GROUP lj  PHASE I
                SUPPLEMENT TO DEVELOPMENT DOCUMENT FOR
                 EFFLUENT LIMITATIONS GUIDELINES AND
                EW SOURCE PERFORMANCE STANDARDS FOR THE
                     CORN VET MIIUNG SUBCATEGORY
                          GRAIN PROCESSING
                           SEGMENT OF THE
                            GRAIN MILLS


                       POINT SOURCE CATEGORY
                            WGUST 1975

            i   "•$• EIWIfOffNTAL PROTECTION AGENCY
                      WASHINGTON, D, C.  20460

-------
f  >
f *
  S*

-------
   SUPPLEMENT TO DEVELOPMENT DOCUMENT

                  for

    EFFLUENT LIMITATIONS GUIDELINES

                  and

    NEW SOURCE PERFORMANCE STANDARDS

                for the

      CORN WET MILLING SUBCATEGORY
    GRAIN PROCESSING SEGMENT OF THE
   GRAIN MILLS POINT SOURCE CATEGORY
            Russell E. Train
             Administrator

             James L. Agee
        Assistant Administrator
                  for
     Water and Hazardous Materials
              Allen Cywin
 Director, Effluent Guidelines Division

           Richard V. Watkins
            Project Officer
              August 1975
      Effluent Guidelines Division
Office of Water and Hazardous Materials
  U.S. Environmental Protection Agency
        Washington, D. C.  20460
                                               ,.*

-------

-------
                         TABLE OF CONTENTS

SECTION                                                            PAGE

     I     CONCLUSIONS                                               1

    II     RECOMMENDATIONS                                           3

   III     INTRODUCTION                                              5

           A.    Court Findings                                      5

           B.    Court Directives                                    6

           C.    Purpose of this Report                              6

           D.    Organization of this Report                         6

           E.    Sources of Data                                     7

    IV     EFFECTIVENESS OF BIOLOGICAL TREATMENT                   11

           A.    Municipal Wastewater Applications                 11

           B.    Industrial Wastewater Applications                 12

           C.    Corn Wet Milling Industry                          13

                 1.    CPC - Pekin,  Illinois                        19
                 2.    CPC - Corpus  Christi,  Texas                 21
                 3.    American Maize - Hammond,  Indiana           22
                 4.    Clinton Corn  Processing Company -
                         Clinton, Iowa                              23
                 5.    Pretreatment  Facilities                     23

           D.    Summary                                            24

     V     IN-PLANT CONTROLS                                        25

    VI     EVALUATION OF FILTER TECHNOLOGY                          27

           A.    Treatment of Water  Supplies                        27

           B.    Filtration of Treated Wastewater                  28

           C.    Expert Opinion on Applicability of
                   Filtration to Corn Wet Milling Wastes           39

           D.    Siunmary                                            40

-------

-------
                     TABLE OF CONTENTS  -  (cont)
SECTION

   VII     CLINTON CORN WASTE TREATMENT                           43

           A.    Background and  Pollution Abatement
                   Efforts                                         43

           B.    Evaluation of Treatment Plant
                   Performance                                     44

           C.    Performance of  Deep  Bed Filters                  46

           D.    Summary                                           46

 VIII      ABILITY OF CORN WET MILLS  TO MEET NEW SOURCE
           PERFORMANCE STANDARDS                                  51

           A.    Design Criteria                                  51

           B.    Waste Treatment Components                       52

           C.    Performance of  Waste Treatment
                   Facilities                                      53

           D.    Effects of Product Mix  58                       56

           E.    Costs of Compliance  with New Source
                   Performance Standards                          57

    IX     NONWATER QUALITY ASPECTS                               59

     X     SUMMARY                                                 61

           REFERENCES                                              69

           APPENDIX A                                              75
                             ill

-------

-------
                             SECTION I

                            CONCLUSIONS


An  extensive  review  of  available  data  on  the  application  of
activated  sludge  and  deep bed filtration for treatment of various
wastewaters was conducted.  Sources of data included  the  technical
literature,   field  inspections,  special  consultants  and  expert
opinions, corn wet milling companies, and  equipment  manufacturers.
The  data  unequivocally and unmistakably substantiate the fact that
high strength biodegradable organic wastes, such as those  generated
by  corn  wet  mills,  can  be  successfully treated with biological
treatment processes,, particularly  complete  mix  ac-civated  sludge.
With  proper  design and operation of treatment facilities, a stable
high quality effluent can be attained on a  reliable  and  sustained
basis.
                                         i
Ample  evidence  also exists to demonstrate that deep bed filtration
is being applied successfully outside the corn wet milling industry.
Filtration is used to treat potable water  supplies,  effluent  from
domestic sewage plants, and biologically treated effluents from many
industrial waste treatment operations.  Filtration can be applied to
the  corn  wet milling industry, and, in fact, is being successfully
employed by one company in the industry, despite  inadequacies  with
in-plant controls and the preceding biological treatment process.

The  treatment  of  variable,  high  strength  wastewaters is not an
enigma in  sanitary  engineering  and  pollution  control  practice.
Indeed,  numerous  treatment  applications  have  been  made in many
industries similar to corn wet milling, all with successful results.
Not only have  high  degrees  of  pollutant  removal  been  achieved
through biological treatment, but additional pollutant reductions or
effluent  "polishing"  have  been demonstrated with filtration.  The
performance of these measures with corn wet milling  wastes  can  be
reasonably and unmistakably predicted.

On the basis of this study, it is concluded that the New Source Per-
formance  Standards  as  promulgated  can  be  met  by  applying the
technology prescribed for the corn wet milling  subcategory  in  the
Development  Document  for  the  Grain Processing Industry.  Data on
treatment of similar wastes strongly  indicate  that  the  standards
quite  probably can be achieved through biological treatment without
filtration.  Deep bed filtration, as with other "polishing" devices,
provides additional assurance of maintaining a high quality effluent
and minimizes or reduces the effects of biological  treatment  plant
upsets.   Such  upsets  can  usually  be attributed to poor in-plant
control,  faulty  design,  improper  operation  and   human   error.
Polishing  mechanisms  such as filtration reduce the effects of such
circumstances and are well demonstrated in  achieving  high  quality
effluent on a long-term basis.

-------
The  costs  required for new corn wet mills of various sizes to meet
the New  Source  Performance  Standards  were  reevaluated  and  are
indicative of current pollution control technology.   These costs are
based  on  January  1975  dollar  values and include waste treatment
facilities  and  necessary  in-plant  controls  or  cooling   system
designs.   Evaluation  of  nonwater  quality aspects of applying the
recommended technology indicated that  energy,  air   pollution,   and
solid waste impacts will be minimal.

-------
                             SECTION II

                          RECOMMENDATIONS
Based  on  an  extensive review of technical data, it is recommended
that the _New Source Performance Standards for the Corn  Wet  Milling
Subcategory be implemented as promulgated on March 20, 197U.  As the
data  reviewed  demonstrate,  new  plants  employing  best available
demonstrated control technology can readily achieve these standards.
The New Source Performance Standards as recommended are as follows:

                 BQD         Suspended Solids        ]DH

    kg/kkg       0.357            0.179
    Ibs/MSBu    20.0             10.0
    units                                            6-9

-------

-------
                            SECTION III

                            INTRODUCTION


On May 5r 1975 the U.S. Court of Appeals  for  the  Eighth  Circuit,
issued  its  decision on Effluent Limitations Guidelines, New Source
Performance Standards, and Pretreatment Standards for the  Corn  Wet
Milling  Subcategory  of  the  Grain  Milling  Industry.   The Court
determined that it had  no  jurisdiction  over  Effluent  Limitation
Guidelines;  but  indicated  this  is  a matter to be decided by the
District Court.  The Court did find it  had  jurisdiction  over  New
Source Performance and Pretreatment Standards.

A.  COURT FINDINGS

In  reviewing  the New Source Performance Standards for the Corn Wet
Milling Subcategory, the court noted the New Source limitations  are
identical  to  the  1983 guidelines.  The 1983 guidelines assume the
technology  available  to  meet  the   1977   guidelines   will   be
supplemented  by additional technology in 1983.  After reviewing the
basis  for  the  1977  guidelines,  the  Court  concluded  that  the
recommended 1977 technology, when employed in a new corn wet milling
plant,  would  enable that plant to comply with the 1977 guidelines.
However, the Court concluded that the record does not support  EPA's
determination  that  technology  is available to meet the New Source
Performance Standards.  This technology, in the Court's view,  would
be  required  to remove an additional 30 pounds of BOD and 60 pounds
of suspended solids per MSBu (0.536 kg/kkg BOD and 0.714 kg/kkg TSS)
beyond the 1977 guideline  limits  of  50  pounds  per  MSBu  (0.893
kg/kkg)  of  both  BOD  and  suspended solids.  The Court felt that,
according to the record, deep bed filtration was being called on  to
provide  most  of  the  incremental  reduction  in BOD and suspended
solids.  They concluded that within the record there are no concrete
data, test results, literature, or expert opinion in support of  the
"prediction" that filtration will permit a new corn wet mill to meet
the New Source Performance Standards.  To base standards on transfer
technology,  the  Court stated that EPA must:  1) determine that the
technology  is  available,  2)   determine  that  the  technology  is
transferable,   and   3)  make  a  reasonable  prediction  that  the
technology will be capable of removing the increment required by the
New Source Performance Standards.

The Court also concluded that the costs required in adopting the New
Source Performance standards were  not  covered  adequately  in  the
record.   Two  problems  were  cited.   First,  EPA  did not project
separate  capital  and  operating  costs  for   control   technology
implemented  at  new plants.  Second, EPA used 1971 prices in making
cost estimates,  despite  the  fact  that  more  current  data  were
available.

-------
B.  COURT DIRECTIVES

On the basis described above, the Court remanded the New Source Per-
formance Standards to EPA.  Their directive was that EPA, within 120
days,  either furnish support for the present New Source Performance
Standards or establish new standards.

C.  PURPOSE OF THIS REPORT

The purpose of this report is to review the basis for the  Corn  Wet
Milling  New  Source  Performance  Standards,  review the technology
recommended to meet these standards, and recommend further action to
EPA in following the Court's directives.   This  report  presents  a
detailed evaluation of the treatment technologies identified to meet
the  New  Source  Performance  Standards:   namely, activated sludge
followed by deep bed filtration, although other technologies or even
activated sludge alone may be  capable  of  meeting  the  standards.
Other technologies were briefly reviewed and may be applied, but the
identified  technologies  cited  above  clearly  represent  the most
practical and common approach  to  treatment  of  corn  wet  milling
wastes.

The  Court's  conclusion  that  deep bed filtration was being called
upon to remove an incremental 30 Ib of BOD and 40 lb of TSS per MSBu
(0.536 kg/kkg BOD and 0.71U kg/kkg TSS) is not entirely correct.  An
analysis of existing plants and their capabilities of  meeting  1977
and  1983 Effluent Guidelines cannot be extrapolated directly to new
corn wet mills.  New m.~tll in-plant controls will provide  raw  waste
load reduction and stabilization and thus will greatly contribute to
treatment   plant   performance.   This  report  logically  develops
attainable effluent levels from new corn wet  mills  employing  best
available  control  technology  and  compares  these levels with the
present New Source Performance standards.

D.  ORGANIZATION OF THIS REPORT

This report is divided into ten sections.  Section I summarizes  the
conclusions  of  the  report.   Recommendations  to EPA based on the
findings documented herein are presented in Section II.  Section III
summarizes the  decision  by  the  U.S.  Court  of  Appeals,  Eighth
Circuit,  and  reviews  the  sources  of  data  used to prepare this
document.  Section IV is  an  evaluation  of  the  effectiveness  of
biological  treatment  systems,  particularly  the  activated sludge
process, handling corn wet milling wastes and  other  high  c; .rength
organic  wastes.   In-plant  control  technologies applicable ;o new
corn wet mills are discussed in Section  V.   Section  VI  evaluates
deep  bed  or tertiary filtration, a technology for reducing BOD and
suspended solids levels in secondary  effluents.   A  discussion  of
Clinton  Corn  Processing  Company"s (Clinton, Iowa)  waste treatment
facilities is presented in Section VII.  Clinton employs much of the
end-of-pipe treatment technology  recommended  to  meet  New  Source
Performance  Standards,  including filtration, although the plant is
subject to operational limitations as discussed in the section*    In
Section  VIII,  predicted  performance  of  new plants is presented.

-------
Model plants with accompanying treatment  facilities  are  outlined.
Effluent  levels  are  predicted  in terms of pound of pollutant per
unit quantity of raw material.  Associated costs to meet New  Source
Performance Standards are also given.  Section IX evaluates the non-
water  quality  aspects  of  compliance  with New Source Performance
standards, including energy requirements, air pollution,  and  solid
waste  disposal.   Section  X  is  a summary of the findings of this
study.

E.  SOURCES OF DATA

The detailed information on which this report  is  based  was  drawn
from  a  variety  of sources including the literature, field inspec-
tions, special consultants and  expert  opinion,  corn  wet  milling
companies, equipment manufacturers, and the Environmental Protection
Agency.

An intensive review of the literature was conducted on deep bed fil-
tration  with  particular  emphasis  on  filtration  of biologically
treated effluents.  Pertinent findings from this review are cited in
Section VI of this report.  The  entire  Literature  Review  (1)   is
summarized  in the supporting documentation contained in a record on
remand which is available for review.

The use of the activated sludge process in municipal and  industrial
waste  treatment,  including  corn  wet  milling was investigated to
document its effectiveness in handling high strength organic  waste-
waters.   Of  particular  concern  was  the application of activated
sludge to industrial wastes similar to those generated by  corn  wet
milling.  These similar wastes include brewing, distilling, malting,
edible  oil refining, and wine production.  Data were drawn from the
literature and from EPA experience^

Dr. E. Robert Baumann,  Anson  Marston  Distinguished  Professor  of
Engineering,   Iowa   State  University,  was  retained  to  provide
additional  technological  insight  into  advanced  waste  treatment
concepts  and  applications.   Dr.  Baumann  is  one  of the leading
authorities on filtration of secondary  treatment  plant  effluents,
and  is  the author of numerous papers and textbooks in sanitary and
environmental engineering.

Dr. Raymond C. Loehr, Professor of Engineering, Cornell  University,
was  also consulted in the development of this report.  Dr. Loehr is
a recognized expert in the treatment and disposal of  high  strength
organic wastes.

In  addition, Dr. Charles M. Cook, Technical Advisor to the Director
of Monitoring  and  Data  Support  Division  of  EPA,  reviewed  and
evaluated the data and statistics developed in this study.

A specific request (2) was made to a number of corn wet milling com-
panies for data regarding:

     1.    Performance of existing treatment and pretreatment plants

-------
including related production and product-mix figures, and
treatment costs.
     2.
Previous pilot plant studies.
     3.    Relationship between raw aste characteristics (and treat-
ability) and product mix.

     U.    Projection of new plant costs, raw waste characteristics,
additional in-plant controls, and power requirements for
the model 30,000, 60,000 and 90,000 bushels/day corn wet
mills.

Responses  to some or all of the specific information requested were
made by six companies.  Two meetings were held in  Washington,  D.c.
with  representatives  of  Anheuser-Busch,  Inc.,  CPC International
Inc., and Penick 6 Ford, Ltd.; and one meeting was held with  A.  E.
Staley Manufacturing Company in Decatur, Illinois.

Several  field  inspections  were  conducted including two visits to
Clinton Corn Processing company in Clinton, Iowa and a visit to  the
CPC  International,  Inc.  plant  in Corpus Christi, Texas.  A brief
three-day sampling program was conducted at the Clinton waste treat-
ment plant in an attempt to assess the efficacy of  their  deep  bed
filtration  system.   visits  were  also  made  to  the Metropolitan
Sanitary District of Greater Chicago and to Du Page County, Illinois
to discuss their experience with deep bed filtration and to  observe
filter installations.

Much  information was provided by waste treatment equipment manufac-
turers, particularly regarding  filtration.   The  following  filter
manufacturers were contacted and provided data during the study:

           Can-Tex Industries
           Crane-Cochrane
           Dravo Corporation
           Ecodyne Corporation
           Envirotech
           General Filter Company
           Hardinge Division, Koppers, Inc.
           Hydro-Clear Corporation
           Hydromation
           Infilco-Degremont
           Neptune Microfloc, Inc.

Finally,  information  was  provided  by  EPA  on  several pertinent
projects.   Specifically,  detailed  data  were  supplied   on   the
treatment  of  other  high  strength  food  wastes,  such  as  those
described in the Development Documents for the  Miscellaneous  Foods
and Beverages, Fruits and

Vegetables, and Pulp and Paper Point source Categories.  Information
was  also  r«ceived  on  the  EPA  Demonstration Project on deep bed
filtration of treated oil refinery wastes.

-------
All of the information from the above sources was carefully reviewed
and considered in preparing this report.  Materials used directly in
the report are referenced herein, and all data are included  in  the
record which supplements the report.

-------

-------
Distillery wastes have been  shown  to  be  amenable  to  biological
treatment.   Smith   (12)  reported  that  90  percent  or better BOD
reduction can be attained with biological treatment  processes  such
as   trickling  filters  and  activated  sludge.   Raw  wastes  from
distilleries generally contain 500 to 1000 mg/1 BOD and  50  to  200
mg/1  or more TSS.  Effluent data for several distilleries are shown
below (8, 13) :
                 Range of
                 Effluent BOD
                 Concentrations
                    .mg/1
                Range of
                Effluent TSS
                Concentrat ions
                    fflSt/i.
20 -
12 -
20 -
10
3 -
40
40
50

15
25 -
12 -
20 -
20 -
10 -
40
70
50
UO
100
On the following page is shown  Table  119  from  the  Miscellaneous
Foods   and   Beverages  Draft  Development  Document.   This  table
summarizes the  performance  of  11  treatment  facilities  handling
wastes from grain distillers operating stillage recovery systems.

Biological  treatment  has  been  successfully applied to bakery and
winery wastewaters.  One bakery in the mid-south  provides  complete
treatment  of high strength wastes  (14).  Treatment includes equali-
zation,  dissolved  air  flotation,  trickling  (roughing)   filter,
activated  sludge,  and  stabilization  ponds.  High strength wastes
with a BOD concentration in excess of  2000  mg/1  are  consistently
reduced  to less than 10 mg/1.  Treatment performance levels in 1974
were:
                Average
                Influent
              Concentration
               Range of
               Effluent
             Concentrations
     BOD
2210
7-9
                 Percent
                 Removal
  99.6
     TSS
1020
6-15
98.5 - 99.4
High strength wastes from winery operations have been shown amenable
to biological treatment processes (15, 16, 17, 18).   Wineries  gen-
erate medium to high strength organic wastes with BOD levels of 1000
to 2000 mg/1.  BOD removals of 90 to 97 percent and suspended solids
removals  of  67  to  90 percent are documented in the Miscellaneous
Foods and Beverages Draft Development Document.
                            15

-------
                  TREATMENT fiYGTI'M SUMMARY
                       SUBCATEGGRY A22
Plant

85A01


85A02
                                              PercenL Removal
                                             BOD
85A05


85AOJ
8 5 AIT


85A18


85A22


85A27

85A29
             Activated Sludge,
             Bio Disc.

             Aorated Lagoon
             Stabilization Pond

             Aerated Lagoon,
             Stabilization Ponds

             Aerated Lagoon,'
             Stabilisation Ponds

             Activated Sludge

             Bio Disc.
             Activated Sludge,
             Contact Stabilization

             Aerated Lagoon,
             Stabilization Pond

             Aerated Lagoons

             Aerated Lagoons
            _ Trickling Filters,
            - ["Jbabiliaation Ponds
97-5*


37-0


93-3


93-3

91.9
73-8

82.8
              3^.3***
                                           97-3
  *' Activated S.Ludgo Portion
 *•» i^cTore Contact .'."taliiJ.iz.atiori Added
*** Ho Clarification after Aeration
                          16

-------
One winery in New York State employed an aerated  lagoon  system  to
treat  its wastes  (19) .  Average raw waste flow is 0.85 mgd  (3217 cu
m/day) .  Influent BOD ranges from 1000 to  4400  mg/1  and  averages
2300  mg/1.   Yearly  average  BOD reductions for a four-year period
were as follows:  1970 - 94 percent, 1971 -  95.2  percent,  1972  -
95.6 percent, 1973 - 94.7 percent.  Because of problems with ground-
water  intrusion  and seasonal fluctuations in effluent quality, the
winery installed an activated sludge treatment system in late  1973.
Data  from the first six months of operation  (including start-up and
shake-down procedures) indicate an average effluent BOD level of  30
mg/1  and  a suspended solids level of 75 mg/1.  Data obtained later
in 1974, after stabilization of the treatment process, indicate  the
following effluent characteristics  (20) :

                       Effluent Concentrations

                                         Aver age
                 BOD    18-32              21

                 TSS    20-54              40

Another  example  of  effective  treatment  of high strength organic
wastes exists in the frozen specialty product industry.  A  manufac-
turer  in  Virginia  has treatment facilities consisting of screens,
floatable fat and solids removal, dissolved air flotation, anaerobic
lagoons,  trickling  filters,  contact  aeration  tanks    (activated
sludge),  final  setting  tanks, and chlorination of final effluent.
Based on an analysis of 54 daily samples, BOD and  suspended  solids
are  reduced  by 99.7 and 99.3 percent, respectively.  The very high
influent raw waste levels  (3500 mg/1 BOD  and  4500  mg/1  TSS)  are
reduced  to final effluent levels of 15 mg/1 or less BOD and 35 mg/1
suspended solids.  Fat and oily material are reduced from 3000  mg/1
to  1.5  mg/1,  representing  an  overall removal efficiency of 99.9
percent.  Maximum daily BOD and TSS values are 27 mg/1 and 119 mg/1,
respectively  (21) .

To  further  evaluate  the  effectiveness  of  biological  treatment
processes,  data  from EPA experience with the Fruits and Vegetables
Industry were analyzed (22) .  This industry generates a very diverse
array of high strength organic wastes.  Within  the  industry  there
are  numerous biological treatment systems demonstrating the ability
to  successfully  treat  these  wastes  without  further   polishing
techniques such as filtration.  EPA accumulated data on 10 activated
sludge  systems, two trickling filter systems, and 13 aerated lagoon
systems  providing  treatment  of  fruit  and  vegetable  processing
wastes.   These  exemplary treatment facilities handle biodegradable
wastes from  32  product  commodities  processed  at  these  plants ,
ranging from corn, tomato products, and sauerkraut to jams, jellies,
and dry beans.

At  the  fruit  and  vegetable  plants  employing  activated  sludge
systems, raw waste BOD varied from 253 to 4096 mg/1.  The  activated
sludge  systems  resulted  in  long-term BOD reductions from 95.5 to

                               17

-------
99.7 percent.  Sustained effluent  (annual average) BOD levels ranged
from  8  to  62  mg/1  and  averaged  20  mg/1   for   all   plants.
corresponding  effluent suspended  solids levels ranged from 10 to 80
mg/1 and averaged 31 mg/1.  The new source performance standards for
the corn wet milling industry would require a sustained effluent BOD
and suspended solids load  reductions  of  97.4  and  96.9  percent,
respectively, with biological treatment and the addition of deep bed
filtration.   Under  representative waste use and waste flows in the
industry these load reductions would represent final effluent levels
of approximately 20 to 30 mg/1 BOD and 10 to  20  mg/1  TSS.   Under
present  experience  and  application  as  described  in Section VI,
Evaluation  of  Filter  Technology,  deep  bed  filtration  may   be
reasonably  expected  ro  provide  at  least  an additional 50 to 75
percent  removal  of  BOD  and  suspended  solids  beyond  effective
biological treatment.

The  13 aerated lagoon systems within the fruits and vegetables pro-
cessing industry handle raw wastes with BOD levels ranging from  388
to  5642  mg/1  and  averaging  2126  mg/1.  These treatment systems
demonstrated long-term BOD reductions of  90  to  99.8  percent,  an
average of 97.9 percent.

Variability  of  the  treated  effluents  from the treatment systems
described above was analyzed.  The ratio of maximum  daily  BOD  and
maximum  monthly  (30 consecutive days)  BOD values was determined to
be 2.0.  The same ratios for suspended solids were calculated to  be
1.8  for  aerated  lagoons  and  2.6  for  the  activated sludge and
trickling filter systems.

Analysis of the  data  gathered  for  the  Miscellaneous  Foods  and
Beverages and the Fruits and Vegetables Industries indicated that no
statistical correlation existed between influent BOD and TSS levels.
In  other words, influent suspended solids values varied erratically
in relation to BOD.  Despite this  variation,  effective  high-level
waste  treatment  was  clearly  demonstrated  to  be  available  and
practicable.   Statistical  analysis  of   fruits   and   vegetables
treatment  data  also  indicated that no relationship exists between
influent and effluent TSS levels.

A concern raised frequently by the corn wet milling industry is  the
apparent  sensitivity of the activated sludge system to fluctuations
in  the  influent  waste  characteristics.   As  shown  above,  irany
existing  activated sludge systems receiving variable, high strength
organic wastes have been  able  to  consistently  produce  a  highly
stable effluent with low BOD and suspended solids.

A.  w.  Busch  (7)  discusses operational problems at some length and
summarizes as follows:  "In short, most  operational  problems  with
fluidized  systems [activated sludge] are built into design (lack of
ability to control growth rate for example)  or  are  due  to  faulty
operating  concepts,"   Referring  to shock loads, he further states
that "determination of such flow patterns and their anticipation  in
process  and  system  design  is  a  vital  part  of  the engineer's
responsibility."
                              18

-------
The information presented above fully substantiates the  ability  of
biological treatment systems to reliably produce very high pollutant
reductions  on  a sustained basis with high strength food processing
wastes.  With reasonably prudent design and operation, a  biological
treatment  system can effect overall BOD and suspended solids reduc-
tions in excess of 95 percent and can produce effluent levels of  30
mg/1 and less, despite highly variable raw waste characteristics.

C.  CORN WET MILLING INDUSTRY

When  waste treatment has been reguired, the corn wet milling indus-
try has for the most part used the  complete  mix  activated  sludge
process.   There are four mills that treat process wastes and event-
ually discharge treated effluent directly to waterways (one  of  the
four  recycles  its  effluent into the mill for reuse) .  All four of
these mills employ complete mix activated sludge.  Five  mills  pro-
vide  pretreatment of process wastes prior to discharge to municipal
systems, and a sixth mill  is  currently  constructing  pretreatment
facilities.   Three  of  these  six  pretreatment  plants employ the
activated  sludge  process;  the  remainder  use  other   biological
processes.   A  review of the performance of these treatment facili-
ties is presented below in order to provide a  better  understanding
of treatment practices within the corn wet milling industry.

1-   CPC - Pekj.nf Illinois

CPC"s Pekin, Illinois corn wet mill is an older, medium-sized  plant
located  on  the  Illinois  River.   The mill uses a large volume of
water, 20.5 mgd  (77,593 cu m/day), for process and cooling purposes.
Much of this water is used in  once-through  barometric  condensers.
Prior to the late 1960's, all of the wastes from the Pekin mill were
discharged  directly  to  the river without treatment.  In 1968, CPC
applied for and received a Research and Development grant from EPA1s
predecessor, the Federal  Water  Pollution  Control  Administration.
The  grant  provided funds for development, design, and construction
of waste treatment facilities.  The treatment plant was designed  to
handle  only  concentrated  process wastes (less than 1 mgd) and not
the contaminated barometric discharge.  Operation of  the  treatment
facilities  began  in  late  1970.   Included  in  the  scheme  were
equalization, cooling, nutrient  addition,  complete  mix  activated
sludge  (aeration  and clarification with sludge recycle), dissolved
air flotation, and reaeration.   As  the  Final  Grant  Report   (23)
documents,  numerous  mechanical  and  other  problems  plagued  the
plant's first year of operation, and treatment performance has never
reached the original design criteria on a long-term basis.

The following discussion  concerns  the  performance  of  the  waste
treatment  facility  at  Pekin.   The  industry has claimed that the
Pekin treatment plant does not perform satisfactorily;  that  it  is
subject  to  shock loads and variations in raw waste and,  therefore,
is not capable of producing a stable effluent.   Carryover  of  sus-
pended  solids  in  the  effluent  is  a  particular  problem.  This
investigation confirmed the  variability  of  treatment  plant  per-
formance  at  Pekin,  but  it  also established the reasons for this
                            19

-------
variability:  Deficiencies in design and operation of the  treatment
system  and  deficiencies  with  in-plant  waste controls within the
manufacturing plant.

Initially the Pekin waste treatment plant was  operated  at  a  very
high  organic  loading  rate,  specifically  a food-to-microorganism
(F:M) ratio of about 0.8.  Experience has shown  that  treatment  of
corn  wet  milling  wastes  is  effectively  attained  at much lower
loading rates.  For example, the new treatment facilities at Clinton
and Muscatine, Iowa are designed for an F:M of 0.3 to  0.35.   Pekin
has  since  made  in-plant changes to reduce waste loadings, but the
treatment plant is still subject to influent shock loads that exceed
normal design loading rates.  Other design deficiencies  include  an
undersized  clarifier.   The  Pekin  clarifier  was  designed  at an
overflow rate of 600 gallons per day per square foot (24.4 cu m  per
day  per  sq  m).  Although this value conforms with standard sewage
treatment practice, it is considerably higher than loadings used  at
other  treatment  plants  handling  high  strength  organic  wastes.
Overflow rates of 400 gpd/sq ft (16.3 cu m/day/sq m) are more common
in the miscellaneous foods and beverages industry   (8) „   The  Pekin
Final  Grant Report (23) stated that M[a] lower [clarifier] overflow
rate should result in a lower sludge blanket, and less frequent floe
carryover.n  The report also made note of  severe  raw  waste  shock
loads  - shock loads that could be controlled within the plant.  The
conclusion was that "further improvement in  effluent  quality  will
require  stabilization  of  the  waste  load, and improved suspended
solids removal.11

CPC is well aware that the  Pekin  waste  treatment  facility  needs
improvement.   An  EPA  evaluation  of  the plant in 1972 concluded:
"With enlarged plant facilities it appears possible  to  reduce  BOD
and  suspended  solids concentrations below 25 mg/1.  Performance of
the existing waste treatment plant cannot be improved  significantly
by operational control modifications alone (24).H  CPC has performed
its  own  evaluations (25).  studies conducted from November 1972 to
February 1973 concluded that additional equalization (to absorb  the
extremely  high  waste  load  fluctuations) and additional activated
sludge facilities were  required  to  improve  the  plant  effluent.
CPC's  consultant  performed  studies  in 1973 on effluent polishing
techniques, including filtration and dissolved air  flotation.   The
consultant  concluded  that  new  dissolved air flotation facilities
were needed (26).  The consultant's conclusions  on  filtration  are
discussed  in  section  VT of this report.  Based on the findings of
CPC and its consultant,   extensive  improvements  to  the  treatment
plant   are  now  being  implemented.   These  improvements  include
additional equalization basins, new aeration tanks, new  clarifiers,
and  a  new dissolved air flotation unit (27).  Work is scheduled to
be completed by late 1975.

Despite its limitations (and the  present  in-plant  control  inade-
quacies  within  the  corn  wet mill), the Pekin treatment plant has
demonstrated its ability to perform well  for  extended  periods  of
time.   For  example,  between July 1972 and January 1973, BOD reduc-
tion was below 90 percent only 10  percent  of  the  time,  and  the
                               20

-------
effluent  design  level  of 40 mg/1 BOD was met nearly 60 percent of
the time  (23).  More recent data  substantiate  that  the  treatment
facility  can  produce  a  stable  effluent.   From July 24, 1973 to
August 21r 1973, BOD concentrations exceeded 100 mg/1 only once  and
generally  were below 50 mg/1.  From October 8, 1973, to December 3,
1973, the same conditions were met.  Particularly  good  performance
was  recorded  in May and June of 1974.  Average effluent BOD was 35
mg/1 in May and 32 mg/1 in June.  Average TSS was 65 mg/1 in May and
66 mg/1 in June (27).  Improved waste treatment and  additional  in-
plant  controls,  which  would  be much easier to implement in a new
manufacturing plant, would enable Pekin to produce a stable  treated
effluent  on a reliable and sustained basis.

It  should  again  be  noted that the waste treatment plant at Pekin
only receives the high strength waste waters constituting less  than
10  percent  of  the  total mill effluent.  The major portion of the
mill effluent consists  of  the  contaminated  barometric  condenser
cooling  water  discharge  (once-through cooling water) , which repre-
sents a dilute, but significant waste stream.  Total waste loads  to
the  receiving  stream  are about 600 Ib/day  (272.4 kg/day) from the
treatment plant and 6000 Ib/day (2724 kg/day)  from  the  barometric
cooling waters.

2.   CPC - Corpus Christie Texas

CPCfs Corpus Christi plant is a small mill  of  more  modern  design
than  the  mill  at,  Pekin,  Illinois.   in  1970,  the  plant began
installing waste treatment facilities to treat process wastes  prior
to discharge to the adjacent ship channel.  The Corpus Christi plant
suffers  the same failing as Pekin; namely, direct discharge of con-
taminated barometric condenser water without treatment.  The problem
is not as severe as at Pekin, however, since the plant does employ a
number of surface condensers.  Barometric discharge accounts for 3.4
mgd  (12,869 cu m/day) and 264 Ib/day  (119.8 kg/day)  or  16  Ib/MSBu
(0.286 kg/kkg) of BOD on an average basis  (28).

The corpus Christi treatment facility is a piecemeal design.  Begin-
ning  as an unsuccessful batch treatment operation, it was converted
to an activated sludge system over a period of four years.  The sys-
tem originally consisted of three vertical steel  tanks,  each  with
sparged air and a submerged turbine mixer, followed by dissolved air
flotation.    Other   components  were  gradually  added,  including
nutrient addition facilities, an equalization and cooling  basin,  a
primary  tilted plate separator for starch wastes, a fourth aeration
tank (of  larger  capacity  than  the  first  three),  and  a  final
clarifier.   The flotation unit was not effective and is now used as
a second clarifier.  There is  no  skimmer  on  the  clarifier,  and
solids  tend  to  float  to  the  top  and  pass over the weir.  The
aeration tanks are operated  at  a  very  high  MLSS  (mixed  liquor
suspended solids)  level, between 6000 and 8000 mg/1 (29, 30) .

The  limitations  of the Corpus Christi treatment facility result in
an effluent atypical of  a  well  designed  and  operated  activated
sludge  system.   During the period December 1974 to April 1975, the


                              21

-------
effluent averaged 311  mg/1  BOD  and  730  mg/1  suspended  solids.
Despite  this  poor  performance  and  the  discharge  of  untreated
barometric condenser wastes, the total BOD discharge from  the  mill
(including  cooling water) was 57.4 Ib/MSBu  (1.025 kg/JOtg) as a mean
daily average and 64.6 Ib/MSBu  (1.154  kg/kkg)  as  a  mean  30-day
average.   These  levels  are  not  much  above  the  1977  effluent
guideline of 50 Ib/MSBu  (0.893 kg/kkg) for  a  maximum  month.   BOD
discharged  from  the  waste treatment plant was 41.8 Ib/MSBu (0.747
kg/kkg) as a mean daily average and 48.3 Ib/MSBu (0.863 kg/kkg)  as a
mean 30-day average  (23) .

As at Pekin, a stable effluent (in terms  of  BOD)   is  obtained  at
Corpus christi during certain periods.  For example, from February 9
to  21,  1975, effluent BOD ranged from 9 to 90 mg/1 and averaged 43
mg/1.  From March 14 to 29, 1975, effluent BOD values ranged between
10 and 64 mg/1 and averaged 35 mg/1 (27) .

Because of alleged violation of state water  quality  standards  and
failure of the waste treatment facility to achieve acceptable efflu-
ent quality, CPC had extensive evaluations performed during 1974 and
early 1975 (29) .  Concurrently, CPC intensified its effort to reduce
waste  loads within the elaijt.  Additional facilities were installed
to improve operation and divert, remove, or pick up wastes  in-plant
and  to  reduce the potential for accidental discharges.  CPC's con-
sultant recommended extensive modifications to the  waste  treatment
plant,  and  work  has  begun  to  implement  the new measures.   The
modifications include additional seeded  equalization  and  a  final
stabilization lagoon.

         ican Maize - Hammond, Indiana
The American Maize corn wet mill is an older plant located  on  Lake
Michigan.   The  plant  discharges  once-through  cooling waters and
treated process wastes to the lake.   A  major  pollution  abatement
program  was  undertaken during the late 1960"s.  Surface condensers
replaced most of the barometric condensers.  Only  two  small  baro-
metric  condensers remain in the syrup refinery, one of which is not
presently being used.

American Maize's waste treatment system is a series of three lagoons
that was converted to essentially an activated  sludge  system  u'ith
polishing  ponds.  Aerators were installed in the first lagoon and a
clarifier with sludge return was added in 1968, process and sanitary
wastes were segregated in 1969, and aerators were installed  in  the
two polishing lagoons receiving activated sludge effluent in 1970.

The 1977 effluent guidelines would allow American Maize to discharge
a maximum of 3250 Ib/day (1475.5 kg/day) of BOD and suspended solids
for  any 30 consecutive days and 9750 Ib/day (4426.5 kg/day) for any
one day.  A look at American Maize NPDES reports (31) shows that the
plant is currently  meeting  these  levels.   Discharge  levels  for
January through March 1975 are shown below:
                                 22

-------
                    Average for Month
                  BOD              TSS
                               Maximum for Month
                            BOD              TSS*
              Ib/day  kg/day  Ib/day  kg/day  Ib/day  kg/day  Ib/day  kg/day
Jan

Feb

Mar
1977 Effluent
Guidelines     3250

NPDES Permit
Levels         2000
1722
1418
421
782
644
191
1995
2002
2080
906
909
944
3819
2193
1313
1734
996
596
5600
6756
5286
2542
3067
2400
1476    3250    1476    9750    4427    9750    4427
 908
Calculated   by   assuming   that   maximum
concentration occurred on same day.
10,000    4540     —

flow  and  maximsn  TSS
The current discharge levels for American tfeize are well under  both
the  NPDES  permit  levels  for  the  plant  and  the  1977 effluent
guideline levels.

The total discharge levels from American Maize  do  not  necessarily
indicate  that  all  components  of  the  waste treatment system are
operating well.  For example, during the period April  1974  through
March  1975, monthly average values of TSS in the clarifier overflow
ranged from 66 mg/1 to 1292 mg/1.  Averages for TSS during the first
three months of 1975 were 628, 651, and 277 mg/1, respectively (32).
The major portions of  the  American  Maize  treataent  system  were
originally designed for an entirely different process and, hence, do
not  reflect  current design standards for activated sludge systems.
The  polishing  lagoons,  however,  provide   additional   treatment
capacity.   Thus, although the total system does not include a well-
designed activated sludge process, American Maize is easily  meeting
1977  effluent  guidelines with current technology, despite frequent
upsets in their adapted activated sludge system.

^'   Clinton Corn Processing Company - Clinton, Iowa

Clinton Corn has recently installed new waste  treatment  facilities
that  include  deep  bed  filtration,  part  of the 1983 reccmnended
control technology.  Accordingly the  Clinton  treatment  system  is
discussed separately in Section VII.

5.   Pretreatment Facilities

There are five corn wet mills that presently provide pretreatment of
wastes prior to discharge  to  municipal  collection  and  treatment
systems.   These plants include Anheuser-Busch - Lafayette, Indiana;
Cargill - Cedar Rapids, Iowa; Penick & Ford - Cedar Rapids;  and  A.
E.  Staley  -  Decatur,  Illinois  and Morrisville, Pennsylvania.  A
sixth mill. Corn Sweeteners - Cedar  Rapids,  is  constructing  pre-
treatment  facilities.   The  two Staley plants and Corn Sweeteners1
plant  employ  the  corplete  mix  activated  sludge  process,   and
                            23

-------
Anheuser-Busch  operates  an aerated lagoon system.  Penick & Ford's
system is a unique fungal digestion process, with a  final  clarifi-
cation  step  soon  to  be  added.  Cargill's pretreatment is rather
limited, consisting of settling tanks and some aeration.

All five of the existing pretreatment facilities are operating  suc-
cessfully  from  the standpoint that requirements of the municipali-
ties involved are being met.  For example, Anheuser-Busch1s  aerated
lagoon system consistently removes 50 percent and more of the BOD in
the raw waste (33) .

D.  SUMMARY

Biological  treatment  processes, particularly activated sludge, are
proven and effective means  for  handling  domestic  and  industrial
wastewaters.   The activated sludge process has been demonstrated in
thousands  of  applications  as  a  versatile  pollution   abatement
technology and is well-suited for treatment of high strength organic
wastes.

Many  experts in the field of environmental and sanitary engineering
attest to the applicability of activated  sludge.   Furthermore,  an
extensive  evaluation  of biological treatment systems in industries
generating wastes similar to corn  wet  milling  proves  that  these
systems  can effectively treat variable high strength wastes and can
produce stable,  high  quality  effluents.   With  good  design  and
operation,  such  a  treatment  system  will consistently produce an
effluent containing 30 mg/1 BOD and TSS, despite  raw  waste  varia-
tions.  This has been demonstrated with a wide variety of industrial
wastes,  including those generated by breweries, edible oil refiner-
ies, malting plants, distilleries, bakeries, wineries, and fruit and
vegetable processing plants.

The activated sludge process has been applied by a  number  of  corn
wet  mills  in  the  treatment of their wastes.  No plant has imple-
mented a properly designed and operated treatment  facility  coupled
with  good  in-plant control.  Thus there is presently no example of
best practicable control technology within  the  industry.   Factors
such as limitations in design and operation of the treatment facili-
ties,  lack  of in-plant controls, and discharge of untreated wastes
have prevented treatment plants from attaining  optimal  performance
on  a long-range basis,  A treatment facility at a new corn wet mill
that incorporates good in-plant controls in its design will  not  be
subject to upset conditions and thus will achieve long-term perform-
ance comparable to the high levels achieved by other industries.


On  the  basis  of  experience  of treating similar waste from other
industries it is generally concluded with  reliance  that  with  the
most  careful  design,  operation,  and  in-process control, the New
Source Performance standards can be  achieved  through  demonstrated
performance  of  biological  treatment systems.  Deep bed filtration
provides an  additional  assurity  as  to  the  achievement  of  the
required high quality affluent on a reliable and sustained basis.
                          24

-------
                             SECTION V

                         IN-PLANT CONTROLS


As  the  Development  Document for the Grain Processing Industry (9)
indicated, there are  many  water  recycling  and  reuse  techniques
presently  employed  in  corn  wet  mills.   These  techniques  have
resulted from efforts to improve product recovery and simultaneously
to reduce raw waste loads.  In-plant controls are continuously being
developed and improved.

The Development Document also pointed out that not all existing corn
wet mills employ every available in-plant control to conserve  water
and  reduce  waste loads.  Age of the plant and physical constraints
such as space within the  plant  and  land  availability  are  often
determining  factors.   This is not the case with new plants.  A new
corn wet mill will be able to incorporate, in design, many  in-plant
controls  that  are  less  readily retrofitted into existing plants.
These controls would include segregation  of  process  and  sanitary
wastes,   recirculating   or   noncontact   cooling  systems,  spill
containment facilities or overflow tanks, holding tanks  to  collect
discharges  of  concentrated  wastes  or acids, clean-in-place (CIP)
cleaning systems, and instrumentation to monitor and control process
variables and resultant waste streams.

After consultation with the industry  (34) and with process equipment
suppliers, it has been determined  that  a  new  plant  designed  or
constructed  today  would  incorporate the majority of available in-
plant controls as a matter of good  practice  with  or  without  the
demands  of  pollution control regulations.  In fact, with some pro-
cess equipment such as evaporators, there would  be  little  choice.
Process  equipment  with  the  most  modern  water  conservation and
pollution control features would also be the most attractive from an
economic or product yield point of view.

The only exception that might be taken regards cooling  systems.    A
new  plant  built today that did not face stringent effluent limita-
tions might employ a once-through cooling water system  using  baro-
metric condensers.  The mill would thus discharge a high volume,  low
concentration,  contaminated  waste  stream.   Faced  with  effluent
limitations, the new plant designer has  two  choices.   One  is  to
install  barometric  condensers utilizing a recirculating water sys-
tem.  A cooling tower would be needed, and  the  blowdown  from  the
tower would require treatment along with process wastes.  The second
choice  is  to  install  surface  {noncontact)   condensers,  with or
without recirculation,  that  would  prevent  contamination  of  the
cooling water stream.  Surface condensers are more costly to buy and
operate than barometric condensers.

The lack of adequate in-plant controls, coupled with treatment plant
deficiencies at existing corn wet mills, contribute to the inability
to  identify  a truly exemplary treatment system meeting potentially
achievable performance in the industry.  These same deficiencies   in


                             25

-------
current  practice also make it very difficult to extrapolate current
treatment performance to new  sources.   .One  cannot  conclude  that
existing  treatment  systems are performing as well as possible.  It
would be more valid to  look  at  periods  of  good  performance  at
existing   treatment   plants   reflecting   the  absence  of  upset
conditions, since a waste treatment plant serving a new  mill  would
experience  less  raw waste variability and shock loads.  Better in-
plant controls will reduce raw waste fluctuations.

In summary, a high degree of  in-plant  control  can  be  reasonably
expected at any new corn wet mill.  Such control will be an integral
part  of  the  basic  process  design and, if supplemented with flow
equalization, will largely eliminate major raw  waste  fluctuations.
Elimination  of  raw  waste  fluctuations  will allow the subsequent
waste treatment processes to produce a stable,  high-quality  efflu-
ent.   In  considering  the  cost  of  the in-plant controls for new
sources, only the cost of  a  recirculating  or  noncontact  cooling
water system is directly attributable to pollution abatement.
                             26

-------
                      TABLE   9-1  (CONTINUED)
                      RESULTS  OF  STUDIES  OH
   FILTRATION Of  EFFLUENTS FROM SECONDARY BIOLOGICAL TREATMENT
                          i j A K A C T C R I S T f C .
'"••u i"-. '•• i;. :L* ^UJ«LE ML:DIA size DLPTM
type mm in
' : "•• • '•'•'- • ' •-'•:- '• • " '• '• • •- = ^ ••-'-• . A, t : vatcd i.-.nd 1-2 60 i,
'•';• r lo« \ _S • jJ ,;e
• ,•••-•••' '! J - -• >'•••• , i :,. -\c t i •„ a t ed Sand 0.5-2.5 - 1
• .;..,.-, •: i-'--j;= -• L j .. ,'-,.-! f 1 o^' S i ud^e
..-!». -l^l.MlJ
'• M , . :K..--, . M . ;. . i . .;: ..vi t v TncU i r.g Sand 1,1 - 1
D J v, n f i o w F i 1 c e r
OJ 'J'-iri .,o Hui-Na . :;r,iyi t v Trickling Sand 1,5-3 - 1
*~" ' " 	 :-- ;; , '•-• '•'• sJ "• -' '•' '- ;.'c wn f 1 o w Filter
•••• i'-., !.:.«;]. uid Si^ucc:' Tricklir.^ Sand 1-2 - . 4
?. a j i a 1 Flow Filter
•: -« -.t- 3 Ue , tnsl anJ Ir,.^sJium Activated Sand 1-3 63 5
Pressure Sludge 3
U p f 1 o w 5
5
': -.r.i-i L Je , fcn^Kmd Per^iutit Activated Sand 0.60-1.20 57 3
!lpf low Sludge 3
5
5
• . .> •:> <.- s i ,1 c , C n g 1 a /i d 3 i m a t e r Activated Sand 0.5-1 3
Ri dial Flow SI uJ^e 3
5
5
' , : ; .i i] \i ri j ._• r • [./ n ^ , I nn; e d i urn Tr i c V. 1 i n ^ Sand 1-2 60 4
1 U •« A u L : L
g p n f r :
.3

.3-;. 4


-3

. 6-3. 2
-6

. 3
. 3
. 0
. o
. 3
. 3
.0
.0
. 3
. 3
.0
. 0
.5-5.0
IN1
.„.'!
1 ?

i 2


20

21
22

9
46
a
37
9 '
32
1 1
22
1 1
51
i ^
24
30
	 . 	 ,-~^-— —
OL'T
.,»/!
-

C


5

5
9

2
B
6 .
10
1
7
4
5
3
7
4
10
8
kEMOVAL LENGTH
percent h f
60

Sg


75

75
60

74
84
20
74
86
73
60
ft3 '
74
86
62'
08
30
REFERENCE

( 2 - )

(22)


(22)

(22)
(22)

(23)



(23)



(23)



(24)
i, f lo^V
           ilter

-------
treatment to  meet  increasingly   stringent  effluent  requirements.
Filtration applications in wastewater treatment include:

1.   removal of biological floe from secondary effluent,
2,   removal of precipitates in the process of phosphorus removal,
     and,
3.   removal of solids remaining after chemical treatment (36) .

Filtration  as  a tertiary treatment method following secondary pro-
cesses has long been practiced in  Great Britain.  While the  British
practice  tends  toward  the  upflow pressure mode, their experience
includes gravity downflow, bi-flow, and radial-flow methods, many of
which are patented processes.  In  the United states the tendency  is
toward  the  conventional  down flow  gravity  method,  although some
pressure filters are presently in  use  (35, 40) .
    EPA Process Design Manual  for  suspended  solids  Removal   (35)
devotes  considerable  attention  to granular media filtration.  The
manual states that  '•this  process  [ filtration J,  long  applied  in
treatment  of  municipal  and industrial water supplies, is becoming
widely used for wastewater  treatment  both  in  upgrading  existing
conventional  plants  and  in  designs  of  new  advanced  treatment
facilities."  Data  on  numerous   filter   installations   treating
biological  or  secondary effluents are given in the manual in Table
9-1, which is reproduced in  this  report  on  the  following  three
pages.   Note  that  in  all  of  these  examples,  filter  effluent
suspended solids never exceeded 10  mg/1,  despite  influent  levels
which varied as high as 50 mg/1.  Removals were above 60 percent and
often  were  80  percent  and  higher.  In developing x:he New source
Performance Standard for the corn wet milling industry,  deliverance
ot  a biologically treated effluent to the deep bed filtration stage
at a concentration  lex'el  of  75  mg/1  BOD  and  TSS  is  assumed,
eventhough  experience  with biological treatment for similar wastes
for other food processing industries indicate achievement of  a  BOD
and  5E*SS  effluent  level  of  30  mg/1  is  generally achievable by
biological treatment.  An additional 50 to 75 percent removal of BOD
and TSS is attributable to deep  bed  filtration  beyond  biological
treatment,

Many  tests  have been conducted on both pilot plant and plant scale
levels to determine the effectiveness of the filtration process  for
suspended solids and BOD removal, to further refine the process, and
to  develop  design  criteria.  Gravity downflow filtration has been
found to be a cost-effective means of reducing suspended  solids  in
the  effluent  from  wastewater  treatment facilities.  Both conven-
tional and special design filter beds have been tested.   Filtration
of  wastewaters  has been successfully demonstrated without chemical
addition.  In some instances, chemical  addition  has  improved  -the
filterability  of  biologically  treated  wastewaters  and  thus has
increased BOD and suspended solids removals.

Hsiung and Cleasby (41) performed studies to develop  a  simple  and
rational method for design and operation of water treatment filters.
Optimum  cost  for filtration was obtained by developing performance

                               32

-------
curves with the flow rate determined by consideration of either fil-
trate quality or head loss.  They concluded that the results of this
study may be applicable to sewage solids and that such applicability
should be investigated.

Lynam et al (42), in studies conducted at Chicago Metropolitan Sani-
tary District's Hanover plant, found rapid sand filtration of  acti-
vated   sludge   effluent   to  be  a  cost-effective  process  that
consistently produced effluent suspended solids of less than 5 mg/1.
Removals of 76.5 percent were  obtained.   The  study  reports  that
chemical  addition  (alum followed by polymer) provided insufficient
improvement of filtrate quality over plain filtration to warrant its
inclusion in future designs.

Tchobanoglous  (43) evaluated the performance of various filter  con-
figurations  including  conventional and special bed designs and the
effects of chemical addition.  His conclusions were:  1)  filtration
efficiency  without  chemical  addition  is a function of filter bed
grain size, 2) in most dual-media filters as presently designed  the
sand   underbed  contributes  little  to  overall  suspended  solids
removal, and 3) polyelectrolytes can be used to aid  in  removal  of
suspended  solids.   At  a  filtration  rate  of 5,15 gpm/sq ft (3.5
liters/sec/sq m)  and  with  polyelectrolyte  addition,  an  average
influent  suspended solids concentration of 23.5 mg/1 was reduced to
1 to 3 mg/1.  Filter operating periods  (run lengths) were between  4
and  5  hours.   Suspended  solids  removals  ranged  from  87 to 96
percent.

Tchobanoglous and Eliassen  (44) investigated filtration of activated
sludge  effluent  in  a  pilot  plant  study.   They   developed   a
generalized  rate  equation  based on size of filter medium, rate of
filtration, influent characteristics, and  the  amount  of  material
removed  within the filter.  Using 0.488 mm diameter sand, suspended
solids were reduced from 6.3 mg/1 to 2 mg/1, a 68  percent  removal.
Filtration  rate  was  5.8  gpm/sq  ft  (3.9 liters/sec/sq m) and run
length was 6.25 hours.  The top  1-inch   (2.54  cm)  of  the  filter
removed  75  percent  of  the  suspended  solids, and no solids were
retained below a depth of 6 inches (15.2 cm).

Baumann and Huang (45) conducted a pilot plant study using  effluent
from the Ames, Iowa standard rate trickling filter,  objectives were
to  determine  the  feasibility  of  using granular filtration as an
effluent polishing step and to develop a method and pilot plant test
procedure  to  be  used  in  the  design  of  filters  for  tertiary
wastewater  treatment.   They  concluded  that  use  of a dual-media
filter (anthracite over sand) increases filter capacity and provides
for better utilization of the filter depth and that it is  desirable
to have the anthracite as coarse as possible and the sand as fine as
possible.   Best results were obtained by a bed of coarse anthracite
overlying fine  sand,  which  produced  a  better  quality  effluent
without significant head loss development as compared with a coarser
sand topped by the same size anthracite.
                              33

-------
Culp  and  Hansen  (46) found that up to 98 percent of the suspended
solids found in an extended aeration treatment plant  effluent  with
21-hour  aeration of domestic sewage conoid be removed by filtration.
Turbidities as low as 0.3 jtu  were  obtained  without  the  use  of
coagulants.

Culp  and  Culp   (47)  report that filterability of sewage solids is
affected by the degree of flocculation  attained  in  the  secondary
process  and  that  activated  sludge achieves the best flocculation
results  as  compared  to  trickling  filter  and  physical-chemical
processes.   Flocculation  is  proportional  to  aeration  time  and
inversely proportional to FrM ratio.  Aeration basin MLSS variations
in the normal operating ranges of 1,500 to 5,000 sag/1 do not  affect
filterability.  With domestic wastes, suspended solids removals from
70 to 98 percent can be obtained at aeration times of 6 to 10 hours,
respectively.   Biological  processes,  in  general,  produce a more
fully developed floe  than  chemical  coagulation,  and  the  higher
removal percentages were obtained with the effluent from an extended
deration  plant.   Culp  and  Culp  present  the following levels as
guides to the suspended solids concentrations that might be achieved
when filtering a typical  secondary  effluent  without  addition  of
chewical coagulants:

                 process                 Effluent TSS

     High-rate tricltling filter          10-20 mg/1

     Two-stage trickling filter          6-15 mg/1

     Contact stabilization               6-15 mg/1

     Conventional activated sludge       3-10 mg/1

     Activated sludge with load factor
       (FrM) less than 0.15              1-5 mg/1

Culp  and  Culp  indicate  that,  although  mixed-media  filters can
tolerate higher suspended  solids  loadings  than  other  filtration
processes,  there still is an upper limit at which economically long
runs can be maintained.  The authors state that:

With activated sludge effluent suspended solids loadings  of  up  to
120  mg/1, filter runs of 15-24 hr at 5 gpm/sq ft [3.4 liters/sec/sq
m] have been maintained when operating to a terminal head loss of 15
ft [4.6 «} c>f water.  Suspended solids concentrations of 300 rag/1 or
more will lead to uneconomically short filter runs, even when  using
a  mixed-media  filter...Should  the secondary plant involved have a
history of frequent, severe upsets resulting in  secondary  effluent
suspended  solids  concentrations  of  200-500 mg/1, an intermediate
settling tank between the secondary clarifier and  the  filter  with
provision  for  chemical  coagulation during upset periods should be
made (47).
                             34

-------
One advanced wastewater treatment  plant  that  includes  dual-media
filtration  has  been in operation since 1968 (48).  The effluent is
of such high quality that turbidity is monitored  rather  than  sus-
pended solids.  The effluent from this plant is used to recharge the
water  table in Nassau County, New York.  The process as reported by
Vecchioli et al  provides  for  the  removal  of  suspended  solids,
phosphates,  dissolved  organics,  and  MBAS  (i.e.,  detergents  or
surfactants).  Treatment after activated  sludge  includes  clarifi-
cation   (where  alum is added), dual-media filtration, and activated
carbon adsorption.  Suspended solids  removals  of  99  percent  are
obtained.

Most  filtration rates reported in the literature vary from a low of
1 to a high of about 10 gpm/sq ft  (0.7 to 6.8 liters/sec/sq m).   In
optimization  studies, the rate has been established at about 6 or 7
gpm/sq ft  (U.I to 4.7 liters/sec/sq m).  Studies of ultra-high  rate
filtration  of activated sludge effluent in Cleveland (49) concluded
that no significant relationship existed  between  filtration  rates
and  effluent BOD, COD, and TSS concentrations in the range of rates
(8 to 32 gpm/sq ft or 5.4 to  21.7  liters/sec/sq  m)  used  in  the
studies.    The   investigation  also  revealed  that  for  influent
concentrations of less than 30 mg/lr the filter  effluent  generally
remained  in  the  range  of 1 to 12 mg/1 with or without polymer or
coagulant and polymer  addition,  but  for  influent  concentrations
above  60  mg/1, filtration with coagulant and polymer addition pro-
duced a higher quality effluent.

During the ultra-high rate filtration study in  Cleveland  (49),  an
upset  condition  caused by the breakdown in the sludge incineration
process forced the recycling of digested sludge through  the  plant.
The  filter  effluent  rapidly deteriorated from 2 to 10.2 mg/1 sus-
pended solids.  To correct the problem, chemical  addition  of  alum
and polymer restored the effluent quality to an acceptable 5.1 mg/1.
Filter  efficiencies  were  88, 54, and 93 percent before the upset,
during the upset, and after chemical addition, respectively.

In evaluating granular media filtration as a  method  for  upgrading
waste  treatment  facilities,  Middlebrooks  et al (50)  state:  "The
simple design and operation of this process makes it  applicable  to
wastewater  streams  containing  up  to  200  mg/1 suspended solids.
Filtration rates can range from 25 to 50 gpm per sq ft [17.0 to 39.9
liters/sec/sq m] for coarse solids to 2 to 5 gpm per sq ft  [1.4  to
3.4  liters/sec/sq m] for colloidal suspensions.  The versatility of
filter bed designs (media sizes and depths)  is such that nearly  any
effluent quality can be achieved."

Tertiary  filtration  has been applied successfully to many types of
industrial wastes.  Industries that are either employing  filtration
or  have  successfully  tested  and  are installing filters include:
steel manufacturing, petroleum refining, brewing, corn wet  milling,
wine processing, and food processing  (51, 52, 53, 54, 55, 56).
                                                  /
Multi-media  filtration has had particular application in the petro-
leum refining industry, where at least three refineries are known to
                           35

-------
successfully treat refinery wastewaters after biological  treatment.
These  applications represent some of the longest-term use of modern
filtration facilities employed with biological treatment systems for
treatment of strong industrial waste (BOD up to 800 mg/1 and TSS  up
to 300 mg/1) (57) .

Ample  supportive  data is available on the long-term performance of
these systems.  These plants in  which  multi-media  filtration  has
been applied include Amoco Oil Company, Yorktown, Virginia; Marathon
Oil  Company, Robinson, Illinois; and Southwestern Oil and Refinery,
corpus Christi, Texas.  At the Amoco Oil Company, Yorktown, Virginia
plant, the filtration system follows an  aerated  lagoon  biological
system.   The  plant  served as an EPA demonstration project.  Quar-
terly reports submitted to EPA  establish  the  underdesign  of  the
aerated  lagoon  and substantiate poor performance of the biological
system.  Upsets in performance of the biological system in turn gave
rise  to  operational  difficulties  with  the  subsequent   filter,
resulting  in  shorter  filter runs and performance reduction.  Even
under these limitations, the filter has been effective  in  reducing
influent  TSS  of 25 mc;/l to 10 mg/1 in the final effluent on a sus-
tained basis, a pollutant reduction of 60 percent.

The successful application of filtration is also  quite  well  docu-
mented  at  the Marathon Oil Company, Robinson, Illinois plant.  The
filter handles effluent from an activated sludge  biological  treat-
ment  system  and is designed for a maximum influent TSS of 40 mg/1.
In recent months (after installation of the  filter)  effluent  from
the  biological  system has averaged 49 mg/1 TSS, and the filter has
been successful in reducing the waste load to a  TSS  of  11.2  mg/1
over  the  last  18-month period - a total TSS reduction through the
filter of approximately 77 percent.  It is important  to  note  that
this  filter  performance resulted with an average overload in rela-
tion to filter  design  loading  of  25  percent.   Also,  prior  to
installation of the filter, average effluent TSS from the biological
system  was  19  mg/1  as  compared  to  49  mg/1  after  the filter
installation - seemingly indicating less emphasis on  proper  opera-
tion  and maintenance of the biological system after installation of
the filter and increased reliance on the filter for production of  a
high-quality   effluent.   Similar  results  were  attained  at  the
Southwestern Oil and Refining Company,  Corpus  Christif  Texas,  as
shown by 4 months of data  (22).

A  fourth  refinery,  Clark  Oil at Hartford, Illinois, has recently
installed treatment facilities  that  include  an  activated  sludge
system  followed by dual-media filters.  The plant was started up in
early 1975 and presently meets the Illinois  EPA  effluent  require-
ments  of  20  mg/1  BOD  and  25  mg/1  TSS, without the dual-media
filters.  The filters have yet  to  be  put  on-line,  but  will  be
shortly in anticipation of more stringent effluent requirements.

Average  BOD  levels  in  the  biological effluent were 14.1 mg/1 in
April 1975 and 7.7 mg/1 in May 1975 (58).
                             36

-------
Another example of successful application  of  biological  treatment
and  multi-media  filtration  to high strength organic wastes can be
found at Welch Foods' Brocton, New York grape processing plant.  The
treatment  system  consists  of   activated   sludge   followed   by
filtration.  The seasonal average BOD concentration in the raw waste
is  4096  mg/1.   A  summary  of the filter performance is presented
below:

                      Activated Sludge          Multi-media Filtration
                         Effluent                       Effluent
Season Average        20.6        28.1             8.1           8. a

Maximum 30-Day        34.0        78.1             13.0          9-2

Maximum Day          114.0       216.0             32.0         20.0

The data indicate the following filter performance levels:

                             Percentage               Percentage
                                                      TSS Removal
Seasonal Average               61                        70

Maximum 30-Day                 61                        88

Maximum Day                    72                        91

The data also indicate that BOD  and  TSS  levels  in  the  filtered
effluent  were 13.0 mg/1 and 9.2 mg/1, respectively, for the maximum
month  (30 consecutive days).  Maximum day  BOD  was  2.5  times  the
maximum  month  value, and maximum day TSS was 2.2 times the maximum
month value.  On a seasonal average basis, BOD  reduction  was  99.5
percent through the activated sludge system and 99.8 percent through
the activated sludge and filter combined  (22,59).

The  table  on  the  following  page  was taken from the Development
Document for the Renderer Segment of the Meat Products and Rendering
Processing  Point  Source  Category  (60) .   It  is  a  summary   of
filtration  results for a variety of biologically treated effluents.
The Development Document concluded that "the rapid sand  filter  has
also  been  receiving more extensive application in municipal sewage
treatment for tertiary treatment; thus, its use in  tertiary  treat-
ment  of  secondary  treated  effluents  from  any  type  of meat or
rendering processing plants appears to  be  a  practical  method  of
reducing  BOD  and  suspended  solids to levels below those expected
from conventional secondary treatment."

A positive feature of filtration is the ability to handle reasonable
fluctuations  in  solids  loadings  without  serious  impairment  of
effluent  quality.   These fluctuations can be absorbed by increased
backwash frequency.  This point is discussed by E.  R.   Baumann  in
D§§i2G Qt filters for Advanced Wastewater Treatment (61) :

                              37

-------
                                     Toble 13A.
                                     ration of
               Ce
  Tvce
                             Gravi ';;,'
ilter  Infl
TOD	
                 1 5-

                                                                           -i>" r • TI.'
W
OQ
Activated Sludge

Extended
Aeration plus
setllipy

Trickling
Filter

Activated. Sludge
with Clarifier

Contact
Stabilizdtion
 (raw waste
  includes
  cannery)

Miscellaneous
       Trickling
       Filter with
       Nitrification
pressure,
f!u! ti -media
Gravity,
Sand

wulti-media
sand
 (slow and
  rapid)
sana
                 11-50

                  7-36



                 15-130
                                            10-50
                                                         30-2180
                                                          3-75
                               18
                              (AVE)
                              15-75
                                                                     3-8
                                          1-4
                                          2-74
                                                                     2-4
                             2-6
                             9-20
                                                                                    1
                                           -20
                                          ^-27
                                          2.4
                                          AVE)

                                          '?-R
                                                        _I n
                                                          *

-------
The  process  of  filtration is one which is unusual in that it will
never, if properly designed, provide a grossly inferior  quality  of
effluent.   In general, if a filter can provide the desired effluent
quality under normal conditions, upsets  in  pretreatment  processes
wil1  provide  shorter  filter  runs  and  not  signif icantly poorer
effluent quality.  Thus, if under normal conditions the effluent  SS
are  running at 18 mg/1 and suddenly increase to a level of 30 or 40
mg/1, the principal effect will be a  significant  decrease  in  run
length but a relatively lesser increase in effluent SS.

C.   EXPERT  OPINION  ON  APPLICABILITY  OF  FILTRATION  TO CORN WET
     MILLING WASTES

CPC International, inc. has submitted a consultants report  on  the
application   of  deep  bed  filtration  as  an  effluent  polishing
technique at the Pekin, Illinois corn wet  mill.   The  report  (26)
presents  the  results  of  pilot studies conducted in 1973.  It was
concluded that in-depth filtration was technically  not  a  reliable
method  for  meeting the required Illinois EPA effluent standards of
25 mg/1 TSS and 20 mg/1 BOD.  The Eighth Circuit  Court  cited  this
report in their decision (footnote at 39) .

The  filtration  report and test data were reviewed by Dr. E. Robert
Baumann of Iowa State University, a recognized expert on  filtration
of  water  and  wastewater.   The  complete  text  of  Dr. Baumann's
analysis, including comments  on  the  ability  of  the  recommended
technology to meet the New Source Performance Standards, is appended
in this report  (62).

Dr.  Baumann  concluded that the filter test results in the consult-
ant's report to CPC do not lead to the conclusion that filtration is
an unacceptable technology for meeting the  New  Source  Performance
Standards.   In  reviewing  the  test  data, he noted that "it would
appear that with only a few exceptions the filter  runs  were  being
made  during  periods while the activated sludge process was upset."
In other words, the effluent being tested was not characteristic  of
effluent  from a well-run treatment system.  Dr. Baumann stated that
"it must be concluded that on the basis of the few runs relating  to
near  normal  activated  sludge  process  performance  in the Weston
report that dual  media  and  multi  media  filters  both  performed
adequately  with  respect  to  effluent  polishing."   Dr. Baumann1s
observations included the following points:

1.   "The failure of the coarse dual media filter was obviously  due
to  improper  media  selection...The  performance of the multi media
filter is not at a level that would be expected with proper design."

2.   Anionic polymers apparently were effective, "but their use  was
not  followed  with  an experimental design that employed a rational
basis for selection of dosages."

3.   contact time for formation of a filterable floe was  apparently
not provided.
                            39

-------
**•'   nh redesigned multi media filter using the top two media  about
like  that  in the fine dual media filter and a smaller sized garnet
compatible with them should provide  excellent  operation  at  rates
even as hic[h as 3 to 4 qpm/sq ft provided that adequate contact time
is provided with the arionic polymers."

Dr.  Baumann  also  concluded that the ability of deep bed filter to
function as a back-up system during a biological treatment upset was
clearly demonstrated by the  test  data.   During  the  filter  runs
involving very high influent suspended solids  (415, 900, 1490 mg/1) ,
both  the dual-media and multi-media filters produced removals o£ 94
to 99 percent.

Dr. Baumann summarized his findings on the CPC filtration report  as
follows:

In  general,  the  Weston report demonstrates the fact that deep bed
filters can consistently provide from 50 to 70  percent  removal  of
suspended  solids  from  a  wastewater  that has been given adequate
pretreatment.  The report also demonstrates the fact that  deep  bed
filters will, in general, provide an adequate back-up system for the
trapping  of  most  of  the solids that may be carried over from the
activated  sludge  process  when  upsets  involving  bulking  sludge
occurs.

Dr.  Baumann also evaluated the ability of the recommended treatment
technology to meet the 1983 effluent guidelines.  He concluded  that
the guidelines can be met provided that:

1.   The activated sludge system is designed for a low loading  rate
or F:M ratio.

2.   Adequate controls, for sludge bulking are  provided  in  design,
i.e.,  polymer  feed  capability, hydrogen peroxide addition, deeper
clarifier design.

3.   The deep bed filter media is properly sized  and  matched,  the
wastewater   is  chemically  pretreated  prior  to  fi3tration,  and
adequate contact time is provided for floe removal.

In assessing the 1983 and New Source effluent  levels.  Dr.  Baumann
stated  that  «i  would  expect  such a plant [properly designed and
operated activated sludge-granular media  filtration  with  chemical
pretreatment]  to  be  able to meet the BAT effluent standards for a
medium size corn wet mill".

D.  SUMMARY

Deep bed filtration is a demonstrated and available technology.   It
is  universally  applied  in  the  treatment  of  water supplies for
municipalities and industries and can produce  effluents  containing
essentially  no  suspended solids in these applications.  Filtration
has been successfully employed in the  treatment  of  municipal  and
industrial   secondary   (biologically   treated)   effluents.   The


                             40

-------
literature and other data clearly document this fact.   The  process
has  been  sufficiently  demonstrated  to  enable the development of
meaningful  design  criteria  and  has  been  employed  in  numerous
applications.  In municipal applications, a well-designed filter can
be expected to reduce suspended solids by 80 percent and more and to
produce  an  effluent  containing  less  than  10 mg/1 TSS.  Similar
results can be anticipated with industrial wastes if  the  preceding
biological  treatment  system is well designed and operated.  It has
also been demonstrated that  deep  bed  filtration  performs  as  an
effective  back-up  system  during  periods  of biological treatment
upset and thus helps to  insure  that  a  stable  effluent  will  be
produced.

Filtration  is  applicable  to corn wet milling wastes.  The process
has been demonstrated  with  other  high  strength  organic  wastes.
Expert  opinion  supports this applicability, as well as the ability
of  a  well  designed  treatment  system  to  meet  the  New  Source
Performance  Standards.   Furthermore,  as  will be shown in Section
VII, deep bed filtration has been successfully applied in  the  full
scale treatment of wastewaters at an existing corn wet mill.
                             41

-------

-------
                            SECTION VII

                    CLINTON CORN WASTE TREATMENT


The Clinton Corn Processing Company's corn wet mill is of particular
concern  since  it  has  installed much of the end-of-pipe treatment
technology  recommended  to  meet   the   New   Source   Performance
Standards ,;   namely,   activated   sludge   followed  by  deep  bed
filtration.  It is important to evaluate efforts at Clinton  in  the
proper  perspective.  Although some general conclusions can be made,
there are many conditions at Clinton that do not apply  directly  to
projected  conditions  at  new  plants.   The  following  discussion
summarizes Clinton* s  pollution  abatement  efforts,  evaluates  the
current  performance  of Clinton's new waste treatment facility, and
finally specifically evaluates the performance of Clinton's deep bed
filters.

A.  BACKGROUND AND POLLUTION ABATEMENT EFFORTS

The Clinton mill is a large, old plant located  on  the  Mississippi
River  in Clinton, Iowa.  The manufacturing plant includes corn pro-
cessing facilities and a distilled spirits operation.  The plant has
undergone almost continuous improvement and expansion.

Prior to 1973, all wastes from the  Clinton  plant  were  discharged
directly  to  the  Mississippi  River.   A major pollution abatement
program was undertaken to reduce waste loads to the river.   Several
barometric  condensers within the plant have been replaced with sur-
face condensers, thereby reducing pollutant loads.  Clinton plans to
make further  condenser  replacements  and  improvements.   Sanitary
wastes  have  been  segregated from process wastes and routed to the
municipal sewage treatment plant.  Steps have been taken  and  still
are  being  taken  to  reduce  raw waste loads and spills within the
plant.

A major portion of the pollution abatement program was the construc-
tion of new waste treatment facilities.  Land is limited at the corn
mill site, and the treatment plant had to be located almost  a  mile
away.  The treatment facilities include cooling towers, a biological
packed  tower or trickling filter with synthetic media, complete mix
activated sludge (aeration  and  clarification),  chlorination,  and
dual-media  filtration.   screening  and limited equalization of the
raw waste are provided at the manufacturing plant.   The  biological
tower  was installed in September 1973, the filters were placed into
service in November 1974, and the full design waste load  was  being
treated as of April 1, 1975-

The  treatment  plant was designed to handle 3 mgd (11,355 cu m/day)
of process wastes.  The treated effluent is returned to the mill for
cooling uses.  Ultimately, spent barometric condenser water is  dis-
charged directly to the river from the mill.
                               43

-------
Clinton  is  still  implementing  in-plant  controls to reduce waste
loads.  Clinton believes that their present program  and  new  waste
treatment  facility will allow the plant to meet by January 1976 the
standards of the NPDES permit issued April 17, 1975 (63).  The NPDES
permit levels are identical to the 1977 Effluent Guidelines.

B.  EVALUATION OF TREATMENT PLANT PERFORMANCE

In evaluating data from the Clinton waste treatment facility,  there
are  a number of factors that must be considered.  At issue here are
New Source Performance Standards and the ability of a new  corn  wet
mill  with  proper  in-plant  controls, cooling water practices, and
waste treatment  to  meet  the  New  Source  Performance  Standards.
Although Clinton's treatment facilities include activated sludge and
deep   bed   filtration,  the  following  points  must  be  strongly
considered:

1.   Clinton  is  an  older  mill  lacking  many  in-plant   control
measures.   Clinton  still  experiences  in-plant accidental spills,
such as acid spills  (June  2)   and  sugar  spills  (June  8),  that
materially  affect  raw  waste  load variability and treatment plant
performance (63) .

2.   Although changes are planned, Clinton continues  to  use  large
volumes  of  once-through barometric condenser water that contribute
greatly to total waste load discharge to the river.

3.   The Clinton waste treatment plant is receiving higher suspended
solids loadings in the raw waste than were  anticipated  in  design.
Presently,  only screening of the raw waste prior to equalization is
provided.   Additional  pretreatment  facilities  are   planned   to
alleviate the problem of excessive raw waste suspended solids.

4.   Limited raw waste equalization is provided at Clinton.   Deten-
tion  time  provided in the equalization basin is only 5 to 6 hours.
The Development Document for the Grain Mill Industry (9)   recommends
12  to  18  hours, and several mills have provided up to 24 hours of
equalization.

5.   The  Clinton  waste  treatment  facility  is  new   and   still
undergoing  shake-down  operations.   Operating  parameters  such as
amount of sludge recycled, dissolved oxygen levels in aeration,  and
amount  of  recycle  through the biological tower (trickling filter)
are still being varied to determine optimum operating conditions.

6.   Aeration basin dissolved oxygen levels were found to be  insuf-
ficient  in  the  original  design.   On  June  14,  1975 additional
aeration equipment was installed  in  the  basins.   Improvement  in
plant    performance,    particularly    regarding    mixed   liquor
settleability, has been observed since that time.

7.   The existing data record  does  not  necessarily  indicate  the
performance  of  the deep bed filters.  A malfunctioning check valve
has permitted flow to bypass the  filters  at  various  times.   The


                             44

-------
limited  historical data on treated effluent, therefore, include the
contribution of unfiltered wastewater during undeterminable  periods
of time.

8.   Total discharge to the river from  the  Clinton  mill  includes
water  treatment sludge.  This results from treatment of river water
prior to in-plant process uses.  Clinton has installed  a  clarifier
and  filter  system  for  river water treatment, and the sludge will
eventually be disposed of at a landfill.  This will  greatly  reduce
the  suspended  solids and a portion of the BOD presently discharged
to the river from the mill,

Despite the limitations discussed above, the Clinton waste treatment
facility  is  functioning  quite  well,  producing  a   high-quality
effluent  in terms of BOD and turbidity.  Clinton has supplied daily
sampling data for their waste treatment facility (64).  This data is
the most current and representative information available,  particu-
larly after April I, 1975, when the full design waste load was being
treated.   The data indicate that during the period November 1974 to
May 1975, average monthly BOD loadings  to  the  treatment  facility
ranged  from  1264  mg/1  to 1691 mg/1 in terms of concentration and
27,409 Ib/day to 36,668 Ib/day (12,443 to 16,647 kg/day) in terms of
load.  Monthly suspended solids levels ranged from 336 mg/1  to  600
mg/1 and 7286 Ib/day to 11,009 Ib/day (3308 to 4998 kg/day).
Treatment  plant
summarized below:
effluent  levels for November 1974 to May 1975 are
                        BOD
              Clinton Waste Treatment Efffluent

                   Turbidity  	TSS*
Month
Nov 1974
Dec 1974
Jan 1975
Feb 1975
Mar 1975
Apr 1975
May 1975
mg/1
54
44
86
39
17
16
26
Ib/day
1091
844
1864
846
369
334
585
Ib/MSBu
10.9
8.4
18.6
8.5
3.7
3.3
5.9
jtu
-
8.3
18.4
20.6
5.2
6.8
4.9
mg/1
133
74
1U7
54
70
101
135
Ib/day
2440
1419
3087
1172
1518
2106
3040
Ib/MSBu
24.4
14.2
30.9
11.7
15.2
21.1
30.4
     *Not necessarily filter effluent.

Suspended solids levels in the effluent are higher than anticipated,
but it must be recalled that these levels do not reflect filter per-
formance, since part of the secondary effluent bypasses the filters.
                              45

-------
If Clinton Corn was to eliminate its direct  discharge  of  contami-
nated  cooling  water,  either  by  recirculation or installation of
surface condensers, the above treatment  data  show  that  the  1983
Effluent  Guidelines   (20 Ib/MSBu BOD and 10 Ib/MSBu TSS for maximum
month) are well within reach at an existing plant  with  present-day
technology, even without filtration of the total secondary effluent.
Since   the   1983  guidelines  are  identical  to  the  New  Source
Performance Standards,  Clinton,  an  existing  facility,  would  be
meeting standards for new sources.

C.  PERFORMANCE OF DEEP BED FILTERS

From an operational point of view, Clinton indicates that their deep
bed  filters  are  performing quite satisfactorily  (63),  There have
been no problems with clogging, excessive head  loss,  or  excessive
backwashing.   Backwasn  cycles  are  normally 12 to 16 hours apart.
Shorter filter runs do occur when clarifier overflow solids increase
and thus solids retained in the filter increase,  demonstrating  the
ability of the filters to handle variations in solids loading.

Filter  operation was monitored at Clinton between November 25, 1974
and February 16, 1975.  Reductions in COD,  BOD,  suspended  solids,
and  turbidity  were determined (64).  The data are presented on the
following page.  Suspended solids removals ranged  from  45  to  100
percent  and averaged 77 percent.  BOD reductions through the filter
alone ranged from 6 to 53 percent and averaged 29 percent.

Additional limited sampling was performed by EPA's contractor during
June 9-11, 1975.  A summary of the data collected is included in the
second table that  follows.   Results  of  this  sampling  indicated
suspended  solids  removals  of 50 to 68 percent attributable to the
filter.  BOD removals were 0 to 32  percent.   Abnormally  high  raw
waste suspended solids and an aerator failure in one of the aeration
basins  may  have  been  the cause of increased effluent BOD and TSS
values during the sampling period (65).

D.  SUMMARY

Data from the Clinton Corn waste treatment facility is  of  interest
since  deep  bed filtration is used following complete mix activated
sludge.  Despite the fact that the corn wet mill is  quite  old  and
both  the  mill  and the treatment plant suffer from several limita-
tions, the following conclusions can be made:

1.   The Clinton Corn wet mill will meet the  1977  Effluent  Guide-
lines in 1976.

2.   The complete mix activated sludge process, when applied to corn
wet milling wastes, can produce a stable, high quality effluent with
BOD concentrations of less than 30 mg/1.

3.   Deep bed filtration is transferable to  the  corn  wet  milling
industry and has been demonstrated as an effective treatment process
at an existing corn wet mill.


                             46

-------
                                   O
 "i  O
 rb
 on  >j

 O. '.)'
 CT  --J
13  ,-t-

 J
-n o
O>  -J~
cr in
    o
    ;*r

-H <
r.v tn

(/i  <
    fD
QJ  ("
o
o    h
O  CJ

^3
                                           co      r^j
                                           O      -P.
•^-j     en     01
                                      Co     ITJ
                                      Cn     c.n
                                                                                                cri      -.j     rn     (..*j
                                  •'O

                                   CO
                                   r\.>
                                           
-------
!
!	1
                                                 H
                                                 -C--
                                                 O
                                                                              '.T.'  y
                oo
                to
                                                 l-1
                                                 CO
                CQ
                o
                             ••_rt     \O
                             O      O
                                     ro
                                                                                .
                                                                              O  rr

-------
                            SECTION VIII
To  determine  achievable  effluent limitation levels and to develop
associated cost data, three model new corn wet milling  plants  were
synthesized.   Since the present trend within the industry is toward
production of sweeteners, the new mills  are  logically  assumed  to
convert  the  large  part of their grind into syrups.  To evaluate a
range of plant sizes  for  cost  evaluation  and  pollution  control
impact,  three  mills  of  30,000;  60,000;  and  90,000 bushels/day
capacity were developed.

A.  DESIGN CRITERIA

Important parameters used to develop the three model plants included
wastewater flow to be treated, BOD  loading,  and  suspended  solids
loading.   Data received from industry  (64, 66, 67) on projected raw
waste flows for new  mills  ranged  from  8333  gal/MSBu  to  66,667
gal/MSBu (1.2 to 9.9 cu m/kkg).  Data from more recently constructed
and  operating  corn  wet  mills  in  the  30,000  bu/day size range
indicate that a waste flow of  1  mgd   (3785  cu  m/day)  or  33,333
gal/MSBu  (5.0  cu  m/kkg) may reasonably be anticipated for a plant
employing recirculation of cooling water.  Since  the  latter  value
falls  within  the  mid-range of values projected by industry and is
demonstrated at mills presently in operation, it was chosen  as  the
design  value.   This  value  reflects wastewater loadings resulting
from in-plant controls presently practiced at  newer  mills  in  the
industry.  These plants do not necessarily reflect the incorporation
of  all in-plant pollution reduction controls that are known to have
application within the industry.

Industry projections of new  sweetener  plant  raw  waste  BOD  load
ranged  from  233  to  500  Ib/MSBu (4.2 to 8.9 kg/kkg).  For design
purposes, a load of 400 Ib/MSBu  (7.1 kg/kkg) was selected for a  new
plant,  a value that corresponds to the level selected for the model
plant in the  Development  Document  (9) .   The  cost  of  treatment
facilities  for a new plant generating only 250 Ib/MSBu (4.5 kg/kkg)
of raw waste BOD (at the lower end of the load  range  projected  by
industry)  was  also  evaluated to determine the significance of the
design value upon treatment facility capital and operating costs.

Projections of suspended solids loads for new corn  wet  mills  sub-
mitted  by  the industry ranged from essentially zero to 166 Ib/MSBu
(3.0 kg/kkg).  Information on existing mills indicates average  sus-
pended  solids  loadings of 200 Ib/MSBu  (3.6 kg/kkg), and this value
was chosen as a conservative number.

New plant raw waste criteria used in this evaluation are  summarized
below:
                            51

-------
    Capacity    _____ Flow __  _  _____ BOD ____ _  _____ TSS
                qal/MSBu  mgd   Ib/MSBu ~lb/day.~   Ib/MSBu   Ib/day.
     30,000      33,333   1.0       400    12,000       200    6,000

     60,000      33,333   2.0       400    24,000       200  12,000

     90,000      33,333   3.0       400    36,000       200  18,000

B.  WASTE TREATMENT COMPONENTS

As  discussed  in Section IV, data  available from treatment applica-
tions involving high strength organic wastes indicate that  a  well
designed  and  operated activated sludge  system  (or other biological
process) preceded by equalization and  coupled  with  good  in-plant
waste  controls  will produce a stable, high quality  effluent.  This
has been demonstrated within numerous industries and  at least at one
corn wet mill.  Furthermore, available data indicate  that  deep  bed
filtration   is  an  effective  means  of  polishing  effluent  from
biological treatment  systems.   This  is  discussed  previously  in
Sections VI and VII.

Based  on the above information, the treatment systems for the model
new corn wet mills are assumed to contain the  following  components
or unit processes:

1.   Grit removal

2.   Flow equalization

3.   Nutrient addition

4.   pH control

5.   Complete mix activated sludge

6.   Secondary clarification

7.   Chlorination of effluent

8.   Chemical coagulant addition

9.   Deep bed or mixed-media granular filtration

10.  Sludge thickening

11.  Sludge centrifugation

An  aerated  grit  chamber is provided with a design  capacity of 2.5
times the average wastewater flow.  The detention time  is  selected
as  3  minutes,  with  5 cubic feet per minute (cfm)  of air (0.14 cu
m/min)  being provided per linear foot (0.305  m)   of  grit  chamber.
The  detention  time in the equalization basin is chosen as 18 hours
at average daily flow.  Aeration and mixing are also  provided.   The

                             52

-------
activated  sludge  process  design  is  based on an F:M of 0.3, MLSS
concentration of 2000 mg/1, and aeration requirement of 1500  cfm/lb
 (93.6  cu  m/min/kg) of BOD applied.  The clarifier overflow rate is
chosen at 350 gpd/sq ft (14.2 cu m/day/sq m) with a  detention  time
of 4 hours.  The filters are sized for a loading rate of 4 gpm/sg ft
 (2.7  liters/sec/sq m).  All of the above design values are based on
widely used and accepted sanitary engineering design criteria.

C.  PERFORMANCE OF WASTE TREATMENT FACILITIES

Information gathered on treatment of brewery, distillery,  and  many
other  food  processing  wastes  and information on stable operation
attainable with treatment of corn wet milling wastes provide a sound
basis for predicting corn wet milling waste treatment facility  per-
formance.   The  data  indicate  that a well designed, operated, and
maintained system treating wastes from a  new  corn  wet  mill  will
produce  an  effluent containing 30 to 40 mg/1 of BOD and total sus-
pended solids  (TSS) on a long-term average basis.  For the model new
corn wet mills outlined above, these levels  correspond  to  8.3  to
11.1  Ib/MSBu   (0.1U8  to  0.198  kg/kkg)  of  BOD and TSS.  Data on
filtration of effluents from biological treatment of  municipal  and
industrial  wastes  provide an equally firm basis to conclude that a
properly designed deep bed filter will reduce the above effluent BOD
levels by 50 percent and  the  effluent  TSS  levels  by  70  to  80
percent.   These reductions are predicated on good biological treat-
ment preceding the filter.  The filter effluent for the model  plant
will  contain  a  BOD  load  of  4.2  to 5,6 Ib/MSBu  (0.075 to 0.100
kg/kkg)  and a TSS load of 1.7 to 3.3 Ib/MSBu  (0.030 to 0.059 kg/kkg)
on  a  long-term  average  basis.   A  schematic  diagram -  of   the
recommended  treatment  system  and  a  tabulation of the achievable
effluent levels for a 30,000 bushel/day corn wet mill are  presented
on the following page.

The effluent values presented above represent long-term averages and
must  now be related to the maximum monthly and maximum daily levels
in the New Source  Performance  Standards.   The  corn  wet  milling
industry   has  suggested  that  these  relationships  may  best  be
established by the use of  calculated  variability  factors.   While
such  statistical  calculations  are  not  without  merit, there are
important limitations  in  utilizing  this  approach  that  must  be
recognized  for the corn wet milling industry.  First, the data base
must  reflect  good,  stable  operation,  including  good   in-plant
controls  and  end-of-pipe treatment,  second, the data base must be
applicable to new plants.  The  data  on  existing  treatment  plant
performance  in  the corn wet milling industry are deficient in both
respects.  All existing complete treatment plants suffer from either
severe design and  operating  deficiencies  or  inadequate  in-plant
controls.    Plants   operating   under  such  1es s  than  des ir able
conditions will invariably experience greater effluent  fluctuations
and  more  frequent  biological system upsets.  In other words, such
plants will inherently generate a  higher  variability  factor  than
well-designed  and  operated  facilities at mills with good in-plant
controls.  Moreover, not one  of  the  mills  for  which  meaningful
                             53

-------
             EFFLUENT PERFORMANCE LEVELS FOR A
            NEW 30,000 BUSHEL/DAY CORN WET MILL
                                        Nutrientc/pII control
Haw W.-ustc
1.0 mgd
1440 mg/1 BOD
720 mg/1 TSS

Grit
Chamber



\
Aerated
Equalization




Complete Mix
Activated Sludge



                           Chloririat ion/Coagulants
To Receiving Water
1.0 mgd
15-20 mg/1 BOD
6-12 mg/1 TSS

Deep" Bed
Filtration

'
1.0 mgd
30-40 mg/1 BOD
30-40 mg/1 TSS

Secondary
Clarification



                                                                        "
Raw waste characteristics
Flow  1.0 mgd
BOD   1440 mg/1 = 12,000 Ib/day = 400 Ib/MSBu
TSS    720 n.g/1 =  6,000 Ib/day = 200 Ib/MSBu
Waste characteristics after grit removal, equalization,
activated sludge, and clarification
Flow   1.0 mgd
BOD   30-40 mg/1 - 250-334 Ib/day = 8.3-11.1 IVMSBu
TSS   30-40 mg/1 = 250-334 Ib/day = 8.3-11.1 Ib/MSBu
Waste characteristics after deep bed filtration
Flow  1.0 mgd
BOD   15-20 ir-g/1 - 125-167 Ib/day - 4.2-5.6 Ib/MSBu
TSS    6-12 mg/1 -  50-100 Ib/day - 1.7-3.3 Ib/MSBu
                        54

-------
                             SECTION IX

                     NON-WATER QUALITY ASPECTS
Factors such as air pollution control and solid waste  disposal  are
discussed  in  the  Grain Processing Development Document  (9).  This
discussion of nonwater quality aspects remains valid based  on  this
additional study.

One  industry  representative  has  provided  data  (67) on projected
power requirements for new corn wet mills.  These projected  demands
in terms of kilowatt-hours  (kwh) are presented below:
     Capacity
     (bu/day)

      30,000

      60fOOO

      90,000
Power Requirement
    (kwh/bushel) ~

        4.5

        4.0

        3.8
 Power
(kwh/day)

 135,000

 240,000

 342,000
The  same  industry  representative  estimates  that waste treatment
power requirements to meet New Source Performance Standards will  be
10  or  15  percent of total mill power requirements.  This estimate
appears to be high based on analysis of data provided  by  the  corn
wet  milling  industry in 1973 and projections of power requirements
for new treatment facilities.  These data indicate  that  a  maximum
increase  of  only  3  to 5 percent would be reasonably expected for
treatment facilities at a new corn wet mill.

-------

-------
                             SECTION X

                              SUMMARY
INTRODUCTION

Public  Law  92-500  requires  the  Agency  to  publish  performance
standards  for  new  sources  under authority of Sections 306 (a) and
307 (c).

Effluent Limitations for Existing Sources,  New  source  Performance
Standards  and  Pretreatment Standards for New Sources for the Grain
Mills Point Source Category were proposed on December 4, 1973  (38 FR
33438) and promulgated on March 20, 1974  (39 FR 10512).

On May 5, 1975, the U.S. Court of Appeals  for  the  Eighth  Circuit
remanded  to the Agency the new source performance standards and the
pretreatment standards for new sources  for  the  corn  wet  milling
subcategory  of the Grain Mills Point Source Category (40 CFR 406.15
and 406.16) promulgated by EPA under Section  306  and  307  of  the
Federal Water Pollution Control Act Amendments of 1972.

The new source standards for TSS and BOD were identical to those for
1983.   They  are  based  on  the  availability  of  the  technology
underlying the 1977  limitations  plus  the  addition  of  deep  bed
filtration.    The   Court  upheld  the  availability  of  the  1977
technology and the ability of  that  technology  to  meet  the  1977
limits in new plants.  However, it concluded that the record did not
demonstrate  that  deep bed filtration would achieve the incremental
reduction necessary to meet the new source standards.

Moreover, the Court held that the Agency's  analysis  of  the  costs
associated  with  the  new  source  standards  were deficient in two
respects.  First, the capital and operating costs for the new plants
were not separately prepared  but  developed  by  reference  to  the
incremental  cost  of  modifying existing sources to go from 1977 to
1983 levels.  Second, the costs were based on 1971 prices.

The  Court  remanded  the   new   sources   standard   for   further
substantiation   of   both   the  technical  aspects  and  the  cost
calculations.  The Agency was to promulgate  revised  standards  and
additional  supportive evidence for the present new source standards
within 120 days (i.e., September 2, 1975).

The purpose of this document is to provide supportive  evidence  for
EPA's  tentative  conclusion not to revise performance standards for
new sources in the corn wet milling subcategory of the  grain  mills
point source category in accord with the Court's remand order.

TECHNICAL BASIS

An  extensive  analysis  was  conducted  of  available  data  on the
application  of  activated  sludge  and  deep  bed  filtration   for


                             .61

-------
treatment  of  various  waste  waters.  Sources of data included the
literature, field inspections, special consultants, expert opinions,
corn wet milling companies, and equipment manufacturers.  The review
focused on the technology identified by EPA to meet the  New  Source
Performance  Standards;  namely,  good  in-plant  waste  control and
maximum water recycling, activated sludge biological treatment,  and
deep  bed  filtration.   The  data  unequivocally  and  unmistakably
substantiates the fact  that  high  strength  biodegradable  organic
wastes,   such  as  those  generated  by  corn  wet  mills,  can  be
successfully   treated   with   biological   treatment    processes,
particularly  complete mix activated sludge.  With proper design and
operation of treatment facilities, a stable  high  quality  effluent
can  be  attained  on  a  reliable  and  sustained  basis.  The data
strongly support the new source performance standards as  originally
promulgated.

Corn  wet  milling  wastes  may  originate  from  a  number  of unit
operations  in  the  wet  milling  process—steepwater  evaporation,
modified  starch  production,  and  syrup  refining  as well as less
pollutant producing processes of feed  de-watering,  oil  extraction
and refining, and general plant cleanup.

Waste  waters  from  the  industry can generally be characterized as
high-volume, high-strength discharges.  Based on summary  data  from
12  of  the  17  corn wet mills, BOD varies widely, from 255 to 4450
mg/1, with a corresponding range in COD.  Those plants with very low
BOD5 values typically have barometric condensing systems using once-
through cooling water.  At the other extreme, the very  concentrated
wastes  are  from  plants  using  recirculated cooling water (either
surface or barometric condensers).

Suspended solids levels in the  total  waste  streams  show  similar
variations  ranging  from  81  to  2458  mg/1.   The plants with low
suspended  solids  concentrations   are   those   using   barometric
condensers  with once-through cooling water.  The inter-relationship
of pollutant loads, pollutant concentrations, waste flow, and  plant
production is discussed in Appendix A.

BOD5  in  terms of raw material input (shelled corn)  ranges from 2.1
to 12.5 kg/kkg (119 to 699 Ibs/MSBu) , and averages 7.4  kg/kkg  (415
Ibs/MSBu).  Similarly, the suspended solids in the total plant waste
waters range from 0.5 to 9.8 kg/kkg (29 to 548 Ibs/MSBu) and average
3.8  kg/kkg  (211  Ibs/MSBu).   These  data emphasize again the wide
variation  in  waste  characteristics  from  the  corn  wet  milling
industry.    The waste water flows vary from 3.1 to 41.7 cu m/Kkg (21
to 280 gal/SBu) with an average of  18.3  cu  m/kkg  (123  gal/SBu).
Those plants with lower waste flows per unit of production are those
that employ recircula-ting cooling water systems.

Most  plants  segregate their major process waste water from cooling
water prior to treatment.  Once-through cooling  water  systems  are
being replaced with recirculating systems, in several instances.
                            62

-------
In  the development of New Source Performance standards for the Corn
Wet Milling Subcategory, a specific methodology  was  followed.   In
determining representative raw waste loading for new corn wet mills,
a  medium-sized mill with a daily grind of 1524 kkg (60,000 SBu) was
selected.  This  hypothetical  mill  would  practice  good  in-plant
control  and  incorporate  use  of  recirculated cooling water.  The
waste water characteristics of this  mill  reflect  actual  industry
practice  based on average data received from existing mills.  These
waste water characteristics would be as follows:

     Flow                 11,355 cu m/day       (3.0 mgd) (50 g/SBu)
     BOD5                 7.14 kg/kkg           (400 Ibs/MSBu) 960 mg/1
     Suspended Solids     3.57 kg/kkg           (200 Ibs/MSBu) 480 mg/1

The pollutant potential of the raw waste  discharge  from  a  60,000
SBu/day  corn wet mill is equivalent to the untreated waste expected
from a domestic population of 138,000.

In the development of the new source performance standards, a number
of  alternative  treatment   systems   were   identified   for   the
representative  corn  wet  mill.  The investment and annual cost for
each alternative, and the resultant pollutant load  reductions  were
identified  (9).

The specific technology identified to facilitate compliance with the
recommended New Source Performance standard for the Corn Wet Milling
Subcategory,  Grain Mills Point Source Category was a combination of
biological/physical treatment.  For the Corn Wet Milling Subcategory
of the grain milling industry, the new source  performance  standard
and  best  available  technology  economically  achievable  comprise
improved solids separation following activated sludge or  comparable
biological treatment.  Improved solids separation can be represented
best  by  deep  bed  filtration  although alternative systems may be
available.  The "exemplary" technology includes 12 to  18  hours  of
aerated  equalization  ahead  of  a  complete-mix  activated  sludge
process  with  associated  chemical  feed,   sedimentation,   sludge
dewatering facilities  (centrifugation), grit removal,  pH adjustment,
nutrient  addition,  and  deep  bed  filtration of the biologically-
treated effluent.  BOD5 and suspended solids concentrations of 20 to
30 mg/1 and 10 to 20 mg/1 respectively, are expected in the effluent
from this  series  of  treatment  processes.   These  concentrations
correspond  to  effluent  loads  of 0.15 to 0.22 kg/kkg  (8.3 to 12.5
Ibs/MSBu) of BOD5 and 0.07 to 0.15 kg/kkg (4.2 to 8.3  Ibs/MSBu)  of
suspended solids.  BOD5 and suspended solids reductions expected are
about 97.4 and 96.9 percent respectively.

In  achieving  the  recommended  new  source  performance  standards
through this technology, deep bed filtration of biologically treated
waste attributes  to  an  additional  50  to  75  percent  BOD5  and
suspended   solids   removal  beyond  biological  treatment,  alone.
Biological treatment, alone, attributes  to  producing  an  effluent
before  filtration  of 75 - 125 mg/1 BOD5 and suspended solids, with
corresponding overall BOD5 and TSS reductions of 90 and 80 percent.
                            63

-------
In meeting  the  New  .Source  Performance  Standard,  the  pollutant
potential of the representative plant is reduced from the equivalent
of 138,000 people to 6,700 people in terms of waste load.

In  evaluating waste water control in the corn wet milling industry,
it is essential to  evaluate  both  in-plant  control  measures  and
effluent  treatment  systems.   Good  in-plant  controls can greatly
reduce  the  total  raw  waste  load  and  improve  treatment  plant
efficiency.   Various  in-plant  controls  identified for New Source
Performance Standards include:

     1.    Isolation and treatment of all process waste waters.   No
           process  wastes  should  be  discharged  without adequate
           treatment.

     2.    Elimination of once-through  barometric  cooling  waters,
           especially  from  the  steepwater  and syrup evaporators.
           This change can be accomplished  by  recircalating  these
           cooling  waters  over  cooling  towers  or  replacing the
           barometric condensers with surface condensers.

     3.    Isolation  of  once-through  noncontact  (uncontaminated)
           cooling  waters  for  discharge directly to the receiving
           waters  or  provision  of  recirculating  cooling   tower
           systems  with  the • blowdown  directed  to  the treatment
           plant.

     U.    Installation of dikes at all  process  areas  subject  to
           frequent  spills  in  order  to  retain  lost product for
           possible reuse or by-product recovery.

     5.    Installation  and  maintenance  of   modern   entrainment
           separators in steepwater and syrup evaporators.

     6.    Surveillance and monitoring of  major  waste  streams  to
           identify and control sources of heavy product losses.

     7.    Provision of extensive waste treatment for the  resulting
           process  waste  waters  consisting  of:  flow and quality
           equalization, neutralization, biological  treatment,  and
           solids  separation.   The  biological  treatment  methods
           available i.nclude activated sludge, pure oxygen activated
           sludge, bio-discs,  and  possible  combination  of  other
           biological systems.

     8.    Institution of  maximum  water  reuse  practices  at  all
           plants over and above the current levels of practice.

     9.    Provisions to improve solids recovery at individual waste
           sources.

In-plant housekeeping and good operation can have a major impact  on
the  raw waste loads from a mill.  Diking of spill areas, monitoring
and careful operation have been reported to reduce raw  waste  loads

                             64

-------
by  25  to  50  percent in some plants.  The combination of in-plant
controls and proper waste treatment constitutes a practicable  means
for achieving the specific effluent limitations.
         AND DETERMINATIONS
The  end-of-pipe  treatment  technology recommended for new corn wet
mills, activated sludge and  filtration,  was  reviewed  within  and
without the corn wet milling industry.  The activated sludge process
has  been  proven  as  an  effective  treatment  method  in numerous
applications.  These  applications  include  municipal  or  domestic
wastes  and  high strength organic industrial wastes.  Of particular
emphasis in this study were applications  of  the  activated  sludge
process  to  food  processing wastes with characteristics similar to
those generated by corn wet mills.  The data  established  the  fact
that  the  activated sludge process can effectively treat these high
strength wastes and produce an acceptable long-term average effluent
quality.

In industries generating wastes similar to corn wet milling  wastes,
such  as  brewing,  distilling,  and  malting,  the  data  show that
consistently high levels of effluent reduction  are  attained  on  a
long-term  basis.   High  quality  effluents  with BOD and suspended
solids less than 30 mg/1 are consistently and routinely produced  at
many  treatment  facilities.  This has been demonstrated with a wide
variety  of  industrial  wastes,  including   those   generated   by
breweries,  edible  oil  refineries,  malting  plants, distilleries,
bakeries, wineries, and fruit and vegetable processing plants.

Within these industries raw waste loads range generally from 750  to
6700  mg/1  of  BOD5  and  50  to  4000 mg/1 total suspended solids.
Biological  treatment  with   or   without   filtration   has   been
demonstrated  to  result  in  overall  BOD  and TSS removals ranging
between 87 and 99.4 percent BOD, and 67.0  and  99.4  percent  total
suspended  solids,  generally  greater  than  96 percent for BOD and
suspended solids.  Effluent levels as low as 10 to 20 mg/1  BOD  and
suspended  solids  have consistently been achieved.  For purposes of
comparison of waste strength, corn wet milling wastes ranges between
250 to 4450 mg/1 BOD, and 80 to 2450 mg/1 total suspended solids.

Data and experience indicate that these industries  are  subject  to
most of the same elements of waste load variability, waste strength,
and required treatment mechanisms as experienced within the corn wet
milling industry.  In fact, a detailed statistical evaluation of raw
waste  load  variability  for  both the brewing and corn wet milling
industry performed in this study indicate an unmistakable similarity
in waste characteristics, waste load variability, and general nature
and biodegradability of the waste.  Waste waters  generated  by  two
breweries  which  now  discharge  treated wastes to navigable waters
substantiate  the  treatability  of  brewing  waste  to  very   high
pollutant  reduction  levels.   Corn wet milling waste is not unique
among many  food  processing  wastes,  and  presents  no  enigma  in


                            65

-------
accomplishing  high  levels  of  treatment  under  present  sanitary
engineering practice.

Activated sludge has yet to be  applied  to  its  fullest  potential
within  the corn wet milling industry because of poor in-plant waste
controls, design inadequacies,  and  operational  deficiencies.   If
these   were   eliminated,   the   activated  sludge  process  could
successfully treat corn wet milling wastes and reliably  produce  an
effluent with BOD5 and TSS of 30 mg/1 or less on a long-term average
basis.   The  New  Source  Performance  Standards  can  be met at an
existing corn wet mill employing present day technology.

peep bed filtration is a demonstrated and available technology.   It
is  universlly  applied  in  the  treatment  of  water  supplies for
municipalities and industries and can produce  effluents  containing
essentially  no  suspended solids in these applications.  Filtration
has been successfully employed in the  treatment  of  municipal  and
industrial   secondary   (biologically   treated)   effluents.   The
literature, expert consultants, and other data clearly document this
fact.  The process has been sufficiently demonstrated to enable  the
development  of  meaningful design criteria and has been employed in
numerous applications.  In municipal applications, a  well  designed
filter  can  be expected to reduce suspended solids by 80 percent or
more and to produce an effluent containing less than  10  mg/1  TSS.
Similar  results  can  be  anticipated with industrial wastes if the
preceding biological treatment system is well designed and operated.
It lias also been demonstrated that deep bed filtration  performs  as
an  effective  back-up system during periods of biological treatment
upset and thus helps to  insure  that  a  stable  effluent  will  be
produced.

In  addition  to applications outside the corn wet milling industry,
filtration may be successfully applied to corn wet  milling  wastes.
The  process  has been demonstrated with other high strength organic
wastes.  Expert opinion supports this applicability, as well as  the
ability  of  a well designed treatment system to meet the New Source
Performance  Standards,   One   corn   wet   milling   company   has
successfully   applied  deep  bed  filtration  following  biological
treatment on a full-scale basis.  At Clinton Corn Processing Company
in Clinton, Iowa, filtration has been demonstrated as  an  effective
means of further reducing BOD and suspended solids levels in treated
corn wet milling effluent.  Suspended solids reductions of 50 to 100
percent  and averaging better than 75 percent were demonstrated with
deep bed  filtration  of  treated  corn  wet  milling  wastes.   BOD
reduction of up to 50 percent was demonstrated.

Industries that are either employing filtration or have successfully
tested  and  are  installing  filters include:  steel manufacturing,
petroleum refining, brewing, corn wet milling, wine processing,  and
food   processing.    Multi-media   filtration  has  had  particular
application in the petroleum refining industry, where at least three
refineries are known to successfully  treat  refinery  waste  waters
after  biological  treatment.   These applications represent some of
the longest-term use of modern filtration facilities  employed  with


                          66

-------
biological  treatment  systems  for  treatment  of strong industrial
waste.

The costs required for new corn wet mills of various sizes  to  meet
the  New  Source  Performance  Standards  were  reevaluated  and are
indicative of current pollution control technology.   These costs are
based on January 1975 dollar  values  and  include  waste  treatment
facilities   and  necessary  in-plant  controls  or  cooling  system
designs.

Based on an extensive review  of  technical  data,  the  New  Source
Performance Standards for the Corn Wet Milling Subcategory should be
implemented   as  promulgated  on  March  20,  1974.   As  the  data
demonstrate,  new  plants  employing  best  available   demonstrated
control technology can readily achieve these standards.

A reasonable prediction of waste treatment facility performance at a
new  corn  wet  mill  indicates  that  the  New  Source  Performance
Standards  can  be  readily  attained   with   currently   available
technology.   There  is  adequate  provision  in  the  standards for
anticipated fluctuations in effluent quality.

There  are  no  increased  economic  costs,  intermediate   effects,
programatic  or  energy  consequences  expected  as  a result of the
technology reevaluation for the  New  Source  Performance  Standard.
Evaluation   of   the  nonwater  quality  aspects  of  applying  the
recommended technology indicated that  energy,  air  pollution,  and
solid waste impacts will be minimal.

Based  on  the  above outlined study and analysis, the Administrator
has evaluated data and the performance of biological  treatment  and
deep  bed filtration, waste load variability, industrial application
of pollution control techniques for  grain  processing  as  well  as
similar  types  of  wastes,  and  has  concluded that the New Source
Performance Standards as originally promulgated are proper and well-
founded.

The Administrator has determined that the  technology  is  available
and  transferrable  with  reasonable  prediction that the technology
will be capable of removing the increment required by the New Source
Performance Standards.  The costs associated with the technology  do
not preclude or adversely affect its effective use and application.
                            67

-------

-------
                             REFERENCES


1.   Sverdrup 6 Parcel, Literature Review on Deep Bed Filtration,
     June 1975.

2.   Letter to Mr. Charles F. Lettow, Record File 4823-EPA-2,
     May 22, 1975.

3.   Metcalf 6 Eddy, Waste5/ater_Engineering, McGraw-Hill,
     New York, 1972.

4.   Roy F. Weston, Inc., "Process Design Manual for Upgrading
     Existing Wastewater Treatment Plants", for U.S. EPA Technology
     Transfer, October 1971.

5.   McKinney, Ross E. , Microbiology for Sanitary Engineers,
     McGraw-Hill, New YorkT 19627

6.   Nemerow, Nelson L.r Liquid Haste of Industry - Theories,
     Practices f and Treatment, Addison-Wesley, Reading,
     Massachusetts, 1971.

7.   Busch, Arthur W. , Aerobic Bilogical_ Treatment of Waste Waters,
     Oligodynamics Press7 HoustonT 1971.

8.   U.S. EPA, "Draft Development Document for Effluent Limitations
     Guidelines and New Source Performance Standards for the Miscel-
     laneous Foods and Beverages Point source Category," March 1975.

9.   U.S. EPA, "Development Document for Effluent Limitations Guide-
     lines and New Source Performance Standards for the Grain Process-
     ing Segment of the Grain Mills Point Source Category," March 1974.

10.  Cook, Charles, Memorandum to John Riley - Fluctuations in raw
     waste load BODS from two breweries, July 15, 1975.

11.  issac, P. G. , "Malting Effluents," Effluent and Water Treatment
              November 1969.
12.  Smith, A. J. , "Waste Treatment in the Liquor Distilling Industry,"
                     , March/April 1972.
13.  Burkhead, C. E. , Lessig, C. A., Jr., Richardson, T. R. ,
     "Biological Treatment of a Distillery Waste," 23rd Industrial
     Waste Conference, Purdue University, May 7-9, 1968.

14.  McKee, J., "$300,000 Waste Treatment at McKee Baking," Baking
     Industry, October 1972.

15.  Pearson, E. A., et al, "Treatment and Utilization of Winery
     wastes," Proceedings of the 10th Industrial Waste Conference,
     Purdue University, May 1955.
                              69

-------
16.  Schroeder, W. D., et al, "Biological Treatment of winery
     stillage". Proceedings of the Uth National Symposium on Food
     Processing Wastes, December 1973.

17.  Ryder, R. A., "Winery Waste Treatment and Reclamation,"
     Proceedings of the 28th Industrial Waste Conference, Purdue
     University, May 1973.

18.  "Winery Innovates Waste Treatment,11 Food Engineering.
     June 1972.

19.  Tofflemire, T, J., et al, "Unique Dual Lagoon System Solves
     Difficult Wine Waste Treatment Problem", Water & Wastes
     Engineering/Industrial^ November/December 1970.

20.  Data on Taylor Wine Company waste treatment, Hammondsport,
     New York, received from EPA July 3, 1975.

21.  Data on Morton Frozen Foods waste treatment, Crozet, Virginia,
     received from EPA July 18, 1975.

22.  Data on fruit and vegetable processing waste treatment -
     statistical analysis; effluent data on Southwestern Oil and
     Refining, Corpus Christi, Texas; data on Welch Foods treatment
     system, Brocton, New York; received from EPA July 10, 1975.

23.  Brown, D. R., and Van Meer, G. L., "Biological Treatment of
     Wastes from the Corn Wet Milling Industry," Final EPA Grant
     Report on Pekin Waste Treatment, August 30, 1973.

24.  west, A. W,, "Report on April 19, 1972 Investigation of the
     Wet Corn Milling Waste Treatment Plant, CPC International Inc.,
     Pekin, Illinois,"'EPA, Cincinnati, Ohio, May 1972.

25.  Repta, R. J., "Activated Sludge Treatment of a Corn Wet Milling
     Waste," M.S. Thesis, Illinois Institute of Technology, December
     1973.  (also titled Progress Report - Improve Pekin Waste Treat-
     ment Plant I, November 1972 - February 1973.)

26.  Correspondence with CPC International Inc., May 29, 1975.

27.  Correspondence with CPC International Inc., June 6, 1975.

28.  correspondence with CPC International Inc., June 9, 1975,

29.  Roy F. Weston, Inc., "Process Evaluation Report for Upgrading
     Existing Wastewater Treatment Facilities," for CPC International
     Inc., Corpus Christi, Texas, March 15, 1975.

30.  Memorandum - visit to CPC*s Corpus Christi corn wet mill. Record
     File 4823-EPA-28, June 23, 1975.

31.  Correspondence with American Maize-Products Company, June 18,
     1975.


                            70

-------
32.  Correspondence with American Maize-Products Company, June 9,
     1975.

33.  Correspondence with Anheuser-Busch, Inc., June 6, 1975.

34.  Memorandum - meeting with EPA and corn wet millers. Record File
     4823-EPA-29, June 23, 1975.

35.  U.S. EPA, "Process Design Manual for Suspended Solids Removal,"
     Technology Transfer Office, January 1975.

36.  Weber, Walter J. , Jr., Phvsicochemical Process es for Water
     Quality Control, Wiley-Interscience, New York, 1972.

37.  Fox, David M. , and Cleasby, John L. , "Experimental Evaluation
     of Sand Filtration Theory," Journal of the Sanitary Engineering
              ^ASCE, SA5, October~1966.
38.  American Water Works Association, Inc., Water_2uality._and
              , Third Edition, McGraw-Hill, New YorkT 197l7
39.  Correspondence with Neptune Microfloc, Inc., June 24, 1975.

40.  Cleasby, J. L. , and Baumann, E. R. , "Wastewater Filtration Design
     considerations," U.S. EPA Technology Transfer Seminar Publication,
     July 1974.

41.  Hsiung, K. Y. , and Cleasby, J. L. , "Prediction of Filter Per-
     formance," Journal of the Sanitary Engineering Division, ASCE,
     94, December~19687

42.  Lynam, B. T. , Ettelt, G. , and McAloon, T. , "Tertiary Treatment at
     Metro Chicago by Means of Rapid Sand Filtration and Micros-trainers,"
     Journal Water Pollution Control Federation , 41, February 1969.

43.  Tchobanoglous, G. , "Filtration Techniques in Tertiary Treatment,"
                                     Fe deration , 42, Apr i 1 1970.
44.  Tchobanoglous, G. , and Eliassen, R. , "Filtration of Treated Sewage
     Ef f 1 uent , " Journal of the Sanitary Engineering Division, ASCE,
     April 1970.

45.  Baumann, E. R., and Huang, J. C. , "Granular Filters for Tertiary
     Wastewater Treatment," Journal Water^ Pollution ^Control Federation,
     46, August 1974.

46.  Gulp, G. L. and Hansen, S. P., "Extended Aeration Polishing
     by Mixed Media Filtration," Water_and_ Sewage_Works ,
     February 1967.

47.  Culp, R. L. , arid Gulp, G. L., Adyancedjtfagtewater Treatment,
     Van Nostrand Reinhold, New York, 1971.

                            71

-------
48.  Vecchiolo, Jr., et al, "Wast.ewat.er Reclamation and Recharge,
     Bay Park, New York," Journal Sanitary Engineering Division.
     A§CE, April 1975.

49.  "Ultra-High Rate Filtration of Activated Sludge Plant Effluent,"
     EPA-R2-73-222, April 1973.

50.  Middlebrooks, E. J., et al, "Evaluation of Techniques for Algae
     Removal from Wastewater Stabilization Ponds,n Utah Water Research
     Laboratory, Utah State University, Logan, Utah, January 1974.

51.  Correspondence with Dravo Corporation, Water and Waste Treatment
     Division, May 30, 1975.

52.  Memorandum-discussion with Infilco-Degremont, Record File
     4823-EPA-23, June 18, 1975.

53.  Correspondence with General Filter Company, June 24, 1975.

54.  Memorandum - discussion with Welch Foods, Record File
     4823-EPA-17, June 17, 1975.

55.  Memorandum - discussion with Robert Kerr Laboratories, Ada,
     Oklahoma, Record 7ile 4823-EPA-21, June 18, 1975.

56,  Savage, E. S. , "Deep-bed Filtration of Steel Mill Effluents"
     Proceedings of 17th Ontario Industrial Waste Conference, Niagra
     Falls, Ontario, June 7-10, 1970.

57.  Data on deep bed filtration of refinery wastewaters:  Amoco,
     Yorktown, Virginia; Marathon Oil, Robinson, Illinois; Marathon
     Oil, Texas City, Texas; received from EPA July 21, 1975.

58.  Memorandum - discussion with Clark Oil and Refining, Hartford,
     Illinois, Record File 4823-EPA-36, July 15, 1975.

59.  Data on Welch Foods treatment system, received from EPA
     May 30, 1975.

60.  U.S. EPA, "Development Document for Effluent Limitations
     Guidelines and New Source Performance Standards for the
     Renderer Segment of the Meat Products and Rendering Process
     Point Source Category," January 1975.

61.  Baumann, E. R., "Design of Filters for Advanced Wastewater
     Treatment," Project 1002-S, Engineering Research Institute,
     Iowa State University, Ames, Iowa, June 1973.

62.  Correspondence with Dr. E. R. Baumann, Iowa State University,
     July 22, 1975.

63.  Memorandum - meeting with Clinton Corn Processing Company on
     June 2, Record File 4823-EPA-14, June 12, 1975.


                            72

-------
64.  Correspondence with Clinton Corn Processing Company, June 17,
     1975.

65.  Memorandum - Clinton sampling results. Record File 4823-EPA-32,
     June 20, 1975.

66.  Correspondence with A. E. staley Manufacturing Company,
     June 19, 1975.

67.  Correspondence with CPC International Inc., June 26, 1975.
                                                      i
68.  Brief for Respondents, Case 74-1448, U.S. Court of Appeals for
     the Eighth Circuit, January 1975.

69.  U.S. EPA, "Development Document for Proposed Effluent Limita-
     tions Guidelines and New Source Performance Standards for the
     Animal Feed, Breakfast Cereal, and Wheat Starch Segment of
     the Grain Mills Point Source Category," September 1974.

70.  Patterson, w. L., and Banker, R. F. , Black 6 Veatch Consulting
     Engineers, "Estimating Costs and Manpower Requirements for
     Conventional Wastewater Treatment Facilities," Report for the
     Office of Research and Monitoring, U.S. EPA, October 1971.

71.  Koon, J. H., Adams, C. E., Jr, Eckenfelder, W. W., Jr.,
     "Analysis of National Industrial Water Pollution Control Costs,"
     for Office of Economic Analysis, U.S. EPA, May 21, 1973.

72.  Smith, Robert, "Cost of Conventional and Advanced Treatment of
     Waste Waters," Federal Water Pollution Control Administration,
     U.S. Department of the Interior, 1968.

73.  Smith, Robert, and McMichael, W. F., "Cost and Performance
     Estimates for Tertiary Waste Water Treating Processes,"
     Federal Water Pollution Control Administration, U.S.
     Department of the Interior, 1969.
                             73

-------

-------
                                             APPENDIX A
4
 *
                                            75

-------

-------
                             APPENDIX A
           EXPLANATION OF INTER-RELATIONSHIP OF POLLUTANT
           CONCENTRATION, UNIT POLLUTANT LOAD, PROCESSING
                     RATE. AND WASTE WATER FLOW

The  expression of pollutant loads in terms of concentrations  (mg/1)
is useful for identifying and  evaluating  the  characteristics  and
performance  of treatment measures for organic biodegradable wastes.
Concentration for biodegradable organic  waste  waters  expresses  a
commonality  for  comparison  of  waste  water  characteristics  and
treatability results regardless of the source of the similar organic
biodegradable waste.  For effluent limitations guidelines  purposes,
the  specific  limitations  are  commonly expressed in terms of unit
pollutant per unit of raw material processed or  product,  whichever
is  most  applicable.   The new source performance standards for the
corn wet milling subcategory are expressed  as  unit  production  of
pollutant  e.g.,  BOD5  or  TSS  per 1000 standard bushels  (MSBu) of
shelled corn (raw material).  Concentration levels in terms of waste
loads can be related as follows:

     Waste water concentration, mg/1 = U x P
                                       8.34~x F

     Where:

    , U = unit pollutant load per unit of production,
         Ibs/MSBu of shelled corn

     P = Daily grind rate, MSBu of shelled corn/day

     F = representative or actual flow rate, million
         gallons per day  (MGD)
                              77

-------

-------
1-

-------
U.S. ENVIRONMENTAL PROTECTION AGENCY (A-107)
WASHINGTON, D.C. 20460
           POSTAGE AND FEES PAID
ENVIRONMENTAL PROTECTION AGENCY
                        EPA-335

-------