-------

-------
                                                      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

-------