600281092
Technical Report  No.  251
October 1980
                CHLORINE DIOXIDE FOR WASTEWATER DISINFECTION:
                          A FEASIBILITY EVALUATION
                                     by

      Paul V. Roberts, E. Marco Aieta,  James  D.  Berg,  and  Bruce M.  Chow
                                Supported by

                 Municipal Environmental  Research  Laboratory
                    U.S.  Environmental Protection Agency
                           Research Grant R-805426
                     Department of CIVIL ENGIOSrEJEJiRIlsrG
                                     STANFORD UNIVERSITY

-------
                 Department of Civil Engineering
                       Stanford University
                   Stanford, California  94305
          CHLORINE DIOXIDE FOR WASTEWATER DISINFECTION:
                    A FEASIBILITY  EVALUATION
                                by

Paul V. Roberts, E. Marco Aieta, James D. Berg, and Bruce M. Chow
                     Technical  Report  No.  251
                           October  1980
                 This research was  supported  by

           Municipal Environmental Research Laboratory
              U.S.  Environmental  Protection Agency
                     Cincinnati, Ohio  45268
                     Research Grant R-805426

                         Project Officer
                         Mark C. Meckes

-------
                                   ABSTRACT
     Chlorine  dioxide  was  compared  with  chlorine  for  the  disinfection  of
wastewater  in  laboratory  experiments.   Disinfection with chlorine dioxide  was
also demonstrated  at a full-scale  wastewater treatment  plant.   Criteria  for
comparison  included  coliform  kill,  inactivation of poliovirus and other  indi-
cators, and formation of halogenated organic  byproducts.

     Laboratory  experiments  were  conducted  using  a  4-liter  batch  reactor
characterized  by intense  mixing.   The  experiments were conducted according to
a full factorial design, with mass  dose of disinfectant (3  levels) and  contact
time (3 levels) as independent variables.  The fractional survival of coliform
bacteria was  correlated with the product  of  disinfectant residual times  con-
tact time.

     In  general,   chlorine  dioxide  accomplished  a given  fractional  kill  of
total coliforms with a  smaller product (residual  x time)  than  chlorine.   For a
given contact  time,  the residual required  to achieve a given fractional  kill
of coliforms was 2 to 70 times smaller for chlorine dioxide  than  for chlorine.
The disinfectant demand is greater  for chlorine dioxide than for  chlorine  in a
conventional activated  sludge effluent, but less  than for chlorine in a nitri-
fied, filtered effluent,  when residuals  are expressed on a  mass  concentration
basis and  conditions  are  chosen to  result  in equal  disinfection performance.
Considering  both   required  residual and  demand,  the  required  doses  of  the
disinfectants  were estimated  to  satisfy three  assumed coliform disinfection
levels with two  types  of  effluents:   conventional activated  sludge and  fil-
tered, nitrified activated  sludge.   The required mass  doses of the disinfec-
tants  were  approximately equal  for  treating  conventional  activated-sludge
effluent—approximately  2 mg/1 to  satisfy a  standard of  1000  total coliforms/
100 ml, 2.5 mg/1 to satisfy a 200 coliforms per 100 ml  standard,  and 8  mg/1 to
satisfy a standard of 2.2 coliforms per 100 ml.   The  required  dose of chlorine
was  approximately  2  to 10 times  greater  than  that  of  chlorine dioxide  for
treating  filtered, nitrified  effluent,  depending  on  the  coliform standard.
The results of studies  conducted at a full-scale  plant  generally  agreed within
a factor of two with the predictions from laboratory  studies, when compared on
the  basis  of  the  product (residual  x time)  required  to  accomplish  a  given
fractional kill.

     For the  case  likely  to be most typical  in practice—the disinfection of
conventional secondary  effluent to  meet  a  total coliform standard of 1000  per
100 ml—disinfection  with chlorine  is  estimated  to cost  0.3 to 1.7 cents  per
m  (1.1 to 6.4 cents per 1000 gallons), depending  on  plant  size, compared with
1.5 to 3.5  cents per  m   (5.8  to 13.4 cents per 1000  gallons)  for disinfection
with chlorine dioxide (1979 price levels).

                                      ii

-------
     Chlorine dioxide  was  found to be more  effective for inactivating  Polio-
virus I  and  natural populations  of  coliphage in  both non-nitrified and  fil-
tered, nitrified  wastewater effluents.  Chlorine  dioxide treatment formed  no
measurable amounts  of  trihalomethane  byproducts,  whereas  chlorine treatment
formed  0.5 to  5  yMol per  liter of  trihlomethanes, chiefly  chloroform,  in
experiments using wastewater effluents.   The measured amount of total organic
halogen (TOX) formed by chlorine dioxide disinfection was so  small  as to be  on
the margin of statistical  significance;  chlorination formed at least 10 to  20
times more undesirable halogenated byproducts measured  as  TOX than did chlo-
rine dioxide  treatment.   These advantages of  chlorine dioxide should be  con-
sidered, along with  the  cost-effectiveness comparison based on coliform kill,
to reach decisions as  to when and where to employ  chlorine dioxide  as a  disin-
fectant in wastewater  treatment.

     This  report  was  submitted in fulfillment of  Grant  No. R-805426 by Stan-
ford  University  under  the sponsorship  of the U.S.  Environmental  Protection
Agency.  This report  covers the period September  5,  1977 to  June 5, 1980, and
work was completed as  of June 5, 1980.
                                      iii

-------
                                   CONTENTS
Abstract	       ii
Figures	       vi
Tables	       ix
Acknowledgments  	     xii

    1.  Introduction 	        1
            Background	        1
            Objectives  	        2
    2.  Conclusions  	        4
    3.  Recommendations  	        6
    4.  Materials and Methods   	        7
            Wastewater characterization   	        7
            Microbiological analyses  	        9
            Disinfectant chemistry 	       10
            Halogenated organics analyses   	       12
    5.  Chlorine and Chlorine Dioxide Chemistry and Generation  ....       16
            Measurements of chlorine dioxide and chlorine species   .  .       16
            Measurement of chlorine dioxide generator product
              composition	       20
    6.  Models of Disinfection Kinetics	       30
            First-order kinetics 	       30
            Decreasing rate and lag time	       30
            Retardant reaction approach	       32
    7.  Disinfection Experiments with Conventional Coliform Indicators       37
            Experimental design  	       37
            Statistical analysis of experimental results 	       46
            The effect of process sequence on bacterial numbers   ...       54
            Comparisons among wastewaters   	       57
            Effects of treatment level on the disinfection process  .  .       63
            Comparison of disinfection with chlorine and chlorine
              dioxide	       68
            Predicting full-scale performance from laboratory
              experiments	       71
    8.  Virus Inactivation  	       76
            Results	       76
    9.  Recovery Following Disinfection   	       83
            Experimental 	       83
            Results	       83
    10. Halogenated Organics Produced During Disinfection  	       86
            Chlorine dioxide versus chlorine 	       86
            Mixtures of chlorine dioxide and chlorine  	       95
                                      iv

-------
    11. Cost Comparison Between Chlorine Dioxide and Chlorine for
        Wastewater Disinfection  	      100
            Basis of comparison	      100
            Capital costs  	      101
            Operation and maintenance costs—excluding chemical
              costs	      102
            Chemical costs 	      104
            Cost summaries	      105
            Discussion	      112

References	      114
Appendices	      122

-------
                                    FIGURES


Number                                                                    Page

  1    Chlorine dioxide generator using acid activation of a sodium
         chlorite solution 	      10

  2    Diagram of TOX analysis	      14

  3    Reduction of chlorine dioxide as a function of pH	      16

  4    Chlorine dioxide generator, acid-chlorite process 	      26

  5    Types of bacterial survival curves  	      31

  6    Multi-media filtration column 	      38

  7    Wastewater treatment plant flow schemes 	      40

  8    Experimental dose-time matrix for chlorine-chlorine dioxide
         comparison	      42

  9    A rapid-mix, rapid-sampling, 4-liter batch reactor  	      43

  10   Disinfection contact tank at the Dublin-San Ramon wastewater
         treatment plant ............. 	      44

  11   Results of tracer study of disinfectant contact tank  	      45

  12   Dye tracer concentration measured at overflow weir  	      47

  13   Comparison of chlorine and chlorine dioxide bactericidal
         effectiveness at 30-minute contact time 	      53

  14   Chlorine and chlorine dioxide residual die-away in Palo Alto
         secondary effluent  	      54

  15   Comparison of chlorine and chlorine dioxide bactericidal
         effectiveness at 10 mg/1 dose	      55

  16   Coliform inactivation by free and combined chlorine in San Jose
         nitrified effluent  	      61
                                      VI

-------
Number
Page
  17   Coliform inactivation by free and combined chlorine in San Jose
         filtered effluent 	       62

  18   Coliform inactivation by combined chlorine in San Jose
         wastewater at three sampling points 	       62

  19   Coliform inactivation by chlorine in San Jose wastewater  ...       64

  20   Coliform inactivation by chlorine in filtered and unfiltered
         secondary effluent,  Palo Alto	       65

  21   Coliform inactivation by chlorine dioxide in San Jose
         wastewater effluents  	       66

  22   Coliform inactivation by chlorine dioxide in filtered and
         unfiltered secondary effluent, Palo Alto  	       67

  23   Coliform inactivation by chlorine and chlorine dioxide in
         1978 experiments with unfiltered secondary effluent from
         Palo Alto	       68

  24   Coliform inactivation by chlorine and chlorine dioxide in
         1979 experiments with unfiltered secondary effluent from
         Palo Alto	       69

  25   Coliform inactivation by chlorine and chlorine dioxide in
         1979 experiments with filtered secondary effluent 	       70

  26   Coliform inactivation by chlorine and chlorine dioxide in
         San Jose secondary effluent	       70

  27   Coliform inactivation by chlorine and chlorine dioxide in
         San Jose nitrified effluent	       72

  28   Coliform inactivation by chlorine and chlorine dioxide in
         San Jose filtered effluent	       72

  29   Coliform inactivation by chlorine and chlorine dioxide in
         laboratory experiments with Dublin effluent 	       73

  30   Coliform inactivation by chlorine and chlorine dioxide in
         field experiments with Dublin effluent  	       73

  31   Coliform inactivation by chlorine in laboratory and field
         experiments with Dublin effluent  	       74

  32   Coliform inactivation by chlorine dioxide in laboratory and
         field experiments with Dublin effluent  	       74
                                      vii

-------
Number                                                                    Page
  33   Comparison of in situ coliphage and an inoculum of Poliovirus I
         in non-nitrified secondary effluent 	      78

  34   Total coliform survival in non-nitrified secondary effluent
         comparing chlorine dioxide and chlorine at three doses and
         three contact times	      79

  35   In situ coliphage (E_. coli B host) survival in non-nitrified
         secondary effluent, comparing chlorine dioxide and chlorine
         at three doses and three contact times  	      80

  36   Responses of several in situ organisms to chlorine dioxide
         and chlorine at very short contact times in non-nitrified
         secondary effluent  	      81

  37   Total coliform and Poliovirus I survival in non-nitrified
         activated-sludge effluent (Palo Alto) and in nitrified,
       ,•  filtered effluent (Dublin)	      82

  38   Five-day recovery following disinfection  	      84

  39   Procedure for bench-scale determinations of halogenated
         organics formation  	      88

  40   TOX and THM production in non-nitrified effluent
         (Experiment 4—Palo Alto wastewater)  	      90

  41   TOX and THM production in filtered, nitrified effluent
         (Experiment 6—Dublin-San Ramon wastewater) 	        91

  42   Disinfectant residuals in Experiment 8 (Palo Alto wastewater) .      93

  43   Disinfectant residuals in Experiment 6 (Dublin-San Ramon
         wastewater)	      93

  44   THM production resulting from various mixtures of chlorine
         and chlorine dioxide in Experiment 2 (Dublin-San Ramon
         wastewater	      97

  45   TOX and THM production resulting from various mixtures of
         chlorine and chlorine dioxide in Experiment 5 (Dublin-
         San Ramon wastewater)	      98
                                     viii

-------
                                    TABLES
Number                                                                    Page
  1    Estimated Standard Deviation of THM Analysis   	       13

  2    Precision of TOX Measurement	       15

  3    Blank Correction for TOX Calculation  	       15

  4    Comparison of Amperometric and lodometric Methods for
         Determining Concentration of ClO^ 	       17

  5    Comparison of Amperometric and lodometric Methods by Analysis
         of Variance	       17

  6    Amperometric Titration Evaluation 	       18

  7    Analyses of Variance for Amperometric Titration  	       19

  8    The Effect of Chlorite on the Determination of Chlorine Dioxide
         by Amperometric Titration 	       20

  9    Equivalent Weights for Calculating Concentrations on a Mass
         Basis	       23

  10   Composition of Reactants and Products in Laboratory Generation of
         Chlorine Dioxide by Sulfuric Acid Activation of Sodium Chlorite     24

  11   Composition of Reactants and Products from Chlorine Dioxide
         Generation in a Continuous, Full-Sized Acid-Chlorite Reactor
         Using H2S04	       27

  12   Composition of Reactants and Products from Chlorine Dioxide
         Generation in a Continuous, Full-Sized Acid-Chlorite Reactor
         Using HC1	      28

  13   Yield of Chlorine Dioxide from Acid Activation in Field
         Experiments	       28

  14   Characterization of Wastewater Effluents Used in Disinfection
         Experiments	       41
                                      ix

-------
Number                                                                    Page
  15   Mean Log^Q Surviving Total Coliform Counts in 1978 Batch
         Experiments with Palo Alto Effluent .............       49

  16   ANOVA for Bacterial Analysis Before Disinfection  .......       50

  17   ANOVA for Examining Experimental Reproducibility  .......       50

  18   ANOVA for Comparison of Chlorine and Chlorine Dioxide as
         Bactericides in Palo Alto Effluent  .............       52

  19   Mean Log [Survival Ratio] of Total Coliforms:  Comparison of
         Disinfectants at Different Combinations of Dose and Contact
         Time  ............................       52

  20   Initial Total Coliforms in Undisinfected Effluents from the
         San Jose Wastewater Treatment Plant .............       56

  21   ANOVA of San Jose Initial Bacteria Numbers  ..........       56

  22   Logs of Initial Bacterial Numbers at the Palo Alto Wastewater
         Treatment Plant .......................       57

  23   ANOVA of Palo Alto Initial Bacterial Numbers  .........       57

  24   Values of Fitting Constants for Chlorine and Chlorine Dioxide
         in the Disinfection Model ..................       59

  25   Chlorine to Ammonia Ratios for San Jose Experiments  ......       60

  26   Comparison of Model Coefficients, Segregating Disinfection
         with Free vs Combined Chlorine  ...............       61

  27   One-Hour Chlorine Demand  ...................       64

  28   One-Hour Chlorine Dioxide Demand  ...............       67

  29   Typical Characteristics of Wastewater Effluents Used to
         Measure Formation of Halogenated Organics ..........       87

  30   Wastewater Characteristics in Field Experiments at the
         Dublin-San Ramon Chlorine Contactor .............       88

  31   THM and TOX Formation by Chlorine and Chlorine Dioxide in
         Bench-Scale Experiments ...................       89

  32   Summary of THM and TOX Field Experiments  ...........       91

  33   Mean Values and Standard Deviations of 24-Hour Halogenated
         Organic Byproduct Formation with Chlorine Dioxide  and
         Chlorine  ..........................       92

-------
Number                                                                    Page
  34   Disinfectant Demand 	       95

  35   A Summary of the Results and Conditions of Experiments Comparing
         THM and TOX Formation by Mixtures of Chlorine and Chlorine
         Dioxide	       99

  36   Cases for Evaluation of the Relative Costs of Disinfection
         with Chlorine and Chlorine Dioxide   	      101

  37   Summary of Estimates of Required Disinfectant Dosages Used in
         Cost Evaluation Cases 	      103

  38   Amounts of Disinfectants Required to Achieve Total Coliform
         Standards	      104

  39   Unit Costs of Chemicals Required in Wastewater Disinfection .  .      105

  40   Disinfection Cost Summary for Case A in Thousand $ per Year .  .      106

  41   Disinfection Cost Summary for Case B	      107

  42   Disinfection Cost Summary for Case C	      108

  43   Disinfection Cost Summary for Case D	      109

  44   Disinfection Cost Summary for Case E	      110

  45   Disinfection Cost Summary for Case F	      Ill
                                      xi

-------
                                ACKNOWLEDGMENTS
     The  active assistance  of  the staff  of  the  Dublin-San  Ramon  Services
District, particularly  Robert  Swanson,  Plant Superintendent, was essential  in
carrying  out  the full-scale field  tests of chlorine  dioxide.   The Rio  Linda
Chemical  Co.  made  available chlorine  dioxide generating  facilities  for  the
full-scale demonstration.

     The  staff  of  the Palo Alto  Regional  Water Quality Control Plant and  the
San Jose-Santa  Clara Water  Pollution  Control Plant provided  samples  for  the
laboratory studies.   The Fischer-Porter Company  generously loaned  an ampero-
metric titrator.   Dr. Gary Stevenson gave  useful advice on residual  measure-
ment techniques.

     Gary D.  Hopkins  assisted  in the full-scale demonstration  experiments  and
determined total organic  halogen concentrations.   Annie M. Godfrey, Christoph
Munz, Paul Dandliker, John Gonzales, Laurie Lapat, Mette Horn, Niambi Loud,  and
Christian Drozier performed many  of the microbiological and chemical analyses.

     We  are   grateful  to  Professor  Robert Cooper,  School of  Public Health,
University of California, Berkeley, and Professor  Carleton Schwerdt, Depart-
ment of  Medical Microbiology,  Stanford University,  for  the virus assays per-
formed in their laboratories.
                                      xii

-------
                                   SECTION 1

                                 INTRODUCTION
BACKGROUND

     Disinfection of  wastewaters has  long  been practiced  in  areas where  the
use  of  the  receiving waters  for  public water  supply,  contact  recreation,
and/or shellfish  habitat is to  be  protected.  Many wastewater-effluent  dis-
charge permits allow  a maximum geometric mean of 1000 total coliforms  per 100
ml,  or  200  fecal  coliforms per  100 ml,  in  a 30-day  period.   The State  of
California permits only  2.2 total coliforms per 100  ml if  an effluent is  to be
used for groundwater  recharge.

     Chlorine has been  the favored disinfectant  in  both water and  wastewater
treatment by virtue of  its bactericidal effectiveness, low cost,  convenience,
and  relatively  long-lived residual.   However,  chlorination as normally  prac-
ticed in water treatment recently has  been found  to  result  in  the formation of
trihalomethanes and  other chlorinated  organics that are undesirable from the
viewpoint of public  health and water  pollution  control  in general.  Although
the  formation  of chlorinated  organics  during disinfection of wastewater  by
chlorine previously has  not been investigated  thoroughly,  there is reason to
believe  that  such hazardous  byproducts  are  formed. In view  of the residual
organic contamination of  even  highly treated  wastewater effluents  and  the re-
sulting high doses of chlorine needed  to  provide  a chlorine residual in waste-
waters, it can be expected that undesired, chlorinated organic byproducts will
be  formed  to a  greater  extent  during wastewater chlorination than has  been
reported for water chlorination.

     Thus there is a  need to  find a substitute for  chlorine as a disinfectant
in wastewater treatment that fulfills  the following  criteria:

     1.  Cost effectiveness  as  measured by  the  relationship  of the
         total cost (amortization plus operating  costs)  to the effec-
         tiveness as  a  bactericide  under the  conditions  encountered
         in practice.
     2.  Absence of participation in side reactions  that  yield unde-
         sirable  byproducts,   particularly chlorinated  organic  com-
         pounds .
     3.  Safety and  convenience of  use  and  ease of  installation in
         existing wastewater treatment  plants.
     4.  No  residual  toxicity to aquatic organisms   in  receiving wa-
         ters.

-------
     Of the  possible substitutes for  chlorine,  ozone has  attracted the most
attention.  Ozone is a  powerful  disinfectant  but is more expensive  than chlo-
rine and does not provide a stable residual.

     Chlorine dioxide (C102), although more expensive than chlorine, may offer
an attractive alternative because it:

     1.  Is a powerful disinfectant over a broad pH range,
     2.  Provides a  residual that  is  easily  measurable  for  purposes
         of control,
     3.  Does not react with ammonia to form chloramines,
     4.  Does not enter  into reactions with organic material  to form
         some  classes   of   chlorinated organic  compounds  considered
         hazardous to health (e.g., trihalomethane compounds), and
     5.  Does  not react  to form  chlorinated phenols  to the  extent
         that chlorine does.

     The second factor mentioned  above gives  chlorine dioxide an inherent  ad-
vantage over  ozone  from the viewpoint of process control.    All  the other
factors are  advantages  compared  with  chlorine.   Factor (3),  the   absence  of
reactivity with  ammonia,  constitutes  an  especially  important  advantage  for
chlorine dioxide  over chlorine in wastewater  disinfection,  if a free chlorine
residual is required;  the  chlorine demand to  reach the breakpoint  may exceed
100 mg C12 per liter.


OBJECTIVES

     The purpose of this project is to evaluate chlorine dioxide as  an  altern-
ative to conventional chlorination for the disinfection of wastewater.

     The specific objectives are:

     1.  To assemble  and evaluate the  available  information concern-
         ing the  chemistry  of  chlorine dioxide generation  and of  its
         behavior  in aqueous  solution,  the  technology and  costs   of
         manufacture,  its  effectiveness  as a  disinfectant,   and  the
         possible side effects of its  use.
     2.  To  establish  the  dose-effectiveness  relationship  for  chlo-
         rine dioxide as a  disinfectant of wastewater after secondary
         treatment and after  various   stages  of advanced  treatment,
         using as a criterion the survival of coliform bacteria.
     3.  To  compare  the effectiveness  of  chlorine dioxide with that
         of chlorine as a disinfectant.
     4.  To  develop  a  concept  for a continuous  reactor  system   to
         fulfill  coliform   requirements   corresponding  to  standard
         secondary treatment,  and advanced treatment  for  the purpose
         of wastewater reclamation and  reuse.
     5.  To prepare  a  preliminary design  for  treatment  plants of  1-,
         5-, 10-, 50-,  and  100-mgd  capacity and to estimate the costs
         of construction and operation.

-------
6.  To obtain  preliminary evidence  as  to whether  the  formation
    of chlorinated  organic byproducts  during wastewater  disin-
    fection conforms to  the results  from studies of water disin-
    fection.

-------
                                   SECTION 2

                                  CONCLUSIONS

     The use of chlorine  dioxide  as  a disinfectant in wastewater treatment  is
technically feasible.  This  was  demonstrated  in short-term, continuous opera-
tion  at a  full-scale wastewater treatment  plant,  which  served  to confirm
laboratory results.

     Results of chlorine  dioxide disinfection  can be  conveniently correlated
using the  logarithm of the  surviving fraction of organisms  as the dependent
variable and the  logarithm of the product  of  the disinfectant residual  times
the contact time as the independent variable.   The form of the correlations  so
obtained for chlorine  dioxide is similar to  that for  disinfection with  chlo-
rine.   Compared with  chlorine,  however, a given degree of  disinfection was
achieved with a smaller  value of the  product  (residual x time) when chlorine
dioxide was used.   Based  on  this general finding,  it  is  concluded that, in a
technical sense, chlorine dioxide is superior to  chlorine for the disinfection
of wastewater.

     The chlorine  dioxide disinfection process can be controlled conveniently
and with adequate  precision  by adjusting the  dose to provide a desired resid-
ual.   Proven  residual measurement techniques,  such  as automated amperometric
titration,  are suitable for  measuring the  chlorine  dioxide residual for con-
trol purposes.

     To  satisfy a  given  coliform standard at  a  given contact  time,  the re-
quired  chlorine dioxide   residual  is only  one-third to one-tenth  the corre-
sponding chlorine  residual (mass  concentration basis).   The ratio of required
doses (mass ClO^imass  Cl~)  ranges from  approximately  0.05  to 1.0.  Hence,  in
general less chlorine  dioxide than  chlorine must  be used to satisfy the  coli-
form standards usually applied in wastewater disinfection.

     The unit cost of  chlorine  dioxide  is  approximately  twenty times that  of
chlorine.  This difference is attributed almost  entirely  to the high cost  of
the sodium chlorite raw material.

     The cost of  disinfection with  chlorine dioxide to achieve a given  coli-
form  standard  is   greater than  or  equal to that with chlorine  in all  cases
evaluated;  the higher  unit cost   of  chlorine  dioxide generally more than off-
sets its lower  dose requirement.  For  the  most  typical case—disinfection  of
conventional activated-sludge effluent to satisfy a  total coliform  standard  of
1000  per  100  ml—chlorine  dioxide  is two  to  five  times more  expensive than
chlorine.

-------
     Chlorine dioxide  forms  no  significant amounts of undesirable chlorinated
organic byproducts under  the conditions of wastewater  disinfection.   No mea-
surable amounts of trihalomethane  compounds (THMs) were formed when  secondary
effluents were  dosed with  20 mg/1  or  40  mg/1  ClO^-   Total organic halogen
(TOX) was formed  to  a small  (but  statistically  significant) extent  in one  of
four  experiments  with  ClO^-   Chlorine  at equal  mass doses  formed copious
amounts of  THMs  (0.5  to  5  yMol  per liter)  under the  same  conditions.   The
average TOX  formation from  chlorine was  10  to  20 times  that  from chlorine
dioxide at the same mass dose of disinfectant.

-------
                                   SECTION 3

                                RECOMMENDATIONS

     The evaluation of chlorine  dioxide  as an alternative to chlorine for  the
disinfection of wastewater has resulted in a decision-maker's dilemma.  On  the
one hand,  chlorine dioxide is  found  to be superior  to  chlorine according  to
technical criteria.  On the other hand, chlorine dioxide is not cost-effective
when the customary performance criterion (total coliform standard) is applied.

     Two important classes of questions remain:  Are the unquantified benefits
of  chlorine  dioxide compared  with chlorine—namely,  negligible formation  of
potentially hazardous  chlorinated  organic byproducts and superior virus  inac-
tivation—worth the additional cost?  Can  the unit cost of chlorine dioxide be
reduced to a level  at  which  chlorine  dioxide  would be economically attractive
in  the  customary  sense,  even  without  considering its additional, unquantified
advantages?  Both types of questions deserve further study.

     The  chemical composition  of  disinfection byproducts  and  endproducts  of
all disinfectants  should  be  studied systematically.   There  is  an urgent need
for better  understanding of the halogenated  organic byproducts of wastewater
chlorination.

     The apparent advantage of chlorine dioxide with regard to virus  inactiva-
tion  needs  to  be documented more  convincingly.    Comparison of  the kill  of
coliforms with  that of other  indicators,  using  alternative disinfectants,  is
worthy of further investigation.  Also of  potential  significance is the devia-
tion of disinfection data from the simple  first-order model usually assumed to
apply.  The  causes underlying these deviations  should be studied with a view
toward  re-evaluating the  adequacy  of  the total coliform count as an  indicator
of disinfection performance.

     Development  of processes  to  produce chlorine  dioxide more  cheaply  at
treatment  plant  sites  should  be  encouraged  by  supporting  demonstration  of
candidate concepts.

-------
                                   SECTION 4

                             MATERIALS AND METHODS


WASTEWATER CHARACTERIZATION

2H

     The  pH  was  determined using a pH meter  (Corning  Model 610A),  with a pre-
cision  of ± 0.05 pH unit.   The pH meter was standardized  using a  pH-buffered
standard  at  a pH near  the  value expected in  the sample.

Alkalinity

     Alkalinity was determined  by  potentiometric titration to pH 4.3.

COD

     COD  was determined  using the  low-level  COD procedure outlined  in Standard
Methods,  Section  508  (1).   A  potassium acid phthalate standard was  run with
each set  of  determinations to assure  accuracy.   Samples  were stored as for the
ammonia nitrogen test.

Non-Filterable Residue
     Non-filterable  residue  dried  at  103-105°C was  determined by the procedure
outlined  in  Standard Methods  (1),  Section 208A,  using Gooch  crucibles.   Sam-
ples were tested  on the day of the experiment where  possible.  If storage was
required,  the  storage  procedure used was  the  same  as  for  the ammonia nitrogen
test.

Ammonia Nitrogen

     Two  methods  were  used  to  determine ammonia  nitrogen.   From October 1977
to  January 1979 a specific  ion probe made by Orion Co. was used.   The proce-
dure was  as  follows:

     1.   A standard  curve (semi-log)  was  prepared  at the beginning  of
          each  set  of tests  and every two hours afterwards using 100-,
          10-,  and  1-mg/l ammonium  chloride standards.  The  standardi-
          zation  procedure was the  same  as  for  measuring  ammonia  in
          samples  as well.    The  meter  was  zeroed  using  the  10-mg/l
          standard.

-------
     2.  Readings were  made  as follows:   (a)  100-ml samples were pre-
         pared;  (b)  1  ml  of  10N_ NaOH  was added  and  the  sample was
         stirred  continuously until  a  stable meter  reading  was at-
         tained, which took approximately  15 min.

All  samples  were allowed  to  temperature equalize  for  at  least 1/2 hr  before
the NaOH was added.

     After January 1979  samples  became  too numerous, and a  change was made to
the  distillation  followed by  acidimetric  titration  procedure  described  in
Standard Methods (1), Sections 418A and B.   In addition, total  Kjeldahl  nitro-
gen was  measured by the method  listed under  "Nitrogen (Organic)" in Section
421  in Standard Methods  (1).   An ammonium chloride  standard  and  a glycine
standard  were  used  during  each set  of  determinations to  assure  accuracy.
Samples  were acidified  to below pH  2 using  IkSO, and refrigerated  at  4°C.
Samples were analyzed within two weeks.

TOG

     TOG analysis  was  done  using  a Dohrmann  Envirotech DC-52 Organic  Carbon
Analyzer with  an Oceanographic  International  Ampule Crushing Unit adapted to
it.  Sample preparation was as follows, using  10-ml glass ampules:

     1.  1 ml  of  saturated  potassium  persulfate  was  added  to  each
         ampule.
     2.  Sample  water  was  added  (0.5  to 1.5 ml  for secondary efflu-
         ent).
     3.  Distilled water was  added to  fill  the  ampules  to  no more
         than 10 ml total volume.
     4.  0.2 ml of 10% phosphoric acid solution was added.
     5.  The  ampules  were  purged  for  7-9  min   using pure  oxygen
         cleaned  by  passing  the   02   over  a  475°C  cupric  oxide
         catalyst.
     6.  The ampules were  sealed and autoclaved for 4 hrs.
     7.  The special  crushing unit  was used  to  open  the  ampule and
         purge the ampule  with a helium carrier gas which passed into
         a flame-ionization detector.

Note that  the  initial  purging in step 5  to  remove inorganic carbon also  re-
moved volatile organics.   Therefore the TOG value  is actually  a  non-purgeable
TOG.   A calibration curve was made by running sucrose standards in the  same
manner as the samples.  Reagent  blanks were  also run in  this manner.

     Samples awaiting  sealing in ampules  were stored  in  teflon  sealed  hypo-
vials  with  aluminum seals.   These  samples were also  acidified to pH 2  using
HC1 and stored at 4°C.

     The detection limit of  the  TOG procedure  is 0.1 mg/1 C.  The precision is
approximately ± 5% at TOG  concentrations greater than  1  mg/1 C.

-------
MICROBIOLOGICAL ANALYSES

Bacteria

     Samples  for  microbiological  analysis  were  collected  in sterile  bottles
containing  sodium thiosulfate  to  neutralize  any  residual  oxidant and  were
analyzed  immediately.    Three field  experiments were  run  at  the  Dublin-San
Ramon  Wastewater  Treatment  Plant.   During  these  experiments,  samples  were
collected, stored on  ice,  and analyzed after return to  the  lab,  approximately
one hr after the completion of a 6-hr set of  replicate  runs.  A  set  of  samples
that  had been  disinfected by chlorine and  chlorine  dioxide,  as  well as  an
untreated influent sample, were analyzed immediately for total coliform bacte-
ria  at  the  field  site  and  compared with  the  iced and transported  samples.
There  were  no significant differences  between  the  analyses conducted  in the
field  and  in the  laboratory-analyzed samples.   Therefore,  it is assumed  that
the microbiological integrity of  the samples was maintained  during  the course
of a long experiment.

     Most probable number  (MPN)  and membrane filter (MF) methods of bacterial
enumeration were done  according  to Standard Methods (1).  MPN media used  were
lactose  broth as a presumptive medium and brilliant green bile  broth for  con-
firmation  of  total coliforms  and  EC  medium for confirmation  of fecal coli-
forms.   Gelman GN-6 0.45-micron filters were  used for  the MF tests with M-Endo
Agar  for total  coliforms using   the  single-step  direct technique and  M-FC
medium for  fecal  coliforms.   KF  streptococcal  agar was  used to  assay  for the
fecal streptococcus group.

     MF  tests and MPN  presumptive tests were incubated  at 35°C ± 0.5°C for 24
± 2 hrs.   Confirmed MPN tests and  fecal streptococcus  tests  were incubated at
35°C ± 0.5°C for 48  ± 3  hrs.   Fecal coliform analyses, either by MPN or MF
method,  utilized  an  incubation  period  of  24  ± 2  hrs  and  a  temperature  of
44.5°C ± 0.2°C in a circulating water bath.

Viruses
     Two  bacterial  virus  methods  were  evaluated  to  find  a suitable  in  situ
model for virus  inactivation by C102-  One method,  that  of Kott  (2),  also has
been used  in  earlier studies  (3).   The method uses Escherichia coli  B as the
host.   Samples were  enriched  overnight in nutrient broth at 35°C  and plated
for  observation of  plaques as  a  positive test.    The  procedure employs  MPN
tables;   low  numbers  of  coliphage,  <  2  plaque-forming  units/100  ml,   are
detectable. The  procedure  is  time-consuming,  however.   A second method,  the
Reverse Phage  Titer  Rise Reaction  (RPTRR), as developed by  Atlantic  Research
(1978)  (4), was  also  used.   The RPTRR method  uses   _E_.  coli C as  the host  bac-
teria,  and yields results  in 4 to 6 hrs.  The host  E_.  coli C is  thought to be
susceptible to a broader range  of coliphage than is  J^.  coli B.

     The animal  virus used in  the  study was  Poliovirus I  LSC strain  that had
been grown in Buffalo Green Monkey Kidney  cells (BGM).   The virus  were harves-
ted from the  cells by freezing, thawing, and  centrifugation  to a  titer of ap-
proximately 109  plaque-forming  units per milliliter (PFU/ml).  The  virus  were
recovered after  experimentation using  a  standard  plaque  assay.    The  animal

-------
virus assays  were performed  in  the  laboratories  of  either  Dr.  Robert Cooper,
Sanitary Engineering  Research Laboratory,  University of California at Berke-
ley, or Dr.  Carleton  Schwerdt,  Department of  Medical  Microbiology, Stanford
University.
DISINFECTANT  CHEMISTRY
Chlorine Dioxide Generation
     Chlorine dioxide was generated according to the method described  in Stan-
dard Methods (1), Section 411A,  diagrammed in Fig. 1, except that  the  acid and
sodium chlorite  concentrations were doubled  to  obtain higher  chlorine dioxide
concentrations in the stock solution.   The concentration  so obtained  was 1000
to  1500  mg/1  as CIO™.   Spectrophotometric  analysis at  360  nm  was  used to
verify that the  chlorine species in solution was indeed chlorine dioxide.

Chlorine Generation
     Aqueous  chlorine  stock solutions were made by bubbling 99.99%  chlorine
gas through chilled water.  The solutions so prepared  contained  chlorine at a
concentration of 2000 to 4000 mg/1 as  Cl.
      AIR
   SUPPLY
             REACTION
              VESSEL
                                VENT
                                 TO
                                HOOD
NaCIO,
                                 SALT
                               TOWER
                                                      C102 STOCK SOLN
    Figure 1.   Chlorine dioxide generator  using acid activation of a sodium
               chlorite solution.
                                     10

-------
Standardizing of Disinfectant Stock Solutions

     Disinfectant stock solutions were standardized on  the day  of  each  experi-
ment by the iodometric method described in Standard Methods  (1), Sections  409A
and 411A.  Standard sodium thiosulfate titrant at O.OlN^was  used,  standardized
by the dichromate method.   Blank titrations  were omitted because  it was found
that blank contributions were negligible.

Residual Measurement for Routine Analyses

     Both chlorine  and  chlorine dioxide  residuals were measured by amperomet-
ric titration (1).  The amperometric titrating unit used was a  prototype model
made by  Fischer  and  Porter, similar  in function  to  the Fischer and Porter
Model  17T1010  amperometric  titrator.   The  back-titration  procedure was  used
for both disinfectants.  It consists of the  following steps:

     1.  A known  volume of  0.00564N^ phenylarsine oxide  solution  was
         added to a 250-ml beaker in excess of  the  expected chlorine
         or chlorine dioxide residual.
     2.  2 ml of  the  appropriate pH buffer  was  added.   A pH-4 buffer
         (acetate)  was  used for chlorine  analysis  and  a  pH-7  buffer
         (phosphate) was used for chlorine dioxide analysis.
     3.  A  1-ml  aliquot   of 5%  weight-per-volume  potassium  iodide
         solution was added.
     4.  A  known volume   of sample  was  carefully   poured   into  the
         beaker, approximately  150-175 ml.
     5.  The  titrator  was  turned  to the   "Total"  position and  the
         above solution was  titrated  with  0.00564N^ iodine solution to
         an amperometric  endpoint.   For this  particular amperometric
         titrator,  the  endpoint was  denoted by  sudden  needle deflec-
         tion.

Residual was calculated by the  following formulas:

                                           0.00564N. x 35453  mg  Cl/equiv
       mg/1 Cl = [PAO - (I2 - blank)  x CF]


     mg/1 C102 = [PAO - (I2 - blank)  x CF]
       vol sample (ml)

0.00564N x 67500 mg Cl/equiv
                                                  vol sample  (ml)

where PAO   = volume of 0.00564^ phenylarsine oxide used  in ml

      I2    = volume of 0.00564N_ I2 titrant used in ml

      blank = volume of I2 found for blank (see below)

      CF    = correction factor  for I2 titrant (see below).

The correction factor was found  as follows:
                                      11

-------
     1.  A beaker was prepared with 2 ml of the phenylarsine oxide solu-
         tion, the appropriate pH buffer, and potassium iodide solution.
     2.  150 ml of deionized water was added.
     3.  The sample was titrated as in Step 5 above.  This procedure was
         repeated for each pH.


                         _ vol phenylarsine oxide
                             volume !„ titrant


The blank was found  by  the  same procedure as for the correction factor except
that the appropriate amounts of undisinfected sample water were added:


              Blank  = I~ volume - I^ volume  from CF  determination


A blank was determined for each pH and for each different sample volume used.


HALOGENATED ORGANICS ANALYSES

Trihalomethane (THM) Analysis

     THM analysis was  accomplished by liquid-liquid extraction to concentrate
the organics, and subsequent  gas-liquid  chromatographic analysis according  to
the procedure of Henderson  et  al.  (5).   Liquid-liquid extraction consisted  of
the following steps:

     1.  1  ml  of  specially  cleaned  pentane  [purchased  as  Basic
         Alumina  (Activity  1) grade] was injected  into  each of  the
         sealed  60-ml  hypovials  containing the  samples  (displaced
         water was discarded).
     2.  A  prescribed amount  of 1,2-dibromoethane  internal  standard
         was  injected  so  that  its  concentration in the  hypovial was
         near that expected for THMs  (between 10 and 30 ug/1).
     3.  The  samples were agitated at 250 to 400 rpm for  about 30 min
         on a gyrorotatory agitator.
     4.  A  5-yl  aliquot  from the pentane phase was  injected  into the
         gas-liquid chromatograph.

     The  gas-liquid chromatograph  was  equipped  with a  linearized  electron-
capture detector  (Ni6   beta  source)  with a 20-pg detection limit.   The column
was made  up of 10%  squalane  on 80/100-mesh Chromosorb W.  Column  temperature
was held  constant at  75°C  while the detector  temperature  was held at 275°C.
Carrier gas  and  makeup gas were  argon/methane  mixtures.   Column gas  flow was
60 ml/min;  makeup  gas flow was  40 ml/min making a  total  100 ml/min gas  flow.
Concentrations were determined by  integrating the peak area and comparing with
the peak area of  the internal standard.   Unknowns were identified by  matching
their  retention times to those  of  known standards.
                                      12

-------
     The following definition of total THM (TTHM) is used:
            [TTHM] = 3[CHC13] + 3[CHBrCl2] + 3[CHBr2Cl] +  3[CHBr3]
with  results  presented  in ymol  per liter  as halogen  atoms.   The  standard
deviation of this procedure is given in Table  1.
            TABLE 1.  ESTIMATED STANDARD DEVIATION OF  THM ANALYSIS

Sample
1
2
3
4
5
6
7
8
9
10
Average
Std. Dev.
TTHM
ymol/l-x
0.351
0.330
0.301
0.288
0.307
0.310 TTHM produced = (TTHM of sample)
0.311
0.316 - (TTHM of unchlorinated sample)
0.337
0.241
0.309
0.03
        'TTHM produced
'TTHM  sample     TTHM unchlor.  sample
                   /2  x 0.032  = 0.04  nmol/1
TOX Analysis

     TOX  analysis  was performed  on a prototype  instrument made available  by
Dohrmann Envirotech.  The procedure is diagrammed  in  Fig.  2 and  is  detailed  as
follows:
         Organics  recovery  was achieved by two activated-carbon  filters
         connected  in series.   Each filter  contained  40 milligrams  of
         powdered  activated carbon.   Secondary  effluent  samples  filtered
         were 25 milliliters  in volume.
         The  carbon filters were washed  with a potassium nitrate  solu-
         tion containing  5000 milligram  per  liter  as nitrate to  remove
         inorganic  halide.  Initially, a  four-milliliter aliquot  of this
         solution  was  used  for each  set  of  filters.  Subsequently,  the
         volume of  nitrate used was  reduced to  2 milliliters.
         The  carbon was  extruded from  the  filters  and  placed  into  a
         quartz boat.   Under  a   C02  atmosphere the  boat was   introduced

                                      13

-------
                                             CO,
               ACTIVATED
               CARBON
               FILTRATION
STEP 1.   \^^\
                     NITRATE
                      WASH
         STEP
       ORGAN ICS
       RECOVERY
                                             STEP 3.
                                           PYROLYSIS
                             FILTER
                               T
                               STEP 2.
                            INORGANIC
                       CHLORIDE REMOVAL
                                           STEP 4.
                                     MICROCOULOMETRIC
                                         TITRATION
     4.
                    Figure 2.  Diagram of TOX analysis.
into a 200°C  zone to vaporize the volatile organics.   Next, the
boat was pushed into the 800°C zone.   The  C02  atmosphere al-
lowed more complete recovery of bromides.   Finally  the atmo-
sphere was  changed to  a CO^/O^  mixture  for completing  the
combustion.
A microcoulometric  titration unit served as  the detector  to
measure the  amount  of chloride and bromide released  in  the py-
rolysis process.   In  the  microcoulometric titration  method,
silver ion was  titrated  in a  70% acetic acid electrolyte solu-
tion.
     Generally, 25-ml samples were filtered.
clogging  due to particulates.
                                   Higher volumes often  resulted in
     The  precision  of  the  TOX method was investigated  by determining  TOX  in
ten replicate  samples  of  secondary  effluent  (Table 2).   The  average TOX was
4.25 jjMol/1-X;  the  standard deviation of  the  individual measurements  was  0.69
yMol/l-X,  or  16 percent of the mean.  The calculation of the  amount  of TOX
formed during  disinfection entails  subtracting  the TOX concentration  in the
unchlorinated sample from that in the chlorinated sample;  both quantities in-
corporate  a blank correction (Table 3).  The variance of estimated quantity  of
TOX formed by chlorination  is  given as:
                                    14

-------
                    TABLE 2.  PRECISION OF TOX MEASUREMENT

Sample
1
2
3
4
5
6
7
8
9
10
Average
Std. Dev.
TOX in
Umol/l-X
5.92
4.76
3.85
3.89
4.16
4.27
3.85
4.43
3.87
3.45
4.25
0.69
                      TABLE 3.  BLANK CORRECTION FOR TOX
                                CALCULATION
                                              TOX in
                         Sample              yMol/l-X

                            1                   1755
                            2                   1.54
                            3                   1.46
                         Average                1.52
                        Std. Dev.               0.05
                   _  / 2              2                       2
       STOX formed  ~  '/ STOX sample   STOX unchlor.  sample     TOX blank


In the case  of  our measurements, the  standard deviation was taken to be 0.69
uMol/l-X for both the treated and untreated samples (Table 2), whereas that  of
the blank was 0.05 joMol/1-X (Table  3).  Hence, the standard error of estimat-
ing the amount of TOX formed was calculated as:
             STVW  f     A  "    x °-69   + 4 x 0-05   " °-98
              TOX  rormed

This value was  used in judging whether  observed amounts of TOX formed  should
be considered statistically significant.
                                      15

-------
                                   SECTION 5

            CHLORINE AND CHLORINE DIOXIDE CHEMISTRY AND GENERATION


MEASUREMENTS OF CHLORINE DIOXIDE AND CHLORINE SPECIES

     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 well understood.   Many
of these same methods are also applicable to the measurement of  chlorine  diox-
ide residuals, with some modifications.   Two methods for residual measurement
of chlorine dioxide were chosen for investigation in this  study:   (1) Diethyl-
p-phenylene  diamine,  ferrous  ammonium sulfate  (DPD-FAS)  Titrimetric Method,
and  (2) Amperometric Method.

Comparison of Residual Measurement Methods

     In attempting  to  measure  residual chlorine dioxide  by the DPD-FAS  tech-
nique  (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 .    The discrepancy  between the DPD-FAS and
the iodometric methods was observed not only in wastewater, but  also  in  deion-
ized water, distilled 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 (8,9,10) have reported that  the  pH used in the
iodometric analysis (pH  ~  1.8)  is  not low  enough to assure complete  reduction
of the  chlorine  dioxide  to Cl~ (see  Fig.  3).   If,  indeed,  this is  the  case,
the iodometric  results would give a  low value  for the concentration of  chlo-
rine dioxide in the stock  solution.   The DPD method, which is performed  at pH
6.2 to  6.4,  is assumed to measure the reduction of chlorine dioxide to  chlo-
rite,  and would  therefore  indicate a  higher  value of  chlorine dioxide  than
that given by  the  iodometric analysis.  It was hypothesized that  the iodomet-
ric analysis was not measuring all the chlorine dioxide and hence  the discrep-
ancy between the two methods.


                    CIO2 + e~  - -   CIO2~          pH 7
                                        Cr + 20 —   pH<2
         Figure 3.  Reduction of chlorine dioxide as a function of pH.

                                       16

-------
     To  test  this hypothesis,  two  experiments were  performed.   In  the  first
experiment, the iodometric analysis was performed at  various  pH  values  ranging
from 2.0  to  0.8.   No difference  in  titration value was found.  In  the second
experiment, the  iodometric analysis  performed at pH  1.8  was compared  to  the
amperometric  analysis  at pH  7.   The amperometric  method  is analogous  to  the
iodometric method with these  exceptions:

     1.  In  the  amperometric  analysis  the endpoint  is detected by  ob-
         serving the change in current flow in the  sample while  adding  a
         strong reducing agent as titrant.
     2.  In the  amperometric analysis, phenylarsine  oxide (PAO) 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 electron transfer is possible  for
         chlorine dioxide.

     Since at pH 7,  chlorine dioxide  is  reduced only  to  chlorite, a higher
concentration  of  chlorine  dioxide  as  measured by  the amperometric  analysis
would  indicate  that  chlorine dioxide was  not being fully reduced  to chloride
as  assumed in  the  iodometric  method.    The   analyses were  run in  distilled
water.   The  results  appear in  Table  4,  and a statistical analysis  appears in
Table 5.

          TABLE 4.  COMPARISON OF AMPEROMETRIC AND IODOMETRIC METHODS
                     FOR DETERMINING CONCENTRATION OF C100
Trial
1
2
3
4
5
Mean
Std.Dev.
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/1  as  C102'
        TABLE 5.  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-Ratio
1.12*
                TOTAL
397.5
 Not significant  (p < 0.05).
                                       17

-------
     As seen in Table 5, the difference between the values obtained  by  the  two
methods is not significant.  It can be inferred from these data that the  iodo-
metric  method  accurately  measures the  concentration  of  chlorine dioxide  in
stock solutions.   The inconsistency must  then  be in the assumed  reactions  of
the DPD-FAS method.   It  is hypothesized (although untested at this  time)  that
the chlorite from  the reduction of chlorine  dioxide  reacts with  the DPD  rea-
gent  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 Back-Titration

     The  amperometric titration measures  iodine  (I~)  released  into  solution
when  iodide  is oxidized by  any powerful  oxidant  such  as chlorine  dioxide  or
chlorine.  If this released  I~  comes into  contact with organic constituents  in
the sample  (as  in  the case  or  wastewater),  the 1^ will react with  the organ-
ics.  To prevent this, a back-titration procedure  is recommended,  in which the
liberated Io is  immediately reacted 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 I~ solution  of
known concentration.   The concentration of  chlorine  dioxide (or  chlorine)  is
then determined by difference.  To evaluate  the  need for this back-titration
procedure,  chlorine  dioxide  residual  measurements  were  made  in  Palo  Alto
secondary effluent,  comparing the amperometric method  with and without  back-
titration.

     The results of  this experiment are reported in Table 6.  The  experiment
entailed  dosing  2  liters  of Palo Alto  secondary effluent  to  4.50 mg/1  with
pure chlorine dioxide.  The  sample was then contacted for  2 minutes.  After



Sample
1
2
3
4
5
6
7
8
Mean
Std.Dev., s
s2
TABLE 6.
Secondary
Forward
Titration
1.90
1.86
1.90
1.84
1.81
1.73
1.67
1.91
1.80
±0.083
0.0069
AMPEROMETRIC TITRATION EVALUATION
Filtered
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
Forward
Titration
1.62
1.62
1.52
1.52
1.52
1.39
1.39
1.27
1.49
±0.109
0.0119
Unfiltered
Back
Titration
1.88
1.97
2.07
1.88
1.88
1.88
1.88
1.79
1.90
±0.077
0.0060
      Results  are given  as mg/1  C102  at 2-minute  contact  time; dose  = 4.50
mg/1 CIO?-  Titrations were performed at  pH  7.

                                      18

-------
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  ^ back-
titration, sufficient  PAO was  added  to swamp the released  !«•   Eight  repli-
cates were  run at  approximately  one-minute  intervals.   The  experiment  was
repeated for filtered  Palo  Alto secondary effluent.   The values reported have
been corrected for blank titrations.  In the case of the forward titration,  no
correction was necessary,  but for the  I~  back-titration more 1^ solution was
required in the blank  than  was  indicated by the amount of PAO added.   (The  1^
solution was  standardized immediately  prior  to  use.)   Although not shown  in
Table 6,  the  blank  titrations  (done in triplicate)  showed the same  variation
as the forward or backward  titration.   The standard deviations shown in Table
6 are not statistically different from  one another as measured by the F-test.

     An analysis of variance (ANOVA)  comparing  the forward titration with the
I~ back-titration  showed  that  there  is a significant  difference  between the
values measured by the two  titrations  (Table  7).  Based on the ANOVA results,
the amperometric back-titration method  was chosen for the measurement of chlo-
rine dioxide residuals in wastewater.
           TABLE  7.   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
ANOVA for Unfiltered Secondary
Effluent:
  Between forward and back-
  titrations
  Error

                   TOTAL
     1
    14
    15
     1
    14
    15
0.54023
0.09135

0.63158
0.67651
0.14274

0.81924
0.54023
0.006525
 82.79
 0.67651
 0.01020
 66.35
 P < 0.001  that F is exceeded.
Chlorite Interference

     To investigate the possible interference of chlorite  in  the  determination
of chlorine dioxide by  the  amperometric method, a solution of  sodium chlorite
was added  to  a sample containing  chlorine  dioxide in distilled water.   Since
the sodium  chlorite  used was only  ~ 95% 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  (7).  The  results  of
this  experiment  (Table  8)  showed  no  significant  difference  caused  by the
presence of chlorite.
                                      19

-------
       TABLE  8.   THE  EFFECT  OF  CHLORITE  ON THE DETERMINATION OF CHLORINE
                       DIOXIDE  BY AMPEROMETRIC TITRATION
                                           C102 + 4.33 mg/1 C102"
Sample
I
2
3
4
5
cio2
5.47
5.55
5.72
5.61
5.59
+ Glycine
5.55
5.51
5.55
5.61
5.61
                  Mean          5.59                5.57

      Two ml of 100 g/1 solution of glycine added to 200 ml of sample.


     The amperometric back-titration method for measurement of chlorine  resid-
uals in wastewater is a widely used and well-accepted method  (11).   Since  this
method was found acceptable for chlorine dioxide residual measurements,  it was
also chosen for the measurement of chlorine residuals.  The analyses  for chlo-
rine residuals  (as  Total Available Chlorine)  are  performed at pH 4  to  ensure
quantitative response from  all  forms  of combined chlorine (11).  Results  from
our laboratory  indicate  that  at  pH 4 there is no need for a  blank correction.
The reaction of PAO with I~ must be somewhat  slower at  pH 7  than at pH 4,  so
that some  of  the added I2  titrant  would be available for reaction with other
sample constituents; hence  the  need for blank correction in  determining chlo-
rine dioxide residuals at pH  7.
MEASUREMENT OF CHLORINE DIOXIDE GENERATOR PRODUCT COMPOSITION

     Chlorine  dioxide  for  use  in  wastewater  disinfection is  generated  by
either a  chlorine-sodium  chlorite reaction or  by  a mineral acid-sodium  chlo-
rite reaction.   Impurities  of the reactants  and incomplete reaction  necessi-
tate a  methodology to  identify  all  possible  chlorine  species.  The  possible
chlorine species in the final products are:

                   1.  Chlorine              - C12,  HOC1
                   2.  Chlorine Dioxide      - C102
                   3.  Chlorite              - C102~

                   4.  Chlorate              - C103~
                   5.  Chloride              - Cl~

Analysis of Chlorine  Species

     The  analytical methodology developed  for  this study to identify reactor
product  composition  is essentially  an iterative  process in which  successive
measurements  include  one  more chlorine species  than  the  previous measurement.
The method is  as follows:

                                       20

-------
     1.  Measurement of  chlorine  dioxide:   Chlorine dioxide absorbs  ultravio-
let radiation in the 240-440 nm wavelength.  A wavelength of 360 nm was  chosen
for chlorine  dioxide  determinations because  this corresponds  to  the peak  of
the absorbance  curve;  also,  there  is  no interference  from  other   chlorine
species at this wavelength.

     A calibration  curve was prepared by  appropriate  dilutions of a  standar-
dized  stock solution  to  yield concentrations in  the 5 mg/1 to 60 mg/1  range.
The response of C102 is  linear in this concentration range  and  an  equation can
be  derived  relating  concentration  to  absorbance.   A  typical concentration
equation with correlation coefficient r  is given below:


                   mg/1  C102 = 58.26 x (absorbance) - 0.255

                                  r2 = 0.998

     The calibration  equation will  depend on the  instrument  and path  length
used and should be checked often.

     A sample  of reaction  product was  diluted so  that  the concentration  of
C10? to be  analyzed was  in  the linear  range  (5-60 mg/1).  The instrument was
set to zero absorbance with  a blank of the dilution water just  prior  to  sample
analysis.    The  concentration of  C102  in  the diluted solution was calculated
using  the  calibration equation  and the  appropriate dilution  factor was  ap-
plied.  This was "Reading A."

     2.  Measurement  of  chlorine dioxide  and  chlorine:   Chlorine dioxide and
chlorine were measured  by forward amperometric  titration at pH 7.   An  appro-
priate sample size  was chosen so  that  no  more than  5 ml  of phenylarsine oxide
were used in  the  titration.   To approximately 200 ml of  distilled water,  were
added  2 ml pH 7 buffer,  the  sample, and 2  ml of  5% KI  solution, in that  order.
This mixture  was  titrated with 0.00564N_ PAO  and the resulting volume (ml)  of
PAO solution was recorded as  "Reading B."

     3.   Measurement  of chlorine  dioxide, chlorine,  and chlorite:    Chlorine
dioxide, chlorine, and chlorite were measured by  iodometric titration at pH 2.
To 50  ml of distilled water  were  added 5  ml of concentrated acetic acid,  5 ml
of sample and  approximately  1 g of  KI.   This  mixture was  allowed to  react  in
the dark for 5 min and was then titrated to a starch endpoint with 0.1N_ (stan-
dardized)  sodium thiosulfate.  This was "Reading  C."

     4.   Measurement  of chlorine dioxide, chlorine,  chlorite, and  chlorate:
This procedure was essentially as in Step  3 of this  enumeration, with the  fol-
lowing exceptions.   To  a  50-ml  erlenmeyer flask was  added 5 ml  concentrated
hydrochloric acid and 0.5 g  KI. To this mixture was  added 5 ml  of  sample.   The
mixture was sealed  and  allowed to react  in the  dark in a  closed  flask  for  10
min.   A blank was also  carried through  the  procedure since KI is oxidized  by
oxygen in this  low  pH environment (pH <  0.1).   After 10 min,  the mixture was
diluted to  ~ 50  ml and titrated  with 0.1N_ sodium  thiosulfate.  This reading
minus  the blank value was "Reading D."


                                      21

-------
     5.  Measurement of chloride:  This procedure was essentially the Mercuric
Nitrate Method (408B) as given  in  Standard Methods (1) with the following ex-
ception:  when measuring the chloride concentration in a reaction mixture con-
taining high  levels of  chlorine dioxide,  the  sample  was diluted  to  yield a
final  chlorine  dioxide concentration  of  about  10 mg/1.   To 100  ml  of this
solution was added  4.0 ml  of  indicator-acidif ier reagent instead of 1.0 ml as
given in (1).  The  sample  was then titrated as specified.  This reading minus
the blank value was "Reading E."

     All calculations,  with  the exception  of  the  chlorine dioxide/ spectro-
photometric measurement and the chloride measurement were based on  the equiva-
lents of reducing  titrant  required to react  with the  equivalents  of oxidants
present.   The  equivalent weight of  a compound was defined  as  that weight of
the compound which  contains one gram atom of available electrons,  i.e., equi-
valent weight  equals  molecular weight divided  by number  of electrons trans-
ferred.  For chlorine dioxide,  the equivalent weight is pH-dependent , because
at pH 7 chlorine dioxide is reduced  to chlorite, while at pH < 2 chlorine di-
oxide reduces  to  chloride.   Chlorite and  chlorate also exhibit an equivalent
weight pH-dependency .  A listing of  the  equivalent weights used in the calcu-
lations is shown in Table 9.

Evaluation of Chlorine Dioxide Reactor Yields and Reactor Stoichiometry

Acid Activation of  Sodium Chlorite —

     When  chlorine  dioxide is  generated  from  sodium  chlorite  by  acid acti-
vation, chlorine dioxide  is not the  only  end product.  The final  composition
is dependent on  several variables, among  which are:   sodium chlorite concen-
tration; purity of  sodium chlorite used; acid concentration  and pH  of reaction
mixture; reaction time; and temperature of  reaction.

     Several stoichiometries have been reported  in the literature for acid  ac-
tivation of a  chlorite  solution (12).  The two  stoichiometries which are most
widely accepted are (shown as sulfuric acid activations):
                  4NaC102 + 2H2S04 = 2Na2S04 +  2C102 + HC1 + HC103  + H20   (1)

and

                  10NaC102 + 5H2S04 = 8C102 + 5Na2S04 +  2HC1 + 4H20        (2)


     The  first reaction  is  catalyzed  by the  chloride  ion,  which is  also  a
product  of  the reaction.  Not  only  has it been reported that the  reaction is
accelerated  by the product,  chloride ion,  but  also that the Stoichiometry is
altered  to  that  of  the  second  reaction (12).   Although  chlorine  has  been
reported  among  the  reaction  products   of  chlorine dioxide  generation,  most
evidence  appears  to be against  the  formation of significant amounts of chlo-
rine  (12).
                                       22

-------
 TABLE  9.  EQUIVALENT  WEIGHTS  FOR CALCULATING CONCENTRATIONS ON A MASS BASIS
pH
7
2, 0.1
7, 2, 0,
2, 0.1
0.1
Species
Species
C102
C102
.1 C12
C102~
cio3~
Calculations :
Molecular Weight
(mg/mole)
67,450
67,450
70,900
67,450
83,450

Electrons
Transferred
1
5
2
4
6

Equivalent
Weight (mg/eq)
67,450
13,490
35,450
16,863
13,908

  Chlorine dioxide:  mg/1 CK>2
  Chlorine:  mg/1 C12
              6     *
  Chlorite:  mg/1 C102"
                        35'45°
                           eq
         58.26 (absorbance) - 0.255 = A


        L (-B(ml) x normality(eq/l) _  mg/1 C1°2   >
          ^      sample(ml)          67,450 mg/eq'
_ 16,863 mg j-C(ml) x normality(eq/l)
    eq      *•      sample (ml)

      mg/1 C12      mg/1 C102
  ~ 35,450 mg/eq ~ 13,490 mg/eq-'
  Chlorate:  mg/1 ClOo"
  13,908 mg /-D(ml) x normality(eq/l)
     eq     *•      sample(ml)
                                             mg/1 Cl,
                                   mg/1 CIO,
 mg/1 C102
16,863 mg/eq ~ 35,450 mg/eq ~ 13,490 mg/eq-
  Chloride:  mg/1 Cl" - 35.450 mg/eq x E(ml)  x normality(eq/l)
              °                       sample(ml)

  where A, B, C, D, E denote the respective Readings A, B, C, D, E, as  defined
  in the text.
Reaction Stoichiometry in Laboratory Generation —

     To determine reaction  stoichiometries,  a mass balance on chlorine around
a reactor  must be made  using the analytical  technique previously  described.
Prior  to  examining a  commercial reactor, a laboratory reaction mixture was
prepared,  a  mass  balance performed, and  a reaction Stoichiometry determined.
There  are  significant  differences between laboratory  reaction conditions and
those  typically  found  in  full-scale  commercial  reactors.    Therefore, the
stoichiometries observed  would not necessarily  be the same.   In the labora-
tory,  the  concentration of  sodium chlorite in the reaction vessel was 0.083M,
while  in commercial reactors  the concentration is typically 2.0M.   Concentra-
tion not only  affects  the time  until  the reaction is  complete but  may affect
                                      23

-------
the stoichiometry  as  well.   In commercial  reactors the concentrated  reaction
mixture is allowed  to  react  for  approximately 5 to 25 min and  is  then diluted
(within the reactor) before being fed to the process stream.

     The initial and final compositions of the laboratory reaction mixture  are
shown  in  Table 10.   Under  these conditions the  laboratory  reaction  was com-
plete within  3 hrs; no change in composition  was  found after this time.   The
recovered mass of  chlorine  represents  104% of the initial  mass  of  chlorine;
considering experimental error, this value is not significantly  different from
the expected value of  100%.

     If  the  initial  composition of  the reaction  mixture is  equated to  the
final composition,  the following equation results:


     0.083 C102~ + 0.0004 C103~ + 0.0177 Cl~

                 - 0.0037 C102~ + 0.0206 C103~ + 0.0424  Cl~

                                + 0.0380 C102 + 0.0007 C12                  (3)

   TABLE  10.   COMPOSITION OF  REACTANTS AND  PRODUCTS IN LABORATORY GENERATION
      OF CHLORINE DIOXIDE BY SULFURIC ACID ACTIVATION OF SODIUM CHLORITE
Species
Reactants
cio2
C102~
C103~
C12
Cl"
Total
Final Composition (pH 1.75)
C102
C102~
C103~
C12*
HOC1*
Cl"
Total
Chlorine Atoms
Moles per liter
0.0
0.083
0.0004
0.0
0.0177
0.1011
0.038
0.0037
0.0206
0.0004
0.0005
0.0424
0.1056
 Calculated  assuming  equilibrium is  attained in chlorine hydrolysis,
 KH  =  3  x  10~4.

                                       24

-------
Collecting terms  and  relating all species  to  a reaction involving  four  moles
of chlorite,  analogous to Reaction 1, yields:


           4 C102~ = 1.92 C102 + 1.02 C103~ +  1.25 Cl~ + 0.04  C12           (4)


This stoichiometry is essentially that of Reaction 1.

Reaction Stoichiometry in Full-Scale, Continuous Generation—
     A reaction  mixture  from a  continuous,  full-size generator  (commercially
available  from  Rio-Linda Chemical Company)  was analyzed.   This reactor  uti-
lized the  sulfuric  acid  activation of sodium  chlorite.   The  reactor  consists
of  a  section of  PVC  pipe,  100  mm in diameter and 660 mm  long, packed  with
turbulence-inducing packing.   Its  nominal  capacity is approximately 2  kg/h
C102 (Fig.  4).    The  composition of reactants  (calculated  from the flow rate
and composition of reactant  streams) and products (obtained  from determination
of  reactor product  composition)  are shown  in  Table 11.   The  sample was  taken
four hrs  before  the  analyses were  performed,  and it  is   not  known  to  what
extent the reaction may  have proceeded during  transportation.   Therefore,  the
values reported  in Table  11 should be  taken  as equilibrated values  for  the
reaction mixture  at the  time of analysis.

     The recovered mass  of chlorine atoms represents 104% of the initial  mass;
this is not significantly different from 100%.  The  stoichiometry related to a
reaction involving four  moles of C102~ is:


           4 C102~ = 2.63 C102 + 0.53 C103~ +  1.01 Cl~ + 0.05  C12           (5)


which can be compared  to the result  from the laboratory reaction (Eq. 4),  or,
on  the basis of ten moles of C102~ reacting:


          10 C102~ = 6.58 C102 + 1.32 C103~ +  2.52 Cl~ + 0.12  HOC1          (6)


     Comparing Eqs.  5 and 6 with Eqs. 1  and  2, it  appears  that the reactor
stoichiometry falls between  stoichiometries 1  and  2.   This may  indicate that
the overall  reaction  in  an  acid-chlorite  generator  is a combination  of  Reac-
tions 1 and  2.    In addition,  from  Table  11,  it can be  seen  that only 78% of
the added  chlorite  is participating in the  reaction at  all,  and of that 78%,
66% is  converted to  chlorine  dioxide.  This  gives an overall  yield  of  51%,
i.e., 51% of added  chlorite  is converted  to chorine dioxide,  10%  is converted
to  chlorate, 22%  is unreacted  chlorite,  and the remainder is  accounted for as
chloride and  a  trace  of chlorine.   Under  the most  favorable conditions,  the
maximum yield of  chlorine dioxide found from  the  sulfuric  acid  activation of
sodium chlorite was in the 50-55% range.
                                      25

-------
            SODIUM
            CHLORI
                N
              No CIO,
             REACTION
             COLUMN
           PRESSURE
           REGULATOR
       WATER
       FLOW
n
ii
n
ii
ii
                         VENTURI
FLOW
RATE
METERS
           HCI or
 No CIO VACUUM  LINE
      2
              -ACID  VACUUM LINE
                      SIGHT  TUBE

                   tt&C


                       DRAIN
                  REACTION   COLUMN
                DIAMETER
                LENGTH
                CAPACITY
               10 cm  I.D.
               66 cm
                5 liters
        Figure 4.  Chlorine dioxide generator, acid-chlorite process.
     Since it had been reported (12) that  the chloride ion catalyzes the reac-
tion and  alters the stoichiometry to the  more  favorable Reaction 2, the com-
mercial reactor under study was converted  to hydrochloric acid activation from
sulfuric acid activation.  To  evaluate this generation process, a mass balance
on chlorine atoms was again performed around the  commercial reactor.  The re-
sults of this mass  balance are reported  in Table 12.

     The recovered  mass  of chlorine  represents  99.7% of the initial mass.  If
the stoichiometry of this mixture is determined and related to an equation in-
volving 10 moles of chlorite,  the following stoichiometry is found:
              10 C102~ =8.04 C102 +0.28 C103~ +1.58 Cl"

                                   26
                                        (7)

-------
    TABLE 11.  COMPOSITION OF REACTANTS AND PRODUCTS FROM CHLORINE DIOXIDE
   GENERATION IN A CONTINUOUS, FULL-SIZED ACID-CHLORITE REACTOR USING H2S04

                                                      Chlorine Atoms
      Species                                          Moles  per liter

  Reactants
      C102                                              0.0

      C102~                                             0.0214

      C103~                                             0.0003

      C12                                               0.0
      Cl~                                               0.0021
  Total                                                 0.0238

  Final Composition (pH 2.1)

      C102                                              0.0109

      C102~                                             0.0048

      C103~                                             0.0025

      C12*                                            < 0.0001

      HOC1*                                             0.0002

      Cl~                                               0.0063
  Total                                                  0.0247

 Calculated assuming equilibrium is attained in chlorine hydrolysis,
    = 3 x 10~4.
     Reaction 7 very closely approximates Eq. 2, the  stoichiometry  reported  by
Gordon et al. (12) as optimum for acid activation of  sodium chlorite.

     From Table  12  it  can be seen  that  97% of  the chlorite has reacted,  com-
pared  to  only 77%  previously reported  for activation  of  sodium chlorite  by
sulfuric acid.   Of  the feed  chlorite  78.2% is  converted to chlorine  dioxide.
The average conversion of feed sodium chlorite observed  in three field experi-
ments  (Table  13)  was 78.5%.   In  view  of the agreement  between laboratory and
field  data,  it  is reasonable to  expect  consistent  reactor yields of  78%  con-
version of sodium chlorite to chlorine dioxide  by acid  activation with hydro-
chloric  acid.    The complete stoichiometry  expected from  hydrochloric  acid
activation of sodium chlorite is as follows:
                 5NaC102 + 5HC1 = 4C102 + 5NaCl + HC1 + 2H20                (8)
                                      27

-------
    TABLE 12.   COMPOSITION OF REACTANTS AND PRODUCTS FROM CHLORINE DIOXIDE
    GENERATION IN A CONTINUOUS,  FULL-SIZED ACID-CHLORITE REACTOR USING HC1
                                                    Chlorine Atoms
     Species                                       Moles per liter
  Reactants
      C102                                              0.0
      C102~                                             0.0110
      C103~                                             0.0000
      C12                                               0.0000
      Cl~                                               0.0260
  Total                                                 0.0370
  Final Composition (pH 2.2)
      C102                                              0.0086
      C102~                                             0.0003
      C103~                                             0.0003
      C12 and HOC1                                    < 0.0001
      Cl~                                               0.0277
  Total                                                 0.0369
         TABLE 13.  YIELD OF CHLORINE DIOXIDE FROM ACID ACTIVATION IN
                              FIELD  EXPERIMENTS
Date
1-16-79
2-5-79
2-13-79
Average
C102~ - Feed (mg/1)
2111.4
2072.3
1454.3
conversion
C102~0utput (mg/1)
1573.4
1581.0
1231.0

% yield
74.5
76.3
84.6
78.5
Chlorine-Chlorite Generation of Chlorine Dioxide—
     An alternative generation method  for  chlorine dioxide utilizes the reac-
tion of  sodium chlorite with chlorine.   A mass balance  on chlorine around a
reactor using  this reaction was  attempted  but time limitations of the analyt-
ical procedure precluded a meaningful mass balance.  The  simplified  stoichiom-
etry for this reaction is:
                                      28

-------
                        2NaC102 + C12 - 2C102 + 2NaCl                       (9)

     From Reaction 9, theoretically  100% of  the sodium chlorite can be  reacted
to chlorine dioxide.  An excess of chlorine above the stoichiometric require-
ment as predicted by  Reaction 9 has been found to increase yields of chlorine
dioxide substantially and ensures complete conversion of sodium chlorite.   The
excess chlorine  (predominately as  HOC1) presented the analytical problem men-
tioned  above.    Hypochlorite  ion  reacts with  chlorine  dioxide  over  time  to
produce chlorate ion according  to Reaction 10:


                   2C102 + OC1~ + H20 = 2C103~ + Cl~ + 2E+                 (10)


     While  this  reaction is  not instantaneous,  it  is  sufficiently  rapid  to
cause  significant  decreases  in  the  chlorine  dioxide  concentration  between
sampling time and analysis for  mass  balance.  Therefore, samples were analyzed
immediately for  chlorine dioxide to determine yields of chlorine dioxide,  and
chlorite measurements were made within  6 hours to estimate extent of reaction.

     Chlorine dioxide yields were found to  vary  from 93 to 98% based  on  feed
chlorite.   Chlorite conversion amounted to  97%  of that fed  to  the reactor.
Chlorine feed was  approximately 4% above the stoichiometric requirement based
on chlorite.

Chlorine Dioxide from Sodium  Chlorate—

     Chlorine  dioxide  may   also  be  generated directly from  chlorate (13),
rather than passing through  the intermediate step of reducing the chlorate to
chlorite.  Generation of chlorine  dioxide  directly from chlorate is practiced
widely in  the  pulp and   paper  industry, but  not in water treatment.  Problems
of  reactor instability reportedly  are encountered.    There  is  an economic
incentive  for  developing a  generation  process based on chlorate  for  use  in
water  and  wastewater treatment.   The  chemical costs are  considerably lower
than in processes based  on chlorite.
                                      29

-------
                                   SECTION 6

                        MODELS  OF  DISINFECTION KINETICS


FIRST-ORDER KINETICS

     To compare  the relative disinfection  capabilities  of chlorine and  chlo-
rine dioxide, an  empirical model relating  degree  of  bacteria inactivation  to
the contact time and residual concentration was used in  this  study.  The  model
was  first  applied  to  disinfection  to  approximate  kinetics  that deviated
greatly from  the  often assumed first-order  disinfection kinetics referred  to
as "Chick's law."

     In 1908, Harriette Chick published her  "Investigation of  the Laws  of Dis-
infection"  (14).   She examined  the destruction of several bacterial organisms
in buffered distilled water  by  phenol and other compounds.   It  is interesting
to note  that  in only one  experiment, the  destruction of unicellular  anthrax
spores by phenol,  did she  observe the rate expression which  carries her  name.
That expression is the familiar first-order kinetic rate  expression:


                                   || = - kN                               (11)


or in integrated form:


                                   N     e~kt                              (12)
                                  N(0)


where N    = number of surviving organisms at time = t;

      N(0) = number of organisms at time = 0;

      k    = rate constant.


DECREASING RATE AND LAG TIME

     In  the  majority of  experiments,  however,  Chick observed  that the  rate
constant k did not remain constant and in fact decreased over  the course  of  an
experiment.  This type of relationship is generally what is  observed in waste-
water  disinfection:   namely,  a  decreasing  rate  of  bacterial inactivation  as
the  disinfection  process  proceeds (15).   In wastewater disinfection, however,
the  disinfection process begins after some finite time interval, the lag

                                      30

-------
     TYPES  OF  BACTERIAL SURVIVAL  CURVES
cr
ID
en
     CHICK'S

     OBSERVATIONS
WASTEWATER

DISINFECTION
                                    CHICKS  LAW
                LAG  TIME
                         TIME
      Figure 5.  Types of bacterial survival curves,



                     31

-------
time.   (Chick  also  observed a lag  time with  some pure cultures in clean  sys-
tems.)  These relationships are shown in Fig. 5.

     Several hypotheses have been  formulated  to explain both the lag  time and
the decreasing rate of disinfection:

     1.  The lag time is  indicative of  the diffusion to and transport
         across the cell  membrane  and the  reaction time for inactiva-
         tion within the cell (16).
     2.  The organisms are  not isolated  but are associated in groups,
         so that the internal organisms  are protected by those on the
         periphery  of  the  agglomeration.   The  lag time  is  then re-
         lated  to  the size  of  the  aggregation and is  indicative  of
         the time to reach the inner-most organism  (17).
     3.  The decreasing  rate is due  to  a decrease  in the germicidal
         properties of the disinfecting agent, i.e.  residual  (18).
     4.  The bacterial  population has an  ab origine  distribution  of
         susceptibilities  to the disinfecting  agent so that  the ob-
         served  decreasing  rate  is  the  sum  of several  first-order
         rate constants (19).
     5.  The decreasing rate is  the result of the bio-chemical disin-
         fection  reaction  inducing an  increased  resistance  in  the
         population of organisms, i.e. induced heterogeneity  (20).

     Chick postulated  several  mechanisms of  bacterial destruction to  account
for the decreasing rate of inactivation (14), and attempted to  test them.   She
found that the rate of inactivation she observed was related  to the age of the
pure culture used and the seed material from which  the  culture  was grown.
RETARDANT REACTION APPROACH

     In  the  1930s,  Fair  and Thomas  (21) observed  the same  type of  kinetic
behavior  in  studies  of  biochemical  oxygen demand,  (BOD).    Fair applied  an
empirical model,  called  the  "retardant"  formula,  to  his  observations of  de-
creasing rate of  reaction.   As  applied to the BOD reaction, this  rate expres-
sion is:


                                  dL      k   T
                                              L
                                  dt ~ 1 + at

and in integrated form is :
                                   LQ(1 + at)"k/a                          (14)
where L  = BOD;
      LQ = ultimate BOD;

      Lt = BOD at time t;

                                       32

-------
      k  = velocity constant;

      a  = coefficient of retardation; and

      t  = time.

An intuitive rationalization  of  this  expression for BOD is based  on  the  rela-
tive rates of oxidation  by  bacteria of the organics present  in a  test  sample.
The least stable  compounds  are oxidized at a high rate early in the  test.   As
these compounds are  depleted, the rate  of  reaction becomes a function of  the
more stable organics  and the  overall  rate of reaction slows.  This same  rela-
tionship was also applied by  Fair and Geyer  (22) to the removal of BOD  through
a trickling filter.

     Phelps (23)  reported on  the  self-purification of streams  and rivers  as
regards bacterial numbers.  It was found that the bacteria disappear  according
to a decreasing rate of  inactivation and that Fair's retardant formula  (21)  is
applicable to this situation  also.

     In 1957,  at a Ciba  Foundation Conference (24), a discussion between  Jonas
Salk, the discoverer of  polio  vaccine,  and Sven  Card,  a Swedish researcher,
centered on the kinetics of  chemical  virus  inactivation.  There was more than
just an academic interest in  this area since children in the  U.S.A. were  being
vaccinated with  formaldehyde-inactivated polymyelitis virus.  Salk  contended
that  the  inactivation  proceeded according  to  first-order  kinetics   (Chick's
law), while  Card   presented  evidence  that  the  rate of  inactivation  deviated
from first order in that  the  rate was decreasing over time.   Card  then  applied
a model identical to Fair's  retardant  formula  (although  Card did not  call  it
that,  nor  did  he reference  it)  to his  data  and  showed  excellent  agreement
between his  experimental data  and  this model.   In Card's  formaldehyde-virus
system, he observed  that both an increase  in  formaldehyde and an increase  in
reaction time had the same effect  on  inactivation.   He reasoned  that  concen-
tration (R) and time  (t)  should be interchangeable and  should  appear in  the
reaction equation together as a  factor "Rt,"  This is equivalent to the state-
ment  that  for a  given  level  of inactivation  Rt  = constant  .   Card merely
stated this conclusion with no explanation or evidence to support  it.

     This relationship between  concentration  and time had  been discussed  by
both Chick (14) and  Watson  (25).  Watson formulated as a general  relationship
that   Rnt  = constant  .    He  interpreted  this   relationship as  a   chemical
reaction between  one  molecule of one  substance with  n  molecules of  a second
substance,  or analogous  to a  chemical process:


                                  [R]°[t] = K                              (15)


with   [t]   as an artificial  concentration.   In kinetic  terms, n  depicts  the
empirical reaction order with respect  to disinfection concentration.   Fair  et
al. (26) reported some experiments on  the effect of concentration  on  bacterial
inactivation and  found  n  to  vary from 0.8 to  1.4  for  free chlorine and   n
to  equal  1.3 for chloramines.   The  value  of   n  is  probably  temperature-

                                       33

-------
dependent and  organism-specific.   Incorporating  this  concept into the  retar-
dant formula, Card's model (taking  n = 1) as a rate equation for  inactivation
is:


                                dN _      k
                                dt   1 + a(Rt)


or in integrated form


                          N(t)  = N(0)  (1  + a(Rt))~k/a                      (17)


where N    = number concentration of target organism;

      N(t) = concentration at time t;

      N(0) = initial concentration;

      k/a  = velocity coefficient;

      R    = concentration of chemical species held  constant over  time;

      a    = coefficient of retardation; and

      t    = time.

It is  important to note  that  in the integration  of the above rate  reaction,
the concentration  (R) is  assumed to  be constant during  the entire time  inter-
val of integration.

     Selleck et al. (27) have applied Card's model to  the disinfection process
in primary effluent.  They  have, however, made some modifications and relaxed
the assumptions required for integration of Card's rate  equation.

     Typically  in  wastewater disinfection,  there is  a lag  time, t',  during
which no disinfection or bacterial inactivation takes  place.   This is true for
constant concentration  of chemical disinfectant.   If, however, the  disinfec-
tant concentration  R  is  allowed  to  vary,  then a minimum product  (Rt)'  must
be exceeded  before inactivation takes place.   Selleck et al.  (27) have taken
the basic rate  expression as:


                                  || = - kN                               (18)


where


                           k -  0 for  (Rt) <  (Rt)'                        (19)

                           k = k'  for  (Rt) = (Rt)'                        (20)

                                      34

-------
and
Equations 20  and  21 above are  equivalent  to  the statement that  (Rt)' =  a  .
Integrating the rate expression for the general  case and applying the  boundary
condition that at  Rt = (Rt)' ;  N(t) = N(0)  yields:

                                        ,_..   -k'/a
                            N(t) = N(0)[|gjr]                             (22)


                        N(t) = N(0)  for   (Rt) <  (Rt)1                     (23)
where all variables are defined as before.

     An implicit assumption made  in  the integration of the above  rate  expres-
sion is  that  the concentration  (R)  remains constant.   In  the case of waste-
water disinfection with either chlorine or  chlorine dioxide,  the concentration
(R) decreases with time.  Selleck et al.  (27) use a mathematical extrapolation
to estimate the  active  residual  concentration of chlorine after the  immediate
demand of the wastewater  has  been satisfied and assume the bactericidal  prop-
erties of the residual to remain constant.

     In this  study,  the modified retardant  formula  has been used  to describe
the disinfection process  in  several  wastewaters of varying quality.  In  order
to incorporate  into  the model  some  means  of adjusting for the difference  in
quality among wastewaters  and hence the  oxidant  demand,  the residual  oxidant
is measured at the actual exposure time of  the bacteria.  This  allows the more
general application  of  the model and  renders more meaningful  the  comparisons
between wastewaters  of  different levels  of  treatment.   In addition, measure-
ment of residual at the end of the organism exposure  time yields a lower  limit
on (Rt) for the  measured  level  of inactivation.  The model used in this  study
is given by Equations 24 and 25:
                        N(t)  = N(0)P^-]     for Rt > b                   (24)


                           N(t) = N(0)   for  Rt  < b                        (25)


where N(0) = initial total coliform bacteria;

      N(t) = total coliform bacteria at  time = t;

      R    = residual oxidant in mg/1 as measured at  time  t;

      b    = lag time coefficient (Rt)'  in mg-min/1;


                                      35

-------
      k    =-k.'/a = velocity coefficient; and

      t    = time in minutes.

     In all  of  the  above rate expressions,  the assumption has been made  that
the disinfectant  concentration is  not  a function of  time.   This  is not  the
case,  but  without this  simplification, an  expression for  the  residual as  a
function of  time is  required and  to  date  no  satisfactory,  generally valid
formulation exists.

     It must be emphasized  that  the model presented for use in this  study  has
no rational, mechanistic  basis in describing disinfection by  chemical agents.
Nonetheless, it does  approximate empirically the  behavior of  the real  system
and as such provides a useful design tool.
                                       36

-------
                                   SECTION 7

        DISINFECTION EXPERIMENTS WITH CONVENTIONAL COLIFORM INDICATORS
EXPERIMENTAL DESIGN

     For evaluation  of  chlorine dioxide as a wastewater disinfectant,  experi-
ments were  performed in  wastewaters from  three treatment  plants  in the  San
Francisco Bay Area.   Identical  experiments were run in parallel with chlorine
as the  disinfecting  agent  in  order that  the  relative efficiency of  chlorine
dioxide  could  be determined.   The  wastewaters  used were  from the Palo  Alto
Water  Pollution  Control  Plant  (Palo  Alto),   the  Dublin-San  Ramon  Services
District Wastewater  Treatment  Plant (Dublin),  and the  San Jose-Santa  Clara
Wastewater Treatment Plant  (San Jose).

Treatment Plant Descriptions

Palo Alto—
     The  Palo  Alto Water  Pollution  Control  Plant  is  a  non-nitrifying
activated-sludge  plant  with  a  capacity  of  1.5 nr/s  (35 mgd).   Samples  for
disinfection experiments and water-quality characterizations were of  secondary
effluent before chlorination.   Two  sets  of  disinfection experiments were  run
on Palo Alto wastewater.  The first  set  (Palo  Alto  - 1978)  were experiments in
which either chlorine or chlorine dioxide were dosed to the secondary effluent
as received from  the treatment  plant.  For both  chlorine  and chlorine dioxide,
eight separate  experiments  were performed on  five  different days over a  two-
week period  in  July  of 1978.   On  three days,  two experiments were run on the
same wastewater.  The  second  set (Palo Alto - 1979) were experiments  in  which
one disinfectant, either  chlorine or chlorine dioxide,  was dosed to  two  sam-
ples of  Palo Alto wastewater,  one  of  which  had  been filtered.   The filter
utilized was  a laboratory-scale  multimedia filter.   The  filter  is shown  in
Fig. 6.   The  filtration rate  was   20  liters/hr.   The  filter was backwashed
immediately before  each use  and  ripened by allowing  20 liters of wastewater
sample  to  pass  to waste  before collection  of 20 liters for  experimentation.
The filter was  then  rinsed with  tap water,  drained to the  top of the  anthra-
cite layer, and allowed  to  stand until it was needed  for the  next experiment.
The empty-bed volume is  ~ 1.5 liters.  These  experiments were  run on six days
over a three-week period in August  1979.  Three  experiments  were run  for  chlo-
rine and four  for chlorine dioxide,  comparing their performance between  fil-
tered and non-filtered secondary effluent.
                                      37

-------
overflow
  33cm(l3in)
  20.3cm(8in )
  I6.5cm(6.5in)
                     diameter
                 5.1 cm  ( 2 in )
  I    anthracite

  2   fine  sand

  3   garnet
                                              Reservoir
                                          mixer
53.3cm ( 21 in  )
                                     69.9cm  (27.5 in)
 valve to control
    filtration rate
       filtered
        effluent
        Figure 6.  Multimedia filtration column.


                       38

-------
Dublin—
     The  Dublin-San Ramon  Wastewater Treatment  Plant  is  a  combined-process
activated-sludge/nitrification plant.   The effluent from the  secondary  clari-
fier is filtered through a dual-media anthracite-sand  filter.  The  capacity  of
the plant  is  0.22 m^/s  (5  mgd). Samples  for laboratory disinfection experi-
ments  and water-quality characterizations  were  taken  after  filtration and
before chlorination.   These laboratory experiments were run on three days  in
January and February 1979.

     Experiments were also conducted at full-scale  at  the Dublin Plant.   Plant
flows were controlled  manually,  as were either  the chlorine  feed  rate  or the
chlorine dioxide feed rate.  For experiments  with chlorine  dioxide  as  the dis-
infectant, the chlorine  feed was turned off and chlorine dioxide feed  solution
was introduced  into the  rapid  mixing chamber through the same header  system
used for  normal  aqueous chlorine feed.   This assured that both disinfectants
would be  subjected  to  the identical conditions of mixing with the  wastewater.
Experiments using chlorine  dioxide  were  performed on the same days as experi-
ments with  chlorine, allowing  sufficient equilibration time between experi-
ments.  Chlorine dioxide was  generated in a commercially available continuous
packed reactor,  utilizing the HCl-acid activation  technique  (2 kg/hr maximum
capacity).  For doses higher than 2.5 mg/1 chlorine dioxide,  2 generators were
used in  parallel and plant  flow was diverted to  a holding pond as required.
Samples  for  bacterial  analysis  and residual  analysis  were  taken  at  the en-
trance to the contact tank, the mid-point of  the contact tank and at the over-
flow weir.   These points  had been chosen to provide reasonable contact time
for the  doses used.   These  experiments  were run  on five  days  from January
through April 1979.

San Jose—
     The  San  Jose-Santa Clara Wastewater  Treatment Plant  is a 7.0 m3/s (160
mgd) plant with separate activated  sludge and nitrification processes  followed
by multimedia filtration.  Samples  were taken of secondary  effluent, nitrified
effluent, and  filtered  effluent  for laboratory  disinfection experiments and
water-quality characterizations.  Each sample point was  sampled once each week
(not on  the  same day  nor at the  same time) and  the experiment was  repeated
three times for each sample point in June and July  of  1979.

Process Sequence and Wastewater Characteristics

     The  flow  diagrams   for these  three  treatment  plants  and the  points from
which samples were  taken for  laboratory disinfection experiments are  shown  in
Fig.  7.   Average  wastewater  characteristics for  each  plant at   each  sample
point are shown  in Table 14.   The  values given are  the means of  all experi-
ments at  the  particular sample point ±  the  standard deviation of  the parame-
ters over the period in which experiments were performed.

Design of Experiments

     Laboratory  experiments  were   performed  in  a  randomized  complete   block
design.  Each experiment consisted  of dosing wastewater samples with  chlorine
or  chlorine  dioxide at  three  concentrations,  usually 2,  5,  and  10   mg/1.
Samples were taken at  5,  15, and  30-min  intervals  for both total   coliform

                                      39

-------
        PALO  ALTO  WWTP
RAW
WASTE
PRIMARY
TREATMENT



CONVENTIONAL
SECONDARY
TREATMENT



DISINFECTION

                                               sample point
         DUBLIN -SAN RAMON  WWTP
          RAW
        WASTE
EXTENDED  AERATION
SECONDARY TREATMENT
WITH  NITRIFICATION
FILTRATION
f
DISINFECTION
                                                             sample
                                                              point
         SAN JOSE-SANTA CLARA WWTP
RAW
WASTE

PRIMARY
TREATMENT

*•
CONVENTIONAL
SECONDARY
TREATMENT


1
samp
poin

NITRIFICATION
PROCESS
s
le
t
I
am
po

FILTRATION |«j-

DISINFECTION
pie sample
nt point
              Figure 7.   Wastewater treatment plant flow schemes.
analysis  and  residual  analysis.    Samples for  bacterial  analysis  were  also
taken at  time 0.   All of the six runs within an  experiment (two  disinfectants
each at three doses) were done in a randomized order.   From three to eight ex-
periments were  performed at  each  sample point  (Palo  Alto, Dublin,  and  three
points  at San Jose).   Wastewater  for experiments  was collected  immediately
before an experiment was conducted.  The basic dose-time matrix for  an experi-
ment is  shown in Fig.  8.   In some cases,  more  than one replicate  experiment
could be  conducted  on one  wastewater sample on one  day.  However, most repli-
cate experiments  were conducted on separate days,  each with  fresh  wastewater
samples.  Therefore,  the wastewater characteristics  would not  be  identical for
all replicates of an experiment.  The  data  from these  experiments were used to
evaluate  the  constants  of  the disinfection  model  discussed  above.   In  some
instances,  the  number of dose  levels  was  augmented to include  an  additional
experiment  at a  dose of 1  mg/1,  and  the contact  times  were augmented  with
samples at  1  and 60 min.  This was  done to provide an adequate  range of  data
for use in  the model.

     The  Palo   Alto - 1979 experiments were designed  to  compare  one disinfec-
tant between  filtered and unfiltered  wastewater.   The question  of  concern in
these  experiments was  the  effect of  the  filtration  process  on disinfection
kinetics.

Laboratory-Scale  Reactor Design

     The  reactor  for  all laboratory  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
                                       40

-------
          TABLE 14.  CHARACTERIZATION OF WASTEWATER EFFLUENTS USED IN DISINFECTION EXPERIMENTS
Wastewater
Palo Alto Secondary - 1978
Palo Alto Secondary - 1979
Chlorine, unfiltered
Chlorine, filtered
Chlorine Dioxide,
unfiltered
Chlorine Dioxide, filtered
Dublin Filtered - nitrified
Laboratory experiments
Field Experiments
San Jose
Secondary
Nitrified
Filtered
SS*
mg/1
17.4 ±

13.8 ±
4.4 ±
13.8 ±
3.3 ±

3.0 ±
3.6 ±

42.5 ±
18.3
3.9 ±
8.6

6.5
0.5
3.7
1.2

3.7
2.6

35.3
5.4
COD* Alkalinity*
mg/1 mg/1 as CaCO,
32.7 ± 6.2

47.6 ± 2.7
37.0 ± 4.2
52.6 ± 2.9
45.9 ± 6.0

33.4 ± 1.2
23.3 ± 6.3

68.2 ± 37.5
41.5 ± 4.6
38.2 ± 6.9
234.5

89.8
90.3
112.8
114.3

115.6
161.2

216.3
146.3
184.7
± 10.1

± 6.3
± 7.0
± 15.8
± 13.6

± 38.8
± 57.6

± 117.5
± 79.7
±84.1
PH
(Range)
6.7

6.8
7.0
6.7
7.0

6.9
6.8

7.0
6.9
7.0
to 7.4

to 7.3
to 7.4
to 7.2
to 7.1

to 7.1
to 7.3

to 7.4
to 7.3
to 7.2
NH
mg/1
30.4

24.1
23.9
29.3
29.0

0.4
0.2

16.3
0.8
0.6
oN Total Coliforms
as N log of #/100 ml
± 2.7

± 1.9
± 1.7
± 7.0
± 7.3

± 0.6
± 0.4

± 7.4
± 0.8
± 0.9
6.21

4.97
4.86
5.38
5.27

4.41
4.36

6.51
4.76
4.81
± 0.48

± 0.21
± 0.35
± 0.27
± 0.62

± 0.47
± 0.39

± 0.23
± 0.39
± 0.98
Mean ± standard deviation.
One value only.

-------
                              CHLORINE
                                 DOSE
                                 mg/l
                                        10
  CHLORINE
   DIOXIDE
    DOSE
     mg/l
2     5    10
                   5 MIN
              o   15 MIN
              <
              Z
              8   30 MIN
    Figure 8.  Experimental dose-time matrix for chlorine-chlorine dioxide
               comparison.
the most  active mixing zone  of  the reactor, and all  glass  or teflon connec-
tions.  The reactor was maintained  at  184 kPa (12 psig) and both disinfectant
injection and sample delivery were  pressure driven.   Stirring was provided by
a 60-mm (2-1/2") magnetic  stir bar  (cylindrical)  with rotational speed of ap-
proximately 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-tracer  study  was  performed.   The  dye-tracer  study in-
volved dosing  three liters  of  deionized water  with  a methylene  blue stock
solution of 500 mg/l.  Sufficient dye  was added to make a final concentration
of 5  mg/l in the  reactor.   These  conditions  mimicked those  in  which actual
disinfection experiments were performed.   Samples were taken at 2, 5, 10, 20,
40, 60, and  120 sec and  analyzed spectrophotometrically  for  methylene blue.
The results of this experiment indicate that the half-time of response is less
than 5 seconds.   This reactor is shown in Fig. 9.

Characterization of the Dublin (Full-Scale) Disinfectant Contact Tank

     Since the  residence  time  varies with the  flow  rate  in  a fixed-volume
reactor,  the contact time  in the disinfectant  contact  tank  can be determined
for any flow if the contact  time is known for a known flow rate.  The contact
times are related by
                                         t2Q2
                           (26)
where t. = contact time at flow
                                      42

-------
                              DISINFECTANT
   DELIVERY
   PRESSURE
   BULB
   SUPPLY
    12 psig

   INDENTATIONS
   FOR  HIGH
   SHEAR MIXING
SAMPLE
  PORT
   CONSTANT-
   TEMP.
   H20  BATH
                                                    STIR BAR
       Figure 9.   A rapid-mix, rapid-sampling 4-liter batch reactor.
                 12 psig = 184 kPa absolute pressure.
    A tracer  study was performed  in the Dublin chlorine  contact tank  at a
constant flow  rate of  0.14 nr/s (3.25 mgd).   Sufficient  rhodamine-WT dye was
injected through the chlorine injection system to yield a concentration of 100
yg/1 based on dilution in the entire volume of the contact tank.  The chlorine
feed was turned off for  15 min before injection and 10 min  after injection to
avoid destruction of the rhodamine-WT tracer by high  concentrations of chlo-
rine.   Figure 10 shows  the physical  layout of the chlorine contact system.
Figure 11  shows  the concentration of  dye as a function of  time  at the three
                                 43

-------
54" PIPE
           PROCESS
            WATER
INJECTION
 POINT
SAMPLE
  POINT A
          kJUNCTION
          (STRUCTURE
           INFLUENT
                              120
                                       C=H

                               MPLE POINTS
                                                 SAMPLE POINT C
                          MEASURING  WEIR
                           SAMPLE POINT D
      Figure 10.  Disinfection contact tank at the Dublin-San Ramon wastewater treatment plant.

-------
         30 i—
K
W
O
O
K
O
         20
10
          0
           0
               ENTRANCE TO

               CONTACT TANK
     MID-POINT OF

     CONTACT TANK
                                               OVERFLOW WEIR
               20
    40            60

TIME—MINUTES
80
                   Figure 11.  Results of tracer study of disinfectant contact tank.

-------
sample points,  inlet  to contact tank, midpoint  of contact tank, and  overflow
weir.  The areas under the three curves (i.e., the mass of dye) are not equal.
In fact,  the  mass  of dye as  calculated  under the first  peak  is  about 80%  of
the mass  injected.   This initial loss of 20%  of  dye  is due to the removal  of
water between  the  injection point and the  first  sample point for use as  pro-
cess water  throughout  the plant.  The continued  decrease of dye mass through
the remainder  of the  contact  tank is believed to be caused by photodecomposi-
tion and  destruction of  the  dye by  chlorine remaining  in  the  contact tank.
Mean residence times  were determined  for the influent  to  the tank, the  mid-
point,  and  the overflow  weir.    They  are  12.3  min,  34.6 min,  and 70.1  min,
respectively.

     In addition to determining  the  average contact time of the contact tank,
the  dye  study  also  provided a  measure  of the  hydraulic efficiency  of  the
contact tank.   Figure  12 shows  the  concentration of dye versus  time for the
overflow weir  sample  point.   The initial appearance of tracer (t^n) indicates
the minimum  contact  time of the  tank.   In  this  case, t.  is approximately  52
min.  An  exact determination cannot be made  due  to loss of dye as previously
mentioned.  The theoretical contact time  for this  tank  at the  constant flow at
which the dye  study was performed is 78.9 min, while the average contact  time
calculated  from Fig.  12  is  70.1 min.   The ratio of average  contact time  to
theoretical contact  time  is a measure of dead space or  inactive space within
the tank.  For  the Dublin-San Ramon contactor  this ratio  equals 0.89.  (In the
absence of any  inactive space the ratio would  equal 1.0.)

     Another measure  of hydraulic efficiency  that  is widely used  in  chemical
engineering is  the reactor dispersion number "d".  The  value of "d" approaches
0 for ideal-plug flow and infinity for complete mix.   (For a  complete discus-
sion of  the  reactor  dispersion  number,  see  Levenspiel (28).)   The reactor
dispersion  number for  this  contact  tank is 0.0045.  A study  performed by the
California  Department of Health in  which  dye  studies were  performed  on  15
chlorine  contact  tanks reports  the  range of  values  for dispersion number  as
0.005-0.36 and  the range of values for the  ratio average  to theoretical deten-
tion times as  0.63-1.05 (29).

     In  summary,  the Dublin-San Ramon disinfectant  contact  tank closely ap-
proximates a plug-flow  reactor with  very little dead space, and as  such is an
example of a well-designed disinfectant contactor.
STATISTICAL ANALYSIS OF EXPERIMENTAL RESULTS

     The  raw  data from the experiments  described above are given  in Appendix
A.  The data  from the  Palo Alto - 1978  experiments were chosen  for a statist-
ical analysis  intended to investigate the  sources  of experimental  error,  the
relationships of  dose  and contact time with bacterial survival,  the variation
between replicate experiments on  the same wastewater  (i.e.,  two  experiments on
one day  performed on  different  aliquots of  the  same wastewater sample),  the
variation  between experiments  on wastewater  from  one source  but   sampled  on
different days, and finally the difference  in  bacterial response as a function
of disinfecting agent, i.e., chlorine  or chlorine dioxide.
                                      46

-------
 TRACER  STUDY OF  DISINFECTANT CONTACT TANK


      3.0
      2.5
K
W
O
O
K
o
      2.0
      1.5
1.0
      0.0
   —  OVERFLOW WEIR
                  50
                      60        70

                    TIME—MINUTES
80
90
                 Figure 12. Dye tracer concentration measured at overflow weir.

-------
  TABLE 15.  MEAN LOG1Q SURVIVING TOTAL COLIFORMS  IN  1978  BATCH EXPERIMENTS
                           WITH PALO ALTO EFFLUENT
r*/"\Ti t'ftf* i"
V^VJLl Ldl. I
Time,
min.
Disinfectant: C102
2 mg/1 Dose 0*
5
15
30
5 mg/1 Dose 0#
5
15
30
10 mg/1 Dose 0#
5
15
30
Disinfectant: C12
2 mg/1 Dose 0#
5
15
30
5 mg/1 Dose 0*
5
15
30
10 mg/1 Dose 0#
5
15
30
Mean
Value" of log^Q
(N/100 ml)
Experiment Number


5.
5.
5.
4.
5.
3.
3.
3.
5.
2.
1.
0.

5.
4.
4.
<4.
5.
3.
3.
2.
5.
2.
1.
0.
1

90
52
13
39
48
73
14
27
74
02
74
44

88
89
48
34
58
92
16
60
96
86
85
71
2

6.12
4.52
4.41
4.33
6.15
<2.0
2.00
1.86
6.25
0.87
0.54
<0.34

6.18
5.52
4.48
3.30
6.23
3.94
2.19
1.69
6.30
2.25
1.16
0.58


6
4
4
3
6
3
1
1
6
1
0
0

6
5
4
3
6
3
2
1
5
2
1
0
3

.27
.42
.31
.81
.26
.04
.70
.27
.18
.71
.32
.34

.09
.82
.86
.53
.23
.88
.80
.80
.96
.31
.62
.74
4

6.73
5.01
4.96
4.87
6.83
3.64
2.49
3.02
6.83
2.52
2.11
0.32

6.58
6.32
5.46
4.47
6.79
4.19
3.48
3.05
6.70
3.77
1.32
1.42
5

6.79
5.02
5.21
5.01
6.76
3.85
2.98
3.27
6.70
2.00
1.58
0.32

6.81
6.14
4.61
3.85
6.83
4.26
3.50
3.21
6.84
2.64
2.04
0.30
6

6.50
5.14
5.20
5.02
6.52
3.24
3.57
3.37
6.73
2.28
1.45
1.13

6.63
6.29
5.58
4.83
6.67
4.53
4.52
2.86
6.67
3.16
1.97
0.92


6
5
5
5
6
3
2
3
6
2
1
0

6
6
5
4
6
4
3
3
6
3
1
1
7

.48
.11
.23
.25
.67
.66
.52
.08
.93
.28
.19
.90

.62
.34
.48
.77
.63
.08
.80
.05
.60
.24
.97
.24
8

5.84
5.12
4.73
5.03
5.55
3.40
2.76
3.45
5.84
1.72
0.95
0.39

5.76
5.46
4.59
3.98
5.91
3.99
3.34
2.63
5.59
2.84
1.70
0.43
 The raw data from which the mean values were calculated are given in Appen-
 dix A,  Tables A-l and A-2.

'Mean value of 8 replicates.
#.
 N(0), prior to disinfection
Statistical Analysis

     Preliminary statistical  analyses  were done on  the  data from Palo Alto  -
1978 shown in Tables A-l and A-2 (Appendix A).  To investigate any bias in  the
data due to the order of experiments or sampling technique,  the data  from time
                                      47

-------
Experimental Results — Palo Alto 1978

     All  samples  for bacterial analysis were  analyzed by the membrane-filter
technique for total  coliform bacteria (1).   Generally, four dilutions  of  each
sample were analyzed in  order  to  obtain an incubated  sample plate with a  sig-
nificant number of colonies  to count, but a sufficiently small number  so  that
the colonies do not  overlap (usually  20-80 colonies are adequate).   It  is  pos-
sible to have from one to four measurements of coliform bacteria  for each  sam-
ple.   The mean values of  the  logarithms of the  numbers  of bacteria for  each
sample (Palo Alto 1978) are given in  Table 15.

Statistical Analysis

     Preliminary statistical analyses were done  on  the  data from Palo  Alto  -
1978 shown in Tables A-l and A-2 (Appendix A).  To investigate any bias 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 variables
used were wastewater, disinfectant,  and disinfectant dose.   In setting  up the
ANOVA in this way, the difference  between coliform density at time  0 with re-
spect to wastewater  could  be  tested  for all experiments.  The results  of  this
ANOVA are shown in Table 16.

     The  only  effect that  is  significant  is  that  of wastewater,  indicating
that the  variance  in bacterial numbers 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  filterable residue,  COD,  alkalinity,  pH, ammonia-nitrogen)
and bacterial counts.

     The next step in the  statistical analysis was to investigate any  differ-
ences between experiments performed on  the same wastewater sample, 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 17.  The  dependent variable
was logiQ[N(t)]. As  seen in  this ANOVA, all variables except experiment had  a
significant effect on  the  number of  surviving  bacteria.   (These effects  will
be  examined  in  subsequent ANOVAs.)   This implies  that  the duplicate  experi-
ments  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  variations seen between experiments in subsequent statist-
ical 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 statistical analyses  shown   in Tables 16 and 17 have been  calculated
using the logarithms of bacterial numbers from each sample (Appendix A)  as the
dependent variable.  In  order  to  compare the bactericidal efficiency of chlo-
rine 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.
                                      48

-------
  TABLE  15.
MEAN LOG1Q SURVIVING TOTAL COLIFORMS IN 1978 BATCH EXPERIMENTS
              WITH PALO ALTO  EFFLUENT
                      Contact-
                       Time,
                       min.
                                       Mean Value1" of Iog10 (N/100 ml)
                                Experiment Number
                               3456
Disinfectant:
2 mg/1 Dose Of
5
15
30
5 mg/1 Dose 0*
5
15
30
10 mg/1 Dose 0#
5
15
30
5
5
5
4
5
3
3
3
5
2
1
0
.90
.52
.13
.39
.48
.73
.14
.27
.74
.02
.74
.44
6.
4.
4.
4.
6.
<2.
2.
1.
6.
0.
0.
<0.
12
52
41
33
15
0
00
86
25
87
54
34
6
4
4
3
6
3
1
1
6
1
0
0
.27
.42
.31
.81
.26
.04
.70
.27
.18
.71
.32
.34
6.73
5.01
4.96
4.87
6.83
3.64
2.49
3.02
6.83
2.52
2.11
0.32
6.79
5.02
5.21
5.01
6.76
3.85
2.98
3.27
6.70
2.00
1.58
0.32
6.50
5.14
5.20
5.02
6.52
3.24
3.57
3.37
6.73
2.28
1.45
1.13
6.48
5.11
5.23
5.25
6.67
3.66
2.52
3.08
6.93
2.28
1.19
0.90
5.84
5.12
4.73
5.03
5.55
3.40
2.76
3.45
5.84
1.72
0.95
0.39
Disinfectant:
2 mg/1 Dose Off
5
15
30
5 mg/1 Dose 0#
5
15
30
10 mg/1 Dose 0#
5
15
30
5
4
4
<4
5
3
3
2
5
2
1
0
.88
.89
.48
.34
.58
.92
.16
.60
.96
.86
.85
.71
6.
5.
4.
3.
6.
3.
2.
1.
6.
2.
1.
0.
18
52
48
30
23
94
19
69
30
25
16
58
6.09
5.82
4.86
3.53
6.23
3.88
2.80
1.80
5.96
2.31
1.62
0.74
6.58
6.32
5.46
4.47
6.79
4.19
3.48
3.05
6.70
3.77
1.32
1.42
6
6
4
3
6
4
3
3
6
2
2
0
.81
.14
.61
.85
.83
.26
.50
.21
.84
.64
.04
.30
6.63
6.29
5.58
4.83
6.67
4.53
4.52
2.86
6.67
3.16
1.97
0.92
6.62
6.34
5.48
4.77
6.63
4.08
3.80
3.05
6.60
3.24
1.97
1.24
5.76
5.46
4.59
3.98
5.91
3.99
3.34
2.63
5.59
2.84
1.70
0.43
 The raw data from which the mean values were  calculated  are  given in Appen-
 dix A, Tables A-l and A-2.

 'Mean value of results from one  to  four dilutions.

 *N(0), prior to disinfection
                                       49

-------
    TABLE  16.   ANOVA FOR BACTERIAL ANALYSIS BEFORE DISINFECTION
                USING LOG1Q[N(0)] AS DEPENDENT VALUE
Source of
Variance
Wastewater (Xi)
Disinfectant (X2)
Dose (X3)
Interactions
\X-t ) x ^Xn )
/v \ v /v "\
\^]_) x \A3/
(X2) x (X3)
(Xt) x (X2) x (X3)
Error
Degrees of
Freedom
4
1
2

4
8
2
8
51
Sum of
Squares
13.2259
0.0002
0.1625

0.4077
0.4897
0.1686
0.2568
2.0911
F-Value
80.64
0.00
1.98

0.25
. 1.49
2.06
0.78

Probability F
Exceeded
0.0001
0.9516
0.1483

0.9092
0.1831
0.1384
0.6197

N(0) = number of bacteria before disinfection.
     TABLE  17.   ANOVA FOR EXAMINING EXPERIMENTAL REPRODUCIBILITY
Source of
Variance
Time (XL)
Level (X2)
Disinfectant (X.,)
Waste (X4)
Experiment (Xe)
it
Interactions
(V x (X2)
(X,) x (X3)
/v \ x fy \
{A.-^} x ^A4J
(Xj_) x (Xc)
(X2) x (X3)
(X2) x (X4)
(X2) x (X5)
(X3) x (X4)
(X3) x (X5)
Error
Degrees of
Freedom
2
2
1
4
3


4
2
8
6
2
8
6
4
3
355
Sum of
Squares
48.247
781.544
7.569
40.366
0.553


9.229
17.296
2.607
2.038
3.102
2.896
1.774
4.430
1.5064
39.428

F-Value
217.20
3518.42
68.15
90.86
1.66


20.77
77.86
2.93
3.06
13.97
3.26
2.66
9.97
4.52

Probability F
Exceeded
0.0001
0.0001
0.0001
0.0001
0.1735


0.0001
0.0001
0.0035
0.0063
0.0001
0.0013
0.0154
0.0001
0.0042

 Interactions of order higher than 2 are included in the error  term.
The survival ratio was defined as:
                     Survival ratio = N(t)/N(0)
(27)
                                 50

-------
where N(t) = number of bacteria measured at time t,

      N(0) = number of bacteria measured at time 0.

The logarithm of survival ratio is then


                   log[N(t)/N(0)] = log[N(t)] - log[N(0)]  .                (28)


Log[N(t)] was calculated as  the  arithmetric mean of the logarithms of  bacter-
ial numbers  from each sample  (see  Table  15) for each  time  t.   The  log[N(0)]
was similarly  calculated at  time 0.   This calculation  yields the  geometric
mean of the replicate analyses for bacterial numbers.

     Experiment, disinfectant, disinfectant dose,  and time were used as  inde-
pendent  variables.   The  resulting ANOVA  is shown  in  Table 18.   This  ANOVA
reveals that all of the main effects—disinfectant, disinfectant dose,  contact
time, and  experiment—were  highly significant, as  evidenced by the  fact  that
P(F > Fcalc) = 0.0001  in all cases.

     With respect to the comparison between chlorine  dioxide and chlorine,  the
ANOVA (Table  18)  implied that there was  a significant difference between  the
two disinfectants.   The implications of  this  significant difference are  dis-
cussed below.  The significant effects of  contact time  and dose were  expected:
The survival ratio declined with  increasing disinfectant  dose and  contact time
for both disinfectants.  The significant difference between  experiments is  be-
lieved to be attributable to variations in properties of  the wastewater.

     Several  of  the  second-order interactions  were found  to  be significant
(Table 18),  notably those between  the  disinfectant  variable on the one  hand
and the disinfectant dose, the contact time, and experiment  on  the other  hand.
Clearly, the difference between disinfectants with  respect to coliform  inacti-
vation was not the  same  at  all doses,  at  all contact times,  and in all waste-
water samples.   Also, the  interaction  between disinfectant dose and  contact
time was found to be  significant, indicating that the  effect of dose on  coli-
form inactivation by  a given disinfectant depended significantly on the  con-
tact time  and  vice  versa.   None  of the  third-order interactions was found to
be significant.

     The meaning  of the  significant  effects can be  interpreted by  comparing
the means of  log[survival ratio]  for chlorine dioxide with  those  for chlorine
for various  combinations of dose and contact  time  (Table 19).   The  values in
Table 19 represent  means of the  eight experiments.  The mean  values in  Table
19  show  the observed  differences between chlorine  and  chlorine  dioxide  and
between dose levels and  contact  times  for a given disinfectant, corresponding
to the ANOVA main effects  in Table 18.   From Table  19, the  difference  between
disinfectant  doses  is obvious,  as is the difference  between  contact  times.
The significant difference between disinfectants seen in  the ANOVA of Table 18
is modified  by  the  ANOVA interactions.   The relationship  of  the difference
between  disinfectants  to the  values of  dose  and contact time can  be better
understood with the help of  Figs. 13, 14,  and 15.

                                      51

-------
       TABLE  18.   ANOVA FOR COMPARING CHLORINE AND CHLORINE DIOXIDE AS
                      BACTERICIDES IN PALO ALTO EFFLUENT
Source of
Variance
Disinfectant (X^)
Dose (X2)
Experiment (X3)
Contact Time (X4)
it
Interactions
(X^ x (X2)
(Xj^) x (X3)
(X1) x (X4)
(X2) x (X3)
(X2) x (X4)
(X3) x (X4)
(X,) x (X2) x (X3)
(X,) x (X3) x (X.)
(X2) x (X3) x (X4)
Error
Degrees of
Freedom
1
2
7
2


' 2
7
2
14
4
14
14
14
28
32
Sum of
Squares
4.858
279.297
21.869
37.341


1.339
2.109
6.423
2.463
3.403
1.253
0.904
1.788
3.134
4.052

F-Value
38.36
1102.76
24.67
147.43


5.29
2.38
25.36
1.39
6.72
0.71
0.51
1.01
0.88

Probability F
Exceeded
0.0001
0.0001
0.0001
0.0001


0.0104
0.0445
0.0001
0.2146
0.0005
0.7515
0.9096
0.4691
0.6277

      Fourth-order interaction neglected.
    TABLE 19.  MEAN LOG[SURVIVAL RATIO] OF TOTAL COLIFORMS:  COMPARISON  OF
       DISINFECTANTS AT DIFFERENT COMBINATIONS OF DOSE AND CONTACT TIME

                           Mean* of  log[Survival  Ratio]  = log[N(t)/N(0)]
                             Chlorine Dose
                                 mg/1
Chlorine Dioxide Dose
        mg/1
^UUCctUL
Time
5 minutes
15 minutes
30 minutes
2
-0.49
-1.38
-2.18
5
-2.23
-2.98
-3.72
10
-3.44
-4.62
-5.53
2
-1.34
-1.43
-1.61
5
-3.01
-3.69
-3.50
10
-4.40
-5.09
-5.81
      Calculated means of eight experiment replicates.
     Chlorine  and  chlorine  dioxide are  compared by  dose at  30-min  contact
times in Fig.  13.   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/l doses, while
chlorine dioxide  gives a lower  log[survival ratio]  at 10-mg/l dose.  This
illustrates the disinfectant dose interaction mentioned above.   The reason for
                                      52

-------
                       (.49)
              o
              DC

              CO
              o   -
•     CHLORINE
	 DIOXIDE
|^ CHLORINE
 (  ) = RESIDUALS
                            1.18)
                                         .78)
    Figure 13.
    0


   -1


   -2


   -3


   -4


    5
i


  -6



    7      2           5           10

            DISINFECTANT DOSE (mg/l)

 Comparison of  chlorine and  chlorine dioxide  bactericidal
 effectiveness  at  30-min contact  time.
                                                  (8.08)
this  dose-dependent  disinfectant  difference  can  be  found  in the  die-away
curves of chlorine and chlorine dioxide.   Figure 14 shows typical chlorine and
chlorine dioxide die-away curves  in secondary effluent.  The  initial oxidant
demand reduces the disinfectant concentration  significantly.   Hence,  the rate
of disinfection will be slowed.

     When disinfectants are  compared with respect to contact times,  the disin-
fectant-contact time  interaction  can be evaluated.   In Fig.  15  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 chlo-
rine dioxide.  Chlorine dioxide effects  a major portion of  bacteria inactiva-
tion at shorter times, while chlorine requires longer contact times  to achieve
the same degree of disinfection.

Summary

     The conclusions  concerning the comparative bactericidal  action  of chlo-
rine and chlorine dioxide derived from the statistical analysis of the data on
Palo Alto effluent are as follows:
                                      53

-------
                  14
               •=•  8
               <
               D
               ==  6
               V)
               LLJ
               cc
                                 I
I
                                           CHLORINE
                                           DIOXIDE
I
                                10    15    20

                               TIME IN MINUTES
           25
           30
                      INITIAL DOSE OF 12.5 mg/l FOR BOTH
                       CHLORINE AND CHLORINE DIOXIDE
  Figure 14.  Chlorine and chlorine dioxide die-away in Palo Alto secondary
              effluent.
     1.  Both chlorine  and chlorine  dioxide  give decreased  survival
         ratios  when dose or contact time is increased.
     2.  Although some  variations  exist, chlorine and  chlorine  diox-
         ide give essentially  the  same survival ratios when  compared
         on a mass dose basis at 30-minute contact time.
     3.  Chlorine dioxide is a more rapid  disinfecting agent,  effect-
         ing  greater  bacterial  inactivation  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.
THE EFFECT OF PROCESS SEQUENCE ON BACTERIAL NUMBERS

     The experiments with  San Jose effluents provided data  for  evaluation of
process sequence effects on disinfection kinetics.   Additionally,  analysis of
the bacterial concentrations at each sample point permitted evaluation of pro-
cess sequence effects in removal of bacteria before disinfection. Intuitively,
                                      54

-------
                  -3
              o
              I-
              cc
              cc
              13
                  -4
              O  -5
                  -6
                           (8.64)
  CHLORINE
  DIOXIDE
  CHLORINE
= RESIDUALS
                      (6.
                           5           15          30

                           CONTACT TIME (MINUTES)
Figure 15.  Comparison of chlorine and chlorine dioxide bactericidal effec-
            tiveness at 10-mg/l dose.   Values of residual concentration are
            given in mg/1.
decreases in  bacterial numbers would  be expected  to  result  from  additional
treatment, due to  physical  processes  that remove or select  for  particulates.
For example,  the clarification  step of  activated-sludge  treatment selects for
those microorganisms  that  aggregate  into particulates with  a  density greater
than that of  water.   Therefore, any bacteria  (including  total coliform) that
exhibit  aggregation  or  "clumping"  would  be  expected  to be  removed  in the
process.

     The  logarithms  of the  initial  bacteria  counts  from each of  the sample
points  at San Jose were compared.   Summary statistics are  shown in  Table 20
and an ANOVA is shown in Table 21.
                                      55

-------
    TABLE 20.  INITIAL TOTAL COLIFORMS IN UNDISINFECTED EFFLUENTS  FROM THE
                      SAN JOSE  WASTEWATER TREATMENT PLANT
Sample Point

Statistic
Mean, log10[N(0)]
Variance
Standard Deviation
Number of Samples
Secondary
Effluent
6.51
0.0548
0.2342
18
Nitrified
Effluent
4.76
0.1501
0.3875
18
Filtered
Effluent
4.81
0.9516
0.9755
18
             TABLE  21.   ANOVA OF SAN JOSE INITIAL BACTERIA NUMBERS
Source
Treatment Level
Error
Degrees of
Freedom
2
51
Sum of
Squares
35.5562
20.8183
Mean
Square
17.7781
0.4082
F-Ratio
43.35
       Total
53
56.3745
     As shown by  the  F-value  [P(F > 43.35) < 0.001] of  the ANOVA in  Table  21,
there is a highly significant difference in initial  bacteria  numbers  among  the
treatment levels.   To  identify  statistically which means  are different from
others, critical  sums of  squares  were computed  and comparisons were made  (a_
posteriori test) on the means of Table 20.  These tests  showed a  statistically
significant difference between bacterial numbers  in  the  secondary effluent  and
nitrified or  filtered  effluent.   No  difference  was  found  between  nitrified
effluent and filtered effluent.  This result was  surprising.  While removal  of
bacteria through  the  nitrification process might  be due to the  final clarifi-
cation step or  predation,  it  was expected that greater  reductions would  occur
in filtration  than were observed.   If the  physical processes of particulate
removal are analyzed  and filter efficiency  is  computed as a function of par-
ticulate size (30), a minimum efficiency  of removal is  found for particulates
on the  order  of  1  ym.   An  individual coliform  bacteria  cell  is in  the size
range of 1-5 urn (31).  It can be inferred  from this  analysis  that the bacteria
in the nitrified effluent exist as singlets or small aggregates.

     A  similar  analysis  for  the  Palo Alto  - 1979  experiments  is  shown  in
Tables 22 and 23.  As shown, there is no significant difference  [P(F  > 0.6588)
= 0.429] between  the  initial bacteria numbers  of secondary effluent  and lab-
filtered secondary effluent.  It can be inferred  from  these data  that the bac-
teria in secondary effluent are also mono-dispersed  or nearly so.

     At most  wastewater treatment  plants (all  in   this study),  it   is common
practice to  use chlorinated  finished  water  as  process water  throughout  the
treatment process.  For example, both the  Dublin  and San Jose plants  use water
from the chlorine contact tank (8-10 mg/1  total chlorine residual) to backwash
                                      56

-------
             TABLE  22.   LOGS  OF  INITIAL BACTERIAL NUMBERS AT THE
                     PALO ALTO WASTEWATER TREATMENT PLANT
      Statistic
Secondary Effluent,
    Unfiltered
Secondary Effluent,
   Lab-Filtered
Mean
Variance
Standard Deviation
Number of Samples
       4.97
      0.1199
      0.3462
         9
        4.86
       0.0423
       0.2056
         9
            TABLE  23.   ANOVA OF  PALO ALTO INITIAL BACTERIAL NUMBERS
Source
Treatment Level
Error
Total
Degrees of
Freedom
1
16
17
Sum of
Squares
0.0601
1.4593
1.5194
Mean
Square
0.0601
0.0912
F-Ratio
0.6588

the filters.  This  backwash  water is settled and blended with  the  influent  to
the primary  sedimentation basin.   Dublin also  uses a small  fraction of the
chlorinated  water  (12-14  mg/1  total  chlorine  residual)  from  the  chlorine
rapid-mix tank for foam control in the activated-sludge process.  The  dose and
form  of  chlorine that  is  applied to  the wastewater  in  processes  other  than
disinfection is difficult, if not impossible, to determine.  Additionally, the
possible  effects  of  low  doses of chlorine  species on the  morphology of the
coliform  bacteria group and  the  potential consequences for disinfection kine-
tics  are largely matters  of  conjecture.   One  area of  concern in comparing
chlorine  and chlorine  dioxide as  disinfectants in real wastewaters is the
question  of  induced  resistance.    If a  bacterial   population  has  an  induced
resistance to  chlorination due to  low-level exposure  (such  as cited  above),
the bacteria may  appear relatively  more sensitive to chlorine  dioxide than  to
chlorine, a bias which may disappear after long-term use  of chlorine dioxide.
COMPARISONS AMONG WASTEWATERS

     The  data  (Appendix A) from  the  experiments  described above were  used to
determine the  coefficients  of  the model used in this study.   The coefficients
were determined by linear regression  of  log(residual-time  product)  on log sur-
vival ratio.   This regression yields:
                     log[N(t)/N(0)] =  log  bf + k  log(RT)

                                    =  k  log[RT'(b')1/k]
                                                (29)

                                               (29')
or
                                       57

-------
                           N(t)/N(0)  = [RT'(b')1/k]k                       (30)


where

                                   V = b~k


so that


                               N(t)/N(0)  = [^|]k                           (31)


     An analysis  of the  standardized residuals from  the  regressions of  each
experiment  showed  in all  cases  except chlorine dioxide  disinfection of  Palo
Alto wastewater - 1979, the simple regression model


                               y = 3Q + exx + e                            (32)


was  adequate  in  describing  the  relationship  between  survival  ratio  and
residual-time  product  (i.e.,  the  standardized  regression residuals  were  nor-
mally distributed).  For  the chlorine dioxide disinfection of  Palo Alto  waste-
water  - 1979, the  doses  and  sample times  provided  low  enough  residual-time
products, so that bacterial samples were  taken  before  any  inactivation had oc-
curred.   To remove these  points  from the linear regression,  a 95% confidence
interval  around   N(t)/N(0)  = 1.0    was  estimated  using  the  variance  of  the
membrane-filter  technique, as  calculated  from replicates  of zero-time  bac-
terial  numbers.   Any survival  ratios which fell  into this  interval  were  ex-
cluded  from  the  regression,  and  analysis  of  the  resulting   standardized
regression  residuals showed conformity to the regression model.

     The  coefficients  of  the disinfection model determined  by linear regres-
sion are  shown in Table  24.   Log-log plots of the model  predictions  and  data
points  for  each experiment are shown  in Appendix B.

     The  San Jose  treatment  plant  had   brought  their separate  nitrification
process and  the filtration process on line only six months prior  to our  labor-
atory disinfection  studies.  Complete nitrification of the wastestream was not
always  achieved  during  the sample period,  so  that low levels of  ammonia  were
measured  in both  the nitrified effluent  and the filtered  effluent.   The model
coefficients reported in  Table 24 for chlorine  disinfection  of San Jose  nitri-
fied and  filtered  effluents  are from linear   regressions with  all  data  from
these sample points.  If  the experiments  are segregated according to the chlo-
rine: ammonia ratios, then the experiments  in which  free chlorine  is  the  pre-
dominant  disinfecting agent  can be  separated  from those  experiments  in which
disinfection is due to  combined  chlorine.  Extensive  research has been  repor-
ted  in  the literature  on the reactions  of  chlorine  with ammonia,  the  break-
point phenomenon (11).  For complete  oxidation  of ammonia  to nitrogen gas,  the
following stoichiometry follows:


                                      58

-------
           TABLE  24.   VALUES OF FITTING CONSTANTS FOR CHLORINE AND CHLORINE DIOXIDE IN THE DISINFECTION MODEL
VO
Chlorine Dioxide
Experimental
Description
Palo Alto - 1978
Palo Alto - 1979
Unfiltered
Palo Alto - 1979
Lab Filtered
95% CI*
b for b k
1.56 0
0.67 0
0.004 0.
Palo Alto Unfil- 1.29 0
tered 78 & 79 Comb.
Dublin-San Ramon
Lab
Dublin-San Ramon
Field
San Jose
Secondary
San Jose
Nitrified
San Jose
Filtered
CI = confidence
^For definitions
r = correlation
**
0.14 0
0.57 0
0.89 0
0.78 0
0.003 0.
interval.
of b and
coefficient
.99 to
2.47
.32 to
1.41
.002 to
0.010
.88 to
1.91
.07 to
0.28
.37 to
0.87
.33 to
2.43
.18 to
3.30
,001 to
0.01
k , see
-2.90
-2.30
-1.21
-2.75
-1.84
-2.20
-2.18
-2.06
-1.13
Eqs .
95% CI t1i
for k r' n b
-2.49 to 0.86 72 5.17
-3.30
-1.78 to 0.86 29 2.21
-2.83
-1.03 to 0.89 44 0.84
-1.40
-2.42 to 0.86 101 3.95
-3.07
-1.15 to 0.78 22 0.65
-2.51
-1.86 to 0.84 69 1.67
-2.54
-1.41 to 0.78 21 4.06
-2.95
-1.23 to 0.75 19 2.70
-2.88
-0.74 to 0.81 18 0.59
-1.51
24 and 25.
95% CI
for b
2.72 to
7.62
0.61 to
5.82
0.38 to
1.83
2.62 to
5.95
0.37 to
0.86
0.56 to
4.93
1.50 to
10.98
0.83 to
8.81
0.13 to
2.75

Chlorine
95% CI
k for k r
-3.15 -2.82 to 0.92
-3.47
-2.22 -1.67 to 0.83
-2.76
-2.10 -1.65 to 0.86
-2.54
-2.79 -2.50 to 0.88
-3.07
-1.62 -1.31 to 0.91
-1.92
-1.79 -1.32 to 0.93
-2.27
-2.82 -2.14 to 0.86
-3.51
-2.58 -1.67 to 0.79
-3.48
-2.12 -1.16 to 0.69
-3.07


**
n
72
31
32
103
26
12
26
24
21

for log-log transformation of raw data.
          n = number of observations.

-------
                       2NH3 + 3HOC1 * N2 + 3H30+ + 3C1                     (33)


The theoretical weight  ratio of C1:N is  7.5  for completion of this  reaction.
However, the  stoichiometry is a  simplification of a  complex series of reac-
tions, all of which show some pH dependency (11,32).  Weight ratios of C1:N up
to 10:1  are  required in  some  situations to  achieve  breakpoint,  depending on
the pH  of  the reaction and  the aqueous  system involved.   For our purposes, a
C1:N ratio of less than 6.25 was chosen as indicative of combined chlorine re-
sidual, while a C1:N ratio of 12.0 was interpreted as indicating predominately
free chlorine.

     The chlorine  to ammonia nitrogen  ratios for the  San Jose nitrified and
filtered experiments are  shown  in Table 25.   Based on these data, the follow-
ing decisions were made  to classify the  San Jose effluent  samples into the
"free  residual"  or  combined  residual  categories.    In  nitrified   effluent,
Experiment  1 at  all chlorine  doses  and Experiment  2 at  2 mg/1  dose  were
considered to have a combined chlorine residual;  the  rest with the  exception
of Experiment 2  at 5 mg/1 dose were considered  to have  free chlorine  resid-
uals.    Experiment  2  at  5  mg/1  dose  was excluded from the analysis altogether
because  of the uncertainty of residual composition near  breakpoint.  In  fil-
tered effluent, all  Experiment 1  doses were  considered  as  below breakpoint,
the remaining experiments  were  used  as  beyond breakpoint.  Linear regressions
of these subgroups yielded model coefficients which  are  compared in Table 26
with the model coefficients of all experiments within the wastewater  of  inter-
est.  These  relationships  are  shown  graphically in Figs.  16 and 17 for  nitri-
fied and filtered San Jose wastewater, respectively.

     The nitrified and  filtered experiments  in which  the C1:N ratios indicate
disinfection with combined chlorine are compared with disinfection by chlorine
in secondary effluent (C1:N ratios « 6.0, i.e. combined chlorine) in Fig. 18.
The three  curves  are essentially identical.   To  simplify further discussion,
"secondary effluent"  will be  used in  referring  to disinfection  by combined
chlorine whether in partially nitrified or in  conventional secondary  effluent,
and will be represented by the curve from the  secondary effluent  experiments.
    TABLE  25.   CHLORINE TO AMMONIA NITROGEN RATIOS FOR SAN JOSE EXPERIMENTS
Chlorine : Ammonia-N
Experiment
Nitrified Effluent
1
2
3
Filtered Effluent
1
2
3
NH3-N
mg/1
1.7
0.7
< 0.04
1.6
0.12
< 0.04
2 mg/1
1.18
2.86
50
1.25
16.67
50
Dose
5 mg/1
2.94
7.14
125
3.13
41.67
125
10 mg/1
5.88
14.29
250
6.25
83.33
250
                                      60

-------
    TABLE 26.  COMPARISON OF MODEL COEFFICIENTS,  SEGREGATING  DISINFECTION
                       WITH FREE VS COMBINED CHLORINE
Experiment
San Jose Nitrified
All data
Combined residual data
Free residual data
San Jose Filtered
All data
Combined residual data
Free residual data
b*

2.70
6.74
1.33

0.59
5.24
0.44
k*

-2.58
-3.17
-2.40

-2.12
-3.25
-2.16
rt

0.79
0.91
0.87

0.69
0.80
0.77
n

22
11
11

21
6
15
 For definitions  of  b   and  k  , see Eq. 24.

'r = correlation  coefficient.
"Nitrified effluent"  and "filtered effluent"  will be understood to imply dis-
infection by free  chlorine and will be represented by the free chlorine curves
in €Figs.  16 and 17.
            10"
            10
              -2
       o
       55
       O
            10
              -3
10
  -4
            10
              -5
       g
       co
10
              -6
                       ALL DATA
FREE
CHLORINE
RESIDUAL
COMBINED
CHLORINE
RESIDUAL
                                        10J
                                        10
                                                       2
                         RESIDUAL—TIME   IN MG-MIN/L
  Figure 16.  Coliform interaction by free  and  combined  chlorine  in San Jose
              nitrified effluent.
                                      61

-------
          10"1 b-
          10'
          10'
s    io"
          10"
      g   10"
                             ALL DATA
                                           \
                                              \
                                               \
FREE
CHLORINE
                                                 \
                                                      COMBINED
                                                  \    CHLORINE
                                    RESIDUAL    \\\  RESIDUAL
                                i  i i i 11
                                                    i  i  i i 111
                         10°         IO1         IO2
                      RESIDUAL—TIME  IN MG-MIN/L
Figure 17.  Coliform inactivation by free and combined  chlorine in San Jose
           filtered effluent.
      10-1

      10~2

      10~3
      10'
     10
       -6
                        SECONDARY \\\
                        EFFLUENT
                                                NITRIFIED EFFLUENT
                                                (COMBINED CHLORINE)
                                FILTERED
                                EFFLUENT             \
                                (COMBINED CHLORINE)
                                                       i   i i i  1111
                       10
                         0
                                     101
             10*
Figure 18.
                   RESIDUAL — TIA1E   IN MG-MIN/L

          Coliform inactivation by combined chlorine in San  Jose wastewater
          at three sampling points.
                                 62

-------
EFFECTS OF TREATMENT LEVEL ON THE DISINFECTION PROCESS

Disinfection with Chlorine

     The San  Jose  experiments  provide comparisons of disinfection efficiency
in wastewater effluents  of  progressively  increasing  quality.   As discussed
above, a significant difference  was  found in the initial bacterial concentra-
tion between  secondary and nitrified  wastewater at San  Jose.   This finding,
coupled with  the observation  that  combined chlorine disinfection in partially
nitrified  wastewater  and in  partially  nitrified,  filtered  wastewater was
nearly identical  to disinfection in  secondary  effluent,  suggests that disin-
fection kinetics are not  affected by  variations  in initial bacterial concen-
tration of the magnitude encountered in this study.

     Figure 19  compares  the  disinfection process by  chlorine at three treat-
ment  levels  in San  Jose   wastewater, after adjustment  of  the  data base  by
deleting combined  chlorine residual data points  in the "nitrified" and  "fil-
tered" categories.  Hence  the "nitrified" and "filtered"  categories correspond
to a free chlorine residual,  whereas the  "secondary" category corresponds to a
combined chlorine residual.  Superior  disinfection  efficiency is  observed when
the wastewater has been nitrified compared  to conventional  secondary effluent.
This  increased  efficiency is  believed to  be  due  to  the superior germicidal
properties of free chlorine in nitrified  effluent (11).   A  further increase  in
disinfection  efficiency  is found in comparing  the  nitrified and filtered ef-
fluents.  A decrease in the "non-target organism" chlorine  demand (that demand
not  associated  with bacterial  inactivation) might account for the  increased
efficiency in filtered effluent  vs  nitrified effluent.   The one-hour chlorine
demands, however,  were found to be  equivalent  or slightly higher in the fil-
tered effluent.   The one-hour demands  for  chlorine in filtered  and nitrified
effluents are shown in Table 27.

     The Palo Alto - 1979 experiments also indicate an  increase in  disinfec-
tion efficiency due to filtration,  shown in Fig. 20.   Hence,  the increase  in
disinfection  efficiency of chlorine  achieved by  prior  filtration is not depen-
dent on  the  form  of chlorine responsible  for  the  bacterial inactivation and
cannot be  explained by the removal  of bacteria associated with  solids,  since
no difference was  found  in  initial  bacterial  concentration between  nitrified
and  filtered  effluents in San Jose wastewater,  or  between  Palo Alto  secondary
and Palo Alto filtered secondary.  Moreover, the  increased  efficiency is  not a
result of  decreased chlorine demand,  since comparison of one-hour demands  in
both San Jose and  Palo Alto  wastewaters  (see Table 27)  indicate that the fil-
tered  wastewater   demand  is  equal  to—if  not  greater than—the non-filtered
demand.
      Disinfection  efficiency  is  defined  in  relation  to  the  residual-time
 product  required  to  achieve  a  given  level  of  bacterial  inactivation  as
 measured  by  the  survival  ratio   N(t)/N(0)  .   An  increase  in  disinfection
 efficiency  corresponds  to a  decrease  in  residual-time  product required  to
 achieve the  same  survival  ratio.

                                       63

-------
      10
      10
         -1
         -2
      10"
o     10
         -4
      10
         -5
      10
         -6
CO
                     NITRIFIED  \
                                         SECONDARY
                                                              j	I
                           0
                         10
                     RESIDUAL—TIME   IN MG-MIN/L

Figure 19.   Coliform inactivation by chlorine in San  Jose wastewater at
            various levels of treatment after segregating data base
            according to the form of chlorine residual.

                  TABLE 27.   ONE-HOUR  CHLORINE DEMAND
Demand, mg/1

Wastewater and Treatment Level
San Jose, Nitrified
San Jose, Filtered
Palo Alto - 79, Unfiltered Secondary
Palo Alto - 79, Filtered Secondary

2 mg/1
1.71
1.73
1.01
1.38
Doses
5 mg/1
4.36
4.28
1.72
2.32

10 mg/1
6.52
6.92
3.12
4.05
   The one-hour  demands  are calculated in the manner of Feben and  Tarus
  (33) from residual measurements made over  the  course  of  the  experi-
  ments.   This method involves calculation of the demand at  time  t  as
  dose minus  residual. A regression analysis of log(chlorine demand)  on
  log(time) yields an equation for demand  of  the form:

                              ~  one-hour  *-
  where  D = chlorine (or chlorine dioxide) demand;  Done-hour =  one~
                                          a  = a coefficient.
hour demand,  mg/1;  t = time, hours; and
                                  64

-------
         10-1

         10"2

         10~3

         10-4
         10
           -6
                    \

                        \UNFILTERED
                          \
              FILTERED \  x
                                \
                                  \
                                    \
                                      \
                                                                \
                       i i
                                  i 111 in
                                            I	I
10"1       10°
RESIDUAL — TIME
                                                101        102
                                             IN MG-MIN/L
   Figure 20.   Coliform  inactivation by chlorine in filtered and unfiltered
               secondary effluent, Palo Alto.
Disinfection with Chlorine Dioxide

     The results of the San Jose  experiments  with  chlorine  dioxide as disin-
fectant are  shown  in  Fig.  21.  In  contrast  to  the corresponding experiments
with chlorine as disinfectant, no  increase in disinfection efficiency is seen
when wastewater  quality  changes  from that  of secondary effluent  to  that  of
nitrified effluent.   This  is  expected,  because the  form  of the disinfecting
agent,  chlorine  dioxide, is  not  affected by  NH3~N  concentration (34).   This
observation  further supports  the  conclusion  that variations  in initial bac-
terial  concentration of the order of  magnitude  observed  in  this study (Table
20) do not affect the  disinfection process, in view of the fact that the coli-
form  concentrations in  the  undisinfected nitrified effluent  were  nearly  one
hundred times lower than in  non-nitrified  effluent.

     The  similarity of  disinfection  in  nitrified  and secondary effluent  by
chlorine dioxide (Fig. 21)  is  in good  agreement  with the results  obtained with
chlorine when the  experiments with  chlorine  in  nitrified  effluent are segre-
gated by Cl:NHo-N  ratios.  When the species of  chlorine is  unchanged, no dif-
ference  in  disinfection efficiency  is  seen  between  secondary  and nitrified
effluents (Fig.  18).
                                      65

-------
        10
           	1
         10
           -2
         10
           -3
         10
           -4
         10
           -5
         10
           -6
                                    SECONDARY
                    FILTERED
                                                             NITRIFIED
                            10
                              0
                              10-
10
                                               2
Figure 21.
            RESIDUAL—TIME   IN MG-MIN/L

Coliform inactivation by  chlorine dioxide in San Jose wastewater
effluents.
     The San Jose  experiments  with  chlorine dioxide as disinfectant show  an
increase in disinfection  efficiency  in filtered wastewater as compared with
either  secondary  or  nitrified  effluent,  similar  to  that  seen for  chlorine
disinfection but of greater  magnitude.  This increase in efficiency  is  again
seen in the Palo Alto  -  79 experiments,  shown  in  Fig.  22,  between unfiltered
and filtered secondary effluent,  similar  to that  previously shown  (Fig.  20)
for chlorine disinfection in Palo Alto wastewater.   However,  the  increase  in
efficiency for chlorine dioxide  is  again of greater magnitude than  that  for
chlorine.

     The one-hour  chlorine  dioxide  demands are shown  in  Table 28.  Judging
from the data  in Table  28,  the  increase in disinfection efficiency  using  chlo-
rine dioxide seen between  unfiltered and filtered wastewater cannot be attrib-
uted to a decrease  in demand.
                                     66

-------
 O
 i—i
 E-i
       10
         -3
o     10
         -4
       10
         —5
       10
         -6
                                          \
                                          i i i 11 nl   i  i  i i r i nl    i  i i i i i ii
                      10
                        -1
                                 10
                                   0
 Figure 22.
                      RESIDUAL — TIME   IN MG-MIN/L

             Coliform inactivation by chlorine dioxide in filtered and
             unfiltered secondary effluent,  Palo  Alto.
                TABLE 28.  ONE-HOUR CHLORINE DIOXIDE DEMAND
Demand, mg/1

Wastewater and Treatment Level
San Jose, Nitrified
San Jose, Filtered
Palo Alto - 79, Unfiltered Secondary
Palo Alto - 79, Filtered Secondary

2 mg/1
1.45
1.80
1.99
1.94
Doses
5 mg/1
3.56
3.85
4.46
4.57

10 mg/1
4.86
4.19
7.46
6.79
Demand in mg/1 calculated as described in Table 27,
                                   67

-------
COMPARISON OF DISINFECTION WITH CHLORINE AND CHLORINE DIOXIDE

Secondary Effluent

     Chlorine and  chlorine  dioxide  disinfection are  compared in  Palo Alto
secondary effluent (1978 experiments) in Fig.  23.  Chlorine dioxide is a more
efficient bactericide than  chlorine  when  compared on  the  basis of residual-
time product required for  a given survival  ratio.   This is true for the full
range of Rt products (or equivalently survival  ratios)  of the experiments.

     Because of  the nature of the log-log representation, however, it is dif-
ficult to visualize how the  chlorine dioxide advantage varies with changes in
survival ratio.   For illustrative purposes, a ratio of  chlorine  dioxide resid-
ual to chlorine  residual  (calculated assuming  one-hour contact time)  will be
presented at two  levels  of  inactivation,  survival  ratio of 10~^ and survival
ratio of 10~->.  A value  of  required  residual ratio (RR)  less than unity indi-
cates superiority for chlorine dioxide compared to chlorine.  For example, in
the case  of Palo Alto  secondary effluent (Fig.  23)  at a  survival ratio of
10  ,  the residuals  required are 0.06 and 0.18 mg/1 for chlorine dioxide and
chlorine, respectively,  yielding a  C102: Clo  residual ratio of  0.33.   At a
survival ratio of 10" , the  residuals are 1.38 and 3.33, respectively, yield-
ing a residual ratio of 0.41.   The implication is  that at  lower residual-time
         10
            -1
          1CT
   O
         10
            	O
         10
            -4
         10
            -5
         10
            -6
   CO
                                    \
        \
         \
           \
CL02 AS     \
DISINFECTANT \
CL2 AS
DISINFECTANT
                 \
                                                  \
                                                   \
                       \
                                                       \
                                                         \
                             10
                               0
                           10'
                         RESIDUAL—TIME   IN MG-MIN/L
  Figure 23.  Coliform inactivation by chlorine dioxide  in  1978  laboratory
              experiments with unfiltered secondary effluent from  Palo Alto,
                                      68

-------
products, chlorine dioxide has a greater advantage in disinfection efficiency
over chlorine than at higher residual-time products.   This convention will be
used for  the comparison of  chlorine  dioxide and  chlorine  disinfection pre-
sented below.  The residual  ratios at  survival ratios of lO"-*- and 10"^ will be
referred to as RR1 and RR2,  respectively.
     Disinfection in  Palo
ments) is compared in Fig.
Alto unfiltered
24, with  RR  =
                                            secondary
                                            0.29   and
                                                   effluent  (1979  experi-
RR
                                                          0.25
                                                                     Chlorine
dioxide is  again shown to  be  a more  efficient  bactericlde over  the  entire
range of Rt products  studied and  the advantage remains relatively constant.

     Figure 25 shows  comparison of chlorine and chlorine dioxide disinfection
of filtered Palo Alto secondary effluent (1979 experiments).  The RR, and RR2
in this case  are 0.01  and  0.26,  respectively.   The large advantage  realized
for chlorine  dioxide due to filtration  is  apparent in this comparison.  This
advantage,  however, decreases with increasing Rt product to the advantage seen
in unfiltered Palo  Alto secondary effluent.

     The results of disinfection experiments with chlorine and chlorine diox-
ide in San Jose secondary effluent are .compared in  Fig. 26.  In this case, RR,
and RR2 are  0.28 and  0.73,  respectively,  which indicates a decrease in chlo-
rine dioxide advantage with increasing Rt product.
O
•—i
E-!
   CO
      10
            -1  L-
         10
            -3
   o     10
            -4
          10
            -5
          10
            -6
                                                      CHLORINE
                                             \
                                               \
                                                 \
                                                \
                                    CHLORINE    \
                                    DIOXIDE        \
                                                      \
                                                           \
                                                             \
                                                           \
                                I  I I I IHI
 Figure  24.
                      10"1       10°        101        102
                      RESIDUAL—TIME   IN MG-MIN/L

         Coliform inactivation by chlorine and chlorine dioxide  in 1979
         experiments with unfiltered secondary effluent from Palo Alto.
                                      69

-------
  o
  I—I
  E-«
  CO
                                                 CHLORINE
                                                          10'
                       RESIDUAL—TIME    IN MG-MIN/L
Figure 25.
Coliform inactivation by chlorine and  chlorine dioxide in 1979

experiments with filtered secondary effluent from Palo Alto.
10~2



10~3



10-4
               10
                 -5
               10
                 -6
                            CHLORINE

                            DIOXIDE
                                           \
                                            \
                                              \
                                               \
                                     \
                                      CHLORINE
                                                  \
                                                    \
                                                      \
                                                       \
                        -i—\  i i 11 nl	1—i  i i 11 in	i	i  i i 11 nl   i  i  i i 11 ii
                                          10J
                                           10-=
                           RESIDUAL—TIME   IN MG-MIN/L

Figure 26.   Coliform inactivation by chlorine  and chlorine dioxide in  San Jose

            secondary effluent.
                                     70

-------
     In general, chlorine dioxide is a more efficient bactericide  in  secondary
effluent than chlorine when  comparison  is  made on a residual or residual-time
product basis  at a given  level of inactivation.   The  advantage  for  chlorine
dioxide is greatest  in filtered secondary effluent  and at lower  Rt values  or
correspondingly higher survival ratios.

Nitrified Effluent
     A comparison of  bacterial  inactivation in San Jose nitrified  effluent  is
shown  in  Fig.  27.   The advantage of  chlorine dioxide over  chlorine seen  in
secondary effluent  has disappeared  due  to the  increase  in efficiency of  the
free chlorine  residual in the  nitrified effluent.  RR^  and  RR2 are 0.69  and
1.29, respectively for  these experiments.

Nitrified-Filtered Effluent
     The results  of  experiments with San Jose  filtered effluent are  compared
in Fig.  28,  with  corresponding RR.^ and  RR2 of 0.02  and 0.88, respectively.
The tremendous advantage of  chlorine dioxide disinfection seen due  to  filtra-
tion at low values of Rt decreases with increasing Rt product.  This result  is
consistent with  the  results  of experiments  in  Palo  Alto filtered  (secondary)
wastewater as discussed above.

     The correlations  based  on Dublin  lab  experiments  are  shown in Fig.  29.
In this  case,  however,  the  RR,  and RR2 of  0.18  and  0.09  indicate that  the
chlorine dioxide  advantage remains  relatively constant over the full  range  of
Rt products considered.  The Dublin  field experiments confirm  this  observation
as seen in Fig.  30.   The RRj^ and RR2 for field experiments  are 0.27 and 0.10,
respectively.

     In general,  chlorine dioxide is a more efficient  bactericide  than chlor-
ine in filtered, nitrified wastewater.  The advantage found  for chlorine diox-
ide is  due  to filtration  and not nitrification.   The advantage for  chlorine
dioxide  decreases at  lower  survival  ratios  in San  Jose  nitrified-filtered
effluent, while  remaining  relatively  constant in  Dublin  nitrified-filtered
effluent.

Summary

     The  results obtained in this  work  indicate  that  chlorine  dioxide  is a
generally superior disinfectant  when compared  to  chlorine on  the basis  of
residual  *  time  product required to achieve  a given  total coliform  survival
ratio.   This conclusion is  in agreement with  previously reported  laboratory
investigations (34-39) and field  studies  (40-43).
PREDICTING FULL-SCALE PERFORMANCE FROM LABORATORY EXPERIMENTS

     A  comparison  of  the  laboratory and  field  (full-scale)  experiments  in
Dublin  wastewater  are  shown  in Figs.  31  and 32  for  chlorine dioxide  and
chlorine,  respectively.    In  both  cases,  the  regression  equations from  the
laboratory  experiments slightly  underestimate the  survival  observed  at the

                                      71

-------
       o
      1
10


10


10


10


10
               2
              -3
              -4
              -5
                  CHLORINE       \\
                  DIOXIDE
                                               CHLORINE
                            1 1 ii   i _ i  i i 1 1 1 il _ i _ i — i i 1 1 ill _ i — i — i i 1 1 ii
                                          _ _ —
                            10           101          102
                         RESIDUAL — TIME   IN MG-MIN/L
Figure 27.  Coliform inactivation by  chlorine and chlorine dioxide in San Jose
            nitrified effluent.
            10-1

            ID"2

            10~3
o    10
§
_l    10
              -4
              -5
            10
              -6
                         \
                          \
                            \
                              \
                                            \
                                             \
                             CHLORINE
                             DIOXIDE
                                     \
                                        CHLORINE
                                                   \
                           10°         101          102
                        RESIDUAL—TIME   IN MG-MIN/L


Figure 28.   Coliform  inactivation by chlorine and chlorine  dioxide in San Jose
            filtered  effluent.
                                     72

-------
      o
      I—I
      E-i
           10
             -1
           10
             -2
10~3

10-4

10"5
           10
             -6
                            CL02 AS
                            DISINFECTANT
        CL2 AS
        DISINFECTANT
\
                           10^         10^          10^
                        RESIDUAL—TIME  IN MG-MIN/L
     Figure 29.  Coliform inactivation by chlorine and chlorine dioxide in
                 laboratory experiments with Dublin effluent.
Figure 30.
            10
              -1
            10
              -2
            10
              -3
       o    1Q-
            10
              -5
            10
              -6
                              \
                    \
                                   \
                                     \
                                         \
                CL02 AS
                DISINFECTANT
                                            \
                                              \
          CL2 AS
           )ISINFECTANT
                                                \
                                                  \
                                             j—i i i i..i nl	1	i i  i 11 ti
                10°         101         10
             RESIDUAL—TIME  IN MG-MIN/L
Coliform inactivation by chlorine  and  chlorine dioxide in field
experiments with Dublin effluent.
                                     73

-------
             10-1

             10~2

             10~3

             10-4

             10~5

             io~6
                               \
                           x
          X
             X
       LAB     X
       EXPERIMENTS v
                                        X
                                           FIELD
                                           ^EXPERIMENTS
                                          x
                                            X
                                              X
                                                X
                                             -I	I  I I I I III
                                                          I  I  1 I I 111
    Figure 31
                    10°         101         102
                 RESIDUAL—TIME  IN MG-MIN/L

        Coliform inactivation by chlorine in laboratory  and  field
        experiments with Dublin effluent.
o
P
Figure 32.
             10
               ,-1  1-
             10
       -2
             10
               -3  __
             10
               ,-4
             10
               -5
             10
               -6
                         \
                           x
                             \
                               X
                                 x
                                   X
          X
LAB         X \
EXPERIMENTS    x
                                            X
                                              FIELD
                                              EXPERIMENTS
                                              x
                                                X
                                  _l	I I  IJ I 111
                                       I  I I I I I II    I  1 I I I III
                     ±3°         101         102
                  RESIDUAL—TIME  IN MG-MIN/L

    Coliform inactivation by chlorine dioxide  in  laboratory and field
    experiments with Dublin effluent.
                                     74

-------
full-scale treatment plant.   The  regression lines,  however, are not  statisti-
cally different from each other  for  either chlorine dioxide or chlorine.  The
agreement  is  actually quite  good when  factors  such  as  temperature,  contact
time, and  chemical mixing  are considered.   In  field experiments,  it is ex-
tremely  difficult  to  reproduce  the  ideal mixing,   constant  temperature, and
controlled oxidant  generation chemistry found in  the laboratory experiments.
From Figs. 31  and  32 it can  be  concluded  that laboratory bench-scale  experi-
ments of the  type  described in this  report  would be a valuable design aid  in
predicting the  residual  and/or contact  time required to  achieve an  effluent
bacterial  standard.   Bacterial  response  to disinfectants  is  sufficiently
varied in  different wastewaters and  within the  same  treatment plant at  dif-
ferent levels  of  treatment that  no  universal disinfection  equation could  be
expected to yield  accurate  predictive results.   The coefficients of  the  model
used to  describe  the experimental results  in  this  study are highly  dependent
on wastewater quality and treatment history of the wastewater under  considera-
tion, as well as on the choice of disinfectant.
                                       75

-------
                                   SECTION 8

                              VIRUS INACTIVATION
     In  view of  the  controversy over  the  validity of  the  coliform group  as
adequate indicators of disinfection using chlorine  (44-49) or  chlorine  dioxide
(50,51),  additional  indicators  should  be compared  in evaluating a  disinfec-
tant.   The  use  of animal  viruses  is  suggested,  since  several authors  have
shown that viruses, such  as the widely studied poliovirus, are  typically  more
resistant  to disinfectants  than are  enteric bacteria  (52,53).  Poliovirus,
while not  of epidemiological significance in  the  United States, may serve  as
an indicator for viruses,  such as  the Hepatitis virus,  that  can be dissemi-
nated via  the water route.   Katzenelson and  Kedmi  (54), however, noted  that
poliovirus may  not be a  suitable indicator since the virus  could not be de-
tected  consistently  in urban  sewage.   For that reason,  and  also for  ease  of
recovery  and analysis, other  groups such  as Kott  et al.  (55) and Atlantic
Research Corp.  (4) have  proposed observing coliphages in wastewater  disinfec-
tion studies.  These bacterial  viruses  are generally present in  number  concen-
trations approaching those  of  the coliforas,  and their structure more  closely
approximates that of pathogenic viruses than do coliform  bacteria.  The use  of
viruses  to monitor water  quality in receiving waters  has also been suggested,
since human  enteric  viruses have been  observed  in water that were determined
to be safe by bacteriological  standards (56-58).

     Analogous protective mechanisms offered by particulates to  bacteria apply
toward the protection of viruses  against inactivation.  Virions  may be  protec-
ted  against   chlorine  by  aggregation  (59-61),  adsorption  onto particulates
(62), or occlusion within organic solids  (63).  Also,  the physical and chemi-
cal  conditions  that  permit viral  aggregation may  not  only  protect  against
inactivation, but  may  also  allow partially inactivated virions  to reassociate
and  to  cooperatively  infect host cells (64).   The same  protective mechanisms
apply for  disinfection with chlorine dioxide (65) and may be  even more impor-
tant, because chlorine dioxide can  react  very quickly and be  consumed  in  com-
peting reactions with other  sources of  demand  in the wastewater.
RESULTS

Palo Alto (Non-Nitrified) Effluent

     Selected experiments were conducted  to assess  the viricidal  effectiveness
of  chlorine dioxide  as compared  with chlorine.    In  situ  coliphage  and  an
inoculum of Poliovirus  I (10° PFU/100 ml)  in Palo Alto secondary  effluent  were
dosed with  5  mg/1  chlorine  dioxide and chlorine.   Samples were taken at  2-,

                                      76

-------
5-,  and  10-minute  contact  times  and analyzed  for coliphage  (Kott Method),
Poliovirus (standard plaque assay),  and  for fecal coliforms (MF method).   The
results are  shown in Fig.  33.   In  both trials, chlorine  dioxide was a much
more  effective  viricide  than  was chlorine,  although  the  coliform survivals
would indicate that both disinfectants were performing  equally well.

     Also, the log  reductions  of the coliphage  and poliovirus by each of  the
disinfectants were  similar,  a  response  that had  been  observed in preliminary
experiments  not  reported.   Chlorine dioxide  inactivation  of both  phage  and
virus was more rapid and  of a  much greater magnitude than was inactivation of
these indicators by an  equivalent  dose  of chlorine.  Since the test procedure
for  poliovirus  is much  more expensive  and time-consuming than  the recovery
procedure  for  coliphage, the  coliphage was chosen as an  indicator of virus
response to both disinfectants.  Therefore, in situ phage, which may be a more
realistic  indicator of  native  populations  of other  pathogens,  was  used  as
opposed to using inocula of laboratory-grown strains of virus•

     The results  of  the phage  analyses  that were run parallel with the total
coliform analysis are  shown in Figs. 34 and 35.   Once again, the total coli-
form  bacteria  respond  similarly  to   the  stress  of a  given disinfectant dose
(Fig. 34), whereas all three doses of chlorine dioxide  (2,  5, and  10 mg/1)  are
much more  effective  at  inactivating  the  coliphage  than even the highest dose
(10 mg/1)  of chlorine  (Fig. 35).  The  curve for the  10 mg/1 dose of chlorine
dioxide  represents  the  detection limits of  the  Kott assay  for coliphage.
Therefore  this  curve is  an  upper  limit for the  survival  ratio  and should be
interpreted  not  as  equivalent  to,  but  rather as  less than or  equal to  the
survival at  5  mg/1  chlorine dioxide  dose.   This suggests that accepted indi-
cator organisms  (total  coliform bacteria) yield conservative performance data
for the viricidal effectiveness of chlorine dioxide in  secondary effluent.

     Another set  of experiments was  conducted to  assess  the biocidal effec-
tiveness of  each disinfectant  at very  short contact  times, since the results
of  Fig.  35  indicate  that  CK^ reacts  rapidly, within  less than  5  min,  to
inactivate  phage.    These experiments  were run  at 5-mg/l doses  of chlorine
dioxide or  chlorine  and sampled at  15,  45,  90,  and 120 seconds.  The results
(Fig. 36)  again  indicate that  chlorine dioxide  is  a more effective viricide
than  a  bactericide, and  that  chlorine  dioxide  performs better  in secondary
effluent than chlorine within a short contact  time  (< 120 seconds).

Comparison of Virus Inactivation in Palo Alto  and Dublin Effluents

     Additional  experiments  were  conducted to  compare  several indicator orga-
nisms with  the  conventional total  coliform group in  different  wastewaters.
Two  effluents,   a  non-nitrified  (Palo   Alto)  and  a  nitrified,  sand-filtered
(Dublin)  secondary  effluent were compared  in  laboratory  studies.   The nitri-
fied, sand-filtered  effluent was used  in bench-scale  and full-scale studies,
whereas  the  non-nitrified   effluent  was  used  only  in  bench-scale  studies.
Wastewater characteristics  for  the Palo  Alto and  Dublin effluents  are shown in
Table 14 and Fig. 37.  Microbiological analyses  consisted of:  total coliforms
by  MF method,  coliphage ($g) by  the  Kott method, coliphage (*c) by the RPTRR
method,  and inoculated  Poliovirus 1  (NQ = 10°  PFU/100  ml)  by the  standard
plaque assay.

                                      77

-------
                  k
                -2
                -4
             2° -6
             O

             g
             n
             _i

             o:
-8
            5 mg/l
             C102
                -2
                -4
                -6
                -8
      LEGEND
FC = FECAL COLIFORM
<£B= COLIPHAGE,E_.CoM   B  HOST
VIRUS = POLIO VIRUS
     \  ^d-
      \
       \
                             	D	
              FC
	D
	D
             VIRUS

              _1
               10
                            5 mg/l
                             CL2
                                                        10
                             CONTACT TIME (min)
Figure 33.   Comparison of in situ  coliphage and an inoculum of Poliovirus I in
            non-nitrified secondary  effluent.  Above, treatment with 5 mg/l
            chlorine dioxide;  below,  treatment with  5 mg/l chlorine.
                                      78

-------
        O   -
        <
        o:
        Q:
o:
£
d
o
i
Figure 34.
            -5
            -6
            -7
                                                                2 mg/l
                                                                5 mg/l
                                                               }10 mg/l
               LEGEND
               O--0 C102
                                     15
                            CONTACT  TIME (min)
                                                30
Total coliform survival in non-nitrified secondary effluent,
comparing chlorine dioxide and chlorine at three doses and three
contact times.
     The nitrified sand-filtered effluent (Dublin) was markedly different from
the non-nitrified secondary effluent (Palo Alto).   Typical characteristics for
each  wastewater  are  summarized in  Fig. 37.   The  results  from  bench-scale
experiments  (Fig.  37)  indicate that  chlorine dioxide  and  chlorine  perform
equally well at  a  2-mg/l  dose in the  non-nitrified effluent, yet  chlorine
dioxide is  more  than twice as  effective as chlorine in  the  nitrified,  sand-
filtered  effluent  as measured  by  total  coliform  survival.    Disinfection as
modeled by  inactivation  of inoculated Poliovirus  1 is much  greater for chlo-
rine  dioxide than chlorine at  the 2-mg/l  dose.   Similar results  were shown
earlier (Fig. 33) for a 5-mg/l dose in the non-nitrified effluent.   Therefore,
the  total coliform group  serves  as  an  adequate,  and possibly conservative,
indicator of virus inactivation for chlorine dioxide in both effluents but may
not be an adequate indicator when chlorine is used as a disinfectant.
                                      79

-------
         CD
         O
I

ct

en

CD
o

LLJ
CD

I
CL
_i
O
o
             -3
             -4
             -5
            -7
           \
                      D
 2 mg/l

 5 mg/l

10 mg/l
                                                 2 mg/l
                                               "I0~mg/
               .-•^B

                   I
                              LEGEND


                              D---0 C102
                                     15


                            CONTACT TIME (min)
                                                  30
Figure 35.   In  situ coliphage (_E.  coli  B host) survival in non-nitrified

            secondary effluent,  comparing chlorine dioxide and chlorine at

            three doses and three  contact times.
                                     80

-------
         -1
    •M
     o
     Q
C£.

<
     O
     O
       \
       \
        \
      h \
         -4
         -5
         -6
         -7
                                           TOTAL COLIFORM
                                        _n
                                           FECAL COLIFORM

                                           TOTAL COLIFORM
    	D
FECAL COLIFORM
                          D
                                    — _  COLIPHAGE  E.COLI B HOST
                  LEGEND

                  O—D C102  5 mg/l
                  •—• C12    5 mg/l
                          J_
         15         45              90
                  CONTACT  TIME (sec)
                                                    120
Figure 36.  Responses of several in situ organisms to chlorine dioxide
           and chlorine at  very short contact  times in non-nitrified
           secondary effluent.
                                 81

-------
                         2mg/l  CI02 or Cl  DOSE
                         TOTAL  COLIFORMS
                             .	  , PALO ALTO  CI02
                                    PALO A"LTO~~C"£~~
                                    DUBLIN  CIO,
                        	D	£	
          -5
                              POLIOVIRUS IN
                              DUBLIN EFFLUENT
                   5       10
                      CONTACT  TIME(min)
TYPICAL CHARACTERISTICS

Palo Alto
Dublin
NH3-N mg/l
35.5

-------
                                   SECTION 9

                        RECOVERY  FOLLOWING DISINFECTION
     The need to disinfect wastewater has been questioned (66,67) for a number
of reasons.   Firstly, the  environmental  hazards posed  by  disinfection  (e.g.
carcinogenic byproducts)  have been suggested to  outweigh  the benefits (e.g.,
public health protection).  Secondly, there is little epidemiological evidence
that non-consumptive contact with receiving waters that have been contaminated
with fecal material can cause human disease.   Although subclinical infections
from recreational  contact  are possible (68), they  are  difficult to associate
with contaminated  water sources.   Thirdly,  there  is evidence  that  more  re-
growth of bacteria can  occur  following chlorination than if the effluent were
discharged without disinfection (69-81).

     The conditions can exist in receiving waters for coliforms to persist  for
long periods of  time  (82-84).  It  is  also  possible for pathogenic strains  of
coliforms, such as Klebsiella species,  to proliferate (85)  and maintain  their
virulence  (86-88).   Several  factors  affect the  survival  of bacteria  and
viruses in receiving waters,  including temperature, light, and predators.   The
disinfection process  may  reduce  the  controlling biological factors  such  as
predation and  competition,  and allow wastewater  flora  to  recover and grow in
the receiving environment.
EXPERIMENTAL

     A  set  of experiments  was designed  to  test for  aftergrowth of  bacteria
following disinfection.   Three  replicate experiments were  run,  each  lasting
five  days.    In  each experiment, evidence  of bacterial growth  was sought  in
three reactor vessels:  a secondary effluent control (no treatment), chlorine-
dioxide-treated  secondary  effluent, and  chlorine-treated  secondary effluent.
Each  of  the  latter two were dosed  at  5 mg/1.   All  reactors were diluted  1:4
with  lake water  after  30 min to simulate receiving water dilution.  All  reac-
tors  were exposed  to the same conditions  of  temperature and light.   Parallel
samples  were taken  for total  coliforms,  fecal  streptococcus,  seeded polio-
virus, and coliphage.
RESULTS

     The  results  (Figure  38)  indicate a  slight decline  in  coliforms in  the
control.  A  slight  increase (approximately one log unit) was observed  in both
chlorine-dioxide- and chlorine-treated samples,  although  for chlorine  dioxide

                                      83

-------
       8
       UJ  2
       O
       2
       O  i
       O  6
           2 -
                      TOTAL COLIFORMS
                                           CONTROL
                                          CHLORINE

                                          CHLORINE
                                          DIOXIDE
                    FECAL STREPTOCOCCUS
                                          CONTROL


                                          CHLORINE
                                          CHLORINE
                                          DIOXIDE
I	I
                         VIRUS
                                          POLIO
                                          CONTROL
                                         COLIPHAGE
                                          CONTROL
            30min
                    TIME  (DAYS)
Figure  38.   Five-day recovery  following  disinfection.


                            84

-------
the initial kill (30 min contact) was greater, and the number of coliforms re-
maining was more suppressed than for chlorine.   However, the rate of regrowth
of  coliforms  after chlorine  dioxide treatment  was  equal to  or greater than
that following chlorine  treatment.   The  untreated fecal streptococcus samples
showed a  marked  decline after  24  hrs that  remained  relatively stable, while
both disinfected samples showed a slight but perhaps insignificant recovery.

     The  Poliovirus  I and  coliphage concentrations declined  steadily  in the
untreated control  samples.   The log values  for  the 30-min treated samples of
polio and coliphage were 1.99  and  1.3,  respectively,  and remained at or below
these detection  limits  for the  duration of the  experiment.   Therefore, only
the control  values for  the  phage  and virus are  shown,  since meaningful data
could not be obtained in disinfected samples.   The inactivation rate (approx-
imately 0.5 log/24 hr)  of  Polio 1  at these temperatures (22-24°C) is slightly
lower than has been  reported for thermal  inactivation  and lysis by bacterial
enzymes  in  a surface water (89).    The  inactivation rate  for coliphage, the
animal virus model,  parallels the polio  inactivation  rate (approximately 0.5
log/24 hr).   The mechanisms  of inactivation may  or may not differ, however,
since the phage  is capable  of  reproduction under these conditions whereas the
poliovirus is not.  The fact that the phage did not increase may mean that the
host coliforms  were not growing actively  enough to allow  reproduction, and
hence a measurable increase, in the  phage population.

     The  general  trends  observed in the  recovery  experiments agree with pre-
viously  reported results  (66-71).    Namely,  a  reduction in  fecal indicator
bacterial  levels, and  presumably  the  levels   of  enteric  pathogens,   occurs
within about 24  hrs  after  discharge  of an untreated wastewater  into a  receiv-
ing water;  conversely,  an aftergrowth of  indicator bacteria occurs following
disinfection by  chlorine.   This same  recovery  pattern is observed for  an ef-
fluent disinfected with  chlorine dioxide.   The specific rates of  inactivation
or  recovery may depend on such experimental variables as temperature, dilution
ratio  (wastewater:receiving water),  concentration of  predators in receiving
water, and water chemistry.  These observations  confirm  the concerns of  previ-
ous workers (65-67) that the need to disinfect effluents discharged into a re-
ceiving water with which humans will not  experience immediate, direct contact,
should be re-examined with epidemiologically  designed experiments similar  to
that of Cabelli et al.  (104).  This  conclusion is  subject  to  uncertainty as  to
whether enteric pathogens generally  exhibit behavior similar  to  the indicators
studied  here with  respect  to:    inactivation  during  disinfection;  in situ
inactivation in receiving waters; and recovery following disinfection.
                                       85

-------
                                  SECTION 10

               HALOGENATED  ORGANICS  PRODUCED DURING DISINFECTION
CHLORINE DIOXIDE VERSUS CHLORINE

Introduction

     The  concern  over  halogenated  byproducts  of  chlorine  disinfection has
stimulated  a  search  for  alternative  disinfectants,  among which  is chlorine
dioxide.  In  this  section chlorine dioxide is compared with chlorine alone on
an  equal  mass-dose basis  with regard  to  the amount  of  halogenated organics
formed in wastewaters.  Halogenated  organics are quantified as total trihalo-
methanes  (THMs)  and  as  total organic  halogen  (TOX).   Comparisons between the
two disinfectants  were  carried out  in bench-scale studies  and in  full-scale
field studies.

     Within normal pH ranges  encountered  in  water  and wastewater  treatment,
chlorine dioxide reduces  to  chlorite when  it reacts with organics (12,13,90).
Chlorine dioxide does not oxidize bromide  ion  as does chlorine  (91).

     Chlorine dioxide apparently does  not  form trihalomethanes  (91,92), but it
can produce halogenated organics  in  some cases.   Compounds with carbon-carbon
double bonds  are one group of compounds which can be halogenated.  Work with
aqueous solutions  of cyclohexane  and methyl  oleate  has  produced chlorinated
products (13).  Chlorine dioxide can also  react with phenol to  produce chloro-
quinones,  chlorohydroquinones,  and  chlorophenols  (13,91).   When chlorine
dioxide is  used  in  excess  with  respect  to  the  organic reactants,  such as
cyclohexane and  phenol,  non-chlorinated organic acids  are formed rather  than
chlorinated products (13,91).

     Studies  dealing  with the nature  of humlc compounds  suggests that phenol
derivatives and  benzene carboxylic  acids are major subunits (93).   Therefore,
considering the  chlorinated  products resulting from chlorine dioxide reaction
with phenol,  one can expect to find  some  halogenated organics resulting  from
reactions in  water containing  humic  substances.   Manka et al.  (94)  have  shown
that humics are  present  in treated  wastewater effluents.  Phenols  themselves
have been  found  in domestic secondary effluent  at 0.1 to 1.0  mg/1  concentra-
tions (95).

Experimental

     The experiments presented in this section examine  the production of  halo-
genated organics in  wastewater by chlorine or by chlorine dioxide.  Two  types

                                       86

-------
of wastewaters  were used  in these  experiments:   secondary  effluent from an
activated-sludge  process  at  the  Palo  Alto  Wastewater  Treatment  Plant and
effluent  from  the Dublin-San Ramon  Wastewater  Treatment Plant  which had re-
ceived extended-aeration,  activated-sludge  treatment (achieving full nitrifi-
cation)  and multimedia  filtration.    The  characteristics  of  the wastewater
effluents are summarized in Table 29.  Each experiment represents a wastewater
sample taken on a different day.

     The  experiments were  primarily  bench-scale  experiments;  however,  some
full-scale  field  experiments were conducted  to supplement  the  findings  from
the bench-scale experiments.  In the bench-scale experiments wastewater -efflu-
ents from both Palo Alto and Dublin-San Ramon were used.  As shown in Fig. 39,
three or  four liters of  wastewater was dosed with either chlorine or chlorine
dioxide.  The  doses for both disinfectants were  40  mg/1 and 20 mg/1 measured
as ClO^ for chlorine dioxide and as Cl for chlorine.  The dosed  effluents were
mixed  and then transferred  into individual 0.5-  and 1-liter glass-stoppered
reactor  bottles within five minutes  after  dosing.   The reactor bottles  were
filled completely to exclude headspace.   These  reactor bottles  were  also cov-
ered with aluminum  foil  to prevent  light from influencing the reactions.  The
secondary effluent  in  these bottles was  allowed  to  react for 1 to  24 hrs or
longer at 25°C.   At the completion of  the  respective reaction periods,  each
reactor  bottle  was  opened, and  samples  for THM and TOX analyses were  trans-
ferred  immediately  to  60-milliliter  hypovials.    Excess  sodium thiosulfate
previously introduced into these hypovials quenched  any disinfectant  residual.
The  remaining  water from the  reactor bottle  was  analyzed  for  disinfectant
residuals.  TOX analysis was by  the Dohrmann method  (Section 4).

     The  field  experiments  were run  at the  disinfection  facilities  of the
Dublin-San Ramon Wastewater Treatment  Plant.  The  influent to the  contact tank
was nitrified, filtered activated-sludge  effluent.   A diagram of the  contactor
is shown  in Fig.  10.  Wastewater characteristics in  the  chlorine contactor are
summarized  in  Table 30.   At steady  state,  THM  and TOX samples  were  taken up-
stream from the disinfectant injection point (Point A in Fig. 10) and  also at
the  end  of the contactor  just  upstream of the outfall  (Point  D in  Fig.  10).
Sodium thiosulfate was used to quench  disinfectant residuals.    Disinfectant
      TABLE 29.  TYPICAL CHARACTERISTICS OF WASTEWATER  EFFLUENTS  USED TO
                   MEASURE  FORMATION OF HALOGENATED ORGANICS

                                                   Ranges
                                       Palo Alto          Dublin-San Ramon
            Parameter               (Non-Nitrified)         (Nitrified)

     COD, mg/1                           20-55                 20-35
     NH3-N, mg/l-N                       20-35              <  0.04-0.2
     Total Kjeldahl  N, mg/l-N            26-42              < 0.8-1.1
     Alkalinity, mg/1 CaC03             190-250               100-300
     PH                                 6.9-7.3               6.9-7.7
     Suspended  Solids, mg/1              10-50                  0-13
                                       87

-------
  4-liter  batch mixer
0.5-or I-liter
   reactors
eOmilliliter
sample  vials
   Figure 39.  Procedure for the bench-scale  determinations of halogenated
              organics formation.
    TABLE 30.  WASTEWATER CHARACTERISTICS  IN  THM AND TOX FIELD EXPERIMENTS
                 AT THE DUBLIN-SAN RAMON  CHLORINE CONTACTOR
                     Parameter
         Range  in  Contactor Influent
             COD, mg/1
             NH3-N, mg/l-N
             Total Kjeldahl-N, mg/l-N
             Alkalinity, mg/1 CaCOo
             PH
             Suspended Solids, mg/1
                   23-33
                   < 0.06
                   0.9-2.0
                   105-296
                   7.0-7.4
                     1-7
doses were chosen to achieve a desired bacterial kill;  hence,  the two  disin-
fectants could not be  compared on  an equal mass basis in  the  field  experi-
ments.  For the  two later field experiments, only chlorine dioxide was used.
TOX analysis was  by the Jekel method (Section 4).
                                     88

-------
Results

     The bench-scale experiments showed that chlorine produced THMs and TOX  in
both wastewaters  tested.   Experimental conditions and  the amounts of THM and
TOX formed  during 24-hr contact are summarized  in Table 31.  Typical results
for the  THM and  TOX formation by chlorine and chlorine dioxide   in Palo Alto
       TABLE  31.   THM AND TOX FORMATION BY CHLORINE AND CHLORINE DIOXIDE
                          IN BENCH-SCALE EXPERIMENTS
            Exp.
Wastewater  No.
                     Chlorine
            Chlorine  Dioxide
              Dose,     Dose,
             mg/1 Cl mg/1 C107
  Average 24-hr
    Production

  TOX       THM
ymol/l-X  ymol/l-X
                                                               Average 24-hr
                                                                 Residual
  Cl,
CIO,
mg/l-Cl  mg/l-C102
1



3



4
Non-
Nitrified
/Tain
Alto)





12



Nitrified 6
(Dublin-
(-1 -n \
ban itamonj1
7
20
none
20
none
40
none
20
none
40
none

20
none

40
none
20
none
40
none
20
none
40
none

20
none
none
20
none
20
none
40
none
20
none
40

none
20

none
40
none
20
none
40
none
20
none
40

none
20
6.0
0.0
—
__
—
— —
10.0
3.0

1.6
.
7.4T
1.7

8.1
-0.8
7.1
0.9
7.1
-0.7
22.4
0.5
26.2
1.4

19.3
-0.2
0.53
-0.03
0.22
-0.06
0.91
-0.02
0.18
0.02
0.85
0.00

0.43T
0.00

1.31
0.10
0.45
-0.01
0.98
0.01
3.97
0.01
4.87
0.03

3.75
0.02
6.1
NA
6.0
NA
9.1
NA
6.5
NA
16.2
NA

6.8T
NA

15.1
NA
9.3
NA
18.5
NA
6.6
NA
19.0
NA

7.5
NA
NA
0.5
NA
8.0
NA
8.0
NA
3.4
NA
3.4

NA

NA
10.4
NA
5.2
NA
3.8
NA
8.4
NA
24.5

NA
11.2
      NA = not applicable

      All pertinent  experiments  are  shown;  missing experiment numbers refer to
     experiments conducted with  other  objectives.
     t,
      #
'23  hrs.
 25  hrs.
                                       89

-------
wastewater are shown  in  Fig.  40 as a function of  contact  time  up to 24 hrs.
The data  in  Figure 40 illustrate the lack  of  production  of  THMs by chlorine
dioxide and  a small  apparent  TOX production by  chlorine  dioxide.   However,
other bench-scale experiments in the Palo Alto wastewater show no TOX produc-
tion by chlorine dioxide at 24-hr contact  time.  THM and TOX production in one
of the two bench-scale experiments which used the  Dublin-San Ramon wastewater
is  shown  in  Fig.  41.   As with  the  Palo  Alto wastewater,  chlorine dioxide
caused no production  of  THMs,  but caused  a slight TOX production as shown in
Fig. 41 for the 40-mg/l chlorine dioxide dose.

     Results  of  the field  experiments, in  which   THM  and  TOX formation were
measured during approximately one-hr retention  in  the Dublin-San  Ramon contact
tank, are summarized in Table 32.   These experiments were conducted at disin-
fectant doses having  approximately  equal disinfection  effectiveness (survival
ratio); the  dose of chlorine dioxide is  nearly  ten times less  than  that of
chlorine.   Under these conditions, chlorine  dioxide clearly formed no signifi-
cant amounts  of  either THM or  TOX, whereas chlorine  increased  the values of
both parameters substantially.

     Chlorine dioxide  in some  cases  produced  small amounts  of  TOX in batch
experiments at doses  of  20 and  40 mg/1.  Of the  10 values given in Table 31,
one  value (Experiment  4,   20  mg/1 dose)  showed  TOX  production by chlorine
dioxide greater  than  two standard deviations  (page 15)  for  the Dohrmann TOX
procedure, i.e., TOX  > 1.96 pmol/1. .The mean values  and standard deviations
of TOX  and  THM production  in  the bench-scale  experiments  are shown in Table
33.  Apparent negative TOX production  values are less than two standard devia-
tions and are therefore statistically  indistinguishable from  zero production.
      TOTAL  THM
      PRODUCTION
                              mg/1
                           20 mg/1 Clz
                                   CIO,
                     40 mg/1 CI02
    0    5   10   15  20   25
      TIME  IN   HOURS
30
                                          Z30
                                          O25-

                                          I*H
                                            -5
                   TOX
                   PRODUCTION
                                                              4Omg/ICI2

                                                                  20 mg/1
                                                                          CIO
                                    40*1*1/1
0   5   10  IS  20  25
  TIME  IN  HOURS
30
        Figure 40.  TOX and THM production in  non-nitrified effluent
                    (Experiment 4—Palo  Alto wastewater).
                                      90

-------
                TOTAL THM PRODUCTION  35T
  TOX  PRODUCTION
                       40mg/l
               ^^=^-====?==&
                 ^*     on.*t«*/i r*iA
                       20mg/l CI02
              0   5   10   15   20  25

                TIME IN  HOURS
     40mg/1CI02
     ^— ^^" ^^B«BB*^^_

     ^rngFobj"
5   10   15   20  25
TIME IN HOURS
     Figure 41.   TOX  and THM production in filtered, nitrified effluent
                 (Experiment 6—Dublin-San Ramon wastewater).

             TABLE 32.  SUMMARY OF THM AND TOX  FIELD EXPERIMENTS
Average
Production
Field
Exper .
Date
2-6-79
2-15-79
Chlorine
Dose
mg/l-Cl
18.7
none
19.6
none
Chlorine
Dioxide
Dose
mg/l-C102
none
3.1
none
2.3
Contact -
Time
Minutes
55
55
49
46
at Point D
TOX
ymol
10.2
-0.5
8.1
0.4
THM
pmol/l-X
1.02
-0.04
0.94
0.04
Average
Residual
at Point D
mg/l-Cl
12.8
NA
1.0
NA
mg/l-C102
NA*
0.6
NA
0.1
      NA = not applicable.
Except for  the  20  mg/1 chlorine dioxide dose in Experiment 4,  the  other  in-
stances of significant TOX production mentioned are near  to  the two standard
deviation limit.   Nonetheless,  it cannot  be excluded from  possibility  that
chlorine dioxide can  form  halogenated  organics  in wastewater.   More  experi-
ments with greater precision in TOX  analysis  will  be  required to make a firm
conclusion as to whether  chlorine dioxide does produce halogenated organics in
wastewater.   More  important, however, is the  fact  that  chlorine dioxide pro-
duced markedly less halogenated  organics,  if  indeed any,  compared to chlorine
in the two types of wastewaters  tested.
                                     91

-------
     The  means  and  standard  deviations of  THM production  shown in Table  33
support previous  findings (91,92)  that  chlorine dioxide  does  not form  THMs.
Of all  of the bench-scale  experiments (Table  31),  only the 40  mg/1  chlorine
dioxide dose  in  Experiment 8  showed a THM  production exceeding  two  standard
deviations for THM  production,  i.e.,  THM > 0.08 pmol/l-X  (Table  1).   This  THM
production value is  the  only  instance exceeding two standard deviations  among
many other results to the contrary.  Also, this  value  exceeds the two  standard
deviation limit  by only  a  slight margin.   Therefore,  it  is safe to  conclude
that  chlorine dioxide did  not produce THMs  in the wastewaters  tested.    The
apparent  negative  THM production  values  in  a  few  experiments  are within  two
standard  deviations  of  zero  production and are  therefore  statistically indis-
tinguishable  from zero.
     TABLE 33.   MEAN VALUES AND STANDARD DEVIATIONS OF 24-HOUR HALOGENATED
        ORGANIC BYPRODUCT FORMATION WITH CHLORINE  DIOXIDE AND  CHLORINE

                                          Increase During  24-Hr  Contact
Mass ^
Dose
Effluent mg/1 Disinfectant
20
20


40
20
20
Dublin 4Q
40
Chlorine
Chlorine

Chlorine
Chlorine
Chlorine
Chlorine
Chlorine
Chlorine
Dioxide
Mean
0
.005
0.39

Dioxide

Dioxide

Dioxide


0
1
0
3
0
4

.05
.15
.015
.9
.03
.9
TTHM

Std.Dev. n~
0
0

0
0
0
0


.02
.15

.06
.23
.007
.14
-
—
4
4

2
2
2
2
1
1


Mean
1.
7.

-1.
7.
0.
20.
1.
26.
3
6

2
6
15
8
4
2
TOX
Std.Dev
1.3
1.7

0.6
0.7
0.5
2.2
-
—

. nf
4
4

2
2
2
2
1
1
      Increase  calculated  as  concentration  measured  after  24  hours  less
concentration measured before.

      n = number of replicates.
Halogenated Organics Yield Versus Disinfectant Dose

     The TOX  produced in both  the  Palo Alto  and  the Dublin-San Ramon  waste-
waters by chlorine  is  on the order  of one percent of  the  chlorine  dose.   This
result is  the same as  that  found by  Jolley (96) in  his  work with  secondary
effluent.  However, the  total  chlorine residuals had  decreased to  half  of the
initial  chlorine  dose or less  after  24 hrs in  nearly all of the  bench-scale
experiments (Figs.  42  and 43).   This  result suggests  that  99% of the chlorine
demand is in the form of redox-type reactions, whereas only 1% of the chlorine
demand is in the form of halogenation reactions.
                                      92

-------
                            DISINFECTANT  RESIDUAL
                                          45i
               ,	-a  40 mg/l CI02
                                                 20 mg/l CI02
                                                              20 mg/l
             0   5   10  15  20  25  30
                TIME  IN  HOURS
                        5  10   15 20  23  30
                        TIME  IN HOURS
Figure 42.   Disinfectant residuals in Experiment 8 (Palo Alto  wastewater)
                            DISINFECTANT   RESIDUAL
                  5   10
                  TIME
15  20 25 30
IN  HOURS
                                                      	     20 mg/l CI02
                                                     	IJI
                                                               20 mg/l Cl,
0  5  10  15  20 25 30
   TIME  IN HOURS
   Figure 43.  Disinfectant residuals  in  Experiment 6  (Dublin-San Ramon
               wastewater).
                                       93

-------
     Since  TOX  production by  chlorine  dioxide is  not  well documented by  the
experiments conducted, it is not appropriate at this time  to estimate  how much
of  the  initial  chlorine  dioxide demand  is  in the  form of halogenation pro-
ducts.  However, it is worthwhile to point out that  chlorine dioxide residuals
also  decreased  to half (or  less in some  cases)  of the value of the  initial
chlorine dioxide doses (Figs.  42 and 43).  This result  indicates  that  chlorine
dioxide  does  enter into  redox  reactions  with the  substances in wastewater,
consistent with  the findings of  Noack and Doerr (97).

     Figures  40  and 41 show  that  higher chlorine  doses result in higher  THM
productions in  both wastewaters.   In  contrast,  from  Table 31 (Experiments  8
and 12), it can  be seen that in  the Palo Alto wastewater,  there is no  signifi-
cant difference  in  TOX production  between the two chlorine doses used (20  and
40 mg/1).   The  single experiment (Experiment 6) using  Dublin-San Ramon waste-
water and  two chlorine  doses  does  show  a significantly different TOX produc-
tion between  the two  chlorine  doses.   The difference in THM production in  the
Palo Alto wastewater  experiments is too small to be detected  by  TOX analysis.
Therefore, the dose effect on  halogenated organics  production  in  the Palo Alto
wastewater is very  small,  indicating  that the 20-mg/l  chlorine dose is suffi-
cient  to nearly exhaust  the  amount  of  precursors of  halogenated  organics
available  to  it.   In order  to  observe  a  less ambiguous dose effect, lower
chlorine doses should be used  for the Palo Alto wastewater.

The Relationship Between TOX and THM Production

     The results  summarized  in Tables  31 and 32 show the  24-hr TOX production
to  be between five and  fifty times greater  than THM  production in  the Palo
Alto  wastewater  and approximately  five  times greater  than THM production  in
the Dublin-San  Ramon  wastewater.   These  results  are consistent with  the fact
that TOX represents a broader  spectrum of halogenated organics than does THM.

The Behavior of  Disinfectant Residuals
     At  a given  chlorine  dioxide dose,  the chlorine  dioxide residuals  were
lower than  the  chlorine residuals in  the  Palo  Alto wastewater (Fig. 42).   In
contrast, the chlorine  dioxide  residuals were higher than the  chlorine  resid-
uals in  the  Dublin-San  Ramon wastewater (Fig. 43).  Of  course, the  results  of
comparing the residuals of different disinfectants  depends on the  units  chosen
to express  the  residual concentration.   In  this work, a mass  basis  generally
was  used,  because  of  the direct  relation to disinfectant  dose  and hence  to
cost.    A  better  comparison  of  disinfectant   behavior is  on  an   electron-
equivalent basis  for disinfectant  demand as  shown  in Table  34.  From Table  34
it  is  clear  that the  disinfectant  demand  for  chlorine is  greater than for
chlorine  dioxide  in both  wastewaters; that  is,  more electron-equivalents  of
chlorine  than  of  chlorine dioxide are used  up  in  both  wastewaters.  This  is
relevant  to  questions of  reactions between oxidants and organics, because the
number of electron  equivalents  of  oxidant that  is  reduced is  a measure  of the
extent to which redox reactions have occurred.
                                      94

-------
                        TABLE 34.  DISINFECTANT DEMAND
                                          24-hr Demand in meq/1
                                 20  mg/1  Dose"1"               40 mg/1
Wastewater
Palo Alto




Dublin-
San Ramon
Experiment
1
3
4
8
12
6
7
ci2
0.39
0.39
0.38
0.37
0.30
0.38
0.35
cio2
0.29
0.18
0.25
-
0.22
0.17
0.13
ci2
_
0.59
0.67
0.70
0.61
0.59
~
cio2
—
0.52
0.54
0.44
0.54
0.23
—
      Assumes chlorine dioxide  undergoes  a single electron transfer, i.e.,  it
     is reduced to chlorite ion.
     t20 mg/l-Cl =0.56 meq/1; 20 mg/l-C102 =0.30 meq/1.

     #40 mg/l-Cl -1.13 meq/1; 40 mg/l-C102 = 0.59 meq/1.


MIXTURES OF CHLORINE DIOXIDE AND CHLORINE

Introduction

     One of the methods used to produce chlorine dioxide is the  chlorine/chlo-
rite process.  Using excess chlorine could lead to a product  that  is  a  mixture
of chlorine  and chlorine dioxide.   Miltner (98) has  shown  that a mixture  of
chlorine  dioxide  and  chlorine  would  produce  less  THMs  than  would chlorine
alone in the same concentration as for the chlorine used in the  mixture.

Experimental Design

     The  experiments  using  mixtures  of   chlorine  and  chlorine dioxide  were
bench-scale experiments.   The mixtures of  chlorine  and chlorine  dioxide  were
based  on  the mass  ratios  of  chlorine dioxide (as  C102) to chlorine (as  Cl).
These mass  ratios  were 1:5, 1:1,  and  5:1; experiments with 100%  C12 and  100%
C102 were run  for  comparison.   The chlorine dose was  kept constant  at  5 mg/l-
Cl.   Dublin-San Ramon wastewater  was added directly  to the 0.5- and  1-liter
glass-stoppered  reactor  bottles  identical to  those used  for the  experiments
presented under "Chlorine Dioxide Versus  Chlorine."  The chlorine  and chlorine
dioxide doses were  added  with mixing,  and the reactor bottles were  closed and
allowed to  react at 25°C  for  one hour.    This short  reaction time was chosen
because it  approximated  the  detention time observed  in the chlorine  contact
basin  at  the Dublin-San Ramon  facility.  After the  one-hour reaction  period,
samples were  taken in  the same  manner   as  for the  bench-scale  experiments
described earlier.  TOX analysis was by the Dohrmann method.
                                       95

-------
Results

     Results of two  experiments  using mixtures of chlorine and chlorine  diox-
ide are shown  in  Figs.  44 and 45 and are summarized in Table  35.   In  general,
both experiments  show a decrease in THM production as the proportion  of  chlo-
rine dioxide increases  (Fig. 44).  The same phenomenon is seen for  TOX in Fig.
45.   Because of  analytical  difficulties,  the TOX  data  for  Experiment 2 were
unreliable;  therefore,  TOX production values  are not shown  for  that  experi-
ment.   The decrease in THM and TOX  production  appears  to  require equal mass
doses of chlorine dioxide and  chlorine before these decreases  become  signifi-
cant.

     Figure 44 shows an  initial  increase in THM  production from the point with
no chlorine  dioxide  to  the point where the chlorine dioxide to chlorine  ratio
is 1:5.  Although  the increase is greater than two  standard deviations for THM
analysis, the general behavior of the rest of the data in Figs. 44  and 45 sug-
gest that this increase  is probably due to an operator error in sampling  or in
performing the THM analysis.

     If  the  mechanism  of THM decrease  is  by chlorine  dioxide reacting with
precursors of halogenated organics  to prevent THM production,   as Miltner pro-
posed  (98),  the  decrease in TOX production  would indicate that this  interac-
tion with precursors may prevent halogenation altogether and  not just prevent
THM  production.    However, there  still remains a possible  reaction  between
chlorine and chlorite as proposed by  Noack and  Doerr (97) and also a  possible
direct reaction between  chlorine and  chlorine dioxide noted by Miltner (98).

     Better understanding of the properties and  reactions of mixtures  of  chlo-
rine  and  chlorine  dioxide is  necessary.    Experiments  need  to  be conducted
using  longer reaction  times.   Non-nitrified  as  well  as nitrified  wastewaters
need to  be  studied.   Finally, more  investigations into the precise nature of
possible interactions  between chlorine and  chlorine  dioxide  in  terms of  the
species formed and their  disinfection capabilities  are required.
                                      96

-------
             THM    PRODUCTION
    %  CI02  0    10   20   30  40  50   60   70   80   90   100
    BY  WT.
    RATIO
    ClOg/C^
    BY  WT.
1=5
Figure 44.  THM production resulting from various mixtures of chlorine and
           chlorine dioxide  in Experiment 2 (Dublin-San Ramon wastewater),

                                  97

-------
              THM   PRODUCTION
      %CI02 0    10   20  30   40  50   60  70   80  90   100
      RATIO  CKCl   1^5             hi             5^1
              TOX   PRODUCTION
      RATIO
cio2/q
10  20   30  40   50  60   70  80   90  100
i   $5M              5M
Figure 45.  TOX and THM production resulting  from various mixtures of
           chlorine and chlorine dioxide in  Experiment 5
           (Dublin-San Ramon wastewater).
                                98

-------
                                  SECTION 11

           COST COMPARISON BETWEEN CHLORINE  DIOXIDE AND CHLORINE FOR
                            WASTEWATER DISINFECTION
BASIS OF COMPARISON

     An  economic  comparison is  presented between  disinfection of wastewater
with chlorine dioxide and with chlorine.  In the analysis five  treatment plant
sizes, 0.044,  0.22,  0.44,  2.2, and 4.4 m3/s  (1,  5,  10,  50, and 100 mgd),  are
considered.  A nominal  contact time  of 1  hour is assumed for the disinfection
process.   Chemical costs are  compared between chlorine  and chlorine  dioxide
generated from the chlorine-chlorite process.

     Three  levels  of disinfection are compared,  corresponding  to 2.2/100  ml,
200/100  ml,  and 1000/100  ml total coliforms.   The latter case (1000/100 ml
total coliforms was included in the analysis at the request of  the EPA  Project
Officer  [M. C. Meckes], who felt that it more nearly corresponds to the efflu-
ent  standard  imposed  for  secondary  effluent in  states such  as Ohio.    The
1000/100 ml total  coliform standard is approximately  equivalent to a  200/100
ml fecal coliform standard if  the ratio of total to fecal coliforms is  approx-
imately  5:1 as  found by Hunt  and Springer  (99).   The 1000/100 ml total coli-
form  standard also  is  that   reportedly  recommended  by  ORSANCO,  and widely
emulated in the U.S. (100).

     Two  levels  of  prior  treatment  are  considered:   conventional activated-
sludge treatment as typified by the Palo Alto 1978 laboratory experiments,  and
nitrified-filtered  activated-sludge  treatment as  practiced at  the Dublin-San
Ramon treatment plant.   It  is  believed that the conventional activated-sludge
effluent  represents  the majority of  biologically treated wastewater effluents
in  the   U.S.A.,  whereas the  nitrified,  filtered  effluent corresponds  to  a
higher degree of treatment, such as might be practiced for  water reclamation.

     All costs, except  chemical costs, have been updated to third-quarter  1979
costs  by use  of  the  Environmental  Protection  Agency  Sewage  Treatment  Cost
Index  (101,102).    The  EPA-STP index  for third quarter  (September)  1979  was
337.8.   Chemical  costs are based on  competitive supplier  bids, f.o.b. docks,
January  1980  (103,104).  For  simplicity, the  comparisons  will be referred to
as cases A through F, summarized in Table 36.
                                      100

-------
  TABLE 36.  CASES FOR EVALUATION OF THE RELATIVE COSTS OF DISINFECTION WITH
                         CHLORINE AND CHLORINE DIOXIDE

    Type of EffluentDisinfection Standard
          to  Be                                    as Total Coliform Count
       Disinfected               Case                     (N/100 ml)
Conventional
Activated Sludge


Filtered, Nitrified
Activated Sludge


A
B
C
D
E
F
2.2
200
1000
2.2
200
1000
CAPITAL COSTS

     Capital  costs were  estimated  from generalized  correlations  (101,102).
These estimates were adjusted for inflation as described above.   Capital  costs
are amortized over 20 years at 11 percent.

Contact Basins
     The  common  current practice in  designing  disinfectant contact basins  is
to provide a  serpentine routing of flow  following mixing of the disinfectant
and  wastewater.    For the  purposes  of this  report,  costs  are estimated  for
basins which  generally provide a  flow routing length  to channel width ratio
between 20:1 and 25:1.  The Dublin disinfectant mixing and  contact  facility  is
typical of the design considered in this analysis  (see Fig.  10).

Chlorine Feed and Storage Facilities

     The feed used most widely in wastewater  treatment plants is chlorine  gas.
The  required  equipment  and  storage  facilities   are  well  known;  commercial
equipment is readily available.

     Feed facility costs include distribution panels, cylinder  chocks,  instal-
lation, manufacturers'  preparation of  shop  drawings,  installation check and
startup, and contractor's overhead  and profit.   Chlorinator costs  include one
standby  chlorinator   and  chlorine  evapotators  on  systems   having  a capacity
greater than  4,000 Ibs per  day.   Also included  are  miscellaneous piping and
valves, which typically  amount to 5-10% of  the installed chlorination equip-
ment costs.

     Storage  facility  costs   include  housing,  hoist equipment,   ventilation
equipment and safety monitors.


                                      101

-------
Chlorine Dioxide Generation and Feed Systems

     For the purpose of  this  analysis,  it is assumed that chlorine dioxide  is
generated by  the chlorine-chlorite  process.   The  stoichiometry measured for
this process in a full-scale plant has been reported in Section  5.  In  theory,
1.34 Ibs  of  pure sodium  chlorite  and 0.5  Ibs  of chlorine  react to give one
pound  of  chlorine dioxide.   However,  sodium chlorite is  normally  purchased
with a purity of 80%.  Therefore, assuming 100% conversion of NaClOo. and  100%
yield  of  CK^  based on  NaC102  reacte<* with  4%  excess  chlorine  feed (Section
5), 1.68 Ibs of  80%  sodium chlorite and 0.70 Ibs of chlorine are required per
pound  of  chlorine  dioxide generated.  This  basis is somewhat more optimistic
than generally assumed, but is justified according to the results in  Section 5
of this report.

     Generation  and  feed  systems  costs are  based  on the costs for a  sodium
chlorite mixing  and metering  system,  a  chlorine dioxide generator,  and the
appropriately sized chlorine feed system.  The sodium chlorite system consists
of a polyethylene day tank, a mixer for the day tank, and a  dual head metering
pump.   The chlorine  dioxide generator  is a  PVC tube filled  with  porcelain
Raschig Rings or other transport-enhancing media, and is sized for a  detention
time of 20 min.

Disinfectant Feed Capacity—
     The  sizing  of generating,  feed,  and storage  facilities  is based  on the
required  disinfectant  feed  capacity in  kg  or pounds  per   day.   (A detailed
example of this procedure appears in Appendix C.)

     In determining the feed capacity required, the procedure is as  follows:

     1.  Determine  the  initial  bacterial concentration in  the  waste
         stream to be disinfected (Table  14).
     2.  Calculate  the  required survival ratio,  N(t)/N(0), from the
         disinfection standard  N(t) and  initial  bacterial   concentra-
         tion, N(0).
     3.  Calculate  the  required residual-time  product using  Eqs.  24
         and 25 and the model fitting constants from Table 24.
     4.  Calculate the  required residual, based  on  the  provided con-
         tact time, namely one hour.
     5.  Calculate  the  disinfectant  dosage needed  to  provide the
         required residual disinfectant concentration

     A summary of the results from  steps  1-5  from above for  Cases A-F is given
in Table  37.   The quantities of disinfectant required per  day  shown in Table
38 were calculated from  the required dosages  given  in column 5b  of Table 37.
OPERATION AND MAINTENANCE COSTS—EXCLUDING CHEMICAL  COSTS

     Operation  and maintenance  (O&M)  costs  include electrical  power,  mainte-
nance material, and labor costs  (101,102).
                                      102

-------
           TABLE 37.  SUMMARY OF ESTIMATES OF REQUIRED DISINFECTANT DOSAGES USED  IN  COST  EVALUATION CASES
o
u>
(1)
Initial Bacterial
Concentration
„ * ., (-total coliform
Case log^ .
A 6
B 6
C 6
D 4
E 4
F 4
00 ml
.21
.21
.21
.41
.41
.41
(2) (3) (4) (5a)
Residual-Time Residual Required 1-hr Disinfec-
Product at 1-hr Contact tant Demand
Survival Ratio (mg-min/1) Time (mg/1) (mg/1)
) log(N(t)/N(0)) C12 C102 C12 C102 C12 C102
-5
-3
-3
-4
-2
—.1
.87
.91
.21
.07
.11
.41
376.58
90.11
54.01
210.73
13.02
4.82
164.45
34.79
19.95
22.74
1.96
0.82
6.28
1.50
0.90
3.51
0.22
0.08
2.74
0.58
0.33
0.38
0.03
0.01
1.61
0.95
0.80
7.64
2.39
2.52
5.18
2.32
1.84
5.14
0.57
0.13
(5b)
Required
Disinfectant
Dose (mg/1)
ci2 cio2
7.89
2.45
1.70
11.15
2.61
2.60
7.92
2.90
2.17
5.52
0.60
0.14
            See text for explanation of cases, procedure, and calculations.

-------
           TABLE 38.  AMOUNTS OF DISINFECTANTS REQUIRED  TO  ACHIEVE
                           TOTAL COLIFORM STANDARDS

                          Disinfectant Usage, pounds   per day

                              Treatment  Plant Size  - mgd'
               1             5             10            50
100
Case A
ci2
cio2
Case B
ci2
cio2
Case C
ci2
cio2
Case D
ci2
cio2
Case E
ci2
cio2
Case F
C17
cio2
66
66
20
24
18
14
93
46
22
5
22
1.2
329
330
102
121
90
70
465
230
110
25
110
6
658
661
204
242
180
140
930
460
220
50
220
12
3290
3303
1022
1209
900
700
4650
2302
1100
250
1100
60
6580
6605
2043
2419
1800
1400
9300
4604
2200
500
2200
120
      1 pound = 0.4536 kg.
     ^l mgd = 0.0438 m3/s.
     Electrical requirements  include power for  the  gaseous chlorination  sys-
tem, the  sodium chlorite metering  and  mixing systems,  and building heating,
lighting, and ventilation.

     Maintenance material requirements are  based on  experience with  gaseous
chlorine  systems  and liquid metering  systems.    Costs  for chemicals required
for residual monitoring are included.

     Labor requirements consist of labor for the  gaseous chlorination systems,
plus the  labor required  to mix the  sodium chlorite  solution,  to adjust  its
feed rate, and to maintain the mixing and metering equipment.
CHEMICAL COSTS

     All  chemical  costs are  f.o.b. works.   It  is  assumed  that  the 50-  and
100-mgd  plants  will  be  able  to  take advantage  of  the  lower chemical  costs
                                      104

-------
associated with  larger  quantity purchases.   Chemical  unit costs are  shown  in
Table 39.
    TABLE 39.  UNIT COSTS OF CHEMICALS REQUIRED IN WASTEWATER DISINFECTION

           Chemical                         Unit Cost, $ per pound

Chlorine - one ton cylinders                         0.155

Chlorine - rail cars                                 0.075

Sodium chlorite (80%), less than
  250,000 pounds annually                             2.67

Sodium chlorite (80%), more than
   250,000 pounds annually                            1.98

      Prices are f .o.b. works; 1 pound - 0.4536 kg.
      Source:  References (103,104).


     The price  given  for sodium chlorite  (Table  39) and used in  this  evalua-
tion is based on manufacturers'  quotations for a solution of 25%  sodium chlo-
rite content.   The  equivalent price  is  $4.36/kg  or  $5.88/kg ($1.98  per pound
or  $2.67  per  pound)  of  sodium chlorite  content,  depending on the quantity
purchased.   The price  of dry sodium chlorite is considerably lower:  $2.67/kg
($1.21  per  pound)  of  sodium  chlorite  if  purchased  in drum  quantities  at
January  1980 prices  (105).   The  higher  price  quoted  by  manufacturers  who
supply sodium chlorite  to the water  industry  has  been chosen to  be  conserva-
tive.  However, it should be recognized that sodium  chlorite possibly could be
purchased for approximately 45 to 60% of the values  given in Table 39.
COST SUMMARIES

     The cost  of  wastewater disinfection are  summarized in Tables 40  through
45,  for  Cases A-F,  respectively.   Disinfection with  chlorine  dioxide is  at
least as  expensive as,  and  generally more  expensive  than, disinfection  with
chlorine under the assumptions embodied in the cases compared here.   The rela-
tive cost  factor  for disinfection with chlorine dioxide compared to chlorine
ranges from a high of 15 to one (Case A, 4.4 m Is  [1QO  mgd]  plant) to a low  of
1.0  to  one (Case  F,  0.044 m3/s  [1-mgd]  and 0.22 m3/s [5-mgd]  plants).  The
predominant factor influencing the higher cost  of chlorine dioxide  disinfec-
tion is the cost of the  sodium chlorite.

     Case  C  (conventional  activated-sludge  effluent,  1000/100  ml total coli-
form standard)  is believed  to  represent  most nearly  among the cases  studied
the typical situation of disinfecting secondary  effluent in the U.S.   In  Case
C, the cost  of disinfection with  chlorine  dioxide ranges from 2 times as ex-
pensive  (at  0.044 m /s  [1 mgd])  to  5 times  as expensive  (at  4.4  m  /s  [100
mgd]) compared to  the cost of disinfection with  chlorine.
                                      105

-------
TABLE 40.  DISINFECTION COST SUMMARY FOR CASE A




Annual
Cost
, Thousand $
Treatment Plant Size

Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
ci2
NaC102
Total Cost
$1000/yr
(t/1000 gal)

ci2
14.7

1.7
7.4

3.7
-

27.5
(7.5)
1
cio2
14.7

6.1
9.9

2.6
108.1

141.4
(38.7)

ci2
36.0

4.0
10.5

18.6
-

69.1
(3.8)
5
cio2
36.0

10.0
18.9

13.0
540.3

618.2
(33.9)


ci2
54

6
13

37
-

Ill
(3
.8

.3
.2

.2


.5
•1)
10
cio2
54.8

12.7
22.5

26.0
1082.2

1198.2
(32.8)
per Year
, mgd

ci2
172.3

18.0
17.1

90.1
-

297.5
(1.6)

50
cio2
172.3

25.9
55.0

63.3
4010.3

4326.8
(23.7)


ci2
313.0

29.7
18.9

180.1
-

541.7
(1.5)

100
cio2
313.0

35.5
81.0

126.6
8019.4

8575.5
(23.5)

-------
TABLE 41.  DISINFECTION COST SUMMARY FOR CASE B




Annual
Cost, Thousand $
Treatment Plant Size

Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
ci2
NaC102
Total Cost
$1000/yr
(t/1000 gal)

ci2
14.7
1.6
6.1

1.1
23.5
(6.4)
1
cio2
14.7
4.5
7.8

0.9
39.3
67.2
(18.4)

ci2
36.0
2.0
7.9

5.8
51.7
(2.8)
5
cio2
36.0
7.1
12.2

4.7
198.1
258.1
(14.1)

ci2
54.8
3.0
9.0

11.5
81.3
(2.2)
10
cio2
54.8
8.7
17.1

9.4
396.2
486.2
(13.3)
per Year
> mgd

ci2
172.3
8.1
14.8

28.0
223.2
(1.2)

50
cio2
172.3
16.6
30.0

23.1
1467.9
1709.9
(9.4)




ci2
313
14
16

55
399
(1
.0
.1
.1

.9
.1
•1)

100
cio2
313.0
21.1
43.0

46.2
2937.0
3360.3
(9.2)

-------
                                    TABLE 42.  DISINFECTION COST SUMMARY FOR CASE  C
o
oo




Annual
Cost, Thousand $
Treatment Plant Size

Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
C12
NaC102
Total Cost
$1000/yr
(t/1000 gal)

ci2
14.7
1.8
6.0

1.0

23.5
(6.4)
1
cio2
14.7
3.7
7.1

0.6
22.9

49.0
(13.4)

ci2
36.0
3.5
7.9

5.1

52.5
(2.9)
5
cio2
36.0
6.4
10.0

2.8
114.6

169.8
(9.3)

ci2
54.8
4.6
8.6

10.2

78.2
(2.1)
10
cio2
54.8
9.4
12.2

5.5
229.2

311.1
(8.5)
per Year
> mgd

ci2
172.3
10.8
15.5

24.8

223.4
(1.2)

50
cio2
172.3
20.5
25.0

13.5
849.9

1081.2
(5.9)


ci2
313.0
17.7
23.0

49.6

403.3
(1.1)

100


cio2
313
27
33

27
1699

2100
(5
.0
.3
.0

.0
.8

.1
.8)

-------
                                    TABLE 43.   DISINFECTION COST SUMMARY FOR CASE D
o
VO




Annual
Cost, Thousand $
Treatment Plant Size

Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
ci2
NaC102
Total Cost
$1000/yr
(t/1000 gal)

ci2
14.7
1.80
7.8

5.3
29.6
(8.1)
1
cio2
14.7
5.5
9.0

1.8
75.3
106.3
(29.1)

ci2
36.0
5.0
11.2

26.3
78.5
(4.3)
5
cio2
36.0
8.7
16.5

9.1
376.6
446.9
(24.5)

C12
54.8
7.8
15.8

52.6
131.0
(3.6)
10
cio2
54.8
10.9
20.8

19.3
753.1
857.9
(23.5)
per Year
, mgd

ci2
172.3
21.9
17.9

127.3
339.4
(1.9)

50
cio2
172.3
20.5
45.0

44.1
2794.9
3076.8
(16.9)


ci2
313.0
36.0
19.8

254.6
623.4
(1.7)

100


cio2
313
28
68

88
5589
6087
(16
.0
.6
.0

.2
.9
.7
•7)

-------
TABLE 44.  DISINFECTION COST SUMMARY FOR CASE E





Annual
Cost, Thousand $
Treatment Plant Size

Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
ci2
NaCl02
Total Cost
$1000/yr
Ct/1000 gal)


ci2
14
1
7

1


25
(6
.7
.7
.5

.2
-

.1
.9)
1
cio2
14.7
2.6
7.4

0.2
8.2

33.1
(9.1)

ci2
36.0
4.1
10.8

6.2
—

57.1
(3.1)
5
cio2
36.0
4.5
7.8

1.0
40.9

90.2
(4.9)

ci2
54.8
6.6
13.6

12.4
-

86.4
(2.4)
10
cio2
54.8
5.5
9.0

2.0
81.9

153.2
(4.2)
per Year
, mgd

ci2
172.3
20.0
17.3

62.2
—

271.4
(1.5)


50
cio2
172.3
9.3
16.6

4.8
303.5

506.5
(2.8)




100
ci2
313.0
30.5
19.1

124.5
-

487.1
(1.3)
cio2
313.0
11.1
21.8

9.6
607.1

962.6
(2.6)

-------
TABLE 45.  DISINFECTION COST SUMMARY FOR CASE F
Annual Cost, Thousand $
Treatment Plant Size

Cost Item
Contact Basin
Generation and
Feed Equipment
0 & M Cost
Chemical Cost
ci2
NaC102
Total Cost
$1000/yr
(471000 gal)


C12
14.
2.
6.

1.
-

24.
(6.
7
2
2

2


3
7)
1
cio2
14.7
1.4
6.0

0.1
2.0

24.2
(6.6)
5
ci2
36.0
3.82
8.0

6.2
-

54.0
(3.0)
cio2
36.0
2.6
6.8

0.2
9.8

55.4
(3.0)
per Year
, mgd
10
ci2
54.8
5.1
9.1

12.4
-

81.4
(2.2)
cio2
54.8
3.5
7.1

0.5
19.6

85.5
(2.3)
ci2
172.3
11.9
16.3

30.3
-

230.8
(1.3)


50
cio2
172.3
6.1
9.5

1.2
72.8

261.9
(1.4)




100
ci2
313.0
21.8
26.2

60.6
-

421.6
(1.2)
cio2
313.0
7.8
12.0

2.3
145.7

480.8
(1.3)

-------
     The costs summarized  in  Tables  40 through 45 include the cost of contact
facilities.  The  economies of scale in  contact  basin construction are an  im-
portant  reason  why the  unit  cost of  disinfection  decreases  with increasing
plant size.  In our cost analysis, we  have not differentiated between contact
basin designs  for chlorine dioxide  and  chlorine.   Rather,  we have expressed
the differences between disinfectants as differences in required dose.  Hence,
the capital  costs of contact  basins are identical  for  the two disinfectants
for all cases and plant sizes.

     For large  plants,   chemical  usage will  be  sufficiently  large  to permit
purchasing at lower unit costs•   This cost  advantage is greater for chlorine
than for chlorine dioxide on a percentage basis.   Hence, the results of this
analysis show greater cost advantages for chlorine at large plants (50 and  100
mgd) than at smaller ones.

     Freight costs  have not  been included  in these cost  comparisons.   Were
freight costs to  be considered, the  cost advantage of chlorine might be less-
ened in  some  cases  (e.g., Cases E and F)  in view of the lesser quantities  of
chemicals that must be transportd when chlorine dioxide is used (Table 38).
DISCUSSION

     When  the  costs of  disinfection of municipal  wastewater effluents  using
chlorine dioxide and chlorine are compared as above  (using a  coliform  standard
as the disinfection criterion and assuming presently available technology  and
current market  prices),  chlorine  dioxide  generally is  not  cost-effective  in
the usual  sense.    However,  several factors should  be  considered to  put  this
conclusion into proper perspective.'  These are:   (1) the potential  for  reduc-
ing the costs  of  chlorine dioxide  generation,  and   (2) the  intangible  advan-
tages  of  chlorine  dioxide  that are  not  taken into account  in the "cost-to-
kill-coliforms" comparison.

Potential for Cost  Reduction

     The  possibility of  reducing  the  unit cost  of chlorine  dioxide  hinges
largely on  the potential  for reducing the  unit cost of  the sodium  chlorite
reactant, or for developing a process suitable for generating chlorine dioxide
from chlorate at  the wastewater treatment  plant site.   Investigation  of these
possibilities was beyond  the  scope of this work, and hence we  can only  specu-
late as to the probability of success in these directions.

     Sodium  chlorite is  produced  commercially  from the reduction of  chlorine
dioxide  (produced  from  chlorate)  by hydrogen  peroxide in  sodium hydroxide
solutions  (105).  With the increase of interest in  chlorine  dioxide for water
and wastewater  disinfection,  an  economic  incentive  exists  for increasing  the
efficiency of the current  commercial  production  techniques for  sodium  chlorate
or for developing a new technology for  sodium chlorite production  from chlo-
rate.  It  is conceivable  that  an expansion of the market for  sodium  chlorite
in water  and wastewater  treatment  could result  in  a relative decrease  in  the
unit cost.    This  could occur  as  a result of  economics  of scale  in  production
                                      112

-------
and marketing, as well  as  by virtue of increased competition.  Presently,  one
supplier dominates the market.

     There is  a  strong  motivation  for  developing  a feasible process to  gene-
rate chlorine  dioxide  from chlorate at the  wastewater treatment site.   Based
on raw material  costs,  the unit cost  of  chlorine  dioxide could be reduced  by
approximately  half  if  chlorate rather  than chlorite  were used.   Processes
based on  chlorate  are  widely used  in  the pulp and paper industry where  large
quantities of chlorine dioxide are  required.  However,  there are reputed  to  be
substantial problems of smaller-scale reactor instability,  resulting occasion-
ally  in  minor  explosions.   From  the perspective  of the  required operator
training,  such a generation  facility  probably  would  not  be  feasible in  the
context of wastewater treatment.  Hence,  the wide-spread use of chlorate-based
generation technology would require  further  process development,  the evalua-
tion of which is beyond the scope of our work.

     There is  no appreciable  potential  for  improving the  yield  of chlorine
dioxide from chlorite.   In the  simple,  commercially available reactor studied
in this work, the chlorine dioxide  yield  closely approximated that correspond-
ing to accepted reaction stoichiometry.

Intangible Advantages of Chlorine Dioxide over Chlorine

     The  additional  advantages  of  chlorine  dioxide compared to chlorine  that
were not  considered  quantitatively  in  the cost analysis  in this  section  are
negligible amounts of halogenated  byproducts and superior virus inactivation.
These advantages of chlorine dioxide compared to chlorine have been demonstra-
ted  experimentally  in  this  work.    The superiority  of chlorine dioxide  with
respect to virus  inactivation can  be concluded  only tentatively from the  re-
sults of  this  work;  a program of experimental verification is necessary.   The
advantage of chlorine dioxide over  chlorine with respect to formation of  halo-
genated byproducts  is  unequivocal.   It should weigh heavily in the ultimate
decision as  to whether  to employ chlorine  dioxide  as an alternative to  chlo-
rine, in  view  of  the concern over  halogenated  and other hazardous pollutants
in the nation's water supplies.
                                      113

-------
                                 REFERENCES
 1.   APHA-AWWA-WPCF,  Standard Methods for the Examination of Water and Waste-
     water,  14th ed.,  American Public Health  Association,  Washington, D.C.,
     1976.

 2.   Kott,  Y.   Estimation  of Low Numbers of _E_.  coli Bacteriophage by Use of
     the MPN Method.   Appl. Microbiol., 14(2):141, 1966.

 3.   Berg,  J., E.  M.  Aieta,  and P. V.  Roberts.  Effectiveness  of Chlorine
     Dioxide as  a  Wastewater  Disinfectant.   In:   Proceedings  of  National
     Symposium on Wastewater Disinfection, A. Venosa, ed.  U.S. Environmental
     Protection Agency, Cincinnati,  Ohio, 1978.

 4.   Atlantic   Research Corp.   Phage  Testing and  Evalutiori  Program, Final
     Report.   Alexandria,  Virginia,  Oct., 1978.

 5.   Henderson, J.  E.,  G.  R. Peyton, and W. H.  Glaze.   A Convenient  Liquid-
     Liquid  Extraction Method  for the  Determination  of  Halomethanes in Water
     at the Parts-per-Billion Level.   In:    Identification and  Analysis of
     Organic Pollutants in Water, L.H.  Keith, ed.  Ann Arbor Science Publish-
     ers,  Inc., Ann Arbor,  Michigan,  pp. 105-112.

 6.   Palin,  A. T.    Methods  for  the  Determination,  in  Water,  of  Free  and
     Combined   Available Chlorine,  Chlorine Dioxide and  Chlorite,  Bromine,
     Iodine  and  Ozone, Using  Diethyl-p-phenylene Diamine  (DPD).   J. Inst.
     Water Engr., 21:537,  1967.

 7.   Palin,  A.  T.    Analytical  Control  of  Water Disinfection  with  Special
     Reference to  Differential DPD  Methods for  Chlorine,  Chlorine Dioxide,
     Bromine,  Iodine and Ozone.  J.  Inst. Water Engr., 28:139-154, 1974.

 8.   Miltner,  R.  J.   Measurement of  Chlorine  Dioxide  and  Related Products.
     In:  Proceedings  AWWA Water Quality Technology Conf., Dec.  6-7, 1976,
     San Diego, California.  Paper No.  2A-5; AWWA, Denver, Colorado, 1977.

 9.   Granstrom, M.,  and G. Lee.   Generation and Use of Chlorine Dioxide in
     Water Treatment.  J.  AWWA, 50:1453,  1958.

10.   Ingols, R.,  and G. Ridenour.  Chemical Properties of Chlorine Dioxide in
     Water Treatment.  J.  AWWA, 40:1207,  1948.

11.   White,  G.   Handbook  of Chlorination.   Van  Nostrand Reinhold, New York,
     1972.

                                     114

-------
12.  Gordon,  G.,  R.  G.  Kieffer,  and D.  H.  Rosenblatt.   The  Chemistry  of
     Chlorine Dioxide.  In:  Progress in Inorganic Chemistry, Vol.  15.,  S.  J.
     Lippard, ed.  Wiley-Interscience, New York, 1972.

13.  Masschelein,  W.   J.    Chlorine  Dioxide:    Chemistry  and  Environmental
     Impact  of  Oxychlorine  Compounds.    Ann  Arbor  Science,  Ann  Arbor,
     Michigan, 1979.

14.  Chick,  H.   An Investigation  of the Laws of  Disinfection.   J. Hygiene,
     8:92, 1908.

15.  McKee,  J.  E., et al.   Report  on  the  Disinfection  of  Settled  Sewage.
     California Institute of Technology, Sanitary Engineering Research,  1957.

16.  Collins,  H.  F.,  and  R.  E.  Selleck.    Process Kinetics  of  Wastewater
     Chlorination.  SERL  Report No.  72-5, University of California,  College
     of Engineering and School of Public Health, Berkeley,  California, 1972.

17.  Wei,  J.  H.,  and  S.  L. Cheng.   A Multi-Poisson Distribution Model  for
     Treating Disinfection Data.  In:  Disinfection—Water  and Wastewater,  J.
     D. Johnson, ed.  Ann Arbor Science, Ann Arbor, Michigan, 1975.

18.  Collins, H. F.,  et al.   Problems in Obtaining Adequate Sewage Disinfec-
     tion.   In:   Proceedings of the  National  Specialty Conference  on Disin-
     fection.  American Society of Civil Engineers,  1970.

19.  Hall, E. S.  Quantitative  Estimation  of  Disinfection  Interferences.   J.
     of Soc. for Water Treatment and  Examination, 22(3):153, 1973.

20.  Wyss, 0.  Theoretical Aspects of Disinfection by Chlorine—Public Health
     Hazards  of  Microbial Pollution  of Water.   Proc. Rudolphs  Res.  Conf.,
     Rutgers University, New Brunswick, New Jersey, 1961.

21.  Fair, G. M., and H. Thomas.  Natural Purification of River Muds and Pol-
     lutional Sediments.  Sewage Works J.,  13(270):756, 1941.

22.  Fair,  G.  M.,  and  J.  C. Geyer.   Water Supply  and Wastewater  Disposal.
     Wiley, New York,  1954.

23.  Phelps, E. B.  Stream Sanitation.  Wiley, New York, 1944.

24.  Card,  S.   Chemical .Inactivation of Viruses.    In:   Nature  of Viruses.
     CIBA Foundation Symposium, 1957.  pp.  123-146.

25.  Watson, H. E.  A Note  on  the  Variation of the Rate of Disinfection with
     Change  in the Concentration of  the Disinfectant.  J.  of Hygiene, 8:536,
     1908.

26.  Fair, G. M.,  et  al.   The  Behavior of  Chlorine as a Water Disinfectant.
     J. AWWA, 40:1051, 1948.
                                     115

-------
27.  Selleck, R. E., B. M. Saunier, and H. F. Collins.  Kinetics of Bacterial
     Deactivation with Chlorine.   J.  Env.  Eng. Div.,  ASCE,  104(EE6):1197-
     1210, 1978.

28.  Levenspiel, 0.    Chemical  Reaction Engineering,  2nd  ed.   John  Wiley  &
     Sons, New York, 1972.

29.  Sepp, E.   Tracer  Evaluation of Chlorine Contact  Tanks.  State of Cali-
     fornia,  Health and Welfare  Agency,  Department  of Health, Sanitary Engi-
     neering Section, Berkeley, California, 1977.

30.  Yao, K.  M.,  M. T. Habibian,  and C.  R.  O'Melia.   Water  and Wastewater
     Filtration:   Concepts  and Applications.    Environ.  Sci.  Tech., 5:11,
     1971.

31.  Buchan,  R. E.,  and N. E. Gibbons, eds.  Bergey's Manual of Determinative
     Bacteriology,  Williams and Williams Co., Baltimore, Maryland, 1974.

32.  Saunier, B., and R. E. Selleck.  Kinetics of Breakpoint Chlorination  and
     of Disinfection.  Report  No.  76-2,  University  of California, College of
     Engineering and School of Public Health, Berkeley, California, 1976.

33.  Feben,   D.,  and  M.   J.  Taras.   Chlorine  Demand  Constants.    J. AWWA,
     43:922,  1951.

34.  Berndt,  H., and H. J. Linneweh.  Comparative Quantitative Studies on  the
     Bactericide Activity of  Chlorine  Dioxide  and  Chlorine.    Arch. Hyg.
     Bakteriol., 153:41, 1969.  In:  Chem. Abs., 71:19528e,  1969.

35.  Ridenour, G. M., R. S. Ingols, and E. H. Armbruster.  Sporicidal  Proper-
     ties of Chlorine Dioxide.  Water and  Sewage Works, 96(8):279, 1949.

36.  Ridenour, G. M.,  and  R.  S.  Ingols.   Bactericidal Properties  of Chlorine
     Dioxide.  J. AWWA, 39:561, 1947.

37.  Walters,  G. E.    Chlorine  Dioxide  and Chlorine:   Comparative Disinfec-
     tion.  M.S. Thesis, Johns-Hopkins University, Baltimore, Maryland, 1976.

38.  Benarde,  M.  A.,  B.  M.  Israel,  V.  P.  Olivieri,  and  M.  L.  Granstrom.
     Efficiency  of  Chlorine  Dioxide as  a  Bactericide.    Appl.   Microbiol.,
     13(5):776, 1965.

39.  Benarde,  M. A.,  W.  B. Snow,  V.  P.  Olivieri,  and B. Davidson.  Kinetics
     and  Mechanism  of  Bacterial  Disinfection  by  Chlorine Dioxide.   Appl.
     Microbiol., 15(2):257, 1967.

40.  Longley,  K.E.,  B. E. Moore, and C. A. Sorber.   Comprison of Chlorine  and
     Chlorine  Dioxide  as  Wastewater Disinfectants.   Presented at  Water Poll.
     Control Fed. Annual Meeting,  Anaheim, California, Oct.  1-6,  1978.
                                     116

-------
41.  Aieta, E. A., J. D. Berg, P. V. Roberts, and R.C. Cooper.   Comparison of
     Chlorine Dioxide and  Chlorine  in the  Disinfection  of Wastewater.   Pre-
     sented at Water Poll.  Control  Fed.  Annual Meeting, Anaheim,  California,
     Oct. 1-6, 1978.

42.  Adams, A. P.,  et al.   Bacterial Aerosols from Cooling Towers.   J.  WPCF,
     50:2362, 1978.

43.  Moffa, P.  E.,  and J.  E.  Smith.  Bench-Scale High-Rate Disinfection of
     Combined Sewer Overflows with Chlorine and Chlorine Dioxide.  EPA-670/2-
     75-021,  U.S.  Environmental  Protection Agency,  Cincinnati, Ohio,  April
     1975.

44.  Englebrecht, R. S., and  E.  0.  Greening.  Chlorine-Resistant  Indicators.
     In:   Indicators of Viruses  in  Water and Food,  G.  Berg.  ed.   Ann  Arbor
     Science Publishers, Inc., Ann Arbor, Michigan, 1978.

45.  Geldreich,  E.   Bacterial  Populations  and Indicator  Concepts in Feces,
     Sewage,  Stormwater and  Solid  Wastes.   In:   Indicators  of  Viruses  in
     Water  and  Food,  G. Berg.  ed.   Ann  Arbor  Science Publishers, Inc.,  Ann
     Arbor, Michigan, 1978.

46.  Hackney, C.  R.,  and  G. K.  Bissonnette.   Recovery of Indicator  Bacteria
     in Acid Mine Streams.   J. WPCF, 50:775, 1978.

47.  Matson,  E.  A., S.  G.  Horner,  and J.D.  Buck.   Pollution  Indicators  and
     Other Microorganisms in River Sediment.  J. WPCF, 50:13, 1978.

48.  Berg,  G.,  et  al.   Validity of  Fecal Coliforms,  Total  Coliforms,  and
     Fecal  Streptococci as  Indicators  of  Viruses.   Appl. and Environ. Micro-
     biol., 36(6):880,  1978.

49.  Berg,  G., and  T. Metcalf.   Indicators  of Viruses in Waters.  In:   Indi-
     cators of  Viruses  in  Water and Food,   G.  Berg.  ed.   Ann  Arbor  Science
     Publishers, Inc., Ann Arbor, Michigan,  1978.

50.  Cronier, S., P. V. Scarpino, and  M.  L. Zink.  Chlorine Dioxide  Destruc-
     tion of Viruses and Bacteria in Water.  In:   Water Chlorination:   Envi-
     ronmental Impact and Health Effects, Vol. 2, R.  Jolley et  al., eds.   Ann
     Arbor  Science Publishers, Inc., Ann  Arbor, Michigan,  1978.

51.  Berg,  J.,  E. M. Aieta,  and P.  V.  Roberts.   . Effectiveness of  Chlorine
     Dioxide  as  a  Wastewater  Disinfectant.    In:   Progress  in Wastewater
     Technology.   Proceedings of the  National Symposium,  Cincinnati,  Ohio,
     September 1978, A. D.  Venosa,  ed.  EPA-600/9-79-018, U.S.  Environmental
     Protection Agency,  Cincinnati,  Ohio, June 1979.

52.  Snead, M. C., et al.   Biological Evaluation of Benefits of  Maintaining a
     Chlorine Residual in Water  Supply Systems.  Presented at American Chemi-
     cal Society Meeting,  Anaheim, California, March  1978.
                                     117

-------
53.  Kenyon, K. F.,  and  S.  A. Schaub.   Free Available Chlorine Disinfection
     of  Monodisperse  Enteric Organisms.    Presented  at  American Chemical
     Society Meeting, Anaheim, California, March 1978.

54.  Katzenelson,  E.,  and S.  Kedmu.  Unsuitability  of  Poliovirus as  Indica-
     tors  of Virological  Water  Quality.   Appl.  and  Environ.  Microbiol.,
     37(2):343,  Feb. 1979.

55.  Kott, Y., et  al.   Bacteriophages as  Viral  Pollution Indicators.  Water
     Res., 8:165,  1974.

56.  Goyal,  S.  M.,  C.  P.   Gerba,  and J.  L.  Melnick.    Prevalence of Human
     Enteric Viruses in Coastal Canal Communities.  J. WPCF, 50:2247,  1978.

57.  Goyal, S. M., et al.  Human Enteroviruses in Oysters and Their  Overlying
     Waters.  Appl. and Environ.  Microbiol., 37(3):572, March 1979.

58.  Vaughn, J. M.,  et al.    Survey  of Human Enterovirus Occurrence in Fresh
     and  Marine  Surface  Waters  on Long  Island.   Appl.  and Environ. Micro-
     biol., 38(2):290, 1979.

59.  Floyd, R., D.  G.  Sharp,  and J.  D.  Johnson.   Inactivation by Chlorine  of
     Single Poliovirus Particles in Water.   Environ.  Sci.  Tech.,  13(4):438,
     April 1979.

60.  Floyd, R., and D.  G.  Sharp.   Viral Aggregation:   Buffer Effects in the
     Aggregation of  Poliovirus and Reovirus  at  Low and  High  pH.   Appl. and
     Environ. Microbiol., 38(3):395, 1979.

61.  Sharp,  D.  G.,  et  al.    Virian  Aggregation and  the Kinetics  of Water
     Disinfection  by Chlorine and  Bromine.    Presented  at  American Chemical
     Society Meeting, Anaheim, California, March 1978.

62.  Stagg, C.H.,  C. Wallis,  and C. H. Ward.   Inactivation of Clay-Associated
     Bacteriophage  MS-2  by   Chlorine.     Appl.  and  Environ.  Microbiol.,
     33(2):385, Feb. 1977.

63.  Hejkal, T. W.,  F. M. Wellings,  P.  A. LaRock, and A. L. Lewis.  Survival
     of  Poliovirus Within  Organic  Solids During  Chlorination.    Appl.  and
     Environ. Microbiol., 38(1):114, 1979.

64.  Young, D. C., and D. G.  Sharp.  Partial Reactivation of Chlorine-Treated
     Echovirus.  Appl. and Environ. Microbiol.,  37(4):766, April 1979.

65.  Scarpino, P.   V.,  et   al.    Effect  of  Particulates on  Disinfection  of
     Enteroviruses  in Water  by  Chlorine Dioxide.    EPA-600/2-79-054,  U.S.
     Environmental Protection Agency, Cincinnati, Ohio, July 1979.

66.  Buxton,  G.   V.,  and  S.   A.  Ross.    Wastewater  Disinfection—Toward  a
     National Policy.  J. WPCF,  51(8):2033, 1979.
                                     118

-------
67.  Kinney, E. C.,  D.  W.  Drummond,  and N.B. Hanes.  Effects  of  Chlorination
     on Differentiated Coliform Groups.  J. WPCF,  50:2307,  1978.

68.  Cabelli, V.  J.   Relationship of  Microbial  Indicators to Health  Effects
     of Marine  Bathing Beaches.   Amer. Jour.  Public Health, 69(7):690-696,
     1979.

69.  Davis,  E.  M.,  et  al.   Bacterial Relationships  in Storm Waters.   Water
     Res.Bull., 13(5):895, Oct. 1977.

70.  Eliassen,  R.   Coliform Aftergrowth in  Chlorinated Storm Overflows.   J.
     San. Eng. Div., ASCE, 94(2):371,  1968.

71.  Kittrell,  F. W.,  and S. A.  Furfuri.   Observations of Coliform  Bacteria
     in Streams.  J. WPCF, 35(11):1361, 1963.

72.  Mitchell,  R.,   and C.  Chamberlin.    Survival  of  Indicator  Organisms.
     In:   Indicators of Viruses  in  Water  and Food,  G.  Berg.  ed.  Ann  Arbor
     Science Publishers, Inc., Ann Arbor, Michigan, 1978.

73.  Grabow, W. 0. K., 0. W. Prozesky, and J. S. Burger.   Behavior  in a  River
     and Dam  of Coliform  Bacteria with Transferable  or Non-Transferable Drug
     Resistance.  Water Res., 9:777, 1975.

74.  Gordon, R. W.,  and C. B. Fliermans.   Survival and Viability  of _E_.  coli
     in Thermally Altered Reservoir.   Water Res.,  12:434,  1978.

75.  Deaner, D. G.,  and K.  D.  Kerri.  Regrowth of  Fecal Coliforms.  J.  AWWA,
     61:465, 1969.

76.  Camper, A. K.,  and G.  H.  McFeters.   Chlorine  Injury  and  the Enumeration
     of  Waterborne  Coliform  Bacteria.    Appl.   and  Environ.  Microbiol.,
     37(3):633, 1979.

77.  Hendricks, C. W.,and  S.  M.  Morrison.   Multiplication and Growth of Se-
     lected  Enteric  Bacteria in  Clear Mountain  Stream Water.   Water  Res.,
     1:567, 1967.

78.  Smith, P. R., E. Farrell, and K.  Dunican.  Survival of R  j^. coli in Sea
     Water.  Appl. Microbiol., 27(5):983, 1974.

79.  Verstraete, W., and J.  P. Voels.   Comparative Study  of _E_. coli  Survival
     in Two Aquatic Ecosystems.  Water Res.,  10:129,  1976.

80.  Cooke, E. M.  Escherichia coli  and Man.  Churchill Livingstone,  London,
     1974.

81.  Welch, W.  J.,  and J.  S. Lee.   Modeling Techniques for Estimating  Fecal
     Coliforms in Estuaries.  J.  WPCF, 50:862, 1978.
                                     119

-------
82.  Standridge, J. H., et al.   Effects of Waterfowl (Anas platiphynehas)  on
     Indicator  Bacteria  Populations  in  a  Recreational  Lake  in  Madison,
     Wisconsin.  Appl. and Environ. Microbiol., 38(3):547,  1979.

83.  Seidler, R. J.,  J. E. Morrow,  and S.  T. Bagley.  Klebsiella in Drinking
     Water  Emanating  from  Redwood  Tanks.    Appl.  and  Environ. Microbiol.,
     33(4):893, April 1977.

84.  Knittel,  M.  D.,  et  al.    Colonization of the  Botanical Environment  by
     Klebsiella  Isolates  of  Pathogenic Origin.   Appl.  and  Environ.  Micro-
     biol., 34(5):557, Nov. 1977.

85.  Seidler, R.J.  Coliforms in Drinking Water Emanating from  Redwood Reser-
     voirs.     EPA-600/2-79-049,   U.S.  Environmental   Protection  Agency,
     Cincinnati, Ohio, July 1979.

86.  Brown, C.,  and R.  J. Seidler.   Potential Pathogens in  the Environment:
     Klebsiella pneumoniae, a Taxonomlc and Ecological  Enigma.  Appl. Micro-
     biol., 25(3):900, June 1973.

87.  Katzenelson,  E.   Survival  of Viruses.   In:   Indicators  of Viruses  in
     Water  and  Food,  G.   Berg.  ed.   Ann Arbor  Science  Publishers,  Inc., Ann
     Arbor, Michigan,  1978.

88.  Bitten,  G.   Survival of Enteric Viruses.  In:   Water  Pollution Micro-
     biology, Vol. 2,  R. Mitchell, ed.  John Wiley, New York, 1978.

89.  Obrien,  R.  T.,  and  J.  S.  Newman.    Inactivation  of Polioviruses and
     Coxsachieviruses  in  Surface  Water.    Appl.  and  Environ. Microbiol.,
     33(2):334, 1977.

90.  Myhrstad, J.  A., and Samdal,  J. E.  Behavior and Determination of  Chlo-
     rine Dioxide.  J. AWWA, 61(4):205-208, 1969.

91.  Miller, W.  G., R. G. Rice, C. M. Robson, R. L.  Scullen, W. Kuhn, and  H.
     Wolf.   An  Assessment of  Ozone  an  Chlorine  Dioxide  Technologies for
     Treatment  of  Municipal  Water   Supplies,  EPA-60012-78-147,  Municipal
     Environmental Research  Laboratory, Office of  Research  and Development,
     U.S.  Environmental   Protection  Agency,  Cincinnati,  Ohio,  August  1978.
     pp. 225-290.

92.  Symons,  J.  M.  Ozone, Chlorine  Dioxide,  and  Chloramine as Alternatives
     to   Chlorine   for   Disinfection   of   Drinking  Water.     In:      Water
     Chlorination:    Environmental  Impact  and  Health  Effects, Vol.  2,  R.
     Jolley,  H.  Gorchev,  and   D.   H.  Hamilton,  eds.    Ann  Arbor   Science
     Publishers, Inc., Ann Arbor, Michigan,  1978.  pp.  555-560.

93.  Schnitzer,  M.  Humic Substances:   Chemistry and  Reactions.   In:   Soil
     Organic  Matter,  M.  Schnitzer and  S.  U.  Khan,  eds.  Elsevier  Scientific
     Publishing Company,  Amsterdam,  1978.   pp. 1-64.
                                     120

-------
 94.  Manka, J., R. Menachem,  Mandelbaum,  and A. Bortinger.  Characterization
      of Organics  in  Secondary  Effluents.   Environ.  Sci.  Tech., 12(8):1017-
      1020, November 1974.

 95.  Rook, J.  J.,  A.  A. Gras, B. G. Van der  Heyden,  and J. de Wee.  Bromide
      Oxidation and Organic Substitution in Water Treatment.  J. Environ.  Sci.
      and Health, A, 13(2):91-116, February 1978.

 96.  Jolley,  R. L.   Chlorine-Containing Organic  Constituents in Chlorinated
      Effluents.  J. WPCF, 47(3):601, 1975.

 97.  Noack, M. G., and  R. L.  Doerr.  Reactions of Chlorine, Chlorine Dioxide
      and Mixtures  Thereof  with Humic  Acid:   An Interim Report.   In:  Water
      Chlorination:   Environmental  Impact  and Health  Effects,  Vol.  2.   Ann
      Arbor Science Publishers, Inc., Ann Arbor, Michigan,  1978.  pp.  49-58.

 98.  Miltner,  R.  J.    The  Effect of  Chlorine Dioxide  on  Trihalomethanes  in
      Drinking  Water.    M.S.   Thesis,  Department  of  Civil  and Environmental
      Engineering,  University of Cincinnati, Cincinnati, Ohio, 1976.

 99.  Hunt, D.  A., and  J.  Springer.   Preliminary Report  on  a Comparison  of
      Total Coliform  and  Fecal  Coliform  Values   in  Shellfish  Growing  Area
      Waters.    In:  Proceedings 8th National  Shellfish Sanitation Workshop.
      U.S.  Dept. of Health, Education,  and  Welfare; Food and Drug Administra-
      tion, January 1974.  pp. 97-104.

100.  McKee, J. E., and H. W. Wolf.  Water Quality  Criteria.  California State
      Water Quality Control Board, Sacramento, California,  1963.  p. 119.

101.  Process  Design,  Performance and Economic  Analysis Handbook, Biological
      Wastewater Treatment  Process.    Gulp,  Wesner,  Gulp, El  Dorado Hills,
      California, 1977.

102.  Estimating Costs for Water Treatment as a Function of  Size and Treatment
      Plant  Efficiency.    EPA-600/2-78-182,  U.S.  Environmental  Protection
      Agency,  Cincinnati, Ohio, 1978.

103.  Estimate  from Olin Water Services, Overland  Park,  Kansas.   W. J. Ward,
      Director, Technology, April 1979.

104.  Estimate  from  Rio  Linda  Chemical  Company,  Rio  Linda,  California.
      J. Hicks, President, January 1980.

105.  Chemical Marketing Reporter, 7 January 1980.

106.  Gall, R.  Chlorine Dioxide.  An Overview of Its Preparation, Properties,
      and Uses.   In:    Ozone/Chlorine  Dioxide Oxidation Products  of  Organic
      Materials, Rice and Contruvo,  eds.   Ozone Press Internation, Cleveland,
      Ohio, 1978.
                                      121

-------
                                  APPENDIX A

                          COLIFORM INACTIVATION DATA
Table A-l.  Log[Surviving Bacteria] in Eight Replicate Experiments Using
            Chlorine Dioxide as Disinfectant

Table A-2.  Log[Surviving Bacteria] in Eight Replicate Experiments Using
            Chlorine as Disinfectant

Table A-3.  Full-Scale Experiments Conducted at Dublin-San Ramon Treatment
            Facility on Nitrified, Filtered Secondary Effluent Using  Chlorine
            Dioxide and Chlorine

Table A-4.  Results of Laboratory Experiments with Dublin Filtered Nitrified
            Activated-Sludge Effluent

Table A-5.  Results of Laboratory Experiments with Unfiltered and Filtered
            Palo Alto Secondary Effluent Using Chlorine Dioxide

Table A-6.  Results of Laboratory Experiments with Unfiltered and Filtered
            Palo Alto Secondary Effluent Using Chlorine

Table A-7.  Results of Experiments with San Jose Secondary Effluent

Table A-8.  Results of Experiments with San Jose Nitrified Effluent

Table A-9.  Results of Experiments with San Jose Filtered, Nitrified  Effluent
                                      122

-------
                         TABLE A-l.  LOG[SURVIVING BACTERIA] IN EIGHT REPLICATE EXPERIMENTS

                                        USING CHLORINE DIOXIDE AS  DISINFECTANT
               DOSE :
        EXPERIMENT:
N>
CO
               0 MIN
               5 MIN
I-
o
           O  15 MIN
           O
              30 MIN
                      2mg/l
5 mg/l
10 mg/l
1|2|3|4|5|6|7|8
5.91 6.05 6.09 6.73 6.79 6.41 6.48 5.85
5.90 6.18 6.45 6.67 5.91
5.90
5.70
508 4.41 434 5.00 499 521 5.08 503
5 95 4 36 4 50 4 97 4 99 5 02 4 94 5 04
4.30 5.08 5.11 5.34 5.41 5.30
5.00 5.00 5.00 5.00 5.00
5.26 4.38 4.38 5.06 5.25 5.17 5.20 493
5.00 4.38 4.26 4.95 5.44 5.00 5.09 4.61
4.48 4.30 4.78 5.15 5.48 5.38 4.60
5.04 5.00 5.15 5.26 4.78
4.00 4.20 3.60 4.78 4.81 5.09 5.15 4.91
4.78 4.32 3.04 4.81 4.90 4.92 5.05 5.08
4.48 4.30 4.90 5.26 5.08 5.45 4.11
4.30 5.00 5.08 5.00 5.36
1
5.83
5.60
5.00

399
3 91


3.23
3.20
3.00

3.33
3.37
3.11

2
60?
6.28


2.00



?no



1.78
1.90
1.90

3
6.26



278
3 30


1.70



1.20
1.30
1.30

4 I
6.83



3.76
390
3.60
3.30
2.38
2.27
2.30
3.00
2.89
3.06
3.11

5 I
6.76



345
375
4.15
4.04
2.89
3.15
2.90

3.15
3.15
3.50

6
6S?



341
330
3.00

in?
3.82
3.72
3.70
3.12
3.15
3.83

7 i 8
6.67 5.77
5.88
5.00

3.98 3 20
4 08 3 30
3.30 3.70
3.30
2.28 2.68
2.28 3.04
3.00 2.30
3.00
2.99 3.46
3.24 3.56
3.00 3.32

1 | 2 I 3 |
5.87 6.12 6.15
5.60 6.38 6.20


1 90 1.30 1 78
1 95 1 00 1 78
2.18 0.30 1.56
2.04
1.98 0.60 0.30
1.75 0.48 0.34
1.49

0.31 0.34 0.34
0.60
0.40

4 | 5 |
6.83 6.70



2.62 2 15
246 2 15
1.70

2.11 1.40
1.68
1.66

0.34 0.34
0.30 0.30


6 | 7
6.73 6.93



2.26 2.45
2 56 2 43
2.32 1.95
1.98
1.60 1.18
1.56 1.30
1.20 1.08

1.11 0.90
1.15


8
5.93
5.75


1.60
208
1.65
1.53
0.70
1.00
1.15

0.48




-------
                    TABLE A-2.  LOG[SURVIVING BACTERIA] IN EIGHT REPLICATE  EXPERIMENTS


                                      USING CHLORINE AS DISINFECTANT
      DOSE:
                   2mg/l
5 mg/l
10 mg/l
EXPERIMENT:
       0 MIN
       5 MIN
  H

  O
  o
  o
15 MIN
      30 MIN
1
5 .85
5.90


4.78
5.00


4.48



4.34



1213,4,5,6,7,8
6.09 6.00 6.43 6.81 6.63 6.62 5.89
6.28 6.18 6.74 5.90
5.78
5.48
5.75 5.95 6.16 6.13 6.30 6.34 5.37
5.30 5.70 6.48 6.15 6.28 5.40
5.60

4.70 4.92 5.34 4.72 5.45 5.41 4.60
4 62 4 79 5 25 4 46 5.62 5.55 4.58
4.30 4.90 5.72 4.78 5.68
4.30 4.84 5.53 4.48
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
1 I 2
5.78 6.14
5.66 6.32
5.30

4.09 3.83
4.08 3.90
3.60 4.08
3.95
2.84 2.28
3.15 2.00
3.48 3.30

2.71 1.88
2.48 1.90
1.30

1 3
6.10
6.36


3.96
4.08
3.78
3.70
2.36
3.23


1.76
1.30
2.15
2.00
,4,5
6.79 6.83



4.15 4.15
4.06 4.15
4.41 4.30
4.14 4.49
2.78 3.49
3.34 3.70
3.48 3.30
4.30
2.91 3.23
2.87 3.21
3.38 3.20

,6,7
6.67 6.63



4.53 3.90
4.26


4.26 3.85
4 79 3 75


2.86 3.30
2.89 3.11
2.84 3.00

1 8
5.83
6.30
5.60

3.86
4.06
4.08
3.95
3.01
340
3.34
3.60
2.34
2.73
2.86
2.60
1
5.92
6.00


791
2.81


??7
1 81
1.76
1.57
1.04
0.60
0.48

,2,3
6.30 6.08
5.84


2.51 2.34
2.26 2.40
2.16 2.42
2.07 2.09
1.00 1.74
1.41 1 76
1.20 1.50
1.04 1.48
0.60 1.15
0.60 0.34
0.54

I 4
6.70



3.77



1.34
1.30


1.53
1.30


I 5 | 6
6.84 6.67



281 3 24
2.69 3.08
2.42

2.04 2.24
1 91
1.75

0.30 0.95
0.70


1 7 | 8
6.60 5.73
5.75
5.30

3 24 3 00
2.67


2.22 2.04
1 79 1 66
1.91 1.41

1.23 0.48
1.26 0.34
0.48


-------
      TABLE A-3.  FULL-SCALE EXPERIMENTS CONDUCTED AT DUBLIN-SAN RAMON
     TREATMENT FACILITY ON NITRIFIED, FILTERED SECONDARY EFFLUENT USING
                        CHLORINE  DIOXIDE  AND CHLORINE


Date
1/16/79




2/6/79




2/13/79

2/13/79


Repli-
Disin- Dose cate
fectant mg/1 No.
ClOo 4.76 1
2
3
4
5
6
C102 3.12 1
2
3
4
5
6
C102 2.25 1
2
3
Cl, 4.60 1
2
3
Average


0
4.28
4.38
4.30
4.30
4.45
4.48
3.82
3.90
3.79
3.87
4.04
4.93
4.70
4.75
4.99
4.84
4.69
4.59
Log Surviving Total
log10[N/100ml]
Contact

8
<0.30
0.78
<0.30
<0.30
<0.30
<0.30
0.30
<0.30
0.30
0.30
<0.30
0.30
2.41
2.72
2.33
2.70
2.72
2.51
Time , min

24
<0.30
<0.30
0.30
0.30
<0.30
0.30
<0.30
0.30
0.30
0.30
0.30
<0.30
2.41
2.27
2.36
2.45
2.40
2.61
Coliforms


48
<0.30
0.30
<0.30
<0.30
0.30
0.30
0.30
<0.30
0.30
0.30
0.30
<0.30
2.28
2.22
2.39
2.24
2.36
2.34
2/13/79
Cl.
19.60
4.83
2.08
1.04
0.30
                                     125

-------
TABLE A-4.  RESULTS  OF  LABORATORY  EXPERIMENTS  WITH DUBLIN FILTERED
               NITRIFIED, ACTIVATED-SLUDGE  EFFLUENT

Disin- Dose
fectant mg/1
C102 2

5


10


C12 2

5


10


Repli-
cate
No.*
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Average

0
3.74
4.08
5.18
3.78
4.20
4.90
3.93
4.08
4.99
6.07
4.24
4.98
3.93
4.54
4.95
3.88
4.43
4.86
Log Surviving Total
log10[N/100ml]
Contact
5
<0.30
4.01
4.38
<0.30
2.37
2.06
<0.30
0.30
0.88
3.30
3.84
3.85
_t
2.69
2.08
1.65
2.02
1.37
Time, min
15
<0.30
4.20
4.43
<0.30
2.39
2.09
0.95
<0.30
0.98
3.65
3.87
3.64
2.23
2.76
1.90
1.38
0.30
0.26
Coliforms

30
<0.30
4.18
4.13
1.30
2.35
2.26
<0.30
<0.30
0.34
3.30
3.68
3.74
1.60
2.26
2.62
1.10
<0.30
0.79
 Dates  of  replicate  experiments:  1,  1/9/79;  2,  2/16/79;  3,  2/22/79.
'-  signifies  sample  lost.
                                126

-------
 TABLE A-5.  RESULTS OF LABORATORY EXPERIMENTS WITH UNFILTERED AND
    FILTERED PALO ALTO  SECONDARY EFFLUENT USING CHLORINE DIOXIDE
Average Log Surviving Total Coliforms
log10[N/100ml]
Type
of Dose
Effluent mg/1
Unfiltered 2



5



10



Filtered 2



5



10



Repli-
cate
No.*
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4

0
5.80
5.06
5.02
5.55
5.84
5.24
5.29
5.72
5.24
5.26
5.26
5.31
5.82
5.00
4.82
5.07
6.22
5.01
4.72
5.31
6.68
4.58
4.86
5.15
Contact
1 5
4.35 4.32
4.40 4.30
4.50 4.36
4.80 4.88
<2.0
4.62
3.14
3.34
1.00
1.84
1.59
2.54
3.04 3.35
3.48 3.28
3.09 2.74
3.98 3.19
<1.0
0.78
0.30
1.64
<1.0
1.00
<0.4
<0.4
Time,
15
4.82
4.13
4.43
4.82
<2.0
2.56
2.46
2.52
<0.4
1.30
0.59
<0.4
3.09
3.46
2.86
3.50
<1.0
0.60
<0.4
0.98
<1.0
<0.4
<0.4
<0.4
min
30
4.04
3.86
4.22
4.60
<2.0
2.52
1.84
2.35
<0.4
<0.4
<0.4
<0.4
3.00
3.22
2.74
2.65
1.0
<0.3
0.3
0.79
<1.0
<0.4
<0.4
<0.4

60
_t
-
-
—
<2.0
2.5
1.08
2.84
—
-
-
—
—
-
-
—
<1.0
<0.3
0.6
0.98
_
-
-
—
 Dates  of  experiments:   1,  8/8/79;  2,  8/14/79;  3,  8/17/79;  4, 8/21/79,
'-  signifies  that  experiment was not conducted.
                                127

-------
 TABLE A-6.  RESULTS OF LABORATORY EXPERIMENTS WITH UNFILTERED  AND
        FILTERED PALO ALTO SECONDARY EFFLUENT USING CHLORINE
Average Log Surviving Total Colifonns
log10[N/100ml]
Type
of
Effluent
Unfiltered








Filtered








Repli-
Dose cate
mg/1 No.
2 1
2
3
5 1
2
3
10 1
2
3
2 1
2
3
5 1
2
3
10 1
2
3

0
4.69
4.92
4.95
4.82
4.78
5.04
4.96
5.23
5.38
5.78
4.50
4.65
4.81
4.74
4.77
4.66
4.89
4.89
Contact
1 5
4.80 4.39
4.48 4.67
5.10 4.48
3.19
2.95
3.34
0.79
1.03
2.72
4.64 4.23
5.09 4.40
4.57 4.42
1.53
1.86
1.64
0.39
0.83
1.47
Time,
15
3.04
3.52
2.04
1.17
1.00
1.40
<0.3
<0.3
1.46
2.69
3.23
3.50
0.67
1.06
1.38
0.70
<0.3
0.66
min
30
3.12
3.30
2.95
0.39
1.00
1.44
<0.3
0.80
2.20
1.30
1.47
2.70
<0.3
0.24
1.22
<0.3
0.15
0.26

60
_t
—
—
0.30
1.00
1.26
—
-
—
_
-
—
<0.3
0.15
0.30
—
-
—
*Dates of duplicate experiments:  1, 8/23/79; 2, 8/24/79; 3, 8/22/79.

 - signifies that experiment was not conducted.
                                 128

-------
TABLE A-7.  RESULTS OF EXPERIMENTS WITH SAN JOSE SECONDARY EFFLUENT

Disin- Dose
fectant mg/1
C102 2

5


10


C12 2

5


10


Repli-
cate
No.*
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Average

0
6.08
6.81
6.67
6.40
6.63
6.58
6.16
6.75
6.49
6.30
6.76
6.45
6.27
6.94
6.30
6.29
6.69
6.54
Log Surviving Total
log10[N/100ml]
Contact
5
5.11
3.10
5.71
3.64
4.57
4.56
1.48
2.60
2.59
5.80
6.51
6.42
5.70
4.62
5.06
3.18
2.49
2.15
Time , min
15
4.70
6.60
5.62
3.40
4.21
3.76
_t
1.32
2.76
5.71
5.43
3.23
2.52
2.28
1.70
2.10
1.54
Coliforms

30
4.30
6.30
5.40
3.52
3.77
3.43
—
1.48
—
4.00
5.55
4.45
3.84
2.41
1.78
_
2.30
1.40
 Dates  of  experiments:   1,  6/28/79;  2,  7/10/79;  3,  7/17/79.
 -  signifies  that  sample was  lost.
                                129

-------
TABLE A-8.  RESULTS OF EXPERIMENTS WITH  SAN  JOSE  NITRIFIED EFFLUENT

Disin- Dose
fectant mg/1
C102 2

5


10


C12 2

5


10


Repli-
cate
No.*
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Average

0
4.98
4.98
4.25
4.96
4.66
4.53
5.34
4.73
4.21
4.78
4.72
4.35
5.67
4.72
4.40
4.87
5.26
4.30
Log Surviving Total
log10[N/100ml]
Contact
5
3.88
3.55
3.04
2.08
1.04
1.11
1.30
0.85
0.84
4.76
4.43
3.60
3.38
1.78
2.04
1.78
0.84
0.90
Time, min
15
3.85
3.41
3.78
1.50
1.38
0.30
0.30
0.30
0.30
3.77
4.19
3.15
2.00
-
—
0.30
<0.3
0.64
Colifonns

30
3.84
3.47
2.78
1.48
1.95
0.30
_t
0.30
<0.3
3.77
3.35
3.04
_
0.30
2.11
0.30
<0.3
0.75
*Dates of experiments:  1, 7/7/79; 2, 7/12/79; 3, 7/19/79.
'- signifies that sample was lost.
                                 130

-------
TABLE A-9.  RESULTS OF EXPERIMENTS WITH SAN JOSE FILTERED, NITRIFIED EFFLUENT

Disin- Dose
fectant mg/1
C102 2

5


10


C12 2

5


10


Repli-
cate
No.*
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Average

0
2.79
5.46
4.73
4.90
5.38
4.72
4.81
5.71
4.59
4.75
4.91
4.76
1.90
5.84
4.86
4.89
5.84
4.78
Log Surviving Total Coliforms
log10[N/100ml]
Contact
5
_t
4.12
—
0.48
1.11
<0.3
<0.3
0.45
<0.3
4.4
4.26
2.30
0.30
2.53
1.11
0.20
1.94
0.69
Time,
15
1.30
2.95
—
<0.3
0.85
<0.3
<0.3
0.30
<0.3
3.59
3.18
2.00
0.30
0.84
0.95
0.30
1.51
0.65
min
30
3.18
2.00
<0.30
0.75
<0.3
0.30
<0.3
<0.3
2.41
1.74
—
0.3
0.45
0.84
<0.3
1.21
0.52
     *Dates of experiments:  1, 7/2/79; 2, 7/13/79; 3, 7/20/79.
      - signifies that sample was lost.
                                      131

-------
                                  APPENDIX B

    LOG-LOG PLOTS OF MODEL PREDICTIONS AND DATA POINTS FOR ALL EXPERIMENTS
 Figure B-l.  Inactivation of coliform bacteria by chlorine dioxide in Palo
              Alto wastewater - 1978.

 Figure B-2.  Inactivation of coliform bacteria by chlorine in Palo Alto
              wastewater - 1978.

 Figure B-3.  Inactivation of coliform bacteria by chlorine dioxide in Palo
              Alto wastewater - 1979.

 Figure B-4.  Inactivation of coliform bacteria by chlorine in Palo Alto
              wastewater - 1979.

 Figure B-5.  Inactivation of coliform bacteria by chlorine dioxide in
              filtered Palo Alto wastewater.

 Figure B-6.  Inactivation of coliform bacteria by chlorine in filtered Palo
              Alto wastewater.

 Figure B-7.  Inactivation of coliform bacteria by chlorine dioxide in Palo
              Alto wastewater - 1978 and 1979.

 Figure B-8.  Inactivation of coliform bacteria by chlorine in Palo Alto
              wastewater - 1978 and 1979.

 Figure B-9.  Inactivation of coliform bacteria by chlorine dioxide in Dublin
              lab experiments.

Figure B-10.  Inactivation of coliform bacteria by chlorine in Dublin lab
              experiments.

Figure B-ll.  Inactivation of coliform bacteria by chlorine dioxide in Dublin
              field experiments.

Figure B-12.  Inactivation of coliform bacteria by chlorine in Dublin field
              experiments.

Figure B-13.  Inactivation of coliform bacteria by chlorine in San Jose
              secondary effluent.

Figure B-14.  Inactivation of coliform bacteria by chlorine dioxide in San
              Jose filtered effluent.

Figure B-15.  Inactivation.of coliform bacteria by chlorine dioxide in San
              Jose nitrified effluent.

Figure B-16.  Inactivation of coliform bacteria by chlorine dioxide in San
              Jose secondary effluent.


                                      132

-------
                10
                  -1
                10
                  -2
                10
                  -3
                10
                  -4
                10
                  -5
                10
                  -6
 N(T)/N(0)=[(RT)/1.56]
"r=0.86
                                             -2.90
                                                                 i i  i I i I
                             10
                               -1
                          10J
         10*
                             RESIDUAL — TIME IN MG-MIN/L
Figure B-l.  Inactivation of coliform bacteria by chlorine dioxide in Palo
             Alto wastewater - 1978.
                 io-i  ,
            o
            K
            t)
            CO
      10'
                                        10
10J
                              RESIDUAL — TIME IN MG-MIN/L

Figure B-2.  Inactivation of coliform bacteria by  chlorine  in  Palo Alto
             wastewater - 1978.
                                    133

-------
            10-1


            10~2


            10~3
       S    iO-4

            10
              -5
            10
              -6
                                 4-
                     N(T)/N(0)=[(RT)/0.67]-2-30
                    "r=0.86
                     1  I I 1 M 111	I I  I 1 1 I 1 11	1 I  I 1 I I I 11	I  I I 1 I I I 11	I  I I I I I I I
                         10
                           -1
                                                       10-=
                         RESIDUAL — TIME IN MG-MIN/L
Figure B-3.  Inactivation of coliform bacteria  by  chlorine  dioxide  in  Palo
             Alto wastewater - 1979.
        O
        I—I
        HI
             10
               	1
             10
               -2
             10
               -3
             10
               -4
             10
               -5
             10
               -6
                     N(T)/N(0)=[(RT)/2.21]-2'22
                     "r=0.83
                                         i  i i 1 1 1 i
                          10
                            — 1
10
10
                                                        10
Figure B-4.
                          RESIDUAL — TIME IN MG-MIN/L

             Inactivation  of coliform bacteria by chlorine in Palo Alto
             wastewater -  1979.
                                    134

-------
      o
      I
           10
             ~2
10"3

10-4

10~5

10"6
                   N(T)/N(0)=[(RT)/.004]'1-81
                   "r-0.89
                        10"1      10°       101       10£
                        RESIDUAL — TIME IN MG-MIN/L
Figure B-5.   Inactivation of  coliform bacteria by chlorine dioxide in
             filtered  Palo Alto wastewater.
            10
              -1
            10
              -2
            10
              -3
            10
              -4
       $    10
              -5
            10
              -6
       : N(T)/N(0)=[(RT)/0.84]
       Fr-0.86
                                        -3.10
                         IQ-I      10°       IQ!       io2
                         RESIDUAL — TIME IN MG-MIN/L
Figure B-6.  Inactivation of  coliform bacteria by chlorine in filtered Palo
             Alto wastewater.
                                    135

-------
              io-i   ._
              io
I—I
i
3
              10-4  1_
              10'
               10
                -6
                                         + \, + +   +
            :  N(T)/N(0)=[(RT)/1.29]

            Fr=0.86
                                                              j—i  i 11
10-
                             10
                                                10
10*
                           RESIDUAL — TIME IN MG-MIN/L



Figure B-7.   Inactivation of  coliform bacteria  by chlorine dioxide in Palo

             Alto wastewater  -  1978  and 1979.
              10-1



              10~2



              10~3



              10-4



              10~5


                _6 r   N(T)/N(0)=[(RT)/3.95]~2-78

              10    ^~r=0.88


                      _j	i i 111
ic
                      1
                            10
                                       °
                                                         io
                           RESIDUAL — TIME IN MG-MIN/L



Figure B-8.  Inactivation of coliform  bacteria by  chlorine  in Palo Alto

             wastewater - 1978 and 1979.
                                    136

-------
               10"2


               ID"3


               10-4


               10~5
               10
                 -6
          N(T)/N(0)=[(RT)/0.
         "r=0.78
                        i i  i i mil	i	i i i mil	1—i i i ii
                            10
                              -1
                        10
                          0
        10
          2
                            RESIDUAL — TIME IN MG-MIN/L
 Figure B-9.   Inactivation  of coliform bacteria by chlorine dioxide in Dublin
              lab  experiments.
                 10-1

                 10"2
                 10
                 10
                   -3  _
                   -4
                 10~5  fe-
                 10"6  fer-
           N(T)/N(0)=[(RT)/0.65]-1-62
           "r=0.91
                             10
                               -1
                          10
                            0
10
  1
10*
Figure B-10.
                RESIDUAL — TIME IN MG-MIN/L

Inactivation of coliform bacteria by chlorine in Dublin lab
experiments.
                                     137

-------
         g
          cfl
Figure B-ll.
               10
                 -1
               10
                 -2
               10
                 -3
 10
   -4
               10
                 -5
               10
                 -6
         N(T)/N(0)=[(RT)/0.57]-2'20
         'r=0.84
                        I  I I
                            10
                              -1
                                  i ii	i	i	i i 111
                        10
                          0
10J
10
            2
              RESIDUAL — TIME IN MG-MIN/L

Inactivation of coliform  bacteria by chlorine dioxide in Dublin
field experiments.
               10
               10
                 -2
               10
                 -3  _
               10
                 -4
               10
                 -5
               10
                 -6
         N(T)/N(0)=[(RT)/1.67]-1'79
         "r=0.93
                            10
                              -1
                        10°
10J
Figure B-12.
              RESIDUAL — TIME IN MG-MIN/L

Inactivation of coliform bacteria  by  chlorine  in Dublin field
experiments.
                                     138

-------
       g
io-i

10~2

10~3

10-4

10~5

10~6
                    N(T)/N(0)=[(RT)/4.06]-2'82
                    "r-0.86
                         10
                           -1
10
          101
                                           10<
                         RESIDUAL — TIME IN MG-MIN/L

Figure B-13.  Inactivation of coliform bacteria by chlorine  in  San Jose
              secondary effluent.
            10
            10
              ~2
            10'
            10
              -4
            10
              -5
            10
              -6
      :  N(T)/N(0)=[(RT)/.003]
       "r=0.81
                                        -1.13
                        10
                          -1
10
                         °
                                          10*
                        RESIDUAL — TIME IN MG-MIN/L

Figure B-14.   Inactivation of  coliform bacteria by chlorine dioxide in San
              Jose filtered effluent.
                                     139

-------
           10
             -1
           ID
             -2
           10
               3
      g
           10
             -5
           10
             _6
                    N(T)/N(OH(RT)/0.78]-2-06
                   "r=0.75
                    i  i i 1111 ii   i  i i 111 ni   i  i i 111 n    i i  i 111 ni   i i  i 11111
                        10
                          -1
                                                      10*
Figure B-15.
                        RESIDUAL — TIME IN MG-MIN/L

              Inactivation of coliform bacteria  by  chlorine  dioxide  in  San
              Jose nitrified effluent.
       g
            10
            10
              -2  _
            10
            10
            10-
              -4  _
            10
              -6
                   -  N(T)/N(0)=[(RT)/0.89]
                   Fr=0.78
                                         -3.18
                         10"1       10°       101       102
                         RESIDUAL — TIME IN MG-MIN/L

Figure B-16.   Inactivation of coliform bacteria  by  chlorine dioxide  in  San
              Jose secondary effluent.
                                     140

-------
                                  APPENDIX C

       A DETAILED EXAMPLE OF THE CALCULATIONS TO DETERMINE THE REQUIRED
                  DISINFECTANT FEED CAPACITY—POUNDS PER DAY
     These example  calculations are for  the  disinfection by chlorine dioxide
of a conventional activated-sludge treatment effluent with a  1000/100 ml  total
coliform standard at the  treatment  plant  outfall.  This corresponds to Case  C
discribed in the main body of this report, Table  36.

     1.  Determine  the  initial bacterial  concentration in  the waste
         stream to be disinfected.
             For  conventional  activated-sludge  treatment,  the value
         read from Table 14 is  1.62  x 106 total coliforms/100 ml.
     2.  Calculate  the  required  survival  ratio,  N(t)/N(0),  from the
         initial  bacterial concentration,  N(0).    For  Case   C,  the
         disinfection standard  is  1000  total coliforms/100 ml.   The
         survival ratio is then:

              1000 total coliforms/100 ml
            1.62  x  106  total  coliforms/100  ml  = 6'17 X 10

     3.  Calculate  the required residual  time product using  Eqs .  24
         and 25 and the model-fitting constants from Table 24.


                         N(t)


                       N(t) = N(0)   for  Rt < b

         where N(0) =  initial total coliform bacteria;

               N(t) =  total coliform bacteria at time t;

               R    =  residual oxidant in  mg/1 as measured at  time  t;

               b    =  lag time coefficient  (Rt)' in mg'min/1;

               k    =-k'/a = velocity coefficient; and

               t    =  time in minutes.
                                      141

-------
         The model  fitting  constants  for  Case  C, disinfection  with
         chlorine dioxide are:

                               b - 1.56

                              k = -2.90

         The required residual-time product is then:

                          N(t)/N(0)  = [^]

         Solving for Rt:


                  Rt =  [N(t)/N(0)]1/k  • b


                  Rt =  [6.17  x 10~4]~1/2'90 •  1.56


                  Rt =  19.95  mg C102  * min/1
     4.  Calculate  the  required  residual,  based  on  the  provided
         contact  time,  namely  one hour.   The  required residual  is
         then:

                      Rt     19.95 mg C102 * min/1
                    60 min           60 min


                          R = 0.33 mg C102/l

     5.  Calculate  the  disinfectant  dosage  needed  to  provide  the
         required residual disinfectant concentration.

     In order  to be  able  to calculate the disinfectant  dosage, an  empirical
relationship relating disinfectant dose to residual  has  been developed  using
the observed values of oxidant  demand (see Table 27 and  Reference 33).   The
dose required is then given by the sum of the required residual  to  achieve  the
disinfection standard plus the one-hour  oxidant demand.   The one-hour  demand,
however, is  also a function of  dose,  as it must  be  since demand  equals dose
minus  residual.    A  linear  regression  of  log[1-hour   oxidant   demand]   on
log[dose]  yields the  equation:


                              Dl-hour ' 3(Dose)a

where Di_ho r = one-hour oxidant demand-mg/1,

       3,a     = constants, dependent on the oxidant and the wastewater,

      Dose    = oxidant dose-mg/1.

                                      142

-------
     The equation to be solved is:
                     Dose = Residual required +  3(Dose)
                                                        a
The values of  3  and  a  and the regression correlation coefficients  for  Palo
Alto and Dublin Wastewaters are shown in Table C-l.
        TABLE  C-l.   CONSTANTS FOR ONE-HOUR DISINFECTANT DEMAND EQUATION
                             AS A FUNCTION OF DOSE
Case
Type of Effluent)
A, B, and C
(conventional
activated sludge)
D, E, and F
(filtered, nitrified
activated sludge)
. cio2
gar
0.99 0.80 0.97


0.94 0.995 0.99


ci2
6 a
0.63 0.46


1.11 0.80


*
r
0.81


0.98


          correlation coefficient
     For Case C, the required dose is:
                                                 0.80
                         Dose = 0.33 + 0.99(DOSE)

                         Dose =2.17 mg/1 C102


The required feed capacity in pounds per day  is  given by:


      2.17[g/m3] x l/454[lb/g]  x l/264.2[m3/gal]  x  106[million gal/gal]

             =  18.1  [pounds  C102 per day]/[mgd wastewater treated]


The corresponding  values are entered  in the  row designated  "Case C, Cin"  i-n
Table 38.
                                      143

-------