15th International Symposium on Wastewater Treatment and Drinking Water,
Montreal, Canada, November 17-19, 1992.
"The Impact of Ozonation and Biological Treatment on Disinfection By-Products."
   THE IMPACT OF OZONATION AND BIOLOGICAL TREATMENT  ON
                 DISINFECTION BY-PRODUCTS

           Hiba M. Shukairy,  R. Scott Summers
      Civil  and Environmental Engineering Department
                 University of Cincinnati
               Cincinnati, Ohio 45221-0071

                           and

                    Richard J. Miltner
             Drinking Water Research Division
          U.S. Environmental Protection Agency
                  Cincinnati, Ohio 45268

INTRODUCTION
     Disinfection of drinking water leads to the formation
of disinfection  by-products  (DBFs).  Organic and inorganic
DBFS are  formed  by the reaction of the  disinfectant with
natural organic matter (NOW)  and with inorganic compounds,
respectively.   Some of  these DBFs  are  .regulated or are
being  considered  for  regulation  because  of  the  health
concern   associated  with   them.     In  the   seventies,
trihalomethanes  (THMs) were  the first group of  DBFs that
were regulated with a maximum contaminant  level (MCL)  of
100 M9/L-  This MCL  may  be reduced in the near future and
MCLs may be set for  the  individual species.
     In an  effort to  minimize  DBF formation while  still
providing microbially safe water, modifications in drinking
water treatment have become  necessary.   These  include the
use of  alternative  disinfectants  and more  efficient DBF
precursor removal.
     Microbial regrowth  in the  distribution system can be
controlled  by limiting  nutrient  availability and by the
presence  of   residual  disinfectant.    Ozonation  often
increases the biodegradable  organic  carbon in  the  water
which would serve as  a  nutrient  for the microorganisms.
(Langlais et al., 1991).   Aldehydes and other oxidation by-
products  are  among these biodegradable ozone DBFs.  Ozone
also  reacts   with  bromide   and  generates  bromate,  an
inorganic DBF that may be regulated in the future.
     In the presence of  bromide,  the halogenated DBFs are
a mixture of brominated and chlorinated products. Ozone or
chlorine  can  oxidize bromide, to form  free bromine which
can react with NOM.  Typically, the  speciation shifts to
the more  brominated  DBFs (Amy et al., 1991) as the rate of
bromine substitution is believed to be faster than chlorine
substitution  (Symons et  al., 1987).  Bromate is formed by
the further oxidation of hypobromite.
      Ozonation used in  drinking water  treatment does not
seem   to   significantly   affect   the   organic   carbon
concentration, but does  change the nature of the precursor
compounds.   The  concentration of  halogenated DBFs formed
by subsequent chlorination is often decreased when compared

-------
to that formed with no pretreatment (Reckhow and Singer,
1984, Shukairy and Summers, 1992a, Miltner et al., 1992).
Upon ozonation, the bromide concentration decreases because
of the formation of bromate,  and consequently the decrease
in the ratio of bromide to organic matter leads  to a shift
in DBF speciation to the more chlorinated products.
     The  use of  biotreatment  in  United States drinking
water  treatment  plants  is  very  limited.     Biological
treatment affects the nature of NOM.  The dissolved organic
carbon (DOC) is decreased and the reactivity  to  subsequent
chlorination may be  decreased  as a  consequence  of  the.
oxidation  of  the  precursor compounds.     Biotreatment
downstream  of  ozonation may biodegrade some of the ozone
DBFs and chlorine  DBF precursors and minimize substrate
availability for regrowth (Miltner and Summers, 1992).   A
review of the use of  biotreatment for  the control  of  DBFs
is given  by Shukairy  and Summers (1992b).
     Biotreatment reduces the DOC concentration, while the
bromide  concentration  remains  constant,  resulting in  an
increase  in the bromide to DOC (Br/DOC) ratio.  An increase
in  Br/DOC has been shown to shift the  speciation to the
more brominated DBFs  (Amy et al., 1991).  Upon subsequent
chlorination,  DBF  formation  and speciation  seem to  be a
function of: DOC,  oxidation of  the precursor  compounds,
bromide  concentration and the chlorination conditions that
are  used,   i.e.  pH,   temperature,  holding  time  and the
chlorine dose.   At high chlorine to  bromide  ratios, the
speciation shifts to the more chlorinated byproducts.

OBJECTIVES                              '    .  ,.  „ _ 4.h_
      This paper  summarizes  several  recent studies on the
 impact  of  ozonation  and  biological  treatment  on DBF
 formation.    Ozonation was characterized by the format ion
          .
 of ozonation  DBFs  such as aldehydes, •••1»ll«££no^gg"'
 carbon (AOC) ,  biodegradable dissolved organic carbon (BDOC)
 and the oxidation of bromide to bromate.   The oxidation of
 the  DBF  precursor  compounds  and  the  effectiveness of
 biotreatment  for the control of DBFs were monitored by the
 formation potential  (FP)  for total  organic halogen (TOX),
 total  THMs  (TTHMs)  and  total measured  haloacetic  acid
  (THAAs) .    Special attention was also given to the impact
 of bromide concentration  on the speciation of  the  DBPs.
      A specific objective of this paper was to  compare DBF
 control by batch biological treatment at the bench-scale to
 that by a continuous flow sand filter  at the pilot-scale.


                            was used  in both  the  bench. and
  pilot-scale  biotreatment studies.   Ozone was applied ma
  pilot-scale  countercurrent reactor in both  studies.   For
  the  bench-scale biotreatment,  the experimental design was
  set  un as a  6 by 3  by 2  matrix.   Ozone  was dosed at five
  lelectellevelsand a^ontrol.  Three bromide  levels  5C ).7
  ug/L (ambient),  258, 550  M9/L, were selected.  Each ozone -
  bromide combiAation was  run with a ^without b^otreat ment
  Biotreatment was carried out  in mixed batch  reactors.  The
  water quality parameters examined were:  chlorine  demand,

-------
DOC,  BDOC,  AOC,  spectral absorption  coefficient  (SAC),
bromide and brornate, aldehydes, TOX , THMs and HAAs.
     In comparison, for the pilot-scale biotreatment, raw
Ohio River water was, ozo^ati|d ;a§ an ozone to DOC ratio of
0.7 mg/mg.  After coagulation"^•%locculation and settling,
the flow was split to eignt parallel  filters.  (Miltner and
Summers, 1992).  The results from a fixed bed single-pass
filter with sand will be reported in this paper.

MATERIALS AND METHODS
     Raw Ohio River water was  trucked from the Cincinnati
Water Works to the USEPA facility.  Ozonation was conducted
at  room  temperature   in a  15  cm  (6  in.)  diameter
countercurrent  flow-through   contactor.    The  transfer
efficiency was  greater than 94 percent.   This ozonation
system has been described  in detail (Miltner et al.,  1990).
After each  change in ozone  dose,  samples were collected
after  steady  state  conditions were reached,  which  was
conservatively  chosen  at  two times the T10p time based on
tracer studies.  Applied gas to the contactor and off-gas
from the  contactor  were measured by UV.  Dissolved ozone
from  the contactor  was  measured spectrophotometrically
using the indigo trisulfonate method  (Bader and Hoigne,
1981).      For the  bromide  spike,  a  concentrated stock
potassium  bromide solution  was  added  to the  raw water
before ozonation.  To provide the required concentration as
bromide, the rate of flow  of the stock solution was varied.
Samples were collected before and after contact with ozone.
     For   the   pilot-scale   system,    ozonation,   alum
coagulation,    flocculation,   sedimentation   and   sand
filtration were used.   The sand filter (Filter 5), was  a
3.8 cm (1.5 in.) diameter  glass column containing 76 cm  (30
in.) of  sand (ES= 0.45 mm) supported by 20 cm  (8 in.) of
gravel and 10 cm  (4  in) of coarse sand.
     For the bench-scale biotreatment,  batch reactors were
used.   The  reactor  is a  modification  of the Joret-Levi
reactor  (Joret et al.,  1988).   To provide enough sample  for
formation potential  determination, a 2  L  Erlenmeyer flask
containing 300  ml  of Ohio-River-water-acclimated-sand  was
used with a  sample volume of  1 L.  Vacuum was applied to
the  reactor forcing  the  incoming air  through two water
traps  for scrubbing and  humification.    This  air  was to
provide  enough oxygen  for the bacterial  growth and  for
mixingV The samples were kept in the bioreactors for five
days, and before  formation potential tests, were  filtered
using  prewashed  0.45  /m pore  diameter  (Millipore  HV)
membrane  filters.   Biodegradation  was measured  by  the
difference between the DOC before and after biotreatment.
     DOC was measured by EPA  Method 415.1 and UV absorbance
was  measured  at  254 nm  and  reported  as SAC.   AOC  was
measured  using the  methods  of  van  der  Kooij  et  al.
 (1982,1987).    Two strains  of heterotrophs,  Pseudomonas
fluorescens  P17 and Spirillum NOX,  were employed  in  AOC
determinations.   Bromide and  bromate concentrations were
measured  by ion chromatography by the method of Hautman and
Bolyard (1991).  The minimum  detection  level  for bromate
was  approximately 7 ^g/~L.   Aldehyde concentrations were

-------
measured  by a  modification  of the  PFBOA derivitization
method described by Miltner et al.  (1991).
     Samples were collected for formation potential tests
before and after  ozonation and after biotreatment.   For
formation potential determination, a  12 mg/L chlorine dose
was  used,   at  ambient  pH  (7.5-8.0),  at  20  °C  and  held
headspace-free   in  the  dark  for  seven  days.    After
quenching,  samples were  collected for DBF  analyses  that
were  based on  USEPA methods:  THMs  were measured  by EPA
Method 551; TOX  by EPA Method 450.1 and HAAs by EPA Method
552.
     For  quality  assurance ten  percent  duplication was
used.   Error bars  in  the figures represent the range of
results from the duplicate biotreatment or ozonation tests.
Error bars for  biotreated samples represent the range of
results from two separate reactors.

RESULTS AND DISCUSSION
     The    experimental   matrix   of   the   bench-scale
biotreatment  study  is  summarized by the  DOC results in
Figure 1 which show the impact of ozonation and biological
treatment  on  DOC concentration.  For all  bromide levels,
ozonation  did not  have a  significant impact  on the DOC
concentration.    Biological treatment of  the  raw  water
(O3/DOC ratio equal to 0 mg/mg)  resulted  in  a 13 to 14
percent  removal  of  DOC.   With  ozonation  followed  by
biotreatment, a  range of  20 to 40  percent DOC removal was
observed,  with   the majority of the  effect occurring at
O3/DOC ratios less than 0.8 mg/mg.
    2.0
                                             OZONATION
  cn
  E
    1.2
 8 0.8
 Q
    0.4
    0.0
                                      OZONATION + BIOTREATMENT
 BROMIDE = 50.7 fj.g/L
 BROMIDE = 258 fj.q/1
 BROMIDE = 550 /ug/L
       0
        1                2

TRANSFERRED OZONE/DOC (mg/mg)
Figure 1. BENCH-SCALE BIOTREATMENT:  IMPACT OF OZONATION
          AND BIOTREATMENT ON DOC  CONCENTRATION AT THREE
          BROMIDE LEVELS.

-------
     Biological treatment not only removed the DOC but also
seemed  to lower the chlorine demand  (Figure  2).   In this
case,  ozonation alone  enhanced chlorine  demand  removal,
biotreatment alone decreased the demand by 43 percent, and
the combination of ozonation and biotreatment decreased the
demand  by a range of 60 to  70  percent.
     7
                             OZONATION
                               OZONATION + BIOTREATMENT
        DOC0 = 1.68 mg/L

        BROMIDE = 258 /j,g/l
      0
        1                2

TRANSFERRED OZONE/DOC (mg/mg)
Figure 2. BENCH-SCALE BIOTREATMENT:  IMPACT OF OZONATION
          AND BIOTREATMENT ON CHLORINE DEMAND.

     The halogenated DBFs were also decreased by ozonation
and by  biotreatment.   The  total THM formation  potential
results are presented in  Figure 3.  Ozonation decreased the
TTHMFP by a range of 27 to  46  percent,  and a  28  percent
reduction in TTHMFP was observed  by biotreatment alone.  In
comparison, at all ozone doses,  the effect of biotreatment
was another reduction of 12  to 19  percent.    TOXFP  and
total measured HAAFP  results from this study,  though  not
shown, were also  decreased by ozonation and by biotreatment
(Shukairy et al.,  I992b).

Ozonation DBPs
     BDOC and AOC:  The  formation of biodegradable organic
matter by ozonation is shown in  Figures 4  to 7.   The BDOC
increased  with  increasing   ozone  doses  at  all  bromide
levels, Figure 4.   The  most significant increase  in BDOC
was at O,/DOC ratios less than 0.8  mg/mg.   At higher ozone
doses, the  increase in  BDOC was minimal.     Similar site
specific BDOC behavior  at  higher ozone  doses have  been
reported (Langlais et al.,  1991).   AOC  measured  by both
strains NOX and P17 increased with increasing ozone dose as

-------
    350
     50
      0
                                               OZONATION
                             OZONATION -1- BIOTREATMENT
- DOC0 = 1.68 mg/L

  BROMIDE0 = 550 /j.g/1
       0
                 1                 2
         TRANSFERRED OZONE/DOC (mg/mg)
Figure 3. BENCH-SCALE BIOTREATMENT: IMPACT OF OZONATION AND
           BIOTREATMENT  ON TOTAL THM FORMATION POTENTIAL.
                                      O   BROMIDE = 50.7
                                      A   BROMIDE = 258 M9/L
                                      D   BROMIDE = 550
     0.0
                         1                 2-
                TRANSFERRED OZONE  / DOC  (mg/mg)
Figure 4. BENCH-SCALE BIOTREATMENT:  IMPACT OF OZONATION ON
          BDOC AT THREE  BROMIDE LEVELS.

-------
shown  in  Figure  5.   Most  of
O3/DOC ratios below 1.2  mg/mg.
increases were minimal.
                         this  increase occurred  at
                         At higher ozone doses,  AOC
      1000
  LJ
  X
  o
   I
  O
   CP  400 -
  X
  O
  -z.
   I
  o
  o
      800 -
      600 -
200
V  P17
O  NOX
                                                      200

                                                   UJ
                                                150 O
                                                   CT
                                                   
-------
    2.0
    1.6
    1.2
 O
 O  0.8
 CD
    0.4
    0.0
          OGC0- (1.5 - 52) mg/L
                                     O OZONATEO
                                     * SETTLED
                                     a  FILTER 5 (SAND FILTERED)
                    50
                                 100

                                DAY
                                              150
                                                           200
Figuire 6. PILOT-SCALE BIOTREATMENT: IMPACT OF OZONATION AND
           SAND FILTRATION ON BDOC.
     800
                                                     SETTLED
                                                     FILTER 4
                                                  O FILTER 5
                                                  O FILTER 6
                                                  D FILTER 7
                                                  A FILTER 8
                   50
100
150
200
250
                          RUN TIME,  days
Figure 7. PILOT-SCALE  BIOTREATMENT:  CONTROL OF AOC-NOX IN
           BIOLOGICAL FILTERS.
                                 8

-------
     Aldehydes:   An  increase in  aldehyde concentrations
with ozonation  of Ohio River water  is  shown in Figure 8.
All aldehydes detected, methyl glyoxal, propanal, pent anal,
acetaldehyde and methyl glyoxal, were subsequently reduced
by biotreatment, at all ozone doses,  to concentrations less
than that in the raw water.  Similar decreases in aldehyde
concentrations   attributed  to  biotreatment   have  been
reported  (Shukairy et al., 1992a; Miltner et  al.,  1990),
indicating   that   these   simpler   molecules   are  very
biodegradable.
  UJ
  CJ
  -z.
  a
  o
  Ld
  Q
  >-
  X
  LU
  Q
  _J
                                           DOC. = 1.68 mg/L
O  • METHYL GLYOXAL
V  T PROPANAL
O  • PENTANAL
A  A ACETALDEHYDE
O  * GLYOXAL
   OZONATION + BIOTREATMENT
                       1                2

                TRANSFERRED OZONE/DOC (mg/mg)
Figure 8. BENCH-SCALE BIOTREATMENT:  IMPACT OF OZONATION AND
          BIOTREATMENT ON  THE  FORMATION AND CONTROL OF
          ALDEHYDES.

The biodegradability of aldehydes  in the  pilot-plant is
also apparent from the behavior of  these compounds in the
sand filter  (Filter 5) .    Figure 9  shows the  results for
methyl glyoxal.  For the control of this aldehyde, little
or no time was  required to bioacclimate these filters as
removal  was  near   100  percent  in  the  first  week  of
operation.   The other aldehydes examined  but not shown,
formaldehyde,  glyoxal  and  acetaldehyde,   were  not  as
biodegradable as methyl glyoxal  but the trends  in their
control were similar to that of methyl  glyoxal.

-------
   CH
  _J
  <
  X
  a
o
_j

X
I—
Ld
     ° r
    12
       8
                      A
           V
                                                       •-i
         o
               50
100
150
200
250
                        RUN  TIME, days
Figure 9. PILOT-SCALE BIOTREATMENT: CONTROL OF METHYL
          GLYOXAL IN BIOLOGICAL FILTERS.

     Bromata:  In the presence of bromide, ozone oxidizes
the bromide  to hypobromite ion and then  to  bromate.   At
ambient bromide concentrations,  50.7 M9/L, bromate was not
detected at O3/DOC ratios  less than  0.8 mg/mg.  Increasing
ozone doses  caused bromate levels  to  increase while the
bromide concentration decreased simultaneously (Figure 10).
At the highest O,/DOC ratio studied,  2.54 mg/mg, 50 percent
of the  initial bromide  concentration  had reacted to form
bromate.  At the highest  bromide level investigated, 550
/ig/L, bromate was detected at  the lowest ozone dose in this
study and showed continued increase with increasing ozone
dose to 300 Mg/L,  Figure 11.  At this highest dose studied,
34 percent of the initial bromide concentration had reacted
to form bromate.

DBF Speciation
     The  presence of  bromide in  the water  affects the
speciation  of the DBFs.   Upon  ozonation,    the bromide
concentration decreased due to bromate formation (Figures
10 and  11) .   The DOC of  the  water was not significantly
altered  however (Figure  1).   This led to  a  significant
decrease in  the Br/DOC  ratio.  Figure 12 shows the impact
of ozonation and biotreatment on the Br/DOC ratio.  At all
bromide  levels, the ratio decreased by ozonation compared
to the raw water by as much as 50 percent.
                               10

-------
                                          AMBIENT BROMIDE
                                          DOC = 1.68 mg/L
                       1                2                3
               TRANSFERRED OZONE/DOC (mg/mg)
Figure 10. CONVERSION OF BROMIDE TO BROMATE WITH OZONATION
           AT THE AMBIENT BROMIDE LEVEL.
    600
                                       500 /ig/L BROMIDE ADDED
                                           DOCQ = 1.68 mg/L
                TRANSFERRED OZONE/DOC (mg/mg)'
 Figure 11. CONVERSION OF BROMIDE TO BROMATE BY OZONATION AT
            A TOTAL INITIAL BROMIDE CONCENTRATION OF 550jLig/L.
                                11

-------
With  biotreatment alone, the  DOC decreased significantly
 (Figure   1),  while  the  bromide  concentration  remained
unchanged.   Therefore the Br/DOC ratio increased as shown
in  Figure 12.   The impact  of bromate formation  at high
ozone doses was also apparent in the biotreated samples.
The Br/DOC ratio decreased with  increasing ozone dose as
compared to the raw biotreated water.  In both ozonated and
unozonated samples, biotreatment increased the Br/DOC ratio
in comparison to the nonbiotreated sample.
     o.o
       0.0
                               OZONATION

                                 O
OZONATION +
BIOTREATMENT
BROMIDEQ
50.7 M9/L
258 M9/L  -
550
               0.5      1.0       1.5      2.0      2.5

                TRANSFERRED OZONE/DOC (mg/mg)
                 3.0
Figure 12. BENCH-SCALE BIOTREATMENT:  IMPACT OF OZONATION
           AND BIOTREATMENT  ON  THE BROMIDE TO DOC RATIO.

     The  change  in Br/DOC ratio  is very  important  as it
affects  the  speciation  of  the  DBFs.    As  the  ratio
increases, a  shift to the more highly  brominated species
occurs.   Figure  13  and  14  show the  impact of  bromide
addition on THM and HAA speciation,  respectively,  for the
raw untreated Ohio River water.  Chloro-substituted species
formation decreased as the Br/DOC ratio  increased,  while
the  more bromo-substituted  species  increased.   At  the
lowest Br/DOC ratio  (50.7  M9/L  Br),  chloroform was  the
dominant  species  and was decreased  by  80 percent at  the
highest  bromide  concentration  studied.   At the  highest
Br/DOC ratio,  the dibromochloromethane (CHClBr2) became the
most dominant species, and chloroform the least.  Bromoform
increased significantly with  increasing  Br/DOC (Figure 13) .
 For HAA species,  Figure 14, diehloroacetic acid (DCAA)  was
the  most  dominant  species  at  low  Br/DOC  ratios  and
decreased by 71 percent as the  ratio  increased.  At  the
                              12

-------
^
-z.
o
<
 LU
 O.
 o
 Q_
 O
 a:
 O
    200
     150
    100
    50
RAW OHIO RIVER WATER ""'.
      0.0
          0.1
0.2
                                                 A CHCIjFP
                                                 Q CHCI2BrFP
                                                 O CHCIBr2FP
                                                 O CHBr-jFP
                                                     = 1.6S mg/L
0.3
0.4
                        BROMIDE/DOC  (mg/mg)
Figure 13. IMPACT OF  BROMIDE  ADDITION ON THM  SPECIATION OF
            UNTREATED WATER.
      60
   CT>
      40
   UJ
   o
   Q_
   O
   !<  20

   o
       0.0
              RAW OHIO RIVER WATER
               OOCQ= 1.68 mg/L.
                                            A OCAAFP
                                            Q TCAAFP
                                            O BCAAFP
                                            O OBAAFP
                    0.1            0.2            0.3

                      BROMIDE /  DOC (mg/mg)
                                                              0.4
Figure 14.  IMPACT OF BROMIDE ADDITION ON HAA SPECIATION FOR
             UNTREATED WATER.
                                   13

-------
highest Br/DOC ratio, bromochloroacetic acid  (BCAA) became
the most dominant  species.
     This   shift  in   speciation  would  also   be  more
significant  in  full distribution systems  or in simulated
distribution system studies where the chlorine dose would
be much lower than in these FP studies .  At low chlorine to
bromide ratios the speciation is more shifted to the bromo-
substituted  species.
     The impact of the change in Br/DOC ratio by ozonation
and biotreatment on DBF  speciation can also be inferred
from the individual species behavior.  Figure 15 shows the
control   of   formation   potential  by   ozonation   and
biotreatment of chloroform (CHC13) and bromodichloromethane
(CHClpBr)   at ambient  bromide.   The percent removal  is
calculated relative to the concentration of the particular
species in the control sample,  i.e. the raw water.   For
this study,  chloroform represents 80 percent of the TTHMs
and  its  behavior  is  similar  to  that  of  the  TTHMs.
Oxidation of the precursor compounds by ozonation resulted
in 22  percent removal  of CHC13FP.   This removal  was not
enhanced by  increasing ozone dose.   Similar behavior upon
ozonation was observed for CHCl-BrFP.   With biotreatment,
DOC removal and  biological oxidation
resulted in improved removal of CHC13FP,
raw  water   and   up  to  50  percent
biotreatment .
                                        of the  precursors
                                        30 percent for the
                                        by  ozonation  and
  UJ
  O
  CE
  LU
  a.
      100
       75
      50
     -25



     -50
          DOCQ = 1.68 mg/L
          BROMIDE
                  50.7
        0
                               OZONATION +
                        OZONATION BIOTREATMENT
                          A
                          cr
CHCI-jFP  CQ = 150
CHBrCI2FP CQ = 27.7
                         1                2

                 TRANSFERRED OZONE/DOC (mg/mg)
Figure 15. BENCH-SCALE BIOTREATMENT: PERCENT REMOVAL OF THM
           SPECIES BY  OZONATION AND BIOTREATMENT.
                               14

-------
     The behavior of bromodichloromethane upon biotreatment
was slightly different however.   An increase in the Br/DOC
ratio due to the decrease in DOC with biotreatment resulted
in  17  percent  increased  formation of  CHCl2Br.   As the
bromide  concentration  decreased,  with increasing  ozone
dose, because of bromate formation  (Figure 10) , the Br/DOC
ratio decreased  resulting in  removal of CHCl,BrFP.
     The  more  highly  brominated  DBF   species  displayed
behavioral   trends  very  different  from  that  of  the
chlorinated DBFs.  For dibromochloromethane, at the ambient
bromide concentration and at an O3/DOC ratio of 0.53 mg/mg,
there was an increase of 26 percent in formation as shown
in Figure 16.  Bromate was not detected  at this ozone dose
(Figure  10) ,  hence the bromide was  oxidized  to form free
bromine   and   react  with   the   organic   matter   upon
chlorination.    As  the  ozone doses  increased,  bromate
concentrations  increased,  bromide  decreased,  the  Br/DOC
ratio  decreased and  the speciation  shifted  to  the more
chloro-substituted species  (59  percent decrease in the
CHBr2ClFP at an O3/DOC ratio of 2.54 mg/mg).
     With  biotreatment  at ambient bromide  levels,  the
Br/DOC ratio was even higher than in the raw water leading
to   the   formation  of   CHBr2Cl.      The   trend   upon
ozonation/biotreatment  was similar to  the trend observed
for ozonation only.  A 25 percent  increase in the formation
of CHBr2Cl was observed at an O3/DOC ratio of 1.78 mg/mg.
     For  bromoform, the  behavior was similar  to that of
CHBr2Cl  upon  ozonation.     With  biotreatment  however,
bromoform was removed at all  ozone doses.   As the organic
matter   concentration  was   significantly  decreased  by
biotreatment bromoform  was no longer detected.
      100
          DOCQ = 1.68 mg/L

      75 (- 8ROWIDE0 = 50.7 /j.g/L
 o
 s
 UJ
 UJ
 o
 ct:
 UJ
 CL
-25
     -50
     -75
    -100
        OZONATION 4-
OZONATION  BIOTREATMENT

   O       •
   O       *
                                   CHCI8r2FP CQ = 6.91 /j.g/1:
                                   CHBr^FP  CQ = 2.05 /J.g/1
        0                1.2               3

                 TRANSFERRED OZONE/DOC (rng/mg)

 Figure 16. BENCH-SCALE BIOTREATMENT: PERCENT REMOVAL OF THM
            SPECIES BY OZONATION AND BIOTREATMENT.

                                15

-------
     HAA speciation  trends,  similar to those of the THMs,
were observed at all bromide levels.   Figure 17 shows the
percent control of DCAAFP and DBAAFP  at  ambient bromide.
Similar behavior was also observed for TCAAFP.  The highly
chlorinated  HAA   species  responded   to   ozonation  and
biotreatment in  a manner similar  to those  of  the highly
chlorinated THM species.
     Low ozone doses resulted in 16 to 18 percent removal
of DCAAFP,  and  biotreatment, with  and without ozonation,
resulted in improved removal (50 to 70 percent).  For the
DBAA species, at  a low O3/DOC ratio there  was  35 percent
increase  in  formation.   As the  ozone  dose  increased
however,  the  bromide  concentration decreased  and  a  100
percent removal of DBAAFP was observed at an O3/DOC ratio
of 2.54 mg/mg.  This effect  was  even more pronounced with
biotreatment.
     In  comparison,  on  a pilot-scale  for  Filter  5,  the
behavior  of DCAAFP   is  shown in Figure  18.    Filter  5
averaged 41  percent removal  of  DCAAFP when steady state
conditions were reached.  It  took near  50 days to establish
steady  state  removal  of DCAAFP  in   this  sand  filter.
Generally it took  one and a  half to two and a half months
to establish steady state biodegradation of precursors for
THMs, HAAs and TOX.  This is longer than the time required
to establish control of aldehydes  (only days)  or AOC-NOX
(one  month).    Precursors  of halogenated   compounds  are
larger molecules than the smaller, more biodegradable ozone
DBP molecules.
     100
 LU
 UJ
 O
 OH
 UJ
 Q_
    -100
       OZONATION
          +
OZONATION BIOTREATMENT
                                        •   DCAAFP  50.0 M9/L
                                            DBAAFP  0.40 /j.g/1
               TRANSFERRED OZONE  / DOC (mg/mg)
Figure 17. BENCH-SCALE BIOTREATMENT: PERCENT REMOVAL OF HAA
           SPECIES  BY  OZONATION AND BIOTREATMENT.

-------
     60
  O1
  =1
 cr
 I_L_
 <
 <
 O
 Q
                                             O  FILTER 5

                                             O  FILTER 6

                                             D  FILTER 7

                                             A  FILTER 8
20
      10 -
      0
                 50
                    100
150
200
250
                       RUN TIME,  days
Figure 18. PILOT-SCALE BIOTREATMENT: CONTROL OP DCAAFP IN
           BIOLOGICAL FILTERS.

Bench-Scale versus Pilot-Scale Biodearadation
     The bench-scale bioreactors used in this study provide
five  day  of  contact  time  at  20   °C for  water  and
bioacclimated  sand surfaces.  Mogren et al.  (1990)  and
Joret et al. (1988)  found five days sufficient to provide
optimum biodegradation.   In these  studies  that  provided
greater than 95 percent biodegradation of aldehydes  and 30
to 76 percent of DBF precursors, as seen in Table 1.  These
levels  of  control  were  not  found  in  the pilot-scale
biological sand  filter having 10 minutes contact  time and
temperatures ranging from 17° to 27 °C.  Table 1 shows for
smaller molecules like aldehydes the bench and pilot scale
systems were comparable,  but for  more complex molecules
like DBF precursors  poorer  control was observed under the
pilot-scale real world conditions.
                               17

-------
                         Table l.
          Control of Drinking Water Parameters by
                  Biologically Active Sand
Parameter
Methyl Glyoxal
Forma Idehyde
AOC-NOX
DOC
THMFP
CHC1,FP
HAAFP
DCAAFP
TOXFP
Percent Removal"
Bench-Scale
Bioreactorse'f
99,99
96, d
d
8,31
36,20
58,32
65,68
76,51
43,39
Pilot Plant
Filter 5b'c-9
97 +/- 2
88 +/- 7
47 -I-/- 14
16 +/- 9
17 +/- 5
25 +./- 5
37 +/- 6
41 +/- 6
26 +/- 7
a  O3/DOC approximately 0.7 mg/mg
b  after steady-state control established
c  based on 163 days of operation
d  no data
e  based on Miltner, Shukairy and Summers (1992)
f  based on Shukairy, Summers and Miltner (1992b)
g  based on Miltner and summers  (1992)

SUMMARY
     Ozonation of organic matter leads to the formation of
more biodegradable compounds and to ozonation DBFs.  In the
presence of bromide,  ozonation  also leads to significant
formation of bromate.  For a source water that has a high
bromide  concentration,  the  use  of   ozonation  must  be
carefully   considered  because   of  the   high  bromate
concentrations that could be generated.
     Aldehydes, ozonation DBFs,  are very biodegradable and
can be controlled by biof iltration at both bench and pilot-
scales.
     Ozonation and  biotreatment provide  good control for
halogenated  DBF  precursors.*   There  seems  to  be  no
improvement in  the control by  ozonation  above  a certain
ozone to DOC  ratio.  For Ohio  River  water,  the ratio is
between 0.7 and 1.0 mg/mg.  For  the individual THM and HAA
species, biotreatment is less effective for the control of
the bromo-substituted compounds  because of the increase in
bromide  to DOC   ratio  and   this  effect  is   even  more
pronounced at lower chlorine doses.
     The batch-scale  bioreactors using sand,  with 5 days
residence time, represent ultimate biodegradation. Similar
control  by   biotreatment  was   attainable   for  easily
biodegradable aldehydes in pilot-scale sand filters where
residence times are in the order of minutes.  However, for
                              18

-------
the  more  complex  compounds  like  DBF  precursors, poorer
control was observed in pilot-scale biological filters than
in bench scale bioreactor?.
     On a  pilot-scale, :,biqdegradation of  aldehydes was
established in a matter of days, AOC-NOX in  about a  month
and DBF precursors  in 1.5'to 2; 5 months.
                          *•     *
ACKNOWLEDGEMENTS
     The authors thank  the staff of the Cincinnati  Water
Works,  without  whom  this  study  would   not have  been
possible.   The authors also appreciate the efforts of all
the staff at the USEPA pilot plant.  This work was funded
by USEPA in-house funds and by a  Cooperative  Agreement CR-
816700 between the USEPA and the  University of Cincinnati.
Although the research described was funded by USEPA, it has
not been subjected to agency review and therefore  does not
necessarily reflect the view of 'the agency.

REFERENCES

     AMY,   G.L.;  TAN, L. ;  &  DAVIS, M.K.   The Effects of
     Ozonation  and  Activated Carbon  Adsorption  on THM
     Speciation.  Water Research. 25:2:191 (1991).

     BADER, H. & HOIGNE, J.  Determination of Ozone  in
     Water by the  Indigo Method.  Water Research. 15:449
      (1981).

     HAUTMAN, D.P.  & BOLYARD, M.   Analysis  of Inorganic
     Disinfection  By-products Using  Ion  Chromatography.
     Proceedings.  1991 AWWA, WQTC, Orlando,  Flor.

     JORET,  J.C.  ET AL.   Rapid  Method  for Estimating
     Bioelliminable Organic Carbon in Water.   Proceedings.
     1988 AWWA Conf., Orlando, Flor.

     Ozone in Water Treatment; Application  and Engineering
      (B. Langlais,  D.A.  Reckhow and D.R. Brink,  editors)
     Lewis Publ.,  Chelsea, Mich.  (1991).

     MILTNER, R.J.;  RICE,  E.W.;  &  STEVENS, A.A.   A Pilot-
     Scale  Investigation  of  the Formation and Control of
     Disinfection   By-products.  Proceedings.  1990  AWWA
     Conf., Cincinnati, Ohio.

     MILTNER, R.J.; RICE, E.W. ;  & SMITH,  B.L.   Ozone's
     Effect on Assimilable Organic Carbon, Disinfection By-
     products,  and  Disinfection  By-product Precursors.
     Proceedings.  1991 AWWA, WQTC, Orlando,  Flor.

     MILTNER, R.J.; SHUKAIRY, H.M.;  & SUMMERS,  R.S.
     Disinfection  By-product  Formation  and Control  by
     Ozonation and Biotreatment.  accepted for publication
     Jour. AWWA  (1992).
                              19

-------
MILTNER, R.J.; & SUMMERS, R.S.  A Pilot-Scale Study of
Biological Treatment.  Proceedingsr  1992 AWWA Conf.,
Vancouver, BC.

MOGREN,  E.M.;~ SCARPING,   P.V.   &  SUMMERS,  R.S.
Measurement of  Biodegradable Organic Carbon  in
Drinking Water.  Proceedings. 1990  AWWA Conf.,
Cincinnati, Ohio.

RECKHOW, D.A.;  & SINGER, P.C. Removal of Organic
Halide Precursors by Pre-ozonation  and Alum
Coagulation.  Jour. AWWA. 76:4:151  (1984).

SHUKAIRY,  H.M.  &  SUMMERS,  R.S.   The   Impact  of
Preozonation  and Biodegradation on Disinfection  By-
product Formation. Water Research. 26:9:1217  (1992a).

SHUKAIRY, H.M.  & SUMMERS, R.S.  Biological Treatment
for  the  Control of  Disinfection  By-products:    A
Review.  In  Strategies and Technologies for Meeting
SDWA  Requirements  (R.  M.  Clark  and R.  S.  Summers,
editors) Technomics Publ.,  Lancaster, Penn.  In  press;
(1992b).

SHUKAIRY, H.M.; MILTNER, R.J.;  &  SUMMERS, R.S.
Control of Disinfection By-products and Biodegradable
Organic Matter  through Biological Treatment.   Revue
des Sciences    de  L'Eau.  accepted  for  publication
(1992a).

SHUKAIRY, H.M.; SUMMERS, R.S.; & MILTNER, R.J. Control
and   Speciation  of   Disinfection  By-Products  by
Biological Treatment.  Proceedings. 1992b.  AWWA Conf.,
Vancouver, BC.

SYMONS, J.M.  FU, P.;  DRESSMAN,R.  & STEVENS A.    The
Influence  of Bromide   Ion  on  Organic  Chlorine  and
Organic Bromine Formation  During Free Chlorination.
JOUR. AWWA. 79:9:114  (1987).

VAN  DER  KOOIJ,  D;  VISSER,A.;   &  HIJNEN,  W-iA.M.,
Determining the Concentration of Easily Assimilable
Organic  Carbon in  Drinking  Water.    Jour.   AWWA„
74110:540  (1982).

VAN  DER  KOOIJ,  D.  The   Effect  of  Treatment  on
Assimilable  Organic  Carbon  in  Drinking Water.  In
Treatment of  Drinking Water for Organic Contaminants
(P.M. Huck and  P. Toft, editors)  Pergamon  Press,  New
York NY. (1987).
                        20

-------