svEPA
             United States
             Environmental Protection
             Agency
             Municipal Environmental
             Research Laboratory
             Cincinnati OH 45268
EPA-600/2-78-029
March 1978
             Research and Development
Full-scale Demonstration
of Nitrogen Removal
by Breakpoint Chlorination

Environmental Protection
Technology Series

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.   Environmental Health  Effects Research
      2.   Environmental Protection Technology
      3.   Ecotogical Research
      4.   Environmental Monitoring
      5.   Socioeconomic Environmental Studies
      6.   Scientific and Technical Assessment Reports (STAR)
      7.   Interagency Energy-Environment Research and Development
      8.   "Special" Reports
      9.   Miscellaneous Reports

This report has been assigned  to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                         EPA-600/2-78-029
                                         March 1978
FULL-SCALE DEMONSTRATION OF NITROGEN REMOVAL
          BY BREAKPOINT CHLORINATION
                       by

                 Richard W. Stone

            Sacramento Area Consultants
            Sacramento,  California 95814
                       for
    Sacramento Regional County Sanitation District
            Sacramento,  California 95827
            EPA Grant No. S-803343-01-0
                 Project Officers

                James J. Westrick
                Francis L. Evans III
           Wastewater Research Division
    Municipal Environmental Research Laboratory
              Cincinnati, Ohio  45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
      OFFICE OF RESEARCH AND DEVELOPMENT
    U.S. ENVIRONMENTAL PROTECTION AGENCY
             CINCINNATI, OHIO  45268

-------
                           DISCLAIMER


    This report has been reviewed by the Municipal Environmental
Research Laboratory,  U.S.  Environmental Protection  Agency,  and
approved for  publication.   Approval  does  not signify  that  the
contents necessarily  reflect  the  views  and policies  of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products  constitute endorsement or recommendation
for use.

-------
                           FOREWORD
     The Environmental  Protection Agency was created because  of
increasing public and  government concern about the  dangers  of
pollution to the health  and welfare of  the American people.
Noxious air, foul water,  and  spoiled  land are tragic testimony
to the deterioration of  our natural environment.  The complexity
of  that  environment and  the  interplay  between its  components
require a  concentrated  and  integrated  attack on  the  problem.

     Research and development  is that  necessary first step  in
problem solution and it  involves  defining  the problem, measuring
its impact, and searching  for  solutions.   The Municipal Environ-
mental Research Laboratory develops new and improved technology
and systems  for the prevention,  treatment, and management  of
wastewater and  solid and hazardous waste pollutant  discharges
from municipal  and  community  sources, for the preservation  and
treatment  of  public drinking water  supplies,  and  to  minimize
the adverse economic,  social,  health,  and aesthetic effects
of  pollution.   This publication is one  of  the  products  of
that  research; a most  vital cpmmunications link between  the
researcher and the user  community.

     This  report summarizes the  results  of  a demonstration
program which was carried out to establish  design and operating
criteria  for  breakpoint chlorination nitrogen  removal  systems
for use  in municipal wastewater treatment applications.   The
development and demonstration of such processes provides addi-
tional  tools for use by  water pollution control agencies  in
their  efforts  to  maintain  and enhance the  quality of  the
environment.
                     Francis T.  Mayo,  Director
                     Municipal  Environmental Research Laboratory
                              Xll

-------
                           ABSTRACT
    This work constitutes  the first large-scale demonstration  of
breakpoint  chlorination  for removal of  ammonia nitrogen  from
municipal wastewater.   The report  includes  a discussion  of
breakpoint chlorination  process chemistry,  an evaluation  of
chemical  consumption, and  recommendations for  process  control.

    Breakpoint chlorination can  occur  in wastewater  effluents
following addition of sufficient chlorine for  the  oxidation  of
ammonia nitrogen tor principally, nitrogen gas.  Theoretically,
7.6 mg/1 of chlorine are required for the conversion of 1.0  mg/1
of  ammonia  nitrogen  to  nitrogen gas.  In  the breakpoint
chlorination process evaluation  at  Rancho Cordova, an  average
dosage of  10  mg/1 chlorine was required for the conversion  of
1.0 mg/1  ammonia  nitrogen to  nitrogen gas.   The difference
between theoretical and actual  chlorine  dosages was found to  be
largely  due  to  chlorine  consumption  from the oxidation  of
ammonia nitrogen to nitrate nitrogen,  a  competing side reaction
with a less desirable end  product.  Overall,  about 96 percent  of
the chlorine  dosed  to  the Rancho Cordova system was  accounted
for in  reactions between  chlorine  and  nitrogenous specias  in
specific, identified  chemical  pathways,  and  free  chlorine
residual remaining in solution following breakpoint.

    Tests with  mechanical  mixing  devices showed  that mixing
intensity at the  point of  chemical addition had little  effect  on
chemical  consumption and  efffluent  quality.    The  rate of the
breakpoint reaction was found  to vary  depending upon  the system
pH, with fastest rates  at  a pH  of 7.0.   At pH 7.0,  about 60 sec
to  90 sec elapsed between chemical addition and completion  of
the chemical  reactions.   The Rancho Cordova  evaluation showed
that the process requirement for alkalinity  supplementation (pH
control)  was  almost 'exactly equal to  that  predicted from the
chemical  stoichiometry.    A compound loop control  strategy was
recommended for  dosage   control of chlorine  and alkalinity
supplement.

    This  report was  submitted  in  fulfillment  of Grant
No.  S-803343-01-0  by Sacramento  Area  Consultants  for  the
Sacramento Regional  County Sanitation  District.   The  project
was carried out under the sponsorship  of the  U.S. Environmental
Protection  Agency.  This report covers work  completed as  of
June 30, 1977.
                              xv

-------
                          CONTENTS

Foreword ........ 	 ....  iii
Abstract	   iv
Figures	vii
Tables	viii

     1.   Introduction	• •  1
     2.   Conclusions	2
     3.   Chemistry of the Breakpoint
            Chlorination Process	  5
               Chemistry of ammonia in water 	  5
               Chemistry of chlorine in water	6
                    Formation of chloramines 	  6
                    The breakpoint reaction	7
                    The breakpoint curve  	  8
                    Other reactions encountered in
                      breakpoint chlorination	9
               pH and Alkalinity Considerations	10
     4.   Experimental Methods  .	13
               Breakpoint Chlorination Facilities	13
               Control and Monitoring Equipment	17
                    Free chlorine residual analyzer.  ... 17
                    Ammonia nitrogen analyzer	18
                    pH monitor	  . 18
               Laboratory Testing Procedures 	 18
                    Ammonia nitrogen (NH3 -N)	19
                    Nitrate nitrogen (N0~ -N).  . .  .  .  .  .19
                    Chlorine species 	 19
     5.   Results and Discussion	20
               Process Influent Characteristics	20
               Initial Mixing	20
               Reaction Rates	  i  . 22
               Nitrogenous Residuals	.  . 27
                    Nitrogen trichloride  	  ... 27
                    Nitrate	29
                    Organic nitrogen	30
               Overall Chlorine Consumption	30
               Alkalinity Supplementation	31
               Chlorine Injector Water
                 Considerations	.32
               Breakpoint Model Predictions	34
     6.   Process Control	 35
               Process and Component Descriptions.  .  .  .  . 35
                    Ammonia analyzer 	  .35
                    Flowmeter	36
                              v

-------
Contents (continued)

       6.   Process Control (continued)
                      Chlorinators	36
                      Free chlorine analyzer 	 38
                      pH analyzer	39
                      Sodium hydroxide feeder	39
                 Control Requirements	39
                 Control Systems	.	40
                      Simple feedback control	40
                      Simple feedback plus flow paced
                        control	 . 44
                      Flow modified feedback control
                        plus flow paced control	46
                      Ammonia nitrogen mass flow paced
                        control	48
                 Recommended Control System	48
                      Alkalinity supplement feed
                        control	48
                      Chlorine feed control	49
                 Instrumentation Requirements	51
                      Control components	 . 51
                      Drift in zero and linearity	51
                      Feedback controller scaling	51
                      Signal redundancy	51
                      Calibration signals	51
                 Summary	52

  References	53
  Appendices

     A.  Rancho Cordova Breakpoint Chlorination
         Demonstration Program - Data Summary	   54
     B.  Rancho Cordova Breakpoint Chlorination
         Demonstration Program - Breakpoint
         Model Predictions	   58
                              VI

-------
                            FIGURES

Number                                                      Pag,

 3-1   Theoretical Breakpoint Curve	8
 3-2   Conceptual Model for pH Change in Breakpoint
         Chlorination	't   ^2
                                                 i
 4-1   Process Flow Schematic	     14
 4-2   Flow Splitter Box	!.*!!!! 15
 4-3   Automatic Chlorine Transfer Piping.  ......... ie
 4-4   Chlorine and NaOH Application Point	! ! 17
 4-5   Sampling System	! ! ! 18

 5-1   Process Influent Variations	21
 5-2   Ammonia Removal with Breakpoint Chlorination
         in a  Pipe Reactor - pH  Set Point 6.5	23
 5-3   Ammonia Removal with Breakpoint Chlorination
         in a  Pipe Reactor - pH  Set Point 7.0	23
 5-4   Ammonia Removal with Breakpoint Chlorination
         in a  Pipe Reactor - pH  Set Point 7.0	24
 5-5   Ammonia Removal with Breakpoint Chlorination
         in a  Pipe Reactor - pH  Set Point 7.3	24
 5-6   Ammonia Removal  with Breakpoint Chlorination
         in a  Pipe Reactor - pH  Set Point 7.5	25
 5-7   Ammonia Removal  with Breakpoint Chlorination
         in a  Pipe Reactor - pH  Set Point 7.7	25
 5-8   Ammonia Removal  with Breakpoint Chlorination
         in a  Pipe Reactor - pH  Set Point 8.0	26
 5-9   Ammonia Removal  with Breakpoint Chlorination
         in a  Pipe Reactor  - pH  Set  Point 8.5	26
 5-10   Effect  of Breakpoint  Chlorination  on Organic
         Nitrogen.  .  .    	      30
 5-11   Summary  of  Cl2:NH4  -N Observations 	  [31
 5-12   Breakpoint  Chlorination Chlorine Consumption.'  !  !  '  *  32
 5-13   Lb NaOH/Lb  C12 Added	  !  33

 6-1    Simplified  Process  Diagram.  ...                      36
 6-2    Control Panel	!!.*.*.*''  37
 6-3    Typical pH  Characteristics	..!.!.!.!  41
6-4    Simple Feedback Control - Chlorine Feeders.'  .*!.*!!  42
6-5   Control Response - Simplified Control Loop	.'43
6-6   Simple Feedback Plus Flow Paced
        Control - Chlorine Feeders	44
6-7   Simple Feedback Plus Flow Paced
        Control - Alkalinity Supplement  	 45

                             vii

-------
Figures (continued)

Number

 6-8   Effects of Flow Responsive Control
         on Free Chlorine Residual . . 	
 6-9   Flow Modified Feedback Plus Flow Paced
         Control - Chlorine Feeders. .......
 6-10  Recommended Control - Alkalinity Supplement
 6-11  Recommended Control - Chlorine Feeders.  . .
                                                      46

                                                      47
                                                      49
                                                      50
Number

 3-1

 3-2

 3-3


 5-1

 5-2
 5-3
                             TABLES
Common Reactions Encountered in Breakpoint
  Chlorination.	
Chlorine Requirement for Chemical Transformations
  Commonly Encountered in Breakpoint Chlorination
Acidity and Alkalinity Considerations in
  Breakpoint Chlorination 	 ....
Secondary Effluent Quality, Rancho Cordova,
  California	,	
Reaction Rate Data Summary. .... 	
Formation of Nitrogen Trichloride and Nitrate .
 6-1   Summary of Process Control Signals,
 9

10

11
22
27
28

38
                             viix

-------
                            SECTION 1

                          INTRODUCTION


     Chlorination can  be  used to remove  ammonia  nitrogen  from
wastewater.  It requires the addition of chlorine, and makes use
of the rapid, specific reactions  between chlorine and ammonia in
dilute  aqueous  solution which  can  lead  to  the oxidation of
ammonia  to  nitrogen gas  and other  end  products.   The  term
"breakpoint" has been applied to the process because a point of
minimum chlorine residual  occurs  at the chlorine dosage required
to complete the chemical reaction.

     Testing of breakpoint  chlorination  for  nitrogen  removal
from  sewage has, to date, largely been  carried out  in small
pilot-scale systems.  Data  on  breakpoint chlorination  from
pilot-scale  systems  have contributed greatly to our understand-
ing of process chemistry,  but actual application data and design
criteria for full-scale breakpoint systems  have not been readily
available  to the design engineer and  planner.   The  work  dis-
cussed here  was carried out  to  provide  information  and design
recommendations from a full-scale breakpoint chlorination system
in the specific  areas  of process control techniques and instru-
mentation,  chemical  consumption, and  breakpoint  reaction  end
products.  The system tested during this program was designed to
accept variations in flow  rates  and  chemical composition which
are  characteristic of effluent from  nonnitrifying  secondary
wastewater treatment plants.

     The full-scale demonstration of breakpoint chlorination was
carried  out at  Rancho Cordova, California,  between December,
1975  and  March, 1976.   Costs  of  the evaluation were  borne
jointly  by  the  U.S.  Environmental  Protection  Agency,  the
Sacramento Regional  County  Sanitation District and the State of
California.

-------
                           SECTION 2

                          CONCLUSIONS
     A  full-scale  breakpoint  chlorination  facility was
constructed  at the  Rancho Cordova Sewage  Treatment Plant  in
Rancho Cordova, California.   The system was operated  24  hours
per  day,  5 days  per week from  December,   1975 until March,
1976.  Process flow rates varied,from about  0.1 mgd to 1.2 mgd.
Influent ammonia nitrogen concentrations were ordinarily in the
range of 15 mg/1 to 25 mg/1.

     Flexibility was incorporated into the breakpoint chlorina-
tion chlorine dosage and pH control systems to  facilitate  the
testing of several control strategies.  Data from automatic on-
line analyzers and from  laboratory analysis of samples collected
manually provided  the   basis  for process chemistry  investiga-
tions.

     A number  of specific observations and conclusions made as a
result of the Rancho Cordova breakpoint chlorination demonstra-
tion program are enumerated below:

     1.   The  dosage of  chlorine at  Rancho  Cordova required  to
          reach breakpoint and  maintain a  controllable  free
          residual in the  process  stream averaged 10 mg/1  for
          each 1.0 mg/1  ammonia nitrogen present in the  process
          influent.

     2.   Approximately  70 percent  of  the  breakpoint  chlorine
          dosage was  consumed to produce nitrogen gas (N2) from
          ammonia  (NH4>   at  pH  set points between  pH 7 and  8.
          The  oxidation  of  ammonia to nitrate consumed 8  percent
          to  19 percent of  the total  chlorine  dosed  to  the
          system.   Overall,  about  96 percent of  the  total
          chlorine dosage was accounted  for in reactions  between
          chlorine and  nitrogenous  species  in specific chemical
          pathways  and   free  chlorine  residual  remaining  in
          solution following breakpoint.

     3.   Nitrate   (N03)  production  in  breakpoint  chlorination
          was  not  found  to be  pH sensitive,  with about 1.0 mg/1
          of nitrate  (as N) produced from ammonia across  a  final
          system pH range of pH 6.5 to  8.5.  The production  of

-------
     nitrate  from  nitrite  was  wholly  dependent upon
     influent  nitrite concentration.

4.   Nitrogen  trichloride  production  was  observed  to  be
     fairly insensitive  to  pH across  a  range  of final
     system pH values from pH 7 to 8.  The median  value  for
     nitrogen  trichloride  production was about 0.4 mg/1  (as
     N)  when breakpoint effluent  was used as the  source of
     chlorine  injector  water.  While the amount of chlorine
     consumed  in its formation was relatively small (4 per-
     cent to  6  percent of total  chlorine dosed),  nitrogen
     trichloride generation  affects  the  minimum  ammonia
     concentration  that  can  be  achieved in breakpoint,
     since it decays slowly  in  dilute solution and it is
     converted to ammonia  upon dechlorination with sulfite.

5.   The  concentration of organic  nitrogen  compounds  was
     not affected by breakpoint chlorination.

6.   The rate of  reaction for  breakpoint chlorination  was
     found  to vary depending  upon the  pH control point
     (final system pH  ),  with  fastest rates observed at a
     pH set point  of  pH   7.0.  The  time  to completion  was
     found to be between  60 sec and 90 sec at pH  7.0.   The
     reaction  rate  slowed  considerably at  a pH set point of
     6.5, and also  became progressively  slower as  pH  was
     increased from pH  7.3 to pH 8.5.

7.   Variations  in the  amount  of mechanical mixing
     intensity in the  zone of breakpoint  chemical applica-
     tion  had  no  effect upon  overall  system   chemical
     consumption and  effluent quality.  Mechanical mixing,
     to facilitate a  rapid and  thorough blending of process
     chemicals  and  influent  stream,  was  important  in
     damping  oscillations in free  chlorine concentration
     for control purposes.

8.   Sodium  hydroxide (NaOH)  was  used  throughout  the
     study  as an alkalinity  supplement.    The amount  re-
     quired to  neutralize all  breakpoint acidity (1.53 Ib
     NaOH/lb Cl2)  was  essentially  identical to that
     predicted from chemical  stoichiometry.

9.   In breakpoint chlorination systems which have gaseous
     chlorine  as  the chlorine  source,  the  volume  and
     chemical composition of the  chlorine injector water
     may  be important  factors in  process performance  and
     control.    At  Rancho Cordova, when  secondary effluent
     (breakpoint process  influent)  was used as the injector
     water source, more nitrogen trichloride formed in the
     injector water than  when breakpoint  effluent with low

-------
      ammonia  content was used.  Specific design  recommenda-
      tions  for  breakpoint  chlorination chlorine  injector
      water  systems are given in the text.

10.    The  breakpoint chlorination control system  recommended
      here,  while  comprised  of  commercially  available
      components,  can  provide  satisfactory  nitrogen removal
      with  minimum chlorine   consumption  only if  it is
      tailored to  the individual  application.   Specific
      recommendations for breakpoint process chlorine dosage
      control  and pH control are made in the text.

-------
                           SECTION 3

        CHEMISTRY OF THE  BREAKPOINT CHLORINATION PROCESS
     The use  of  chlorine  as a  disinfectant  in  water treatment
began  in  about 1800  in Europe.  In  about 1893, chlorine  was
applied as a wastewater disinfectant in New York.  In the early
1920's, Houston1 observed that high dosages of chlorine removed
tastes and odors from water.   The use of high dosages of chlo-
rine  in water  treatment  led some  operators  to observe  the
increase,  disappearance and subsequent reappearance of chlorine
residual as chlorine dosage was incrementally increased.  These
observations  gave  rise  to  the  term "breakpoint" chlorination.
Calvert^ recognized  in  1940 that the  breakpoint phenomena  was
the result  of the  oxidation  of ammonia  nitrogen  in solution.

     This  section presents  the chemistry  of chlorine and ammonia
in aqueous solution,  and  describes  the breakpoint  reactions
which can lead to oxidation of ammonia nitrogen to end products
composed principally of nitrogen gas.   Compounds  of chlorine and
of nitrogen may  occur  in a variety of species and proportions.
For this reason quantitative  terms herein are, unless otherwise
noted,  expressed  in terms  of  the elemental  forms,  chlorine or
nitrogen.


CHEMISTRY OF AMMONIA IN WATER

     The actual chemical form of ammonia  in dilute aqueous solu-
tion  depends largely  upon pH and  temperature.  The relative
distribution of nondissociated ammonia and ammonium ion  may
be defined according to the equation below:
                           t
                 NH3 +  H+          K = 5 x 10-10 at 20 C    (1)


where   NH3 = nondissociated ammonia

        NH4 = ammonium ion

        H+  = hydrogen ion

        K   = dissociation  constant

-------
According to  the dissociation constant,  the  pH value at  which
nondissociated ammonia  and ammonium  ion are  present in  equal
proportions (pK)  is  about  pH 9.3 at  20 C.   Above  pH 9.3,  non-
dissociated ammonia  predominates.   Below pH 9.3,  ammonium  ion
predominates.


CHEMISTRY OF CHLORINE IN WATER

     When  chlorine  gas is  dissolved  in water,  a hydrolysis
reaction occurs according to the following relationship:

             C12 (GAS) + H2O *+—» HOC1 + H* + Cl~            (2)

followed by a dissociation reaction:

      .  HOC1 *  » OC1"  + H+        K =  3.3 x 10~8  at  20  C    (3)

where   HOC1  =  hypochlorous acid

        OC1"  =  hypochlorite ion

        H"1"    =  hydrogen ion

        K     =  dissociation constant

     The  pH value at  which  hypochlorous acid  and hypochlorite
ion are present  in equal proportions  (pK)  is  about 7.5 at  20 C.
Above  pH 7.5, hypochlorite  ion predominates.  Below  pH  7.5,
hypochlorous acid predominates.  The total concentration of both
hypochlorous acid and  hypochlorite  ion is commonly termed  "free
available chlorine."

Formation of Chloramines

     When  chlorine  and  ammonia  are  present  in  dilute  aqueous
solution, reactions  can occur according to a number of chemical
pathways.   Reactions between chlorine  and  ammonia which result
in the formation of chloramines are described below:

       NH4  + HOC1 	»* NH2C1  (monochloramine)  + H20 + H+    (4)

       NH2C1 + HOC1 	9- NHC12 (dichloramine) + H20         (5)

                              and

       NHC12 + HOC1 	*• NC13 (nitrogen trichloride) + H20   (6)

The  relative  amounts of the  chloramine species  which exist in
solution  depend upon  certain process  variables,  including pH,
temperature,  contact time and  the initial chlorine  to ammonia

-------
ratio (Cl2:NH4 -N).  The total of the chloramine  species  pre-
sent in solution is termed  "combined"  chlorine  residual.

     The ammonia nitrogen concentration  of  a  wastewater effluent
is  typically in  the range  of  10 mg/1  to 40 mg/1,  unless an
ammonia  removal  process such  as biological nitrification has
been employed.  When chlorine  is added  to such an effluent for
disinfection purposes (dosage of  2 mg/1  to  15 mg/1), the predom-
inant  chlorine  residual  species  in solution  is ordinarily
monochloramine (NH2C1).

The Breakpoint Reaction

     Breakpoint chlorination occurs  when sufficient chlorine has
been added to a water or wastewater  sample  to cause the chemical
oxidation of  the  ammonia in solution to nitrogen gas and  other
end products.   Some  disagreement  still  exists  in the  literature
concerning the  actual  chemical  pathway(s)  for breakpoint  chlo-
rination, but  the following set  of  reactions  appear  to be the
most reasonable:

          2NH4 +  2HOC1 	»- 2NH2C1  +  2H20  + 2H+            (4)

          2NH2C1 + 2HOC1 	+~ 2NHC12 +  2H20                 (5)

          NHC12 + H20 	»- NOH +  2H+  +  2C1~                 (7)

          NHC12 + NOH 	»- N2 +  HOC1  +  H+  +  Cl~             (8)
                 3HOC1 	»- N2  + 3H2O + 5H+ + 3C1~          (9)

where     NOH  =  catalytic intermediary compound.

Combined with equation (2), the overall reaction is:

                                                            (10)

Note that breakpoint  occurs through  the sequential  formation  of
monochloramine  and  dichloramine with  the  subsequent  catalytic
decomposition of dichloramine  to end products of nitrogen  gas,
with a  partial  return of  free  residual  (HOC1)  to the  solution.
These reactions confirm  that  1.5 moles (gram molecular weights)
of  chlorine  are  required to  oxidize 1.0 mole of  ammonia  to
nitrogen gas.

     Stoichiometrically,  the  breakpoint  reaction  requires  a
weight  ratio  of chlorine to ammonia nitrogen   (Cl2:NH4 -N)   at
the breakpoint of 7.6:1, as shown below:

          Molecular weight Cl2  =  70.9

          Moles Cl2 required    =   1.5

-------
          Molecular weight NH4  =  14.0 (expressed as N)

          Moles NH4 required    =   1.0

Therefore, C12:NH4 -N = (1.5)(70.9):(1.0)(14.0) = 7.6:1.

The Breakpoint Curve

     The breakpoint chlorination  curve  is  a graphic representa-
tion of chemical relationships which  exist as varying amounts of
chlorine  are added  to dilute  solutions of  ammonia nitrogen.
The theoretical breakpoint curve shown in Figure 3-1 has several
characteristic features.  The  characteristics of the breakpoint
curve shown  in  Zone  1 include  principally  the reaction between
chlorine and ammonium indicated in Equation  4.   The hump of the
breakpoint curve occurs,  theoretically,  at  a chlorine to ammonia
nitrogen weight ratio  of  5:1  (molar  ratio of  1:1).   That ratio
corresponds  to the point at which  the reacting chlorine and
ammonia molecules are present in solution  in equal numbers.
    ,
                                           CHLORINE
                                           APPLIED
            AMMONIA-N CONG
                                              MEASURED
                                              CHLORINE
                                              RESIDUAL
                       BREAKPOINT-
                  IRREDUCIBLE
                  MINIMUM CHLORIN
                  RESIDUAL
•:
*

I
                            WEIGHT  RATIO
               Figure 3-1  Theoretical Breakpoint Curve

-------
     The  chemical  equilibria  of Zone  2  favor the  formation of
dichloramine  and the  oxidation  of  ammonium  according  to Equa-
tions 5,  7  and 8.   These reactions  proceed to, theoretically, a
chlorine to  ammonia  nitrogen  weight ratio  of  7.6:1.   At the
breakpoint, the  ammonium concentration is minimized.

     To  the right  of breakpoint,  Zone 3  chemical equilibria
require  the buildup of free chlorine residual.

     In  practical   applications of  breakpoint chlorination,
reactions  occur which  result  in the  formation of  nitrate,
nitrogen trichloride  and other  end products.   These reactions
consume  chlorine,   cause  the Cl2:NH4 -N  ratio to exceed the
theoretical  value  of  7.6:1 and  affect the shape  of  the break-
point  curve.   All  of  these  observations   are  discussed in
detail in subsequent sections of this report.

Other Reactions  Encountered in Breakpoint Chlorination

     Breakpoint  chlorination  of wastewater effluent  can result
in chemical reactions other than the direct oxidation of  ammonia
to nitrogen gas. Reaction products and chlorine consumption for
such  reactions  are  governed  by factors  such as  the  type and
degree  of  pretreatment,  initial  chlorine to ammonia nitrogen
ratio,  pH and  alkalinity.   A  summary of  the common reactions
encountered  in  breakpoint  chlorination  is given  in  Table• 3-1.
     Table 3-1.  Common Reactions Encountered in Breakpoint Chlorination
       Description
          Reaction Stoichiometry
  "Breakpoint Reaction"

  Nitrogen trichloride
   formation

  Nitrate formation

      (1) from ammonia

      (2) from nitrite

  Reaction with other
   inorganics

  Chloro-organic
   reactions
2 NH + 3 HOC1-**N2 + 3 H2O + 5 H + 3 Cl
   + 3 HOC1 — NC13
  * + 4HOC1 — NC>3 + H20 + 6 H+ + 4
NO
inorganics + HOC1—•* oxidized inorganics
organics + HOC1—"-oxidized organics
(9)


(ID


(12)
(14)


(15)

-------
     The total  amount  of  chlorine residual consumed (by weight)
for  each of  the common  reactions  encountered in breakpoint
chlorination  has  been  summarized in Table  3-2.   For example, a
total of 20.3 mg/1  of  chlorine residual (as Cl2) is consumed in
the  conversion  of 1.0  mg/1  ammonia nitrogen to 1.0 mg/1 nitrate
(as  N).   A similar conversion from  nitrite to nitrate requires
only 5.1 mg/1 chlorine.
       Table 3-2.  Chlorine Requirement for Chemical Transformations
                 Commonly Encountered in Breakpoint Chlorination
Initial
species
NH+-N
NH^-N
NHj-N
NO~-N
*
Other
inorganics
Organic s
Reaction
product
N2
NC13
NO^-N
NOg-N

Oxidized
inorganics
Oxidized
organ ics
Cl2:N molar
ratio required
1.5:1
3:1
4:1
1:1

Varies
Varies
C12:N weight
ratio required
7.6:1
15.2:1
20.3:1
5.1:1

Varies
Varies
     In effluents from biological secondary treatment processes,
reduced  inorganic  compounds such  as sulfides  and  ferrous iron
have become oxidized by the aerobic treatment.  Hence, reactions
of inorganic  compounds  with chlorine are  usually  not quantita-
tively  significant/   Partially  oxidized and  slowly degradable
organic matter  is  present, both in  solution  and in particulate
suspended matter,  which  does react,  particularly with  free
chlorine.   Such  matter  includes  nitrogen-containing  protein-
aceous matter.
pH AND ALKALINITY CONSIDERATIONS

     The nature and concentration of the breakpoint chlorination
end products,  chlorine  dosage required to  reach  breakpoint and
the rate of the  breakpoint  reaction are  all affected  by the
initial pH (following chemical addition) and the pH change which
                               10

-------
occurs as  the  breakpoint  reaction proceeds.  The initial pH  in
the reaction zone  and  pH  change  through  breakpoint  depends  upon
the pH  and alkalinity of  the process influent  stream,  ammonia
concentration and  chlorine dosage,  and the  amount of  alkalinity
supplementation.

     Acidity is  generated in breakpoint chlorination  applica-
tions from both  the  hydrolysis and  dissociation  of  chlorine gas
(when gaseous  chlorine is used), and the oxidation  of  ammonia
nitrogen as shown in Table 3-3.   Stoichiometrically, three moles
of hydrogen  ions are liberated in  the hydrolysis and dissocia-
tion  of  sufficient  chlorine  for  the oxidation  of  one  mole  of
ammonia nitrogen,  assuming  the  initial pH  in the reaction  zone
is alkaline.  One mole of hydrogen ions  is liberated in the
oxidation of ammonia to nitrogen gas.


  Table 3-3.  Acidity and Alkalinity Considerations in Breakpoint Chlorination
Acidity source
Hydrolysis and dissociation of chlorine
Oxidation of ammonia
Total
Alkalinity source
Lime
Sodium hydroxide
3 C12 + 3 H2U^ OCP +- 6H* + 3C1"
2 NH^ + 3 OCI~-^N2 + 3 H2O + 2H+ + 3 CI~
3 C10 + 2 NH"!J— »N0 + 8H+ + 6 CI~
4 CaO + 4HnO-^4Ca + 8 OH~
8 NaOH-^8 Na+ +8 OH~

(16)
(17)
(10)
(18)
(19)
     When an alkalinity source is added to neutralize breakpoint
acidity,  the  acidity from chlorine  hydrolysis  and dissociation
is neutralized  immediately.   The acidity  from  the oxidation  of
ammonia is released as the breakpoint reaction progresses.  As a
result, the  pH  in the breakpoint reaction chamber decreases  as
reaction time increases.  At high initial ammonia concentrations
and  low  system  alkalinity,  the  pH  excursion in  the breakpoint
reaction  zone  may be  relatively  large.   Figure 3-2 shows a
conceptual model for  the  range of  pH values which would  be
encountered  in  breakpoint  chlorination  of  20  mg/1 ammonia
nitrogen   in  a  secondary  effluent  with  moderate  pH  buffer
capacity.
                               11

-------
11
10
 9
 8
 6
    ESTIMATED pH
    IMMEDIATELY FOLLOWING .
    ADDITION OF BREAKPOINT '
    CHEMICALS
                                       OXIDATION OF AMMONIA (1 H+)
                                PREDICTED PATH OF
                                pH CHANGE DURING
                                BREAKPOINT CH LOR I NATION
                                        CHLORINE HYDROLYSIS (3 H+)
                                           OH" ADDITION (4 OH")
                                                  TYPICAL pH
                                                  TITRATION CURVE
INITIAL CONDITIONS

pH          -  7.4
ALKALINITY  -  165 mg/l ( as CaCO3)
AMMONIA    =  20 mg/l (as N)
  6
   4          2

   OH~ ADDITION, meq/l
    2          4

H+ ADDITION, meq/l
   Figure 3-2  Conceptual Model for pH Change in Breakpoint Chlorination
                                12

-------
                           SECTION 4

                      EXPERIMENTAL METHODS


     The breakpoint chlorination demonstration program described
here was carried out  at the Rancho Cordova Wastewater Treatment
Plant.   The Rancho  Cordova  Plant,  located  in  the  Sacramento,
California area, serves an area of 3,500 acres and a population
of nearly 25,000.   The average daily flow of about  2.4 mgd  is
composed primarily of  domestic sewage.   Secondary treatment  is
provided  through  primary  sedimentation  and  a  Spiro-Vortex
activated sludge process.  The  secondary effluent is chlorinated
for disinfection and discharged to the American River.


BREAKPOINT CHLORINATION FACILITIES
                                                         v
     Breakpoint chlorination  facilities  were  constructed  at
Rancho Cordova  as  indicated  in the  process schematic of  Figure
4-1.   A flow splitter box  was constructed downstream of  the
secondary clarifier  so that a predetermined fraction of  the
plant  effluent  could be passed through  the breakpoint system.
V-notch  weirs  in  the splitter  box were  designed to pass  30
percent of the plant flow to  the breakpoint  system (Figure 4-2).
That  percentage was  observed to be  constant across  the  full
range  of  plant  flow  rates.   A capacitance probe, visible  in
Figure 4-2, was  installed to sense liquid level upstream  of the
weir  and  electronic  components were  provided to  perform  flow
computations.

     Two existing chlorinators  were used  in  the breakpoint study
to  control and  meter the flow of chlorine to the breakpoint
process.   The  chlorinators  were  2,000 pound per day capacity
Wallace and Tiernan variable V-notch devices.  One was equipped
with  a 2,000 pound per  day orifice and rotameter,  the other had
a 1,000 pound per  day orifice  and rotameter.  Each had variable
vacuum and  automatic  V-notch positioner controls.   Chlorine gas
was fed directly from ton cylinders,  arranged in a dual manifold
configuration with  four  cylinders per manifold.   The automatic
transfer system  pictured in  Figure 4-3 assured an uninterrupted
flow  of chlorine  gas upon depletion of the cylinders on one
manifold.

      Sodium hydroxide was  utilized  throughout the  study  as  an
alkalinity  supplement.   It was purchased at  50 percent strength


                               13

-------
and stored  in  a tank of approximately 8,000 gal. capacity.  NaOH
freezing problems encountered during  cold weather were solved  by
insulation  of  the  storage  tank and  heating of  exposed  caustic
feed piping.   A Wallace  and Tiernan  variable speed chemial feed
pump,  equipped'  with  an SCR  drive, was  used  to provide the
required  dosages of  sodium hydroxide to the breakpoint  system.
                          -SAMPLE TAPS (TYPICAL)
                                                     PH
                                                     CONTROL
                                                     SIGNAL
                   BREAKPOINT
                   CHLORINATION
                   BY-PASS 27"»
                                                             NaOH
                                                             BULK
                                                            STORAGE
           CI2 '
           DOSAGE
           CONTROL
           SIGNAL
                                                      NaOH
                                                      FEED
                                                      PUMP
                                                  IN-LINE
                                                  MIXER
                                                  (VARIABLE
                                                  SPEED)
                                                       NaOH SYSTEM
                               FLOW SPLITTER
                               BOX
                     INJECTOR
                      WATER
                  CI2 BULK
                  STORAGE
                                                 SECONDARY
                                                 CLARIFIER
CI2GAS
             Cl 2 SYSTEM
                   Figure t-1  Process Flow Schematic


      The  50  percent  sodium  hydroxide  stream was  added  to  the
 chlorine  injector water   immediately  upstream  of  the  chlorine
 application point,  as  shown  in Figure  4-4.   The chemicals were
 introduced into  the process  flow through  a 4-inch diameter  PVC
 injector  pipe  which  extended  across  the  27-inch  diameter  in-
 fluent pipe  from the  crown  to the  invert.   Three-eighths inch
 diameter orifices were  spaced along the injector  pipe to allow a
 uniform  application  of the  100  gpm  chlorine-sodium  hydroxide
 chemical solution  to the process  influent stream.  The orifices
 provided a head-loss of about 10 feet which assured near-equal
 distribution of  the chemical solution  across the  influent pipe
 cross-section  and  provided an  increment  of   initial  mixing
 energy.
                                 14

-------
It!


  Figure 4-2   Flow Splitter Box
              15

-------
             Figure 4-3  Automatic Chlorine Transfer Piping
     A  3  hp  variable  speed mixing  device  (Figure 4-4)  was
installed  in  the process  stream immediately  downstream  of the
chemical application  point.   Two mixer  propellers  installed on
the mixer  shaft  were  capable of delivering mixing intensity, as
measured by  the  mean velocity  gradient  (G), of  up  to 1,500 to
2,000 sec"* in the chemical addition zone.

     Eight sample  taps  (extending  to  the center of  the break-
point process pipeline)  and two sampling pumps allowed selective
withdrawal of  samples from the  process pipe.  The exact contact
period  could  be  determined for  each of the samples collected by
means of the sampling system.  A photo of the  sampling system is
given in Figure  4—5.
                               16

-------
             Figure 4-4  Chlorine and NaOH Application Point

     A  sulfur dioxide  (SC>2)  dechlorination system  was  in-
stalled  to provide  emergency  dechlorination  for  the  Rancho
Cordova plant effluent when  measured  effluent  chlorine residuals
exceeded  tolerable  levels.   Breakpoint process control  was
maintained throughout the study  so  that emergency dechlorination
was required on very  few occasions.
CONTROL AND MONITORING EQUIPMENT

     Several important primary  elements  in  the overall process
control  and monitoring  system are discussed below.   Design
considerations  which can contribute to the successful operation
of  each  of the  elements  in  the control system are  given  in
Section 6, Process Control.

Free Chlorine  Residual Analyzer

     The  free  chlorine  residual  analyzer used  throughout  the
breakpoint program was a Fischer and Porter "Anachlor" unit.   A
small chemical  feed pump  provided a  continuous  feed of pH buffer
                               17

-------
and bromide to the measuring cell.  The flow of  sample  from  the
constant head  tank to  the  measuring  cell  was  controlled by  a
needle  valve.    Free  chlorine  residual was determined  ampero-
metrically from the amount of bromide  converted  to bromine under
acidic conditions in the measuring cell.
                    Figure 1-5  Sampling System

Ammonia Nitrogen Analyzer

     The  concentration  of  ammonia  nitrogen  in  the  process
influent was analyzed continuously  using  an  automatic analyzer,
Technicon Type  I.    An  alkaline phenate  colorimetric procedure
was used.

pH Monitor

     A continuous  analysis  of  breakpoint  effluent pH was  main-
tained using a Leeds and Northrup system.


LABORATORY TESTING PROCEDURES

     Chemical  and  physical  analyses  were  performed according
to Standard Methods for the Examination of Water and Wastewater3
14th  edition,  except as  noted below for specific analyses.
                               18

-------
Ammonia Nitrogen (NH3 -N)

     The ammonia nitrogen determinations performed on-site were
made using a  commercial ammonia  nitrogen probe  (Orion Research,
Inc., Model 95-10)  combined with  a pH-specific  ion meter  (Leeds
and Northrup  Model  7417).  The probe was calibrated daily using
ammonium chloride standard solutions.

     Samples  collected  and  preserved  for subsequent testing of
ammonia nitrogen at a commercial  laboratory  were analyzed, using
distillation and titration procedures  for ammonia levels greater
than 1  mg/1,  and by distillation and nesslerization at ammonia
concentrations below 1 mg/1.

Nitrate Nitrogen (NC>3 -N)

     Nitrate  analyses were  performed according to  brucine
procedures, except for  chlorine injector  water samples which
interfered with  the brucine  analysis.   Injector water samples
were analyzed using cadmium reduction  techniques.

Chlorine Species

     Both  free  and combined chlorine  residual  species were
analyzed routinely throughout  the breakpoint  study  using  a
Wallace  and   Tiernan  Amperometric Titrater.    lodometric tech-
niques were used to measure the high chlorine residuals  found in
samples of the chlorine injector water.
                               19

-------
                            SECTION  5

                     RESULTS AND DISCUSSION
     Breakpoint chlorination testing was  carried  out at Rancho
Cordova,  California,  during a three-month period,  from mid-
December, 1975 through March 15, 1976.  Test data were collected
to  provide  insights  into process  chemistry and control.   The
system was operated  continuously,  24 hours  per day,  5 days per
week.  Technicians  were  present at all times to collect opera-
ting data and adjust the system as needed.  Specific results of
the  breakpoint  chlorination testing program are  presented and
discussed below and  presented  in  Appendix A.   Additional data
from continuous recording  devices, too bulky  for  inclusion in
this report,  are retained on file by the engineers.


PROCESS INFLUENT CHARACTERISTICS

     The breakpoint chlorination system at Rancho Cordova func-
tioned on-line, with full  effluent  flow and  chemical  quality
variations.    Diurnal variations  in  flow  caused  breakpoint
process  flow  rates to  vary from  about 0.1  mgd  to 1.2  mgd.
Ammonia nitrogen concentrations were also observed to vary on a
diurnal basis over a range of about 15  rog/1 to  25 mg/1.  Ammonia
mass flow, the product of  process  flow  and ammonia concentra-
tion, was found to  vary  on  a typical day  according to the plot
of Figure 5-1.  As therein indicated,  the ratio of ammonia mass
rate maximum  to minimum was  typically about 8 to 1.

     Rancho  Cordova  Wastewater Treatment Plant  effluent  is  a
high  quality  secondary effluent, as evidenced  by the  data
summarized in Table 5-1.  The concentrations of nitrogen species
and  the  average  alkalinity of  165 mg/1 are  typical  of  non-
nitrified secondary effluent.


INITIAL MIXING

     A series  of  tests were conducted in which  the mechanical
mixing intensity at the point of chemical application was varied
and effects  upon  process performance  were monitored.   None of
the  observations  showed  any  effect upon  chemical consumption
or  effluent  quality as  the mechanical mixing intensity was
increased from zero (mixer turned  off)  to the maximum (G - 1500
                               20

-------
to  2000 sec 1)  level attainable.  With  the mechanical mixer
off, the  free  residual analyzer  sensed  somewhat greater varia-
tions  in  free  chlorine  residual  as  compared to  that  with the
mixer  on  full  speed; however,  the mid-point in the oscillation
was  the same in  both cases.   While  the increased amplitude of
the  free  chlorine  residual  excursions  did  not  affect chemical
consumption  or  effluent  quality, process control functions were
made more  complicated and,  for  this  reason, mechanical mixing
was employed throughout the  study.
   2.0
   1.5
 5  1.0
 o
   0.5



                                                    3l
                                 2C

                                                          -i 200
                                        150
                     £

                     U

                     "•
                     J

                     2
                     :
                     S
                                                               •
100 eg
                                                               :

                                                          -I 0
     M
            0400
0800
        1200

        TIME
1600
                                           2000
                                                   ,-:
                 Figure 5-1  Process Influent Variations
     From the  Rancho Cordova data, the total dosage of  chlorine
necessary to  reach breakpoint is not measurably affected by  the
intensity of  initial mixing.   However,  process control  consid-
erations dictate that segregation  of  the  reactants  be minimized.
For  design  purposes, sufficient  hydraulic or mechanical energy
should be provided to facilitate rapid and thorough blending of
the  chlorine  solution,  pH  adjustment   chemical  and  process
influent.    Rapid  mix design criteria  from  water  treatment
practice should serve as a design  guideline.
                                21

-------
      Table 5-1.  Secondary Effluent Quality, Rancho Cordova, California
Parameter
BOD
COD
PH
Alkalinity (as CaCOJ
NH| (as N)
NOg (as N)
NOg (as N)
Total organic nitrogen (as N)
Soluble organic nitrogen (as N)
Turbidity
Average value
8 mg/1
32 mg/1
7.2
165 mg/1
20.6 mg/1
0 . 6 mg/1
0 . 8 mg/1
2 . 4 mg/1
1.2 mg/1
4 JTU
Range
6-10 mg/1
11-45 mg/1
7.0-7.4
150-175 mg/1
15-25 mg/1
0.01-1.4 mg/1
0.07-1.4 mg/1
1-5 mg/1
1-2 mg/1
2-10 JTU
REACTION RATES

     The time required for the breakpoint reactions to reduce  to
a minimum the ammonia nitrogen  in solution depends upon the
initial concentration of  ammonia,  chlorine  dosage,  pH,  alkalin-
ity  and  temperature.   The consistent  chemical  quality of the
Rancho Cordova  effluent  did not  permit testing  of all  process
variables  across  a wide  range  of  conditions.   Test data were
collected  which  clearly  documented the  effect  of  pH set point
(final system pH) upon the rate of reaction.

     Figures  5-2 through  5-9  present  breakpoint  chlorination
reaction  rate data  collected  at Rancho Cordova  for final  pH
values between pH  6.5 and pH 8.5.   For each test, the  control
system was set to provide a free residual chlorine  concentration
in the process  effluent of 12  mg/1.    This  was  done to ensure
sufficient chlorine  residual  in the combined  breakpoint and
nonbreakpoint stream to facilitate  adequate disinfection  of
the  plant  effluent.    Influent  ammonia nitrogen concentrations
exceeded  15  mg/1  except in two test  periods (pH set point 7.3
and  7.7)  when the  breakpoint effluent  used  as  injector water
diluted  the influent ammonia  concentration  at  low flows  to
concentrations of about 9 mg/1 to  15 mg/1.   The  implications  of
the chlorine injector water source are discussed  in a  subsequent
section of this  report.

     The  reaction rate  data are  summarized  in  Table  5-2.   The
rate of the breakpoint reaction was observed to  be  fastest at a
pH set point  (final pH)  of 7.0.  The reaction rate slowed con-
siderably  at pH  set  point 6.5,  with gradual reductions  in rate
observed  across  a  range  of pH  set points from  pH 7.3 to 8.5.
                               22

-------
 0
LU
DC
 CO
z
z
o
QC
LU
Q.
     100
      50
      20
      10
        10
           INITIAL CONDITIONS
           CI2-.NH3-N  -  10.4-12.2 TO 1
           NH3-N    =  18.2 TO 20.3 mg/l
           TEMP     =  ABOUT 15° C
           pH       =  7.4
           ALKALINITY =  165 mg/l AS CaCOj
           NaOH     =  104 TO 122 mg/l
                                                         *    *
                                                                       *   *
NOTE
FREE RESIDUAL SET POINT - 12 mg/l
BREAKPOINT EFFLUENT AS INJECTOR WATER
                   20       30    40  50  60  70 80   100

                                    TIME, SECONDS
                                                                              200
                               300
    100
           Figure  5-2   Ammonia Removal with Breakpoint Chlorination
                          in a Pipe Reactor - pH Set Point 6.5
1
LU

-------
    100

C3    50
111
tc.
 co
I
Z
u
ec
ui
Q.
     20
      10
       10
              INITIAL CONDITIONS
              CI2:NH3-N  =  9.BO-11.2TO1
              NH3-N    =  14.0 TO 19.0 mq/l
              TEMP     =  ABOUT 15° C
              pH      =  7.4
              ALKALINITY-  165 mg/l AS Ca COj
              NaOH     =  147 TO 198 mg/l
                                                       NOTE
                                                       FREE RESIDUAL SET POINT = 12 mg/1

                                                       SECONDARY EFFLUENT AS INJECTOR WATER
                                                                       **   **
                      20       30    40   50  60  70 80  100

                                       TIME, SECONDS
                                                                          200      300
       Figure 5-4  Ammonia  Removal with Breakpoint Chlorination
                      in a  Pipe  Reactor - pH Set Point 7.0
Z
z
<
LJJ

-------
o
z
z
tr
z
 co
I
Z
H
UJ
O
cc
    100
     50
     20
10
       10
                                                  FREE RESIDUAL SET POINT = 12 mo/I
                                                  SECONDARY EFFLUENT AS INJECTOR WATER
          INITIAL CONDITIONS
         NH3-N
         TEMP
         pH
         ALKALINITY
         NaOH
= 9.55-11.1 TO 1
• 16 TO 18 mg/l
= ABOUT 15° C
- 7.4
- 165 mg/l AS CaCOj
- 163 TO 201 mg/l
                 20       30    40   50  60  70 80  100

                                  TIME, SECONDS
                                                                           200      300
        Figure 5-6   Ammonia Removal  with  Breakpoint Chlorination
                       in a Pipe Reactor - pH Set Point 7.5
    100
z
z
<
UJ
CC
 f)
I
Z
H
LU
O
tc
UJ
0.
            INITIAL CONDITIONS
            CI2:NH3-N   = 10.8 TO 12.6 TO 1
      NH3-N     = 8.6 TO 15.0 ma/I
      TEMP      = ABOUT 15° C
      pH       = 7.4
      ALKALINITY = 165 mg/l AS CaCO3
      NaOH      = 112 TO 152 mg/l
                                    FREE RESIDUAL CHLORINE SET POINT' 12 mg/l  „.
                                    BREAKPOINT EFFLUENT AS INJECTOR WATER
                                      40   50  60 70 80   100

                                       TIME, SECONDS
                                                                     200      300
       Figure 5-7  Ammonia Removal with Breakpoint Chlorination
                      in a Pipe Reactor - pH Set Point 7.7
                                          25

-------
    100
O    50
z
z
ui
oc
 m
I
UJ
U
OC
UJ
0.
20



10



 5
       10
             INITIAL CONDITIONS
                   9.5 TO 11. 2TO1

                   18.STO21 ms/l

                   ABOUT 15° C

                   7.4

         ALKALINITY -  166 mgll AS CaCO3

         NaOH     -  212TO244mg/l
C12:NH3-N

NH3-N

TEMP

pH
                                        FREE RESIDUAL SET POINT-12 ma/I   *


                                        SECONDARY EFFLUENT AS INJECTOR WATER
                                                           *   *
                   20       30    40  50  60 70 80   100


                                   TIME. SECONDS
                                                                          200
                                                                      300
       Figure 5-8   Ammonia Removal with Breakpoint Chlorination
                      in  a Pipe Reactor - pH Set Point 8.0
UJ
oc
 CO
I
UJ
U
OC
100



 50




 20



 10



  5
       10
INITIAL CONDITIONS

        -  11.1-12.3 TO 1

        *  16.8 TO 17.5 mg/1

TEMP     -  ABOUT 15° C

pH       •=  74

ALKALINITY -  165 mg/1 AS CaCOj

        -  166 TO 180 n<9/l
                                          NOTE                   #

                                          FREE RESIDUAL SET POINT - 12 mgll


                                          BREAKPOINT EFFLUENT AS INJECTOR WATER
                   20       30    40  50  60 70 80   100


                                   TIME, SECONDS
                                                                          200
                                                                      300
        Figure 5-9  Ammonia Removal with Breakpoint Chlorination
                      in a Pipe Reactor - pH Set Point  8.5
                                           26

-------
                Table 5-2.   Reaction Rate Data Summary
pH set
point
6.5
7.0
-7.0
7.3
7.5
7.7
8.0
8.5
Injector
water source
BP
SE
SE
BP
SE
BP
SE
BP
Influent
NH+-N cone.
18.2-20.3 mg/1
19.5-22.5 mg/1
14-19 mg/1
9.5-15.5 mg/1
16-18 mg/1
8.6-15 mg/1
18.5-21 mg/1
16.8-17.5 mg/1
Estimated time to
completion of
breakpoint reaction
> 200 sec
60 sec
90 sec
130 sec
150 sec
180 sec
200 sec
> 200 sec
     1
     BP = breakpoint effluent
     SE = secondary effluent
NITROGENOUS RESIDUALS

     Nitrogenous  residuals  formed  in  breakpoint  chlorinatj.on
include  both nitrogen trichloride (NC13)  and nitrate  (N03).
Nitrogen trichloride may be  formed according to Equation  6,
although Saunier, et al.4 noted that reaction kinetics suggested
that formation occurred through a more complex chemical pathway.
Nitrate  is formed  in breakpoint chlorination through  the oxida-
tion of  nitrite  (NO2)  according to Equation 13, or by oxidation
of ammonia as  shown in Equation 12  (see  Table  3-1).   Formation
of nitrogen  trichloride  or nitrate  is  undesirable, since large
quantities of  chlorine may  be consumed,  nitrogenous residuals
persist  in solution which reduce the nitrogen removal  capability
of the process, and obnoxious odors can result.

Nitrogen Trichloride

     The  residual  ammonia  concentration  achieved  following
breakpoint chlorination  was  shown  by  Saunier,  et al.^  to  be a
function of  nitrogen trichloride formation, since NC13  is  con-
verted to  ammonia  upon  dechlorination  with sulfite.   Nitrogen
trichloride  is  formed in breakpoint chlorination  and,  since  it
has a  relatively slow decay  rate, the amount  formed serves  to
limit  the  degree of  ammonia removal which can be attained  in
breakpoint chlorination.
                               27

-------
     The  amount of  "apparent"  ammonia  nitrogen  remaining in
solution  (see  Table 5-3)  following  breakpoint chlorination was
assumed in this  study  to be  a measure of the amount of nitrogen
trichloride  in  solution.    Nitrogen trichloride  formation did
not appear to  be sensitive to the pH set point (final pH), with
values generally observed  to be  in the range of 0.3 mg/1  to 0.7
mg/1.   This is  contrary  to the observations of  Pressley, et
al.5  and Wei6 who noted  a  strong pH  dependency  for nitrogen
trichloride production.  In a  recent study by Sa.unier4 some
pH  dependency  was noted, but  not  to  the  degree  reported by
Pressley, et al.,  and Wei.
          Table 5-3.  Formation of Nitrogen Trichloride and Nitrate

pH set point
6.5
7.0
7.0
7.3
7.5
7.7
8.0
8.5

Injector
water source
BP
SE
SE
BP
SE
BP
SE
BP
NClg formed.
mg/1 as N
Mean
_
0.43
0.65
0.34
0.57
0.52
0.73
_b
Standard
deviation
_
0.15
0.14
0.12
0.13
0.27
0.22
_b
~" C
NO, formed ,
mg/1 as N
Mean
0.95
-
1.14
0.93
1.18
1.19
0.78
1.25d
Standard
deviation
0.18
-
0.17
0.27
0.14
0.31
0.19
-
   BP = breakpoint effluent.
   SE = secondary effluent.
   Reaction had not progressed to completion in samples collected.
  °This category includes NO. formed from NH . , does not include NO, formed from NO .
   One data point.


      A careful analysis of the nitrogen trichloride data showed
 that the  chlorine  injector water  source influenced the amount
 produced  in the  Rancho Cordova breakpoint  system.   A  detailed
 discussion of  the effect of injector water  on  that  formation is
 given in  a subsequent  section  of  this report.

      One  possible  explanation for the  apparent  discrepancy in
 observations  on nitrogen trichloride  production  is that  the
 choice of sampling  and  analysis  procedures may  be  important in
 the results  obtained.   Nitrogen trichloride  is  known  to be
 soluble to a  very  limited degree in water.    Its  characteris-
 tically  pungent odor was noticeable  at times in the  sample
 building  at Rancho Cordova.  However, any which escaped from the
 breakpoint process  in a gaseous form was not accounted for in
 any of the analytical measurements.  Loss in  gaseous form could
                                28

-------
have occurred inside  the breakpoint contact  pipe.  Nitrogen
gas  from the breakpont  reaction  may have  stripped nitrogen
trichloride from solution as  it  precipitated from  solution  and
migrated, as  bubbles,  toward the  top of  the pipe.  The  milky
appearance of samples pumped  from the breakpoint system  attested
to the presence of gas bubbles in the breakpoint stream.   Also,
NCI3 could have escaped from solution as the samples were  being
taken  at the sample tap.   In  any case, it is possible that
gaseous losses of NC13 in the Rancho Cordova system contributed
to the apparent  discrepancy  in observations on NC13 formation vs
system pH.

     Another potential source of disagreement  in  NC13  observa-
tions  is that  different  analytical techniques give different
degrees  of accuracy and precision.  The  DPD-FAS method  was
observed by Pressley,  et al.5  to  have  poor reproducibility,
whereas  the spectrophotometric method employed  in the Blue
Plains  work was  of questionable  accuracy  due to the carbon
tetrachloride  extraction  and   potential  for  interferences.

     The amount of  chlorine  residual  consumed  in  the oxidation
of  NH4 to NC13 was found to be about 4 to 6  percent of  the
total chlorine dose across a pH  range of pH 7 to 8.  A  complete
data  summary  of  the chlorine  consumption of  the chemical
reactions  encountered  in breakpoint  chlorination  is given  in
Appendix A.

     From the NC13  data and  other  qualitative observations made
at  Rancho  Cordova,  it  appears  that breakpoint systems may  be
designed without  any  provision  for  off-gas  collection  and
treatment  (for  NC13  odors)  if proper attention  is  given  to  the
design of  chlorine  dosage and pH  control systems.   A  complete
discussion of factors  to consider   in the design  of breakpoint
process control  is given in  Section 6.

Nitrate

     Data  on NC>3 formation  in   the  Rancho  Cordova breakpoint
system are  summarized  in Table   5-3.  It  should be noted, how-
ever,  that  the  data of Table 5-3 show the concentration  of  NC>3
formed from NH|, and do not  include that  formed  from NC>2.   High-
ly variable N02 concentrations in the breakpoint influent  stream
tended to  obscure  the rather consistent pattern of  NC>3  forma-
tion which  was  observed  to  occur  through  the  chemical  pathway
from  NH4.   A more  complete presentation of  NC>3 data  is  in
Appendix A.

     Nitrate formation was  not  observed to be  pH  dependent  in
this study.   This is consistent with the experimental  observa-
tions and breakpoint mathematical model predictions of Saunier4,
but  it does not  agree with the findings of Pressley, et al.5
                               29

-------
        e^ amount  of  chlorine consumed  in the  formation  of N03
from  NH4 was significant,  varying  from about  8 percent  to 19
percent of the  total chlorine  dosage.   As such,  it represented
the  second  largest  chlorine consumption  of  the  chemical reac-
tions  identified during the  study.

Organic Nitrogen

     Analysis  of  data collected  to determine  the effect of
breakpoint  chlorination on  organic  nitrogen  compounds  showed
little  change  in  organic nitrogen  concentrations   through  the
process.  A  statistical summary  of total organic nitrogen data
is given in  Figure 5-10.
    4.5
    4.0
3

1
Ul

I
z
o
cc
o
3.0


2.5


2.0


1.5


1.0


OJ5 \-
MEDIAN = 22 mg/l
                                          BREAKPOINT
                                          EFFLUENT
                                      BREAKPOINT
                                      INFLUENT
                                   _L
                                   _L
      0.01
                     10    30  SO  70    90


                    PROBABILITY OF LOWER VALUE, %
                                           99
                                                               99.99
       Figure 5-10 Effect of Breakpoint Chlorination on Organic Nitrogen


OVERALL CHLORINE CONSUMPTION

     The  total  chlorine dosage necessary to maintain  breakpoint
chlorination  at Rancho Cordova  averaged  10.0  mg/l chlorine  for
each 1.0 mg/l of influent   ammonia nitrogen  (Cl2:NH4 -N =  10:1).
                                30

-------
The  statistical distribution  of Cl2:NH4  -N values  observed
at Rancho Cordova is presented  in Figure 5-11.  This value  may
be compared  to  that  predicted  by the  stoichiometry  of the  break-
point  reaction  (Equation  9),   which predicts  only  7.6 mg/l
chlorine  consumed for each 1.0 mg/1  ammonia.   The  difference in
chlorine  consumption between  the  theoretical  and  actual  value
may be attributed largely to  other chemical  reactions  (Equations
11-1^2) which occur between chlorine  and  ammonia  to form nitrate
(NO3-)  and  nitrogen  trichloride (NC13),  as well as  the free
chlorine residual remaining in solution.  A presentation of
Rancho  Cordova  data  showing  the proportionate  chlorine  demand
for  each of  the breakpoint reactions  is given  in Figure  5-12.
    12.0
    11.0
    10.0
CO
X
Z
    9.0
    8.0
          pH SET POINT BETWEEN pH 7 AND 8
          INITIAL NH3 -N 8 mg/l TO 22.5 mg/l
          DATA POINTS ADJUSTED TO FREE CHLORINE
          RESIDUAL OF 8 mg/l
          MEDIAN = 10.0
                               _L
               _L
      0.01
 10    30   50   70     90

PROBABILITY OF LOWER VALUE, %
99
99.99
             Figure 5-11 Summary of CL: NH^-N Observations


ALKALINITY  SUPPLEMENTATION

      An  analysis of the quantity of  sodium  hydroxide  fed  to  the
breakpoint  system for  purposes of  pH control  showed that  an
average  of  1.53 l.b NaOH were  added for each 1.0 Ib chlorine  to
maintain the breakpoint system pH  between pH 7.0 and  7.5.   The
                                31

-------
statistical distribution  on the NaOH dosage data  is presented in
Figure 5-13. The value of  1.53  Ib  NaOH/lb Cl2 compares favorably
with  the  1.50 Ib  NaOH/lb C\2  requirement predicted  from stoi-
chiometry  (Equation 19).   It should be noted  that if lime (CaO)
were  used  for pH  control,  chemical stoichiometry (Equation  18)
predicts a requirement of  1.05  Ib  CaO/lb
 pH SET POINT = 7-8
 INITIAL NH3 = 8-225
(EQUATION 9)
66-73%
                                    (EQUATION 11)
                                    8-19%
                                                    NH3 -~ NCI3
                                                    (EQUATION 10)
                                                    46%

                                                   UNIDENTIFIED
                                                   0-8%

                                                      ^ NO3
                                                  (EQUATION 12)
                                                  0-2%

                                             FREE RESIDUAL REMAINING
                                             5-10%
         Figure 5-12 Breakpoint Chlorination Chlorine Consumption
CHLORINE INJECTOR WATER CONSIDERATIONS

     When  chlorine'gas is used as the  chlorine  source in break-
point chlorination,  the volume  and source (chemical composition)
of the injector water  may become  important in process operation.
Ordinary chlorine  injector  design  criteria  restricts  the chlo-
rine concentration  in  the  injector  water to  3,500  mg/1  or less
under maximum feed  conditions.   As  a  result, the  high chlorine
application  rates   in breakpoint  chlorination  can cause  the
injector water flow to be   a significant fraction  of  the total
process stream at  low breakpoint process  flow rates.  If break-
point  process influent  is used as  injector water, reactions
between chlorine  and  ammonia which  occur in  the  injector water
can consume  chlorine  in  undesirable  side reactions.   If break-
point process effluent  is  used as  injector  water,  the reacting
                               32

-------
constituents are diluted and  reaction rates reduced, though the
reduction would not be noticeable in most  cases.
     1.80 -
  CM
 U
 £
                      TO MAINTAIN pH BETWEEN 7.0 and 7.50
       0.01
 10    30   50   70    90

PROBABILITY OF LOWER VALUE, %
                                                              99.99
                  Figure 5-13
   Lb NaOH/Lb CI2 Added
An  analysis  of Rancho  Cordova  data collected at pH set  points
between 7.0 and  8.0 showed  that approximately 0.2 mg/1 more
NC13 was  present  in the process  stream  when  secondary  effluent
(breakpoint influent) was  used as  injector water than when
breakpoint  effluent was used  for chlorine injection.    Contact
time in the injector system was about 20 sec.

     Samples of injector water  were collected just  prior  to  and
following NaOH additions.   Some  of  the  samples  were vigorously
agitated  during  collection in  an effort to drive off  any NC13
which  had "formed, so  that  the  amount  of NC13 formed in  the
chlorine  injector water  could  be estimated.   The  sampling
techniques and analysis of  injector water  samples confirmed
that NC13 was  formed  in  significant  amounts in the secondary
effluent  used  as  injector  water.   Nitrate concentrations  in  the
injector water were not appreciably increased.
                                33

-------
     Based  upon  the Rancho  Cordova  testing,  chlorine  injector
water in breakpoint chlorination applications  should be  designed
according to the following criteria:

     •    If possible,  locate  the  chlorine injector at the
          chemical application point  and  use process influent as
          injector water.

     •    If the  injector must  be remote from the actual appli-
          cation point, use  process  effluent  as  injector water.

     •    In large  installations, it may be  preferable to feed
          liquid chlorine (rather than gaseous), mixed  with the
          alkalinity supplement  chemical and  sufficient process
          effluent to  yield  a maximum chlorine  residual in the
          feed stream of 8,000 mg/1.
BREAKPOINT MODEL PREDICTIONS

     As  a supplementary part  of this  study, the  breakpoint
chlorination computer model  developed  at  the University of
California by Saunier  and Selleck^ was used to predict process
performance under a  variety  of  conditions, including many which
could  not have  been tested under actual field conditions at
Rancho  Cordova.   Those predictions  are  given  in Appendix B of
this report.

     Figures B-l through  B-8 represent breakpoint model predic-
tions under the exact conditions of the field-scale measurements
presented  in  Figures 5-2  through 5-9.   As such, the curves can
be  compared  directly to  determine the correlation between the
model  and field measurements.    In most  cases, the  correlation
was excellent.

     Figures  B-9 through B-14  show the effect of  alkalinity
supplement addition  (in this case,  NaOH)  on final pH, percent
ammonia remaining in solution and reaction time as  a  function of
C1-2LNH3~  and_C>p.  C-j is the  total  inorganic  carbon (sum of C02,
HCC>3  and CO-j)  in  the  system,  expressed in moles  per liter.
The alkalinity  supplement addition ("NaOH added")  in each case
refers  to the NaOH  added  to solution beyond that required for
neutralization of  the  acidity  from  hydrolysis  and dissociation
of the chlorine dosage (Equation 16,  Table 3-3).
                               34

-------
                            SECTION  6

                         PROCESS  CONTROL
     A breakpoint chlorination  control system should be reliable
and should  facilitate  the efficient use of  chemicals  and  man-
power.   Chlorine overdosages can  be  quite  expensive,  with
additional costs for chlorine  and alkalinity supplement as  well
as dechlorinating  agent.   Chlorine  and  alkalinity  costs  alone
can reach $800 per year  for  an overdosage  of  1 mg/1  in a  1 mgd
application.   Chlorine underdosages can result  in  significant
reductions  in ammonia removal  efficiency  and high  combined
chlorine residuals in the process  effluent.

     The nature  of the breakpoint chlorination process is  such
that  manual  control  is  difficult,  requiring  the  undivided
attention of an  operator.  One  aspect  of this study  was the
evaluation  of  various  process control  systems  and  associated
instrumentation.


PROCESS AND COMPONENT DESCRIPTIONS

     The  breakpoint  chlorination  process  flow  diagram  is
presented on Figure 6-1 and shows  only those signals involved in
control.  The pilot  plant  control panel is shown on Figure  6-2.
Principal signals  generated  and  displayed are  listed  in  Table
6-1.

     The  final  control  elements  include  the  chlorinators and
sodium  hydroxide  feeder, and the  measured variables include
ammonia,  free chlorine residual,  and  pH  analyzers  and  flow-
meters .

Ammonia Analyzer

     This  device  was  a  continuous  on-line  analyzer which
utilized metered  reagents in the colorimetric determination of
ammonia.   The calibrated  accuracy  was  ±0.2 mg/1.   Occasional
drift was observed to  be  as  high  as  2 mg/1 over a 12-hr period.
"Noise" amplitudes as  high as  3 mg/1, principally due to sample
turbidity or occasional air bubbles,  were observed.  Disturbance
durations ranged  from seconds to  several minutes.   A sample
filter was not used.   Total dead  time for the analyzer was about
17 min,  consisting primarily of  analysis  time.  (Dead time, in


                               35

-------
process  control,  is the period  of  time after a  force  is applied
during which  no response is observed;  it  is  the  lag time between
action and response.)
    DATA
    ONLY
               JM-Smgd
2-12 pH
     i
0-30 m»/l
   AMMONIA-
   NITROGEN
   ANALYZER
          PROBE
    UNTREATED STREAM

    0 = 0.25-1.25 ingd
    NH^CONC. 16-26 mgfl
                        2000 LB/DAY
                        CHLOR1NATOR
                               4500 LB/DAY
                               SODIUM HYDROXIDE
                               FEEDER
                                4 INJECTOR WATER FROM TREATED STREAM
                    Figure 6-1  Simplified Process Diagram
Flowmeter

     The  flowmeter  consisted of  a  level-measuring  capacitance
probe and appropriate circuits to convert level  measurement to a
linear  flow signal.   Calibrated  accuracy was  ±1 percent  which
was maintained for periods  from one to  several days.   Zero and
span  drifts up  to 0.2  mgd were  observed.    These were  usually
corrected  by cleaning  the probe  of grease  accumulations.   The
response  of  the  flowmeter was  essentially  instantaneous.
Hydraulic  noise  oscillations with amplitudes  from  7 to  10
percent at  periods  from about 5 to 10  sec were  transmitted.
This  noise  was eliminated by  appropriate electrical damping
circuits.
Chlorinators
     Chlorine  was added to  the  process by vacuum  injection into
  liquid  stream prior to the  mixing chamber.  Chlorine  gas flow
                                36

-------
was  controlled  by  a  linear  variable  orifice positioner in
response  to  signals from  the automatic  controls.   A  constant
but adjustable  differential  pressure was maintained across  the
linear variable orifice.  All observations were  based on flow as
displayed on  the  gas  flow rotameter.   Repeatability was  about
±1 percent.    Maximum  deviation  from  linearity  over  the  range
from 15 to 90  percent  output  was about  1-1/2  percent.   Response
dead band  was about  1 percent.    Full-scale  output change  re-
quired less than 1 minute.

                      Figure 6-2  Control Panel


     Electric  linear  flow  transmitters  were provided  which
operated with a maximum deviation from linearity of  about 4 per-
cent when used from 25 to 100 percent full-scale.

     The chlorinators were  also provided with  dosage  (ratio
control) and vacuum  controls.   These controls were  immobilized
after calibration  because  of the non-linearity they induced in
the response of the variable  orifice control.


                               37

-------
Free Chlorine Analyzer

     This analyzer was a continuous amperometric titration unit,
providing an  electric signal proportional to  free chlorine
residual.  The  analyzer  maintained  an accuracy better  than
±1 mg/1  over  a period  of  several months  and  was free  of  spu-
rious  noise.   The  dead time  for  this analysis consisted  of  a
plant  flow variant  of  12  to 60 sec with an  additional  2 minute
fixed  time for  transit  from the  sampling point through  the
analyzer.

             Table 6-1.  Summary of Process Control Signals
Signal
Free chlorine residual concentration
Total chlorine residual concentration
Sodium hydroxide feed rate
Plant flow (influent)
PH
Gas flow, chlorinator No. 1
Gas flow, chlorinator No. 2
NHj-N concentration
NH^-N flow rate
Total chlorine gas flow rate
Total CL.2 gas flow rate/NHt-N flow rate
Units
mg/1
mg/1
Ibs/day
mgd
2 to 12
Ibs/day
Ibs/day
mg/1
Ibs/day
Ibs/day
(none)
Notes
(M)
(M)
(M)(I)(T)
(M)(T)
(M)
(M)
(M)
(M)
(C)(T)
(C)(T)
(C)
      All of the above signals were recorded
      (T) Integrated - totalized
      (C) Computed
      (M) Measured by transducer
      (I) Inferred from speed measure
     The useful range of the analyzer  is  limited  at  the  low end
to values  about 3 to 4  mg/1  because of  interference  from  com-
bined  residuals  below breakpoint.   This in  turn  limits  the
minimum reliable process control set point  to  values of  about 7
to 9 mg/1  of free chlorine.  These  erroneous  readings appeared
to be consistent and repeatable and were equal  to the sum of the
free chlorine plus about 1/10 of the combined chlorine.

     It is  important  that  system design provide sufficient  dead
time at maximum plant  flow rate to allow the  reaction to go to
completion  before  entering the  analyzer;  however,  it  is  also
important  that this dead  time be  minimized  because  control
                               38

-------
difficulties increase with  dead  time.   Therefore,  the chlorine
residual analyzer  was  located at a  distance  downstream  of  the
mixing chamber such that the travel  time  between the chamber and
analyzer equaled the maximum reaction time.

pH Analyzer

     The pH analyzer consisted of a  glass measuring electrode, a
reference electrode, and a thermistor temperature sensor mounted
on a. flow-through  electrode holder.   A  signal preamplifier was
externally mounted  on the  electrode  holder to provide a contin-
uous high-level temperature-compensated pH signal.

     The analyzer  was accurate to 0.1 pH  units over a period of
several months  and  was  free of spurious  noise.  The accuracy of
the  pH signals was unaffected  by  the  chlorine dosage  either
above or below breakpoint concentrations.

     The dead  time on  the analysis consisted of  a  plant flow
variant of 12 to 60 sec with an additional 20  sec fixed time for
transit from  the  sampling point  through  the  analyzer electrode
holder.  System design  considerations for pH  analyzer dead time
are the same as those for the Free Chlorine Analyzer.

Sodium Hydroxide Feeder

     The sodium hydroxide  feed  system  consisted  of  a metering
pump which fed a 50 percent solution from a liquid  storage tank.
The  metering  pump  was a positive displacement unit with manual
pump  stroke  control,  and powered by  a variable-speed  d-c motor.
Feed  control was  achieved  by automatic motor  speed variation in
response to signals from  the automatic  control  system.  The
local motor  speed controller was  a  closed-loop system that
compared motor speed  as  measured  by  a tachometer with the
incoming automatic  speed order.

      Initial calibrated  accuracy from 5  to 100 percent of output
was  ±1 percent.    Some instability was observed for  outputs
below  5 to 6 percent.  An output zero drift of about  ±10  percent
was  observed  after about 6 weeks of  operation at which time  the
equipment  was  recalibrated  and  no  further drift was observed.

      System response was on the order of  2 sec for  5  to 100 per-
cent  without overshoot.


CONTROL REQUIREMENTS

      The  principal  parameters  for  breakpoint  chlorination
control  are  free   chlorine  residual  for  chlorine dosage  control
and  pH alkalinity supplement dosage control.   The degree of
resolution  and repeatability of  chemical feed  rates is  one of


                               39

-------
the major  factors  which determines how  efficiently  the  break-
point process can be controlled.

     At  Rancho  Cordova, the  chlorine  feed  requirement  varied
from  160 Ib/day  to 2,300 Ib/day.   Control  of free chlorine
residual to  ±3  mg/1 required a control tolerance of  about
±3 percent  at minimum  chlorine feed  rates,  and ±1  percent  at
maximum  feed rates.  The  resolution of  the chlorinators  was
about  1  percent,  or about  10 Ib/day  on the  small unit  and
20 Ib/day  on the large one.   At a minimum  flow of 0.25 mgd,
a  feed-rate error  of 10  Ib/day  would lead  to free chlorine
variations of about 4 mg/1.    At  maximum flows of 1.25  mgd,  a
feed-rate  error  of  10  Ib/day  gives  an  error of about  1 mg/1.

     Previous discussion of the  chemistry indicates  that about
1.5 pounds  of sodium hydroxide is required per  pound  of chlo-
rine.  This ratio is approximate and valid only if the untreated
stream pH  is near  the  desired reaction  end point  (i.e.,  all
breakpoint acidity  neutralized).   The slope  of  the  alkalinity
curve at the  control  point determines the difficulty  in obtain-
ing proper control.   A  family  of  alkalinity  curves is  shown  on
Figure 6-3.   pH  control would  be  more  difficult for  wastewater
with low alkalinity.  The abscissas of these  curves in relation
to the pH  control point could be interpreted  as error  of con-
centration or error of  sodium hydroxide  feed  rate.   At  Rancho
Cordova the minimum feed rate was  about 250 Ib/day, with maximum
feed rates reaching about 3,450 Ib/day.

     If the pH/feed-rate curve  is  very nonlinear, a special non-
linear element should be  inserted into  the measurement  loop to
hold constant gain at all feed  rates.


CONTROL SYSTEMS

     Four systems were evaluated  for their effect on  breakpoint
chlorination control:  simple  feedback control, simple feedback
plus  flow paced control,  flow  modified feedback  plus flow
paced control, NH4 -N mass  flow paced control.

Simple Feedback Control

     Figures 6-4 and  6-5 show  a simple feedback control system.
This control  system functions  solely  in  response to  deviations
of free  chlorine  residual  from the  selected  set point.   Figure
6-5 shows  a  typical response (output) curve to an error signal.
If the controller  generates an output  in response  to a process
error signal  that  will  overcorrect  to the extent  that  the  new
error is equal to or greater than the original eror and the  new
error  is  of  opposite  sign,   then  the  output will  violently
oscillate.  In terms of feed rate, if  the  analysis showed a
deficiency of 10  Ib/day and  this  provided a  positive change in
output of 20 Ib/day or greater, then the system would oscillate.


                               40

-------
                               MODERATE ALKALINITY
                      FEED RATE OF ALKALINITY SUPPLEMENT
                 Figure 6-3  Typical pH Characteristics

     Referring  again  to Figure  6-5,  the output is  continually
changing as  long as error  is  sensed  by the controller.  Since
the output will  continue to  change  until the  results of the
initial change are analyzed and returned to the controller, this
effect must also be evaluated.
     A  conservative  initial  system design will  provide for
approximately half the  feed  rate  correction  to  be made with the
proportional band setting immediately (-100%  x % ERROR, refer to
                                                % PB
Figure  6-4) with  the remaining  correction  to be  made  over a
period  of  time  equal to twice  the  system dead  time.  Thus, an
error  signal indicating a  deficiency of 10 Ib/day feed rate
would  provide  an  increase  in  output of  5  Ib/day  immediately
with the  remaining 5 Ib/day to be  incremented  over  a period
time equal to twice the dead time.
of
     A plant flow rate of 0.25 mgd with a free chlorine  residual
error of  1.0 mg/1 requires  a correction  of  2.08  Ib/day  in  chlo-
rine feed rate.  The same error at a plant flow rate of  1.25 mgd
                               41

-------
                 0-20%	
              = 0-20 mo/1
                          FREE
                          CHLORINE
                          SIGNAL
                                 MANUAL SET
                            FEEDBACK
                            CONTROLLER
                            PROPORTIONAL
                                +
                            INTEGRAL
                                                 0-100% ADJUSTABLE
                                                 SET 10% (10 mft/l)
                                          0-100% % 0-3000 LB/DAY
                    ORDER TO
                    CHLORINATOR
                        1
ORDER TO
CHLORINATOR
     2
% CONTROLLER OUTPUT = C% - 100% x
ERROR
PB
                                            %PBxT
                  % ERROR
•dt ]
    X ERROR  = SET %- SIGNAL %
    % PB     = % PROPORTIONING BAND ADJUSTMENT (ADJUSTABLE 0-200%)
      T      = RESET TIME ADJUSTMENT. MINUTES (ADJUSTABLE 0-20 MIN)
      t       = REAL TIME, MINUTES
      C%     = ADJUSTED VALUE AT t = O
      Figure 6-1  Simple Feedback Control - Chlorine Feeders
                                42

-------
requires  a correction  of 10.41  Ib/day  in  chlorine feed  rate.
Using simple  feedback control,  the  control system  response,  as
limited by  stability criteria,  is  set for the maximum  response
at minimum  flow.   With a desired correction of  2.08 lb/day/mg/1
free  chlorine  error  at minimum flow,  the correction should
be 1.04 lb/day/mg/1 error (immediate)  plus  0.17  lb/day/min/mg/1
error.
                                 ERROR SIGNAL



FEEDBACK
CONTROLLER,
PROPORTIONAL
+
INTEGRAL *
OUTPUT -A
* \ ^


PROCESS



 tc.
 o
    O
                    INTEGRAL TIME

                    •«	T	
'1	j
  100

            IMMEDIATE RESPONSE
            TO ERROR
               TIMED RESPONSE
               RESET FUNCTION, / * % ERROR • dt
        T
              r
                REAL TIME, t
          Figure 6-5  Control Response - Simplified Control Loop
     At  Rancho Cordova,  plant flow  varied  from a minimum to
maximum in  a  period  of;  about five  hours.   These plant  flow
variations  at  median NH4  -N concentrations  required  a  chlorine
feed rate variation  from about 375 to 1,850 Ib/day.  The rate of
change of chlorine feed rate was:
                                = 5 lb/dav/min
     This  value  is conservative  since,  under some conditions,  a
10 Ib/day/min rate of  change was required.  Simple feedback con-
trol set  for stability at  minimum flow with an error  of  3 mg/1
                                43

-------
could only provide  about 3 x  0.17 lb/day/min/mg/1 error  or about
0.51  Ib/day/min.    Therefore,  simple feedback  control could not
keep  up with  required  dosage  variations  and  a compound control
system  was required.


Simple  Feedback Plus Flow Paced Control

      This compound  system incorporated a  flow paced  control loop
in  addition  to the free  chlorine residual feedback loop.   The
control system is  shown  on  Figures  6-6 and  6-7.   This system
  FREE
  CHLORINE
  SIGNAL t
                          PLANT FLOW SIGNAL
      FEEDBACK
      CONTROLLER.
      PROPORTIONAL
          +
      INTEGRAL »
                0-100%
       MANUAL
       RATIO
       ADJUST*
                    REFER TO FIGURE 6-4 FOR
                    CONTROLLER OUTPUT DEFINITION
                    REFER TO FIGURE 6-4 FOR VALUES
                    MULTIPLICATION FACTOR
                    ADJUSTABLE 0.00 TO 1.0

                    PRODUCT OF RATIO ADJUST
                    FACTOR TIMES INPUT SIGNAL
                                       0-100% £^ 0-1.5 mgd
     MANUAL
     RATIO
     ADJUST
                                t- \-
                  SUMMATION
                            0-100% =0-3000 LB/DAY
            ORDER TO
            CHLORINATOR
                1
ORDER TO
CHLORINATOR
    2
   Figure 6-6  Simple Feedback Plus Flow Paced Control - Chlorine Feeders
                                 44

-------
was  satisfactory  for  sodium  hydroxide feed maintaining  the
treated stream within ±0.3  pH units of the set point.
          1.PH
         SIGNAL
 0-100%	
 ^ 2-12.pH
                   MANUAL SET
                   7.2 pH. (52%)
                      i
            FEEDBACK
            CONTROLLER,
            PROPORTIONAL
                +
            INTEGRAL *
             •  REFER TO FIGURE 6-4
                FOR CONTROLLER OUTPUT
                DEFINITION
             A  MULTIPLICATION FACTOR
                ADJUSTABLE 0.00 TO 1.0
             AA  PRODUCT OF RATIO ADJUST
                FACTOR TIMES INPUT SIGNAL
PLANT FLOW SIGNAL
                                              0-100%= 0-1.5mgd
                       0-100%
              MANUAL
              RATIO
              ADJUST A
                                    MANUAL
                                    RATIO
                                    ADJUST A
                                             AA
                                L
                         SUMMATION
                           T
                                   0-100% ^ 0-3000 LB/DAY
                         ORDER TO
                         FEEDER OF
                         ALKALINITY SUPPLEMENT
  Figure 6-7  Simple Feedback Plus Flow Paced Control - Alkalinity Supplement
      Figure  6-8   shows  the effect  of  a Manual Ratio  Adjust
 block  in the  plant flow signal line  prior  to the  summation
          Optimum  adjustment  for direct  plant  flow  paced  signal
         provide  somewhat less dosage  than is required  for  br+eak-
        at minimum NH|  -N  ion concentrations  and C12/NH4  -N
           For a flow change  of  1.0 mgd  in  150 min, the  chlorine
 feed rate would change  at about  6 Ib/day/min.   At median  chlo-
 rine dosages,  a  rate of  change  of chlorine  feed of about  10  lb/
 day/min would  be  required.   The feedback controller would,  then,
 be required  to  supply the  difference,  or  4  Ib/day/min in this
 example.  The  control error  would be  4.0/0.17 or about  24  mg/1.
block.
should
point
ratios.
                                  45

-------
     In these  tests, free chlorine excursions,  using  this chlo-
rination control system at plant flows above 0.5 mgd, were large
and erratic and operation was generally unsatisfactory.
 z
 
-------
 required  increment  of chlorinator  capacity  that  was not  obtain-
 able  through  the flow modified  feedback  control portion.   The
 flow  modified  feedback   control  was  limited  by  scaling  of  the
 free chlorine signal  and the proportional  band adjustment of the
 controller.
FREE
CHLORINE
SIGNAL +
   -0-20% as
    0-20ma/l
   FEEDBACK
   CONTROLLER,
   PROPORTIONAL
       +
   INTEGRAL *
                                  PLANT FLOW SIGNAL
PLANT FLOW SIGNAL
         0-100% =
         0-1.5 mgd
            -0-100%
                             REFER TO FIGURE 6-4 FOR
                             CONTROLLER OUTPUT DEFINITION
                             REFER TO FIGURE 6-4 FOR VALUES
                             MULTIPLICATION FACTOR
                             ADJUSTABLE 0.00-1.0
                             PRODUCT OF RATIO ADJUST
                             FACTOR TIMES INPUT SIGNAL
             MULTIPLIER
             OUTPUT = A x B x 0.67
                          "B"
                      -0-67%
                           NOTE: DUE TO MULTIPLIER SCALING
                           FACTOR, MAXIMUM FEED RATE WAS
                           LIMITED TO 2000 LB/DAY (3000 x 0.67)
                                        0-67% +
                                        WHERE 0-100%
                                        S 0-3000 LB/DAY (NOTE)
                       ORDER TO
                       CHLORINATOR
                          1
             ORDER TO
             CHLORINATOR
                2
 Figure 6-9  Flow Modified Feedback Plus Flow Paced Control - Chlorine Feeders

      Within the  limits of  signal  output capacity,  chlorinator
 control  on flow  modified  feedback  control only (with  no  contri-
 bution   of direct  plant  flow  signal)   provided   satisfactory
 control with  free  chlorine  excursions of  5 to  6  mg/1  from
 control  point.    These  excursions  were   random and not  partic-
 ularly  related  to  plant  flow  variations.    Addition of  direct
 plant  flow control  signal  up  to  about  one-half  of the total
 control  signal appeared to have  little effect on the  5 to 6  rag/1
 excursions from the  control set  point.
                                   47

-------
     Experience at Raneho Cordova  showed  that  required  chlorine
dosages varied from about 12fl rag/1 to  250 mg/1  as  a  function  of
NH4 concentration and the NH4 -N/C12 ratio.   For chlorine  dosage
variation of  this  magnitude over a 10-hour period,  the rate  of
change of chlorine dosage was  (250-120)/(10 x  60)  =  0.217 mg/1/
min.   The  flow modified feedback  controller can  increment 0.68
Ib/day  chlorine  feed/min/mgd plant  flow/mg/1  of  free  chlorine
error, or can correct for  .0817 mg/1  per minimum  concentration
change  for  a free chlorine  residual  error  input  signal  of 1.0
mg/1.  Control offset from set point would then be:
Ammonia Nitrogen Mass Flow Paced Control

    In this control mode, NH4  -N concentration Was measured  and
continuously multiplied by plant  flow to provide an NH4  -N mass
flow rate  (Ib/day).   This flow  rate was  used as a direct feed
forward  signal in lieu of plant flow in several  control  schemes
for chlorine and  sodium hydroxide feed.  NH4 -N concentration,
as a raw  signal,  was  not suitable for control due to the pres-
ence of  noise.   This  signal was  conditioned by  resistance-
capacitance filtering which  added a 5-min time constant to  the
17-min analysis dead time.

     After conditioning,  the signal was used for  control.    No
improvement in control was noted.   Control overshoots were
occasionally noted that appeared to be  coincidental with  changes
in the rate of change  of NH4  -N concentration.   These over-
shoots were  the  probable result  of  the analysis dead time  and
signal conditioning time  which  introduced out-of -phase  signals
into the control system.


RECOMMENDED CONTROL SYSTEM

     Based  on  the  proceeding  discussion  of process  control
systems,  a control scheme  was selected  for both  alkalinity
supplement feed and chlorine  feed.

Alkalinity Supplement  Feed Control

     Sodium hydroxide feed control at Raneho Cordova was satis-
factory with the control system shown on Figure 6-7.  However, a
flow modified feedback  control with plant flow paced control  as
shown on  Figure 6-10 is recommended.  This control configuration
will provide better pH  control in breakpoint applications with
low influent alkalinity.  The direct plant flow paced portion  of
the total control  signal should be less  than  the ratio of break-
point chlorine dosages  for minimum to maximum process require-
ments.
                               48

-------
                     PLANT FLOW
    FEEDBACK
    CONTROLLER
    PROPORTIONAL
       +
    INTEGRAL *
0-100%
           -0-100%
                                    REFER TO FIGURE 6-4 FOR
                                    CONTROLLER OUTPUT
                                    DEFINITION
                                    MULTIPLICATION FACTOR
                                    ADJUSTABLE 0.00-1.0
                                    PRODUCT OF RATIO ADJUST
                                    FACTOR TIMES INPUT SIGNAL
                       PLANT FLOW
        "A"
                       "B"        "C"
               MULTIPLIER OUTPUT = A x B x C
                        - 0-100% x C
                                                0-100%
                              SUMMATION
                                   NOTE: THEORETICALLY THIS
                                   CONTRIBUTION SHOULD BE
                                   REDUCED TO ZERO,
                                   BUT THIS WAS NOT VERIFIED
                                   IN ACTUAL OPERATION
                                     0-100%x
                              ORDER TO
                          FEEDER OF ALKALINITY
                             SUPPLEMENT
          Figure 6-10 Recommended Control - Alkalinity Supplement
Chlorine Feed  Control

      The recommended  system  for control  of chlorine  feed equip-
ment  is  shown  on Figure 6-11.   Two  chlorinators are  used,  one
large  and  one  small.    For  this   configuration ,  the  large
chlorinator  is controlled  solely by  plant now  rate  and  the
small  chlorinator  is controlled by  a flow modified feedback
controller.

      The  recommended control  scheme  of  Figure  6-11  showed
average excursions from  the free residual control point of  about
±3  mg/1  free  chlorine.   The excursions were somewhat  larger  at
low plant  flows  and  low NH4 -N   concentrations than  they were
at  higher flows and concentrations.

      The chlorinator   feed rate resolution  is  suggested  as  an
important element in the superior  performance of the  recommended
                                  49

-------
chlorine  control  mode.    The difference may  occur  because  the
large  chlorinator  in the  recommended  scheme  is  not called  upon
to  make both positive and negative  adjustments in  any short
period  of  time  since  it  is  totally flow  responsive.   The  flow
signal,  as  previously  notedr  was conditioned to  remove  noise.
This would  tend to minimize any errors due  to lack of  resolution
of  the large  chlorinator  feed rate, because  over  any period of
time  changes  in feed  rate would tend  to  be  unidirectional  and
infrequent.
             I
          FREE
          CHLORINE
          SIGNAL
             4—0-30
   PLANT
   FLOW
   SIGNAL
             I   0-30
0-30% a
0-30 mgfl
                               REFER TO FIGURE 6-4 FOR
                               CONTROLLER OUTPUT DEFINITION
                               MULTIPLICATION FACTOR
                               ADJUSTABLE 0.00-1.0
                               PRODUCT OF RATIO ADJUST FACTOR
                               TIMES INPUT SIGNAL
FEEDBACK
CONTROLLER,
PROPORTIONAL
    +
INTEGRAL  •
             0-100%
PLANT
FLOW
SIGNAL
                           0-100%
                             "B"
                           MULTIPLIER
                         OUTPUT = A x B x C
                     "C"
                     (0-100%) xC-
ORDER TO
LARGE
CHLORINATOR (S)
         ORDER TO
         SMALL
         CHLORlNATORtS)
            Figure 6-11  Recommended Control - Chlorine Feeders


     To  obtain optimum  response from the chlorinators,  the
control  vacuum should be  reduced until full linear travel  of the
variable orifice produces only  the minimum quantity required to
meet the current maximum process requirements.   The small chlo-
rinators should be carefully sized to provide 125 to 150 percent
of  the  amount of  chlorine  feed needed  between anticipated
                                 50

-------
maximum  and  minimum  breakpoint chlorine  dosages.  The large
chlorinators  should  be  sized to provide  breakpoint  concentra-
tions of  chlorine  for maximum  plant  flows and minimum NH4  -N
concentrations   and  Cl2:NH4  -N  ratios.


INSTRUMENTATION REQUIREMENTS

Control Components

     Since process  control  is  critical,   requiring  a  chlorine
feed rate resolution  of 1  percent or better, control components
which have dead bands should  be  avoided.

Drift In Zero and Linearity

     Since breakpoint process  control  depends  partly on  feed
forward flow signals,  zero or linearity errors  have  to be  com-
pensated  for  through  the feedback  controller.   Zero drift
in  control components causes  system gain  changes which  can
result  in severe  control problems at  low  plant flow rates.
Components should be selected which  are  as  linear and drift-free
as  possible and  should  be readily  adjustable to eliminate  zero
and linearity errors.

Feedback Controller Scaling

     Control  loop and component characteristics should be  care-
fully analyzed to allow the initial  calculated proportional  band
for  the  controller to  be  about 40 percent  at  the point  where
oscillations  will first commence.   If  this  requirement cannot be
met by appropriate scaling  of free chlorine residual or pH
signal alone,  then  an additional attenuator  (or amplifier)  must
be  added directly in the feedback controller output to  the
multiplier and the  multiplier  (or divider)  appropriately  re-
scaled.   Full  signal range  must be retained in the multiplier
output.

Signal Redundancy

     If the  process is required to run on a continuous  basis,
redundant  automatic  plant flow, free chlorine  residual and pH
input signals will be required to maintain  operation while  these
transmitters  are  routinely  cleaned,  serviced and  adjusted.

Calibration Signals

     To initially  align  the  control system and to periodically
check  its proper  operation  after  startup, calibration input
signals  will  be  simultaneously  required  for free  chlorine
residual, plant  flow and pH.   It is  recommended  that these
calibration signals be permanently  built into the control system


                                51

-------
with appropriate readouts,  stepless 0  to 100 percent  adjustments
and "operate-calibrate" selector switches.   Plant  flow  calibra-
tion  inputs  should have  separate  "operate-calibrate"  switches
for each  input into  each chlorinator control  and  alkalinity
supplement feed control system.
SUMMARY

     The control system can provide satisfactory control  only  if
it is properly  tailored to  the application.   Oversizing  of  feed
capacity  will   proportionately  reduce  the  control  tolerances.
Prior to  finalizing  system design, data should  be  collected  on
plant flow covering maximum and minimum values and maximum rates
of change.  Data should also be available on  NH4 -N  concentra-
tion maximums,  minimums and maximum rates of change.  Alkalinity
data should also be collected to determine the system's response
to addition of alkalinity supplement.
                               52

-------
                           REFERENCES


1.  Houston, Sir A. C. 19th and  20th Annual Reports of the Metro-
    politan Water Board, London, England  (1925 and 1926).

2.  Calvert, C. K. "Treatment with Copper Sulfate, Chlorine,and
    Ammonia," JAWWA, 3_2 (17) : 1155-64, 1940.

3.  Standard Methods for the Examination of Water and Wastewater,
    13th Ed., Amer. Public Health Assoc., New York, 1971.

4.  Saunier, B. M. and R. E. Selleck.  Kinetics of Breakpoint
    Chlorination and of Disinfection.  Sanitary Engineering
    Research Laboratory Report, University of California,
    Berkeley, April, 1976.

5.  Pressley, T. A., D. F. Bishop, and S. G. Roan.  "Ammonia-
    Nitrogen Removal by Breakpoint Chlorination," Environ. Sci.
    and Techn., 6(7):622-28, July, 1972.

6.  Wei, I. W.  Chlorine-Ammonia Breakpoint Reactions;  Kinetics
    and Mechanisms.Ph.D. Thesis, Harvard Univ., Cambridge,
    Mass., May, 1972.
                               53

-------
            APPENDICES
              APPENDIX A

RANCHO CORDOVA BREAKPOINT CHLORINATION
        DEMONSTRATION PROGRAM

            DATA SUMMARY
                  54

-------
U1
U1
Time
C12
feed,
Ib/day
I. pH set point
1450
1510
1530
1545


2030
2100
2075
1900


Flow,
mgd
C12
dose,
mg/l
Inijial
mg/l
(C,2/NH;-N)
Free
residual
remaining ,
mg/l
6.5 Breakpoint effluent used as Injector water, March 10,
1 .204
1.204
1.171
1.108


202
209
212
206
X =207
SD= 4
20.25
20.25
19,31
IB. 27
X =19.5
SD= 0.9
10.0
10.3
11.0
11,3
X =10.6
SD= 0.6
9.5
9.5
15
17
X =13
SD= 4
Final
NH^-N,
mg/l
976
1 .4*
1 .8*
2.3*
1.4*


II. pH set point 7.0 Secondary effluent used as Injector water, February 20,1976
0020
0105
0202
0301
0403
0703
0755
0907
1003
1104
1200


1625
1300
920
670
560
625
1060
1670
1780
1820
1830


0.944
0.774
0.544
0.404
0.324
0.374
0.684
1 .104
1 .174
1 .174
1 .044


206
201
203
199
207
200
186
181
182
186
210
X =1')6
?D= 11
22
22
22.5
22
21 .5
20.1
19.5
19.5
19.5
20.8
23
X =21.1
SD= 1 .3
9.4
9.1
9.0
9.0
9.6
9.9
9.5
9.3
9.3
8.9
9.1
X =9.3
SD=0.3
16
15
16.5
8
12
9
10
11 .5
9
6
r>
X =11
SD= 4
III . pH set point 7.0 Secondary effluent used as injector water , March 3-4 ,
1953
2051
2150
2252
0002
0055
0158
0251
0351
0451
0552
0651
oaoo


1890
1920
1830
1770
1370
1085
780
550
445
375
375
490
800


.124
.074
.174
.174
.094
0.684
0.524
0.354
0.304
0.284
0.284
0.354
0.624


202
214
187
131
150
190
178
186
176
158
158
166
154
X =177
SD= 19
19
19
19
18
17
17
17
16.5
16
16
16
15
14
X =16.9
SD= 1 .6
10.6
11 .3
9.8
10.0
8.8
11 .2
10.5
11 .3
11 .0
9.9
9.9
11.1
11.0
X =10.5
SD= 0.8
12.5
14
15
15
10
11 .5
13
16
18.5
20
12
10
1 2
X =14
SD= 3
0.58

0.34
0.38
0.29
0.38
0.53
0.29
0.47
0.28
0.75
X =0.43
SD=0.15
1976
0.68
0.75
0.82
0.80

0.52
0.54
i).M)
0.73
0.76
0.50
0.43
	 0_.J5
5T =0.65
SD=0.14
Initial
NOj-N,
mg/l

0.27
0.21
0.15
0.07

















0.38
0.20
0.11
0.09
0.08
0.08
0.07
0.07
0.06
0 .08
0.10
0.10
0.27


Initial
NO^-N,
mg/l

0.20
0.17
0.15
0.08












Final
mg/l

1.2
1.4
1.2
1 .3

















0 . 08
0.05
0.03
0.03
0.03
0.05
0 . 06
0.05
0.03
0.04
0.04
0.04
0.10





1.5
1 .3
1 .3
1 .2
-
.1
.2
.4
.3
.2
.2
.2
2.0


NO§
prod . In
Bpolnt,
mg/l

0.73
1 .02
0.90
1 .15
X =0.95
SD=0.18











Chlorine consumed by reaction
NHj~N2
mg/l

143
140
129
128
X =135
SD= 8
















1.04 139
1.05 ' 139
1.1'.
1 .08
-
0.98
.07
.29
.21
.09
.06
.06
1 .63
X =1.14
SD=0.17
138
131
-
125
125
122
116
116
118
111
101
X =123
SD= 12
NHj— • NO3
mg/l

15
21
18
23
K =19
SD= 4















21
21
24
22
_
20
22
26
25
22
22
22
33
X =23
SD=23
NO2 — -NO3
mg/l

1
1
1
0
X =1
SD=1















2
1
1
0
0
0
0
0
0
0
1
1
1
X =1
SD=1
NH+— -NC13
mg/l

21
27
35
21
X =26
SD= 7


9
5
6
4

8
4


4




10
11
12
12
_
3
8
8
11
12
3
7
1 1
' X" = 10
SD= 2
Unidentified
loss,
mg/l

12
10
14
17
X =13
SD= 3















17
2=i
(-3)
1
_
25
10
14
5
(-12)
(-3)
15
(-4)
"X" ~8
SD=12
                               Table A-1.   Rancho Cordova Breakpoint Chlorination-Data Summary

-------
U1

Time
IV. pH
0405
0450
0552
0652
0750
0858
1002
1102
1200


V. PH
2100
2153
2255
0001
0051
0157
0250
0351
0451
0552
0651
0800
0905
1000


VI. pH
0310
0342
0443
0555
0648
0800
0900
1000
1100
1200
1300
1400



feel
Ib/d»y
set polnl
300
300
290
340
1185
1375
1340
1610
1950


set point
2020
1815
1690
1460
1190
800
570
505
470
430
530
1050
1430
1270



Flow,
mgd
7.3 Br
0.314
0.314
0.304
0.362
0.934
1.174
1.104
1 .174
1 .304



C12
dose,
mg/l

Initial
NHj-N.
mg/t

(cn.XHHj-M,

Tree
residual
remaining ,
mg/l
eakpoint effluent used as inlector water, March 10,
115
115
114
113
152
140
I4f,
164
175
Y -137
SI.1- 24
9.5
9.8
9.5
10.0
14.0
14.0
13.9
15.5
18 2
X -12.7
SD« 3 . 1
7.5 Secondary effluent used
1 .114
1 .104
1 .104
0.934
0.774
0.524
0.374
0.324
0.304
0.284
0.364
0.734
1 .014
0.994


217
197
184
187
1B4
183
183
187
185
182
175
172
169
153
X -183
SD" 14
18
17.7
17
17
17
17
17
17
17
17
16
16
16
1 6
X -16.8
SD- 0.6
12.1
11.7
12.0
11 .3
10.9
10.0
10.5
10. h
,j 6
X '11.0
SI1'- 0.9
20
15
10
f,
15
It,
12.5
11
10
X M3
•.Hi" 4
as Injector water, March 2-3
12.1
11.1
10.8
11.0
10. B
10.8
10.8
11.0
10.9
10.7
10.9
10.7
10.6
9 6
X -10.8
SD» 0.5
15
17
14
14
12.5
12.5
12.5
15
14
20
12.5
15
14.5
12.5
X =14
SD- 2

rinal
NHj-N.
mg/l
976
0.39
0.29
0.33
0.15
0.48
0 . VI
0.20
0.37
_
X -.0.34
31. 0.12
, 1976
0.75
0.75
0.76
0 . 72
0.65
0.5S
0.39
0.49
0.44
0.50
0.46
0.52
0.53
0.45
X= 0.57
SD=0 . 1 3
set point 7.7 Breakpoint effluent used as Injector water, Marchll, 1976
300
250
230
200
250
940
1230
1290
1580
1620
1570
1800


0.374
0.324
0.304
0.304
0.374
0.824
1.104
1.164
1 .304
1.234
1.019
1.044


96
93
91
79
80
137
134
133
145
157
185
207
X -128
SD= 42
9.8
8.9
9.4
7.9
9.2
11.8
12.2
11.8
12.9
15.0
16.3
17.7
X -11.8
SD" 3.2
9.8
10.4
10.8
10.0
8.7
11 .6
11 .0
11 .3
11.2
10.5
11.3
11.7
X -10,7
SD- 0.9
7
7
12
12
9
16
13.5
12.5
10
10
11.5
1 3
X =11
SD- 3
0.70
0.70
0.17
0.64
0.16
0.57
0.62
0.43
0.31
0.88
0.86
0 14
X -0.52
SD-0.27

Initial
NCj-N,
nrsg/l

0.23
0,28
0.34
0.41
0.48
0.67
i). n
!I.B4
0.83



0 . 01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.02
0.02



0.07
0.07
0.06
0.03
0.01
0.01
0.01
0.10
0.27
0.15
0.03
0.03



Initial
NO,-!;,
mg/l

0.27
0.26
0.26
0.31
n.37
0 . 59
'i.79
1 .05
1 .30



0 . 09
0.07
0.06
0.06
0.05
0.05
0.05
0.06
0.07
0.06
0.06
0.07
0.08
0.08



0.05
0.06
0.06
0.06
0.06
0.06
0.06
0.13
0.19
0.20
0.21
0.20



Final
mg /t

1.1
1 .5
1 .0
.*>
2.1
2.5
2.4
2.4
_



1 .4
1 .4
1 .2
.2
1 .1
0.92
1 .1
1.2
1 .2
1.3
1 .3
1.3
1.5
1 .3



1.1
1.2
1.2
1.2
1.1
1.1
.1
.6
.7
.4
,6
.1



NO5
prod , In
B'polnt ,
mg/l

0.60
0.96
1.00
0.88
1.25
1.24
0.96
O.'il
_
f? -0.93
Sl:-0.27

1.31
1 .33
1 .14
1.15
1 .05
0.87
1.06
1.14
1.15
1.24
1.24
1.23
1 .40
1 .70
X =1.18
SD=0.14

0.98
1.07
1 .08
1 .11
1 .04
1.04
1 .04
1.37
1.24
2.05
1.36
0.87
X =1.19
SD-0.31
Chlorine consumed by reoctlon

mg/l •

69
72
70
75
103
103
104
115
_
;•: = 89
SD- 19

131
129
124
124
124
125 •
126
126
126
126
118
118
118
118
X" =124
SD* 4

69
62
63
55
69
85
88
86
96
107
117
134
X - 86
SD= 24
KHj -NO,
mg/l

12
19
20
IB
25
25
19
10
_
X -19
,'iD- 5

27
2T
23
23
21
18
22
23
23
25
25
25
28
34
X =25
SD* 4

20
22
22
23
21
21
21
28
25
42
28
18
X =24
SD= 6
NO! "-NOj
mg/l

1
1
2
2
2
3
4
4
4
X =3
sn=i

0
0
0
0
0
0
0
0
0
0
0
0
0
0
X =0
SD=-

0
0
0
0
0
0
0
1
1
1
0
0
X =0
SD=1
"HJ -*NO13
mg/l

f,
A
'T
2
7
8
3
6

X -!.
:;n=2

11
11
12
11
10
R
6
7
7
8
7
B
8
7
X =9
SD=2

11
11
3
10
2
9
9
7
5
13
13
2
X =8
SD=4

Unidentified
loss ,
mg/l

7
4
7
10
IJ
(-15)
3
18
-
5T - -1
SI)" 9

33
13
11
15
16
19
16
16
15
3
12
6
0
(-19)
X =11
SD=12

(-11)'
(-9)
(-9)
(-21)
(-21)
6
2
(-2)
8
(-16)
15
40
X =(-1)
SD=17
                                                           Table A-1.  (Continued)

-------
U1
-J
Time
VII . pH
1655
1753
1013
1855
2055
2155
2255
2354
0052
0148
0249
0348
0450


VIII. PH
170S
1730
1750
1815


Cl,
•feed,
Ib/day
flow,
mtjd
set point H.O
1435
11.00
1550
1750
2080
2010
1950
1370
1135
855
625
515
490


O.R24
0.924
0.924
1 .000
1 .234
1 .174
1 .104
0.884
0.714
0.524
0.3114
0.324
0.284


set point 3.5
1450
1430
1445
1410


0.906
0.876
0.933
0.933


C12
dose .
mg/l
Initial
NliJ-N,
mq/l
(C12/NHJ.N)
Free
residual
remain ing ,
mg/l
Final
NH^-N,
mg/l
Secondary effluent used as injector water, March 4-5, 1976
209
208
201
210
202
205
206
186
191
196
195
191
207
X =201
SD= 8
21
20.8
21
20.5
21
19.8
20
' 19.5
19
19
19
18.7
18.5
St =19.8
sn= i .0
9.9
10.0
9.6
10.2
9.6
10.4
10.3
9.5
10.0
10,3
10.3
10,2
1U.2
X =10.1
SD= 0.4
7
11
12.5
11.5
11.5
14
16
9
13
12. 5
10
7
12
?; =11
SD= 3
0.52
0.54
0.83
t .02
0.97
0.93
0.88
-
0.58
O.Sb
-
0.43
_
X" =0.73
SD=0.22
Breakpoint effluent used as Injector water, March 10, 1976
192
196
186
181
X =189
sn= 7
17.7
17.2
16.9
16.9
X =17.2
SD= 0 . 4
10.8
11 .4
11.0
1^7
X =17.2
SD= 0.4
14
14
19.5
15
X =16
sn* 3
0.52
1 .1*
1 .3*
1.3*


Initial
NOj-N,
mg/1

0.01
0.01
0.01
0.01
0.01
0.05
0.01
-
0.01
0.01
0.01
0.01
0.01



0.07
0.13
0.04
0.03


Initial
NO'-N,
mg/1

0.05
0.05
0.05
0.05
0.05
0.05
0.04
-
0.01
0.04
0.05
0.04
0.05



0.08
-0.09
0.08
0.07


Final
NOj-N,
mg/1

1 .0
1 .0
1.2
1 .0
0.78
0.68
0.58
-
0.64
0.70
-
0.75
0.82



1 .4
1 .2
0.94
0.78


N03
prod . in
B'polnt,
mg/1

0.94
0.94
1.15
0.95
0.73
0.58
0.54
-
0.64
0.66
-
0.70
0.77
X = 0.78
sr= 0.19

1.25
0.98*
0.82*
0.68*


Chlorine consumed by reaction
NHj— N2
mg/1

156
154
153
148
152
144
145
-
140
MO
-
139
_
X =147
SD= 6

131
-
-
-


NH^-»NC>3
mg/1

19
19
23
19
15
12
11
-
13
13
-
14
16
X =1 6
SD= 4

25
-
-
-


NO~— -NOg
mg/l

0
0
0
0
0
0
0
-
0
0
0
0
0
X =0
SD=-

1
-
-
_


NH+— NC13
mg/l

8
8
13
16
15
14
13
-
9
9
-
7
-
X =11
SD= 3

8
8
-
_


Unidentified
loss,
mg/l

19
16
(-1)
15
8
21
21
-
16
21
-
24
~
X =16
SD= 7

13
-
-
_


                                                           Table A-1.   (Continued)

-------
              APPENDICES
              APPENDIX B

RANCHO CORDOVA BREAKPOINT CHLORINATION
        DEMONSTRATION PROGRAM

      BREAKPOINT MODEL PREDICTIONS
                   58

-------
    100
z
<
 CO
X
z
O
tc.
    50




    20



     10



      5




      2



      1
                                                     INITIAL CONDITIONS
       10
        Figure B-l.
                                       pH
20       30    40  50  60 70 80  100
                                     TIME, SECONDS
                                                                    200
                                                                               8  pH
                      Ammonia Removal with Breakpoint Chlorination in a
                      Pipe Reactor-Model Prediction-pH Set Point =6.5
              300
    100



O   50
ui
oc
 n
I
Z
I-

UJ
O
cc
     20
     10
                                                      INITIAL CONDITIONS


                                                      CI2:NH3-N = 9.30 TO 1
                                                      NH3-N

                                                      TEMP

                                                      pH

                                                      ALKALINITY -

                                                      NaOH
21.05mg/l

15° C

7.4

165 mg/l

120 mg/l
                                                                pH
                                                                               8  pH
       10
        Figure B-2.
20       30    40  50 60  70 80   100
                                     TIME, SECONDS
                                                                     200
                      Ammonia Removal with Breakpoint Chlorination in a
                      Pipe Reactor-Model Prediction-pH Set Point = 7.0
              300
                                          59

-------
    100
                                                       INITIAL CONDITIONS
O
z
<
UJ
cc
z
 i
 m
UJ
U
CC
UJ
a.
50
20
 10
CI2:NH3-N
NH3-N
TEMP
pH
ALKALINITY
NaOH
                                  10.5 TO 1
                                  16.9 mg/l
                                  15° C
                                  7.4
                                  165 m«/l
                                  92 mg/l
                                                                              8  pH
       10
                 20
30   40  50 60  70 80  100
                                                                       200
                         300
         Figure B-3.
                                 TIME, SECONDS
                    Ammonia Removal with Breakpoint Chlorination in a
                    Pipe Reactor-Model Prediction-pH Set Point =7.0
o
z
z
<
tu
CC
 m
X
z
UJ
U
flC
    100
50
20
     10
                                                       INITIAL CONDITIONS
                              NHj-N
                              pH
                        O2:NH3-N
                        NH3-N
                        TEMP
                        pH
                        ALKALINITY
                        NaOH
         12.25 TO 1
         12.02 mg/l
         15° C
         7.4
         165 mg/l AS CaCO3
         70 mg/l
                                                                                  8  pH
       10
                 20
30    40  50  60 70 80   100
                200
300
         Figure B-4.
                                 TIME. SECONDS
                   Ammonia Removal  with Breakpoint Chlorination in a
                   Pipe Reactor-Model Prediction-pH Set Point =7.3
                                           60

-------
    100
z
z
<
LU
OC
 CO
X
o
ec
UJ
0.
 O
 Z
 z
 <
 5
 m
X
z
I-
UJ
O
OC
111
a.
                    INITIAL CONDITIONS
                    CI2:NH3-N  =  10.8 TO 1
                     3-
       = 16.7mg/l
TEMP    = 15.0°C
pH      = 7.4
ALKALINITY = 165mg/l
NaOH
                                                                                 8  pH
                                    40  50  60  70 80   100

                                     TIME, SECONDS
                                                                      200
                                                           300
         Figure B-5.    Ammonia Removal with Breakpoint Chlorination in a
                         Pipe Reactor-Model Prediction-pH Set Point - 7.5
    100
                CI2:NH3-N  =  11.7 TO 1
                NH3-N    =  11.83mg/l
                TEMP     =  15° C
                pH      =  7.4
                ALKALINITY =  165 mg/l AS CaCO3
                NaOH     =  94 mg/l
                                                                                  8   pH
                      20       30    40  50 60 70 80  100

                                      TIME, SECONDS
                                                                      200     300
         Figure B-6.   Ammonia Removal with Breakpoint Chlorination in a
                        Pipe Reactor-Model Prediction-pH Set Point = 7.7
                                          61

-------
    100
 o

 Z

 <

 tu
 oc
 CO
 X


 H

 UJ
 o
 oc
 Ui
 Q.
                                       50 60  7080  100
                                   TIME, SECONDS
                                                                            8  pH
                                                  200     300
         Figure B-7.
       Ammonia Removal with Breakpoint Chlorination in a
       Pipe Reactor-Model Prediction-pH Set Point = 8.0
    100
u
z
2
<

iu
K
 n
X
Z
H

UJ

2
UI
o.
a2:NH3-N  ~  11.7 T01

       -  17.2 mg/l

TEMP    -  15° C

pH      -  7.4

ALKALINITY-  165 mg/l

MiOH    -  148 mo/I
                                                                            8  pH
                            30
                  40  50  60 70 80  100



                   TIME, SECONDS
200
                                                                         300
        Figure B-8.   Ammonia Removal with Breakpoint Chlorination in a
                      Pipe Reactor-Model Prediction-pH Set Point =8.5
                                       62

-------
    7.5
Q.
o
I
CC
I-
1
UJ
K
CD
    7.0
6.5
           INITIAL CONDITIONS

                      =  20 mg/l
                      =  15° C
       NH3-N
                      =  7.20
       TEMP
       PH
       ALKALINITY =  175 mg/l AS CaCO3
       CT         =  0.0036 MOLES/LITER
                          CI2:NH3-N = 10:1
          CI2:NH3-N=11:1
                                                  1
             -SYSTEM CLOSED TO ATMOSPHERE-
                                              BEYOND THAT REQUIRED FOR
                                              NEUTRALIZATION OF ACIDITY
                                              FROM HYDROLYSIS AND
                                              DISSOCIATION OF CHLORINE
                                              (EQ. 16, TABLE 3-3).
    6.0
                   25
                         50          75         100

                              NaOH ADDED1, mg/l
125
                                                                          150
          Figure B-9.   Breakpoint Reaction Final pH vs NaOH Added,
                        CI2:NH3-N Variable
                                     63

-------
  8
I
oc
H-
1
OQ
                                         CT = 0.006 MOLES/LITER

                            CT - 0.005 MOLES/LITER

                    CT = 0.004 MOLES/LITER
            Cj = 0.003 MOLES/LITER
                                                  CT = 0.002 MOLES/LITER

                                      INITIAL CONDITIONS
        BEYOND THAT REQUIRED FOR
        NEUTRALIZATION OF ACIDITY
        FROM HYDROLYSIS AND
        DISSOCIATION OF CHLORINE
        (EO. 16, TABLE 3-3).
                 I	I	
                                       CI2:NH3-N
                                       NH3:N
                                       TEMP
                                       PH
                                       ALKALINITY
            =  9:1
            =  20mg/l
            =  15° C
            =  7.20
            =  175 mg/l AS CaCO3
                                       -SYSTEM CLOSED TO ATMOSPHERE-
                25
                            50
75
                                                  100
                                                              125
                                  150
                                NaOH ADDED1, mfl/l
     Figure B-10.  Breakpoint Reaction Final pH vs NaOH Added, CT Variable
                                     64

-------
   20
   10
<
III
ec
u
oc
                              CI2:NH3-N
          1
   0.5
           BEYOND THAT REQUIRED FOR
           NEUTRALIZATION OF ACIDITY
           FROM HYDROLYSIS AND
           DISSOCIATION OF CHLORINE
           (EQ. 16, TABLE 3-3).
                                                 INITIAL CONDITIONS
                       NH3-N

                       TEMP =15°  =
                       pH
                       ALKALINITY =
                          20 mo/I

                          15°C
                          7.20
                          175 rag/I ASCaCO3
                          0.0036 MOLES/LITER
             -SYSTEM CLOSED TO ATMOSPHERE-
                25
50
 75        100


NaOH ADDED1, ma/I
                                                          125
                                         150
             Figure B-11  Ammonia Remaining  (NC13 Formed) vs NaOH
                          Added, C12:NH3-N Variable
175
                                       65

-------
i   3
ui
K
I
Ul
o
K
                                   CT = 0.002 MOLES/LITER
1 BEYOND THAT REQUIRED FOR
 NEUTRALIZATION OF ACIDITY
 FROM HYDROLYSIS AND
 DISSOCIATION OF CHLORINE
 (EQ. 16. TABLE 3-3).
                 CT = 0.003-0.006 MOLES/LITER
                   INITIAL CONDITIONS
                   NH3-N
                   TEMP
                   pH
                   ALKALINITY
                        =  9:1
                        =  20mg/T
                        =  15° C
                        -  7.20
                        =  175 mg/l AS CaCO3
                                            -SYSTEM CLOSED TO ATMOSPHERE-
               25
                 50
125
150
                                   75         100

                                 NaOH ADDED1, mg/l

Figure B-12.  Ammonia Remaining (NCI3 Formed) vs NaOH Added, CT Variable
                                      66

-------
   350
   300
   250
o
ai
u
   200
    150
    100
     50
                             INITIAL CONDITIONS
                                 20 mg/l
                                 15° C
                                 7.20
                                 175 mg/l AS CaC03
                                 0.0036 MOLES/LITER
NH3-N
TEMP
pH
ALKALINITY
1 BEYOND THAT REQUIRED FOR
 NEUTRALIZATION OF ACIDITY
 FROM HYDROLYSIS AND
 DISSOCIATION OF CHLORINE
 (EQ. 16, TABLE 3-3).
              -SYSTEM CLOSED TO ATMOSPHERE-
                  25
                   50         75         100

                        NaOH ADDED1, mg/l
                                125
150
      Figure B-13.  Time to Reach Minimum Ammonia Concentration vs NaOH
                   Added, CI2:NH3-N Variable
                                     67

-------
   400
   350
   300
   250
M
ui  200
    150
    100
     50
          1
BEYOND THAT
REQUIRED FOR
NEUTRALIZATION
OF ACIDITY FROM
HYDROLYSIS AND
DISSOCIATION OF
CHLORINE
(EO. 16. TABLE 3-3)
INITIAL CONDITIONS
             =  9:1
             =  20 mg/l
             =  15° C
             =  7.20
             =  175 mg/l AS CaCO3
                             CI2:NH3-N
                             NH3-N
                             TEMP
                             pH
                             ALKALINITY
                                  CT = 0.002 MOLES/LITER
                                            CT = 0.003 MOLES/LITER
                                                CT = 0.004 MOLES/LITER
                             0.005 MOLES/LITER

                     Cj = 0.006 MOLES/LITER

                          -SYSTEM CLOSED TO ATMOSPHERE-

                              i	I	1	
                  25
                  50         75         100
                     NaOH ADDED1, mg/l
                                                              125
                                 150
      Figure B-14.  Time to Reach Minimum Ammonia Concentration vs NaOH
                    Added, C. Variable
                                     68

-------
CO
UJ
CO
ill
Z
DC
Q
z
til
o
o
U
_cg
U

<


 E
              LINES REPRESENT CONCENTRATIONS
              PREDICTED BY COMPUTER MODEL
                  20
                        40   60  80 100
200
400  600 8001000   2000 3000
                                        TIME, SECONDS
  Figure B-15 Predicted Breakpoint Chlorination Kinetics in a Plug Flow Reactor,
               pH = 6.80, NH3-N = 2.5 mg/l. Temp = 15 C, CI2/N = 9.0
ui
O
£
CO
UJ
z
£
3  «
X  U
0  CO
z
O
o
U
                                                         LINES REPRESENT CONCENTRATIONS
                                                         PREDICTED BY COMPUTER MODEL
                  20
                        40   60  80 100
200
400  600 8001000   2000 3000
 Figure B-16.
                                        TIME, SECONDS
               Predicted Breakpoint Chlorination Kinetics in a Plug Flow
               Reactor, pH = 7.50, NH3-N = 2.5 mg/l. Temp = 15 C,  CI2/N = 9.0
                                         69

-------
                                   TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
1. REPORT NO.
 EPA-600/2-78-029
              3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
 FULL-SCALE DEMONSTRATION  OF  NITROGEN REMOVAL
 BY BREAKPOINT CHLORINATION
              5. REPORT DATE
               March 1978 (Issuing Date)
              6. PERFORMING ORGANIZATION CODE
 . AUTHOR(S)
 Richard W. Stone
 Sacramento Area Consultants
                                                           8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS

 Sacramento Regional County  Sanitation District
 4660 Ecology Lane
 Sacramento,  California   95827
               10. PROGRAM ELEMENT NO.

                1BC611
               11. CONTRACT/GRANT NO.

                S803343-01
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental  Research Laboratory--Cin.,OH
Office of Research and Development
U.S.  Environmental Protection Agency
Cincinnati,  Ohio  45268
               13. TYPE OF REPORT AND PERIOD COVERED
                Final	
               14. SPONSORING AGENCY CODE

                EPA/600/14
15. SUPPLEMENTARY NOTES

 Project Officer:  James  J.  Westrick (513) 684-7652
16. ABSTRACT
 A large-scale  breakpoint chlorination  system was constructed and operated  at  Rancho
 Cordova, CA.   Reliable operation was demonstrated and a number of observations
 regarding process chemistry and engineering were made, including (1) the chlorine
 to ammonia-N ratio required to reach breakpoint and to maintain a controllable free
 residual was 10:1; (2) nitrate production was not pH sensitive in the range  6.5 to
 8.5;  (3) the rate of reaction of ammonia oxidation varied with final pH, with
 fastest rates  observed at pH 7.0.  Slower rates were observed at lower  and higher
 pH;  (4) mixing intensity had no effect upon chemical requirements or effluent
 quality, but was important for smooth  process control; (5)  the amount  of  sodium
 hydroxide used for neutralization was  identical to that predicted from  stoichio-
 metry;  (6)   the control system used here provided excellent control of  the process
 chemistry.
1 <
1
IV. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Chlorination
Ammonia
Oxidation
Waste treatment
Process control
Reaction time
18. DISTRIBUTION STATEMENT
Release Unlimited
b. IDENTIFIERS/OPEN ENDED TERMS
Demonstration
Rancho Cordova
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COS ATI Field/Group
13B
21. NO. OF PAGES
78
22. PRICE
EPA Form 2220-T (9-73)
70
                                                                8OI«tMBt-a60-880/58

-------