United States
                   Environmental Protection
                   Agency
Water Engineering
Research Laboratory
Cincinnati, OH 45268
                   Research and Development
EPA/600/S2-88/032 July 1988
&EPA         Project  Summary
                   Evaluation  of  Biological
                   Treatment  of  Pharmaceutical
                   Wastewater  with  PAC Addition

                   David A. Gardner and Richard A. Osantowski
                    A lacK of information on applicable
                  removal  technologies  for  total
                  chemical oxygen demand (TCOD)
                  prevented promulgation of best
                  available technology economically
                  achievable (BAT) limitations and new
                  source  performance  standards
                  (NSPS) for TCOO for pharmaceutical
                  manufacturing plants  in  1983
                  (EPA/440/1-83/084).  Therefore,  in
                  1984 U.S. EPA conducted a pilot plant
                  study of activated carbon  treatment
                  technologies utilizing pharmaceutical
                  wastewaters from a manufacturing
                  plant that produces fermentation
                  products (Subcategory A)  and
                  chemical synthesis products
                  (Subcategory C). One technology
                  that was  evaluated  was  powdered
                  activated carbon (PAC) addition to an
                  activated  sludge system (PACT). A
                  viscous floating mass of mixed  liquor
                  solids (VFMLS)  developed in  the
                  PACT units and   resulted   in
                  premature termination of  the study.
                  Therefore, in 1987 an additional pilot
                  study was performed at the site of
                  the 1984  study. The purposes of this
                  study were to: 1) attempt to find  the
                  cause of the formation of the VFMLS,
                  2) generate additional research data
                  for  TCOD removal from  phar-
                  maceutical wastewater using  the
                  PACT process, 3)  evaluate the
                  efficiency of PACT in  removing
                  specific  organics, 4) evaluate  the
                   "Registered Patent Process.  Mention of trade
                   names or commercial products does not
                   constitute endorsement or recommendation for
                   use.
effectiveness of PACT  in reducing
efflu-ent aquatic toxicity,  and 5)
evaluate the use of a selector to
improve the settling characteristics
of the mixed liquor. One control unit,
two PACT units, and a unit equipped
with a series of selector basins were
operated.
   This  Project Summary  was
developed by EPA's Water Engineering
Research Laboratory, Cincinnati,  OH,
to announce  key  findings of  the
research  project  that is fully
documented in  a separate report of
the same title  (see Project Report
ordering information at back).

Introduction
   In November 1982. the EPA proposed
effluent limitations and standards for the
pharmaceutical  manufacturing point
source category. The proposed BAT and
NSPS limited  the discharge of  the
nonconventional pollutant,  TCOD, from
existing and new sources.  However, due
to a lack of information  on applicable
TCOD  removal technologies  for
subcategories A (fermentation products)
and C (chemical  synthesis products),
limitations and standards for TCOD were
not promulgated  along  with other
regulations on October 27, 1983.
   By  way of  responding to these
additional information  needs,  from
September to December 1984, EPA
conducted  biological and physical-
chemical  pilot plant evaluations of
activated  carbon  treatment technologies
utilizing  actual  pharmaceutical
wastewaters. The biological treatment
systems  consisted of three  activated
sludge units operated in parallel. PAC

-------
was added to two of the units in selected
dosages. The third unit was used as a
control.  The PAC  enhanced biological
treatment systems operated during the
1984 test period reduced  influent TCOD
concentration  to  levels  below  those
required by  the  1976 best  practicable
control technology  currently available
(BPT) regulation. This regulation required
an average  74%  reduction from  raw
waste levels. PAC also  improved the
settling  rate  of  the mixed   liquor
suspended  solids  (MLSS)  when
compared to the control unit. A VFMLS
developed in the  PACT units near the
end of the acclimation period  for the
second set of PAC dosages  and resulted
in  premature termination  of the  study.
The VFMLS did not appear in the control.
   Due to the formation of the VFMLS, it
was decided that additional information
was required before  a  PACT  system
could  be  considered  as a   viable
treatment process  for subcategories  A
and  C  pharmaceutical  wastewaters.
Therefore, from   March to  July  1987
additional biological pilot  plant  studies
were performed at the site of the original
1984  pilot tests.  Four biological pilot
units were used in the 1987 study. One
unit was operated  as a control, two units
were operated  with PAC addition to the
aeration basins,  and one unit was
operated with a series of selector basins
ahead of the  aeration basin.
   The purposes of this study were to:  1)
follow  up on  the previous work  and
attempt to find the cause of the formation
of the VFMLS, 2) generate additional
research data  for TCOD  removal from
pharmaceutical wastewater using the
PACT process, 3)  evaluate the efficiency
of PACT in  removing specific organics,
4) evaluate the effectiveness of PACT in
reducing effluent  aquatic  toxicity  as
measured by  bioassay  tests,  and  5)
evaluate the  use of a  selector to improve
the settling characteristics of the mixed
liquor.


Quality Assurance/Quality
Control Methods
   Sample  analyses  were  conducted
on-site  in a field  laboratory and in a
base  laboratory  in  another state.  All
samples were collected by the pilot plant
operators using  prelabeled  containers.
Analyses  for pH,   TCOD,  soluble
chemical oxygen  demand (SCOD),
nitrogen  inhibited  biochemical  oxygen
demand (NIBOD), nitrogen inhibited
soluble biochemical  oxygen  demand
(NISBOD),  chloride,  total  suspended
solids (TSS),  and  volatile suspended
solids (VSS) were performed at the on-
site laboratory. All  other analyses  were
performed  at  the  base laboratory.  A
portable computer  was also located  in
the laboratory trailer during the study  to
perform  engineering  and  analytical
calculations and  to  store and transmit
data between the field site  and  home
office.
   All  samples  were iced during
collection,  refrigerated  after collection,
and iced during  shipment  to the  base
laboratory  (samples  that were analyzed
on-site were usually analyzed the same
day  they  were collected  but  were
preserved  for reanalysis if  necessary).
Samples that  were sent to  the  base
laboratory  for soluble  phosphate  (SP),
nitrate + nitrite  (N03 + N02),  and
ammonia (NHs)  analysis and samples
analyzed on-site  for TCOD and SCOD
were preserved with sulfuric acid addition
to pH  2.  All  samples  for  soluble
parameters were filtered through 0.45 pm
filter paper prior to preservation. Samples
for chloride analysis were shipped to the
home laboratory unpreserved.
   TCOD  and SCOD  analyses  were
performed  by the  Hach  colorimetric
method (Hach Water Analysis Handbook,
No.  11-01-84-SED, Hach  Chemical
Company, Loveland, Colorado, 1984). All
other analyses  for the  conventional
parameters were  performed according  to
the procedures described in  Methods for
Chemical Analysis of Water  and Wastes
(EPA/600/4-79-020,   U.S.  Environ-
mental Protection  Agency,  Cincinnati,
Ohio,  1979) or Standard Methods for the
Examination of Water  and  Wastewater,
15th Ed. (1980).  At least  20%  of  all
samples were analyzed in replicate for all
conventional parameters except  NIBOD
and  NISBOD.  One  sample each   of
NIBOD and NISBOD  was duplicated
each day that analyses were performed.
At least 10% of all samples  were spiked
with appropriate reagents for analysis  of
all conventional parameters except TSS,
VSS,  NIBOD,  and  NISBOD. Blanks
and/or standards were analyzed with
each  analytical  run or sample lot  as
appropriate for the analysis.
   QA/QC objectives for precision  were
to have a maximum of 20%  difference
between replicates  for NIBOD, NISBOD,
and VSS. A maximum of 15% difference
between  replicates was sought for all
other conventional parameters. Accuracy
objectives were  ±  20% spiked  sample
recovery for NIBOD and NISBOD, and a
 ± 15% spiked sample recovery for other
parameters.
   Of  2,131 replicate  samples,  2,093
(98.2%) were within the precision
percent difference criteria. Of 190 spiked
samples,  186 (97.9%) were within the
percent recovery objectives. In all, 3,300
analyses  were performed,  with 3.6%  of
the analytical results deleted for QA/QC
reasons. The  project  had  a  95.8%
completion rate for planned analyses.
   During the two intensive test periods,
samples  were  analyzed  for  specific
organic compounds, for  aquatic  toxicity,
and by the toxicity characteristic leaching
procedure (TCLP) at  various  contract
laboratories.  EPA approved methods
were used for all analyses.


Pilot Study Procedures
   This study consisted of three separate
parts: 1) the  PACT study to evaluate the
PACT process,  2) the  selector study  to
evaluate the  use of selector technology
for improved  mixed liquor settling, and  3)
bench tests  to  try and  reproduce the
VFMLS.

PACT Study
   From March to June 1987, a  101-day
study  was   conducted   using  one
biological pilot plant as  the control and
two  biological  pilot  plants  as  the
experimental (PACT) systems. The three
activated  sludge units  were operated  ir
parallel.

System Operation-
   Raw pharmaceutical  wastewater was
collected  just before it entered  the full
scale aeration basins and  was used  a.'
the feed  to  the pilot  units.  This flov
consisted of  equalized  pharmaceutica
wastewater with  a nominal amount (abou
2%) of macerated, chlorinated  sanitar
waste.
   The three activated sludge units wen
identified  as Unit 1 (control),  Unit  .'
(PACT unit),  and Unit 3  (PACT  unit;
Each system consisted of a rectangula
aeration  tank (189 L)  and a  conice
clarifier (47.3 L). Each  aeration tank wa
operated  as  a completely mixed basir
Peristaltic  pumps were  used to suppl
wastewater  and return  sludge to th
aeration basins.  PAC slurry was manual!
added twice  per day to the front end  (
the aeration   basin. A  peristaltic pum
controlled by a timer wasted sludge froi
the discharge end of each aeration basir
   The targeted aeration basin hydraul
retention  times  (HRT's)  and solic
retention  times (SRT's) were 2.2 and  1
days, respectively,  for  all system
During the first test period,  198 and 6J
mg/L of PAC were added to the  influen
of Unit 2 and Unit 3, respectively, and

-------
the second test  period,  396 and  1,190
mg/L of PAC were added to the influents
of Unit 2 and Unit 3, respectively. Daily
composite samples were taken from the
raw feed wastewater and from the three
clarifier effluents  using  automatic
sampling equipment, and from  the
wasted  mixed  liquor  obtained from the
waste solids holding tanks.

Test Phases-
   The planned PACT pilot operation was
scheduled for a series of five phases as
described below:

   Phase  1  -  Start-up—The  three
biological pilot plants were operated for 3
days  after being seeded  with  mixed
liquor from  the full-scale aeration
basins.  The objective of this phase was
to determine  whether the effluents for
each  of the  three  pilot  plants  were
comparable  with  respect to SCOD
removal.
   Phase 2 - Acc//maf/o/7--Following
Phase 1, PAC addition to the influents
was initiated for Unit 2 and Unit 3 at rates
of 198 mg/L and 698 mg/L,  respectively.
Single doses of PAC (as a slurry) were
added to Unit 2 and Unit 3 each day for
the first 4 days to reach the calculated
equilibrium PAC concentration, and the
operation of all three units was continued
for an additional 35 days.
   Phase 3  -  First Intensive   Test
Period—P\\ot plant  operation continued
for 14 days with Phase 2  operational
conditions. Intensive sampling  was
initiated that included the collection of
aquatic bioassay and organic  samples.
Four samples for carbon isotherm testing
were also collected from the control unit
effluent. Daily settling and oxygen uptake
tests were performed  on each of the pilot
plant mixed liquors.
   Phase  4  -  Second  Acclimation
Period—This  phase  lasted 31  days
and  was  used to increase the   PAC
concentrations  in the  aeration basins for
Unit  2  and Unit 3.  The original  PAC
doses  were increased to 396 mg/L  and
1,190  mg/L of  raw  feed,  respectively.
Dver the first four days, additional PAC
was  added in  slurry  form to  each
 teration basin  so as to  achieve an
jquilibrium  PAC concentration  by  the
ourth day.
   Phase  5 -  Second  Intensive   Test
aeriod-P\\ot plant  operation continued
or 14 days with Phase  4  operational
;onditions. Intensive  testing   was
jerformed  similar to  that described
jreviously  for the first intensive test
period.
Selector Study
   The selector pilot unit was operated
over the same time period as the PACT
systems. Unit 1 was used as the control
while  Unit  4 was the experimental
selector system.

System Operation--
   The selector system  was identical to
the control system with the  exception
that a series of selector tanks preceded
the aeration basin.  Initially, the selector
system  consisted of one 7.6 L (2 gal)
anoxic selector tank and two aerated 7.6
L  (2  gal) selector tanks  operated  in
series.  The return sludge  and  feed
wastewater were mixed in the first anoxic
tank and flowed  by gravity through the
two aerated tanks and  into the 189 L (50
gal) aeration tank. The aerobic selectors
were  equipped  with an  air supply and
metering  valves.  During   the  second
phase of the selector study, a third 7.6 L
(2 gal) aerated selector tank was added.
The HRT of each tank was approximately
2 hr based on the raw waste flow.

Test Phases--
   Due to the long time required to reach
equilibrium, only two test conditions were
studied.
   Phase I - The  selector  system  was
seeded  with mixed liquor from the full-
scale plant.  The system operation  with
respect to sludge settling was monitored
by  sludge settling tests.  The stirred
sludge volume  index (SSVI)  was
calculated  for  the  settling tests.
Operation in this mode was continued for
75 days.
   Phase II  - When the data  indicated
the mixed liquor for the selector unit was
not settling much better  than the control
unit, a third aerated selector basin, for a
total of four  selector basins, was added.
The  system was  operated  with  four
basins for the remainder  of the study (26
days).

VFMLS Study
   It was postulated that the cause of the
VFMLS  during the  1984 study was due
to the presence of a material in  the
wastewater that was adsorbable and not
readily  biodegradable.  This  material
might have accumulated on the carbon
until  it reached a concentration where  it
began to act like a high molecular weight
polyelectrolyte  and caused the  MLSS
and  carbon  to flocculate and  form  the
VFMLS.
   To test this  hypothesis, preliminary
bench tests  were conducted in January
at the home laboratory  and more tests
were conducted on-site  during the  pilot
plant study. The basic procedure was to
add spike materials to mixed liquor from
the PACT pilot systems (or mixed liquor
from  a conventional activated  sludge
system  with  PAC  added  for the
preliminary tests), aerate the mixture,
allow  the mixed liquor  to settle, and
observe the results. This  sequence was
then repeated once every 24 hr for the
on-site  tests. A total of  eight  spike
materials  were  tested  at  various
concentrations.

Feedwater Quality
   Table  1 summarizes  the raw  waste
characteristics. The raw waste TCOD was
quite   variable,  ranging  from  3,070  to
8,080 mg/L, and it averaged 5,230 mg/L
for 64 samples. This average TCOD was
significantly less than the  average  TCOD
of the raw wastewater during  the 1984
study  (7,030  mg/L).  Nitrogen and
phosphorous  were present in  adequate
amounts  for  biological treatment  based
on the  effluent concentrations  of
phosphorous,   ammonia,   and
nitrate + nitrite.
Table 1. Summary of Raw
Characteristics
Parameter
TSS, mgIL
TCOD, mg/L
SCOD, mg/L
NIBOD, mg/L
NIS8OD, mg/L
pH, units
Temperature, °C
Average
396 (92)"
5230 (64)
3820 (66)
2700(21)
2200 (20)
-
25.5(112)
Wastewater
Range
90-722
3070-8080
1550-5850
1700-3800
1200-3300
5.9-12.8
(159)
18.0-31.0
"Parentheses  indicate  number  of
observations.
PACT Study
   The operational and  analytical  data
from the  first and second  intensive test
phases  are summarized  in  Table  2.
During Phase 3, the average HRT and
average SRT values were  very close to
the target values and ranged from 2.18 to
2.20  days  and  10.0 to  10.1  days,
respectively. The average temperature in
the aeration basins ranged from 35.0 to
36.5°C while the pH ranged from 6.4 to
8.0. The  effluent  TCOD concentrations
averaged 654,  413, and  344 mg/L for
Units  1, 2, and  3,  respectively, while the
effluent SCOD  concentrations averaged
498, 364, and 220 mg/L, respectively.
   The average HRT  and SRT values
ranged from 2.18 to 2.20 days and 10.0
to 10.4 days, respectively,  during Phase
5. The pH ranged from 7.3 to 8.1  and the

-------
                   Table 2.  Average Operational Parameters and Effluent Quality During the Intensive Test Periods

                                                        Phase 3                   Phase 5
                                                       Day 43-56                  Day 88-101
                                                 Unit 1
                 Unit 2    Unit 3
               Unit 1
               Unit 2    Unit3
Operation Data
Feed Flow, mUm'm
PAC Dosage, mgIL Feed-
Return Sludge Flow, mUmin
Waste Mixed Liquor Flow, Uday
F/M, g TCOD/dayg MLSS"
HRT, days
SRT, days

60.0
0
60.1
17.7
0.58
2.19
10.0

60.3
198
63.5
18.6
-
2.18
10.0

59.9
698
61.2
15.6
-
2.20
10.1

59.8
0
176
17.7
0.54
2.20
10.4

60.2
396
203
18.2
-
2.18
10.1

60.3
1190
59.5
18.7
-
2.18
10.0
                    Mixed Liquor Data

                    Temp.,°C
                    pH (range)
                    Dissolved Oxygen, mgIL
                    TSS, mg/L
                    SSVI, mUg"

                    Effluent Data

                    TCOD, mg/L
                    SCOD, mg/L
                    NIBOD, mg/L
                    NISBOD, mgIL
                    rSS, mg/L
                    VSS, mg/L

                    Removals

                    Average Feed TCOD, mg/L
                    Average Feed SCOD, mg/L
                    TCOD Removal, %
                    SCOD Removal, %
36.5
6.5-8.0
2.7
4030
229
35.0
6.4-7.8
2.4
5620
144
35.0
6.5-7.7
2.0
8050
58.3
34.5
7.4-8.1
2.5
4870
194
35.0
7.3-8.0
2.2
7780
107
36.0
7.3-8.0
1.7
11600
53.0
           654
           498
            11
             6
            77
            63
413
364
  6
 <4
 20
 17
          5030     5030
          3480     3480
           86.7     91.6
           85.5     89.4
344
220
  9
  6
728
123
        5030
        3480
        92.9
        93.6
532
449
 13
  4
 39
 34
        5690
        4210
        90.5
        89.1
244
210
  7
 <3
 23
 22
164
127
<4
<3
 20
 18
        5690     5690
        4210     4210
        95.6      97.1
        94.9      96.9
                    "Food to microorganism ratio.
                   "Stirred sludge volume index.
                    -Not determined.
average temperature ranged from 34.5 to
36.0°C in the  aeration  basins over  the
Phase  5  test period. Average effluent
TCOD  concentrations for Phase  5 were
532, 244, and 164  mg/L, for  Units 1, 2,
and 3,  respectively, and average  effluent
SCOD concentrations were 449, 210, and
127 mg/L, respectively.
   The addition of PAC resulted in some
improvement in MLSS settling rate as the
SSVI data in Table 2 indicates. However,
problems with  floating  sludge  in  the
clarifiers,  especially   Unit 2,  were
encountered. The large PAC dosages
added  to  Unit 3 resulted in a consistently
good settling sludge.

rCOO and SCOD Removal
   The addition  of PAC to  the activated
sludge process improved the removal of
both TCOD and SCOD. The comparison
of the two intensive test  phases  was
complicated  by  the  different  raw
wastewater characteristics for the  two
phases. The influent TCOD  and SCOD
values were greater for Phase  5  than
Phase  3 but the control  unit  effluent
TCOD  and SCOD  concentrations  were
less during Phase 5 than Phase 3. Thus,
the Phase  3 698 mg/L PACT unit
average  effluent SCOD  concentration
(220 mg/L) was slightly greater than the
Phase 5 396 mg/L  PACT effluent SCOD
concentration (210  mg/L). The average
SCOD  removal percentage  of  the  396
mg/L  PACT unit  (94.9%)  was also
slightly greater than the 698 mg/L PACT
unit (93.6%).
   Carbon adsorption  isotherm  tests
were  performed on  four  control  unit
effluent  samples from each of the two
intensive test  phases. All eight of the
tests were combined to produce a single
isotherm plot. The line of best fit for the
data is as follows:
                           M = PAC dosage, g/L
                         Ce  = equilibrium  effluent SC<
                                concentration, mg/L

                       The  SCOD  data  from  the PA
                    systems can also be  presented or
                    Freundlich plot. It was assumed that
                    additional SCOD removal by the PA
                    systems compared to the control wh
                    was  operated  in  parallel  vv
                    accomplished by the PAC. This allowe
                    calculation  of  the amount  of SC
                    removed per unit of carbon added in
                    SCOD per gram of PAC. The data  <
                    evaluated with the resulting line of t
                    fit:

                                                   (2)
       — =(0.80)CU4
       M           e
                                d)
where:
       X =  amount of SCOD removal
            attributed to the PAC, mg/L
                       Figure  1  compares  the  isothc
                     derived from the pilot plant results
                     that derived  from the jar tests. In
                     case both isotherms are very sim
                     which suggests that isotherms deriv
                     from the jar tests would provide a £
                     prediction  of the PAC requirement:

-------
Combined Jar Tests

  X _

 M
                                                 Carbon Adsorption Isotherm
                                                   Freundlich Equations
                                                   Pilot Plant Results
                                                                              M
                        10000
                          1000
                     g
                     1
                     8
                     r
                     I
100
                           10
               f lot Plant Data
                 •+•

                                                                4


                                      Combined Jar Test Results .
                                                              11
                             10
                      100                 1000
                           Ce (SCOD), mg/L
                                                                                       10000
                    Figure 1.    Comparison of the isotherm derived from the pilot plant results with that derived
                                from the combined jar tests.
produce any desired effluent  SCOD
joncentration (down to Ce - \27 mg/L).
"he similarity between the adsorption
sotherms also  suggests that there was
10 enhanced biodegradation of refractory
;ompounds in the PACT systems.

Specific Organic Compounds
Removal
  Based on preliminary samples of raw
/astewater and the plant raw product list,
ight compounds  were  selected for
uantification using GC/MS  techniques
)r intensive test samples. In addition,
omputer searches to  tentatively identify
              and quantify  all  detected peaks  were
              completed. Of the six volatile compounds
              analyzed for (total  xylenes,  ethyl
              benzene,  acetone,  acrylonitrile,
              methylene chloride,  and ethyl  acetate),
              only two effluent samples and one waste
              mixed  liquor  sample  contained
              quantifiable amounts of any of the  six
              compounds. No differences between the
              control and PACT units with respect to
              removal of  these  compounds  or
              concentration of these compounds in the
              waste sludges were found.
                 Two semivolatile  compounds, alpha-
              picoline and  dicyclohexylamine,  were
              specifically quantified in all  samples.
                                                      Alpha-picoline  was  found  more
                                                      frequently in Unit 1  (7 of 8) waste mixed
                                                      liquor samples than Unit 2 (6 of 8) or Unit
                                                      3 (4 of 8) and in  higher concentrations.
                                                      However,  with  the  exception  of  one
                                                      control  sample with  1,600 ng/L and one
                                                      control  sample with 84  ng/L all samples
                                                      were below 50 pg/L. Alpha-picoline was
                                                      found in 7  of  8 control  unit  effluent
                                                      samples but not in  any  effluent samples
                                                      from either PACT  unit. Thus alpha-
                                                      picoline was slightly higher in the control
                                                      effluent samples as well  as the waste
                                                      mixed liquor samples. All of the control
                                                      effluent samples contained less  than 30
                                                            of  alpha-picoline,  however.

-------
Dicyclohexylamine was found in 7 of 12
waste mixed  liquor  samples  from the
second intensive test period but was not
measured in any samples during the first
period.  The  concentrations  were
somewhat higher (up to 700 ng/L) in the
control samples than the PACT samples
(up to 150 pg/L). Dicyclohexylamine was
only found in one effluent sample (Unit 1
at IIOng/L).
   A wide variety of compounds  were
identified in the computer searches of
the effluent and  waste  mixed  liquor
samples.  However,  there   was  no
apparent  relationship between  PAC
concentration  and tentatively  identified
volatile organic compounds in  either the
effluents or waste mixed liquors.
   The TCLP analysis of the waste mixed
liquors determined  that  none of the
samples analyzed would be classified as
toxic as  defined under  proposed  U.S
EPA  hazardous  waste  identification
regulations  (40 Code  of   Federal
Regulations  261.24).  No   organic
compounds were detected in any of the
samples.

Aquatic Toxicity Removal
   Bioassays  of the samples  from the
first intensive test period did  not  show
any  significant difference in toxicity
removal between the control and PACT
units. The  average acute  (C. dub/a)
toxicity  (medium lethal concentration,
LCgo) of the effluent samples was  46%,
51%, and 47% for Unit 1, Unit 2 and Unit
3, respectively, while the average toxicity
of the raw waste was 0.81%. There was
also no obvious differences between the
chronic toxicity of the various pilot plant
effluent samples.
   The average LCso of the raw  waste for
Phase 5 was 1.0%, which was similar to
the Phase 3 LC$Q. The effluent samples
from the second  intensive  test period,
however, were more toxic than from the
first test period. The average  LCgo
values from Unit 1,  Unit  2, and Unit 3
were 14%,  5.8%, and 0.82%.  The data
for Phase  5  indicate  that  the toxicity
appeared to  increase  with  increasing
PAC concentrations.  Only limited chronic
toxicity  data  were  obtained due to
problems experienced  with the bioassay
test procedure.

Selector  Study
   The selector  unit was operated to
determine if improvements in the settling
characteristics of the mixed liquor  could
be achieved. The  selector unit  mixed
liquor  showed some improvement in
settling  characteristics as measured by
the SSVI test when compared to the
control over  the  entire Phase I test
period. The  average  SSVI values  over
Phase I were 222 and 163 mL/g for Unit
1 (control unit) and Unit 4 (selector unit),
respectively.  The selector  unit mixed
liquor SSVI,  however,  was equal to or
greater than  the control unit values  for
most of Phase II. During Phase II, Unit 1
and Unit 4 SSVI's averaged 194 and 208
mL/gm, respectively. The most obvious
difference between  the two  phases
(besides the  additional  selector  basin
added for Phase  II) was that  the  DO
concentrations  in the  selector  basins
were generally lower during  Phase II as
compared  to Phase I.  This observation,
along  with  the   presence   of
Haliscomerobacter hydrossis found in
the microscopic examination, suggests
that the lack of DO in the selector basins
was the cause of the poor settleability of
the selector  mixed liquor. Pure oxygen
was used for the last 6 days of operation
of  Phase  II,  and  the SSVI  tests
performed during  this  period indicated
some improvement  in settling  of  the
mixed liquor.

VFMLS Study
   None of the bench tests completed in
January, May,  or June  1987  resulted in
the formation  of a VFMLS.  Either  the
individual spike materials tested were not
the cause  of the original VFMLS or the
proper conditions for the formation of the
material were not simulated in the bench
tests. The task of  recreating the VFMLS
was  made difficult by  the fact  that the
product mix at the plant changes, which
results  in   variable wastewater
characteristics. It is  possible  that  a
unique pharmaceutical wastewater matrix
existed during the 1984 study  that did
not exist during this study.

Conclusions
   The pilot  study project conclusions
are summarized below:

  1. Efforts to  identify  the cause of the
    VFMLS that formed during the 1984
    study were unsuccessful.
  2. Effluent  SCOD concentrations  were
    significantly reduced by  the addition
    of PAC   to the activated sludge
    system  when compared to  the
    control unit. For  the first intensive
    test  phase,  effluent  SCOD
    concentrations were reduced  by
    26.9% and 55.8% compared  to the
    control unit effluent SCOD  when
    adding 198 mg/L  and 698  mg/L of
    PAC,  respectively. Effluent SCOD
    concentrations  were  53.2% and
    71.7% lower  than  the control  unit
  value  when  396 mg/L and  1.19(
  mg/L   of  PAC  were  added
  respectively.
3. Isotherms derived  from this plant';
  wastewater treatment  plant effluen
  are a  good  predictor  of  the  PAC
  requirements to produce any desirei
  effluent SCOD concentration (at leas
  down to Ce  =  127 mg/L) by thi
  following equation:
          M
             = (0.80)CL14
4.  PAC addition  improved  the  settlin
   characteristics  of  the  MLSS
   However,  all  three  system
   experienced  floating  and  bulkin
   sludge during the study.
5.  There was no significant difference i
   the removals of acetone, acrylonitrih
   methylene  chloride,  ethyl  acetati
   ethyl  benzene, total xylenes,  alphi
   picoline,  and  dicyclohexylamin
   between the  control  unit and tf
   PACT units.
6.  Effluent aquatic toxicity  was n<
   reduced by the PACT systems ar
   in some cases  the PACT effluen
   were apparently more  toxic than tt
   control unit effluents as measured t
   the acute test.  The effluents  we
   substantially  more  toxic  for tt
   second intensive test  period  i
   compared to  the first as measun
   by the acute toxicity test.
7.  The  study  did  not  show ai
   correlation of COD  with  aqua'
   toxicity, BOD,  or specific identifial
   organic   compounds  in  th
   pharmaceutical wastewater.
8.  The TCLP analyses  of  the  cont
   and two PAC units'  mixed  liqi
   samples  did   not  identify a
   contaminants  above the regulat
   limits.
9.  The selector unit mixed liquor settl
   somewhat better than the control i
   mixed liquor at certain times  but :
   exhibited filamentous  bulki
   characteristics. The presence of
   filamentous organism  Haliscome
   bacter  hydrossis in  one selec
   mixed liquor sample collected dur
   the  second  intensive test  per
   suggested  there was  a  shortage
   dissolved  oxygen in the  selec
   basins. A  shortage  of dissoh
   oxygen may have  contributed to
   failure  of  the  selector system
   control the growth of  filament
   organisms during this period of tir

-------
    The full  report was submitted  in
fulfillment  of  Contract  Number 68-03-
3371 by  Radian Corporation under the
sponsorship  of  the  U.S.  Environmental
Protection  Agency.

-------
  David A.  Gardner and  Richard A.  Osantowski are with Radian Corporation,
    Milwaukee. Wl 53214.
  Clyde Dempsey  is the EPA Project Officer (see below).
  The complete report consists of two volumes, entitled "Evaluation of Biological
    Treatment of Pharmaceutical Waste wafer with PAC Addition:"
    Volume I: (Order No. PB 88-212 527/AS; Cost: $25.95, subject to change)
    Volume II: (Order No. PB 88-212 5351 AS; Cost $32.95, subject to change
  The above reports will be available only from:
           National Technical Information Service
           5285 Port Royal Road
           Springfield, VA 22161
           Telephone:  703-487-4650
  The EPA Project Officer can be contacted at:
           Water Engineering Research Laboratory
           U.S. Environmental Protection Agency
           Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
     BULK RATE
POSTAGE & FEES PAID
         EPA
  PERMIT No. G-35
Official Business
Penalty for Private Use $300

EPA/600/S2-88/032
                           0000329    PS

                           U S EimR PROTECTION  ASfHCY
                           REGION  5  LISRARt
                           230 S OIARftCRK  STREET
                           CHICAGO              It   60404

-------