EPA/600/A-94/191
PILOT PLAHT INVESTIGATIOH OP ALTERNATIVE TREATMENT METHODS

                            by

                      Mark H. Griese
         Presented  at  the 82nd Annual  Meeting  of  the
      Indiana  Section  American Hater Works  Association,
         February 20-22,  1990, Indianapolis, Indiana
             Evansville Water and Sewer Utility
                     Evansvilie, Indiana
                       February, 1990

-------
    PILOT PLANT INVESTIGATION OF ALTERNATIVE TREATMENT METHODS



                          Mark H. Griese







                           INTRODUCTION



Not unlike every other utility within the State, the Evansville



Waterworks Department is somewhat apprehensive about the impact



which the Amendments to the Safe Drinking Water Act will have on



it's Utility.  In addition to those increased costs associated



with the more stringent monitoring requirements, the promulgation



of several specific  regulations 'may necessitate significant and



costly changes in Evansville's current treatment practices and



treatment plant design.  To more fully evaluate this impact and to



determine the most cost effective approach toward insuring



continued compliance, Evansville has initiated a treatment process



study utilizing a pilot plant testing facility.  Although some



attention is being given to  the  optimization of all current unit



processes, specific  attention is being devoted to the evaluation



of  ozone and to increased  chlorine dioxide capabilities  for the



further  reduction of disinfection by-products.  Evansville has



already  altered its  disinfection scheme  in its full-scale plant to



meet  existing  tribalomethane regulations.  More stringent



impending regulations,  however,  require  additional  consideration



of  viable disinfection  alternatives.



The purpose  of this  paper  is to  describe the process by  which



Evansville implemented  it's  pilot plant  project, to indicate the



extreme  degree of  flexibility which  a pilot unit affords a Utility



 in determining treatment  alternatives, and to discuss some of the

-------
preliminary data which Evansville has generated and how this data
has altered the originally designed approach to the project.
This process study is not Evansville's first experience with pilot
plant application.  In the late 70's and early 80's, the
Evansville Utility engaged in a cooperative effort with the EPA to
evaluate the use of chlorine dioxide and to determine the
effectiveness of granular activated carbon for removal of organic
compounds present in the source water as well as any formed after
chlorine dioxide disinfection.  The pilot plant selected for this
earlier project was a single train 100 gpm Neptune Micro-Floe unit
utilizing the conventional treatment processes of rapid mix,
floceulation, settling, and filtration.  Although rather large and
lacking the  flexibility required  for the current study, this unit
did produce  water which compared  remarkably well with our full-
scale system.   The data derived from this project convinced
Evansville  officials that pilot-plant results could be used to
accurately  project the effects of full-scale plant modifications.

                   IMPLEMENTATION  OF  PROCESS STUDY
As already  indicated, increased speculation that more stringent
 disinfection byproduct regulations could be promulagated as early
 as 1992 led Utility  officials  to  the conclusion that more
 information was needed concerning possible disinfection
 alternatives.   Although.the  low  concentrations of chlorine  dioxide
 currently used as a  pre-treatment measure have proven valuable  in
 maintaining Evansville's  annual  THM concentration well below  100
 ug/L, this treatment methodology  would  not adequately insure

-------
consistent compliance if this MCL is drastically reduced.  It was
the desire to address this issue and to obtain data concerning
those byproducts associated with other disinfection alternatives
that prompted the current project.
Due to the fact that full-scale testing is inefficient,  cost
prohibitive, and could expose the public to a quality of water
that is  less than adequate, it was pre-determined that a major
component of the process study would be a pilot plant evaluation
of potential treatment changes or plant modifications.
Specifications were developed for the procurement of  the needed
professional services necessary to evaluate the current  treatment
processes at Evansvilie's Hater Filtration Plant and  to  develop
treatment and operational recommendations based upon  present  and
proposed federal and state standards.  After  reviewing proposals
by a number  of qualified  engineering  firms with specialized
experience  relating  to  treatment  process studies, the engineering
firm  of  Camp Dresser &  McKee was  selected by  the Utility.  Based
upon  Evansville's objectives and  Camp Dresser & McKee's
 recommendations, the following  Project Approach was developed.
    *  Review Raw Mater Quality data  and Plant  Operations  Records
    *  Raw Water Quality  Characterization
    *  Assess Future  Treatment Requirements
    *  Optimize  Existing  Treatment  Processes
    *  Evaluate  Alternate Oxidants  and  Disinfectants
    *  Implement  Pilot Study
    * Evaluate  Results and Develop Recommendations
    * Develop Conceptual Design  and  Cost  Estimates

-------
                        PILOT PLANT DESIGN

To address what was believed to be the most critical component of
these project objectives,  a pilot plant was constructed by CD&M
that would not only simulate Evansville's conventional methods of
treatment, but would also provide the advantages of parallel
treatment trains, ozonation capabilities, and multiple filtration
columns for evaluating a variety of filter media combinations.
Upon arrival in Evansville, the pilot plant was installed and a
raw water connection was made to an existing low service main.
Provisions were made to, utilize the chlorine dioxide,  chlorine,
and alum  solutions employed within the  full-scale plant.  This was
done  to insure a  representative comparison and  to eliminate  any
inconsistencies which might occur by  using dissimilar  treatment
chemicals.   All water produced by the pilot system  would be  run  to
  •
waste.
The pilot plant  itself consists  of a  number of  individual modules
 that  may  be  interconnected in a  variety of different  ways.   It was
 this  feature that afforded the flexibility  required for  this
 project.   A description  of the individual modules  comprising this
 pilot plant system follows.
 * The Electrical  Distribution Module  provides  and  distributes
   power to the rest  of  the Pilot Plant.
 * The Influent Module performs the function of dividing  the  raw
   water stream into two  paths to supply the parallel  treatment
   systems and is equipped with an in-line turbidimeter for
   continuous monitoring  of the source water.
 * The Ozonation unit is  self-contained and is comprised of  an

-------
 ozone generator, supply-air drier and filter,  ozone monitor,



 ozone destruct unit,  and two 6-inch diameter contact columns.



 The  system  is counter-current with water entering at the top  and



 ozone/air being  introduced at the bottom of each column.  The



 ozonated stream  is then directed to the next downstream process.



 The  availability of two columns permits multi-step ozonation  for



 one  treatment  train or comparative testing in both treatment



  systems.



* The two parallel Rapid Mix and Flocculation Modules are



  constructed entirely of Plexiglas and contain two rapid mix



  compartments and three flocculation chambers.  Each of  the five



  basins has its own independent mixer which provides the



  capability of achieving tapered  flocculation.  Injectors and a



  static mixer are provided  prior  to the  rapid mix  for chemical
 •


  pre-treatment.  Provisions  have  also been made to  feed  chemicals



  between the individual basins  if  so  desired.



 * Since sedimentation  is  one of  the most  difficult  water  treatment



  processes  to model on  a pilot  plant  scale,  the Sedimentation



  Module  uses tube  settlers  to achieve adequate turbidity removal.



  Utilizing  this method  of  sedimentation  for  the 2  gpm flow which



  is  characteristic of both treatment  trains,  the  one hour



  detention  of this  module  achieves settled  water  turbidities very



  representative  of  those in Evansville's full-scale plant.  Like



  the raw water influent module, each  of  these units is  equipped



  with a  combination turbidimeter/recorder for continuous



  monitoring of treatment efficiency.



 *  Settled water is  pumped to four individual 4-inch diameter

-------
 filter columns.  Each filter control  module  is equipped with a
 variable-speed pump which controls  the  flow  through the filters
 and,  like the other key treatment points  throughout the system,
 a combination turbidimeter/recorder.   Each filter  is provided
 with  taps located  at various depths of  filter media which may be
 used  to  monitor  headless  development and/or  turbidity
 penetration.  For  Evansville's project, two filters were filled
 with  sand and anthracite  to simulate the full-scale plants
 present  design.  The  other  two filters contain granular
 activated carbon that  can be operated in parallel  with the  dual
 media filters  or in series  with them as a final  treatment
 measure.

                        BENCH SCALE TESTING
Prior to the  initiation of pilot plant testing,  bench scale
testing was performed to determine if the current  unit  processes
of  mixing, coagulation, and flocculation in the full-scale  plant
were being performed with optimum results for both turbidity
removal and THM precursor reduction.  A  variety of mixing rates,
coagulant dosages, and coagulant types were tested with no
significant improvement over existing  full-scale plant production.
In an  attempt to gain some preliminary data concerning ozone  and
chlorine  dioxide dosages and associated  THM reduction,  bench  scale
testing was also performed using various ozone and chlorine
dioxide dosages and then monitoring  the  treated water for 3-day
trihalomethane formation potential.   Although the reductions  in
THMFP  were somewhat less than expected for  ozone, the results of

-------
this test did confirm the effectiveness of both oxidants in
reducing tribalomethane  formation and gave some preliminary
indication of the  results which  could be expected by utilizing
various dosages  in the pilot  plant process.(Figures 1 & 2)

                     PILOT PLANT  TESTING PROGRAM
With the pilot plant in  place and preliminary bench scale testing
completed, continuous  flow  pilot plant testing was initiated to
further  evaluate various treatment options.  The initial testing
period was timed to coincide with that time  of year in which the
highest  trihalomethane formation potential is experienced by the
Evansville Utility.  The key objectives  of this portion  of  the
process  study are to:

            1. Attain Primary Disinfection
            2. Minimize Disinfection  Byproducts
            3. Lower Annual  Average THMs  to < 50 ug/L
            4. Select Secondary Disinfectant

 To accomplish these objectives, the  following  appeared  to be
 Evansville1s treatment options:

 1. Eliminate Chlorine Dioxide — Use Ozone
 2. Keep Chlorine  Dioxide — Add Ozone
 3. Keep Chlorine  Dioxide -- Add Granular Activated Carbon
 4. Keep Chlorine  Dioxide -- Add Ozone &  Granular  Activated  Carbon

-------
           OZONE DOSAGES  VS.  THMEP
      EVANSVILLE - BENCH SCALE  TEST //2
en
CL
b_
        no
        100
        70
        60
        50
                                                  CO
         0
0.5
1.5
                   OZONE DOSAGES (mg/L)

-------
  CHLORINE DIOXIDE  DOSAGES VS.  THMFP
EVANSVILLE - BENCH SCALE TESTS #4 & #5
     no
     100
                                            «M
       0
0.5
1.5
             CHLORINE     DOSAGES (mg/L)

-------
Chlorine dioxide had certainly proven advantageous in reducing
THM concentrations in the full-scale system.  Speculation,
however, that the combined residual of chlorine dioxide and the
byproduct ions of chlorite and chlorate would be regulated at a
level significantly below the current rscommended maximum of 1
mg/L, indicated that increasing the current chlorine dioxide
capabilities alone would not be adequate for addressing the key
objectives.  The increased chlorite ani chlorate concentrations
resulting from elevated chlorine dioxide dosages would need to be
effectively  removed prior to entering the distribution system.
Since a limited amount of information is available that indicates
GAG is  relatively effective in accomplishing this removal, it was
believed that any treatment scheme utilizing chlorine dioxide
alone as the primary disinfectant would have to  also include GAC
capabilities.  The  other  options available  would be to maintain
the current  chlorine dioxide  levels and supplement pre-
disinfection and  pre-oxidation with ozone,  or  to eliminate
 chlorine dioxide  entirely from the treatment scheme.  With
 preliminary  estimates  that the addition of  ozone or combined ozone
 and GAC capabilities could cost the Utility in excess of  10
 million and  50  million dollars respectively, the reason  for
 effectively  reviewing  and optimizing  each  treatment option was
 apparent.
 Each individual  pilot  plant  test was  designed  to last  from 3 to  5
 days.   With  the wide variety  of treatment  options  to be  addressed,
 and the fact that a distinct  impact would  have to  be observed by
 any one treatment combination to meet the  desired  objectives, this

                              10

-------
schedule would  provide  the roost optimum use of the pilot plant
during that time of  year when the DBF formation was greatest.
Although not a  surrogate for all disinfection byproducts,
trihalomethane  formation potential was selected as that parameter
to be used to determine the preliminary "success" of each
treatment option.  In addition to trihalomethane reduction being
one of  the key objectives of the project, in-house analytical
capability for this parameter would expedite data turn-around and
the refinement of future scopes of work.  Upon the generation of
sufficient data to determine the most optimum methods  of
treatment, longer test periods would be implemented to verify
earlier results, to optimize all unit processes involved  in  these
particular methods, to gather data on other disinfection
byproducts, and  to determine the impact of raw water seasonal
variations.
Test  No. 1  of  the pilot  plant study was a  three day evaluation  of
THM precursor  oxidation  utilizing elevated levels  of chlorine
 dioxide as  a pre-oxidant.   Current application of  chlorine  dioxide
 in Evansville's  full-scale  treatment system seldom exceeds  1.0
 mg/L.  This level of treatment  has been selected  because  it is
 adequate for maintaining annual  THM  levels below  the current MCL,
 and in consideration of  minimizing residual chlorite and  chlorate
 in the finished water.   Since  this dosage  has proven inadequate
 for reducing annual  average THM formation  levels  below the  50  ug/L
 level specified as  a project objective,  concentrations of 1.5  mg/L
 or higher would be  applied during the  process study.
 One treatment  train of the pilot plant  was pre-treated using an
                               11

-------
average dosage of 2 mg/L of chlorine dioxide while the second
train was pre-treated with sufficient chlorine to maintain a free
residual throughout the process.  Finished water samples were
collected from each of the two treatment trains, stored in the
presence of chlorine for a period of three days at a temperature
currently representative of the distribution system, and then
analyzed for THM formation potential.  A river water sample was
also analyzed to monitor changing formation potential  conditions.
Figure  3 shows the  results of Test  fl.  The three  day  THMFP for
the  river water  was approaching 300 ug/L and  the  finished water
from the pre-chlorinated train  exhibited a  formation potential of
174  ug/L.   The sample  collected from the pilot  plant system
utilizing pre-chlorine dioxide  oxidation,  however,  showed a
significant reduction  in  formation  potential  with a  three day
average of  only  26.8 ug/L.  This  reduction apparently  indicates  a
 considerable oxidation of  THM precursor material  by  the elevated
 chlorine dioxide dosage of 2  mg/L.
 Test No.  2  was  designed to provide  a direct comparison between
 a water treatment system using  pre-ozonation and one using  pre-
 chlorine dioxide addition.  The raw water  supplying one of  the two
 pilot treatment trains was diverted through the ozonation module
 where it was treated with an average ozone concentration of  1.2
 mg/L.   The second pilot train was again pre-treated with chlorine
 dioxide.  To provide a more direct  comparison of the two
 preoxidants, the chlorine dioxide dosage was reduced to 1.5 mg/L.
 Figure 4 indicates the results achieved in Test 12.   Although
 exhibiting a distinct reduction in formation potential as compared
                               12

-------
    THM FORMATION POTENTIAL  COMPARISON


      EVANSVILLE - PILOT PLANT TEST #1
L
i_

E
       300
       250
       200
        150
        100
        50
         0
                RIVER
 PRE-CI2   PRE-002


SAMPLE TYPE

-------
TIIM  FORMATION POTENTIAL COMPARISON
  EVANSVILLE -  PILOT PLANT TEST
    350




    300




    250




    200




    150




    100




     50




     0
             RIVER
                              o
PRE-OZONE   PRE-002



SAMPLE TYPE

-------
to the earlier tested pre-chlorination  process, the ozone treated
water still exhibited a THM formation potential in excess of that
water pre-treated with chlorine dioxide.   The  increase in THM
levels in  the chlorine dioxide system compared to those in Test 1
was more than likely due to both  the increase  in the  raw water
formation  potential and the 0,5 mg/L dosage reduction.
To again determine if the  conventional  treatment methods of
coagulation  and  flocculation  could be  further  optimized for the
removal of THM precursors, Tests  3 &  4  of the  pilot study were
geared  for a direct comparison  of ferric chloride and the
currently  employed aluminum sulfate.   Test 3 compared the two
coagulants directly with  pre-chlorination being  practiced in both
treatment  trains.  Nearly  identical THM formation potential was
observed  for both systems.  In  Test 4,  the water in  both pilot
treatment  systems was pretreated with ferric chloride and the pre-
disinfectant scheme  utilized  in Test 2 was repeated.   As in Test
 2,  both systems  exhibited significant reductions in  THMFP as
 compared to the raw  water.  Although a slight improvement was
 noted for the system pretreated with ozone, the difference  was
 minimal and it was determined that a significant reduction  in DBP
 formation could not  be expected by altering coagulants.
 The treatment applications used in Pilot Test No.  5  would again
 compare identical dosages of chlorine dioxide and ozone.   In  this
 test, however, ozone treatment would occur, not as a pretreatment
 measure,  but after the treatment processes of coagulation,
 flocculation, and sedimentation.  This study was designed to
 determine if the efficiency  of ozone oxidation could be increased

                               15

-------
by altering its point of application.  As can be seen in Figure 5,
that treatment train being pre-treated with chlorine dioxide again
exhibited the greatest reduction in disinfection byproduct
formation potential.  Although the THM levels of both systems were
somewhat higher due to tie increasing formation potential of the
raw water, pre-oxidation using chlorine dioxide still revealed the
greatest efficiency in reducing this disinfection byproduct.
Tests  6 & 7 compared chlorine dioxide treatment to a treatment
system utilizing both pre- and settled ozone application.   Due to
a  continued increase in  raw  water  formation potential,  both
systems were  treated with  2.5 mg/L of the respective
disinfectants.  The  entire 2.5 mg/L  dosage of chlorine  dioxide
was  applied as  pre-treatment while the ozone was applied  as a pre-
treatment measure  and  in the settled water at 1.5 mg/L  and  1.0
mg/L respectively.   Figure 6 shows the results  of Test  7.   The
elevated  raw  water formation potential was impacting  both
treatment  systems.   By  comparison, however, the chlorine  dioxide
 treated system still  revealed  significantly lower THM levels.
At this point in  the project,  it  was beginning  to appear  that, at
 least for  Evansville's  raw water  source,  chlorine dioxide was
 going to be the pre-disinfectant  of  choice.  The Evansville
 Utility had anticipated ozone  oxidation  to prove more efficient  in
 reducing disinfectant byproduct  formation.  Although  an estimated
 10 million dollar capital investment would be needed  to implement
 ozone treatment in Evansville's  full-scale plant, this  alternative
 was obviously more attractive  than the 40 to 50 million that would
 be needed for GAG to remove  the  chlorite and chlorate associated

                               16

-------
THM  FORMATION  POTENTIAL COMPARISON
  EVANSVILLE - PILOT PLANT TEST #5
    500
    400
    300
    200
    100
     0
                                                M
                                                U,
            RIVER  SETTLED-OZONE PRE-CI02
                   SAMPLE TYPE

-------

CL
    THM FORMATION POTENTIAL  COMPARISON




      EVANSVILLE - PILOT PLANT TEST #7
        500
        400
        500
        200
        100
         0
                                                    vO
                                                     CO
                 RIVER   TWO STAGE OZONE   PRE-CD2
                       SAMPLE TYPE

-------
with increased chlorine dioxide capabilities.
Pilot Test No. 8 repeated the comparison made in Test 2 with ozone
and chlorine dioxide both being applied as a pretreatment measure
at the higher 2,5 mg/L level.  This test was performed to
reconfirm that single stage ozonation was as effective at reducing
THMFP as was  the dual stage application performed  in  the previous
tests.  As  in Test. 2, chlorine dioxide again exhibited the  greater
reduction at  a percentage similar  to those in the  previous  test
modes.
Hot  to  be dissuaded  from our  original perception,  one last  attempt
was  made  to improve  those results  obtained by ozone oxidation.   A
number  of studies have  indicated  that small  amounts of  hydrogen
peroxide  used in  conjunction  with ozone  application greatly
 enhance oxidation capabilities.   To test  the applicability  of
 "Peroxone"  technology in Evansvilie's  process,  Tests 9  &  10
 directly compared a  treatment system using pre-ozonation  to one
 using pre-ozonation  and hydrogen  peroxide addition.  Figure 7
 reveals the results  of  Test 10 after the hydrogen peroxide  dosage
 used in Test 9 was  increased from 0.5 mg/L to a concentration  of
 1.0 mg/L.  This test was performed for an entire week with every
 effort made to identify some improvement in the ozonation/peroxide
 process.  As evident in Figure 7, no improvement was noted.
 Although both trains exhibited formation potential variations
 which compared remarkably well with those in the raw water, the
 numbers generated for both treatment processes were almost
 identical.
                               19

-------
    TI-IM  FORMATION POTENTIAL COMPARISON
      EVANSVILLE  - PILOT  PLANT TEST #10
jr~ ^.


en
Q.
LJL.
        500
        400
300
        200
        100
         0
                                             RIVER
                                              03
                                           03 & H202
                      TEST DATES

-------
                         NEW  SCOPE  OF WORK
When the Pilot Plant study was  still in  its  infancy, Evansville
made a request to the  EPA Research  Center  in Cincinnati, Ohio  for
supplemental analytical  support for disinfection byproducts.   In
consideration of the EPA's plans for the development of
regulations concerning these  byproducts, it  was believed that  the
data generated could assist  the EPA in  its regulatory  development.
As  the data began  to indicate the viability  of chlorine dioxide
for primary disinfection,  the EPA informed Evansville  of bench
scale work being performed by Dr. Gil  Gordon of Miami  University
in  Oxford, Ohio  on minimizing the amount of  chlorite and chlorate
in  drinking water  treated with chlorine dioxide.   Dr.  Gordon had
discovered  that  sulfur dioxide could be used to quantitatively
 remove  the  chlorite ion concentration to below the 0.1 mg/L level.
This  process,  coupled with the minimizing of chlorate
 concentrations by proper chlorine dioxide generation,  would permit
 the use of  higher concentrations of chlorine dioxide with  the
 elimination and/or removal of  chlorite  and  chlorate ion residuals.
 Evansville now had a  fifth treatment option.  Although this method
 of treatment, to our  knowledge,  has not been performed outside of
 the laboratory, the implications of this technology warrant our
 investigation.  Because this research could determine  the
 viability of chlorine dioxide  as a  Best Available Technology
 under the Disinfection Byproduct Rule,  the  EPA is now supporting
 Evansville both financially  and  analytically in this endeavor.
 In consideration of the previously  generated data and the mutual
 benefits which  could  be derived by  both Evansville and the OSEPA,

                                21

-------
a new scope of work was developed  to  fully  evaluate the use of
chlorine dioxide as a primary  disinfectant  and  the reduction of
its byproducts by sulfur  dioxide application.
Prior to the initiation of  the revised scope  of  work, a number of
additions were made to the  Pilot Plant to better evaluate this
treatment process.
A  30 pound  per day chlorine dioxide generator was donated to the
Utility  by  Rio Linda  Chemical  Company to be used specifically for
the  pilot study.   In  tests  1 - 10, chlorine dioxide from the full-
scale generation  system  had been transferred to a day  tank  for
application to the pilot  plant.  Tests indicated that  an increase
 in chlorate formation was occurring during this period  that was
probably due to  the  presence of a low level chlorine  residual in
 the batch solution.   Since control of the chlorate  ion  would need
 to be  accomplished by optimizing the generation process and by
 removing the chlorite ion prior to post-chlorination,  this  new
 generator was installed with the majority of the production stream
 going  to the full-scale plant and a small, but  accurate,
 percentage going to  the pilot plant.  This "slip streaming"
 procedure was accomplished by utilizing  chemical resistant
 variable-area flowmeters.  Based upon the analytically determined
 concentration of the original  chlorine dioxide solution,  the
 proper milliliter per minute setting is  made to achieve the
 desired disinfectant dosage.   This method of chlorine dioxide
 application is more accurate  than the previously employed
 procedure  and it takes advantage of the  newly installed
 generator's efficiency.

                                22

-------
Secondary settling basins have also been incorporated into the
pilot plant to add an additional 90 minutes of settling and
contact time to the treatment system.  This treatment addition was
made to more closely simulate Evansville's full-scale plant, to
provide a location for pH adjustment if so desired, and to provide
additional contact time for the chlorine dioxide prior to the
addition of sulfur dioxide.  This additional contact time would
helf insure the bacteriological integrity of the water supply and
assist in monitoring the presence of a chlorine dioxide residual
beyond primary settling.
A sulfur dioxide  system is also being added to the Pilot Plant.
Based upon the same principal as the chlorine dioxide system,
precision  rotometers will be utilized to "slip stream" a small
percentage of an  inductor generated  sulfur dioxide solution.  This
stream will be added immediately prior to the dual media filters.
The fourth and final addition to the Pilot Plant was the
attachment of four  20  gallon  clearwells to each filter effluent.
These  clearwells  will  permit  post-chlorination for the removal  of
excess sulfur dioxide  and sampling points representative of
finished water quality.
A fifth  piece of  equipment  was  installed, not in the Pilot  Plant,
but in the laboratory.   An  ion  chromatograph  for the detection,
monitoring,  and  accurate  quantitation of chlorite  and  chlorate  has
been obtained.   Generating  chromatorgrams similar  to that produced
 in gas chromatography,  this instrument can accurately  detect both
 ions at  levels  below 0.1  mg/L.   Since it was  nearly impossible  to
 monitor  these  types of levels with the previously  employed
                               23

-------
amperometric measurements, this instrument was essential for
determining the success of the new treatment process.
With these additions in place, the revised scope of work was
initiated.
The revised scope of work consists of five basic components.
1- Optimize Generator Ifficienc^ - A sufficient number of analyses
   on  the chlorine dioxide production stream will be performed  to
   insure that the generator  is producing chlorine dioxide with
   minimal chlorine and chlorate residuals.
2. Bench  Scale Studies fjojr Reduction/Reroova 1 of. Chlorine Dioxide
   and Chlorite - Bench scale studies will be conducted by the  EPA
   to  determine optimum dosages of sulfur dioxide for achieving
   chlorite and chlorine dioxide removal.  Other dechlorinating
   agents such as PAC and ferrous ion may be evaluated to
   determine  their potential  for accomplishing the same task.
 3. Demonstrate the Best Method vs. GAC -  The major operational
   scheme of  this phase will  consist of  adding the sulfur dioxide
   prior  to the dual media for reducing  chlorine dioxide and
    chlorite.   Post chlorination will then occur to remove any
    excess sulfur dioxide.  This process  will be compared to  one
    utilizing  the GAC columns  for removal of the same constituents.
 4.  Comparison of Chlorine Dioxide and Ozone - Since  chlorine
    dioxide  and ozone are  probably the only disinfectants that  can
    be applied to the raw  water and  still meet the disinfection
    byproduct  regulations, they will  be  evaluated in  parallel in
    this phase.
                               24

-------
5.  Optimization of Unit Processes - The purpose of the Phase of



   the pilot study will be to optimize the most desirable



   application by varying other unit processes.  The effects of



   alternative coagulants, varied pH ranges, and various points of



   disinfectant application are examples of those unit processes



   that will be examined during this study period.





Throughout the project, formation potential using chlorine will be



done on collected samples to determine the impact of post



chlorination.  Compounds and surrogates that will be evaluated



include trihalomethanes, haloacetonitriles, haloacetic acids,



chloral hydrate,  chloropicrin, chloropropanone, total organic



halide, and  total organic carbon.  Other byproducts, such as the



aldehydes and  ketones,  will also be monitored  and evaluated.



Phases  1 and 2 of the  Project  are now  ongoing.  The EPA  is



currently involved in  the bench scale  evaluation while Evansville



is optimizing  the chlorine dioxide generation  efficiency and



installing  the sulfur  dioxide  system.  Completion of the project



is expected  by late summer.







                            CONCLUSION



 In a rapidly changing  regulatory environment,  this  project  is



 viewed by the  Evansville  Utility as an investment in the future



 that could  now potentially save  the department millions  of



 dollars.  Although not applicable  to  everyone, the  specific



 research being conducted  by Evansville certainly  underscores the



 flexibility which a pilot plant  facility  provides a water utility.
                               25

-------
Whether it is used to optimize a current treatment process, to
determine the feasibility of retrofitting a new treatment
technique into an existing plant, or to design and size a new
facility, pilot plant testing certainly provides a utility with a
cost-effective approach for addressing these issues.
                               26

-------
                                    TECHNICAL REPORT DATA
                            /Please read Instructions on the reverse be/ore comnletm
1 REPORT NO,
  EPA/600/A-94/191
4. TITLE AND SUBTITLE
   Pilot Plant Investigation of Alternative Treatment
   Methods
                                      6. PERFORMING ORGANIZATION CODŁ
                                                             3. R
                                                             5. REPORT DATE
7. AUTHOR(S)

 Mark H. Griese
                                                            8, PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
   Offic$ of Research and Development
   Drinking Water Research  Division
                                                             10. PROGRAM ELEMENT NO.
   US  EPA
   Cincinnati, Ohio
                                                             11. CONTRACT/GRANT NO.
45268
12. SPONSORING AGENCY NAME AND ADDRESS
                                                             13, TYPE OF REPORT AND PERIOD COVERED
   Risk  Reduction Engineering Laboratory,  Cincinnati,OH
   Office of Research and Development
   OS Environmental Protection Agency
   Cincinnati, Ohio  45268
                                           Symposium Paper
                                      14. SPOroscJMING AGENCY CODE
                                                       :YC(
                                      EPA/600/14
15. SUPPLEMENTARY NOTES
   Presented at the 82nd Annual Meeting  of the Indiana  Section American Water Works
   Association, February 20-22, 1990, Indianapolis, Indiana  PO=Ben1amin Lvkins	
16. ABSTRACT                  TPgT^ZT^                             (513) 559.7450
   Describes the process by which Evansville implemented  it's pilot plant project, to
   indicate the extreme degree of flexibility which a pilot unit affords  a Utility
   in determining treatment alternatives,  and t,Q~ discuss some of the preliminary
   data which Evansville has generated and how this data  has altered the  originally
   designed approach to the project.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                               b.lDENTIFIEHS/OPEN ENDED TERMS
                                                       COSATI Fiefd/Group
18. DISTRIBUTION STATEMENT

  Release to Public
                         19. SECURITY CLASS (This Report}
                          Unclassified 	
                         20. SECURITY CLASS (This page)
                          Unclassified
21. NO. OF PAGES

    2S_
                                                    22. PRICE
EPA form 2220-1 (R»». 4-77)   PREVIOUS EDITION is OBSOLETE

-------