzerdoes
need maintenance.
The last one, before anybody makes
recommendations particularly regarding the materials
we use with ozone, they should have some experience.
For the last six years we have been hearing PVC is
good for ozone, unplasticized. You use it. You write it
in specification. Your problems are solved. This
is not true. Estes Park has the experience, we
have the experience. Please do not specify
unplasticized PVC for ozone use.
DR. RICE: On that last point, as a general rule of
thumb, from our survey, the people in Europe say dry
ozone in dry air can be handled in anything. Wet ozone
in wet air should be handled only in a very costly
stainless steel environment but in the long run you do
not have those maintenance problems.
222
-------
SECTION 6. VIRUSES AND ORGANICS
23.
VIRUS INACTIVATION IN WASTEWATER EFFLUENTS
BY CHLORINE, OZONE, AND ULTRAVIOLET LIGHT
R. A. Fluegge,* T. G. Metcalf,** and C. Wallis***
*The Carborundum Company, Niagara Falls, New York
**University of New Hampshire, Durham, New Hampshire
***Baylor University, Houston, Texas
ABSTRACT
Five wastewctter treatment plants were compared in their ability to
inactivate naturally occurring enteroviruses. Two of these five plants
used chlorine, two used ozone, and one used ultraviolet light in the
disinfectant stage of wastewater treatment. Correlation coefficients
between measured virus concentrations and total coliform, fecal
coliform, turbidity, TSS, COD and TOC are .presented along with a
study of the diurnal variation of virus emissions in both winter and
summer at one plant. Polio, Coxsackie A, and Coxsackie B viruses
have been identified and their relative concentrations are presented
for all five wastewater treatment plants. Virus isolation rates and
virus concentrations were shown to be significantly reduced by all
three modes of disinfection.
INTRODUCTION
This report is a summary of findings for a field and
laboratory program conducted by The Carborundum
Company to determine the virucidal effects of chlo-
rine, ozone, and ultraviolet light under actual operat-
ing conditions at five wastewater treatment plants.
The work is sponsored by the U.S. Environmental
Protection Agency (EPA) and greatly aided by the
inputs of Mr. Albert D. Venosa, EPA Project Officer,
his associates and staff, as well as the on-site personnel
with whom we have worked at each Wastewater Treat-
ment Plant. Mr. Robert Fluegge is Principal Investi-
gator. Dr. Theodore Metcalf, University of New
Hampshire (UNH), Dr. Craig Wallis, Baylor Univer-
sity, and Dr. Joseph Melnick, Baylor University, are
consultants to this program. Carborundum Field
Engineers responsible for all field work include Messrs.
Richard Thacker, Ben Moore, Kevin McConnaghy,
and Charles McGee.
The field portion of the program required extended
sampling at four wastewater treatment plants. These
were located in New Jersey, Colorado, Ohio, and
Massachusetts. The laboratory assays were completed
at facilities in New Hampshire. Program coordination
was located in Niagara Falls, New York.
WASTEWATER TREATMENT PLANTS
Four wastewater treatment plants were studied
during this program. A total of five separate sample
collection periods were completed between November,
1976, and May, 1978.
223
-------
VIRUSES AND ORGANICS
Walclwick (Northwest Bergen County) Operations
The Northwest Bergen County Waste water Treat-
ment Plant, Waldwick, New Jersey was sampled dur-
ing the spring of 1978. It used a conventional and or
step aeration activated sludge system followed by
disinfection with ultraviolet radiation (for this study)
with the final effluent treated with chlorine. A block
diagram for this plant is shown in Figure 1.
This plant had a design flow of 3.2xl04m1 d (8.5
mgd). However, the total plant flow during these
studies was only 1.9xl04m1/d (5.0 mgd). More details
of this plant are given by O. Scheible (2).
DESIGN AVG FLOW 3 2«lo'm'/0 la 5 mgd)
UDGE DEWATERING: CENTRIFUGATION
DISINFECTANT UV LIGHT
SAMPLING POINT A 3 METERS (10 II )
FROM UV CONTACT TANKS
SAMPLING POINT B AT EXIT POINT OF UV LIGHTS
FIGURE 1. NORTHWEST BERGEN COUNTY SEWAGE
TREATMENT PLANT: N.J
The ultraviolet light source was a prototype unit
manufactured by Pure Water Systems, Inc., Fairfield,
New Jersey. It housed 400 100-watt kimps with 30%
power at the germicidal wavelength of 254nm. Each
lamp was contained in quartz jackets measuring 1.5m
(60 inches) in length bay 23mm (0.9 inches) in outside
diameter. The quartz jackets were spaced 12.7mm (0.5
inches) apart at the surfaces. The contact time was
3.6 seconds at a flow rate of 1.9xl04m\ d (5.0 mgd).
After passing through the ultraviolet disinfection
stage, the liquid effluent was treated with chlorine as
required by the State of New Jersey before it was dis-
charged into Ho-Ho-Kus Brook.
Estes Park Wastewater Treatment Plant Operations
The Estes Park Wastewater Treatment Plant, Estes
Park, Colorado; consisted of a conventional activated
sludge plant followed by nitrification, tri-media filtra-
tion, and disinfection by ozone (air was used for ozone
production). The design flow was 5,700 m-'/d (1.5
mgd), but the actual flow was 2800 m-'/d (0.75 mgd).
It was sampled during the spring of 1977. A block dia-
gram for this plant is shown in Figure 2. A detailed
description of the plant is given by Rakness and Hegg
: (1).
Ozone was produced on-site from air using a Wels-
FIGURE 2. ESTES PARK SEWAGE TREATMENT PLANT:
COLORADO
bach corona discharge generator. The contact tank
was a rectangular diffuser chamber divided into nine
equal sized' compartments by baffles. Wastewater
effluent flowed over and under the baffles in serpen-
tine fashion, providing cocurrent and countercurrent
gas-liquid contacting. The contact time was 37
minutes during the virus testing period. The average
dosage applied during the period was 4 mg/1.
Muddy Creek Sewage Works and R. A. Taft
Laboratory Operations
The Muddy Creek Sewage Works, Cincinnati, Ohio
was a secondary wastewater treatment facility using a
conventional activated sludge system followed by
disinfection with chlorine. Separate tests were per-
formed using ozone as a disinfectant in December,
1976, and chlorine in July, 1977. A block diagram of
this plant is shown in Figure 3.
FIGURE 3. MUDDY CREEK SEWAGE WORKS: OHIO
Ozone was incorporated as an alternative disin-
fectant for part of this study. Unchlorinated secondary
effluent was transported in a 20 m' (5,400 gallon),
epoxy-lined tank truck to EPA's R. A. Taft Laboratory
pilot plant. The effluent was pumped into the labor-
atory building and through a packed column ozone
contact tank. The applied ozone dose during the test
period was 8 mg 1. The average residual was 0.13 mg/1.
Contact time was 30 seconds at a liquid flow rate of
224
-------
VIRUSES AND ORGANICS
75 1/min (20 gpm). A full description of this facility
is given by Venosa et al (3).
Marlborough Easterly Wastewater Plant Operations
The Marlborough Easterly Wastewater Treatment
Plant, Marlborough, Massachusetts was a conven-
tional two-stage activated sludge system with nitrifi-
cation followed by disinfection with chlorine. Effluent
samples from this plant were studied in October, 1977.
A block diagram of this plant is shown in Figure 4.
during the test period.
A summary of effluent quality characteristics for
each of the five study sites is presented in Table I.
Sample Collection and Assay
A Carborundum mobile laboratory was driven to
and remained at each site during the extended sam-
pling periods. Figure 5 is a side view of this labora-
tory on-site at Estes Park, Colorado. The unit was
fully self-contained with onboard generators, water
tanks, air conditioning, sinks, and storage facilities
for chemicals and equipment.
FIGURE 5. CARBORUNDUM MOBILE LABORATORY
FIGURE 4. MARLBORO EASTERLY WASTEWATER ON-SITE AT ESTES PARK, COLORADO
PLANT: MASSACHUSETTS AH sample collection and treatment was accom-
The chlorine contact tank provided a contact time of plished inside this mobile laboratory. The surrounding
50 minutes. The chlorine residual averaged 1.3 mg/l environment was controlled to eliminate spurious
TABLE 1. SUMMARY OF EFFLUENT QUALITY CHARACTERISTICS AT THE FIVE STUDY SITES
Parameter
TSS, mg/l
COD, mg/l
TOC, mg/l
Turbidity, NTU
Disinfectant Dose
Disinfectant
Residual, mg/l
Contact Time,
minutes
Total Conforms
per 100 ml before
disinfection
Total Conforms
per 100 ml after
disinfection
Bergen
5.3
30.2
13.0
2.1
131,000
((j-watt-sec/cm2)
0.06
1.8 x 105
3.7 x 101
Estes
8.5
36.1
—
5.0
4.2 mg/l
37
8.4 x 105
3.8 x 103
Muddy Creek
2.7
27.9
11.6
1.8
1.2
51
1.65 x 105
6.5 x 103
Marlborough
12.6
24.4
6.5
5.0
1.3
47
1.8 x 104
2.3 x 102
Muddy Creek Taft
7.4
39.2
11.2
4.0
7.9 mg/l
0.13
0.5
7.8 x 10b
3.1 x 103
Fecal Coliforms
per 100 ml before
disinfection
Fecal Coliforms
per 100 ml after
disinfection
4.1 x 10"
2.4 x 10'
6.9 x 105
3.3 x 103
7.8 x 10"
1.0 x 103
1.8 x 104
8.4 x 10°
1.9 x 105
1.2 x 103
225
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
inputs from the nearby waters, atmosphere, and gen-
eral on-lookers. As will be evident later, the virus con-
centrations were quite low and hence required care in
sample taking.
Carborundum's Aquella Virus Concentrators were
mounted inside this laboratory and used throughout
the program.
Sample Collection and Virus Concentrations
The Aquella Virus Concentrator was a portable,
self-contained apparatus designed to remove water-
borne virus from large sample volumes on a contin-
uous basis. The purpose during this phase of the exper-
iments was to field process approximately 380 liters
(100 gallons) of wastewater per sample and to concen-
trate the sample to as small a volume of liquid as
practicable for analysis to facilitate handling and ship-
ping problems.
All sample hoses, pumps, filter holders, and tubes
were rinsed, sterili/ed, and rinsed again prior to col-
lecting each sample. Dilute hydrochloric'acid (HCI) at
a pH of 0.5 was used to sterili/e all components. The
HCI contacted the components for 30 minutes prior to
final rinsing. The sterilization procedures have proven
FIGURE 6. WALL MOUNTED AQUELLA™ VIRUS CON-
CENTRATOR UNIT
to be very successful, especially when sampling potable
waters in the vicinity of highly contaminated waters
such as typically found in raw wastewater areas.
Two units are mounted on the side walls of the
Carborundum mobile laboratory used for this pro-
gram. Figure 6 shows one of these units bolted in place
on the passenger side wall of the van shown in the
previous figure. This arrangement is somewhat differ-
ent from the standard configuration in which the free-
standing unit is connected together with hinges and
end panels for ease of handling and shipping. Since
this van was used at all wastewater plants studied, the
units were permanently mounted.
The laboratory walls, ceiling and floor were disin-
fected prior to collecting each sample. Chlorinated
water was sprayed on all surfaces and allowed to dry.
In addition to laboratory sterilization procedures.
Carborundum Field Engineers were .rained to insure
personal cleanliness. This is especially important while
sampling sewers, settling tanks, and aeration cham-
bers. Every effort was made to eliminate cross con-
tamination between the equipment, the laboratory
and the on-site personnel.
Wastewater was collected by placing a submersible
pump in the center of the flow stream being sampled.
Water was pumped through one or several nylon
garden hoses to the inlet of the Aquella Virus Con-
centrator and entered at the right of Figure 6. A flow
rate of 3.8 1 min (1.0 gpm) was typical in this experi-
ment. Water continued through an impeller pump
and was further forced through three clarifier filter
holders which may or may not have contained filters.
The requirement for filters was primarily governed by
solids in the wastewater. Several times during these
experiments, clarifier filters were required due to
particulate loadings. Once through the clarifiers,
water flowed through a chemical proportioning pump
which had several satellite liquid injectors. Con-
ditioning chemicals were added at predetermined rates
which were automatically controlled by volume flow.
This feature became important as filters plugged and
How rates correspondingly decreased.
Sodium thiosulfate, dilute HCI and A1C13 were
added through the satellite liquid injector chemical
proportioning pump. The exact concentrations were
dependent upon the wastewater characteristics, but
our researchers and consultants have shown good
virus recoveries for an adjusted pH between 3 and 4,
A1C13 concentrations near 1.5 mM, and sufficient
sodium thiosulfate to eliminate active chlorine,
ozone, or other oxidizing compounds.
A flow totali/er was mounted in the flow stream just
prior to the two custom designed virus adsorbing
filters shown at the lower left of Figure 6. After passing
through these two final filters, the wastewater was re-
turned to the treatment plant.
Most virus collection occurs on the two final ad-
sorbing filters. However, for those experiments re-
quiring clarifier filters, the latter must also be subjected
to virus removal procedures. These experiments have
226
-------
VIRUSES AND ORGANICS
shown some viruses were adsorbed to clarifier filters.
Viruses were eluted from the filter surfaces using col-
trolled methodologies which have been developed
over several years at Carborundum as well as other
research and governmental laboratories. These pro-
cedures included a saline rinse to remove residual
chemicals and elution with 0.5 M glycine at a pH
above 10. This step typically produced between 1 and
3 liters of solution containing most of the viruses
present in the original wastewater sample.
Further liquid volume reduction was accomplished
in the field using a gel reconcentration technique
which incorporated 3 mM AlC^coupled with Na2C(>j
forming a floe at pH 7 and gravimetrically or centri-
fugally separating the supernatant. Viruses were con-
tained in the floe. Viruses were removed from the floe
by raising the pH above 10 with 0.5 M glycine, collect-
ing it, neutralizing, and adding 10% fetal calf serum.
The final concentrate was quickly fro/en at dry ice
temperatures and shipped to our assay laboratory.
Shipping was in accordance with all Federal, State,
and Carrier regulations.
Special care was taken throughout this process
to eliminate sample contamination by either the
sample collector, the surroundings, the containers,
or shipping techniques.
Techniques
Samples collected during this program were ana-
lyzed for virus concentration and identification, under
Carborundum's direction, in Durham, New Hampshire.
A schematic diagram (Figure 7) depicts the route
taken by each sample within the laboratory.
Until it was placed on test, each sample was main-
tained at -80°C to avoid denaturation and other
changes detrimental to virus recovery and infectivity.
Removal of bacteria, yeast and mold was accom-
plished by either an 8-16 hour ether or an antibiotic
treatment. Diethyl ether was used in our early experi-
ments since enteroviruses have been shown to be ether
resistant under these test conditions. However, ad-
vantages were identified by the autumn of 1977 in
using antibiotics: gentamicin (50 ug ml), penicillin
(100 unit ml), streptomycin (100 ug ml), and funga-
/-onc (1 ug ml). Hence, some changes in absolute
sensitivity may have occurred. However, relative sen-
sitivities remained unaffected.
Following sterilization treatment, samples were
hydroextracted using polyethylene glycol. This pro-
cedure further concentrated the samples prior to test.
Natural virus testing was split between two culture
types: Buffalo Green Monkey (BGM) and Primary
Monkey Kidney - African Green (PMK). This was
I
One-Half Sample
I
Buffalo Green Monkey Culture
Coded, Frozen Sample Received
Storage at -80C Until Analyses Started
Sample Isotonic, Bacteriologically Sterile
I
I
One-Half Sample
I
African Green Primary Monkey Kidney Culture
"Plaque Development Procedures Used-Monolayers Incubated at 37C for 14 to 24 days
. I
I
No Plaques Formed
I
Results-Negative
I
Plaques Formed
I
Plaque(s) "picked" and passed to Fresh
Monolayer (Same Type as for Original Plaque)
Cytopathic Effect Obtained, Virus
Isolant Confirmed
Virus Isolant Identity Determined -
Intersecting Pool, Serum Neutralization
Tests Using Lim, Benyesh-Melnick (IBM) Pools
'Controls for non-virus contaminants, adventitious virus, monolayer integrity-monolayers inoculated with serile diluent and overlaid with overlay media
FIGURE 7. TEST PROCEDURES USED FOR VIRUS ANALYSES OF TEST SAMPLES
227
-------
PROGRESS IN WASTE WATER DISINFECTION TECHNOLOGY
done to obtain as great an isolation sensitivity as pos-
sible. All examinations in cell cultures were made by
means of plaquing methods which allowed separation
and enumeration. Cell culture incubation periods of
up to 24 days were used to maximi/e development of
slow-forming plaques.
Following the formation of a plaque, natural viruses
were recovered by means of "plaque picks". This in-
volved transferring the picked plaques to a fresh cul-
ture monolayer with a liquid overlay medium. The
monolayer was always of the same type as that used
tor the initial recovery. Consequently, the loss of virus
during passage due to the use of a nonsensitive culture
was avoided. Identification of isolates was made by
intersecting pooled serum-neutralization tests carried
out in microtiter plates (Lim-Benyesh-Melnick Anti-
sera Pools).
RESULTS
Virus Isolation Rates and Concentrations
Virus isolation rates for each of the five wastewater
treatment plants are given in Figure 8. The white area
FIGURE 8. MONTHLY VIRUS ISOLATION RATE
represents the isolation rates over the test period
shown. The isolation rate is calculated for both the
treated and treated plus disinfected wastewater by
equally weighting each sample shown to contain one
or more virions. At least forty percent of all samples ol
treated wastewater contained at least one virion prior
to the disinfection stage of treatment.
These results showed no major differences between
seasonal isolation rates, particularly for the two sam-
pling periods completed in July and December at the
Muddy Creek Sewage Works. These two rates were
75% and 60%. respectively, and to identify significant
differences would require more testing. This finding
may be significant in the disinfectant usage plans
being proposed for winters in the Northeast of the U.S.
The virus isolation rates shown as blackened-in
areas under each treated wastewater level indicator
show that significant virus reductions were measured.
Reductions occurred for each of the three disinfectants
(UV, chlorine, and ozone) studied. Virus isolation rates
were reduced four times for UV; 29 times for ozone
(Estes Park); for chlorine, two times in one case, and
infinity in the second; and for packed column ozone
(Taft), eight times. Additional sampling would be
required to obtain a definitive difference between the
three methods of disinfection.
These same data were analyzed to obtain an average
virus concentration per one hundred gallons of both
treated and treated plus disinfected sewage. All
samples taken at each plant were analyzed and the
results averaged. The results represent the average
concentrations for about 300 gallons of effluent
sampled at each site. Figure 9 is a comparison of these
MONTH
FIGURE 9. MONTHLY VIRUS CONCENTRATION
average concentrations. Again, there may be no sig-
nificant seasonal differences, especially when compar-
ing the two Muddy Creek Sewage Works data for July
and December. Both sets of data show about six
virions per one hundred gallons of treated wastewater.
The blackened-in area in Figure 9 shows the virus
concentrations after the disinfection step. The Estes
Park Wastewater Treatment Plant had the poorest
ratio of disinfected vs. treated wastewater of those
plants measured, However, during four of the
days we tested the plant, the ozone dosage rate dropped
below 3 mg, 1, and hence this was likely the cause of
high virus concentrations in the effluent. Neglecting
the data collected on those four days gave the black-
ened-in area shown in Figure 9. The average concen-
228
-------
VIRUSES AND ORGANICS
tration would have been 8 pfu/ 100 gal. with all data
included. A comparison of these isolation rates as a
function of applied o/one dosage will be presented
later.
Statistical Significance of Disinfection
Application of the Student's t Test to all of our
measured virus isolation rates and concentrations is
given in Table 2. The t-statistic is given along with its
found in this study. However, it is not known whether
they were the "wild" or vaccine types. The wild strain
of poliovirus is known to produce meningitis and
paralytic polio in man. Coxsackie A9, Bl, B2, B3, B4,
B5 and B6 were also separated and identified during
this program. These latter viruses cause summer rash,
polio-like paralysis, chest pains and colds. Note that,
except for the CBS virus, all virus types were found in
TABLE 2. STUDENT'S t SIGNIFICANCE TEST FOR VIRUS IN TREATED VERSUS DISINFECTED SEWAGE
Bergen
(UV)
t-Statistic 2.5
Average Discharge
Conc./100 gal. 0.1
Mean of Difference
Conc./100 gal. 1.1
Std. Dev. of Difference
Conc./100 gal. 1.8
Significance
Level 2.2%
Estes Muddy Creek
(Ozone) (Chlorine)
(1.4)*
1.0 2.2
(0.4)
8.0 1.0
(3.7)
3.6 6.1
(8)
14 11
(16%)
16% 2.4%
Marlborough Muddy Creek/Taft
(Chlorine) (Ozone)
2.5 1.9
0 0.07
0.6 6.1
0.9 12
2.2% 3.9%
'Bracketed Values Neglecting Four Problem Days
implied significance level. The smaller this value the
more likely the two virion distributions (before and
after administering a disinfectant) are significantly
different. A low significance level is an indication that
the two levels are significantly different and hence that
the daily decreases due to disinfection are significant
and can be attributed to the addition of UV, ozone,
or chlorine.
The average discharge concentrations as well as the
mean value of the differences in virus concentrates
before and after disinfection are given in Table I. The
discharge concentrates range from "0" for chlorine
added to nitrified sewage to 1.0 virion per 100 gal.
for chlorine added to secondary treated sewage. A
measure of virus reduction is obtained by dividing the
mean concentration difference by the discharge con-
centration. These values range from about six for chlo-
rine added to secondary treated sewage to infinity for
chlorine added to denitrified sewage.
Virus Identification
A summation of all tests from this program to deter-
mine the virus identification distribution is shown in
Figure 10. The relative virus concentrations based on
total virus in the treated wastewater is presented.
Polio and Coxsackieviruses were separated and identi-
fied. All three polioviruses (types 1, 2 and 3) were
20-
0.10
CA9 CB1 CB2
VIRUS TYPE
CB3 CB4 CBS CB6
FIGURE 10. VIRUS IDENTIFICATION: BOTH BEFORE
AND AFTER DISINFECTION WITH CHLORINE, OZONE,
OR UV LIGHT.
229
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
the treated as well as in the treated plus disinfected
wastewater effluent.
Diurnal Variations
Virus concentrations in treated wastewater can vary
as a function of time, as shown in Figures 11 and 12.
2400 02OO 04OO 06OO OBOO
TOOO 1200
TREATED AND
WASTEWATER
1600 1800 2OOO 2200 24OO
FIGURE 11. DIURNAL VARIATION OF VIRUS CONCEN-
TRATIONS IN WASTEWATER EFFLUENTS
DISINFECTANT: CHLORINE
MUDDY CREEK: DECEMBER 19/6
2:00 AM and 4:00 AM did not result in a significant
increase in measured virus levels in the effluents during
this same period. The dashed lines indicate virus con-
centrations for treated disinfected effluent collected
during the same time period as the treated effluent.
Measurement of the virus challenge to a wastewater
treatment plant should take into consideration the
possibility of diurnal variations.
Virus Indicators
The diurnal variation of effluent virus levels dem-
onstrated at the Muddy Creek Treatment Plant pro-
vided a unique opportunity to study the correlation
between the virus levels and the normal wastewater
quality parameters.
The sizable variation in these virus concentrations
would signify that changes in virus concentrations
might be correlated with changes in other parameters,
especially if attempts are made to use these parameters
as virus indicators.
Results of regression tests using virus concentration
as the dependent variable against six separate inde-
pendent variables in the treated effluent from Muddy
Creek during July 1977, are summarized in Figure 13.
H«-TREATED
POSITIVE CORRELATION
2400 0200 0400 0500
FIGURE 12. DIURNAL VARIATION OF VIRUS CONCEN-
TRATIONS IN WASTEWATER EFFLUENTS
DISINFECTANT: OZONE
MUDDY CREEK: JULY 1977
The Muddy Creek Sewage Works was sampled to
study these diurnal variations. Virus concentrations
were significantly higher (as high as 40 pfu per IOO
gallons) between 2:00 AM and 4:00 AM in both the
December and July studies. An increase of almost an
order of magnitude is evident.
This increase remains unexplained at this time. Per-
haps the ultraviolet light from the sun may be a factor
in virus inactivation during daylight hours. No com-
parison is available for other plants since this was the
only plant we visited where staggered sampling was
conducted for all wastewater quality measurements.
The increased virus challenge to the plant between
LU
CC
a.
O
o
— 95% CONFIDENCE
— 98% CONFIDENCE
NEGATIVE CORRELATION
- -
QO tuO
HO U.CJ
O
O
O
o
o
FIGURE 13. CORRELATION COEFFICIENTS FOR VIRUS
IN-TREATED SEWAGE:
MUDDY CREEK: JULY
Total coliform, fecal coliform, chemical oxygen de-
mand, and total organic carbon had a positive correla-
tion coefficient. The lack of correlation for total sus-
pended solids and turbidity was surprising, since many
researchers believe that viruses can attach to particu-
lates and thus are more likely to survive various treat-
ment modes. However, the confidence level was suffi-
ciently low to make any conclusions highly speculative.
230
-------
VIRUSES AND ORGANICS
A similar analysis of wastewater quality data col-
lected in December is shown in Figure 14. Here, no
+0.5
o
o
z
o
o
o
POSITIVE CORRELATION
98% CONFIDENCE
95% CONFIDENCE
NEGATIVE CORRELATION
• TREATED WASTEWATER
S
cc
S
tc
- -;
OO UJO
KO U-0
o
8
o
o
FIGURE 14. CORRELATION COEFFICIENTS FOR VIRUS
IN TREATED WASTEWATER:
MUDDY CREEK
DECEMBER
correlation was found for total coliform and fecal
coliforms while total suspended solids and turbidity
had a positive correlation coefficient. Both COD and
TOC exhibited a positive correlation in December
as well as July.
The picture gets even more confusing when looking
at treated, disinfected wastewater. Figure 15 presents
(-•-TREATED AND DISINFECTED—*H
SEWAGE
POSITIVE CORRELATION
UJ
o
u.
UJ
o
o
o
UJ
IX
cc
o
o
*98% CONFIDENCE
— 95% CONFIDENCE
NEGATIVE CORRELATION
,o .o
FIGURE 15. CORRELATION COEFFICIENTS FOR VIRUS
IN TREATED, DISINFECTED WASTEWATER:
MUDDY CREEK
JULY
the calculated correlation coefficients for disinfected
wastewater at Muddy Creek in July. Here the values
are not only.mixed, but much too low for reasonable
significance.
Coliform Studies
A comparison of total coliforms and fecal coliforms
at each plant site visited during the two year study is
shown in Figure 16. Treated wastewater had measured
106-|
ESTES
OZONE
BERGEN
UV
TREATED
SEWAGE
TREATED
AND
DISINFEC--
TED WASTE-
WATER
MUDDY CREEK
CHLORINE
MUDDY
CREEK
OZONE
MARLBOROUGH
CHLORINE
TOTAL
COLIFORM
\
FECAL
COLIFORM
JFMAMJ JASON D
MONTH
FIGURE 16. MONTHLY TOTAL AND FECAL COLIFORM
CONCENTRATIONS
coliform concentrations between I04coliforms/ lOOml
(Marlborough) and almost 10'1 coliforms/100 ml
(Estes).
As with the virus data, there is no indication that
summer months had higher coliform concentrations
than other times of the year. In particular, the Muddy
Creek data showed a slight increase in total and fecal
coliform counts in December as compared to July.
The blackened-in fraction of the data is a measure
of the coliform count afterdisinfection. Again, Muddy
Creek data indicate ozone and chlorine were similar in
disinfection effectiveness. UV light was highly effective
for both total and fecal coliform reduction.
231
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
=>
u.
OL
z
g
O 20'
O
LINEAR REGRESSION MODEL
= 4.9-57x; r= -0.72
OZONE DOSAGE, mg/l
FIGURE 17. VIRUS CONCENTRATION AS A FUNCTION
OF OZONE DOSAGE
ESTES PARK, COLORADO
MAY, 1977
100,000-1
LINEAR REGRESSION MODEL
LOG Y = 6.7-x; r= -0.65'
10,000' '
O
;. 1,000-
-------
24.
EFFECTS OF CHLORINE, OZONE, AND ULTRAVIOLET LIGHT
ON NONVOLATILE ORGANICS IN WASTEWATER EFFLUENTS
Robert L. Jolley, * Norman E. Lee, * W. Wilson Pitt, Jr., * Mark S. Denton,
James E. Thompson, * Steven J. Hartmann, * and Charles I. Mashni**
*0ak Ridge National Laboratory, Oak Ridge, Tennessee 37830
**U.S. Environmental Protect/on Agency, Cincinnati, Ohio 45268
INTRODUCTION
The severity of water-borne diseases and disease
epidemics has been greatly reduced because of disin-
fection treatment of wastewater and potable waters.
During the last half century, the principal disinfectant
has been chlorine. It was only recently established that
chloro-organics are formed when chlorine is used to
disinfect wastewater effluents and potable waters (3,7,
8, 10, 12, 13). These findings are attributable to the
development of improved and more sensitive analyti-
cal methodologies in the last decade. Because of the
possible harmful effects of exposure to chlorinated
organics (9), there is considerable interest in the use of
possible alternatives to chlorine; currently, the prin-
cipal alternative is considered to be ozone (6, 9). The
chemistry of ozone reactions with specific organic
chemicals has been studied extensively. However, the
effects of ozonation on wastewater and wastewater
treatment plant effluents have been characterized
principally in general terms such as oxidation of or-
ganic carbon, biochemical oxygen demand (BOD),
and color (9). Relatively little is known about the
formation or possible formation of ozonation and
ozonolysis products during disinfection with ozone,
although the first products of ozone oxidation of
organic molecules are presumed to be oxygenated
materials that are water soluble and not highly vola-
tile (9). Additional information is needed before ozone
can be evaluated as an alternative to chlorine. Another
alternate disinfection method is ultraviolet (UV) light.
If this method becomes a viable alternative, irradiation
effects on organic constituents in wastewater must be
studied.
Rapid progress has been made recently in the anal-
ysis of volatile organic constituents in waters of en-
vironmental concern. However, the determination
and characterization of organic constituents of rela-
tively low volatility have not advanced as rapidly even
though the greatest portion of dissolved organic carbon
in wastewater and natural water consists of compounds
in this category (4, 11). This lack of amenability to
analytical techniques may be due to the inherently
greater complexity of the "nonvolatile" molecular
constituents (4). Nevertheless, a complete understand-
ing of possible environmental and health-related effects
which may be attributable to the nonvolatile organic
constituents can only be based on definitive data about
the chemical nature of the nonvolatile molecular con-
stituents.
This paper reports the current experimental results
of the effects of disinfection with chlorine, ozone, or
UV light on the nonvolatile organic constituents in
municipal wastewater treatment plant effluents. Re-
sults of mutagenic activity tests on effluent concen-
233
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
trates will be presented in a corollary paper in these
proceedings (5).
ANALYTICAL METHODS
The basic analytical steps used to examine the efflu-
ent samples are 1) concentration, 2) separation of con-
stituents, and 3) identification of constituents.
Concentration
The lower limit of detection for the high-resolution
liquid chromatographs is in the microgram range,
depending on the UV absorption of the individual
compound. Because the high-resolution chromato-
graphs are limited to 5 ml or less per sample, and the
concentrations of specific contaminants in effluent
samples may be lOu g/ liter or less, it is necessary to
concentrate wastewater effluents by factors S.3000
prior to analysis. The method of low-temperature
distillation followed by two freeze-drying steps, as
shown in Fig. 1, was selected in previous studies as the
most convenient and suitable concentration method
and provided adequate recovery of stable, nonvolatile
organic compounds (10). However, during this investi-
gation we have determined that lyophilization can be
used for the vacuum distillation step and effect a sav-
ings in both personnel time and effort.
WASTEWATER TREATMENT PLANT
EFFLUENT SAMPLE
10-100 LITERS
FILTER
10-100 LITERS
DISSOLVE IN ACETIC ACIO
DISSOLVE IN DILUTE ACETATE
BUFFER AND CENTRIFUGE
TO CHROMATOGRAPH
FIGURE 1. Procedure for concentrating effluent samples.
Separation of Constituents
Liquid chromatography has proved useful for sep-
aration and identification of numerous constituents
in wastewater effluents. In a previous high-pressure
liquid chromatography (HPLC) study of primary and
secondary effluents from municipal wastewater treat-
ment plants, approximately 60 organic compounds
were identified and 100 additional compounds were
characterized with respect to mass spectra and gas
chromatographic properties (10).
In that study, high-resolution anion exchange
chromatographs were demonstrated to have sensitiv-
ity in the microgram range and were capable of detect-
ing and quantifying many individual organic com-
pounds in concentrates of the complex aqueous efflu-
ent samples.
Both a preparative-scale and analytical-scale
chromatograph (Figure 2) are used to separate and
detect UV-absorbing compounds and/or cerate oxi-
dizable compounds (e.g., carbohydrates, organic
acids, phenols). The chromatographs consist primarily
of a heated, high-pressure ion exchange column; a
sample injection valve; a two-wavelength dual-beam
UV photometer; a cerate oxidative monitor; and a
strip-chart recorder. The ion exchange column for
each system is a 50cm length of type 316 seamless
stainless steel tubing (0.45 to 1.0cm ID), usually
packed with strongly basic anion exchange resin. A
0.05 to 5.0ml sample (the volume depending on the
inside diameter of the ion exchange column and the
nature of the sample) is applied to the column by a
six-port injection valve mounted as near to the top of
the column as possible to minimize peak broadening
(10).
The chromatograms are developed by eluting the
sample constituents from the resin column with an
ammonium acetate—acetic acid buffer solution (pH
4.4) whose acetate concentration gradually increases
from 0.015 to 6.0 M- The UV absorbances of the col-
umn effluent are measured at 254 and 280 nm using a
dual-beam flow-through photometer and are recorded
on a strip chart. Subsequently, the cerate oxidizability
of the same effluent is measured and recorded.
The use of the cerate oxidative monitor in series
with the UV photometer permits detection of addi-
tional compounds and a much greater sensitivity to
other organic constituents such as phenols. It also
provides another parameter useful in compound iden-
tification since many compounds that are U V absorb-
ing are oxidizable. In this system, cerium (IV) is re-
duced to fluorescent cerium (III) by oxidizable com-
pounds.
234
-------
VIRUSES AND ORGANICS
WASTE
AIR
PRESSURE
LUOROMETER ,i
.60 nm EXCITATION/'
SSOnm EMISSION !
FLUORESCENCE
RECORDER
FIGURE 2. Schematic of HPLC system.
Identification of Constituents
The preparation of samples for analysis, the separa-
tion of constituents, and the application of analytical
methods to separated fractions involve an integrated
and complex series of manipulations and investigative
techniques.
The preparative-scale liquid chromatograph system,
which is coupled to a fraction collector, is capable of
chromatographing 5ml of sample with a resolution
approaching that of the analytical column. The eluate
representing a component is collected and processed
through the following analytical procedure to identify
and characterize the isolated constituents.
Preparation of Fractions for Analyses. Eluted Frac-
tions corresponding to individual chromatographic
peaks from the ahion exchange separations are frozen
at -60° C and lyophilized for removal of the ammonium
acetate—acetic acid buffer. The samples are then dis-
solved in spectroscopic-grade methanol.
Ultraviolet Spectrometry. For each of the collected
fractions in methanol solution, UV spectra are ob-
tained from 320 to 210 nm on a Beckman DB-G re-
cording spectrophotometer and compared with UV
spectra of reference compounds obtained in the same
manner.
Gas Chromatography. Conversion of the nonvola-
tile constituents to volatile compounds is necessary
for analysis by gas chromatography. The principal
method of forming volatile derivatives of the nonvola-
tile compounds is silylation with bis(trimethylsilyl)-
trifluoroacetamide (10). Although silylation is the
. preferred derivatization method, constituents may
also be gas chromatographed as the methylated deriv-
ative. Methylation is accomplished with diazomethane
in ether solution (14). The samples are analyzed by gas
chromatography on one of several suitable instru-
ments available in this laboratory. A variety of detec-
tion modes may be used, including flame ionization,
electron capture, and electrolytic conductivity.
Mass Spectrometry. Mass spectrometry is usually
235
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
performed on aliquots of the trimethylsilyl (TMS)-
derivatized samples described above.
The TMS-derivatized sample is run routinely on a
Finnigan model-3000 gas chromatograph—mass
spectrometer. This provides data concerning the
molecular weights of the constituents and the number
of active hydrogen atoms per molecule. Comparison
of the fragmentation pattern with that of reference
standards is necessary for absolute identification.
DISINFECTION TREATMENT
OF EFFLUENT SAMPLES
Chlorination
All chlorinated effluent samples were disinfected at
the wastewater treatment plant during the routine
chlorination of effluent.
Ozonation
Wastewater effluent samples are treated with ozone
generated from a Wellsbach T-23 ozonator. A small
fraction of the oxygen feed gas is converted to ozone
which is allowed to react with the sample for a pre-
determined amount of time, depending on the total
dose desired. The reaction takes place in a round-
bottomed flask using a medium-porosity gas diffuser
to bubble the ozone and the oxygen carrier gas through
the magnetically stirred sample solution. Nonreacting
ozone is trapped in a 20% KI solution during the re-
action run; after the run, the headspace above the
effluent sample is purged with air into the trap to
determine the amount of unreacted ozone. A wet-test
meter records all gas flows. A schematic drawing of the
ozonation equipment is shown in Figure 3.
Ozone concentration in the gas is determined
by bubbling the gas through a 20% KI solution
and titrating with sodium thiosulfate in an acidic
medium with starch indicator.
A typical reaction run might have a gas concentra-
tion of 25 mg 03/1 gas and a gas flow of 0.25 1/min.
With a 4.5-1 sample and a reaction time of 8.64 min, a
resultant sample concentration of 12 mg 03/1 effluent
would be produced.
Ultraviolet-Light Irradiation
Samples of primary or secondary wastewater efflu-
ents are exposed to varying dose levels of UV irradi-
ation by forced flow through a UV sterilizer. A model
H-1.0 UV sterilizer with sight port (Pure Water Sys-
tems, New Jersey) is used with an LFE centrifugal
pump (Eastern Industries Div., 1 liter/min max). The
internal volume of the baffled-flow cell is 370 ml. A
quartz tube with a low-pressure mercury lamp, capable
of emitting UV irradiation at 254 nm, is positioned
longitudinally through the center of the cell. As the
sample flows through the cell, it is exposed to a con-
stant UV-irradiation dose level. The amount of energy
impinging on the sample is calculated by multiplying
the retention time in the cell by the measured optical
irradiation intensity.
FIGURE 3. Schematic drawing of ozonation equipment.
.Using the following equation,
V,
T = tR
where
c = volume of cell,
F = flow rate, and
tR = retention time,
one can calculate the retention time of a sample
for each flow rate. The tR value times the optical
irradiation ( M W/cm2) measured with an IL-700
Research Radiometer and Photometer gives the
amount of energy impinging on the sample
during this time ( p W»sec/cm2).
EXPERIMENTAL RESULTS
Ozonation
Initial efforts were directed toward establishing
parameters for ozonation of secondary effluent from
wastewater treatment plants. In the typical secondary
effluent from the Oak Ridge East Wastewater Treat-
ment Plant, an ozone dose of approximately 14mg of
ozone per liter of effluent is required for a .>99.99%
bacterial kill (Figure 4) as established by the standard
plate-count method for determination of bacteria (1).
Initial determinations of the bacterial quality of the
control and treated water samples were made using
this method; however, to facilitate comparison with
other disinfection studies, later determinations used
the total coliform method (membrane filter) (2).
236
-------
VIRUSES AND ORGANICS
5 .10 15 20
OZONE DOSE ( mg 0,/liter effluent)
FIGURE 4. Disinfection of secondary effluent with
ozone.
o
z
<
m
cr
O
A. CONTROL
D. 6.9 mg 03/li1er
10
20 25
TIME (hr)
30
35
40
FIGURE 5. Comparison of UV-absorbing constituents in
secondary effluent sample with samples treated with ozone.
Sample A is the unozonated control. Samples B, C, and 0
were ozonated at ozone doses of 0.7, 3.5, and 6.9 mg/liter
respectively. Chromatograms were determined for 1.0-ml
aliquots of 2500-fold concentrates using standard
UV-Analyzer chromatographic procedure. For each
chromatogram, the bottom tracing represents the UV
absorbance at 254 nm.
Large differences were detected between high-
resolution anion exchange chromatograms of concen-
trates of unozonated wastewater effluent and those
that had been ozonated for 3, 15, and 30-min periods
with an ozone gas concentration of 6 mg per liter of
oxygen gas. This treatment resulted in total doses of
0.7, 3.5, and 6.9 mg of ozone per liter of wastewater
effluent. The chromatograms of the UV-absorbing
and the cerate-oxidizable constituents in the control
sample and the ozonated samples are shown in Figures
5 and 6. The ozone dosages of these samples are labeled
A, B, C, and D on Figure 4 for comparative purposes
to show approximate levels of disinfection, although
bacterial counts were not made for these samples.
It is apparent that ozone initially reacts with organic
constituents that are probably of a higher molecular
weight and are not separated in this chromatographic
procedure, although some may be associated with the
first chromatographic peak. The higher-molecular-
weight constituents are broken down into smaller-
molecular weight fragments that are separated chrom-
Ill
LJ
rr
B. 0.7mg 03/liter
C. 3.5 mg 03/liter
D. 6.9 mg 03/liter
10
15 20
TIME(hr)
25
30
35
FIGURE 6. Comparison of oxidizable constituents in a
secondary effluent sample with samples treated with
ozone. Sample A is the unozonated control. Samples B,
C, and D were ozonated with ozone doses of 0.7, 3.5,
and 6.9 mg/liter respectively. Chromatograms were
determined for 1.0-ml aliquots of 2500-fold concentrates
using standard cerate-oxidimetry chromatographic
procedure.
237
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
atographically and detected as UV-absorbing and
oxidizable peaks. This is indicated by a decrease in size
of the first peak (Figures 5 and 6) with increasing
ozone dosage and an increase of poorly resolved peaks
in the middle of the chromatograms for the inter-
mediate dosages. As ozonation proceeds, these frag-
ments are, in turn, oxidized and decomposed as shown
in Figures 5 and 6 for the highest dosage.
The reaction conditions, chemical data, and disin-
fection data of the ozonation experiments used to
calculate the kill curve in Figure 4 are given in Table 1.
The chemical oxygen demand (COD) and BOD of the
samples trend downward with increasing dose. The
total organic carbon (TOC) apparently increases
slightly, then decreases with increasing dose. No con-
clusion can be drawn regarding total Kjeldahl nitro-
gen (TKN).
Figure 7 shows the chromatograms of UV-absorbing
and cerate oxidizable constituents in another ozona-
tion experiment with secondary effluent from the Oak
Ridge East Wastewater Treatment Plant. Two 6-liter
aliquots of Oak Ridge East Wastewater Plant secon-
dary effluent were ozonated in the laboratory to dos-
ages of 3.5 and 20.0 mg ozone per liter of wastewater.
Reaction conditions were: 25° C; ozone gas concentra-
tion, 19.4 mg ozone per liter gas (oxygen); 12 liter
glass reaction flask; magnetic stirrer; porous glass-frit
gas diffuser, ozone-oxygen gas mixture flow rate of
250 ml/min.
From the chromatograms of the UV-absorbing
constituents, it may be deduced that most of the
destruction of UV-absorbing compounds occurs rap-
idly and within the first 3.5 mg/ liter dosage of ozone.
With respect to the oxidizable constituents, there is an
apparent initial destruction of some compounds and
an increase in numbers and concentration of others
with increasing ozone dosage; for example, the in-
crease in numbers and size of peaks in the 5 to 15 hr
elution period. Chemical data for the control and
ozonated wastewater samples are given in Table 2.
Note that the disinfection data were not run. The usual
ozone dosage for disinfection is 6 to 10 mg ozone
per liter effluent; thus the chemical effects of ozonation
for disinfection may be considered intermediate be-
tween the two laboratory-ozonated samples. The
wastewater samples are representative of effluent from
an activated-sludge secondary treatment plant. The
wastewater was principally domestic, containing very
little industrial waste.
In another experiment, a 20 liter sample of Oak
Ridge West Wastewater Treatment Plant primary
effluent was ozonated in two 10.5 liter batches to a
dosage of 6.9mg ozone/liter effluent. The ozone-
oxygen gas flow rate was 0.1! liter/ min, and the total
reaction time was 29.25 min. The system was purged
with approximately 1 liter of air after each ozonation.
In batch one, the total amount of ozone applied was
72.4mg; 26.5mg ozone was trapped in'the K.I solution
following the reaction train, for an ozone utilization
efficiency of 63%. In the second batch, the total
amount of ozone applied was 72.4mg; 29.6mg ozone
was measured in the KI solution following the reaction
train, indicating an ozone utilization or a reaction
efficiency of 59%. Thus, an average of 61% of the
ozone dose reacted with the primary effluent in our
experimental equipment. The chromatograms of UV-
absorbing constituents in the untreated (control)
sample and those in the ozonated primary effluent
TABLE 1. OZONATION OF SECONDARY EFFLUENT FROM THE OAK RIDGE EAST
WASTEWATER TREATMENT PLANT: CHEMICAL AND DISINFECTION DATA (RUN 1)
(a)
Reaction
time
(min)
Control
5
15
30
60
120
Total ozone
dose
(mg/liter
of effluent)
0.0
0.7
2.1
4.3
8.5
17.1
Total (b)
Bacteria
count
(No./ml)
600,000
300,000
22,500
6,100
1,000
<1
pH
7.5
7.8
7.8
7.9
7.9
7.9
coo
24
26
20
19
20
16
BOD5
9
9
7
<5
<5
<5
TKN
mg/liter
20.0
18.6
18.6
19.0
23.5
18.2
TOC
7.9
7.6
8.3
8.6
7.5
7.2
fa) Room temperature (approximately 25°C); ozone-oxygen gas mixture at 0.08 liler/min containing 8.9-mg ozone per liter gas; 5-liter effluent
sparger with magnetic stirrer
(b) Standard plate-count method (1).
sample; glass-frit
238
-------
VIRUSES AND ORGANICS
OAK RIDGE SECONDARY WASTEWATER EFFLUENT
COMPARISON OF UV AND FLUORESCENCE CHROMA-
TOGRAMS. OF CONTROL AND OZONATED SAMPLES
BOTTOM UV ABSORBANCE AT 254 nm
f EXCITATION 254 nrr>
I EMISSION 360 nm
TOP FLUORESCENCE
CONTROL —
OZONflTED 3.5 mq Os/1
OZONATED 20 mq 03/A
FIGURE 7. Chromatograms of UV-absorbing constituents and cerate-oxidizable constituents
in secondary effluent ozonated with 0-, 3.5-, and 20-mg 03/liter effluent.
TABLE 2. O2ONATION OF THE SECONDARY EFFLUENT FROM THE OAK RIDGE EAST WASTEWATER TREATMENT
PLANT: CHEMICAL DATA (RUN 2).
(a)
Reaction
time
(min)
Control
4.3
24.8
Total ozone
dose
(mg ozone/liter
effluent)
0.0
3.5
20.0
PH
7.61
7.69
7.84
COD
30
28
24
BODg
mg/liter
6
7
5
TKN
18.2
17.4
17.2
TOC
10.8
9.8
10.5
(a) Room temperature (about 25°C): ozone-oxygen mixture at 250 ml/min containing 19.4-mg ozone/liter gas; 6-liter effluent sample; glass-frit
magnetic stirrer.
sparger with
sample are compared in Figure 8. It is evident that
many of the UV-absorbing constituents are destroyed
by ozonation. Note, however, the formation of one
major peak eluting to approximately seven hours.
Ultraviolet Irradiation
Since UV irradiation is being evaluated as a poten-
tially viable alternative disinfection process, studies
are underway to determine its treatment effects on the
nonvolatile organic fraction of wastewater effluents.
This study involves irradiating secondary wastewater
effluent at a fixed optical irradiance (jiW/cm2) at dif-
ferent flow rates (i.e., different retention times within
the UV-flow cell).
Total coliform [membrane-filter (MF) method (2)]
survival curves were determined for UV irradiation of
primary and secondary effluents from the Oak Ridge
Wastewater Treatment Plants (Figure 9). For the non-
irradiated primary effluent, the BOD5 was 80 mg/liter
and the pH was 7.5. The non-irradiated secondary
effluent BOD5 was 19 mg/land the pH was 7.1. The
survival curves were used to establish the parameters
for subsequent irradiation experiments.
The chemical effects of UV-irradiation disinfection
were evaluated in a series of HPLC separations.
Samples of unchlorinated secondary effluent were
collected in 23-liter polyethylene carboys at the Oak
Ridge East Wastewater Treatment Plant, which were
delivered immediately to the laboratory with no refrig-
eration. Ambient temperature during collection and
transportation was approximately 0°C. All samples
239
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
— OZONATED PRIMARY EFFLUENT ( 69 mg 0,/i )
PRIMARY EFFLUENT CONTROL
FIGURE 8. Chromatograms of UV-absorbing constituents in ozonated primary effluent (ozone
dosage, 6.9 mg/liter) compared with undisinfected control primary effluent. Samples were col-
lected from the Oak Ridge West Wastewater Treatment Plant and the ozonated sample was
ozonated in the laboratory- Ozonated sample, 0.77 ml of 790X concentrate. Unozonated sample,
0.77 ml of 700X concentrate.
107
10s
I05
m
3
I03
10'
.-PRIMARY EFFLUENT
^SECONDARY EFFLUENT
i I i I ill | L.
'0 10,000 20,000 30,000 40,000
ULTRAVIOLET IRRADIATION DOSE (MW-««c/cm2)
50,000
FIGURE 9. Disinfection of primary and secondary ef-
fluent with UV irradiation. Total coliform measurements
were made using the membrane-filter method (2).
were filtered through a Millipore filtering system
utilizing a prefilter and an Sum Millipore filter. Filtered
samples were then pumped through the UV sterilizer
at different flow rates to vary retention time within
the cell, and thus the irradiation level. A 20 liter aliquot
from each irradiation level was lyophilized, concen-
trated 2000X, and compared by HPLC.
Chemical and disinfection data for the control and
UV-irradiated samples are given in Table 3. It is in-
teresting to note the apparent decrease in TOC and
TKN with irradiation dosage increase. However, as
indicated by the chromatograms of UV-absorbing
compounds in the effluent concentrates (Figure 10),
UV irradiation seems to have little chemical effect on
the UV-absorbing constituents.
Comparison of Chlorination, Ozonation, and
UV-Irradiation Effects
Sixty liters of unchlorinated secondary effluent and
20 liters of chlorinated effluent (2 mg/1 chlorine dose;
0.5 mg/1 chlorine residual) were collected from the
Oak Ridge East Wastewater Treatment Plant. The
secondary effluent samples were filtered through
Whatman No. 2 filter paper. A 20 liter aliquot of the
unchlorinated effluent was lyophilized and concen-
trated as the control sample. A second 20 liter aliquot
was ozonated in two 10 liter batches at 25° C to an
ozone dose of S.Omg ozone per liter of effluent. Re-
action conditions were: 20.0mg 03 per liter of oxygen
240
-------
VIRUSES AND ORGANICS
TABLE 3. EFFLUENT DATA FOR UV-IRRADIATION EXPERIMENT.
Sample
1
2
3
4
5
Concentration
factor
1905
2000
1818
1800
1667
UV-irradiation
exposure
(UW«sec/cm>)
Control (0)
13,300
18,700
39,400
61,200
(a)
Total conform
(No.,100 ml)
200,000
<1
<1
40(b)
<1
PH
7.4
7.3
7.4
7.4
7.3
BOD5
15
<5
<5
<5
<5
COD
mg lit<
25
24
23
22
23
TOC
»r
20
20
19
16
14
TKN
4.0
3.7
3.6
3.3
3.3
(a) Membrane-filter method (2).
(b) This may represent possible contamination.
OAK RIDGE SECONDARY WASTEWATER EFFLUENT
CONCENTRATED 2000X COMPARISON OF CHEMICAL
EFFECTS RESULTING FROM EXPOSURE TO ULTRA-
VIOLET IRRADIATION
CONTROL SAMPLE
uv IRRADIATED li.300 ^W sec/cm2
UV IRRADIATED 61,200 ,jW sec/cm2
FIGURE 10. Comparison of chromatograms of UV-absorbing constituents in secondary waste-
water effluents irradiated with UV light at 0, 13,300, and 61,200uW»sec/cm2.
gas; gas flow rate, ISO ml/min; reaction, 24.3 min;
12 liter glass flask; and magnetic stirrer. The ozone
utilization efficiency in the reactor was 76%. After
ozonation the sample was concentrated for HPLC. A
third 20 liter aliquot was UV irradiated at 21,200
uW*sec/cm2 at 25°C and concentrated. The 20 liter
chlorinated aliquot was also concentrated by lyophili-
zation for HPLC.
The chemical analyses and disinfection results for
all samples are shown in Table 4. Chromatograms of
UV-absorbing constituents and cerate oxidizable
constituents are shown on Figures 11 and 12. Confirm-
ing previous effects, ozonation at disinfection con-
centrations destroys many nonvolatile organic con-
stituents and produces other nonvolatile organics,
especially oxidizable components. Ultraviolet irradi-
ation apparently had little effect on both nonvolatile
oxidizable organic constituents and the nonvolatile
UV-absorbing components. As also evidenced in
previous experiments, chlorination destroys some
nonvolatile UV-absorbing and oxidizable constitu-
ents and forms several other constituents.
Similarly, effects of disinfection on the primary
effluent were studied. Samples of both chlorinated
and unchlorinated primary effluent were collected in
23 liter carboys at the Oak Ridge West Wastewater
Treatment Plant and delivered to the laboratory with-
out refrigeration. Ambient temperature during collec-
tion and transportation was approximately 1°C. All
samples were filtered with a Millipore filtering system
using an 8 nm filter.
Twenty liters of filtered, unchlorinated primary
effluent were exposed at 27,500 uW»sec/cm2 of UV
irradiation. A second 20 liter sample was ozonated in
two 10.5 liter batches to a concentration of 6.9mg
ozone/liter effluent. The experimental details of this
241
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
TABLE 4. COMPARATIVE DISINFECTION OF A SECONDARY WASTEWATER EFFLUENT SAMPLE: CHEMICAL
AND DISINFECTION DATA
Disinfection
treatment
Control
Ozonated
Ultraviolet
Chlorinated
(a) Standard plate-count
Treatment
dosage or
residual
None
8 mg 03/l effluent
0.5 mg/l residual
method (13).
Bacteria (a>
count
(No./ml)
277,000
500
100
2,200
pH
7.3
7.5
7.4
7.3
COD
45
40
39
44
BOD
17
7
<5
<5
TKN
mg/liter
3.8
8.2
7.9
9.2
TOC
11.5
7.0
9.6
8.9
OAK RIDGE SECONDARY WASTEWATER EFFLUENT COMPAR-
ISON OF UV ABSORBING CONSTITUENTS AFTER EXPOSURE
FOR DISINFECTION TO (t I CONTROL SAMPLE, (2) UV
IRRADIATION.(3) CHLORINATION, ANO(4} OZONATtON
CONCENTRATED
CONTROL UNTREATED 2350X
U V IRRADIATED 21,200 jiW sec/cm2 235OX
CHLORINATED ' 0 5-mq/% RESIDUAL 2220X
OZONATED 3-mg Os/l 2220X
FIGURE 11. Comparison of chromatograms of UV-absorbing constituents in secondary waste-
water effluents disinfected with chlorine, ozone, and UV tight.
OAK RIOGE SECONDARY WASTEWATER EFFLUENT COMPAR-
ISON OF FLUORESCENT CONSTITUENTS AFTER EXPOSURE
FOR DISINFECTION TO (1) CONTROL SAMPLE, (2) UV
IRRADIATION, (3) CHLORINATION, AND (4 } OZONATION.
CONCENTRATED
CONTROL UNTREATED 2350 X
U.V. IRRADIATED 21,200 ^.W sec/cm* 2350X
CHLORINATED 0 5-n*/l RESIDUAL 2220X
OZONATED 3-mg 03/1 • 2220 X
2 13 14 15 16 17 19 19 20 21 22 23 24 25 26 27 28 29 30 31 32
FIGURE 12. Comparison of chromatograms of cerate-oxidizable constituents in secondary
effluent disinfected with chlorine, ozone, and UV light.
242
-------
VIRUSES AND ORGANICS
ozonation were reported in the section on ozonation.
Chemical and disinfection data for the four 20 liter
aliquots are given in Table 5. The'aliquots were con-
centrated and chromatographed. The chromato-
graphic comparison of the chemical effects of chlori-
nation, ozonation, and UV irradiation on the UV-
absorbing constituents in primary effluent revealed
significant chemical differences among the three disin-
fectants, as shown in Figures 8, 13, and 14. In these
figures, the chromatogram for the UV-absorbing
constituents in the control samples are offset above
the chromatograms of the disinfected samples for
comparison purposes. In Figure 13, the control sample
is compared with a primary sample chlorinated at the
Oak Ridge West Wastewater Treatment Plant with a
10 mg/1 chlorine dose to a chlorine residual of 0.9
mg/1. Some chromatographic peaks are apparently
destroyed by chlorination and other peaks are formed
during disinfection with chlorine. As previously re-
ported, some of these constituents formed during chlo-
rination are probably chloro-organics (3, 7, 8, 10, 12,
13). In Figure 8, the control sample is compared with
the primary sample after ozonation with 6.9mg ozone,
liter effluent. The high-oxidation potential of ozone
is clearly evident in the destruction of most of the
chromatographic peaks.
In Figure 14, the control sample is compared with
the primary sample after disinfection with 27,500
TABLE 5. COMPARATIVE DISINFECTION OF PRIMARY WASTEWATER EFFLUENT SAMPLE: CHEMICAL
AND DISINFECTION DATA.
Sample
Control
Chlorinated
(Cl residual, 0.9 mg/liter)
Ozonated
(6.9 mg ozone/liter)
Ultraviolet irradiated
(27,500uW»sec/cm2)
PH
7.6
7.7
7.9
7.6
BOD
14
19
12
18
COD
28
42
32
42
TOC
mg/liter
20.1
20.7
17.6
17.9
TKN
9.2
8.6
8.4
8.8
Total (a)
coliform
(No. 100 ml)
185,000
<1
15
<1
(a) Membrane-filter method (2).
"T
~T
1-
ID
CJ
UJ
O
z
<
00
rr
o
v>
CD
T
CHLORINATED PRIMARY EFFLUENT (0.9-mg/l chlorine residual)
PRIMARY EFFLUENT CONTROL
I 23456
TIME (hr)
FIGURE 13. Chromatograms of UV-absorbing constituents in chlorinated primary effluent
(chlorine residual, 0.9 ppm) compared with undisinfected control primary effluent. Samples were
collected from the Oak Ridge West Wastewater Treatment Plant. Chlorinated sample, 0.77 ml of
800X concentrate. Unchlorinated sample, 0.77 ml of 700X concentrate.
243
-------
CD
CC
O
CO
m
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
~T
UV-IRRADIATED PRIMARY EFFLUENT ( 27,500/uW sec/cm2)
PRIMARY EFFLUENT CONTROL
TIME (hr)
FIGURE 14. Chromatograms of UV-absorbing constituents in UV irradiated primary effluent
(27,500 W»sec/cm2) compared with undisinfected control primary effluent. Samples were
collected from the Oak Ridge West Wastewater Treatment Plant, and the disinfected sample was
UV irradiated in the laboratory. Ultraviolet irradiated sample, 0.77 ml of 660X concentrate. Unir-
radiated sample, 0.77 ml of 700X concentrate.
HW«sec/cm- of UV irradiation. The apparent absence
of chemical effect on UV-absorbing constituents is
clearly evident from the close agreement between the
two chromatograms. This lack of chemical effect by
UV irradiation at levels that disinfect the wastewater
may be highly significant and is being investigated
further.
CONCLUSIONS
Disinfection of wastewater effluents by either chlo-
rination or ozonation destroys many nonvolatile
organic constituents and produces other nonvolatile
compounds. Disinfection of wastewater effluents by
UV-light irradiation has relatively little effect on non-
volatile chromatographic constituents. Methods have
been developed for preparing effluent concentrates
and nonvolatile chromatographic constituents for
testing for mutagenic activity. The next phase of this
research program will emphasize characterization and
identification of compounds having mutagenic activity.
ACKNOWLEDGEMENTS
This research was sponsored by the U.S. Environ-
mental Protection Agency under Interagency Agree-
ments EPA-IAG-D7-01027 and DOE 40-593-76, and
the Division of Biomedical and Environmental Re-
search, U.S. Department of Energy, under contract
W-7405-eng-26 with the Union Carbide Corporation.
The authors wish to express appreciation to E. F.
Barth, U.S. Environmental Protection Agency, O. K.
Tallent and E. J. Arcuri for critically reviewing this
paper, to Vivian Jacobs for technical editing assistance,
and to Ogene Gentry for typing this manuscript. We
also wish to express our appreciation to J. Robinson,
Supervisor, Oak Ridge East and West Wastewater
Treatment Plants for assistance in obtaining effluent
samples.
REFERENCES
I. American Public Health Association. American Waterworks
Association, and Water Pollution Control Federation,
Standard Methods for the Examination of Water and
Wastewater, I4th ed. Washington, D.C. 1976. pp.908-913.
2. Ibid. pp. 928-935.
3. 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.
4. Christman, R. F. and R. A. Minear. 197I. "Organicsin Lakes".
Organic Compounds in Aquatic Environments. S. J. Faust
and J. V. Hunter (eds). Marcel Dekker, Inc., New York.
pp. H9-143.
5. Gumming, R. B., L. R. Lewis, R. L. Jolley, and C. I. Mashni.
"Mutagenic Activity of Nonvolatile Organics Derived
from Treated and Untreated Wastewater Effluents". This
conference.
6. Diaper, E. W.' J. 1975. "Ozone Treatment of Wastewater".
Ozone Chem. Techno/, pp. 55-74.
244
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
1. Glaze, W. H., J. E. Henderson. J. E. Bell and V. A. Wheeler.
1973. "Analysis of Organic Materials in Wastewater Efflu-
ents After Chlorination." J. Chromatogr. Sci. 11:580-584.
8. Jolley, R. L. October 1973. Chlorination Effects on Organic
Constituents in Effluents from Domestic Sanitary Sewage
Treatment />/onW.'ORNL/TM-4290. Oak Ridge National
Laboratory, p. 340.
9. Nebel, C., el a/. 1973 "Ozone Disinfection of Industrial Munici-
pal Secondary Effluents". J. Water Pollul. Control. Fed.
pp. 2493-2507.
10. Pitt, W. W., R. L. Jolley, and S. Katz. August 1974. Automated
Analysis of Individual Refractory Organics in Polluted
Water, EPA-66072-74-076. U. S. Environmental Protec-
tion Agnecy. p. 98.
II. Rebhun, M. and J. Manka. 1971. "Classification of Organics
in Secondary Effluents". Environ. Sci. Technol. 5:606-609.
12. Rook, J. J. 1974. "Formation of Haloforms during Chlorina-
tion of Natural Waters". WaterTreat. Exam.23(2):234-243.
13. Stevens. A. A., C. J. Slocum, D. R. Seegerand G. G. Robeck.
1978. "Chlorination of Organics in Drinking Water".
Water Chlorination: Environmental Impact and Health
Effects. Vol. I. R. L. Jolley (ed.). Ann Arbor Science Pub-
lishers, Inc., Ann Arbor, Mich.
14. Webb, R. G.,elal. 1973. Current Practices in CC/ MS Analysis
of Organics in Water. U.S. Environmental Protection
Agency, EPA R2-73-277.
245
-------
25.
MUTAGENIC ACTIVITY OF NONVOLATILE ORGANICS DERIVED FROM
TREATED AND UNTREATED WASTEWATER EFFLUENTS
R. B. Camming*, L. R. Lewis*, R. L. Jalley*, and C. I. Mashni**
*Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830
**U. S. Environmental Protection Agency, Cincinnati, Ohio 45268
INTRODUCTION
Wastewater disinfection is an important element in
the continuing battle to control infectious disease.
Adequate disinfection becomes more critical as popu-
lation size increases and as people live closer together.
Thus, on a global scale the effectiveness of water dis-
infection is a matter of vital concern.
Disinfection technologies, like other large-scale
technologies, have routine and potential costs as well
as benefits associated with them. It is now well estab-
lished (4) that a large number of chlorine containing
organics are produced during the disinfection of
wastewaters by chlorination. Other disinfection pro-
cesses also have the potential for producing modified
organics, but most of the compounds so produced
have yet to be identified. Some of the nonvolatile chlo-
rine containing organics produced in wastewater by
chlorination are relatively stable and are known to
show great biological activity. If released in sufficient
quantity, they have the potential for damaging the
environment and/or causing adverse effects on human
health.
Disinfection processes which have been proposed as
alternatives to chlorination also have the potential for
releasing modified organics of unknown biological
activity to the environment. That ozonation modifies
the spectrum of nonvolatile organics in wastewater
effluents has been clearly demonstrated (5), but for
other alternative disinfection procedures the situation
is less clear. The identity of biological consequences
of modified organics produced by these alternative
technologies are entirely unknown.
Mutagenicity is one biological endpoint which is
relevant to safety related evaluation of environmental
chemicals. A significant increase in the mutation rate
could be a direct chronic health hazard to humans or
could adversely affect animal or plant populations and
thus profoundly damage the environment. In addition
short-term tests for mutagenicity have been highly
correlated with mammalian carcinogenicity (1) and
thus may give a preliminary indication of danger of an
increase in human cancer.
We have performed several biological tests on
treated and untreated wastewater effluents. The pres-
ent paper is limited to tests for mutagenicity using
several indicator strains of enteric bacteria. While this
biological screening program is in its early stages, the
results to date already indicate that the problem of
mutagenicity of modified organics in disinfected
wastewater effluents is one that needs to be taken very
seriously in planning the disinfection strategy for the
future.
MATERIALS AND METHODS
Bacterial Mutagenicity Testing
The data reported here were obtained using standard
bacterial tester strains and protocols. Excherichia coli
strain WP2 was used according to procedures outlined
by Green and Muriel (3). This strain measures rever-
sion to tryptophan prototrophy by a base-pair substi-
tution mechanism and was used in a suspension test in
which the bacteria were exposed to the presumptive
mutagen for different periods of time and then plated
for survival and mutation frequency. We also used
four tester strains of Salmonella typhimurium devel-
oped by Dr. B. N. Ames and his collaborators which
test for reversion to histidine prototrophy. The strains
used were TA-1535 and TA-100 which are sensitive to
base-pair substitution mutagens, and TA-1538 and
TA-98 which detect mutagens which operate by a
frame-shift mechanism. These strains were used in a
plate test in which the bacteria are included with the
presumptive mutagen in a top agar overlay, and the
effect is scored as mutant colonies per plate. The pro-
cedures used are detailed by Ames et al. (2).
246
-------
VIRUSES AND ORGANICS
Sample Collection, Handling, and Concentration
Samples of wastewater effluents were collected,
quick frozen, and shipped to the Oak Ridge National
laboratory on solid CO2- Once in the laboratory the
samples are concentrated 1000-3000 fold by lyophili-
zation and subsequent rehydration in a suitable vol-
ume of distilled water. The details of the sample handl-
ing and concentration procedures are presented else-
where (5).
Sample Sources
The samples were obtained from several wastewater
treatment plants and sampling was always done before
and after the final disinfection step. The samples were
obtained from the plants listed in Table 1.
TABLE 1. SOURCES OF SAMPLES
RESULTS
Preliminary Experiments
The testing of complex mixtures for mutagenic
potential presents special problems and any separation
technique currently available carries with it the possi-
bility for the loss or inactivation of mutagenic com-
ponents. The most sensitive or useful tester strains for
nonvolatile organic materials cannot be determined
from negative results. We have, therefore,tested stan-
dard Oak Ridge primary and secondary effluent con-
centrates extensively in a series of preliminary experi-
ments using a variety of bacterial tester strains in an
effort to optimize the conditions for the detection of
mutagens among the nonvolatile organics of waste-
TESTED FOR MUTAGENIC ACTIVITY
Designation
ORPE
ORSE
CSE
EPSE
NBCSE
MCSE
LBSE
Name of Treatment
Plant
Oak Ridge West End
Wastewater Treatment Plant
Oak Ridge East End
Wastewater Treatment Plant
Taft Center USEPA Pilot
Wastewater Treatment Plant
Upper Thompson
Sanitation District
Northwest Bergen County
Wastewater Treatment Plant
Meander Creek
Wastewater Treatment Plant
Lyons Bend Wastewater
Treatment Plant
Location
Oak Ridge, TN
Oak Ridge, TN
Cincinnati, OH
Estes Park, CO
Waldwick, NJ
Mahoning County,
OH
Knoxville, TN
Type of
Treatment
Primary
Secondary
Secondary
Secondary
Secondary
Secondary
Secondary
Disinfection
Method
Chlorine
Chlorine
Ozone
Ozone
UV
Ozone
Chlorine
Type of
Wastewater
Primarily
residential
Primarily
residential
Residential
and Industrial
Residential,
relatively clean
Primarily
residential
Primarily
residential
Primarily
residential
High Pressure Liquid Chromatography
A sample which proved to be mutagenic as a crude
concentrate was subjected to HPLC as described
elsewhere (5). Fractions were collected and pooled to
make subsamples. In "one case the 220 fractions were
pooled to give 29 samples, most of which represented
individual UV absorbing peaks on the chromatogram
(Table 3, Figure 2). In the other case the 200
samples were pooled to give four samples (Table 3,
Figure 3). These samples were lyophilized, rehydrated
and tested for mutagenic activity with strain TA-1535.
Presentation of Data
For the purposes of this report tests which showed
2 times the activity of contemporary negative controls
or less are presented as "negative" ( —). Those which
showed from 2 to 5 times the activity of the negative
controls are called "weakly positive" (+). Those which
showed 5 or more times the activity of negative con-
trols are considered positive (+).
water effluents.
In these preliminary experiments we did obtain
positive results with the Salmonella strain TA-1535
in some samples. TA-100 gave only weak response to
samples which were positive for TA-1535, but we re-
tained it in many subsequent tests because for some
types of samples it is similar to, but more sensitive
than, TA-1535. The other Ames tester strains, TA-1538
and TA-98, which detect primarily frameshift muta-
gens, showed no response in our preliminary experi-
ments and we used them rarely in subsequent tests.
The E. coli strain WP2 which, like TA-1535, should
detect base-pair substitution mutations, also showed
no positive response in our preliminary experiments.
We did use WP2 in many experiments because in our
hands it had previously given results similar to TA-
1535 for other kinds of samples.
Testing of Field Samples
Table 2 summarizes results on field samples from 7
247
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
wastewater treatment plants. Most of the tests show
negative results. We obtained no evidence for muta-
genicity for either chlorinated or unchlorinated
samples from the Oak Ridge east end wastewater
treatment plant (ORSE) using a variety of tester
strains. This plant treats primarily residential waste-
water and carries out secondary treatment. In con-
trast, we obtained clear positive results for chlorinated
samples from the Oak Ridge West End Wastewater
Treatment Plant (ORPE) in strain TA-1535. There was
no positive effect for these samples in WP2 and only a
slight effect in TA-100. The positive effect of these
samples is substantial (see Figure 1), and it was re-
peated many times in samples taken at different times
of the year. The positive mutagenic effects in samples
from ORPE is dependent upon chlorination — un-
chlorinated samples showed no mutagenicity. The
Oak Ridge west end plant is a primary treatment plant
and this is the only primary treatment plant included
in this study (see Table 1).
A weak positive effect was seen with or without
chlorination in effluents from the Lyons Bend waste-
water treatment plant from Knoxville, Tennessee.
.This plant serves a primarily residential area. The
effect was seen only with TA-1535. Clearly more test-
ing needs to be done on Lyons Bend. To date, we have
looked at only a single sample from this plant.
The other plants tested (CSE, EPSE, NBCSE, and
MCSE) have to date given only negative results in the
strains used. This does not mean that there are no
mutagenic components in these treated or untreated
effluents, but only that with the strains and procedures
we have used to date, such components have not been
detected.
100 123 150 200
OF CONCENTRATE ADDED PER PLATE
Figure 1. Mutagenic activity tests of effluent concentrates
using Salmonella typhimurium strain TA-1535.
TABLE 2. RESULTS OF MUTAGENICITY TESTS ON SOME
Sample*
Source
ORSE
ORSE
ORSE
ORSE
ORSE
ORSE
ORSE
ORSE
ORSE
ORSE
ORPE
Disinfection
Treatment
none
none
none
none
chlorine
chlorine
chlorine
chlorine
ozone
U.V.
none
Treatment
Dose
—
—
—
—
0.5 ppm residual
0.2 ppm residual
0.5 ppm residual
0.8 ppm residual
8.0 mg/l
21,208uWsec/cm2
—
Sample
Date
2/7/77
9/27/77
10/5/77
2/8/78
8/2/77
9/27/77
10/5/77
1/31/78
10/5/77
10/5/77
9/27/77
WASTEWATER CONCENTRATES
Concentration
Factor
1000
1000
2350
2000
3100
1000
2200
2000
2200
2350
•1000
Tester
Strain
WP2
WP2
TA-1535
TA-100
WP2
WP2
WP2
TA-1535
TA-100
WP2
WP2
TA-1535
WP2
TA-1535
WP2
TA-1535
Mutagenic
Activity
-
-
-
—
-
-
-
-
-
-
-
-
-
~
-
-
248
-------
VIRUSES AND ORGANICS
TABLE 2. RESULTS OF MUTAGENICITY TESTS ON SOME WASTEWATER CONCENTRATES (Continued)
Sample*
Source
ORPE
ORPE
ORPE
ORPE
ORPE
ORPE
ORPE
CSE
CSE
CSE
CSE
ESPE
ESPE
NBCSE
NBCSE
MCSE
MCSE
LBSE
LBSE
Disinfection Treatment
Treatment Dose
none —
none —
chlorine 1.0 ppm residual
chlorine 0.9 ppm residual
chlorine 1.0 ppm residual
ozone 6.9 mg/l
U.V. 27,500fjWsec/cm2
none —
ozone 8.4 mg/l
none —
ozone 8.2 mg/l
none —
ozone ~6 mg/l
none —
U.V. 35,000 uWsec/cm2
none —
ozone ~6 mg/l
none —
chlorine 0.9 ppm residual
Sample
Date
12/12/77
5/19/78
9/27/77
12/12/77
5/19/78
12/12/77
12/12/77
9/12/77
9/12/77
5/29/78
5/24/78
10/19/78
10/19/78
4/21/78
4/21/78
6/15/78
6/15/78
5/3/78
5/3/78
Concentration
Factor
2000
2000
1000
2000
2000
2000
2000
1500
1600
2330
2150
2000
2000
3330
2700
2000
2000
2000
1667
Tester
Strain
TA-1535
TA-1535
WP2
TA-1535
WP2
TA-1535
TA-100
TA-1535
WP2
TA-1535
TA-100
TA-1535
WP2
TA-1535
TA-100
WP2
TA-1535
TA-100
TA-1535
TA-100
TA-1535
TA-100
WP2
TA-1535
TA-100
WP2
TA-1535
TA-100
WP2
TA-1535
TA-100
WP2
TA-1 535
TA-100
TA-1535
TA-100
TA-1 538
TA-98
TA-1535
TA-100
TA-1 538
TA-98
WP2
TA-1535
TA-100
WP2
TA-1535
TA-100
Mutagenic
Activity
—
-
-
+
-
+
+
+
-
-
-
-
-
—
—
-
"
-
-
—
-
~
-
~
—
-
—
—
-
-
-
-
—
~
-
-
-
-
-
-
-
-
-
+
-
- •
+
"
•See Table 1.
249
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
Testing of Fractions Separated by High Pressure
Liquid Chromatography
In an attempt to identify the mutagenic com-
pound(s) in ORPE chlorinated wastewater responsible
for the strongly positive response in TA-1535, we
subjected one such mutagenic concentrate to HPLC.
Samples were collected and pooled to give subsam-
ples of the original concentrate (see Materials and
tested for mutagenicity with TA-1535. Samples 2 and
3 were positive (see Table 3 and Figure 3). In both
cases the total mutagenicity recovered was far less
than present in the original concentrate.
DISCUSSION
The problem of measuring mutagenicity in treated
and untreated wastewater effluents is very complex.
Such effluents contain a large number of organic com-
TABLE 3. RESULTS OF TESTS OF CHROMATOGRAPHIC FRACTIONS OF CHLORINATED PRIMARY EFFLUENT FROM
OAK RIDGE WASTEWATER TREATMENT PLANT (MAY 19, 1978).
Sample
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Chromatographlc Separation No.
(Figure 2)
Fraction Peak (UV)
Number Present
1-7 +
8-10 +
11-15 +
16-20 +
21-23 +
24-28 +
29-32 +
33-37 +
38-43 +
44-47 +
48-51 +
52-58 +
59-62
63-68
69-73 +
74-77
78-80 +
81-88 +
89-94 +
95-107 +
108-114 +
115-124
125-133 +
134-145 +
146-155 +
156-170
171-185
186-200 «•
201-220
1
Result
TA 1535
-
+
+
-
-
—
-
+
-
-
+
-
-
-
-
-
—
+
-
-
-
-
-
_
+
+
-
+
-
Number
of Tests
1
2
2
1
1
1
1
2
1
1
2
1
1
1
1
1
1
2
1
1
1
1
1
1
2
2
1
2
1
Chromatographic Separation No. 2
(Figure 3)
Cut Fraction Result Number
Number Number TA 1535 of Tests
1 1-62 Negative 2
2 64-124 Positive 2
3 125-186 Positive 2
4 187-200 Negative 2
Total 36
Tests run with appropriate positive and negative controls.
8
Methods). Of the 29 samples so obtained from one
such fractionation, eight were weakly positive in a
mutational assay with TA-1535 (see Table 3 and
Figure 2). In another HPLC separation the fractions
were pooled to obtain four samples and these were
pounds, some of which are quite toxic to the test sys-
tems currently available. In addition, there is tremen-
dous variation in the content of effluents from one
treatment plant to another, making generalizations
very difficult. At present, studies must proceed on a
250
-------
VIRUSES AND ORGANICS
plant by plant basis. The primary goal of this type of
work is to evolve general principles for predicting the
production of mutagenic and carcinogenic materials
during the disinfection process, and to achieve an
understanding of the conditions under which such
materials are likely to be produced. This would lead
to the technology necessary for adequate disinfection
without increasing the burden of chronic disease in the
human population.
Mutagenicity of some chlorinated wastewater efflu-
ents has been demonstrated, and in spite of imperfec-
tions in the testing technology, the reality of the effect
is not in doubt. The data obtained in the HPLC separ-
ations suggest the possibility of complex interactions
between various components of the concentrate. Areas
of the chromatogram which are weakly mutagenic in
the first experiment show no effect when tested in
larger groupings in the second experiment (compare
Figures 2 and 3). The total mutagenicity of the frac-
tions from HPLC when tested separately is far less
than the mutagenicity seen in the original crude mix-
ture. This suggests that some of the mutagenic agents
present in the crude concentrates are either lost or
• inactivated in passage through this type of column.
This is an important, but not unexpected, finding with
implications for the planning of further studies. It is
important to identify specific mutagenic compounds
in treated wastewater effluents, but if they are lost or
inactivated by the best available separation tech-
niques, the job becomes extremely difficult.
The data presented here, though preliminary, indi-
cate a potentially serious problem which must be con-
sidered in the planning of the disinfection needs for
society in the future. Therefore, high priority should
be given to further research in this area. It is important
to produce the data which are necessary to make a
complete evaluation of the problem, and given the
complexity of the scientific questions which have
been raised, this will require a considerable period of
time.
CHROMATOGRAM OF OAK RIOGE CHLORINATEC
PRIMARY WASTEWATER EFFLUENT SHOWING
MUTAGENIC FRACTIONS (RUN I ) 1 INDICATES
WEAK MUTAGENICIT
Figure 2. Chromatogram of Oak Ridge chlorinated primary effluent showing mutagenic fractions
(Run 1). ±indicates weak mutagenicity.
NEGATIVE ACTI
POSITIVE ACT
CHROMATOGRAM OF OAK RIOGE
CHLORINATED PRIMARY WASTEWATER
EFFLUENT SHOWING MUTAGENIC
ACTIVITY FRACTIONS (RUN 2)
NEGATIVE ACTIVITY
Figure 3. Chromatogram of Oak Ridge chlorinated primary effluent showing mutagenic fractions
(Run 2).
251
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
Estimating the potential for increased risk to adverse "Carcinogens are Mutagens". A simple test combining liver
human health effects is the ultimate goal of research of £^1" Scl^USA 6^3 128-". 32
the type described here. This implies a quantitative 2 Ames B N j McCann and E Yamasakj ,9?5 ^^ for
estimate which could be used in balancing the risks Detecting Carcinogens and Mutagens with the Salmonella/
with the benefits for the various disinfection technolo- Mammalian-Microsome Mutagenicity Test". Mutation
gies. However, we are now at the stage of trying to gain Res' 3I-347-364-
insights into potential risks on a purely qualitative 3. Green, M.H.L. and W.J. Muriel. 1976. Mutagen testing using
level. This is all that the current state of the scientific TRP+ reversion in EscheriMa coli Mutation Res' 38:3-32'
art will allow. We are still a long way from being able 4 Jolley' R L- October 1973. "Chlorination Effects on Organic
.,- , , , , • , , • Hi Constituents in Effluents from Domestic Sanitary Sewage
to quantify the problems Which are becoming all too Treatment Plants". ORNL/TM-4290. Oak Ridge National
apparent on a purely descriptive level. Laboratory, p. 340.
5. Jolley, R. L., N. E. Lee, W. W. Pitt, M. S. DentonJ. E.Thompson,
REFERENCES S. J. Hartmann, and C. 1. Mashni. "Effects of Chlorine,
Ozone and Ultraviolet Light on Nonvolatile Organics in
1. Ames, B. N., W. E. Durston, E. Yamasaki, and F. D. Lee. 1972. Wastewater Effluents" This symposium.
252
-------
26.
PANEL DISCUSSION OF VIRUS, NON-VOLATILE ORGANICS,
AND MUTAGENIC EFFECTS
Participants in Viruses and Organics Roundtable
Discussion
Edward J. Opatken, Moderator
U.S. EPA-Cincinnati
1. Robert Fluegge
Carborundum Company
2. Robert Jolley
Oak Ridge National Laboratory
3. Robert Gumming
Oak Ridge National Laboratory
DR. LONGLEY: I have a comment first of all.
In the last break I talked to Al Venosa and ap-
parently I did not define myself fiilly enough
when I had talked earlier. We have today a
situation where the EPA and other groups,
ASTM as an example, are putting out standards.
There is no coordination that is very effective
between these groups. On the biological side, I
have to give a lot of credit to Ed Geldreich for
the efforts that he has made to make the EPA
standard methods have some agreement, but in
general, in residual analysis we have no strong
central guilding light to put us all down the same
path and this is creating a lot of problems.
This, I believe and a number of other people I
have talked to believe, is something that we have
to address soon. It is a problem and if we do
not establish coordination, it is just simply going
to hurt the field of disinfection in years to come.
Are we open for questions to the panel now?
MR. OPATKEN: Yes, we are.
DR. LONGLEY: Thank you. Mr. Fluegge, I
have two questions to ask you on the work that
you presented. In particular, you mentioned that
you are using pre-filters. I understand from what
you said that the pre-filters were not used in all
instances, is that correct?
MR. FLUEGGE: That is correct. Whenever the
total suspended solids were sufficiently high that
the on-site field engineer due to his experience
would determine that the flow rate would drop
too far prior to collecting the full 100 gallons,
then we would add the clarifier filter in front of
the addition of all chemicals.
DR. LONGLEY: When you did that, did you
then attempt to elute virus in the solids that you
picked up on the pre-filter?
MR. FLUEGGE: We did in each case. We did
find virus in the clarifiers as well as in the ad-
sorber, as I mentioned in the presentation. The
number of viruses collected on the clarifier are
usually smaller than the number of viruses that
are picked up on the adsorbing filter. This is not
always the case but most of the time.
DR. LONGLEY: In San Antonio at the Rilling
Road plant, which is a 90 mgd plant, we saw the
same phenomenon that you saw. One possible
mechanism leading to that is the fact that the
waters which are entering that plant prior to that
large peak are lower in organic and other
materials which would be adsorbed on the floe,
in the case of the San Antonio plant which is an
extended aeration plant. We are hypothesizing
that one mechanism might be that we are simply
beginning to elute some of the virus off of the
floe that have been adsorbed on the floe in the
plant and we are seeing them in the effluent. I
do not know if that explanation would be ap-
plicable to what you saw . . .
MR. FLUEGGE: You are thinking that the floe
ends up on the clarifier . . .
DR. LONGLEY: I am saying that the floe in
the plant has adsorbed virus and, as the load
coming into that chamber is lower in organics
and other materials which will adsorb to that
floe, there is less competition for sites. We are
beginning to elute for a period of time some of
the virus in the plant off of the standard aeration.
MR. FLUEGGE: Usually the pH has to change
a quite drastic amount.
DR. LONGLEY: That is another factor, too.
MR. VENOSA: Karl, just a quick question on
your San Antonio effluent. Did you find
relatively the same number of viruses as Mr.
Fluegge found?
DR. LONGLEY: I was not looking at the
numbers so closely on his graph. I just saw his
peak. You were isolating animal virus. We were
253
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
looking at phage, so the numbers would not
necessarily be comparable.
Just one other comment. Your animal virus
numbers are much lower, if I am not mistaken,
than the viruses reported sometime ago in Gerry
Berg's book. Am I correct on that?
MR. FLUEGGE: I believe that is correct, yes.
DR. LONGLEY: They are like maybe three or-
ders of magnitude in fact.
DR. JOHNSON: Dr. Gumming, your results
are really impressive. The question is, what kinds
of compounds does this particular Salmonella
mutagenesis test respond to? I know that there
has been a comment made about several of the
Ames Salmonella strains not responding very well
to chlorinated compounds. That is, you could
have a chlorinated compound which would be
mutagenic in animals but not sensitive in the
Salmonella test, which is I guess the standard
test. What kinds of compounds would be expec-
ted to provide the kind of transformations that
you see in your testing?
DR. GUMMING: Well, the strain TA-1535 pre-
sumably detects base pair substitutions mutations.
The Ames base pair substitution strains are
missense mutants and the precise molecular basis
is not as well established as it is for some of the
E. coli strains.
TA-100 has the same mutation as TA-1535 but
has an addition of an episome (an R factor),
which makes it more sensitive to most com-
pounds and very much less sensitive to what we
are seeing in the sewage. That is very similar to a
result observed by others with some relatively
small chlorinated compounds. That is, TA-1535
being more sensitive than TA-100. Specifically,
certain nitrosamines, which are not mutagenic
without metabolic activation, become mutagenic if
they are chlorinated. Those compounds respond
more strongly to TA-15 34 than they do to TA-
100. So, whatever we have is like that. That does
not give any indication as to what the nature of
the beast is, but this does establish a precedent
for certain chlorinated compounds being more
strongly effective in the particular strain we did
see the activity in.
DR. JOHNSON: Can you explain base pair
substitution a little bit? In other words, are you
talking about the nucleic acid natural bases being
substituted, like the classic fluoro-uracil type
compound?
DR. GUMMING: Right. Incidentally, since you
mentioned it, we have looked at chloro-uracil and
see nothing in any of the Ames strains with it,
but we do see a positive effect in E. coli.
Base pair substitutions are the change in the
genetic code by substituting one nucleic acid base
for a wrong one. It's a mis-substitution, and in
general I do not want to get too technical on the
mutagenic theory, but there are two types.
One, so called missense mutation, simply in-
serts a wrong amino acid in a protein, so you have
a defective protein as the end result. The other,
called nonsense mutation, inserts a signal which
terminates the protein, and the E. coli strain we
use is of that type. The Ames strains are mis-
sense; they just give a defective protein.
DR. JOHNSON: Very exciting!
MR. DeSTEFANO: I have a question for Dr.
Gumming. I brought {his up yesterday and had I
known you would be here, I think I would have
reseerved the question. Do you think it would be
possible that, with ultraviolet light or some other
form of disinfection, you could inhibit the ability
of E. coli to ferment lactose and therefore not
appear on the MPN test?
DR. GUMMING: Yes, I presume it is possible.
Certainly one thing that we have to be concerned
about is /that when you disinfect you have a sur-
viving fraction, and the surviving fraction may be
different biologically from what you started out
with. It can be different for a lot of reasons but
mutagenesis is certainly a factor.
MR. BLAINE SEVERIN: I have a question for
Mr. Fluegge. Are we to understand from your in-
formation that none of the three disinfectants are
really killing the virus present and that we cannot
claim that one is better than the other? I have
heard claims that ozone kills more virus than
chlorine and that UV will kill more virus than
chlorine and in this real situation disregarding the
question about solids elutriation and the primary
filter, is there a difference? Is one better than the
other?
MR. FLUEGGE: All three disinfectants ended
up with virus in the effluent to the stream. There
was not one other than chlorine at one plant
which had a low input and we did not pick up
any in the output, but we did find an output in
a chlorine disinfection scheme .at Muddy Creek.
At the beginning of the program we tried to
254
-------
VIRUSES AND ORGANICS
design the experiment to look at one versus the
other and it is not clear to me that we could
pick one kind of disinfectant as being better than
the other, especially if you compared ozone and
chlorine at Muddy Creek. UV looked good at
NW Bergen County. On the other hand, I had
no comparison with chlorine as a base or with
ozone.
DR. BILL CHAPPELL, University of Colorado:
We are doing some work that is similar to the
Oak Ridge work in many ways but in the context
of potable reuse of wastewater, which as some of
you may know Denver is interested in. We are
looking at the output from the Denver metro
plant. The chemistry is being carried out by
Harold Walton of the Chemistry Department at
CU and the Ames testing by Dr. Lee Spalbender
of the American Medical Center. I would like to
relate a few of our own results. We are in a
much earlier stage, but we have seen strong
mutagenic activity in the chlorinated effluent
from the metro sewage plant. It has involved
both base substitution and frame shift mutations,
but there seems to be a strong time variation
because from one time to the next we will see
very strong mutagenic activity and then another
time, essentially no activity at all.
MR. VENOSA: Bill, is the metro sewage plant
a secondary or a primary plant?
DR. CHAPPELL: Secondary.
MR. VENOSA: I think it should be empha-
sized here that the study Bob Gumming reported
positive results on was on primary effluent.
DR. GUMMING: The positive effects were
from primaries and they were dependent upon
chlorination. I do not know whether your results
were also dependent upon chlorination.
DR. CHAPPELL: Yes, they were.
MR. WARRINER: A couple of methodology
questions for the Carborundum procedure. I
gather you passaged 50% of your isolations and,
if that is correct, I would like to know what
fraction of that 50% was confirmed for anything?
MR. FLUEGGE: All of the plaques were
passed from one cell culture to another. The
limitation of the. LBM serum pool was part of
the reason for not identifying all of them as well
as the time associated with that.
MR. WARRINER: Did you find false plaques
in the first time around?
MR. FLUEGGE: False plaques? Yes.
MR. WARRINER: What percentage?
MR. FLUEGGE: I could not say. Dr. Metcalf
would be more able to give a percentage. I do
not end up with that particular percent number
in Dr. Metcalf's reports. I get the positives
coming through.
MR. WARRINER: Could I ask if you do cali-
bration runs using one or another of the viruses
you identified or something related to them?
MR. FLUEGGE: We did a number of seed
runs, as people will call them, putting poliovirus
1 seed into waters and then recovering them.
There are a lot of difficulties in trying to under-
stand what a seed run truly means.
When you put high liters of virus, like 106 per
ml or 107 per ml into 50 gallons and then start
concentrating through the filters, elute, and
analyze and determine how many viruses are
picked up after you've pulled them out, that
takes care of the numerator quite well. You have
to worry about what is in the denominator. How
many did you really put into that water? Now,
you have 1 ml of liquid and you are putting it
into this 50 gallons. You are going to have to
mix that 50 gallons. We try very hard to get a
mono-dispensed set of viruses throughout that 50
gallons. The question is: are there other
chemicals, are there materials within that 50
gallons that can kill those viruses you just put
in? If you just simply said, I had 106 per ml and
now I have this many that I got back, then you
can end up with a very conservative percentage
recovery. With that kind of calculation, we get
numbers coming out somewhere between 5 and
30% of the number put in.
Then, it can go the other way. If you put it
into, say, 5 gallons, and you let that seed sit
around for the 2 to 3 hours that you had spent
collecting your sample, and then you take a grab
sample out, you could get 130% recoveries,
which would indicate to me that we had better
start running. So, it is very difficult to get that
denominator, large or small. If it is distilled
water or distilled water with a little salt solution
in it, you are in good shape. It is when you get
out into the real world, where you may have all
kinds of chemicals that may cause your virus to
clump together or cause them to die or whatever,
where the real rub comes.
MR. WARRINER: You say you do not have a
255
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
recovery factor we can put on the numbers that
you had in your graph.
MR. FLUEGGE: One of the reasons for
designing the experiment the way we did, namely,
before and after the adding of disinfectant, was
that whatever was in the water during that 2
hour period, was true for both cases, those that
were not having the disinfectant added and those
that did have it added. Thus, we could compare
the two. Even though you cannot go to an ab-
solute basis, you can compare the number of
viruses that entered say the UV light versus the
number of .viruses that left the UV light sampling
the chlorine.
MR. ELLNER, Ultraviolet Purification Systems:
I would like to first point out some very serious
inadequacies in the reportage here. In one of the
first discussions, the ultraviolet source was de-.
scribed as a 100 watt lamp with 30% of its out-
put in the ultraviolet range. Now that would be
akin to somebody talking about visible light and
saying I used 100 watt lightbulb with 70% of its
output in visible light. You could have 100 watt
fluorescent lightbulb one inch long with an
extremely high light intensity per inch of length.
You can have 100 watt neon light tube. You
could have 100 watt quartz iodine light tube and
the point is that in describing ultraviolet lights
the method of describing the output would be in
microwatts per square centimeter at a given
distance, because there is no way that anybody
reviewing this information can have any idea as
to the actual output per unit length of the lamp.
Also, the length of the lamp is not given.
We see variations with what is apparently the
same equipment. One reporter claims 180,000
microwatt seconds per square centimeter while
another reporter .talks about 21,000 microwatt
seconds per square centimeter. There is no infor-
mation regarding the ultraviolet transmission, as
measured in a spectrophotometer, of the liquid
being treated, which can have a complete and
total effect on the results. If one of the effluents
had a 90% transmission and the other effluent
had a 10% transmission, this is going to have a
tremendously important and meaningful effect on
what the results are. If we are trying to discuss
quantitative results here, I suggest that before
items are published, a little more serious thought
be given to the descriptions and the use of quan-
titative terms, primarily dosage. Especially, the
dosage information is totally meaningless unless
you relate it to the transmission. The dosage
capability of the system is directly related to the
transmission of the particular liquid. You can
take that device and put a material with zero
color and zero turbidity, let's call it kerosene for
this example . . .
MR. OPATKEN: Sid, this session is devoted to
the questions in regard to the organics and the
viruses and so forth ...
MR. ELLNER: Well, there were statements
made regarding virus and the people here are ob-
viously going to come to conclusions regarding
the ability of ultraviolet, not a given wavelength,
but of the technique to either handle or not
handle virus. Unless you have meaningful . . .
MR. OPATKEN: True, but if you have a
question, you can address it to the board.
MR. ELLNER: My question is this. How did
you arrive at your dose? What was the ultraviolet
transmission? What were the flow rates and what
were the lengths of the lamp?
MR. FLUEGGE: Well, first of all the plant
was described yesterday by Karl Scheible of
Hydroscience in quite a bit more detail than I
tried to cover today. Regarding the number that I
calculated, I just simply took the wattage output
at 30 watts for each bulb. I took into account
the length of the bulb as being five feet and the
diameter of the bulb as being 0.9 inches. I then
calculated the number of photons leaving that
bulb a quarter of an inch away. Now, what
happens is that only about 120 of the 400 lamps
take up the outer part of this whole array. That
means that there are 280 bulbs that are all inside,
and so all of those bulbs are simply radiating to
other lightbulbs. That photon ends up going into
another lightbulb and then re-emitting. I had no
way of determining how many photons are
bouncing around inside of this thing. I do not
see how anybody can really get a handle on how
many photons the viruses are being exposed to
without some sort of chemical actinometry or
chemical calibration. I did not have any way of
doing that in the output of a regular plant.
MR. ELLNER: And that is why I am saying
you- should not report what the dosage was then.
DR. JOLLEY: Mr. Chairman, before you
adjourn, I would like to make a comment, too. I
think I would like to focus back again on the
256
-------
VIRUSES AND ORGANICS
significance of the mutagenic activity in the could not assay in the field to determine the
chlorinated primary effluent. It was somewhat difference.
dismissed by comment that it is just primary. MR. BERMAN: The second question to go
There are many major cities which currently along with this. Did you have any trouble at all
discharge their primary into waterways. For getting samples back to the laboratory? I am
example, St. Louis discharges a considerable curious as to how you shipped it, because this a
amount. potential etiologic agent. Did you just have it
Furthermore, the quality of the primary sewage frozen and then ship it without any special
that we are dealing with in Oak Ridge, for warnmg label or what have you?
example, is equivalent to many secondaries. It is MR. FLUEGGE: We label each of our samples
a very clean, clean sewage. The fundamental as they should be iabeied. We follow the Public
question to me is whether or not indeed there is Health Laws> the Federal Laws and the carrier
something in activated sludge that does reduce law requirements as we ship. It is not easy. I
this mutagenic activity, or whether or not it is a does take a iot of paperwork. You do have to
concentration phenomenon and we are just not fm out all kinds of forms. You have to label it
seeing it and it is still there. correctly and your biggest problem is when you
Then I would like to make a statement relative conduct a seed experiment. How do you get that
to THM's which I missed. The question came up to the field and how do you get it back home
yesterday or the day before concerning There is where your concentrations are known in
production of THM's in these disinfection your sample You put them there
processes.
We measured THM's with respect to our UV
irradiation and found essentially no difference
between the influent and the effluent. With
respect to ozonation at neutral pH we found little.
difference. Some would say that ozone would
convert the chloride into chlorine because of its
potential. This is a very slow reaction. Further-
more, there is a lot of ammonia there. Now
whether or not there would be any THM's
formed in a system without ammonia I do not
know.
With respect to chlorination of sewage, the
same concentrations of THM in influent and
effluent were measured in effluents that had
ammonia above about 1 ppm. In nitrified sewage
there was considerable amount of chloroform
produced on chlorination, and so there is that
potential when you remove the ammonia to pro-
duce THM's on chlorination.
MR. DONALD BERMAN: For Mr. Fluegge. I
am curious . . . was there any die-off during
transport of the sample from the field to the
laboratory?
MR. FLUEGGE: We have looked at this on
other studies, not in this particular study, but we
have looked at survival of virus at dry ice
temperatures. The concentrate is put into a vial
and put at -78° C. At that temperature the die-
off has been very low in our previous studies. I
cannot answer the question in this one because I
257
-------
SECTION 7. PLANNING AND IMPLEMENTATION
27.
PLANNING DECISIONS IN SELECTING WASTEWATER EFFLUENT DISINFECTION
FOR THE OLENTANGY ENVIRONMENTAL CONTROL CENTER
Catherine F. Grissom
EIS Preparation Section, U.S. Environmental Protection Agency, Region V, Chicago, Illinois
(Presented by Mr. Michael MacMullen)
ABSTRACT
Region V of the U. S. Environmental Protection Agency prepared an
Environmental Impact Statement (EIS)for the Olentangy Environmental
Control Center in Delaware County, Ohio. The new interceptors and
treatment plant would provide tertiary treatment for 5,700 m3/d (1.5 mgd)
and later 11,000 m3/d (3 mgd) of wastewater. Discharge will be to a seg-
ment of the Olentangy River noteworthy for its biological quality and
natural character. Mitigation was necessary to protect the biota from
potential adverse chlorine impacts at low flow. Ozone was selected as the
most cost-effective disinfection alternative. The EIS was a successful
tool for selecting a new technology for wastewater disinfection.
INTRODUCTION AND OBJECTIVES
The National Environmental Policy Act of 1969,
known as NEPA (1), established the Environmental
Impact Statement, or EIS, process. Much has been
stated in documents, books, and legal decisions about
what an EIS should be and what it should do (2). The
fine points of this remain a continuing debate, as
shown in the proposed Council of Environmental
Quality's (CEQ) regulations, which apply to the ElS's
of .all Federal agencies (3). At its best, we believe that
the EIS process gives the Federal agency the oppor-
tunity to examine the environmental implications
of its actions, and expands the agency's thinking about
its standard solutions to a given problem.
We have learned from preparing a number of EIS's
for EPA's sewage treatment projects that the planning
constraints are seldom ideal and that the solutions
one can choose are not perfect. These problems lead
to a need for some creative thinking on how to mitigate
the potential adverse impacts of a project, to provide
greater benefits for clean water and the total environ-
ment. In addition to the creative thinking mandated
by NEPA, the Construction Grants Program for
wastewater treatment facilities encourages the use of
new and advanced technology in projects which we
fund (4).
We encountered a number ot problems requiring
creative solutions when preparing the EIS for new
wastewater collection and treatment facilities at the
Olentangy Environmental Control Center, in Delaware
County, Ohio (5). The successful application of the
EIS process is best illustrated in the series of decisions
which lead to selecting innovative effluent disinfection.
Southern Delaware County is an attractive, semi-
rural area located just north of Columbus, Ohio (see
Figure 1). Our study area is divided by three major
streams: the Scioto River, the Olentangy River, and.
258
-------
PLANNING AND IMPLEMENTATION
DELAWARE COUNTY
in 111 in t minium
FRANKLIN COUNTY
MM Planning area
Proposed sewage
treatment plant
Figure 1. Regional Context for the Olentangy Environmental Control Center.
Alum Creek. These streams have been impounded at
several points to provide reservoirs for water supply.
flood control, and recreation. Local soil conditions
severely limit the effective use of septic systems, while
package treatment plants have been resisted by local
authorities because of their poor maintenance history.
A proposal for Delaware County's central sewer
service was presented to EPA in the 1974 Facilities
Plan1. . A 5,700 m'/d (1.5 mgd) design size, two-stage
activated sludge plant, with tertiary sand filters, phos-
phorus removal, and chlorination was to be built .on
the Olentangy River immediately above the county
Burgess & Niple, Ltd. 1974. The sanitary sewerage facilities plan for south-
central Delaware County, Ohio. Burgess & Niple, Ltd. Columbus, Ohio.
Letter from Harlan D. Hirt. U.S. Environmental Protection Agency, Region
V. March 28, 1975. Notice of intent to prepare an environmental impact state-
ment, Delaware County, Ohio, and environmental impact appraisal.
line. Known as the Olentangy Environmental Control
Center, it was proposed to be expanded to 11,000
m'/d (3 mgd) design size by the end of the 20-year
planning period, to serve an expanded sewered area
within south-central Delaware County.
EPA decided to prepare an EIS on this project in
March, 1975. Key issues included the secondary im-
pacts of the population growth to be induced by the
plant, the facility's impacts to an adjacent regional
park, its potential impacts to archaeological resources,
its impacts to the Olentangy River, and public contro-
versy2.
I he Olentangy Environmental Control Center
presented a complex problem, for all of these rea'sons.
In this paper, we will focus on its potential instream
impacts and how we resolved to mitigate them.
259
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
MATERIALS AND METHODS
The Delaware County EIS was prepared by Region
V of EPA, with the assistance of Enviro Control, Inc.,
of Rockville, Maryland
RESULTS AND DISCUSSION
The scope of the EIS included an examination of
existing stream quality and of the environmental im-
pacts of centralized wastewater treatment. Additional
information and public input about the impacts of the
proposed wastewater treatment facilities arose during
the EIS preparation period. The Olentangy River was
designated a State Scenic River and provided both a
valuable natural habitat and a good recreational
resource. Although some impacts to it were noted
from the existing upstream Delaware County wastewater
treatment plant, (which is in the process of planning
new improvements), the stream segment above the
proposed effluent discharge location had recovered
from the effects of the upstream effluent. Dissolved
oxygen levels were good (5). Aquatic insects in this
stretch of a basin-wide survey showed a high diversity,
with about 70% classified as "pollution sensitive" (5).
A great variety of mollusks have been found in the
Olentangy River, including four state endangered
species (5). A fishing census by the U.S. Fish and
Wildlife Service demonstrates both the biological
and recreational quality of the Olentangy as a warm-
water fishery (5). The Olentangy River has regional
significance because of its biological diversity, and
because of the limited number of relatively natural
stream segments remaining in central Ohio. Reser-
voirs, channelization, and pollution have gradually
eliminated many high quality areas of running water.
Our first major decision in the EIS was the location
of the treatment facilities (5). Were there any satis-
factory alternatives to the one proposed in the Facili-
ties Plan? Examination' showed that the other two
streams in the service area were not as suitable for
effluent discharge. The Scioto River has downstream
reservoirs for domestic water supply, while Alum
Creek has an even lower average streamflow than the
Olentangy. Using other sites along the Olentangy
would have led to considerably higher construction
and operation and maintenance costs. In addition,
many of the. Olentangy sites would use similar outfall
locations. Moving the outfall downstream, below the
end of the State Scenic River Segment, was proposed
L. Peltier et al. 1975. Analytical studies for assessing the impact of sanitary
sewage facilities of Delaware County, Ohio, Enviro Control, Inc., Rockville,
Maryland.
in the draft EIS (5). It was eliminated in the final EIS
because of engineering complications, higher con-
struction costs, and a substantial outcry from down-
stream residents. Regionalization of sewage facilities
with Columbus was not possible because the long-term
capacity of the existing interceptors was uncertain,
and because of a variety of political and institutional
problems. The original treatment plant site was
chosen in the final EIS as the most feasible (5).
Within the constraints of using this site, we needed
to decide how to provide wastewater treatment with a
minimal amount of damage to the Olentangy River's
ecosystem. We wanted a facility that would be cost-
effective in the fullest sense. The effluent to be dis-
charged from the plant, as originally proposed, would
have had a chlorine concentration limit established by
its discharge permit of 0.5 mg/1 (5). At average
streamflow, 12.5 mVs (441 cfs), this would dilute to an
instream concentration of 0.003 mg/1 for the 5,700
m'/d facility. The extremely low flow condition, 0.13
m-'/s (4.73 cfs), based on the minimum release guar-
anteed from an upstream reservoir, would give an
instream chlorine concentration of 0.17 mg/1 for the
5,700 m'/d (1.5 mgd) plant and 0.25 mg/1 for the
11,000 mVd (3 mgd) facility (5). These concentra-
tions assume no instream chlorine residual from the
Delaware City effluent which is discharged upstream,
so the actual instream values could be higher.
While we were preparing the EIS we consulted
Dr. William Brungs of EPA's National Water Quality
Laboratory at Duluth, Minnesota, about the effects
of chlorine on freshwater life. He indicated that fresh-
water fish are extremely susceptible to chlorine con-
centrations during long term exposures. He recom-.
mended a total residual chlorine limit of 0.003 mg, 1 to
protect aquatic life. Both coldwater and warmwater
fish species were affected by concentrations above
this limit f The 0.003 mg/1 maximum for receiving
waters had also been recommended in the National
Academy of Sciences' Water Quality Criteria publi-
cation in 1972 (6). EPA's 1976 Quality Criteria for
Water Publication (EPA-440/9-76-023), which was
published and distributed after production of the EIS,
has since recommended a criterion of 0.002 mg/1 for
protection of Salmonid fish and 0.010 mg/1 for other
freshwater and marine organisms. At the average
streamflow, at the 5,700 m3/d (1.5 mgd) size, the
Olentangy treatment facilities could discharge enough
chlorine to reach this concentration within the com-
Memorandum from William A. Brungs, U.S. Environmental Protection Agency.
February 19, 1976.
260
-------
PLANNING AND IMPLEMENTATION
pliance limit of its National Pollution Discharge
Elimination System (NPDES) permit. Because of this •
problem of toxicity to streamlife we decided that
chlorinated effluent would be an inappropriate addi-
tion to this high quality segment of the Olentangy
River.
At this point we obtained further technical assistance
from the Wastewater Research Division of EPA's
Municipal Environmental Research laboratory in
Cincinnati. We asked them to evaluate several disin-
fection alternatives, to decide if there was a better
choice than chlorination. Their report confirmed the
need to use a disinfection method that would produce
less environmental damage than chlorination. Earlier
demonstration studies performed by EPA researchers
at Wyoming, Michigan, showed that effluent dechlo-
rinated with sulfur dioxide and post aerated was non-
toxic to freshwater biota. The same study concluded
that o/onated effluent was also nontoxic. Both disin-
fectants were equally effective in treating tertiary
effluent to meet the bacterial standard for disinfection.
Other disinfectants, such as bromine chlorine and
ultraviolet light, were considered too new and un-
proven, at that time, to be used for this application.5
Treatment costs were essentially similar for either
dechlorination or ozonation, although o/onation was
slightly more expensive. Dechlorination does not
remove chlorinated organics, which have been of
concern in drinking water supplies. Dechlorination
also involves storing and handling chlorine and sulfur
dioxide, which are both toxic chemicals. Ozonation
offers several advantages to dechlorination. It has
been used on drinking water supplies for a number of
years with no observable toxicants. Ozone is produced
as needed, and is not stored at the site.6 Finally,
ozonation does not increase the instream levels of
dissolved solids (5).
CONCLUSIONS AND SUMMARY
We realized that ozone was a relatively new waste-
water disinfection technology. At that time it was
being used at two full scale treatment facilities, in
Ohio and Colorado, while other systems were at the
design or bid stage.7 We wanted to design a waste-
~Memorandum from E. F. Barth, U.S. Environmental Protection Agency.
June 1, 1976.
Iffld.
water treatment facility that would preserve and en-
hance the natural quality of this part of the Olentangy
River. The mandate of both NEPAand the Construc-
tion Grants Program for innovative solutions to pollu-
tion problems led us to investigate alternate disin-
fection technology.
While ozonation costs were higher than those for
chlorination, the definition of a "cost-effective" solu-
tion includes social and ecological factors, as well as
monetary costs (4). Social and ecological considera-
tions were significant in selecting ozonation as the
cost-effective alternative. One of the major con-
clusions made by the Regional Adminstrator in the
1976 final EIS was to employ ozonation at the
Olentangy Environmental Control Center (5).
Since then the facilities have been designed and the
new wastewater system is presently under construction.
Capital costs for ozonation in 1978 are about eight
times higher than the 1975 estimated costs for chlorin-
ation .8 However, this cost for ozonation is less than
1% of the total costs of constructing the first phase of
the interceptors and treatment plant for south central
Delaware County.
The EIS process was a successful tool for examining
disinfection methods and choosing the one which best
met local needs. In order to do this we had to take the
time to investigate a variety of alternatives and to
assess the environmental implications of our routine
funding of chlorination for effluent disinfection.
The need for a thoughtful and creative approach to
resolving water pollution problems applies to Facili-
ties Planning as well as to Environmental Impact
Statements. It is stimulating to attend a symposium,
like this one, which presents the latest developments
in a particular technical field. It is even more challeng-
ing to try to consider and appropriately use this new
technology. I hope we can all work together to solve
pollution problems and to reduce the environmental
damage of human activities.
REFERENCES
1. National Environmental Policy Act. 42 U.S.C.
et. seq.
"4321
40 C.F.R. '35, appendix A.4.C.3.
Personal communication with Gary Gilbert, Delaware County Sanitary Engi-
8
neer. July, 1978.
2. One of the most succinct sources for NEPA developement
is the Annual Report series of the Council on Environ-
mental quality; see Council on Environmental Quality.
1977. Environmental Quality-1977. U.S. Government
Printing Office, Washington, D.C. pp. 116-129.
3. Council on Environmental Quality. 1978. National Environ-
mental Policy Act. Proposed Regulations, 40 C.F.R. "
1500-1508. Federal Register v. 43 no. 112. Friday,
June 9, 1978. Part II. pp. 25230-25247.
261
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
4. Environmental Protection Agency. 1974. Final Construc-
tion Grants Regulations. 40 C.F.R. ' 35.908. Federal
Register v. 39 no. 29. Monday, February 11, 1974.
Part Iir. p. 5256.
5. U.S. Environmental Protection Agency, Region V. 1976-b.
Final environmental impact statement for the Delaware
County, Ohio Board of Commissioners Olentangy
Environmental Control Center and interceptor system.
U.S. Government Printing Office, Chicago, Illinois.
6. National Academy of Sciences - National Academy of
Engineering. 1972. Water Quality Criteria 1972. USEPA,
Washington, D.C. 594 pp.
DISCUSSION
QUESTION: Did I understand you correctly to
say that the average flow of the river is normally used
in Ohio to calculate the dilution factor for a disin-
fectant?
MR. MACMULLEN: No, you did not. I said that
the average flow was documented at 441 cubic feet per
second. I gave what the chlorine residual would be
from a 1.5 and a 3.0 mgd facility at that average flow.
But then I said, "However, this is not a free flowing
stream." Over and above the average and low flow,
you have the problem that this is a regulated stream
and in fact, the critical low flow that we use for plan-
ning purposes is 4.73 cfs.
QUESTION': 1 reali/e that in this case. But in
other cases where you do not have a regulated stream,
would vou use the average flow?
MR. MACMULLEN: No, sir. We use the seven
day, once in ten year low flow.
DR. RICE: I am fascinated by your thought pro-
cesses. 1 think this is the first o/onation sewage treat-
ment plant that 1 have heard of in which o/one was
chosen without apparently doing any piloting work.
Is that really the case?
MR. MACMULLEN: We did not do any piloting
work. That is true.
DR. RICE: There was not a collection system?
MR. MACMULLEN: That is a good point. There
was no collection system there to begin with, so it
would have been somewhat difficult to do pilot work
in southern Delaware County. I presume that we
could have gone to the city of Delaware and estab-
lished a small pilot facility there. The applicability
of results from the city of Delaware to the Delaware
County sewage treatment plant would not be 100 per-
cent necessarily, because there are industrial wastes
in the city of Delaware which we did not anticipate in
southern Delaware County. But in answer to your
question, we did not do any piloting work.
DR. RICE: 1 was not being critical at all. It just
seems to me it is a famous first for o/onation. That is
all.
MR. MONTAGUE, EPA ( Philadelphia, Region
3:) 1 just have a comment. 1 think the meeting has been
very enlightening in terms of what information is now
available, and I have learned a great deal from the
many speakers and information that has been pro-
vided. But frankly, I am also left in a quandary, and
that is to say, which system is most applicable to a
given situation. Since the matrix is so complex, we
find that we really have to look at it very, very uniquely
for each and every case.
But 1 think what is really needed for the engineer.
and particularly for the state and local entity that is
going to have to sign off oh a process, is to have some
basis of comparison initially to make some judgments
as to which direction they would really like to go. 1
think that the system right now as it stands leaves us
somewhat pu//led as to which one is really cost effec-
tive based on what standard.
The comment that was made several times during
the last tew days about "what is disinfection" 1 think
really comes to roost here. I would like to suggest to
Al, I think the region would support him in this regard.
is that we go forward with a joint effort on the part of
not just one company but on the basis of the various
companies that are here represented and those that are
not here in the o/one, UV, or chlorine entities and to
present, if you will, a project that would at least on
some equal level make a determination of what is the
cost effectiveness in a given situation. 1 reali/.e there
will be departures from that. But at least it would give
the engineers and those who are signing off on a con-
tract some insight as to what might be the situation.
I know many may not agree with it. It is one man's
opinion. Thank you.
MR. WHITE: What is the disinfection requirement
and who set it?
MR. MACMULLEN: The disinfection require-
ment was set in the Ohio Water Quality Stan-
dards that we had to deal with. And that was
200/400 coliform bacteria.
262
-------
28.
THE CONSULTANT'S VIEWPOINT ON ALTERNATE DISINFECTANTS
J. L. Pavoni, M. W. Tenney, M. E. Tittlebaum, B. F. Maloy, and T. L. Kochert
TenEch Environmental Consultants, Inc.
Louisville, Kentucky
INTRODUCTION
Disinfection of wastewaters for the control of
enteric and disease-producing organisms has been
common throughout the United States for a number
of years. In almost all instances, the mechanism for
disinfection has been the application of chlorine to
treated effluents. Until recently, little consideration
was given to the possible hazardous effects of effluent
disinfection on the environment. Concern about these
effects has been expressed in recent years, however,
and the use and availability of other disinfectant pro-
cesses is now being more thoroughly investigated. In
light of these developments, it has become increasingly
important for the engineering consultant to formulate
an evaluative procedure for the assessment of disinfec-
tant use in any given situation. This procedure must
take into account not only the relative merits and
disadvantages of chlorine use, but also those asso-
ciated with use of other disinfectants. This paper pres-
ents such a decision making model along with an
actual case study of its use.
AVAILABLE ALTERNATIVE DISINFECTION
TECHNIQUES
Wastewater disinfection processes, as they are util-
ized today, involve specialized treatment for the des-
truction of harmful and otherwise objectionable
organisms. Classically, disinfection processes have
been employed for the purpose of destroying or inacti-
vating disease causing microorganisms, primarily bac-
teria of intestinal origin, contained in domestic
wastewater. Other organisms which are currently the
target of disinfection processes include viruses, intesti-
nal protozoa, and various microorganisms.
The advantages and disadvantages related to the use
of alternative disinfectants or disinfection techniques
should be considered according to the following
factors:
• Technological state-of-the-art.
• Effectiveness in destroying particular types and
numbers of organisms under certain
environmental conditions (pH, temperature,
etc.).
• Safety precautions necessary for transport, stor-
age and application.
• Costs associated with physical facilities and
materials.
• Effects of reaction by-products.
• Availability of practical, accurate assay
techniques for determining disinfectant
concentrations.
• Tendency of the disinfectant to persist in resid-
ual concentrations.
• Energy requirements.
Most of the disinfectants currently being consi-
dered for domestic wastewater treatment have one
or more serious limitations that preclude their
general acceptance and adaptability. The next
section of the paper deals solely with the
disadvantages of chlorine for disinfectant use.
THE NEED FOR ALTERNATIVES TO
CHLORINATION
The use of chlorine to disinfect wastewaters has
undeniably produced a public benefit of great magni-
tude. Wastewaters contain a number of infectious
microorganisms, including Shigel/ae, Salmonellae,
and enteric viruses, that are removed by disinfection.
In the majority of cases, this disinfection has been
accomplished through the use of chlorine. Reductions
in the outbreaks of waterborne diseases have been
apparent since its widespread use. It is obvious that
263
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
chlorine disinfection has served to protect the public
health in the past; however, it may have also contrib-
uted to potential environmental damage.
In recent years much attention has been focused on
the occurrence of certain halogenated organics in
drinking water supplies, both in this country and
abroad.
Technological advances have produced instrumen-
tation sophisticated enough to measure part per bil-
lion quantities of chemicals formerly unknown in
domestic water supplies. Public attention was brought
to this fact when, in November of 1974, the United
States Environmental Protection Agency announced
the presence of small quantities of 66 organic chemi-
cals, some of which were suspected carcinogens, in the
New Orleans water supply. Since then, similar discov-
eries have been made in .all parts of the country (4).
Several of the compounds that are suspected of
being carcinogenic or toxic have been shown by-
numerous researchers to result from disinfection with
chlorine (1, 3, 6). Of these, the following have been
found in chlorinated wastewaters: methylene chloride,
chloroform, trichloroethane, trichloroethylene, dich-
lorobenzenes, trichlorobenzenes, and tetrachloroethy-
lene. Because of the health hazard associated with the
ingestion of these and other halogenated organics,
chlorine can no longer be considered the optimum
disinfectant for all uses.
If chlorine is no longer regarded as the best possible
disinfectant, other alternative disinfectants must be
considered in the planning of wastewater treatment
systems. These considerations should be based on fac-
tors such as the potential uses of the water and applica-
ble state stream standards. Several questions that may
be asked in determining the best disinfectant for a
particular installation include the following:
• Is the stream inhabited by sensitive aquatic life?
(Chlorine residuals- may prove toxic to many
aquatic species.)
• Will the water be used for human consumption?
(The presence of the halogenated organic com-
pounds may be possible.)
• Will body contact recreation take place within
the water? (Disinfection of effluents is
paramount if the answer is yes.)
• Are there in-stream standards regarding chlo-
rine residual? (If so, dechlorination of effluents
may be required- where sensitive aquatic life
exists.)
The appropriate place to ask these questions is in the
preliminary planning stages of treatment plant design.
This is best accomplished as part of the 201 Facilities
Planning Process outlined in the next section.
CONSULTANT'S PLANNING APPROACH TO
EVALUATING ALTERNATIVE DISINFECTION
TECHNIQUES
From a consultant's viewpoint, the selection of
appropriate disinfection techniques for municipal
wastewater treatment facilities must be accomplished
through the 201 Facilities Planning Process. The over-
all goal of this planning process is to develop a strategy
based upon the most cost-effective solution which will
meet established effluent and water quality goals
within a particular planning area while accommodat-
ing recognized environmental and social contraints.
With respect to wastewater disinfection, feasible
alternatives are identified on the basis of effluent lim-
itations and water quality standards which are usually
based on designated water usage and differ from state
to state.
The feasible disinfection alternatives are then tech-
nically evaluated by the consultant in terms of the
following criteria:
• Water quality impact
• Monetary cost
• Implementation feasibility
• Environmental effects
• Resource and energy usage
It should be pointed out that the resource and energy
usage requirements are evaluated primarily in terms of
cost. The preferred disinfection alternative should
then be determined by the consultant with input from
a Facilities Planning Advisory Committee (composed
of local officials, citizens, regulatory agency represen-
tatives, and environmental groups) which assigns an
overall priority to each disinfection alternative based
upon a criteria rating system established or approved
by the committee. This committee process insures
public participation in the development of an over all
facilities plan.
In the past, the 201 Planning Process has usually led
to the idenfication of liquefied chlorine gas or sodium
hypochlorite as being the most suitable disinfectant
according to the established criteria and priority rank-
ing system. In light of recent findings concerning the
environmental effects of these disinfectants, a re-
evaluation of priorities appears to, be warranted. A sug-
gested procedure for evaluating alternative
disinfection processes during the planning process is
presented in the form of a decision tree shown in
Figure 1 As indicated, this decision process is based
264
-------
PLANNING AND IMPLEMENTATION
Is receiving water used as a drinking water supply ?
No
Yes
Is receiving water used for
human contact recreation
or shellfish harvesting?
No Yes
Evaluate feasible alternative disinfection
techniques including chlorine and sodium
hypochlorite with particular concern for
possible health hazards due to formation of
halogenated orgon/cs.
Evaluate need for Hfostewoter
disinfection. Is disinfection
required P
Yes
Yes
No
Is there a potential for residual chlorine
_^Joxicity to aquatic life or do In-stream
chlorine residual standards exist?
No
Yes Consider chlorine, and
sodium hypochlorite for
disinfection
Evaluate feasible alternative
disinfection techniques
including dechlorination
Is chlorine or sodium
hypochlorite feasible for
disinfection ?
No
Eliminate chlorine and
sodium hypochlorite from
consideration
„ Prepare preliminary design as part of
201 Facilities Plan
"*~ Select most cost effective
"*" disinfection technique
on basis of:
e Water quality impact
e Monetary cost
e Implementation feasibility
• Resource and energy usage
FIGURE 1. DECISION TREE PLANNING APPROACH TO SELECTING A WASTEWATER DISINFECTION TECHNIQUE
upon anticipated water usage and the application of
in-stream chlorine residual standards. Therefore,
highest priority is given to the water quality and envir-
onmental effects with particular emphasis on eliminat-
ing the following:
• Health hazards caused by disease producing
organisms.
• Health hazards caused by halogenated organic
materials.
• Potential for residual chlorine toxicity to aqua-
tic life.
The suggested decision process is based on the use of
chlorine and sodium hypochlorite since these are the
most widely used disinfectants and they are currently
in the midst of considerable controversy. It is signifi-
cant to point out that these disinfectants are not elimi-
nated from consideration in the decision making
process unless they are shown to pose a health threat to
man, or if alternative techniques are shown to be more
cost-effective. Considering current technology, the
decision process will usually determine that
wastewater discharges to large rivers,.which provide
sufficient dilutional capacity to minimize health
hazards due to halogenated organics, can be most
cost-effectively disinfected with chlorine.
An additional feature of the proposed decision tree
is that it permits an evaluation of the need for waste-
water disinfection in cases where the receiving water
will not be used for drinking water, human contact
recreation of shellfish harvesting. This feature is con-
sistent with the conclusions of the U.S. Environmental
Protection Agency Task Force (2) finding that in some
circumstances, "such as high dilution and die-off, sea-
sonal recreation, and no downstream reuse potential,
the benefits of disinfection for protection of public
health are minimized and may not be needed."
A major advantage to the suggested decision process
is that it provides the consultant with a well defined
approach for evaluation of alternative disinfection
processes. It is anticipated that the utility of such an
approach will increase as the state-of-the-art of alter-
native disinfection techniques becomes more
established. A case study involving the use of a similar
decision tree process is summarized in the next section.
265
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
CASE STUDY OF DECISION TREE PLAN-
NING FOR ALTERNATIVE DISINFECTION
TECHNIQUE SELECTION
A case study of the decision tree planning approach
to the selection of the optimum disinfectant was con-
ducted in Louisville, Kentucky, for the Louisville and
Jefferson County Metropolitan Sewer District (5). In
this study, a Welsbach ozone pilot plant was put into
operation at the Fort Southworth Sewage Treatment
Plant to evaluate the possibility of ozone use in disin-
fection. At the time the study was conducted, the
Fort. Southworth plant was a primary sewage
treatment facility discharging to the Ohio River
downstream of Louisville, Kentucky. Renovations
were being planned to upgrade the plant to
provide secondary treatment including disinfection,
so the decision tree approach was utilized to aid
in the selection of the disinfectant.
The Ohio River is used as a drinking water source;
therefore, alternate disinfection methods were consi-
dered along with chlorine use. The dilution effect of
the river made the use of chlorine a viable alternative
for consideration. Of the available alternate disinfec-
tants, ozone was chosen for further study since this
process phased in well with the proposed high purity
oxygen activated sludge treatment scheme, it had been
used in Europe for disinfection, and the U.S. Environ-
mental Protection Agency was interested in a pilot
study involving ozone. The experimental approach
was taken and results of the study are outlined below.
In preparation for the study, a 150 m3/d (40,000
gpd) package plant was added to the 4 x 105m3/d (105
mgd) Fort Southworth facility to provide secondary
treatment prior to ozonation. This step was necessary
to evaluate the use of ozone in the treatment scheme
proposed for the plant after upcoming renovations.
The ozone contact reactor was connected to the efflu-
ent weir trough of the package plant. The reactor
consisted of nine rectangular chambers through which
ozone was dosed at approximately 15 mg/1. Samples 5
to 10 ml in size were taken at influent and effluent
points and in each chamber for viral and bacterial
analyses. The results of the study were as follows:
• a 15 mg/1 ozone dose provided 99+% removal
total coliform, fecal coliform, and fecal strepto-
cocci bacteria. Average effluent concentrations
after ozone treatment were:
total coliforms —500/100 ml
fecalcoliforms— 103/100 ml
fecal streptococci — 9/100ml
• Ozone provided nearly 100% inactivation of a
test virus within a 5 minute contact time at a total
ozone dosage of 15 mg/1 and a residual of 0.05
mg/1.
• Oxone provided an average turbidity reduction
of 70%, a COD reduction of 29%, and color
removal.
• Ozonated effluent produced no noticeable
harmful effects on native fish populations while
non-ozonated secondary effluent proved toxic in
a test study.
Obviously, ozone was proven to be an effective disin-
fectant for use in the Fort Southworth treatment
scheme. The decision tree approach was utilized in the
final disinfectant selection as outlined below.
As a result of the ozone pilot study, both ozone and
chlorine were being considered for use as a disinfectant
following the high purity oxygen secondary treatment
process. In the case of chlorine, there were no water
quality standards limiting in-stream levels and adverse
effects on aquatic live were not expected. The costs for
the ozone system were estimated to be slightly higher
than for chlorine disinfection. However, it was also
realized that direct comparison of process costs was
not realistic since ozonation achieved effluent polish-
ing in addition to disinfection. Nonetheless, at that
point in time chlorine technology was well known
while ozone use was rare in this country and a conser-
vative approach was preferred. For these reasons,
chlorine was chosen as the optimum disinfectant for
use at the Fort Southworth Treatment Plant.
SUMMARY
The use of chlorine as a disinfectant in wastewater
treatment systems may not always provide the maxi-
mum benefits to be obtained by bacterial reduction.
Chlorine residuals have been noted to produce adverse
effects upon aquatic life and to form several carcino-
genic or possibly toxic compounds when applied to
wastewaters. For these reasons the use of a number of
alternate disinfectants must be considered in waste-
water treatment systems. These include ozone, ultravi-
olet radiation, bromine chloride, chlorine dioxide, and
dechlorination methods. The availability of various
disinfection alternatives necessitates the use of an
appropriate decision making process to provide a
mechanism for selection of the optimum disinfectant.
The decision tree approach, presented herein, consid-
ers both the environmental and health effects of disin-
fectant use for the evaluation of various disinfection
alternatives. Use of such a technique should avoid
decisional paralysis and incorporate disinfectant selec-
tion as a logical part of the facilities planning process.
266
-------
PLANNING AND IMPLEMENTATION
REFERENCES
I. Bel la r. T.A., J. H. Lichten berg, and C.R. Croner. November.
1974. "The Occurrence of Organohalides in Chlorinated
Drinking Waters" EPA-670 4-74-OOX.
2. EPA Task Force Report. March, 1976. Disinfection of Waste-
water. EPA-430 9-75-012.
3. Jolley. R.W. October. 1973. "Chlorination Effects on Organic-
Constituents in Effluents from Domestic Sanitary Sewage
Treatment Plants". Oak Ridge National Laboratory.
4. Kochert. T.I.. Master Eng. Thesis. University of I.ouis\ille.
Unpublished.
5. Pavoli. .I.E., M.E. littlebaum, H.T. Spencer, M. Fleischman,
C. Nebel, and R. Gottschling. December. 1972. "Virus
Removal from Wastewater Using O/.one". Water anil
Sewage Works. Vol. 1 19. No. I 12.
6. Rook. J..I. 1974. "Formation of Haloforms During Chlorination
of Natural Waters". The Journal of the Society for Water
Treatment anil fi\aniination. Vol. 23. Part 2. p. 234.
DISCUSSION
MR. WHITE: Was that ever characterized as a
well oxidized secondary effluent?
DR. PAVONI: It was a fairly well oxidized
secondary, but it was not filtered which obviously it
now would be.
DR. TITTLEBAUM: There was a nine hour
detention time in the secondary followed by two
clarifiers in series.
MR. WHITE: Was it nitrified?
DR. PAVONI: No. I think that a point to make
here if 1 remember right, is the estimate of overall
treatment costs at that time at that facility was 20 to 25
cents per thousand gallons. The cost of chlorine was
estimated at somewhere around 1.0 to 1.5 cents per
thousand. The cost of the ozone was estimated
somewhere around two to three times that. So it was
two or three times the disinfection costs, but viewed in
terms of the overall treatment costs, it was not a
significant percentage change. I think that if that same
pilot study conducted in 1972 was conducted today
with the concerns about the organics in the Ohio
River, probably ozone would be the disinfection
process at that plant instead of chlorine.
MR. WHITE: 1 just want to make a comment that
1 mentioned to Joe in private. We have the decisions in
California already made for us. But this is very
interesting because it took California about 3D years to
finally wind up with this type of process. So every
discharger in the state is faced with a given set of
conditions that is already spelled out for them in this
selection of what the requirements will be.
MR. FLEISCHER: 1 just have a quick comment
in that I am somewhat familiar with this through the
person who ran the pilot plant. Carl Nebel. You did
not really go over anything but disinfection, and I
realize this is a disinfection seminar. But if anybody
did see that froth forming, and I heard some
comments, that was what I think you refer to as the
polishing, which is suspended solids removal. This is
another aspect of ozone we cannot really discuss here
in disinfection. But in some cases when you have a low
amount of suspended solids, ozone will remove it
perhaps better than a filter will. That is the froth
you saw at the top of the contact chamber.
DR. PAVONI: Right, that is very true. We did not
get into the other aspects besides disinfection.
DR. RICE: I have a quick observation on this.
The contacting system that was used In Louisville was
a nine chamber bubbler, which effectively showed that
after the third chamber, you really did not need the
'other six. This was actually confirmation of the earlier
Blue Plains study in D.C. which also used a nine'
chamber contactor and also showed that after the
third one, you do not really need the other six. Now
five years later, Estes Park puts in a nine chamber
bubbler contactor. I am not really criticizing other
than to say that someday I think we ought to learn
from what we have done in the past. 1 think that was an
expense that did not have to be.
DR. PAVONI: Since there are no other questions,
one final comment is that I realize that California in
most cases is ahead of the rest of the country as far
as implementing environmental controls and proce-
dures. All I think we are getting to here in our
paper is the fact that we work primarily in the mid-
west, having offices in both Louisville and South
Bend. We really have not seen this type of approach
reviewed by either the state agencies or by Regions
4, 5 or 7, in general. We just think that 'it is
something that should be emphasized a little more
strongly, especially now with the increased emphasis
that is being placed on disinfection.
MR. BARTH: Thank you, gentlemen. We have
an unscheduled speaker, Jack Mills from Dow
Chemical. Jack, we are going to let you have five
minutes.
DR. MILLS: There have been a number of people
who have asked about bromine chloride. Since we were
not on the program, I thought 1 might spend a few
minutes just discussing what has happened since the
Wyoming project. Some of the things I am going to say
267
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
quickly here probably should have been said before
lunch. It might have fit in with some of those other
subjects. One of the reasons is that bromine chloride is
a very effective virucide and it has been shown to be
better than chlorine. This is one of the advantages that
it has had.
In the Wyoming project, as some of you know, they
looked at alternatives to chlorine; ozone, bromine
chloride, and dechlorination were the three
alternatives. The conclusion from that trial indicated
that chlorine indeed was a problem, both chronic and
acute toxicity-wise, whereas the alternatives (ozone,
bromine chloride, and dechlorination) were not, as far
as toxicity.
As far as antimicrobial activity or control of
microorganisms, both dechlorination and bromine
chloride were satisfactory on secondary effluent and
ozone on tertiary effluent. Since then we wanted to
conduct some full-scale plant studies. Some of you
know we had some problems with the feeding
equipment at the Wyoming project and we had the
same problem in our first trial. But I am confident we
ironed that problem out. Bromine chloride is a liquid.
It has to be vaporized, which poses a unique problem.
It does feed very similarly to chlorine and you can use
similar equipment, but you do have to use a vapori/er
and that was our main problem.
In our trials we picked both secondary and tertiary
treatment plants. I think mainly secondary is
important because I would guess most plants do not
produce filtered nitrified effluent in this country nor
even primary as mentioned earlier. But to evaluate the
suspended solids problem and this sort of thing, you
really have to look at a typical secondary plant.
Anyway, we have two plant trials going right now —
one in Hatfield Township, Pennsylvania, and this is a
3.6 mgd tertiary treatment plant; and then we are
also involved with the state of Maryland on a 3 mgd
plant at Sykesville, Maryland. This is a secondary
treatment plant. Both of them are activated sludge.
These trials have been running several months. The
data so far indicate that we do control the fecal
coliform levels between zero and 20 per 100 ml as well
or better than chlorine was at lower dosages. In both
cases, I think the d.osage was half the chlorine dosage.
The residuals are running in the neighborhood of
0.1 mg/1 at the outfall, but shortly after downstream
from the outfall, they disappear completely because
bromine chloride residuals continue to drop off. This
is the one significant difference between chlorine and
bromine chloride, that is, that you form an unstable
residual. The bromamines decay much faster than
chloramines and disappear quite rapidly. So we are
meeting those requirements.
The third thing I want to mention is that we are
looking at organics. Now one of the things that I think
is very important here is that haloformsdo not seem to
be formed as much in sewage disinfection as they are in
potable water, and there are reasons for this. Now I
will not get into this, but we are looking at haloforms
in the outfalls of these plants and we are finding less
than 20 parts per billion. Shortly downstream there
are none observed. The reason for this from our
laboratory data is that the haloforms in wastewater
effluent vaporize or are volatile and disappear. That is
not the case in potable water. In potable water you
have a pressurized system and the chloroform
continues to build up in the distribution system. When
you drink a glass of water, you are going to get so
many parts per billion of chloroform. If you drank out
of a stream down from a sewage plant, I am confident
that you will not see any chloroform because the
half-life in water is less than six hours.
We are also looking at nonvolatile organics. Bob
Jolley did not mention this, but Ethyl Corporation ran
a pilot study down at Hampton Roads, Virginia. They
evaluated some of their effluent for brominated
organics. But it does indicate that the brominated
compounds are much less active or much less stable,
anct that is one of the key differences in the halogen
chemistry between chlorine and bromine.
First of all, oxidation is the predominant reaction.
And second, the types of compounds that are formed,
are very low in concentration and are much less stable.
MR. WHITE: Jack, when am I ever going to sell
you on the idea of taking existing chlorinator
installations, use the primary chlorinator to satisfy the
initial demand, and then generate bromine on site with
the stand-by chlorinator and bromide salt. You
would have the whole thing made, instead of
going through all this rigamarole?
DR. MILLS: George, we have considered that
and there are a couple problems. We have tried
it. One problem is that it is more expensive. It is
not as simple as you say.
MR. WHITE: It is simple in France.
DR. MILLS: Well, ammonia interferes with that
reaction between bromide and chlorine. And you do
have a problem of mixing and getting a complete
reaction. I am not saying it is not possible. I know the
French have dojie it.
268
-------
APPENDIX A.
BANQUET PRESENTATION:
"THE LAW'S RESPONSE TO PUBLIC HEALTH HAZARDS'
J. V. Karaganis
Karaganis and Gail, Ltd.
Chicago, Illinois
MR. VENOSA: Mr. Francis T. Mayo, who is the
Director of the Municipal Environmental Research
Laboratory here in Cincinnati, will be introducing our
guest speaker, Mr. Joseph V. Karaganis, attorney
from the state of Illinois.
Mr. Mayo has been our Laboratory Director for
approximately two and a half years. Before he was
appointed director of MERL, he was Regional
Administrator for EPA Region 5 in Chicago for six
years. During that time, he became very well
acquainted with Joe Karaganis. So I think it would be
fitting for Mr. Mayo to introduce Mr. Karaganis this
evening.
MR. FRANCIS T. MAYO: This is indeed a privi-
lege. I am going to call him Joe, not Mr. Karaganis,
because I have known him as Joe for eight years now.
We were reflecting on the occasion when Joe
and I first met. It was Earth Day, Chicago,
1970, at a time when Joe Karaganis was kind of
downy-cheeked and I was a very virginal Regional
Administrator for EPA.
Many things have happened in the ensuing years.
Joe enjoyed a very enviable position for an attorney, I
thought, during those years because of a special
assistant relationship to the attorney general in
Illinois. The attorney general's office there enjoys some
environmental latitudes that most attorney generals
do not. The office has the opportunity to initiate
actions of its own whereas in most states, they come to
the attorney general's office through an environmental
agency.
Bill Scott in Illinois has had a deep personal concern
for environmental issues and a pretty astute political
perspective. With Bill Scott's initiative and Joe
Karaganis' talents, they have built a very, very
enviable reputation for dealing with the giants in
industry and taking on the big ones in some very, very
difficult and very challenging environmental issues.
Through the years between them, they have enjoyed an
outstanding reputation for successes, one of the most
recent ones being the increasingly famous City of
Milwaukee litigation in which the state of Illinois
initiated suit in the federal district court to bring about
some abatements and improvements in the
management and operation in the overall facilities for
the Milwaukee wastewater treatment system as well as
those for some of the intervening municipalities
between Milwaukee and Chicago.
That decision has created great anxiety among
many of the major metropolitan sanitary districts in
the country. It has added to the commentary on the
role of the courts in resolving difficult environmental
issues. It has thrown a scare into many, many
organizations faced with somewhat comparable
circumstances.
In addition to being a special assistant to the
attorney general in Illinois, Joe has a private practice
— the name is Karaganis and Gail. As a firm, they are
achieving an outstanding reputation nationally for
taking on and prevailing in a number of very, very
difficult environmental litigations. I am confident that
you will find Joe's commentary on the Milwaukee case
most interesting, and I think you will find him a very
delightful speaker.
From an attorney's standpoint, Joe has a very, very
great capability to be both a quick learner and to be
able to translate basic principles of science and
engineering experiences to the courtroom scene and to
use the combination of those factors and the law very,
very successfully.
It is my pleasure to offer to you as your speaker this
evening Mr. Joe Karaganis.
MR. JOSEPH V. KARAGANIS: Thank you.
Frank, ladies and gentlemen at the main dinner table
and ladies and gentlemen in the audience.
As Frank suggested, at the time when he first met me
back in 1970, I had been working in the field of
pollution control a relatively short time. I think we
have both received a few bruises and had a few
experiences in that interval of time since then.
269
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
To show you how one accidentally, and literally
accidentally, gets into a field like this, my first
experience with pollution problems came from
Congressman Abner Mikva from Illinois—some of
you may know of him. He was at that time a
congressman from the south side of the city of Chicago
and asked me to go out and help some of his
constituents who were having a fight with an asphalt
batching plant. It was not a water pollution problem;
this was an air pollution problem.
I went out there, and it turned out that the asphalt
batching plant was not only using a sour crude or a
sour petroleum derivative highly loaded with sulphur
for its petroleum base for the asphalt, but they were
also using slag from the Wisconsin Steel Company,
which had again a great deal of sulphur in it, for the
basis of the aggregate (the hard material that made up
the asphalt). Presumably, this was all again theory.
The heating up of this process, be it the sulphur and the
oil or the sulphur and the slag, resulted in an intensely
acrid smell throughout the residential neighborhood
which was surrounding this plant. If you have ever
stricken a match close to your nose and caught that
stench, you know what I mean; only this was on an
eight to ten to twelve hour a day basis.
This phenomenon had occurred in the way that
Chicago politics traditionally worked these things.
What had happened was that environmental control
over the years had been one of zoning—to keep the
offensive odors away from the residences. We do not
work things that way in Chicago. As long as the
industrial land next to this plant was vacant, some
sharp real estate developer got ahold of his local
alderman, the right amount of consideration passed
between them, and the next thing you know we had
zoning for residences right next to this asphalt
batching plant.
The poor yokels who came out to buy their homes—
and home ownership is a very serious and intense thing
among many people in Chicago—came out typically
on a weekend (on a Saturday or a Sunday) and were
told by the honest and forthright real estate developer
that the asphalt batching plant was closed and would
never operate again. Lo and behold, upon coming into
the neighborhood, they found somewhat to their
chagrin that that was not the case.
Well, when I got there, I smelled this odor like they
did. I reported it to the city officials, thinking, of
course, we would get immediate and aggressive
enforcement. Lo and behold, 1 still had a few things to
learn about Chicago politics which, for all the jokes,
we are not that different from the nation as a whole. I
found that I did indeed get an inspector. An inspector
from the local Air Pollution Control Department
came rushing out in his car with his flashing Mars
lights and it looked very official and very impressive.
He sniffed the air and said, "I don't smell anything. If
anything, it smells to me like bacon and eggs. It smells
pretty good to me."
Well, I learned that this inspector subsequently had
the name Gilfoil which was a very helpful name to have
in Chicago at that time because Gilfoil was the maiden
name of a woman by the name of Eleanor Daley, who
was married to the then-mayor of the city of Chicago.
But as it happened, Inspector Gilfoil never did find a
problem. Over a four year period, we had to take this
asphalt batching company to the circuit court of Cook
County and fight it through as both a citizen's suit and
subsequently a state of Illinois suit, until we got relief.
The asphalt plant really gave an example of how the
law relates to environmental hazards. I could not sit
back and tell you what public health hazard that
asphalt company presented. I know from the
standpoint of just a daily living aesthetic nuisance, it
made it impossible to live in the area and enjoy your
home at all. People would try and get out of the area as
soon as they could. But it gave a good example as to
the step-by-step process that one faces when trying to
present this problem to either public agencies for
administrative correction or to the courts for judicial
abatement.
One of the things I remember is that the first step is
always: "Your Honor, we're not causing any problem.
These people are crazy." Finally the court would be
convinced after witness after witness would come on
the stand that yes indeed, there was a problem.
The next thing was to look for the quick and dirty
cure. And the quick and dirty cure—and you can pick
your field, be it water pollution or air pollution- is
always something that is very low cost and the
perpetrator of the act will say "we will immediately
and. permanently cure the problem."
It was kind of funny and sad in a way. This fellow
who ran the asphalt plant, which interestingly enough
we subsequently found had a lot of contracts with the
city of Chicago, which may have had something to do
with why we were getting less than aggressive
enforcement, decided that he was going to solve the
problem by putting perfume into the smokestack.
You can imagine the feelings in the neighborhood
when in addition to the acrid, bitter, biting stench of
the sulphur that this was overlaid with an incredibly
heavy scent of what smelled like soapy perfume
throughout the neighborhood. That was not either the
270
-------
J. V. KARAGANIS
immediate or the ultimate solution to the problem.
But my remarks here tonight are not to deal with
aesthetic nuisance, although that can be a significant
one. I would like to deal with the question of (and it
was a topic that I suggested) the law's response to
public health hazards, and in particular the law's
response, be it judicial or administrative, to the
question of the public health threat presented by
pathogenic organisms in human sewage.
One of the things that I asked when I came here this
evening was to what extent your seminar and
conference had discussed this issue. One of the things
that I find incredibly frustrating in the field today is
almost a schizophrenia within the Environmental
Protection Agency as to what should be the control, be
it intermittent or permanent or nonexistent, at the
discharge point of sewage treatment plants; hopefully
it gets to the sewage treatment plants. In most cases it
does not. 1 will have a little bit to say about that—what
should be the level of control for pathogenic
discharges?
I happen to be a public health advocate. I have yet to
find the expert today anywhere who can sit back and
say to me that I can safely rely upon the forces of
nature to correct the problem, or to rely upon the
municipal officials to take the water in and to treat it so
it will not cause any hazards, or who can give the kind
of guarantees that I feel I am entitled to as an average
citizen in consuming public waters as a resident of the
Chicago metropolitan area or any other area. And I
am an advocate for maximal controls, maximum
elimination of pathogenic organisms at their source.
Containment is a philosophy that I have followed as a
lawyer not simply by one of legal or political
philosophy but one of having cross-examined
probably three to four hundred experts in a whole
variety of fields and finding and being amazed by and
being impressed by not how much we know, but how
little we know with respect to the receiving
environment's capability of assimilating various
pollutant or contaminant discharges. You would be
amazed when you start going down the track of
variables that exist in nature as to how little the most
prominent experts in the field can tell you about the
variability of the concentration and strength of the
discharge, about the efficacy of the treatment methods
that currently exist, about the transport and diffusion
and reconcentratioh of pollutants in the ambient or
receiving environment.
So what 1 am suggesting to you is I want to get it
clear from the start that I have a bias in favor of
containment. But I find frustrating the agency to
which we look for guidance, the big eagle I call it. ...
Now 1 will tell you, it is a great game that lawyers
play, a great game. How many of you have ever seen a
report that is not worth the paper it is written on but it
has a big eagle on the front? "United States
Environmental Protection Agency." Lawyers love to
wave those things. They walk in, "Judge, the United
States Environmental Protection Agency says. ..."
What is a problem is when both lawyers are waving
documents saying, "United States Environmental
Protection Agency. . . ." and they say diametrically
opposed things. And this is one of the problems we are
having in the field of pathogenic control today.
1 cannot identify the source of this dilemma. I do not
know in the field with the public health professionals
that 1 talked to, they are almost universally in favor of
strong controls. In those of whom have said, "Don't
worry. There isn't a threat," I have found they have not
done the level of research in many instances which
allows them to establish definitively that there is not a
threat. I would accept immediately anybody's
statement that: "Yes, there are a lot of unknowns." But
for people to say, "Don't worry about it," is to me
totally unfounded from a standpoint of any empirical
evidence that 1 have been able to see.
And 1 think this dichotomy between "We don't
know; therefore, it won't hurt us," and on the other
side, "We don't know all the answers; therefore,
contain"; is one that has been reflected in the
development of the law's response to public health
hazards.
Many of you have heard about nuisance law as
• being relatively antiquated and not able to respond to
current needs. Let me suggest to you that much of the
development currently within the field of
environmental law is based upon the identical analysis
of public health risk and public health risk
containment that is being developed in the most
sophisticated of recent cases.
When you go back to (and we recently did this in the
Milwaukee case) the federal Appeals Court in
Chicago, which is now hearing the Milwaukee appeal,
they said to us and to Milwaukee, "Really tell us what
common law nuisance is all about." Everybody thinks
they know what nuisance law is all about. Most
laymen do; most lawyers do. We take in the course
called "Torts" and we read a hornbook statement of
what a nuisance is, but we never really have analyzed
what its origins were and why it developed. I would
like to if I can because this is very, very important.
How many of you in dealing with lawyers in any of
these fields have heard the statement by the lawyer,
271
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
"Well, we cannot stop this problem unless we have
clear and convincing evidence that it will certainly
cause harm." This is the standard defense line in any
legal attack, be it quasilegislative through a regulatory
process, be it through a permit process, or be it
through a judicial attack. You must show clear and
convincing evidence that there will be harm. This so-
called clear and convincing burden turns out to have
been developed through an anomaly, a quirk, in the
jurisdictional development of the law between what
are known to lawyers as the fields of law and equity. I
would like to tell you about this in a little while because
I think it will convince you that the law offers remedies
that most of you may have felt in the past have not
been available.
We went back as a result of the mandate from the
7th Circuit Court of Appeals and we went back to such
publications. I will tell you that I had the fortunate or
unfortunate experience of attending a Jesuit
secondary school and I could not remember one ounce
of my Latin, but what 1 could remember I believe were
Cicero and Caesar's commentaries. But what I could
not remember and could not decipher at all were the
medieval bastardization of Latin that came across in
such publications which we had to go to. We had to go
to the Latin translation of such things as "Lord
Glanville—A Treatise on the Law and Customs of the
King of England" which was written in the year 1189
A.D.
Now in Glanville and books'like "Coolie's Treatise
on Blackstone's Commentaries,"and "Hawkins'Pleas
to the Crown," and Stephen's "A Disgest of the
Criminal Law," we were able to gather the roots, the
origins of the field of common law nuisance,
particularly with respect to the public health hazards
and the English response to those public health
hazards. Why are we so concerned about the English
response?
Well, we lawyers love 16 cloak ourselves in these
broad, amorphous things. 1 am always critical of
scientists about using terms like "judgment" rather
than being precise and finding out what is behind the
judgment. Well, we do the same thing. We have this
mantle we call the common law. We would really
ideally like to wrap ourselves in the robes and wigs of
the English judges. We are not able to do that so we
come across with code words that assist us in that.
And in the common law roots of nuisance, we find
that English law being concerned about things like a
plague. . . .one of the first areas of nuisance, or the first
area of nuisance was a field called a purpresture.
Forget that word and forget I ever said it. It is of
little use to you and you do not need to remember
it. But a purpresture was essentially an interference
with the public right. If you built a wall across a
public highway, you were guilty of a purpresture.
You were interfering with one of the King's rights,
which was freedom of movement across a public
highway. And in feudal times the King could emerse
your lands. You would be at the mercy of the
King and he would take your lands.
This was all done by the criminal law. It was not
done in the fields that were called equity or chancery.
It was done by the criminal law. You were prosecuted
for the creation of a nuisance. And part of the
prosecution was to put you in jail and to force you to
abate the nuisance. There was nothing about clear and
convincing or anything else. If you caused a problem,
you were off to jail, your lands were at the mercy of the
King — you were in real trouble.
Now 1 suggest to you that the development of
purpresture then went into the fields of public health.
You may find in the roots of English common law
references to abatement through the criminal process
of the building of tenements. And why? Not because
the tenements caused disease, not because disease was
found in the tenements, but because of the common
sense and public health recognition that in medieval
times when you packed people together in unsanitary
conditions, the likelihood of disease was substantially
increased. It was a risk that was created which would
be abated. It was risk abatement, not disease
abatement, but public health risk containment. So you
find that at the very origins of the common law.
You find in the definition of common law nuisance
again, these are public nuisances now. I am not going
to bore you with a very fundamental, and both
theoretical and practical, distinction between public
and private nuisances. But most of the law that has
been developed with respect to private nuisances has
only sought and resulted in the confusion as to what
our opportunities are under public law and the law of
public nuisance.
There is a description, for example, of a public
nuisance of "one who carries a child infected with the
pox along a public highway or adjacent to populated
dwellings." Again, the risk of infection is there. Thf
risk is what creates the public nuisance subject to
criminal abatement.
The reason that I wanted to take you up to this point
is to show you not only the principle has been
established early in the la w but also to show you where
272
-------
J. V. KARAGANIS
the proof comes in.
When you go before a judge today and you want a
discharge source stopped, the judge is exercising what
is known as the jurisdiction of the Chancellor, the
jurisdiction of courts of equity. In England there were
divisions between the chancery courts and the courts
of law. The chancery courts were able to effectuate
rapid relief, immediate relief, immediate abatement of.
a problem. One of the things that grew out of the
Magna Charta was the right to trial by jury.
As you know, today, we lawyers are famous for our
ability to delay proceedings. Well, the lawyers in those
times were also capable of delay; and the trial by jury
resulted in inordinate delays. So many times the public
prosecutor would come to the chancery courts and
say, "We don't have time to wait for trial byjury. We
don't have time to wait for the facts to be found by a
jury of our peers or a jury of the defendant's peers. We
must have immediate relief."
You see this articulated repeatedly under such
people like Story, who was a famous justice of our
United States Supreme Court. The equity courts
would say, "Well, look, when the facts are in conflict,
when there is Expert A, Dr. Jones, and Expert B, Dr.
Smith, in conflict over whether or not this really is a
hazard, I will demand of you clear and convincing
evidence. And where I have this scientific conflict, I
will very often deny relief because of the very conflict
in testimony."
Most lawyers have stopped there. They have said,
"If you want common law public nuisance relief, you
must show clear and convincing evidence,
uncontroverted evidence or you will not get the relief."
They never bothered to read the procedure thereafter.
The procedure of the courts of equity historically
was to then send the case back over to the jury calendar
in the law courts who would then try the case, find the
facts, and then upon the finding of a hazard, relief
would be granted. What we have done today is to have
eliminated this time-honored distinction between law
and equity.
The concept of clear and convincing evidence of a
hazard, of clear and convincing evidence of actual
harm, for example, that has existed in equity at-best
continues in the field of what is known as"preliminary
injunctive relief" where we go into court tomorrow
and seek an immediate injunction. It has no historical
or practical basis in the area of common law nuisance
or any other area in dealing with a fully litigated
controversy.
I suggest to you as scientists to read a decision which
is both a basis of support for my position as well as one
that creates great food for thought. It is the case of
Ethyl Corporation versus EPA. It has nothing to do
with nuisance, but it has a great deal to do with the
whole question of scientific probabilities in the field ot
judicial review of administrative decision making as
well as the whole question of judicial decision making
on the facts. Judge Skelly Wright is the author of the
principal opinion, and it deals with this whole question
of lead additives in gasoline and the kind of public
health hazard that they presented.
What I am saying to you is that public health
hazards, if clearly articulated, even in the presence of
conflicting or contradictory evidence on the other side,
are the proper subject of judicial scrutiny and judicial
action on the basis of the preponderance of the
evidence. Preponderance of the evidence simply
means: there is evidence on one side that says one
thing; there is evidence on the other that says the
other—which am 1 most convinced by? If I am
convinced by the experts on public health hazards,
then I am going to go forward and 1 am going to grant
relief.
One of the things that is very, very important in this
area is that in the public health field, the concept of
balancing conveniences is looked upon with a great
deal of resistance. The concept of allowing the public
to be exposed to a significant health hazard simply
because it is convenient for the discharger to continue
discharging is looked upon with great jaundice by the
courts.
One of the cases that I think you should be most
familiar with in the whole field of health hazard
protection and environmental law is the case of
Reserve Mining versus United States. Reserve Mining
versus United States is a classic example of how these
things develop and how the law handles them. I have
to say on balance that it was a very good decision. I
disagree with some of the law that was used, but it was
a very good decision ultimately in terms of the result.
Judges lately have created this doctrine and they
think it is new; they think it is something they invented.
They call it risk-benefit.- They say, "What is the
probability of the harm? Is it great? Is it small? What is
the statistical likelihood of the harm occurring?" And
then, assuming we have established the statistical
likelihood: "What is the severity of the harm if that
were to occur?" It is a two-pronged test. It could have a
very low statistical probability but great harm if it
occurred. It could have a very high statistical
probability but be innocuous if it occurs.
Among the first cases to articulate the question of
risk-benefit is the Reserve Mining case and Judge
273
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
Skelly Wright. And it has been articulated primarily in
justifying EPA or OS HA or Department of Labor
regulatory action in such toxic substances or
potentially toxic substances as aldrin-dieldrin. It has
been involved in the field of asbestos in AFL-CIO
versus Hodson. It has been involved in the field of
vinyl chloride control in an OSH A regulation. And, in
a nonregulatory sense, in a courtroom sense, it was
involved in the Reserve Mining case in the field of
asbestos-like fibers.
And I suggest that you all, if you get a ctiance, wade
through the final decision of the 8th Circuit in the
Reserve Mining Case because it shows you how far the
courts will go with the appropriate public momentum
behind them. This is a very important aspect. Courts
do not operate in vacuums. Courts are responsive to
the public mood; the courts are responsive to the
political winds just like anybody else. But they will go
very far forward in protecting the public health.
In the Reserve Mining case, first of all, it started out
as a water pollution case from the standpoint of: Was
Reserve Mining discharge causing anything with
respect to the eutrophication or with respect to the fish
or the biota of Lake Superior? I do recall having
worked with a number of EPA scientists who were
involved with that case and were working with me in
related cases. There was a great deal of frustration
about their ability to prove or disprove the question of
water quality impact.
Then along came public health. We lawyers always
look for this. I mean we have to have it. It's a blood and
guts issue. No question about it. Public health has sex
appeal. Public health is something a judge or any
decision maker can relate to. Very few decision makers
want to be subject to cancer. Very few decision makers
want to be subject to a debilitating disease. So if you
can say to them, "Look, you or your children or your
relatives (or whatever the case may be—you do not say
that to them directly but the level of analysis is there—
they are thinking this) are clearly going to be subject to
a serious health risk," you have come a long way in
establishing the basis for abatement of a discharge.
Here you had Reserve Mining's asbestos-like
fibers—not clearly asbestos, not proven to have all of
the properties of asbestos, subject to a great deal of
conflicting evidence as to the mineral and geologic
composition of these materials as to their properties of
being like asbestos, but nevertheless accepted by the
court as having the same properties as asbestos. All of
the evidence with respect to harm caused by asbestos
relates to inhalingasbestos—ingestion through the air.
There is absolutely not one iota of evidence with
respect to the carcinogenic effects of oral ingestion of
asbestos through water. There is no epidemiologies.1
evidence, noclinical studies, nothing. It is unlikely that
epidemiology could ever develop that evidence.
The court, after working through all of this and
working through the absence of evidence on many cf
these subjects, essentially relied upon the statements
by the medical and other experts testifying in favor of
control, that yes, in their opinion there was a public
health risk. It was impossible to quantify the degree of
the risk. It was impossible to quantify—and I defy
anybody in this room to ever quantify—assuming that
cancer could be contracted, how many people would
contract cancer, what the degree of the severity of the
cancer would be, what the ability to cure the cancer
would be, what the susceptibility to therapy would be,
what the ultimate death rate would be, what the degree
of pain would be. The impacts of these kinds of health
risks are impossible to quantify.
One of the things I want to disabuse anybody in this
room about now is the game of trying to put costs and
benefits on these risks. You cannot do it. I would love
to cross-examine any expert who professes to be
capable of doing it, to simply say, "Well, this risk we
will throw out the window; we will not control because
it's only worth $27.50 and to control it will cost $36.42;
therefore, the costs of control outweigh the benefits of
control and therefore we should not do it." This is a
popular game that is going on these days. We have
other words for it, cost-effectiveness, whatever the
case may be. I suggest to you that it is impossible.
When all of the costs are laid out in the equation, and 1
can always find costs that nobody else can because
there are experts who have not been brought into the
equation, there are variables that have not been
brought into the equation. Then when you start cross-
examining you will say, "Well, yes, I didn't include
that. Yes, I did know about it but I didn't think it was
pertinent." It is always elements of risks, elements of
hazards that are ignored in this process. So I suggest to
you if you think we are going to play some nice little
game that we can plug into a computer and come out
with that lovely word "matrix." They love matrix.
Matrices. I go crazy over matrices.
You are crazy. You cannot quantify these things.
You have to lay it out on the table. You have to lay out
the areas you do not know. You have to do it honestly.
And then it basically comes down to, yes, there is a
risk—nothing more, nothing less than that.
And the court said, "We cannot quantify degree of
274
-------
/. V. KARAGANIS
risk. We cannot quantify the likelihood of degree of
injury assuming the risk takes place, assuming the
statistical probability does actually occur. But we have
identified a risk and we are going to cure it."'They went
the abatement route simply on that basis.
And where did they find their legal hook? They said
at the time, "We do not want to get into the murky
waters of common law nuisance," which really are not
that murky if you do a thorough analysis of it. "We are
going to rely on language that no longer exists in the
current Clean Water Act or the 1972 Amendment, the
language that said where a discharge 'will endanger the
public health.' " That was an interstate control
mechanism that existed in the '65 Act that no longer
exists, unfortunately, in the current amendments. But
it was essentially identical statutory language to the
public common law nuisance concepts that existed for
700 years --endangerment to the public health. You
find it in Reserve Mining; you find it replete in the
Milwaukee case.
In the Milwaukee case again, I ask you: Why am I
typically in favor of thejudicial process rather than the
rule making process in this area? Because I have seen
my ox gored more times than I care to think about by
legislators or quasilegislators that say, "We decided
against you." What evidence did you use to decide
against me? "Well, there was some statement by so and
so." What evidence did you use, what specific logical
progression of analysis did you use to reach the
conclusion? "Well, policy. It was policy." You cannot
get straight answers. There is a crude two word Anglo-
Saxon expletive that refers to what ultimately gets
translated into policy by those who want to defend
some of these decisions. But the fact is that there is no
substitute on the basis of defining where the process of
decision making went, on what it was based, where its
weaknesses are. I will be the first to admit that many of
the proposals that we put forward in litigation on
many of these things have empirical holes. I have yet to
come across a position yet that is totally free from
empirical doubt. The day it is, hire me on—I will do it
for free. The fact is that you have to honestly
acknowledge the areas where there is doubt, where
there are unknowns, and then develop it from that
level of analysis. See where your paths of alternative
action go. But deal with it honestly. Deal with it up
front. Then the beauty of litigation is that you and the
other side are forced to lay out your data, to lay out
your methods of analysis, and to expose your thought
processes to see how good you are, and how bad you
are.
One of the cases that I am dealing with now
concerns a little dam down the Mississippi called Lock
and Dam 26. And we hired the Corps, whom I love.
They are probably the best agency in the federal
government. Why? Because they are unlike EPA. EPA
is afraid of their own shadow. EPA will take its
convictions and run around worried that its
convictions will endanger its appropriations.
The Corps on the other hand is tough. They will look
you straight in the eye and say, "We're robbing the
public blind. We are going to destroy the environment.
And what are you going to do about it?"
We have not done a very good job about it, I will tell
you that straight up. The first thing the Corps told us
in this Lock and Dam situation was cost benefit
analysis (they are pros at that). Cost benefit savs the
economic benefits of replacing this lock and building
this huge new lock and dam far outweigh the economic
costs or the costs of construction which were a mere
S400 million. We questioned that. We heard some
comments that blew the Corps out of the water and
just showed how their economic data was false, and
just could not withstand critical analysis.
The Corps, when challenged, will always retreat to
its bastion of expertise—engineering, civil
engineering. "Well, at one time we did say the existing
structure could be replaced at a relatively cheap cost.
But now we find that it will cost more to fix up the old
structure than to build a new one. It will cost $500
million to patch up the old structure and only $400
million to build a new one. Common sense says that we
should build a new one."
What did we do? We hired the very man that the
'Corps and Bureau of Rec. and everybody else has
hired around the country to look at dam safety. He
came in, and he put in civil engineering terms, the very
same phenomenon that you face in the scientific fields
that you pursue, be it sanitary engineering or virology
or bacteriology, you name it—and we do it in the law
too. It is the phenomenon known as "black boxing. "It
is to put it all in a little black box and say. . . .pat the
lawyers on the head- a dangerous practice, I warn
you. . . .pat the lawyers on the head and say, "If you
only understood, you would appreciate our position.
But rely on us, please, because we're the experts. We
know what we're doing."
And I have found time and time again that the critical
basis of decision making can be distilled into language
and levels of analysis that even the most stup.id or
ignorant layman can understand. If you cannot do
that whether you want to or not, you have a major
275
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
limitation on articulating your scientific position,
because I defy you, whether you are a scientific
advocate or a scientific decision maker, to be the
source of all knowledge.
Many people have said, "Let's have an
environmental court or an environmental decision
making body made up of experts." Do you know what
would happen? You, being a civil engineer or you
name it, will find the only universal language that
assists somehow in the communication of this
sometimes is mathematics. But whatever your
specialty is, you are going to come across fields you do
not understand. You are going to come across fields
you are unfamiliar with, or if you do have a knowledge
of it, it is 30 years old. It is not current. But what you
are going to do, because your ego is involved, is to be
embarrassed about asking what seems to you to be
stupid questions but may be indeed fundamental
questions. And you will let slide statements that have
no empirical foundation but which sound intelligent.
But because you do not want to look stupid, you are
going to let it go.
I remember, just as an anecdote, the dean of the law
school I went to, who is a very famous man, has a habit
of speaking in nonsentences. He is a philosopher. He
would give dinner speeches. It was kind of like the
emperor who wore no clothes. He would give dinner
speeches that made not one ounce of sense. I would sit
there with a puzzled look on my face while everybody
would be nodding in admiration, oohing and aahing
over this wonderful speech. I could not believe it. I
would say after the dinner, "Did you understand a
word of what he was saying?" And the replies would be:
"Of course. Of course. Wasn't it the most lucid thing
you'd ever heard?"
I suggest to you that is exactly the process that
occurs when scientifically trained individuals have to
decide upon their colleagues. They do not want to be
embarrassed in front of their colleagues.
So I suggest again to you that the process is one of
detailed analysis brought down and filtered down and
distilled and analyzed from every perspective on the
level of common sense communication and
understanding. I suggest to you that the field of public
law nuisance is an effective tool in dealing with these
areas where Congress, in all of its wisdom, has
hopefully left us a little room. I suggest to you also that
when you are dealing with a proposal, whether it be a
legislative or an administrative proposal, make your
voices heard.
Now I do not know whether you folks in this room
are in the field of selling mechanisms or chemical;
involved in the destruction of pathogens. I do not
know whether you are in this room from the
standpoint of doing the analysis on it. I do know that I
went through a four month trial with some of the top
experts in the country—with Melnick and Berg and
Sproul and Cliver and Wellings and you name it. It
was like a convention. Walter Mack. And I learned a
hell of a lot more than I ever wanted to know about
viruses, a hell of a lot more than I ever wanted to know
about bacteriology. Ed Geldreich was here from the
agency. I suggest to you that I have tremendous
respect for some of those people and their willingness
to state what they do and do not know. I have a little
bit less respect for the methods of advocacy of some
others. But many of those people would say, "We don't
know certain things and the level of knowledge that we
have is imperfect. But what we do know is that we want
to establish (I think this is a historical principle) what
is known as the 'multiple barrier concept'."
I am just going to close with some basic facts in the
field of pathogens that we developed out of the
Milwaukee case. Anybody who says that an activated
sludge secondary treatment plant is an effective
mechanism for the removal of viruses does not know
what they are talking about.
I have been buried for the last five years in the
operating records of the city of Milwaukee. Assuming
that they took accurate tests, which they did not. . .
they have a little game that goes on down in
Milwaukee. I use Milwaukee as an example. This is
typical of most communities.
First of all, most of the sewage never gets there or a
good chunk of it never gets to the plant. Why? Because
they were building sewers where it is all gravity fed
down to the plant, but they are building them as we say
in the Midwest, ass-back wards. A 44 inch sewer would
be downstream and then a 72 inch sewer would be
upstream. A little common sense tells you what's going
to happen. You are going to have overflows. Well,
there are overflows all over the city. And this is not a
practice that you can turn back to the turn of the
century and say, "Oh, well, those poor dumb sanitary
engineers like Metcalf and Eddy and all those people
did not know what they were doing." Just so you
know, the people who designed the original system for
Milwaukee back in 1911 were Harrison P. Eddy and I
forget who else was involved. But they were prominent
sanitary engineers of the time.
We would get into situations where you would find
memoranda that said sorry, those of you who do
276
-------
J. V. KARAGANIS
or do not know the Milwaukee situation, they built
this plant called South Shore. South Shore was
supposed to relieve the problems at Jones Island.
Except that Jones Island had a little thing that is a very
profitable operation called milorganite. They do not
want to send their high nitrogen wastes down to South
Shore because that goes away in a sludge disposal and
it does not go to the milorganite plant. You literally
have the milorganite operation wagging the dog. It is
"sell high nitrogen milorganite" that the whole goal of
the Milwaukee systems gets to be.
But you would see a memo in the plant records that
would say, "Sorry. In last night's storm I had to send
you about a hundred million gallons. But I assure you,
it never got there."
And it does not get there. The relief gates handle
that problem. So that is one problem.
Now assuming it gets to the treatment plants, let me
suggest to you again on the basis of the records,
Milwaukee would know when the good days and the
bad days were. By Thursday of the week, you were
starting to get some bad days. The best day to test was
on Monday. The industrial loads had not built up to
the point where they were really affecting plant
performance. Monday was your best day to do tests,
because you would get some relief over the weekend.
So all the tests that went to the state were basically
based on Monday's results, particularly in the field of
bacteriology.
That was a grab sample that was highly
questionable as to how good it was and the
methodology for taking the samples was also suspect.
But assuming it was valid, you would have days at an
activated sludge plant, which is supposedly a model
for the nation, where the influent solids and the
influent BOD would be 280, 300. And the effluent
solids and effluent BOD would be 280, 300.
Maybe better.
So anybody could sit there, as one witness did, and
say, "Oh, I looked at those records. That's 90 percent
removal." It may be 90 percent average removal over a
three year period, but it certainly is not day-by-day
removal. That is what we are talking about in the field
of public health protection. We are talking about'the
use of geometric averages which I find to be a fraud.
A geometric average, according to the scientists,
takes into account the wide variability of test results.
The fact is that if counts are coming out of a plant at
300,000 fecal conforms per 100 ml after chlorination,
something is wrong. And when a geometric average on
the basis of monthly samples or monthly averaging
allows you to report to the state that you are in
compliance with EPA .standards, something is wrong.
I tell you that from a legal standpoint; I tell you that
from having investigated it from the public health
standpoint. If you are allowing instantaneous
discharges (which we think to be a low number,
parenthetically), something is wrong with your facility
and something is wrong with the system of regulation
which allows that kind of thing to happen.
I am suggesting to you that activated sludge plants
are not an effective mechanism for viral control and,
the way most activated sludge plants are operated in
this country on a day-to-day or hour-to-hour variation
basis, not a very effective mechanism for bacterial
pathogen control. I suggest to you that EPA, if it is
going to be talking about effective barrier protection
on public health hazards, has to be looking first of all
at effective collection. This is right now one of the
biggest disgraces we have going in this country, both
legally and technically, because everybody is ignoring
it.
Most municipal systems are operating as combined
sewers throughout the metropolitan areas of the
country. EPA for political and financial reasons, and
not for technical reasons, is ignoring it. I suggest that it
has to be faced and faced right up front; and Congress
is going to have to deal with it.
I am an AWT advocate- I am a fan of AWT—yet I
will hear some sanitary engineers saying, "What the
hell are we talking about AWT for when we're
ignoring all of the rest of that stuff coming out?" And
what they say makes quite a little bit of sense. To sit
there and build an AWT plant when you have flows
coming out of those overflows day in and day out is
pound foolish.
Now my approach would be to attack both
problems, not ignore one for the sake of the other, not
to say, "Gee, that allows us to ignore the AWT
responsibility." Attack the one and attack it honestly
and attack it openly and not sweep it under the rug like
the laws have been doing and like the administration
of the program has been doing.
In administration, you cannot lay this at the
administrative doorstep of EPA when the
municipalities do not want the problem addressed
because they know if it is honestly addressed, there is
not enough EPA money to go around and they want to
take the political and legal heat of controlling the
problem. The next aspect is the congressmen. The
congressmen do not want it addressed either. To me it
is probably a private scandal, but it is one that has to
277
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
be dealt with.
The next question is the flow control when it gets to
the plant. I do not know of a major municipal plant in
the country. I have to say that maybe some of the big
MSD facilities can do it because they have so much
reserve capacity. When you are talking about south-
southwest in Chicago which as a 1,750,000,000 gallon
a day capacity, they can get hit with a big flow slug and
absorb it within their reserves. But most plants get hit
with these highly variable flows. They do not know
what the hell to do with it. The honest operators will
tell you, "Well, there just went our sludge blankets out
in the Fake or the river," or whatever the case may be.
And anybody who says under those circumstances
that there is effective pathogen control is again
whistling "Dixie. "It just does not happen. So you have
to have flow equalization.
And then, if anybody can articulate for this
audience or me after the meeting tonight how the
biological process within an activated sludge plant
actually kills—kills, inactivates, whatever term you
want to use, not just takes out by way of the
sedimentation process but actually kills—viral
pathogens (I do not know about the bacterial patho-
gens), then he ought to get a Nobel Prize, because
the process is not understood. The way I understand
the process to be from all the measurements I have
seen, at best in an activated sludge plant, is that you are
sedimenting the stuff out and then you can hopefully
handle it through some kind of sludge treatment
process. But you are not killing it by way of the
activated sludge process, and it is going to go out
into the receiving body if the activated sludge
process acts up.
I am suggesting that there ought to be at treatment
plants, once collection and flow equalization are
attained, a third level of treatment which involves
some physical process, either physical or physical-
chemical process, for balancing out, evening out, the
highly variable performance of activated sludge
sewage treatment plants. I suggest to you that the
incremental costs of doing that are not that significant.
I have been surprised to find in a number of cases that I
have worked on around the country that people are
amazed when they talk about the total package for an
AWT and then they are told what increment of that is
secondary or activated sludge or biological. It turns
out that well over 90 percent is typically in the
secondary, and the so-called filtration or physical
treatment is a relatively small (still a significant but a
relatively small) increment.
So I am suggesting that we have much to do, much
more. We have to do it by way of honest recognition of
what the problems are. We cannot be going around the
country saying that there is not a problem. I suggest
that, those of you who feel that there is a problem, you
have a professional responsibility to speak out,
because the state agencies who are typically tied... .the
tail that wags the dog in the whole process is the
construction grants program. I suggest to you folks
that the state agencies are like little pigs at their
mother's side. "Who's going to give us the money?
What do we have to do to get it?" They'll do
anything—lie, cheat, steal, whatever the case may be.
If the blessing from on high says, "There shall be no
disinfection," we do not need disinfection. No
problems. The public be. damned. There is no
epidemiological evidence that anybody ever got sick.
Listen to a good epidemiologist tell you the horror
stories about reporting mechanisms for most
pathogenic or enteric diseases. We do not have a very
effective system. The question that really has to be
addressed is: Are we concerned about enteric
pathogenic organisms? Are we concerned about
transmission by the water route? This is a system of
protection that was developed 80 years ago, and has
worked, I think, very effectively to date. It needs
addressing because the problems of population
concentration are such that those lucky assimilative
capacities that we have had are being sorely taxed, so
we have to be more rigorous in our point source
discharge protection.
Is it worth it to protect public health? If it is not
worth it, then I think that the people who are advo-
cating it ought to clearly state: "We are willing to
take the risk. We are willing to take the gamble
on you, the private citizen's health." I do not
think that is being done today, and I do not
think the advocates of public health protection
are making their voices heard.
All I get out in the field talking to the state agencies
are, "This is the route we are going. EPA from on high
has disavowed the need for disinfection and disavowed
the need for effective control of pathogens. This is the
route we are going because it is most cost effective."
Sure it is most cost effective. It is less dollars out of
your pocket. Whether it is public health effective or
not is a question I hope we never have to learn the
answer to.
So I am suggesting to those of you in the audience
who are public health professionals that you get about
the business of beating the clarion call or whatever the
278
-------
J. V. KARAGANIS
case may be by making the public aware that there is a
serious problem here that is not being addressed. 1
would suggest to you that all the law suits in the world
will not help that problem. All the law suit will do is
help focus attention on the problem. But until you, as
professionals, get the word out, then the political
sphere is not going to react to it. And I suggest that
maybe the next time around when the Clean Water
Amendments of 1983 come around, or whatever the
case may be, we are going to have a logical process for
addressing the question of pathogenic public health
protection that can be done on a nationwide basis and
not just on the basis of a long, drawnout fight in the
case of Illinois versus Milwaukee.
I was asked tonight if I would open myself up to
questions, and I foolishly said yes. If there are any
questions about my preaching to you_tonight, 1 will be
glad to open myself up. Anybody?
QUESTION: Under the present judicial system, is
it possible that an individual in a regulatory agency
can be sued on a case, for example, a cost effectiveness
statement made by a technical agency regardless of
whether it is federal, state, or what have you?
MR. KARAGANIS: The question was, and 1 am
going to paraphrase, is there potential for individual
agency official responsibility for engaging in and
making the wrong decision in one of these areas of
cost-effective analysis? As the Corps would say, "Glad
you asked that." They are wonderful. They have a
public relations campaign that will blow your mind.
With respect to potential liability of public officials for
making the wrong judgments, this is an area that is
becoming increasingly pursued in the field of dam
safety—a small thing called the Teton Dam failure.
The Bureau of Rec.wisely settled like mad. They went
out there with a pocketful of cash as soon as they could
and were settling claims left and right. So I do not
know what will ever come out with respect to the
Teton failure.
But I am dealing with one now. 1 was telling Frank
at dinner that 1 was involved in Home State Mining
controversy out in South Dakota a few years ago, and
we had successful results in that. But while I was there,
EPA was funding two sewage lagoons in the town of
Spearfish, South Dakota. If you do not know
Spearfish, that is the home of the Passion Play.
Anyway, Spearfish had these two sewage lagoons
and the local farmers out there—please rely on the
common sense of the people in this nation—the
farmers out there would say, "Nope, that thing's never •
going to work. I've tried to build stock watering dams
out here. The ground soaks it up just like that."They
told the local city officials; they told the local
engineers: they told the South Dakota Department of
Environmental Protection; they told EPA. EPA in its
infinite wisdom said, "Oh, yes, it will work. I have the
little black book here that says it will work.""No, it's
not going to work.""Oh, yes, it is. As a matter of fact,
we'll load it with bentonite. We'll pave it over with
bentonite and it will hold."
Now as it happens out there in Spearfish, they have
a thing called gypsum that just is throughout the land.
Gypsum is like putting sugar in your coffee. The
minute that water hits it, it dissolves. It is just
incredibly soluble. And lo and behold, if anybody
bothered to check, the ground underneath the sewage
lagoons was loaded with gypsum.
I remember being out there and ranchers out there
would say- "My God, we are seeing a trickle of sewage
coming up about a mile and a half away from the
sewage lagoons." EPA would say, "No, that is not
sewage—not from this facility." Well, in the fall of last
year, or maybe the spring (I forget which), suddenly
what the farmers had been saying all along happened.
The bottom of the lagoon fell out—just literally fell out.
The entire lagoon is emptying now and it is coming up
on my client's property a mile and a half away. It is
contaminating a considerable extent of the public
water supply in the area and the stock water supply of
the area.
Somebody asked me, "Is there potential
professional liability on behalf of those officials?" I
answered, "Yes, there is." I think that one of the things
that EPA is going to have to start doing is a review,
before they pass their hands over these construction
grants so blithely, of what hazards have been created
here and, if there is any risk, be it to property damage
or public health or otherwise, state that there is the
perspective of potential liability. I cannot say it is
going to occur. I would suggest to you that that kind of
analysis has to be done.
QUESTION: On the same order, if an industrial
official can be hit personally as in a pollution situation,
, why, on a legal basis, can't the reverse be true? For
example, a chemical company with bad water, dumps
a big load. He gets hit by EPA both as a corporation
and also as the president and chief executive officer.
Cannot, because of a precedent, the same ruling go
back to the regulatory agency?
MR. KARAGANIS: Well again, I want you to
understand something. One of the EPA enforcement
tools over the years has been to come riding in and say,
"Thou shall stop polluting and that will be an
enforcement suit." And in the next breath give a grant
279
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
to stop it. That was part of the settlement we discussed
earlier. I only wish I had as thick a wallet as that. I
could have been in good shape.
But having given the grant to stop the pollution, if
that does not stop the pollution, if that causes a
problem, then I think you are right. Now in this
community of Spearfish, for example, these are farm
folk. They are not sewage specialists. Their response to
this whole problem has been. "We only did what EPA
told us to do."
They happen to have a pretty savvy lawyer
representing them—the same guy who represents
Home State Mining Company. He is seriously
thinking of going against, what we call "up the
line," that is, going against the state officials '
who approved this thing; going against the engi-
neering firm who designed it; and going against
EPA officials who recommended it and told
everybody that it would work.
So this potential liability is there. It has not been
demonstrated yet. I do not want you to run out and
buy insurance policies tomorrow, but 1 suggest to you
that it ought to be a major consideration in the design
of your review programs, and it ought to be a
consideration that your legal counsel should get
involved in as a precautionary measure.
One other thing that 1 find to be the case is that most
agency people look upon lawyers as pests. Get them
into your programs. I realize they are a pain. First of
all, most lawyers are lazy and they do not want to
understand, but get them to understand what you are
doing. Educate them as to what you are trying to
accomplish, what the elements of your program are,
and then use them as your own staff. Get them to check
over what you are doing. Do not just send overa grant
application or send it over to regional counsel's office
or whatever and say, "What do you think of this?"
because most of those guys do not know what is going
on. They never have and-they really do not want to
know what is going on. But the ones who are trained,
the ones that you will take the time to train, you will
find they will be conscientious and will do a damned
good job for you in covering you against the very
problem we are talking about. If asked to do it, they
will be the ones who will ask the hard questions. You
have to do that internally before somebody like me
gets hired to do it externally.
QUESTION: Probably the best thing that is
happening then is for EPA to delegate to the states,
just like it just did to Illinois.
MR. KARAGANIS: 1 just responded to a question
like that recently. The question asked if probably the
best thing was for EPA to delegate it to the states. I
responded to that question with this answer which
applies to this question as well as one 1 recently had. If
you believe that, I have a used car outside in the
parking lot that I would like to sell you. You have
certain responsibilities you cannot delegate. Whether
it be under the 201 program, 401 program, whatever
lingo you want to use, 208, etcetera, 1 do not think you
can delegate liability protection. You have an
oversight responsibility here. You are dealing with
your involvement of federal trust. You are the
conservator of the federal law both in terms of grants,
in terms of permits, in terms of enforcement, in terms
of standards, whatever the case may be.
I have this little problem out in Las Vegas, Nevada,
right now where EPA is running around like mad
saying it is the state that did it. It was not the state that
did it. The state is going around like mad saying EPA
told us to do it, and they are probably right. They did
not do it the right way, but you simply cannot get away
with it by saying, "Thou shall follow the law." You
have a responsibility to make sure that whoever you
delegate it to actually does indeed do the job. I am not
telling you anything that the lawyers down in
Waterside Mall do not know about, because they write
the regulations and that says, "Thou shall not delegate
withoul checking up lo see if il is done." You cannol
gel away wilh il in lhal process.
1 really feel that many of these problems are coming
up because there is lhal wholesale delegation without
adequate oversight. 1 want lo say lhal any of my
remarks here tonight are done with the recognition
that there are tremendous manpower problems. You
have enormous programs lo run with loo few-
individuals. There oughl lo be some kind of
mechanism for increasing Ihe efficiency of indivi-
duals, developing safeguard check lisls which move
these things through, hopefully at a rapid adminis-
trative rate, but at the same time making sure that
not too many mistakes slip through the cracks.
QUESTION: Have you done anything to
challenge the coliform standard? This has been coming
up in Ihe conference—whelher or nol Ihe slandards
are valid for environmenl and wastewater.
MR. KARAGANIS: No, 1 have not fora very simple
reason. My ability to do anything is severely limiled by
Ihe same thing that limits most of you folks—money. 1
have to have experts behind me; 1 have to have
resources behind me. I have to have somebody lhal
pushes the butlon in Ihe back of my head and says,
"Go do it." Until I can have that resource, I just deal
with the specific problem, lam familiar wilh the field. I
280
-------
J. V. KARAGAN1S
think it ought to be challenged.
Let me tell you about the so-called 200 geometric
mean. In checking with Ed Geldreich, who I am sure
has had a hand in setting the 200 geometric mean, he
tells me the reason 200 was selected: "Well, at 200, we
found a 7 percent likelihood of bacterial pathogen
occurrence in the samples we analyzed." Nobody ever
attempted to correlate viruses. Those attempts that
have been made have shown there is an inconsistent
correlation—that it is not a positive co'rrelation. But at
least the bacterial pathogen, is 7 percent. I do not
know how many samples Ed did to obtain this. But
when talking to him and hearing his testimony on the
stand, he agreed that the geometric mean was
inappropriate because the geometric mean would
mask individual slug discharges. I think the use of the
geometric mean is an illusion. It is almost a fraud to
say to the public: "Okay, here at our bathing beaches if
. you go out swimming, one day out of the month
we made it." The public buys that. I really think
that it has to be laid out. Most newspaper
reporters buy it. Who cares — geometric mean,
arithmetic average, whatever the case may be.
They just see a bottom line number. You tell
them you meant it, and they think they are home free.
I think it should be attacked. I think it should be
attacked professionally within the agency. I think one
of the things that might be helpful within the agency is
the articulation of responsible dissent. Where you
have a policy that is up top, there ought to be an
ombudsman mechanism or whatever else that gives
responsible scientific dissent a mechanism for voicing
its position so that it can ultimately be made public or
somehow be made available to the public.
DR. REYNOLDS: You talked about public health
risk. It seems to me that what you seem to be telling us
is that, with the current standards, we have accepted a
given level of risk which you are not willing to accept.
It seems to me this whole issue seems to be: What level
of risk is the public willing to pay for in terms of a
public health nuisance?
MR. KARAGANIS: Yes, sir. Would you kindly tell
me how many pathogenic bacteria I am exposed to at a
level of 200 fecal coliform per 100 ml?
DR. REYNOLDS: I understand what you are
•saying. I cannot give you that answer, as you know.
MR. KARAGANIS: Not only do I know, sir, but
the public has never been given that answer. When the
public is told the standard of safety is 200 fecal
coliform per 100 ml, they are not told that they have no
idea how many pathogens are present or the fact that
at any series of readings at 200 or below, any number
of bacterial pathogens may be present or the fact that
considerable quantities of viral pathogens may be
present. The number has had no relationship to the
viral aspects of public health and has had little
empirical relationship^ the standard of bacterial'
pathogens in public health. I suggest to you that from
the standpoint of fecal coliform, I am just telling you
as a lawyer what the experts have told me.
There is a guy running around saying, "It is all the
birds." It is the birds or the fish or whatever the case
may be. I tell you, we had big charts in Milwaukee
saying that any fecal coliform found anywhere in Lake
Michigan could come from 37,000 different sources,
only one of which is human. I am told again within the
limits of knowledge there is this (Geldreich has
suggested and others have suggested) fecal
coliform/fecal strep relationship, whatever the case
may be.
But if we are looking for scientific analogies or
methods of indication and if fecal coliforms are not
satisfactory because of the possibility of interference,
if there can be other indicator organisms that are more
precise, so be it. If there can be other indicator
organisms that would match in pairs between fecal
coliform and fecal strep, so be it. But then once we do
that, I think we ought to have a standard that says "as
close to zero human fecal contamination as possible."
I do know that with current technology, and not at
tremendous cost, most municipal sewage treatment
plants can produce an effluent of zero fecal coliform.
And if all of the sewage gets to the sewage treatment
plant and if the sewage treatment plants can produce
it, we should not find any fecal coliform with a fecal
strep human source relationship in any public water.,
We will find fecal coliform from the ducks, and the fish
and the birds and whatever; but we will not find it from
humans.
Again what I am told is the diseases we are worried
about and the fecal coliform which we use as an
indicator is from human enteric sources. If we collect
and if we treat properly, we should be willing to say
that fecal coliforms or indicators of pathogenic
contamination from human sources should register
zero, because right now nobody can say beyond that
exactly what the hell you are getting. In the virology
field, you folks have to get your act together from a
policy standpoint.
In the viral field, it is a massive fraud to the public to
use that kind of an indicator and say, "It is okay." It's
just not.
281
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
There is one regulation that is being inconsistently
used around the country in terms of public health and
safety. We have an Illinois Department of Public
Health regulation that says: "No more than 20 per 100
ml for two days running." You would not believe the
number of violations. I do not Jcnow what its empirical
base is, but you would not believe the number of
violations that hit at that standard that just slide right
by the geometric mean—just slide right by. We are
keeping our beaches open when they may be subject to
pathogenic contamination.
I will buy all of the business that, yes, it has to get in;
the dosage has to be properly adjusted; to establish a
minimum infective dose for typical individuals,
susceptible individuals, the whole business. What if it
gets in and causes a disease? So what? Somebody is
discomforted for a few days. You can go through all of
those arguments, and 1 can find countervailing policies
as to why you should protect because the risks are so
great and the cumulative or additive effects of those
diseases can be so debilitating and even life
threatening.
Lawyers have developed with some justification—
they have an axe to grind. They are being paid to do it.
The only way I can be effective is, as Frank suggested,
as a translator— someone who is articulating for
laymen and assists experts in articulating their
position. But it has to be the experts that are out in the
forefront. It has to be the experts who are publishing
the documents. It has to be the agency who says, "Cost-
effectiveness—let's recognize it for what it is."
We are waiting out the Milwaukee situation right
now. Milwaukee said, "We will agree to abate our
overflows."The judge and everybody thought that was
great. "We will abate them to that level which is cost-
effective." What level is that?"We do not know." What
level do you think it is? "Just so we cannot find any
observable effect." Does the fact that you cannot find
any observable effect mean that you are not causing
any harm? "Well, no."
One of the things that has happened in the
Milwaukee situation, concerns the Howard Street
water intake. Can you imagine-living in this situation?
I cannot. I find it mind-boggling. The expert from
Milwaukee called up the water treatment plant
thinking he was going to find evidence of how clean
Milwaukee's water is. And one of the big problems we
have in this field is the damn water treatment plants
are still doing total coliforms and not doing fecal
coliforms. And the sewage plants are doing fecal
coliforms and not doing total coliforms. Try and find
out that statistical relationship sometime in a cross-
examination.
Now you have to get some consistency in the
methodology that you are demanding as well. 1 say do
them both. If you are checking your water, do them
both and do them the right way as well. One of the
things you are still doing is following this quick and
dirty method of the membrane filter test which is
causing so many problems. Milwaukee found it was
unusable. Milwaukee found it was giving them low
numbers when the MPN test was giving them
astronomical numbers. You have an honest
bacteriologist in the Milwaukee sewage lab saying,
"Let's stop using the M F method. It's giving people the
illusion that the waters are safe," because the
Milwaukee Health Department continues to this day
to use the MF method. The sewage people are off
doing the MPN test and finding incredible numbers. I
will not go through the long list of interferences which
many of you are familiar with.
The M F test is fine if the water is clean. If the water
is dirty, it is not good. So you use it to find if there is
dirty water? Let's get some consistency. I always look
for the thread of common sense that is running
through these things and sometimes I do not find it.
We are all working towards the same goal. Let's sit
down and get some plans together.
But this is a problem, of logical consistency and
planning. At Milwaukee, do you know how Howard
Edner responds? Everytime it rains, they dunp
carloads of chlorine in the intake. They just dump like
mad.
Now people tell me, and Milwaukee told me, "My
God, you can't use chlorine because chlorine causes
carcinogenic compounds to form which may cause a
problem." And indeed it may. One of the ways to do it
is not assault the treatment plant, the water intake
plant, with pathogenic loads so that you have to ase
that much chlorine.
One of the things that we have found through
sewage treatment and otherwise, is that the clea icr
you make your effluent the less chlorine you need to do
an effective job.
Believe me, if there is an effective way that has been
established and will be accepted by EPA to jse
something other than chlorine which will resolve the
questions that have been raised about chlorine, I am
all for it. But we looked because we wanted to impose
the most effective mechanism for disinfection in the
Milwaukee situation. The only one that all of our
people told us that we could have some certainty on at
282
-------
J. V. KARAGANIS
that stage of the game (that was a year ago) was
chlorination. If you can come up with an effective
thing for large treatment plants to use other
methodology, fine. Let's get into it and let's resolve the
questions about chlorine. But even with chlorine, you
get the organics out and you get the solids out. You
convert the ammonia and you are in good shape with
respect to being able to do an effective job of
disinfection with minimal development of chlorine
compounds that could be potentially carcinogenic.
All the questions you raise are legitimate ones. But
you put them out on a table and subject them to
analysis and you will find tnat many 01 tne things tnat
the experts are telling the public just are misleading as
hell right now.
Anybody else? Yes, sir.
QUESTION: One word in favor of using the
geometric mean. It has evolved not to evade standards,
but because bacterial populations grow according to a
geometric rate. Therefore, in order to find the correct
average, to get a straight line on a graph, you plot it.on
a log-log graph and take your midpoint. So there was
no intent by any biologist to evade regulations.
MR. KARAGANIS: I know that. But you see,
laymen take your tests. Maybe your test and your
geometric mean may have perfectly valid scientific
rationale and may be useful within thecontext of your
scientific analysis, but a layman will distort it. A
layman will take it and run with it. A lawyer will take it
and run with it and say, "My God, I have a test I can
use." He puts it in and it is misleading. Most public
health people that I have talked to want to know about
the worst case. They do not want to know about
averages, because averages are misleading. They want
to know about the worst case—what is the worst
potential case. And "the geometric mean from the
standpoint of receiving waters does mask. It may be
fine for the laboratory basis, but it masks worst case
situations in receiving waters. That is the only point I
am making about geometric mean.
And if you tell me that you get a high result in the
bacteriology and it may be because of a lab error, it
may be because of anomoly as to the way that bacterial
grew and there really were not X thousand bacteria out
there, I will buy that too because the tests are crude,
the tests are ineffective, the te'sts very often will give
you erroneous results. But from a public health
standpoint, recognizing all those limitations, if the test
shows up with a bad number, I want to be able to
prevent, I want to be able to see that the situation does
not develop. And that is the kind of protection we have
to work for.
If there can be more effective tests than MPN MF,
whatever the case may be, so be it. If there can be
effective viral tests, so be it. But within the crude limits
of knowledge we have today, let's maximize public
health protection.
283
-------
APPENDIX B.
SYMPOSIUM RAPPORTEUR
Dr. E. Joe Middlebrooks
Utah Water Research Laboratory
Utah State University
Hogan, Utah
One of the best things about the symposium, in my
opinion, has been the transfer of information and the
audience participation. When I first came here, I
figured this was too large a crowd. You will not get two
people to get up and say anything besides George
White.
But it has been far better than that. I have seen
numerous people get up and make comments. I really
think that is good and I am glad to see that Al
incorporated this into the program.
I guess one of the distracting things is the
incompleteness of several of the studies. This is
invariable when you have projects that are in various
stages of completion. You cannot expect not to find
some of these things. When you are organizing a
symposium you cannot wait until everyone finishes. If
you did, you would never present anything.
There was a limited amount of practical, how-to-do-
it kinds of things presented too. I heard several people
comment about that during the symposium. I think
perhaps the next one, after some of these studies are
finished, will solve that problem. I certainly hope so. I
do not mean to imply that there were not any practical
presentations—certainly there were some. But I think
that was probably one of the limiting factors.
I think the EPA Research Program is addressing the
problems that we face. They found some solutions, but
I think in most cases though, they have asked more
questions than they have answered. That is not
unusual either for research projects. If anyone in this
room has ever conducted a research project, they
know you wind up asking more questions than you
ever answer. But as long as the intentions of the agency
are to try to continue to answer these questions, then
that is fine. We have accomplished our purpose of
being here..
I will make a few specific comments about
some of the questions that came up during the
symposium. The issue of using hypochlorite was
mentioned, and people 'talked about hypochlorite
generators. There is a simple technique for small
systems that appears to work very well, at least in
the one instance where I have taken any data.
This is simply a tablet type feeder where you pass
water over tablets in a tank and allow it to go into a
contact tank. The only one that I mentioned that I
am familiar with is located in a small town in New
Mexico, and they are chlorinating lagoon
effluent. It does a good job, and they have very
low coliform concentrations coming out of the
lagoon system which they then filter. It is an
excellent quality effluent following that.
That is something, if you are dealing with small
systems, you might keep in mind. I am sure it is
not the most effective use of the hypochlorite; but
nevertheless, it does not require a great deal of
talent to keep it running. If you see tablets in
there, it is running. Most operators are capable of
handling that.
There is a great deal of discussion about
regrowth of coliforms. I do not think anyone
reached any conclusions as to what was actually
happening.
I think one of the biggest problems that I
detected, in addition to the few things I
mentioned before, is that we do not know what we
are measuring half the time. We talk about the UV,
we talk about the coliform, we talk about the
various residual techniques. Now maybe in the
future we will be able to solve that problem, and I
saw some indication of that during the course of
the meeting. People are beginning to talk to each
other and there were some comments by the
standard methods people and the other
standards organizations. I hope we can get
around to that -because half the time we are not
talking about the same thing. Someone disagrees
and yet, he does not even know what he is
disagreeing with. I think that is another issue that
should be pursued very actively in the EPA
Research Program.
The fish bioassay was also presented as a
284
-------
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
method of evaluating the impact of the various
by-products of disinfection. It is certainly a good
technique. There is nothing wrong with fish
bioassays. But you cannot tind many people in
this country, including most of the scientists, that
can run a decent fish bioassay. If you have ever
tried to keep those little beasts alive, you will
know what I am talking about. They just get all
kinds of diseases and they do not pay much
attention to what you do. You can'put all kinds of
streptomycin in there and everything else, and
'they belly-up on you. So the fish bioassay is a
tough thing to run.
Maybe we need something simpler for the
operator to detect these types of things. I know
one thing that may or may not apply, but it does
not sound like a bad idea to me, and that is the
simple bottled bioassay technique that has been
used in eutrophication studies. It is about as
simple a test as you can get with the biological
organism. It does have the advantage of
simplicity, and I know many operators that can
run it. Plus or minus 50-60 percent is not bad
considering most bioassayists that you run into. I
think that again is something that might be
considered in the future research program.
Last night we heard a great deal about
standards and the risks that are involved and
associated with certain levels of fecal coliform, or
whatever we are measuring. Regardless of how
clean we make sewage or any other waste
material that we are destroying, there is going to
be some risk associated with the discharge of that
product. Contrary to what anyone wants to say,
there is some risk just sitting in this room. We
have to decide what risk we are willing to accept. I
think to run out and do something simply
because you can do it is foolish. I do not think that
is what was being said last night, but I could
certainly get that implication occasionally. I do
not think that is very smart. I think it is extremely
foolish in terms of fiscal responsibility. We should
decide what reasonable risks are and proceed
from there. I know a lot of people do not like that,
but when you start looking at some of the other
problems in life, it certainly has to be dealt with on
a risk-analysis type basis.
We might take the same approach as the people
who are dealing with the problem of the dam now.
They have been working basically with stochastic
type processes, and it is something that we might
be able to apply to disinfection and control of
diseases. I am not sure, but it would be worth a
shot.
There is obviously a great need to do a lot more
work concerning the side effects of the various
type of disinfectants. I think if nothing else came
out of this morning's session, it certainly shows
that we do not know much about what we are
doing and much about what the side effects are of
any of the disinfectants. We know more about
chlorine perhaps than we do the others, but I am
not sure we know much about any of them when
you really get down to a basic situation. I think
that is something that we could definitely
investigate and improve. I know there is work
going on in that area. I hope there will be more.
The cost 'figures obviously need to be
coordinated. I think we are all at fault when we
present cost figures. We invariably wind up
leaving out things because it is convenient or we
want to put things in and make somebody else
look bad. We go through a lot of procedures and.
although not always intentionally, we still make a
lot of mistakes. I think it would be nice if the
manufacturers would be the first. Don't wait until
somebody beats you over the head or
embarrassses you. Be first and lay those figures
out. I think you were today and yesterday in trying
to get your figures out, and it was certainly an
eye-opener to me because I have seen basically
the same kinds of figures that Ed Opatken had
presented before. It is nice to know that things are
being improved and they are coming down in
cost. I think the best thing to do is to contact
someone with experience. If they know what they
are doing, they can certainly help you a lot when
you start looking at costs.
Two other little quick items and I will quit. I
would like to caution you against using some of
the general formulations that you saw. They were
not presented as general formulations. They were
presented as specific formulations. For example,
Reynolds' and my work certainly should not be
used indiscriminately. You have to use it
carefully, and the same goes for any of the other
formulations that you have seen. Make sure you
know what you have before you start using them.
You can get in more trouble than you would care
to have.
I think the routine monitoring techniques,
which I alluded to a little earlier, need to be
improved and some of the automatic feedback
control type devices need to be improved. Once
285
-------
E. J. MIDDLEBROOKS
those come into play, I think we will have much
better acceptance of some of the newer
technologies.
One item that was briefly mentioned, but very
little was said about it, is our energy crunch. I
think if things stay as they are now and the Arabs
continue to jack up their oil prices and Congress
continues to drag their feet in coming up with
some decent energy policy, that is something that
is going to have a tremendous impact on
disinfection and what process is accepted.
286
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/9-79-018
I. RECIPIENT'S ACCESSION-NO.
4. TITLE ANDSUBTITLE
PROGRESS IN WASTEWATER DISINFECTION TECHNOLOGY
Proceedings of the National Symposium, Cincinnati,
Ohio, Sept. 18-20, 1978.
5. REPORT DATE
June 1979 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
Albert D. Venosa, Editor
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Municipal Environmental Research Laboratory--Cin, OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
10. PROGRAM ELEMENT NO.
1BC611, SOS #3, A/22
11. CONTRACT/GRANT NO.
Inhouse
12. SPONSORING AGENCY NAME AND ADDRESS
s ame as ab ove
13. TYPE OF REPORT AND PERIOD COVERED
Final - 9/18-9/20/78
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Contact: Albert D. Venosa (513) 684-7668
is. ABSTRACT
symposium brought together scientists, consulting engineers, municipal
design engineers, and various technical, state, and local government officials to
listen to presentations on the latest developments in the field of wastewater dis-
infection and to actively participate in floor discussions on the data and interpre-
tations presented therefrom.
Rapid progress is being made in the field of wastewater disinfection, but much
more work is needed before a design manual can be formulated. It appears that con-
siderable savings in chlorine usage is possible with a well designed, optimized
mixing and contacting system. Dechlorination with sulfur dioxide is cost-effective,
but not with activated carbon or holding lagoons. Disinfection of well oxidized,
filtered secondary effluent is best achieved with a bubble diffuser contactor and is
independent of contact time. Total costs of ozone disinfection appear to be twice
the cost of chlorine. Ultraviolet light is finally being recognized as an extremely
effective alternative disinfection process, with costs ultimately promising to be
competitive with chlorine. Chlorine dioxide is somewhat disappointing compared with
chlorine on bench-scale analysis, but pilot testing should reveal the true effective-
ness. Very interesting preliminary results were presented on nonvolatile organic
by-product formation by chlorine, ozone, and ultraviolet light. This report is of
interest to consulting engineers, municipalities, and research scientists pursuing
-pjg-seargLh -in- d-i-5-in-f Action- technology. --
17. KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Disinfection, Disinfectants, Chlorine,
Chlorination, Ozone, Ozonization, Ultra-
violet Radiation, Viruses, Microorganism
control (sewage), Coliform bacteria,
Chlorohydrocarbons, Meetings, Lagoons
(ponds), Wastewater,.Cost effectiveness,
Lie ch lorination
Chlorine dioxide
Photoreactivation
Aftergrowth
Ozone contactors
13B
06M
13. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
295
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
287
« U.S. GOVERNMENT PRINTING OFFICE: 1979-657-060/5311
-------
-------
-------
c o
3 f>
M
o
o
VO
o
53 :
^3 3
Do?
• . O CD
-» 2 o,
^ CD
li.
CD •?
"> X
r-fc CD
C
3 2
0 0'
CD CO
a S
to
CO
O
3
CD
O-
O
Co
O
cs
>
z
m
O
C
>
r
O
TJ
TJ
O
JJ
H
C
z
H
Y EMPL
O
-<
m
33
TJ
m
Z
>
r
H
_x
^s
Tl
O
JJ
TJ
JJ
<
>
H
rn
C
(A
m
(0
GO
O
O
O
Tl
T|
n
>
r~
CD
C
0)
z
m
CO
0)
o
o
(-)
5'
13
fl)
.—
O
zr
5'
45268
m
Z3
<
^
0
1
CD
3
OJ
^3
CD
C/)
CD
Q}
O
IT
B"
ormation
o
CD
^
S"
g
o
CD
O
*'
3D
esea
-1
o
zr
0)
13
Q.
Develop
w
3
CD
^j
c
w
m
z
<
5
O
Z!
^
m
~z.
H
>
|—
TJ
JJ
O
m
0
H
O
>
O
m
CD
0
0
7T
0)
TJ
CD
O
T""
(B
T|
O
C
r-*-
^
0
o>
w
c
0)
m
z
<
3
O
Z
7
^
m
z
5 H
5 >
> r
CO TJ
CO -n
01 g
H
rn
n
TJ
o
en
H
>
Q
m
>
Z
O
"n
m
m
0)
-n
O
JJ
91
r*
(D
CD
m
Z
n
------- |