United States
Environmental Protection
Agency
Robert S. Kerr Environmental Research
Laboratory
Ada OK 74820
EPA-600/2-79-172
August 1979
Research and Development
&EPA
Biological Treatment
of High Strength
Petrochemical
Wastewater
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
' 1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-79-172
August 1979
BIOLOGICAL TREATMENT OF HIGH STRENGTH
PETROCHEMICAL WASTEWATER
by
William J. Humphrey
Enrique R. Witt
Celanese Chemical Company, Inc,
Technical Center
Corpus Christi, Texas 78408
Joseph F. Malina, Jr.
Project Consultant
Austin, Texas 78703
Grant No. 12020 EPH
Project Officer
Thomas E. Short, Jr.
Source Management Branch
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ADA, OKLAHOMA 74820
!;. ,\ ft
230 South Dearborn Street
Chicago, Illinois 60604
-------
DISCLAIMER
This report has been reviewed by the Office of Research and Develop-
ment, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recom-
mendation for use.
U.S. Environmental Protection Agency
-------
FOREWORD
The Environmental Protection Agency was established to
coordinate administration of the major Federal programs designed
to protect the quality of our environment.
An important part of the Agency's effort involves the search
for information about environmental problems, management tech-
niques and new technologies through which optimum use of the
nation's land and water resources can be assured and the threat
pollution poses to the welfare of the American people can be
minimized.
EPA's Office of Research and Development conducts this
search through a nationwide network of research facilities.
As one of these facilities, the Robert S. Kerr Environmental
Research Laboratory is responsible for the management of programs
to: (a) investigate the nature, transport, fate and management of
pollutants in ground water; (b) develop and demonstrate methods
for treating wastewaters with soil and other natural systems;
(c) develop and demonstrate pollution control technologies for
irrigation return flows; (d) develop and demonstrate pollution
control technologies for animal production wastes; (e) develop
and demonstrate technologies to prevent, control, or abate pollu-
tion from the petroleum refining and petrochemical industries;
and (f) develop and demonstrate technologies to manage pollution
resulting from combinations of industrial wastewaters or indus-
trial/municipal wastewaters.
This report contributes to the knowledge essential if the
EPA is to meet the requirements of environmental laws that it
establish and enforce pollution control standards which are
reasonable, cost effective and provide adequate protection for
the American public.
W. C. Galegar
Director
Robert S. Kerr Environmental Research Laboratory
ill
-------
ABSTRACT
The biological treatment of a complex petrochemical wastewater
containing high concentrations of organic chlorides, nitrates and amines
was initially studied using a sequence of anaerobic methanogenesis and
oxygen activated sludge. Bench scale and pilot plant treatability studies
were conducted using various composite samples and process wastewater
blends. The results of preliminary studies indicated the need for stream
segregation and waste reduction programs at the petrochemical plant.
Segregation of components of the combined plant waste stream was re-
quired to eliminate nonbiodegradable materials and pretreatment minimized
the concentration of substances which were toxic or inhibitory to biological
treatment.
Nitrates inhibited methanogenesis in the anaerobic system but
quantitative removal of nitrates was accomplished. Only partial removal
of chemical oxygen demand (COD) was achieved during anaerobic denitrifi-
cation because of the relatively low nitrate/COD ratio. Anaerobic
methanogenic treatment also was unsuccessful in reducing the COD
concentration to any great extent, even after pretreatment by anaerobic
denitr if ication.
The activated sludge system was effective in removing the
biodegradable portion expressed as biochemical oxygen demand (BOD)
of the pretreated combined wastewater stream; but the yellow color of the
effluent was unacceptably dark. The activated sludge system performed
equally well when high purity oxygen, or air was used for aeration.
Therefore, the final treatment sequence included anaerobic
denitrification, and activated sludge using air for aeration and produced
an effluent with the following characteristics; essentially no nitrates,
BOD ~ 50 mg/1, COD ~ 1200 mg/1, suspended solids 200 mg/1, and yellow
in color. The average influent composition was BOD ~ 6000 mg/1, COD
~ 8000 mg/1, nitrates ~ 1000 mg/1, and yellow in color.
This effluent does not meet the limitations imposed for discharge
to surface waters.
IV
-------
This report was submitted in fulfillment of Project 12020 EPH by
Celanese Chemical Company under the partial sponsorship of the U.S.
Environmental Protection Agency. This report covers a period from
September 23, 1969 to February 20, 1976 and work completed July 20,
1978.
-------
CONTENTS
Foreword iii
Abstract iv
Figures lx
Tables xi
1. Introduction 1
2. Conclusions 5
3. Waste Characteristics and Process Selection 7
A. Stream Segregation 12
B. Waste Reduction Program 14
4. Experimental and Analytical Procedures 16
A. Bench Scale Studies 16
B. Pilot Plant 20
1. Stage I 20
2. Stage II 20
3. Stage III 20
C. Analytical Methods 24
5. Results and Discussion of Glassware and Pilot Plant Studies 25
A. Bench Scale Studies 25
B. Bay City Pilot Plant Studies 25
1. De nitrification 26
2. Methanogenic Unit 27
3. Activated Sludge 29
C. Treatability Studies 30
1. Aerobic Treatment 30
2. Anaerobic Systems 31
a. De nitrification 31
b. Treatment of Composite C to Methanogenesis 36
c. Process Limitations and Design Criteria
for Denitrification 37
continued
vii
-------
6. Capabilities of Proposed Treatment Sequence 43
A. Characteristics of the Composites 43
1. Denitrification of the Composites 43
2. Heavy Metal Composite 46
3. Aerobic Composite 46
B. Treatment of Performance 47
1. Denitrification 47
a. Packed Bed Unit 47
b. Fluidized Bed Reactor 48
2. Methanogenesis 49
3. Aerobic Treatment 49
a. Activated Sludge Treatment 49
b. Extended Aeration of Effluent from
Activated Sludge Treatment 52
Appendices
A. Stream Segregation 55
B. Waste Reduction 61
1. Chlorinated Organics 61
2. High Carbon Concentration Wastewater 61
3. Volatile Light Ends 61
4. Heavy Metals 62
5. Amines 62
6. Nitrates 62
C. Description of Bench Scale Reactors 63
1. Aerobic Biological Treatment (Oxygen) 63
2. Anaerobic Methanogenic Treatment 65
3. Anaerobic Denitrification 72
4. Pretreatment 74
D. Aerobic Treatability Performance .Data 76
E. Denitrification in Packed Beds 82
F. Anaerobic Methanogenic Treatment 87
G. Anaerobic Denitrification Fluidized Bed 95
Vlll
-------
FIGURES
Number Pa§e
1 Sequence of Events in Evolution of Project 3
2 Daily Variation in Volume of Plant Effluent Make 11
3 Process Schematic Evolution 13
4 Proposed Sequence 15
5 Initial Bay City Laboratory Studies (October, 1971) 17
6 Initial Celanese Chemical Company Technical Center
Laboratory Studies 18
7 Feed Compositing - Pretreatment and Biological Treatment 19
8 Pilot Plant Configurations 22
9 Pilot Plant Schematic 23
10 Denitrification - Anaerobic - Aerobic Treatment of
Bay City Effluent 32
11 Acid-Base Relationships in the Synthetic Substrate Used
for Denitrification Studies 42
12 Treatment Sequence of Bay City Plant Effluents 44
13 Alternative Treatment Sequence of Bay City Plant Effluents 45
C- 1 First Stage of Two-Stage Bench Scale Oxygen-Aerated
Activated Sludge Unit 64
C-2 Continuous Simulation Reactors 66
continued
ix
-------
Number Page
C- 3 Anaerobic Packed Bed, First Design 69
C-4 Backmixed Anaerobic Filter, Final Design 71
C-5 Fluidized Bed Denitrification Anaerobic Filter 73
E- 1 Packed Bed Denitrification Unit 83
E-2 Packed Bed Denitrification Unit 85
F- 1 Operating Parameters for Backmixed Anaerobic Filter 87
F-2 Operating Parameters for Backmixed Anaerobic Filter 91
G- 1 Anaerobic Denitrification Fluidized Bed 95
G-2 Anaerobic Denitrification Fluidized Bed 96
G-3 Anaerobic Denitrification Fluidized Bed 97
G-4 Anaerobic Denitrification Fluidized Bed 98
G-5 Anaerobic Denitrification Fluidized Bed 99
-------
TABLES
Number Page
1 Composition of Combined Process Wastewater, Bay
City Plant 8
2 Composition of the Process Wastewater Streams at the
Bay City Plant Before Segregation 9
3 Variations in Composition of Combined Process Waste-
Water, Bay City Plant 10
4 Summary of Pilot Plant Data 28
5 Composite of Nitrate-Containing Bay City Wastes,
Composite B 33
6 Bay City Composite, Carbon-Only Wastes, Composite C 34
7 Composition of Synthetic Denitrification Effluent 39
8 Fractional Distillation of Final Bay City Plant Effluent 51
9 Distillation of Fraction- 1 53
10 Analysis of Bay City Plant Extended Aerated Effluent 54
A- 1 Relative Aerobic Biodegradation Rates for Specific
Compounds in the Process Wastewater 56
A-2 Stream Biodegradability Tests 57
A- 3 Soluble Nitrogen Requirements for Waste Treatment 58
A-4 Composition of Combined Process Wastewater at Bay
City Plant After Elimination of Low Volume High
Carbon Stream 59
continued
XI
-------
Number -n
Pa§e
C- 1 Synthetic Starter Feed for Anaerobic Filters 68
D- 1 Composition of Bay City "Aerobic" Composite (Free
of Nitrate- and Organochlorine- Containing Wastes) 76
D-2 Activated Sludge Treatment of Bay City Effluent,
First Stage 77
D- 3 Activated Sludge Treatment of Bay City Effluent,
Second Stage on
E- 1 Composition of Bay City Nitrate-Containing Effluent
Composite, Analytical Summary 82
Xli
-------
SECTION 1
INTRODUCTION
The Bay City plant of the Celanese Chemical Company located in
Matagorda County, Texas produces a wide variety of petrochemicals.
Waste-waters generated within the plant contain significant quantities of
nitrates, amines, chlorinated organics, organic carbon, cyclic compounds,
high molecular weight polymers, and heavy metals. In the early years of
plant operation, the wastewater treatment and disposal scheme consisted
of several hundred acres of waste stabilization and evaporation ponds. The
organic loads to the ponds exceeded the design loading and the facultative
ponds became anaerobic resulting in the production of odors. This
anaerobic biological process indicated that the anaerobic treatment had been
effective in reducing the COD (chemical oxygen demand) by more than 90%.
However, the anaerobic samples were taken as portions of large pond areas
which did not represent actual effectiveness of the entire shallow anaerobic
ponds and the results were not duplicated in laboratory anaerobic studies.
Initial treatability studies of the process wastewater s in laboratory
activated sludge units resulted in less than 50% removal of the COD
concentration. These early observations indicated that some combination
of anaerobic and aerobic processes could possibly produce a high quality
effluent treating a complex petrochemical wastewater. Therefore, bench
scale and pilot scale studies involving combinations of anaerobic and aerobic
biological processes were initiated at the Bay City plant and the Corpus
Christi Technical Center of the Celanese Chemical Company under the
partial support of a demonstration grant funded by the Federal Water
Pollution Control Administration (FWPCA).
The objectives of the studies included: (a) the investigation of the
anaerobic/aerobic treatability of high strength petrochemical waste to
produce a water of reusable quality or suitable for discharge; (b) investi-
gation of the effects of high purity oxygen on the aerobic process, specifi-
cally the activated sludge system; (c) demonstration of the total process
in a pilot scale system; and (d) comparison of the economics of the anaerobic/
aerobic system with injection wells.
-------
The project was divided into three phases. The first two phases
overlapped. Phase I included benchscale studies conducted at the Celanese
Chemical Company Technical Center in Corpus Christi and at the Bay City
plant. Phase II included the design, construction and operation of a pilot
plant located at the Bay City plant using continuous flow of process waste-
waters. The development of design parameters and definition of effluent
quality attainable were completed in benchscale studies conducted at the
Technical Center as Phase III. The sequence of events through which the
project evolved are illustrated in Figure 1.
Several observations made during the early stages of Phase I helped
to define the process sequence. Denitrification occurred under anaerobic
conditions resulting in reduction in nitrates from about 300 mg/1 (milligrams
per liter) to less than 5 mg/1 with a corresponding reduction in organic
carbon. Some of the wastewater streams which were inhibitory or toxic
to the anaerobic system were quite readily degradable under aerobic
conditions. Other streams which were not amenable to aerobic biodegra-
dation, however, were treatable in the anaerobic systems. Therefore,
the wastewater streams at the Bay City plant were segregated into three
components, namely the high nitrate containing streams which were
amenable to denitrification, wastewaters which were anaerobically degraded
and those which could be treated in an aerobic activated sludge system.
The initial biological system used in the bench scale and pilot plant
studies included anaerobic denitrification followed by anaerobic methane
production, and finally followed by treatment in the aerobic activated sludge
system. Bench scale units were operated at the Bay City plant as well as
at the Technical Center. The denitrification system operated with a packed
bed as well as the fluidized bed containing either sand or activated carbon.
Anaerobic methanogenic units were submerged packed beds. The aerobic
process was a complete mixed activated sludge system. Individual waste-
water streams and various composites of different process waste streams
were fed to each of the individual units. Composite samples were routinely
collected at the Bay City plant and shipped to the Technical Center. The
composition of these composite samples varied as the project progressed
reflecting changes in the wastewater characteristics resulting from in-
plant waste reduction, waste segregation and proposed pretreatment of
specific waste streams prior to biological treatment.
The pilot plant (Phase II) constructed at the Bay City plant included
two parallel systems to minimize downtime resulting from potential bacte-
rial inactivation caused by shock loads of specific components in the waste
stream. The pilot plant was designed to be portable. This placed some
constraints on the selection of equipment. Each of the pilot plants consisted
of a three-process system including submerged packed bed denitrification
-------
Jl
Grant
Started
1/6/71
me Jan June
71 72 72
— _ 1 1
Lab Studies
ANA-AER
i
Design % Build
Pilot Plant
-4
Jan June
73 73
— " '""-•• • i
Operate Pilot Plant
ANA-AER -DEN
Develop Instrumenta-
tion
Jan June
74 74
_ Waste Survey ,^_
-* <^
--Sources and
Composition
- -Pretreatrnent
Study
--Stream Segregation
Jan Ju
75 7
Sample
Collection
P ret r eatment
and
Composition
Samples
ne
5
BAY CITY
PLANT
Oxygen
Activated
Sludge
Started
Prior to
Grant
PHASE I
AER- Oxygen
Laboratory Studies
--^ ».
Jr'HAbi; 11 . i-rttt.0£. 110. ,
•* ++* ™
Laboratory Studies
ANA -DEN
Lab
Studies
DEN
Synthetic
on Plant Composites Blanes
Laboratory Study
DEN- ANA- DEN
Develop Design
Data
Define
CELANESE
CHEMICAL
(T»\/rPAT>JV
^^/iVLt fa\ i
TECHNICAL
CENTER
Effluent Quality
LEGEND ANA - Anaerobic Methanogenesis
AER - Aerobic Activated Sludge
DEN - Denitrification
Figure 1. Sequence of events in evolution of project
-------
unit, a submerged packed bed anaerobic methanogenic unit and completely
mixed activated sludge system with a separate clarifier. The activated
sludge system was designed to be operated with high purity oxygen using
a down flow bubble contact aerator for oxygen transfer and auxiliary mixing.
A wastewater storage tank was used to blend the feed for the pilot plant. On-
line instrumentation also was included in the pilot plant to continuously
monitor dissolved oxygen, organic carbon, and copper concentration. The
pilot plant design was based on data from limited bench scale studies. As
a result, pilot plant operations were unstable and produced unsatisfactory
results.
-------
SECTION 2
CONCLUSIONS
1. Effective biological treatment of a complex petrochemical wastewater
requires waste reduction and stream segregation to eliminate non-
biodegradable wastewater streams. Pretreatment is essential for
removal of heavy metals and other compounds -which may be toxic to
or inhibit biological treatment processes.
2. Nitrates inhibit methane production in anaerobic treatment systems.
However, nitrates are quantitatively converted to nitrogen gas during
anaerobic denitrification resulting in the oxidation of 2. 85 parts of COD
for each part of nitrate N reduced.
3. A treatment sequence including anaerobic denitrification and activated
sludge produced an effluent which contained essentially no nitrates
(100% removal), BOD ~ 50 mg/1 (> 99% removal), COD ~ 1200 mg/1
(~ 85% removal), suspended solids ~ 200 mg/1, and yellow in color.
This effluent does not meet the limitations imposed for discharge to
surface water.
4. Data obtained from glassware (bench scale) units simulate the
biological and chemical processes well and provide useful performance
data for various operating conditions. The major difficulty is with
hydraulic scale-up.
5. The design of a pilot plant should be based on complete glassware
(bench scale) data which define process performance and the variables
affecting process performance.
6. Reliable instrumentation which continuously monitor flows and other
parameters which characterize the various influent and effluent
streams is essential to pilot plant operations. However, during the
pilot plant phase of this study such instrumentation was not available
and a considerable effort was directed toward improving on-line
instrumentation to monitor the characteristics of wastewater
streams.
-------
The results of these studies indicate that the activated sludge process
performed equally well when air or high purity oxygen -was used for
aeration. Therefore, the choice of high purity oxygen versus air is
an economic standoff that should be evaluated for each particular
application.
-------
SECTION 3
WASTE CHARACTERISTICS AND PROCESS SELECTION
The Celanese Chemical Company Bay City plant produces a wide
variety of petrochemicals including acetaldehyde, acetic acid, vinyl acetate,
butanol, butyraldehyde, crotonaldehyde, nitric acid, adipic acid, cyclo-
hexanone, hexamethylenediamine, 1, 6-hexanediol, nylon salt and several
other minor by-products. Organic compounds including cyclic organics
and high molecular weight polymers contribute to the high concentrations
of total organic carbon. Nitrogen is present as nitrates, nitrites and
amines. Heavy metals, primarily copper, with small amounts of nickel,
chromium, cobalt and vanadium may be found. The overall composition
of the process wastewaters is presented in Table 1. The combined waste-
water is made up of numerous streams; however, six major contributing
streams are identified in Table 2. The flow rate percent of total flow and
the total organic carbon contribution of the individual streams are also
presented in Table 2. It is interesting to note that Stream 3 contributes
only 8. 6% of the total flow but contains 75.2% of the total organic carbon.
The data presented in Tables 1 and 2 represent only gross averages of the
composition of the total process wastewater, as well as the average of the
flows and total organic carbon of the six individual streams.
The composition and flow vary during the course of the day and from
day to day. Analysis for metals indicate a range of copper concentration
from less than 0. 5 mg/1 to as high as 17 mg/1, although surges of heavy
metals in excess of 100 m.g/1 have been recorded. These surges of heavy
metals usually occur during the shutdown and clean out of a process
reactor. The variability in the composition of the combined process waste-
water is also illustrated in the data presented in Table 3. Although the
pH indicates that the wastewater is primarily acidic, alkaline values of pH
occur. The total organic carbon concentration varied from 1, 200 to
22, 000 mg/1. The average total organic carbon (TOC) for the 25 samples
reported is 15, 680 mg/1.
The variations in the flow of the combined process wastewater is
illustrated in Figure 2. Flow variations of approximately 600 gpm are
not uncommon. The range of flow of the combined process wastewater is
-------
TABLE 1. COMPOSITION OF COMBINED PROCESS WASTEWATER
BAY CITY PLANT
Concentration
ma/1
Total Organic Carbon
Biochemical Oxygen Demand
Chemical Oxygen Demand
Nitrate-Nitrogen
Nitrite Nitrogen
Ammonia-Nitrogen
Phosphate-Phosphorus
Chloroaldehydes
Inorganic Chlorides
Sodium
Copper
Iron
Chromium
Manganese
Palladium
Nickel
Cobalt
PH
Color, APHA
15,000
33,000
48,000
780
27
770
130
2,085
35-400
400
2-100
0.06
0.02
<0.2
<0.2
<0.4
<0.02
w 10,000(b)
dark brown
(a) pH units
(b) color units
-------
TABLE 2. COMPOSITION OF THE PROCESS WASTEWATER STREAMS AT THE
BAY CITY PLANT BEFORE SEGREGATION
'"'
Stream
No.
1
Composition
Acid stream. Organic
Flow
Percent of
GPM Total Flow
244 38.8
Total Organic Carbon
mq/1
2700
Percent of
Total TOC
3.7
chlorides, and chlorinated
acetaldehydes
Acid stream. Phosphoric 41
acid, volatile organic acids,
sodium hydroxide, low mole-
cular weight aldehyde, aldol
products, low molecular
weight alcohols
Acetic acid through hexanoic 62
acid
Sodium hydroxide
GJ-CS esters
Alkaline stream. Amines, 49
ammonia, alcohols, oils and
copper laden solids
Acetic acid through hexanoic 215
acid, 0-7% nitric acid,
succinic, glutaric, adipic acids,
cyclic organics, copper and
vanadium (0-100 mg/1)
Nylon salt, adipic acid, 110
succinic acid, glutaric acid,
copper and vanadium
5.7
8.6
6.8
29.8
15.3
1760
2.4
54300
74.2
3500
1650
4.8
2.1
9200
12.8
-------
TABLE 3. VARIATIONS IN COMPOSITION OF COMBINED PROCESS WASTEWATER
BAY CITY PLANT
Sample No,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
DH
5.2
9.0
6.9
9.4
3.0
5.7
8.3
5.1
5.2
5.4
4.7
5.5
5.5
6.6
5.0
7.2
5.1
3.7
5.1
5.3
7.5
5.5
5.0
5.1
4.5
Mean
Total Organic
Carbon, mq/1
12,000
9,300
3,500
15,000
49,000
11,000
1,200
21,000
15.000
12,000
10,000
16,000
14,000
10,000
20,000
20,000
16,000
14,000
15,000
17,000
22,000
17,000
20,000
15,000
17,000
15,680
10
-------
AUGUST
SEPTEMBER
OCTOBER
NIOVEMBER
DECEMBER
975
900
o
§
750
675
600
525
450
300
AUGUST
SEPTEMBER
NOVEMBER
DECEMBER
Figure
Bay City - 1973
2. Daily variation in volume of
plant effluent make.
-------
between 300 and 900 gpm. Variability of the flow and composition coupled
with the toxic nature of several compounds in the process wastewater
indicate that difficulties with the biological treatment of the combined
process wastewater stream can be anticipated.
A number of physical and chemical processes were considered as
alternative schemes for treatment and/or disposal of the combined process
wastewater at the Bay City plant but each was rejected as a viable alternative
for any one of a,number of reasons. Severe corrosion problems and
inability to oxidize acetic acid which is a major constituent of the waste or
its precursors eliminated wet air oxidation from consideration. Fractional
distillation was considered but rejected because of the cost entailed in
distilling essentially all of the water away from the heavy ends which make
up the bulk of the TOC in the process wastewater. In addition, the distillate
contained a high organic concentration which precluded the reuse of the
water without additional biological treatment. The residue also contained
high salt concentration which would cause reboiler fouling and also lead to
difficulties in incinerating the residue. Two problems were apparent with
reverse osmosis; namely, poor rejection of the relatively low molecular
weight fraction of the organic compounds and the presence of tarry
fractions which cause rapid deterioration of the membrane resulting in
extremely low flux rates. Solvent extraction with removal of chloraldehyde
showed some promise, but extremely high solvent recycle requirements
rendered the process impractical in terms of solvent losses and complete
removal of the chloraldehydes was not possible. Adsorption on activated
carbon was technically possible, but was not evaluated during these studies.
A. STREAM SEGREGATION
An initial treatment scheme was developed which included anaerobic
denitrification followed by anaerobic methanogenesis and final treatment
with an activated sludge system. This treatment sequence is illustrated in
Figure 3-B. Several attempts to treat the combined process wastewater in
the anaerobic/aerobic series of processes indicated the necessity of stream
segregation. The complex mixture of chemicals in the process wastewater
posed antagonistic restraints on the stability and efficiency of the biological
treatment system. Figure 3-A presents a modified process flowscheme
that was developed to utilize the potential benefits of the constituents in the
process wastewater.
This initial segregation resulted in three composite wastewater
streams (Phase I). The composite "A" included all the nitrate containing
wastewater. Composite "B" was a blend of wastewater streams which
contained no nitrates or chlorinated organics. Composite "C" was a
composite of the other wastewater streams containing organics as well as
12
-------
3-A Final Process Flow
De nitrification
Feed
( Nitrate
Waste
Streams)
Denitrification
Filter
Methanogenic
Filter
Methanogenic Feed
(Monobasic acids,ester!
without nitrates or
chlorinated organics)
Activated
Sludge
Reactor
Activated Sludge Feed
(Ammonia, amines,
chlorinated organics)
Clarifier
3-B Initial Proce»» Flow
Feed
Total Plant
1
•*•
Denitrification
Filter
— *
Clarifier
1
1
Methanogenic
Filter
Clarifier
1
t
1
>
Activated
Sludge
Reactor
Clarifier
1
1
Figure 3. Process schematic evolution.
-------
ammonia, amines, alcohols, nylon salts and chlorinated aldehydes. Each
of these three blends of wastewater was fed to the denitrification,
methanogenic and activated sludge units, respectively. A number of
preliminary tests (Appendix A) were performed to define the design
limitations and process components compatibility.
The analyses of the various wastewater streams indicated that
effective biological treatment would be possible only if the following
remedial measures and pretreatment schemes to control future variations
in the composition of the feed were installed. The proposed controls
include: a) incineration of streams containing chlorinated organics, amines,
and high carbon concentrations with the appropriate post combustion
recovery/abatement system; b) reduction of carbon losses by in-unit process
changes to minimize wastewater production and reduce the ash content of
waste to be incinerated; c) process modifications to eliminate continuous
discharge of wastewater into unit sumps; and d) a centralized physical/
chemical treatment system to handle the intermittent flows of all unit
sumps, to reduce variations in carbon, and eliminate heavy metals from
the biological systems.
B. WASTE REDUCTION PROGRAMS
This waste reduction program resulted in significant changes in the
composition of the wastewater streams. The overall wastewater disposal
and treatment system is illustrated in Figure 4. The proposed treatment
or disposal scheme for each of the respective types of wastewater is
described in Appendix B.
14
-------
High Carbon Containing
Streans (>60,000 prm TOC)
TOTAL
INCINERATION
Orqanochloride Con-
talnlng Streams
TOTAL
INCINERATION
Anlne Containing
Streams
TOTAL
INCINERATION
PROCESS
UNITS .
(Requires major
1n-un1t proc-
ess revisions)
HHrate Containing
ALL
SUMP
(Intern
Wastewater
Low Strength Carbon
Containing Wastewater
(<6000 ppm
TOC)
UNIT
FLOWS
ttent)
DENITRIFICATION
UNIT
CONTAMINATED
AEROBIC
UNIT
STORMWATER
POSSIBLE LOCATION
FOR '
DENITRIFICATION I
Solids
TO
RIVER
DISCHARGE
Figure 4. Proposed sequence.
15
-------
SECTION 4
EXPERIMENTAL AND ANALYTICAL PROCEDURES
A. BENCH SCALE STUDIES
Bench scale treatability of aerobic and anaerobic biological treat-
ment processes was conducted at the Bay City plant and the Technical
Center (Phase I). The original objective of these studies was to evaluate
the treatability of the combined plant wastewater; however, as the project
progressed, modifications of the systems were necessary until the following
proposed treatment sequence evolved. Detailed descriptions of the bench
scale reactors are included in Appendix C.
Three bench scale units were operated at the Bay City plant and
included: a) a backmixed anaerobic reactor followed by an activated
sludge unit; b) a biological control unit that consisted of two activated
sludge units in series; and c) anaerobic filter followed by activated sludge
system. These three systems and the typical composition of the wastewater
blends fed to each system are illustrated schematically in Figure 5.
The initial bench scale studies at the Technical Center concentrated
on aerobic biological treatment using high purity oxygen as well as air.
These systems included: a) two^ stage oxygenated activated sludge system;
b) conventional activated sludge system using air; and c) long-term aeration
of activated sludge. A schematic representation of the three systems
operated at the Technical Center and typical wastewater feed blends to each
system are presented in Figure 6.
The bench scale studies at the Technical Center expanded and evolved
to develop design parameters and criteria for anaerobic denitrification,
anaerobic methano gene sis and activated sludge systems. These modifi-
cations were based on results of initial laboratory studies and stream
segregation (Figure 7).
16
-------
Anaerobic
Back mixed
Activated
Sludge
6
Feed Source
Activated
Sludge
Activated
Sludge
Anaerobic
Trickling. Filter
Activated
Sludge
Typical Feed Composite, mg/1
TOC
COD
BOD
15,000
48,000
33,000
Figure 5. Initial Bay City laboratory studies (October, 1971)
17
-------
Oxygen
Oxygen
Activated
Sludge
Activated
Sludge
Feed Source
Air
Extended Aeration
(20 days)
Air
Activated
Sludge
Figure 6.
Typical Feed Comnosite. mr;/l
TOC 15,000
COD 48,000
BOD 33,000
Initial Celanese Chemical Company Technical Center Laboratory
Studies.
18
-------
^Stripper
Treated
Heavy
Metals
rt
«
I
-
O
g
*•
rt
V
n)
U
«-•
00
o
pq
U
H
O
O
U
Denitrifical
Feeds
ion
Y
Combined
Heavy
Metals
Feeds
Metals
Effluent
Experimental System
Effluent
Possible Alternative
System
Figure 7. Feed compositing - pretreatment and biological treatment,
19
-------
B. PILOT PLANT
The pilot plant constructed at Bay City originally was designed to
treat the entire combined process wastewater stream by anaerobic
methanogenic filters followed by high purity oxygen activated sludge system.
Feed equalization and parallel reactor systems were included in the original
design with considerable flexibility designed into the system for later
modifications. The pilot plant phase of the project overlapped the
laboratory bench scale studies. Modifications in the mode of operation of
the pilot plant resulted as additional information became available. The
evolution of the pilot plant work is summarized below.
1. Stage I
Parallel two-stage treatment of the combined feed by anaerobic
methanogenic filters followed by high purity oxygen activated sludge
system was initiated. Difficulties in maintaining acceptable performance
of the biological units led to Stage II operation.
2. Stage II
Streams containing nitrates, amines, and chlorinated organics were
pumped into an anaerobic denitrification filter. The denitrified effluent
was fed into the oxygenated activated sludge system. The remaining streams
were passed through an anaerobic methanogenic filter and the effluent from
this unit could be fed either into the oxygenated activated sludge unit or
discharged from the pilot plant. These operations were characterized by
difficulties in maintaining stable operation and acceptable performance of
the biological processes.
3. Stage III
The pilot plant was operated with segregated streams in a three-stage
sequence consisting of denitrification followed by anaerobic methanogenesis
with the final stage being oxygenated activated sludge system. The high
nitrate streams were directed to the denitrification unit. The feed to the
anaerobic methanogenic filter included the effluent of the denitrification unit
plus those streams which were free of nitrates and chlorinated organics.
The effluent of the anaerobic system was mixed with the third stream which
contained chlorinated organics plus other organic containing streams and
fed into the oxygenated activated sludge system. At this time 13 individual
waste streams were delivered via seven pipes to the pilot plant. Difficulties
were encountered with heavy metals from the high nitrate stream and from
the activated sludge feed stream. The ion exchange guard beds performed
quite well; however, periodic excursions in pH caused acid conditions and high
20
-------
concentrations of heavy metals were released from the ion exchange resins
into the biological reactors resulting in loss of the biological population.
A schematic diagram of the pilot plant in its final Stage III
configuration is shown in Figures 8 and 9. Thirteen streams were segre-
gated and diverted from the general plant area via seven pipes to the pilot
plant feed equalization constant volume head tank. Head tank timer controls
were periodically adjusted to achieve feed blends representative of the plant
wastewater.
High nitrate streams were fed to equalization tank. These streams
were pumped at rates up to 0. 56 gpm to denitrification filter which was
equipped with a 20 gpm recirculation pump. The denitrified effluent was
fed to the anaerobic methanogenic packed bed filter along with those streams
which are free of nitrates and chlorinated organics.
Effluent from the anaerobic reactor was pumped into the aerobic
reactors along with raw aerobic feed. Oxygen was introduced via a down-
flow bubble contact aerator in the activated sludge unit. The reactor was
equipped with a pump to recycle the mixed liquor through a CO2 stripper
and the downflow bubble contact aerator. After clarification, a portion of
the settled sludge was recycled to the aerobic reactor. Periodic sludge
blowdown was done manually.
The pilot plant also was equipped with on-line instrumentation to
monitor TOC, dissolved O2, and copper concentration. In addition to the
variability of hydraulic and organic loading, this system was also plagued
with mechanical problems and difficulties with keeping the on-line instru-
mentation in working order. The gear type metering pumps initially
installed in the pilot plant could not be operated for more than 48 hours
at any time without extensive maintenance and repair. These pumps
were eventually replaced with peristaltic pumps which performed well.
Extensive foaming inHie downflow bubble contact aeration unit resulted in
carryover of volatile suspended solids and reduced the efficiency of this
process. Modifications in the design were ineffective in remedying this
problem.
Two types of dissolved oxygen probes were used in an attempt to
monitor the dissolved oxygen level in the aerobic system. At the time
of the pilot plant evaluation, there was no commercially available dissolved
oxygen (DO) analyzer which could provide adequate service for continued
oxygen measurement or control in the aerobic system. The on-line DO
monitoring system was one of the first installations in the U.S. Attempts
to make the system operable required a great deal of time and effort. Many
of the problems were attributed to the high mixed liquid volatile suspended
21
-------
Chloro. Organics
Initial Startup , ^_
Sept 1972
High
Wastes
DEN
ANA
^.
* "
1
1
i
I
1
i
AER
t
Oxygen
. Y
TOC - 5000 mg/1
NOj - 1000 mg/1
(A)
First High Carbon Wastes ^
Revision (B) Max 50,000 ms/1 TOC
Mar 1973 No Organic Chlorides
(C)
Normal Carbon Wastes
5000 mg/1 TO
(A)
Ion
Exchang<
Second Revision
May 1973
(C)
Legend: DEN - Denitrification
ANA - Anaerobic
AER - Aerobic
DEN
Figure 8. Pilot plant configurations.
22
-------
Live Feed
0. 4 gpm
5 g/1 BOD
Anaerobic Filters (3)
3'»x 12' Packed Bad
S«ries or
Parallel
Metering Chemical.
Pump
Centrifuge/
Filter Tests
Thickened
Sludge Storage
11.5 Scfh
Aerobic Reactors (3)
With DCCA's, 41 x 4' x 7*
Series or Parallel
Figure 9. Pilot plant schematic.
-------
solids concentration (5000 mg/1) as well as grit particles which passed by
the probe. The membrane type DO probe was abandoned in favor of the
thallium probe. Fouling of the membrane was a major drawback of the
membrane type probe and erosion of the thallium limited the applicability
of that probe at high velocities. The system was modified to reduce the
velocity at the face of the thallium probe. This was done to minimize the
problems with erosion but reliable operation was never obtained in the
pilot plant.
The continuous total organic carbon analyzer initially installed could
not be used in monitoring the pilot plant. Once again a great deal of time
and effort went into the development of this system and eventually the
system was modified by the'manufacturer so that the continuous analysis
of total organic carbon was possible. The heavy metal analytical system
to continuously monitor the copper concentration in the effluent stream
suffered from similar problems to those mentioned above. No successful
operation was obtained.
Continuous operation of the pilot plant required the attention of an
operator and instrument mechanic to maintain the instrumentation around
the clock. These experiences indicated that sophisticated and reliable
on-line analyzers were required for the control of the complex biological
system. Unfortunately, these systems were not available during the pilot
plant studies.
C. ANALYTICAL METHODS
The characteristics of the feed material to the various biological
systems are expressed in terms of biochemical oxygen demand, chemical
oxygen demand, total organic carbon, nitrates, pH, suspended solids, and
metals. The analytical procedures used during these studies are described
in detail in EPA Manual of Wastewater Analyses with the exception of COD
which for the most part was deterrained by instrumental analysis rather than
the standard two-hour dichromate reflux method. The instrumental COD
for the wastewaters used in this study was approximately 20% greater than
the reflux COD values.
24
-------
SECTION 5
RESULTS AND DISCUSSION OF GLASSWARE AND PILOT PLANT
STUDIES
A. BENCH SCALE STUDIES
Laboratory scale glassware studies of the treatability of the combined
process wastewater and of various blends of the components of the combined
process wastewater were carried out at both the Celanese Chemical Company
Technical Center at Corpus Christi, Texas and the Bay City plant at Bay
City, Texas. Every effort was made to obtain representative samples of
each of the waste streams. A portion of the composite samples were used
in the Bay City glassware studies and a portion was shipped to the Technical
Center.
The laboratory glassware studies at the Bay City plant were termi-
nated shortly after the startup of the pilot plant; however, the glassware
studies at the Technical Center continued throughout the duration of the
project. Samples for the glassware studies were prepared at the Bay City
plant by collecting samples of the individual streams. The combined waste-
water or any blends which were anticipated as a result of the waste
reduction program were prepared at the Bay City plant and shipped to
Corpus Christi at regular intervals.
B. BAY CITY PILOT PLANT STUDIES
Individual wastewater components of the entire process stream were
fed to the pilot plant during approximately a one-year period. Operating
difficulties led to modifications to the piping, pump replacement,
installation of clarifiers, changes in the size of downflow bubble contact
aerator, and CO2 stripping system.
The pilot plant was seeded with a combination of domestic activated
sludge and acclimated biomass. Acclimation to the process wastewater
was a slow process. Throughout the entire pilot plant operation, one-
half of the system was operated on the plant wastewater while the other
parallel system was fed a synthetic blend of wastewater components to
maintain an available acclimated biological population.
25
-------
The various modifications of the pilot plant system were discussed
previously and the normal configuration of the pilot plant is presented in
Figure 9. Mechanical and instrumentation problems during startup in
the initial phases of the pilot plant study made it almost impossible to
evaluate the performance of any of the biological processes.
All nitrate containing streams were channeled directly to the
denitrification unit, the high carbon streams went directly to the anaerobic
methanogenic treatment unit, and the remainder of the waste stream
containing the chlorinated organics was fed into the aerobic unit. The final
treatment scheme included a series of units employing denitrifying,
methanogenic and finally aerobic bacteria.
In addition to the mechanical difficulties which were experienced
during the startup of the pilot plant, other problems were experienced by
the denitrification unit; namely, a) excessive lime neutralization require-
ments; b) poor gas production and/or accountability; c) plant power failure;
d) excessive instrument maintenance requirements; e) poor performance
of ion exchange resin in removing soluble copper; and f) intermittent feed
supply with widely varying composition.
The variable flow of the wastewater to the pilot plant coupled with
varying concentration in the feed made manual control of changes in the
pH by adjustment of feed rates almost impossible. The changes in nitric
acid concentration resulted in widely varying pH between the feed with
pH = 1 to pH = 3 and the bioreactor which had a pH = 6 to pH = 9. At low
values of pH, copper and heavy metals removed by an exchange were
released into the biological system. The ion exchange system was
subsequently replaced by precipitation of the heavy metals with sulfide.
1. Denitrification
The activity of the denitrifying ecosystem was evaluated by the
addition of massive doses of methanol and nitric acid. The hydraulic
loading also was temporarily increased to 2200 gallons per day resulting
in a hydraulic detention time of about 19 hours. These operating conditions
resulted in a) loss of biomass; b) lowering of the reactor pH to pH = 5. 5;
c) marked reduction in gas production; and d) increase in the methane
content in the off-gas to 20% by volume. Therefore, increasing the nitrogen
load to the denitrification unit by increasing the hydraulic loading at same
nitrate concentrations did not provide a solution to the problem.
The organic carbon content of the feed was continuously analyzed
with an on-line TOC analyzer for a single week. During the test period,
the TOC varied in diurnal cycles between 2000 and 2100 mg/1.. None
26
-------
of the large upsets experienced in previous operations were seen during
this single test period. Operating data for the denitrification unit are
presented in Table 4. The data reported during the early phases of the
pilot plant operation are somewhat erratic; however, only average values
are reported for the first six-month period. The bulk of the data in
Table 4 represents biweekly averages of daily analyses.
During the last three-month period of operation of the pilot plant,
the denitrification unit operated at a mean hydraulic detention time of
about five days and the average nitrogen removal efficiency was 67. 7%.
The feed to the denitrification unit contained an average TOG of 10, 400 mg/I
and BOD 16, 200 mg/1 and nitrate concentration of 600 mg/1. The TOG to
NO3 ratio in the feed averaged 9. 1. The average loading was 0. 15 Ib of
organic carbon per cu ft per day and the removal efficiencies were
relatively low in the order of 32. 2% for TOG and 19. 2% for BOD.
The denitrification process required particular attention to alkalinity
and TOG to NO3 ratio control. Excess nitrates in tiie reactor effluent
usually was correctable by the addition of methanol.
The biweekly averages of the carbon concentrations varied as much
as 80 to 90%. This variability accented by equally variable heavy metal
concentrations in the feed contributed to the reduced nitrogen removal
efficiency during the latter part of the pilot plant operation. The data
observed in the early phases of the pilot plant operation indicated that
nitrate removal efficiencies in excess of 90% were readily obtainable.
The chemical unbalance reduced the operating capability of the unit.
Towards the end of the pilot plant operation, the hydraulic loading
to the denitrification unit was increased to 1, 100 gallons per day which
resulted in deactivation of the unit. The BOD loading during this time
was increased to 0. 5 Ib/cu ft/day which is more than three times the
previously identified stable range of operation.
2. Methanogenic Unit
The operation of the methanogenic unit was the most difficult. Over-
all growth rate of the methane bacteria was extremely low and the amount of
organic conversion was almost negligible. In the early stages of the pilot
plant operation, low rate of activity in the methanogenic unit was attributed
to the solids washout. In the first quarter of operation, the methanogenic
filter was treating waste with a TOG of 1000 mg/1. Shortly thereafter, the
TOG concentration was increased to 2000 mg/1 and progressively to
8000 rag/1. Within four days the gas production was completely stopped. A
period of 30-60 days was required to regain activity in the methanogenic unit.
27
-------
oo
TABLE 4. SUMMARY OF PILOT PLANT DATA
DENITRIFICATION
6PD
290
381
315
335
217
575
11.5
138
220
291
327
132
61
173
61
375
398
TOC
Feed
5,837
11,387
16,443
11,661
8,340
8,898
15,461
2,895
2,843
2,732
3,703
3,803
6,722
5,710
3,947
2,102
3,743
Product
2,625
6,593
13,180
7,145
6,219
7,738
142
622
1,346
1,544
2,206
2,082
1,843
1,302
789
134
1,239
BOD NO--N
Feed Product Feed Product
11,520
21 ,625
23,889
13,242
12,312
14,800
7,269
6,001
4,198
6,223
7,748
10,822
9,029
8,613
3,194
5,861
6,181 609 106.2
10,683 474 148.5
29,306 575 410.0
11,103 891 70.3
10,232 726 214.0
13,600 590 214.0
METHAN06ENIC
1,170
2,065
2,543
2,491
2,645
AEROBIC
2,250
2,682
2,376
484
61
1,120
TOC REMOVAL, % * TOC/
Loading TOC BOD NO?-N NOo-N
15.9
46.4
55.0
37.0
14.5
43.4
1.3
3.3
5.1
6.5
9.8
6.3
3.6
7.7
0.9
7.4
12.1
55.0
42.1
19.8
38.7
25.4
13.0
99.1
78.6
52.7
43.5
40.4
45.3
72.6
77.2
80.0
93.6
66.9
46.3 82.6 9.58
50.0 68.7 2.40
-22.7 28.7 2.86
16.2 92.1 13.09
16.9 70.5 11.48
8.6 63.7 15.08
71.7
57.6
40.7
57.5
71.0
75.2
73.7
94.4
98.1
80.9
*Based on Concentrations.
-------
A synthetic blend of sodium acetate, methanol and nutrients was used
during the reacclimation period to develop a healthy methanogenic bacterial
population. At that time the Bay City plant waste was introduced as feed to
the anaerobic unit.
Key to the operation of the methanogenic unit was frequent attention
to: a) pH of the reactor; b) the cationic concentrations; c) off-gas analysis;
and d) organic loading rates. The mean hydraulic detention time was
nine days and the methanogenic unit was fed a blend containing 5500 mg/1
of TOG and a BOD of 5900 mg/1. This loading is equivalent to 0. 03 Ib
organic carbon/cu ft/day resulting in removals of 50-60% of the TOC and
72. 6% of the BOD. The gas production rate was 68. 5 cu ft/day of which
60- 80% was methane. At a detention time of less than ten days the loading
to the reactor markedly affects the removal efficiency of the methanogenic
unit. The organic carbon loading was generally less than 10 Ib/day.
Operating the methanogenic unit at a hydraulic detention time of
six days resulted in relatively stable operating conditions. The off-gas
contained approximately 70% methane with less than 2% oxygen. Approx-
imately 6. 8 Ib of organic carbon were converted into methane and biomass
per day.
The biomass production rate is equivalent to 500 mg/1 or 0. 16 mg
per mg BOD removed. Periodic inspection indicated that the biomass
concentration in the effluent was 400-500 mg/1 under normal hydraulic
rates. As the hydraulic rate reached 500 gal/day with a detention time of
3.5 days, the volatile suspended solids concentration increased to 2500-
5000 mg/1. This high washout rate contributed significantly to the rapid
deterioration of the process as the hydraulic loading was increased. Within
a 24-hour period at the higher hydraulic loading rates, gas production was
decreased by 90%. However, by reducing the hydraulic loading to
392 gal/day, gas production began to return to normal after seven days
of operation. Therefore, the system was not completely inactivated since
only a portion of the active biomass was lost at the higher hydraulic loading
rates.
3. Activated Sludge Unit
Operation of the activated sludge plant was extremely vulnerable to
malfunctions in the dissolved oxygen controller and to the influx of heavy
metals in the feed blends to the system. Each of these problems caused
at least one major reactor deactivation during the pilot plant operation. In
one case, a complete replacement of the biomass was required.
29
-------
The mean hydraulic detention time was 5. 5 days and the average
TOG loading of 0. 4 Ib/cu ft/day. The average TOG removal was 72. 6%
with a BOD removal of 82. 2%. The characteristics of the feed to the
activated sludge system also are shown in Table 4. The average BOD
was 7500 mg/1 and the average TOG was 4300 mg/1 during the last quarter
of operation. The plant operated on the feed blend which varied in
composition sufficiently to change the biweekly feed concentration averages
by 80-90%. The feed to biomass ratio (F/M) ranged from 0. 2 to 1.3 Ib
of BOD/lb of MLSS. The best operating efficiency was obtained at F/M
equal to 0. 2 to 0. 6 Ib of BOD/lb of MLSS. The BOD to TOG ratio in the
feed to the activated sludge plant varied from 1.5 to 2.2 which provided
for 32% change in the oxygen demand within a change in the total organic
carbon content.
C. TREATABILITY STUDIES
Extensive laboratory treatability studies were conducted at the
Technical Center to define in more detail the effectiveness and limitations
of the anaerobic and aerobic processes in treating the complex petro-
chemical wastewater. Composited pretreated samples of wastewater
were collected at the Bay City plant and shipped to the Technical Center
(Phase III). Synthetic blends of typical components were prepared for
some of these studies. The results of the aerobic studies are presented
in Appendix D, Table D- 1 through D- 3.
1. Aerobic Treatment
Two stage oxygen activated sludge systems were operated at
detention times of four days per stage. At feed concentrations of
13. 5 gm/1 of TOC and 30 gm/1 of BOD overall removals of 85% for
TOG and 95% for BOD were observed. The effluent after clarification
contained dispersed growth (1000 mg/1 suspended solids) and was turbid.
High concentrations of amine by-products from the nylon unit was toxic to
the system.
Pretreatment and distillation eliminated the toxic amines, and the
TOC of the resulting wastewater was approximately 4000 mg/1. This
stream could be treated in the two-stage oxygen system at efficiencies
of 90% for TOC removal and 98% for BOD removal at detention times
of one to three days per stage. The system was unstable and easily upset
attheoneday detention time. In all cases, however, the effluent was
turbid because of poor settling of dispersed biomass and was yellow in
color.
30
-------
Percolation of the effluent through a granular activated carbon bed
(Filtrasorb 300 Calgon) did remove the color. However, additional in-
formation which defines effective carbon life, regeneration capabilities,
etc., are not available. Ozone and hydrogen peroxide also removed some
of the color but the cost of these oxidants is economically prohibitive.
2. Anaerobic System
The high oxygen requirements (approximately 40 tons/day) for
commercial scale aerobic treatment of the combined Bay City plant process
wastewater led to the consideration of anaerobic treatment.
The results observed in the initial anaerobic studies led to the
segregation of the process plant wastewater into the following three
composites: a) a composite containing all the organic chlorides; b) nitrate
containing waste streams; and c) normally biodegradable hydrocarbon waste
which is free of organic chlorides and nitrates (Figure 10). Composite A
was treated aerobically and the results are discussed above. Composite
B containing the nitrates was fed to an anaerobic backmixed filter for
denitrification. The characteristics of composites B and C are presented
in Tables 5 and 6, respectively. The relatively high TOC concentrations
of composite B is caused by a high carbon containing stream which was not
used in the composites fed to the aerobic units.
a. Denitrification
The startup of the denitrification anaerobic filter was almost
immediate as indicated by the data presented in Appendix E, Figures E- 1
through E-2. The operation of the system was relatively stable. Nitrate-
nitrogen is almost completely removed from the feed containing
2. 6 grams/liter and the effluent contained a concentration less than
5 mg/1 of nitrate-nitrogen (analytical detectability). The removal of
2. 6 g/1 of nitrate-nitrogen results in the removal of approximately 2. 5 g/1
of TOC. The organic carbon content in composite B is too low to reduce all
the available nitrate; therefore, approximately 10% by volume of composite
C was added to the denitrification feed in order to provide additional TOC.
A more detailed evaluation of the performance of the denitrification filter
in terms of influent flow gas production, nitrogen production, detention time,
BOD removed per liter of feed, volume ratio of gas produced feed, mole
fractions of nitrogen, methane and N2O as well as pH is presented in
Figures E- 1 through E- 2.
These data indicate that during stable operations the pH was between
pH>7.0 and 7. 5 and constant. The ratio of gas produced per unit of feed was
constant, gas composition was also constant and the effluent was completely
31
-------
Organic Chloride*
Nitrate
Streams
Composite B
Denitrification
Filter
Carbon
Only Streams
Composite C
Methnnogenic
Filter
Activated
Sludge
Unit
Clarifier
Final
Outfall
Figure 10. Denitrification - anaerobic - aerobic
treatment of Bay City effluent.
32
-------
TABLE 5. COMPOSITE OF NITRATE-CONTAINING BAY CITY WASTES
"COMPOSITE B"
Identification
Appearance
pH
TC, mg/1
1C, mg/1
TOC, mg/1
COD, mg/1
Total Chlorides, mg/1
Total Nitrogen, mg/1
Sodium, mg /I
Phosphorus, mg/1
Copper, mg/1
28-1
Clear,
Yellow
0.9
2660
200
2460
6900
<20
2500
2900
<20
10. 1
28-2
Clear,
Yellow
0.8
2050
200
1850
8000
<20
2900
2500
<20
10. 2
13-1
Clear,
Yellow
1.4
3700
380
3320
10500
< 100
1.0
< 10
30
13-2
Clear,
Yellow
< 1. 0
3130
3130
8300
< 100
7. 0
< 10
0.2
33
-------
TABLE 6. BAY CITY COMPOSITE, CARBON-ONLY WASTES
"COMPOSITE C"
Identification
Appearance
PH
TC, mg/1
1C, mg/1
TOC, mg/1
COD, mg/1
Total Chlorides, mg/1
Sodium, mg/1
Total Nitrogen, mg/1
Nitrate Nitrogen, mg/1
P as PO4=, mg/1
Copper, mg/1
26-1
Brown
3. 1
8000
3
38000
80400
35
400
1ZOO
< 10
3200
4.6
26-2
Brown
3.4
28000
10
28000
60600
20
410
1400
<10
36000
2.4
14-1
Amber
2.9
56500
93
56500
184000
< 100
4000
1200
22
14-2
Amber
3. 3
67000
93
67000
208000
< 100
4500
2400
22
34
-------
devoid of nitrate nitrogen. Under upset conditions there was a drop in pH,
an appreciable amount of nitrate-nitrogen appeared in the effluent, and the
composition of the gas produced changed markedly with a reduction in the
mole fraction of nitrogen accompanied by a corresponding increase of N2O
in the off-gas. Recovery of the system after upset apparently could be
quite rapid. The upsets are a result of a high TOG to nitrate ratio in the
absence of sufficient buffering capacity. Under these conditions, a
considerable amount of the residual organic carbon is not oxidized by the
nitrate, and since the influent is quite acid, the pH drops. At the lower
pH, the denitrification reaction proceeds to N2O rather than to nitrogen
gas. Therefore, the amount of oxygen available for destruction of the TOC
is reduced and additional lowering of pH occurs. As the pH drops further,
the denitrification to nitrogen gas practically stops and reaction results in
the generation of NO which leads to an additional decrease in the pH.
Therefore, once the environment is such that denitrification to nitrogen
gas is inhibited or ceases, the situation in the reactor rapidly deteriorates.
Under the low pH conditions, copper present in the feed may dissolve and
add to the toxic effects.
Composite B which contains adipic acid as the main source of
carbon was blended with known amounts of nitrate, TOC and alkalinity
in order to better define the parameter affecting denitrification. Sodium
bicarbonate was added to provide buffering capacity and copper also was
added at a constant concentration of 20 mg/1 to simulate the major heavy
metal present in the waste. The stoichiometry involved in the oxidation
of adipic acid via biological denitrification is expressed below:
2 HNO3 -» H2O + N2 + 50
MW = 126
HOOC(CH2)4COOH + 13 O-* 6CO2+ 5 H2O
MW = 146
26HN03+5HOOC(CH2)4COOH- 38H2O + 30CO2+ 13N2
Approximately two grams of adipic acid were required per gram of
nitrate-nitrogen. The feed blend contained a constant one gram per liter
of nitrate-nitrogen and 4. 5 g/1 of sodium bicarbonate (54 millimoles of
sodium per liter) providing buffering. The sodium bicarbonate was
sufficient to neutralize the adipic acid which was not biologically degraded
during the denitrification process. In addition to the nitrate and buffer
capacity, various amounts of adipic acid were added to the feed, namely,
2, 2,5, and 3, and 3. 5 g/1. The hydraulic detention time in the unit was
35
-------
one day based on the void volume. The response of the reactor to increasing
amounts of adipic acid from 2 to 3. 5 g/1 were as follows: the residual TOC
increased from 150 mg/1 up to 750 mg/1 while the COD increased from
100 mg/1 to 1200 mg/1. No copper was detectable in the effluent and the
nitrate-nitrogen concentration in the effluent was less than 5 mg/1 (detectable
limit) throughout the study except when the feed adipic acid concentration
was 2 gm/1. At that time, the nitrate nitrogen concentration decreased
with simultaneous appearance of between 10 and 15% of N2O in the product
gas. The appearance of the N2O and the breakthrough of the nitrates
resulted from an insufficient amount of organic carbon in the feed.
Methane was present in the produced gas at all feed rates in small
amounts; namely, from 0. 5 to 10% by volume. The volume fraction of
methane in the product gas increased as the concentration of adipic acid
in the feed increased.
Denitrification occurred by complete oxidation of adipic acid to CO2
and water rather than by stepwise oxidation, since no intermediate degra-
dation products of the adipic acid were present in the effluent.
b. Treatment of Composite C to Methanogenesis
Acclimation of the anaerobic unit to Composite C containing a high
concentration of carbon but a very low concentration of nitrates required
a considerable amount of time. Gas production was satisfactory when the
feed was diluted to half strength; however, the system failed when full
strength waste was fed. After failure of the system, the reactor was once
again acclimated to a feed containing equal volumes of composite C and the
effluent of the denitrification unit treating composite B. The volume ratio
of the two streams was estimated to be similar to the actual ratio of the
streams at the Bay City plant. It should be pointed out at this time that
the effluent of anaerobic reactor treating the composite C is decolorized
from an extrapolated APHA color of greater than 10, 000 to APHA color of
150. This color removal represents a marked improvement of the effluent
since the color remained unchanged after the various aerobic treatment
systems.
Approximately 65% of the feed to the anaerobic methanogenic unit
consisted of denitrified effluent and the full strength composite C made up
approximately 35%. Variations in the performance of the methanogenic
units are presented in Figures F- 1 through F- 2. The activity in the unit
remained relatively poor until the amine streams were removed from the
composite C. At that time, the operation of the anaerobic methanogenic
units improved considerably in spite of mechanical troubles.
36
-------
The anaerobic methanogenic unit operated quite well until day 142
when the volume percent of the composite C in the feed to the unit was
increased to 25%. At that time the gas production decreased steadily and
the system did not get back to normal performance until the feed strength
was cut in half to 12. 5% by volume.
The final phase of the anaerobic methanogenic treatment of
composite C used a feed which contained 30% strength composite C diluted
with water to determine the minimum hydraulic detention time for the
anaerobic methanogenic unit. Performance of the anaerobic filter was
relatively marginal at a detention time of two days and the TOC removal
decreased to 60% compared to the 75 to 80% removal of TOC observed at
a four-day detention time. The detention times mentioned in this
discussion are based on the original void volume of the filters and the actual
detention time is considerably less since the void volume was decreased by
growth of biomass in the units.
c. Process Limitations and Design Criteria for Denitrification
Process wastewaters from the Bay City plant contain several streams
which have relatively high concentrations of nitrates. The nitrates inhibit
methanogenesis but can be converted to nitrogen gas (N2) quite effectively
anaerobically. Off-gas from the denitrification filters consists essentially
of N2 and CO2 with only trace amounts of methane, less than 1% by volume.
The denitrification process is affected by the ratio of organics to the
nitrates in the influent. At high concentrations of organics in comparison
to nitrates the system is characterized by a sharp decrease in the effluent
pH and the appearance of N2O in the off-gas. If the system continued to
operate at the relatively high organic to nitrate ratio the N2 in the off-gas
would drop to zero and NO would be detectable in the off-gas. However,
the biomass in the system could be regenerated after being upset by high
organic to nitrate ratio by flushing the system with dilute bicarbonate of
soda folio-wed by the addition of the substrate -which contained more
alkalinity and/or more nitrates in comparison to the organic substrate.
In the presence of sufficient substrate and residual alkalinity, the bacterial
denitrification reaction goes to N2:
2 HNO3 -» H2O + N2 + 5 (O) Reaction I
In this reaction one part of nitrate N is reduced by 2. 857 parts of
substrate COD. At low pH values, the denitrification reaction is not
complete and the N2O appears as an end product:
2 HN03 - H2O + N20 + 4 (O) Reaction II
37
-------
In this case, one part of nitrate N corresponds to only 2. 286 parts of
substrate COD. If the appearance of N2O was caused by an acidic organic
substrate with limited alkalinity the residual substrate concentration would
be higher and result in a further decrease in Hie pH and a larger proportion
of N2O in the effluent. As the pH continues to drop, the denitrification
becomes less efficient:
2 HNO3 -. H2O + 2 NO + 3 (O) Reaction III
Therefore, one part of nitrate N now corresponds to only 1. 714 parts
of substrate COD. Analytically NO is difficult to measure, but its presence
was evident by NO2 fumes which were generated when the off-gas contacted
the atmosphere. The denitrification of the effluent to N2O via Reaction II
can represent a stable mode of operation, for example, when there is
insufficient substrate. Results of laboratory experimentation indicate
that rather than leave residual nitrate in the effluent, the denitrifying
organisms will produce enough N2O in the off-gas to balance the substrate
COD against the oxygen from denitrification.
The process wastewaters from the Bay City plant commonly contain
relatively high concentrations of organics relative to the nitrate. Therefore,
a bench scale evaluation was undertaken to determine if a stable operation
could be maintained in the pH range of 4. 6 as the carboxylic acid. If this
scheme were possible, denitrification could take place without the addition of
large amounts of alkalinity.
A synthetic substrate was used in the bench scale studies, and the
composition is presented in Table 7. Operating data for the anaerobic
denitrification filter are presented in Figures G- 1 through G-5 in.the
Appendix. The various parameters used to describe the performance of
the anaerobic denitrification process are presented graphically.
The synthetic substrate was designed so that denitrification according
to Reaction I -would remove 0. 05 ml/1 of acetic acid and leave a buffer
containing 0. 05 ml/1 of acetic acid and 0. 05 ml/1 acetate. Under these
conditions, the denitrification reaction would be forced to proceed at the
pH of the carboxylic acid- carboxylate buffer and would approach the
conditions of a COD rich denitrification feed similar to that from the Bay
City plant except the process waste would contain C5 to C6 dibasic acid
instead of acetic acid.
After initial startup problems a low feed rate was maintained until
the performance of the system resembled steady state conditions,
approximately 50 days based on washout kinetics for a well mixed reactor.
38
-------
TABLE 7. COMPOSITION OF SYNTHETIC DENITRIFICATION EFFLUENT
CH3COOH =6.0 g/1 = 0. 1 mole /I = 6^
CH3COONa = 4. 1 g/1 = 0. 05 mole /I = 3.2 g COD /I
HN03 = 1. 12 g nitrate N/l s 4. 57 g COD /I = 0.071 moles
acetic acid/1 via denitrification to NH3
= 3. 2 g COD /I = 0. 05 mole acetic acid/1 via
denitrification to N2
= 2. 56 g COD/1 = 0. 04 mole acetic acid/1 via
denitrification to N2O
Alkalinity from CH3COONa = 0. 05 equivalent /I
Alkalinity from NH3 generated via denitrification = 0. 08 g/1
PrCOD of nitrate N = 1:100
P added as H3PO4
CO++ = 2 mg/1
Fe++ = 2 mg/1
5 = 20 mg added as Na2SO4
39
-------
Under the low loading conditions, the pH continued to drop to the
pH= 5. 76, and then began to increase gradually to about pH = 7 when the
calculated washout was achieved. At the same time, the amount of N2 in
the off-gas began to drop and after approximately 25- 30 days the N2
accounted for only approximately 15% of the nitrate N in the feed. The
data indicate that the denitrification was complete at all times. Analysis
of the effluent of the reactor indicated that the total Kjehdahl nitrogen was
relatively high. It was possible that at these low loading feeds denitrification
was resulting in ammonia as an end product.
HN03 + H20 - NH3 + 4 (O) Reaction IV
Denitrification to ammonia represents 4. 57 g COD per part of nitrate
N in the feed, and therefore, a considerably greater quantity of substrate
would be consumed in denitrification to ammonia than would be consumed by
conventional denitrification to N2 gas. The denitrification reaction resulting
in ammonia also produces an appreciable amount of alkalinity which aids
in buffering the unoxidized fraction of the acidic substrate. In this case,
the nitrate N in the feed would be denitrified entirely to ammonia,
consuming 0. 08 moles per liter acetic acid. This compares with
0. 05 moles per liter of acetic acid if the denitrification proceeded to N2
gas. At the same time, the ammonia would result in 0. 08 equivalent of
buffer which could be used to neutralize any acetic acid substrate. Denitri-
fication via ammonia was confirmed by calculating the amount of
ammoniacal N accumulated in the reactor by assuming all the nitrogen
not accounted for as off-gas is converted to ammonia. A nitrogen balance
maintained during these studies verified the various nitrogen products in the
offgas at various carbon loadings.
Under conditions of increased pH in the reactor (pH greater than 6)
an increasing amount of methane was observed in the off-gas. Contribution
of this methane to COD removal is quite significant and by day 40 was
comparable in magnitude to the amount removed by denitrification. At
about the same time, the amount of N2 gas in the off-gas, also began to
increase. It may be possible that methanogenesis can compete successfully
with the relatively feeble energetics of the reaction of N2-NH3. The system
approaches steady state at a relative low loading of COD between
0. 5- 0. 6 gm/l/day and a hydraulic detention time of approximately 15 days.
This resulted in about 70% of the nitrate-N accounted for in the off-gas.
Also, the effluent pH was practically neutral. Under these conditions the N
and methane in the off-gas accounted for approximately 90% of the 9. 6 gm/1
of the COD in the feed to the unit and was initially in agreement with the
removals calculated based on effluent analysis.
40
-------
The feed rate to the unit was increased gradually in an attempt to
develop conditions in the anaerobic filter where methanogenesis would be
overwhelmed by the increased load of nitrates. However, only a small
decrease in methanogenesis was observed under these high nitrate loadings.
The COD load was increased to about 8 gm/l/day (approximately 500 lb/
1000 cu ft/day). More than 95% of the nitrogen in the feed was accountable
as N2 gas in the off-gas, and the effluent is completely denitrified in the
system. The denitrification reaction under these conditions proceeds
entirely to N2 gas; the COD removal is greater than 95% based on the
analysis of the off-gas in terms of methane and nitrogen, A substantial
fraction of the substrate COD is converted to biomass which is lost from
the system. The various denitrification reactions induced by changing the
feed composition and organic loading are represented by the bar diagrams
in Figure 11.
41
-------
Denitrification
to N2 Gas
Denitrification
and
Methanogenesis
Denitrification
to NH3
FREE ACETIC ACID
ALKALINITY
ACETIC ACID DEGRADED VIA METHANOGENTS
ACETIC ACID DEGRADED VIA DENITRIFICATION
0.15
0. 14
0.13
0. 12
0.11
0. 10
0. 09
0.08
0. 07
0.06
0.05
0. 04
0. 03
0.02
0. 01
0.0
Figure 11. Acid-base relationships in the synthetic substrate
used for denitrification studies.
42
-------
SECTION 6
CAPABILITIES OF PROPOSED TREATMENT SEQUENCE
The two biological treatment sequences which were evaluated
are illustrated schematically in Figures 12 and 13. Scheme I involves
denitrification and a concurrent heavy metal removal process. The
denitrified effluent plus other composites were then treated by the activated
sludge process. Scheme II also involved denitrification to remove nitrates
and the denitrified effluent was mixed with the aerobic composite and the
heavy metal composite which was not pretreated for metal removal. These
three streams were fed to an anaerobic process for the removal of most
of the BOD and all of the residual heavy metals. The effluent of the
anaerobic system was fed to an activated sludge unit in order to provide
final polishing and removal of residual organics.
A. CHARACTERISTICS OF THE COMPOSITES
1. Denitrification Composite
The characteristics of the grab samples used for the startup of the
packed bed anaerobic denitrification unit and of the composites used during
the demonstration run showed the variation expected by this sampling
procedure.
Several analytical discrepancies and limitations were evident from
analysis of the data. The nitrate N reported by the Technical Center
differed appreciably from those reported by the Bay City plant laboratory. A
comparison of these data with the "kinetic" nitrate N removal as N2 indicates
that the data reported by the CCCTC more closely approximated the actual
concentration, but in most cases the nitrate concentration reported is still
too low. The total nitrogen concentration determined by microcoulometry
performed at the Tech Center most nearly approximated the actual data
observed in the units more closely. The nitrate N data reported by the
Technical Center were much lower than those reported by the Bay City
laboratory. The data reflect the destruction of nitrite under acidic
conditions in these effluents by the several days of sample storage between
analysis of the Bay City laboratory and the Technical Center. The average
nitrate N calculated from the Technical Center data observed for the
43
-------
" DENITRIFI CATION'1
COMPOSITE
DENITRIFICA TION
"HEAVY METALS"
COMPOSITE
"AEROBIC"
COMPOSITE
HEAVY METALS
REMOVAL
"HEAVY MET
COM
ALS REMOVED"
OSITE
ACTIVATED
SLUDGE
1
FINAL OUTFALL
Figure 12. Treatment sequence of Bay City plant effluents.
G
W
rt
H
44
-------
• DENITRIFICA TION"
COMPOSITE
DENITRIFICA TION
"HEAVY METALS"
COMPOSITE
"AEROBIC"
COMPOSITE
DENITRIFIED OUTFALL
ANAEROBIC
FILTER
ANAEROBI
3 OUTFALL
ACTIVATED
SLUDGE
1
FINAL OUTFALL
Figure 13. Alternative treatment sequence of Bay City plant effluents.
45
-------
denitrification composites was 1. 334 ± 0. 625 g/1 which corresponds to
3. 811 ± 1. 786 g/1 of oxygen equivalent or COD of the sample.
Analyses of the data also indicate that the instrumental (Aquarator)
COD results were markedly affected by the nitrate concentration in the
sample. Attempts were made to correct the observed COD values by
incorporating a "nitrate factor". However, it was not possible to develop
any correlation of the effects of the nitrates on the instrumental COD by
analyzing samples containing known concentrations of COD and nitrates.
Therefore, analyses of these nitrate-rich samples must be performed by
the classical reflux technique. In the few cases when the reflux COD
data were available from both the Technical Center and the Bay City
laboratory, the agreement was quite good as was that of the TOC's.
The TOC concentration of the denitrification composite was 3. 179 ±
2. 994 g/1 which indicates a very strong organic waste. These composite
samples were also quite acidic and contained a relatively high but fairly
constant concentration of copper (about 45 mg/1).
The variability in the organic and nitrate N concentration in these
composites at times resulted in insufficient organic material for the
reduction of the nitrates. Therefore, the feed to the unit was supplemented
with additional "organics" COD from the aerobic composite.
2. Heavy Metal Composite
The characteristics of these two composites should be essentially
identical with the exception that the heavy metals have been removed
from one of the composites. The COD concentration of these composites
is relatively high and extremely variable with an average of 11. 126 ±
9. 282 g/1 of COD for the heavy metal composites and 9. 280 ±6. 905 g/1 of
COD for the composite with heavy metals removed. The ratios of the COD
to TOC were 3. 069 ± 0. 638 and 3. 759 ± 0.662, respectively, for the heavy
metal sample and the composite with the heavy metals removed. These
ratios indicate a relatively low oxidation level for the organic matter in the
two composites.
3. Aerobic Composite
A complete analysis of all the aerobic composites was made to
represent all the streams at the Bay City plant which are free of nitrates,
heavy metals and chloraldehydes. The COD content of these samples is
approximately twice the COD of the heavy metals composites; however,
the variability is less than these samples. The average COD concentration
46
-------
is 21.235 ± 7.958 g/1. The ratio of the COD to TOC is 3. 379 ± 0.511 which
indicates a more oxidized substrate at oxidation state approximating that
of acetaldehyde. Sodium content in the aerobic composites also is
relatively high and relatively constant. The average sodium concentration
is 4.022 ± 1.435 g/1.
B. TREATMENT PERFORMANCE
1. Denitrification
The anaerobic denitrification was carried out in two different modes;
namely, a packed column with recirculation and a fluidized bed biological
reactor with recirculation of the reactor contents. The denitrification
process in both systems resulted in effluent nitrate N concentrations of less
than 10 rag/1. Consistently low nitrate concentration in the effluent is some-
what remarkable in view of the fact that denitrification composites were not
balanced in composition of COD to nitrate N required for the denitrification
reaction to go from nitrates to N2 gas. In general, the COD at 5- 10 g/1
exceeded the oxygen available from the nitrate N at concentration of
1 - 2 g/1. In two cases, the oxidative power of the nitrate exceeded the
COD and required the addition of a calculated amount of the heavy metal
composite to the feed.
a. Packed Bed Unit
The conventional anaerobic denitrification filter was operated at
a hydraulic detention time of about 1.2 days with a corresponding loading
of approximately 300 Ib of COD (nitrate O)/100 cu ft/day for a feed nitrate
N concentration of 2 g/1. The operating data observed for the packed bed
anaerobic filter are presented in Figure E- 1 through E-2 in the Appendix.
Relatively poor nitrate analyses coupled with poor instrumental
results of the COD analyses caused by the presence of high nitrate
concentration samples resulted in difficulty in controlling the ratio of
COD to nitrates N entering the denitrification unit. These difficulties were
particularly significant at the beginning of the demonstration run when the
nitrate O exceeded the COD in the denitrification composite resulting in the
appearance of significant amounts of N2O in the off-gas.
When the ratio of COD to nitrate O in the effluent to the denitrification
unit was approximately balanced (COD only slightly greater than nitrate O)
the off-gas contained primarily nitrogen with very small concentrations,
less than 0. 5% by volume, of methane. Under these conditions, the high
nitrate concentration inhibits methanogenesis. The effluent of the reactor
also contains very little ammonia nitrogen as indicated by the relatively low
47
-------
concentration of total nitrogen as measured by the microcoulometric
technique. However, when the feed contains a much greater concentration
of COD to nitrate O, for example, greater than 1 g/I of COD over and above
the nitrate concentration, significant amounts (approximately 5 to 15% by
volume) of methane are found in the off-gas. The effluent also contains
between 100 and 200 mg/1 of ammonia nitrogen. These data indicate that
a significant concentration of methanogenic organisms are able to survive
in the packing of the reactors in spite of the presence of inhibitory nitrate
concentrations. During these time periods, significant amounts of methane
were found in the off-gas, although the major constituent of the off-gas
was nitrogen.
The amount of ammonia nitrogen in the effluent during these same
time periods also is significant indicating that reduction of some of the
nitrate O to ammonia also takes place as well as the conventional pathway
of converting the nitrate N to N2 gas. At the end of the run, the packed
bed was dismantled and a considerable accumulation of biomass was
observed in the packing. However, this biomass was in a relatively
loose form which should not lead to channeling and was relatively easily
removed from the packing.
b. Fluidized Bed Reactor
The circulated fluidized bed unit was evaluated in view of the success
with denitrification observed in the packed bed system because of the
obvious freedom from channeling and easy removal of the excess biomass
in the fluidized bed system. The data observed during this demonstration
run are presented in Table G- 1 through G-5 in Appendix G.
The fluidized bed denitrification unit was run for only a relatively
short period of time and the data are somewhat erratic; however, the
potential capability of the fluidized bed denitrification unit are clearly
indicated by the data. In fact, during the course of the demonstration run,
the entire contents of the unit was lost and the system was restarted with
relatively few difficulties. The loading to the fluidized bed unit was
approximately 1. 5 Ib of COD (or nitrate O) per cu ft per day without
evidence of overloading. The hydraulic detention time with these
loading conditions is less than 0.2 days. The only significant difference
observed in the fluidized bed system compared to the packed bed system
was the absence of methanogenic acitivity in the fluidized system when
the feed COD concentration exceeded the available nitrate O.
The off-gas in the fluidized system never contained more than 0. 1%
by volume of methane, even under the same feed conditions that resulted
in a generation of about 10% by volume of methane in the off-gas of the
48
-------
packed bed unit. The lack of any significant methanogenic activity in the
fluidized bed system may be attributed to the more violent hydraulics of
the system which tends to wash out the slower growing and more loosely
attached methane organisms while the denitrifying organisms adhered to
the fluidized media.
At the end of the denitrification run with the fluidized bed reactor, a
portion of the fluidized bed material (liquor plus biomass coated activated
carbon) was removed and analyzed. The volatile suspended solids con-
centration was relatively high and account for the capability of the system
in the fluidized bed reactor to handle high loadings. The data also indicate
that the copper which was trapped in the system as insoluble copper sulfide
was uniformly distributed between the biomass and the activated carbon.
The major problem in the control of the denitrification system was
a need for relatively slow reflux COD analysis since the nitrates interfere
with the much more rapid instrumental analysis of COD.
2. Methanogenesis
Performance of the anaerobic system was very poor for reasons
which are not very clear. In view of the poor performance, the system
was shut down after a relatively short period of operation. However, an
anaerobic filter which had been treating effluents from another plant with
COD removal of greater than 90% became available, and the Bay City plant
composite consisting of the aerobic feed, denitrified effluent, and heavy
metals composites were fed to the unit. The COD removals were only in
the order of 30 to 40% which was insufficient to justify an anaerobic treat-
ment system, and this procedure was discontinued for the final study.
3. Aerobic Treatment
a. Activated Sludge Treatment
The activated sludge treatment of the aerobic composites from Bay
City plant was started much sooner than this final demonstration run. The
operating data are presented in Appendix D. In the early parts of the acti-
vated sludge treatment studies, the feed to the units was aerobic grab
samples. During the later stages, the unit began receiving the demon-
stration run composites made up of the denitrified effluent, the aerobic
composite, and the heavy metals removed composite. The components
were blended each week in a ratio which corresponded to the rate of gene-
ration of these wast streams at Bay City. Mechanical operation of the bench
scale activated sludge system posed no problems; however the effluent COD
concentration was greater than 1000 mg/1. The addition of an extended
49
-------
aeration system following the activated sludge system did not result in any
marked reduction in the effluent COD. The cause of the high COD was
attributed to either (a) presence of major non-biodegradable components
in the feed to the system which passed through the aerobic biological
treatment into the effluent, or (b) the generation in the bioreactor of refrac-
tory secondary metabolites.
The first alternative was evaluated by measuring the biodegradability
of all the major contributors of COD in the Bay City effluent. The results of
this study indicate that all of the major components are readily biodegradable,
The second alternative was evaluated by starting a second activated
sludge unit in another area of the laboratory, in order to develop new sludge
flora. The results were inconclusive because the second unit initially
produced an effluent COD concentration which was relatively low but the
effluent COD increased to more than 1000 mg/1 while the first unit improved
and the effluent COD began to decrease. However, the overall performance
of the activated sludge system throughout the demonstration was relatively
uniform. COD removals were as follows: March -- 83.40± 1.45%; April --
81. 62 ± 2.51%; May -- 86.47 ± 1.60%; June -- 87. 63 ± 2.96%. The effluent
BOD analyses were performed by an independent laboratory, and the BOD
concentrations ranged between 24 and 50 mg/1 during this time period.
Nitrification did not occur to any substantial extent, and the effluent
nitrate N concentration was less than 10 mg/1 during most of the time
period. The outfall color and suspended solids, however, were relatively
high. Variations in the mixed liquor volatile suspended solids concentration
(MLVSS) over a wide range of about 2000 up to 9000 mg/1 did not markedly
affect the quality of the effluent.
A portion of the effluent of the activated sludge unit was processed
by distillation to concentrate and segregate the organic constituents in
order to simplify identification. Distillation was performed in a stainless
steel packed column. Three volatile fractions and the residue were
collected , and each sample was analyzed for COD and carbon. The analyses
presented in Table 8 indicate that approximately 31% of the COD in the
raw sample contained volatile components, while 43% of the COD was non-
volatile. The remaining 26% was unaccounted for and was assumed to be
entrapped in the distillation apparatus. Therefore, the volatile fraction
would contain 57% of the original COD. Organic carbon accountability was
poor; however, some useful information can be extracted. The major
portion of the organic carbon was in the residue. The ratio of the COD to
TOC ratio was 11.81 which is relatively high and indicated that the volatile
COD is primarily made up of non-carbon components such as amino nitrogen
containing compounds. Additional distillation was performed on the first
50
-------
TABLE 8. FRACTIONAL DISTILLATION OF FINAL BAY CITY PLANT
EFFLUENT
Sample
Raw Sample
Fraction- 1
Fraction- 2
Fraction- 3
Residue- 1
Total (2-5)
% Accountable
Total
NH3-N Carbon
(mg) (mg)
450 1874
281 306
45
20
1138
1509
81%
Inorganic
Carbon
(mg)
1436
273
34
14
600
921
64%
Organic
Carbon
(mg)
438
36
11
6
538
591
135%
COD*
(mg)
1450
425
13
13
625
1076
74%
* COD Analyzed in the Aquarator.
51
-------
fraction in order to determine the nature of the volatile COD. The fraction
was acidified with phosphoric acid to prevent distillation of any amines.
These data are presented in Table 9. The fractions collected (1-A)
would contain any volatile acids or neutral compounds. The COD in this
fraction was 29 mg. The residue remaining was made alkaline with sodium
hydroxide and redistilled. Under these conditions the amines would be
distilled over. The distillate (l-B)was collected in dilute hydrochloric
acid to prevent loss of volatile amines. The COD of this final fraction was
22% of the raw sample; however, the TOC was only 8 mg.- The inorganic
carbon remained in distillation flash as insoluble carbonates. The data do
not provide any conclusive data to substantiate the proposition of the
accumulation of secondary metabolites in the aerobic reactors.
b. Extended Aeration of Effluent from Activated Sludge Treatment
In view of the high residual COD and turbidity in the effluent of the
activated sludge system, extended aeration lagoon was added to the system
to provide eight day retention. The mechanical operation of this unit was
satisfactory and the clarity of the lagoon outflow was distinctly improved;
however, the yellow color remained unchanged and the effluent COD was
only marginally improved. The amino nitrogen compounds interfere with
the instrumental analysis (Aquarator) of COD and result in COD values
which are higher than those observed using the classical COD reflux method.
This phenomenon may explain the reason for higher instrumental COD values
presented in Table 10 which includes analysis of the Bay City plant extended
aeration effluent.
This final series of laboratory demonstration runs was an effort to
combine the most probable sequence of biological treatment that had been
attempted during the entire study. Segregation of the various types of
compatible feed to biological systems after pretreatment to remove known
toxic compounds or excessive carbon loading provided considerable improve-
ment in final effluent. However, even after the extended aeration, this
system did not obtain satisfactory outfall except for a single value of
7 mg/1 BOD5. The other parameters of color, TOC and COD are excessive
for discharge to surface water.
These results shown in Table 10 are the best results obtained under
ideal operating periods. The final outfall from the Bay City plant to the
Colorado River must meet a maximum concentration of 90 mg/1 TSS,
25 mg/1 as BOD and 50 mg/1 as TOC.
52
-------
TABLE 9. DISTILLATION OF FRACTION-1
Total Inorganic Organic
Organic Carbon Carbon NH3-N COD"
(mg) (ing) (mg) (mg) (mg)_
Fraction- 1
Fraction- 1-A (H3PO4)
Fraction- 1-B (NaOH)
Total (2-3)
*.'s
% Accountable
306
35
8
43
14%
273
23
< 1
<24
9%
36 281
12
8 187
20 187
56% 67%
425
29
320
349
82%
* Loss in carbon accountability primarily due to formation of carbonates
after caustic addition.
** COD Analyzed with an Aquarator.
53
-------
TABLE 10. ANALYSIS OF BAY CITY PLANT EXTENDED AERATED EFFLUENT
(Values as mg/1.)
Sample
Number'1"
1
2
3
T.C.
1874
1639
1387
T.I.C.
1436
1384
1233
T.O.C.
438
255
154
Inst. COD""
1450
1320
625
4Hr
COD*** BOD5 BOD2o
- -
851 7 30
609
* CCCTC Code Numbers are as follows:
1 - 22882-4
2 - 22905-23- 11, 12, 13, and 14 22918 11- 15 and 16
3- 22918-48 23, 26, 28, 30, 2
** Inst. COD = Aquarator COD
*** 4 hr. COD = 4 hr. reflux dichromate method.
54
-------
APPENDIX A
STREAM SEGREGATION
Aerobic biodegradability of the individual compounds was evaluated
and these data are presented in Table A- 1. The results of tests using Petri
cultures treated with each of the wastewater components and published
information were used to select the individual wastewater streams to
be included in the final feed blend for each of the three unit treatment
processes. The biodegradability of individual streams under anaerobic and
aerobic conditions also was evaluated. These data are presented in Table
A-2 for fifteen of the individual wastewater streams which are included in
the combined process wastewater stream. Anaerobic pretreatment of a
wastewater stream would not always promote improved efficiency of waste-
water stabilization. All of the streams with the exception of stream number
15 which contains aldehydes and alcohol were anaerobically biodegradable.
However, a number of the streams which contain amines, ammonia, organic
acids, some chlorinated organic compounds, phosphoric acid, aldol products,
aldehydes and alcohols and one caustic stream were not aerobically bio-
degradable in spite of the anaerobic pretreatment.
The form of nitrogen and relative nitrogen requirements for the
anaerobic and aerobic processes also were briefly investigated. Data
presented in Table A- 3 indicate that the mixed culture in the anaerobic
reactors favored nitrate nitrogen while ammonia or amines was preferred
in the activated sludge process.
The combined process wastewater can be classified in the three
basic categories; namely, (a) streams containing high organic carbon, (b)
streams containing high nitrogen concentrations, and (c) a third component
which generally originates from sumps throughout the plant which contained
high concentrations of heavy metals.
The elimination of one low volume wastewater stream -which contained
about half of the total organic carbon in the combined process wastewater
markedly improved the degradability of the composite wastewater stream
in the anaerobic/aerobic system. The composition of the modified process
wastewater blend is presented in Table A-4.
55
-------
TABLE A-l. RELATIVE AEROBIC BIODEGRADATION RATES FOR SPECIFIC
COMPOUNDS IN THE PROCESS WASTEWATER
ppm-TOC
Compounds Removal /Day
Acctaldchydc 745
Crotonaldchyde 305
Cyclohexanone 25
Acetic Acid 92
Propionic Acid 180
Butyric Acid 49
Adipic Acid 27*
Methanol 250
n-Butanol 280
Cyclohexanol 152
2-Ethylhexanol 76
Vinyl Acetate 168
Hexamethylenediaminc 276
1,6-Hexanediol 235
Nylon Salt 273
*Aerobic studies with adipic acid indicated that the adipic molecule
was consumed entirely as opposed to partial degradation.
56
-------
TABLE A-2. STREAM BIODEGRADABILITY TESTS
Stream
No.
I
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Major
Component
Kerosene
Amines
Nitric, organic acids
Caustic
Phosphoric acid, aidol
products
Amines
Organic acids
Alcohol
Chlorinated organics
Chlorinated organics
Chlorinated organics
Organic acids, aldol
products
Esters
Amines, ammonia
Aldehydes, alcohols
Aerobic
Biodegradable
yes
no
yes
no
no
yes
no
yes
yes
no
yes
yes
yes
no
no
Anaerobic
Biodegradable
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
no
57
-------
TABLE A-3. SOLUBLE NITROGEN REQUIREMENTS FOR WASTE TREATMENT
Anaerobic Filter
Activated Sludge
mg/1 Nitrogen Utilized/ mg/1 Nitrogen Utilized/
Form of Nitrogen mg/1 BOD Removed mg/1 BOD Removed
Ammonia- Nitrogen
NH3-N
0. 11
0. 07
Nitrite- Nitrogen
NO2-N
<0.01
<0. 01
Nitrate- Nitrogen
NO3-N
2. 10
<0. 01
58
-------
TABLE A-4. COMPOSITION OF COMBINED PROCESS WASTEWATER AT BAY CITY PLANT
AFTER ELIMINATION OF LOW VOLUME HIGH CARBON STREAM
Concentration
mo/1
Total Organic Carbon
Biochemical Oxygen Demand
Chemical Oxygen Demand
Nitrate-Nitrogen
Nitrite Nitrogen
Ammonia-Nitrogen
Phosphate-Phosphorus
Chloroaldehydes
Inorganic Chlorides
Sodium
Copper
Iron
Chromium
Manganese
Palladium
Nickel
Cobalt
pH
Color, APHA
5,000
4,500
21,000
VSO
27
770
130
2,085
35-400
20
2-100
°'02
<0.2
<0>2
<0-4
<0-02
7,500-10,000
(a) pH units
(b) color units
59
-------
The results of the batch anaerobic/aerobic treatability test based on
the removal of soluble TOC indicate that all of the individual process waste-
water streams containing organics with one exception can be treated
anaerobically0 One stream also was anaerobically treatable following
neutralization. Approximately 75% (volume) of the high strength wastewater
streams was treatable aerobically and with neutralization this volume was
increased to 80%„ Anaerobic pretreatment increased the aerobic treatability
of four of the individual wastewater streams, but anaerobic pretreatment
reduced the aerobic treatability of three other wastewater streams .
The performance of the three stage anaerobic/aerobic biological
treatment system is markedly affected by variations in the composition and
flow of wastewaters to the system. Therefore, the variability in the carbon
concentration and the concentration of specific components of 62 streams
were evaluated. These data were analyzed in an attempt to develop the best
technical treatment or disposal scheme for each individual wastewater
stream. Streams were classified according to the process which would best
reduce the pollution potential; namely, (a) incineration, (b) conventional
biological treatment; (c) physical/chemical pretreatment; (d) specialized
high strength biological treatment; and (e) reuse within the plant.
Categorization of the wastewater streams was based on the following criteria:
(a) carbon concentration; (b) presence of components potentially toxic to
biodegradation; (c) for potential reclaiming product; and (d) the variability
coefficient of the components in the wastewater stream.
Wastewater streams containing a soluble carbon concentration with
a variability coefficient of more than 50% were considered to require
additional process control prior to being acceptable as feed to a biological
treatment. Toxic streams would either be detoxified, incinerated or
pretreated with the physical/chemical process. Results of this analysis
indicated that nine or ten of the streams considered as feed to the high
strength biological treatment process have carbon variabilities in excess
of 50% of the mean carbon concentration for the individual wastewater
stream. High variability in the carbon concentration coupled with varied
flow rates indicated that the control of the quality of the biological feed blend
was almost impossible. The greatest degree of variability in the concen-
tration of the components as well as flow was observed in those streams
generated from batch operations or cyclic work schedules. The wastewater
from tank car washings, maintenance repair areas, and chemical sumps
exhibited the most variability in carbon concentration, component concen-
tration, and flow volume.
60
-------
APPENDIX B
WASTE REDUCTION
1. CHLORINATED ORGANICS
The toxicity to biological systems of these compounds at concen-
trations normally found in the process wastewater indicates that this type of
waste must be handled by some physical/chemical system. Incineration is
considered as the best alternative. Process modifications to minimize
water utilization result in increasing the carbon concentration in the waste-
water stream to approximately 10% by weight prior to incineration. The
possibility of by-product recovery of HC1 adds incentive for employing
incineration.
2. HIGH CARBON CONCENTRATION WASTEWATER
Most of the process wastewater streams can be classified into two
categories based on TOC. Most wastewater streams with a TOC concentra-
tion in excess of 60, 000 mg/1 will be considered for incineration while those
wastewater streams with a TOC concentration less than 60, 000 mg/1 will be
treated biologically. It is anticipated that the recovery of heat will reduce
the minimum supplemental fuel required to maintain the temperatures
necessary for complete combustion. However, most of the plant wastes
which have high carbon concentrations typically contain large amounts of ash
or nitrogen containing compounds. In-process modification can reduce
sources of ash; however, a NOx abatement system will be required to remove
the nitrogenous combustion products .
3. VOLATILE LIGHT ENDS
Water phases from decanters contain soluble volatile organics and
insoluble oil resulting from incomplete phase separation. The results of
laboratory studies indicate that the installation of strippers can alleviate
this problem. The strippers provide additional product recovery, are
effective in smoothing carbon variations, and can be effective in reducing
the carbon load of wastewater streams although partially offset by energy
required. Wastewater streams containing small amounts of vinyl acetate,
cyclohexane, cyclohexanol and cyclohexanone are suitable for stripping.
61
-------
4. HEAVY METALS
The main sources of heavy metals in the process wastewater are the
process unit sumps. This problem is common to all plant operating areas.
Therefore, all continuous discharges to unit sumps will be eliminated by
transporting these discharges directly to the ultimate destination. Some
type of centralized physical/chemical treatment scheme will be necessary
to handle intermittent sump discharges. The exact physical/chemical
treatment scheme has not been developed, but will more than likely include
one or more of the following processes: pH adjustment, sulfide addition,
coagulation, sedimentation, filtration, and ion exchange. Discharge from
the physical/chemical treatment will be amenable to biological treatment.
Heavy metal residuals will be suitable for disposal in a registered sanitarv
landfill. y
5. AMINES
Amines are typically toxic and/or inhibitory to biological systems
and may be present at high enough concentrations to exceed the nutrient
requirement of the biological system resulting in unacceptable concentrations
of nitrogen in the final treated effluent. Incineration is currently considered
as a means of handling the amine containing streams with the possible
exception of sump discharges. NOx abatement systems to treat the com-
bustion products will be necessary.
6. NITRATES
Nitrates are one of the most common constituents in the wastewater
from the adipic acid unit; however, spills or leaks in other plant units are
also a source of nitrates. The concentration of nitrates usually is in excess
of nutrient demand of biological systems; therefore, anaerobic denitrification
of the nitrate containing streams offers a possible solution. The waste
reduction and stream segregation programs led to three categories of
process wastewaters, and the evaluation of these streams by a series of
anaerobic denitrification, anaerobic methanogenic and aerobic activated
sludge processes.
62
-------
APPENDIX C
DESCRIPTION OF BENCH SCALE REACTORS
1. AEROBIC BIOLOGICAL TREATMENT (OXYGEN)
The two-stage oxygenated activated sludge system consisted of
duplicate aeration tanks and clarifiers operated in series. The first stage
of the two-stage system is illustrated in Figure C- 1. The combined process
wastewater blend with nutrients added at a ratio of COD:N:P equal 100:5:1
was stored in a 20-liter reservoir constructed of a six-inch diameter, four-
ft section of Pyrex glass pipe flanged at the bottom and calibrated to facilitate
accurate measurement of the wastewater feed to the aeration unit. A
peristaltic pump (Cole-Farmer Masterflex) fed the wastewater from the
reservoir to the aeration tank. The wastewater feed was intermittent and
controlled by timer (Tork Timer 60-minute repeater timer). A burette
was located immediately upstream of the pump and provided a means of
quickly estimating pumping rate at any pump speed by feeding from burette
over a controlled time period.
The aeration tank was constructed of a Plexiglas tube which was
8-inch in diameter and 4-ft long with 1/8" wall thickness. The Plexiglas
tube was held between upper and bottom machined stainless steel plates.
Oxygen was introduced near the bottom of the aeration tank through three
fritted glass spargers which were connected by means of Swagelok fittings
to a 1/4" diameter stainless steel tubing manifold. The oxygen was presatu-
rated with moisture in a wash bottle and the flow measured by means of a
rotameter. A pressure gauge was installed in the oxygen line to measure
increases in the pressure drops across spargers as biomass accumulated
on the spargers. Agitation was provided by three shipscrew shaped impellers
mounted on a vertical shaft extending the full length of the aeration tank axis.
These impellers were driven by a Fisher Dyna-Mix variable speed stirrer.
An opening in the top plate provided access for sampling of the biomass and
offgas for introduction of the dissolved oxygen probe and for the returned
sludge line.
Mixed liquor from the aeration tank overflowed through a tube to a
clarifier which also was constructed of 4-inch Plexiglas tubing and fitted
with a conical bottom. The overflow tube was so located that the volume
63
-------
o
Pump
and
Controls
Figure C-l. First stage of two-stage bench-scale
oxygen-aerated activated sludge unit.
64
-------
of the mixed liquor in the aeration tank was 15 liters,, Sludge which
collected at the bottom of the clarifier was recirculated to the aeration
tank by means of a peristaltic pump operated by a timer. A three-way
solenoid valve activated by a timer was installed in the sludge return line,
This device permitted for wasting of the sludge, but during operation the
valve tended to plug and become inoperative„ Therefore, sludge wasting
was done manually. A device consisting of three windshield wiper blades
mounted on a slowly revolving shaft (4 rph) also was installed in the clarifier
to prevent bridging of the sludge in the clarifier,, Clarified effluent from
the first stage was pumped into the second stage or discharged to waste
from the second stage.
This unit was started with a domestic activated sludge which was
initially fed a wastewater consisting of easily degradable organics; namely,
two grams per liter each of malt extract, lactose, dextrose, peptone, and
yeast extract. The combined process wastewater from the Bay City plant
was added to this initial feed material in increasing amounts until feed
consisted of 100% of the process wastewater. This unit also was operated
with a number of other individual process wastewater components.
The aerated activated sludge systems operated at the Bay City plant
and the CCCTC consisted of a reactor similar to the unit illustrated in
Figure C-2. A peristaltic pump was used to transfer the process waste-
water blend with nutrients added to the aeration chamber. Air was intro-
duced to the aeration chamber through a diffuser stone located at the bottom
of the unit. The flow of air was controlled by a needle valve. An adjust-
able baffle separated the aeration chamber from the clarification section.
The volume of the mixed liquor in the aeration chamber was approximately
seven liters and the volume of the clarification zone was three liters. The
clarified effluent was discharged through an adjustable overflow tube.
2. ANAEROBIC METHANOGENIC TREATMENT
The initial attempts at evaluating the anaerobic treatability of the
combined process wastewater were performed at Bay City in an anaerobic
mixed liquor contact system. A 12-liter round bottom flask was the
reactor. The flask was fitted with a heating mantle and an agitator.
Provisions were made for measuring offgas, withdrawal of effluent and
addition of fresh feed. The unit was operated in a batchwise fashion with
daily withdrawal of mixed liquor and daily addition of combined process
wastewater blend. Neutralization of the feed was required to maintain a
pH in the range of pH = 6.7 to pH =7.3 necessary for methanogenesis. The
cost of neutralization resulted in the abandoning of the anaerobic mixed
liquor contact system.
65
-------
FEED BOTTLE
PUMP
INFLUENT
FEED LINE
^-ADJUSTABLE
OVERFLOW
WEIR
EFFLUENT
BOTTLE
Figure C-2. Continuous Simulation Reactors.
66
-------
Additional anaerobic studies were continued at CCCTC using an
anaerobic packed bed. A schematic diagram of the initial design of the
anaerobic filter is illustrated in Figure C-3. The reactor consisted of a
4-ft long section of 6-inch diameter Pyrex industrial glass pipe flanged at
both ends. The pipe was filled with procelain Berl saddles as packing
material. The pipe was wrapped with heating tape and the temperature
controlled by a powerstat. Temperature in the anaerobic filter was
monitored with a thermocouple. A gas separation bulb was located at the
top of the filter and permitted separation of gas from the effluent. The gas
passed through a gas sampling ampoule and the volume of gas produced was
recorded on a wet test meter. The effluent passed through a flow splitter
which diverted a portion of the effluent to a surge vessel which consisted
of a 12-liter round bottom flask. The purpose of the surge vessel was to
provide some neutralization of the process wastewater feed. The logic
behind this concept involved the fact that under steady state conditions the
organic material in the process wastewater would be converted to methane
and carbon dioxide. Therefore, the liquor in the filter would consist of a
solution of bicarbonates which are capable of neutralizing the incoming
acid. The surge vessel also contained a supply of solids from an anaerobic
digester which slowly wash into the filter. Combined process wastewater
blend with some caustic added was pumped into the surge vessel. This
mixture of process wastewater, caustic, recycled filtered effluent, and
digested solids was pumped through the reactor. During startup operations,
the anaerobic filter was seeded with digested sludge from the municipal
wastewater treatment plant and fed a synthetic substrate consisting
primarily of sugars and methanol as indicated in Table C- 1. The system
was considered acclimated when several liters of gas were produced per
liter of feed and the methane to carbon dioxide CH4:CO2 was approximately
3:1. At this point the combined process wastewater blend from the Bay City
plant was introduced in increasing amounts. The gas production dropped off
drastically when the combined process wastewater blend reached about 30%
by volume of the total feed.
The inability of aerobic or anaerobic processes to satisfactorily
treat the combined process wastewater blend resulted in the segregation
of the combined process wastewater stream into three composites; namely,
a) composite A which included the nitrate containing streams but was free
of amines and organic chlorides; b) composite B containing streams which
were free from nitrates, organic chlorides, and amines; and c) composite
C which included streams which contained ammonia, amines, and organic
chlorides. Composite B was considered to be amenable to anaerobic
methanogenesis without pretreatment. Composite A also was considered
to be suitable for anaerobic treatment but only after pretreatment by
anaerobic denitrification. The last composite blend, composite C, can be
treated by aerobic activated sludge. These components were made into a
67
-------
TABLE C-l. SYNTHETIC STARTER FEED FOR ANAEROBIC FILTERS
Solution A: 200 g MeOH in 22 liters tap water
Solution B: malt extract c a
~> 5
yeast extract 5 „
peptone 5 g
urea 9>5 g
H3P04(85%) 4<8g
Na acetate, anhydrous 33. 8 g
MgS04- 7 H20 1>7 g
68
-------
First Design
Wet Test Meter
Gas Separation
Bulb
Gas Sampling
Ampoule
Packing
(porcelain
Berl
saddles)
Powers tat
Surge
Vessel
Heating
Mantle
Recirculation
Pump
Thermocouple
Figure C-3. Anaerobic packed bed.
69
-------
volume of 165 ml with water and mixed with solution A at the rate of
7.5 ml per liter. Preparation of a concentrated solution B gives it shelf life
and largely prevents the spontaneous fermentation of the starter.
A much less complex starter solution was evolved using methanol
and acetic acids as the only organics, and with COD:N:P = 1000:5:1. The
nutrients were supplied as urea for N, phosphoric acid for P and used
sodium sulfate for the sulfur required to form insoluble sulfide.
The segregation program led to reevaluation of the capability of
anaerobic and aerobic processes for treating individual components. A
revised backmixed anaerobic filter was constructed and is described
schematically in Figure C-4. This concept of backmixed anaerobic filter
had been designed and operated successfully using wastewater from a
Celanese Chemical Company plant in Pampa, Texas. The filter proper
consisted of a 4-ft section of 6" diameter Pyrex industrial glass pipe
flanged at both ends. The gross volume of the pipe was 20 liters but the
void volume was reduced to 15 liters after filling the pipe with 1-inch
ceramic Raschig rings. The detention time and volumetric loadings were
calculated on the basis of this void volume although in the course of a run
the accumulation of biomass in the filter reduced the void volume. Heating
tapes connected to a powerstat were used to heat the filter. The entire
pipe covered with glass wool pipe insulation to reduce heat losses as
well as to exclude light which would lead to the appearance of photosynthetic
organisms, and the oxygen produced would be inhibitory to the anaerobic
organisms. Temperature in the filter was measured by a single thermo-
couple which was considered sufficient since the system was backmixed.
The combined process wastewater effluent with nutrients and reserve
alkalinity added was stored in a 20-liter plastic feed tank. A peristaltic
pump actuated by a timer pumped the feed blend into the recirculation loop.
The fluid in this loop was pumped continuously into the bottom of the filter
by a recirculation pump at a rate of about 150 liters per day which
corresponds to approximately 10 void volumes per day. A check valve •was
installed on the recirculation loop immediately before the fluid entered the
filter in order to prevent loss of mixed liquor from the filter in the event
of a leak in the recirculation loop. The filter effluent and gas generated in
the filter passed through a gas separation bulb. The gas passed through a
sampling ampoule and the volume of the gas produced was recorded on a
wet test meter. The effluent of the system overflowed from the gas
separation bulb through a liquid seal. The sampling port was installed in
the liquid seal to permit sampling of effluent for pH measurements. If the
effluent was allowed to remain in contact with the air for even a few minutes
dissolved CO2 was lost and the pH rose between 0. 5 and 1. 0 pH units. The
majority of the liquid leaving the filter reentered the recirculation loop
70
-------
Final Design
Gas Meter Gas
Sampling
Gas
Separation
Liquid Seal
Timer Controller
Pump
Effluent
Heating Tape
Variac
Thermocouple
Figure C-4. Backmixed anaerobic filter.
71
-------
from the gas separation bulb. A funnel and two pinch clamps were installed
in the recirculation loop to permit the introduction of sludge seed, nutrients,
and other liquids to the system bypassing the feed reservoir. The per-
formance of the anaerobic filter was evaluated in terms of gas production
and changes in the concentration of total carbon, inorganic carbon, total
organic carbon, and chemical oxygen demand. The gas produced was
analyzed for CO2 and CH4 by mass spectrometer. Heavy metal and calcium
concentrations in the effluent also were monitored. Calcium carbonate was
added to the feed reservoir to provide buffering. A significant portion of
the calcium carbonate precipitated in the filter.
Startup of the anaerobic methanogenic packed bed involved seeding
with anaerobically digested sludge and feeding a synthetic substrate, the
composition of which is listed in Table C- 1. An acclimated anaerobic
population developed as indicated by production of several liters of
gas per liter of feed and a CH4:CO2 ratio of 3:1 was achieved. A
composite B blend was introduced to the system in increasing amounts until
the feed was made up only of composite B. The anaerobic methanogenic
filter also was fed a blend of composite B and the denitrified composite A.
3. ANAEROBIC DENITRIFICATION
The bench scale anaerobic denitrification studies were conducted in
backmixed filter illustrated in Figure C-4 and described above. The startup
procedure was essentially the same as that described for the anaerobic
methanogenic system with the exception that the synthetic substrate contained
5 grams per liter of sodium nitrate (0. 82 grams per liter of nitrate-N).
Composite A blend was then incorporated in the feed in increasing pro-
portions until the feed was essentially 100% composite A. The feed also
was enriched with composite B to the extent of about 10% by volume since
the degree of the biodegradability of the organics in composite A was
unknown and the concentration of nitrates in composite A was relatively high.
This procedure insured the presence of excess COD relative to the cor-
responding oxidizing capacity of the nitrates. This approach was satisfactory
for some time but some unexplainable upsets occurred. The fact that the
denitrification unit survived these upsets and was able to recover, usually
in a matter of days, was a good indication of potential application of this
process in the prototype system.
The anaerobic denitrification system had been modified and the
packing material was replaced initially with sand, and finally with granular
activated carbon and operated in a fluidized bed fashion. A schematic flow
diagram of the fluidized bed anaerobic is illustrated in Figure C-5. The
main reactor of the system consisted of a cylindrical section about 7 cm in
diameter with a volume of about 1. 64 liters. Cylindrical section was
72
-------
Vent
Recycle
Sample Port
Recycle
Pump
Outfall
1640 ml
30-40 Mesh
Sand
Drain
Pump
Feed
Reservoir
Timer
Figure C-5. Fluidized bed denitrification anaerobic filter.
73
-------
wrapped with a heating tape and the temperature controlled by means of a
thermocouple. The upper end of the fluidized bed section expands into a
bulb made of a 5-liter round bottom flash. An effluent overflow port and a
gas outlet were installed in the overhead section. The gas passed through
a gas sampling ampoule, and the volume was recorded on a wet test meter.
The effluent was recirculated from the overhead section to the bottom of the
fluidized bed by peristaltic pump. A sampling port also was installed in the
recirculation loop. The wastewater feed was stored in a calibrated
reservoir and pumped to the bottom of the reactor by a time actuated
peristaltic pump. The fluidized bed consisted of 20 to 40 mesh Ottawa sand
which filled two-thirds of the volume of the reactor. Fluidization caused the
bed to expand two inches. Sand particles became wedged in the lower portion
of the fluidized bed section and were difficult to dislodge. This wedging
became more pronounced with time and eventually forced the shutdown
of the unit. The unit was rebuilt and granular activated carbon was used as
a fluidized bed.
This fluidized bed system for denitrification was a modification of a
process described by Dr. J. S. Jeris (* ). His system was operated as a
single pass system with fluidized sand bed. The process developed in this
study uses high recycle flow through a granulated carbon bed which
provides dilution of high strength waste to provide high rate denitrification.
This modification may be considered for patent application under the grant
study.
4. PRETREATMENT
The composition of the process wastewaters to be treated as a result
of the waste reduction and segregation programs made it necessary to
operate a laboratory scale stripping column and heavy metals removal
system to simulate the wastewaters which would be treated in anaerobic or
aerobic biological systems. A laboratory scale stripping column was con-
structed with a 30-tray 2" Oldershaw column with feed injected on
Tray 20. The column overhead was equipped with a decanter allowing the
overhead water phase to be totally refluxed. The overhead oil phase was
collected but not refluxed. The primary purpose of the stripper was to
reduce the carbon content in the composite C blend. The input to this
column contained the relatively low carbon stream and the sodium contain-
ing streams from the cyclohexane oxidation and cyclohexanone units. These
streams constituted approximately 55% of the total flow of the composite C
blend. The column feed was composited according to the average flow rates
of the stream.
Jeris, J.S., and R. G.Owens, "Pilot Scale High Rate Biological Denitrifica-
tion at Nassau County, N. Y.". Presented at the New York Pollution Control
Association, Winter Meeting, January 21-23, 1974.
74
-------
Heavy metal removal was accomplished by precipitating the metals
as the metal sulfide and decanting the liquid for further treatment. The
streams requiring heavy metal removal contained relatively low con-
centrations of nitrates and carbon. However, the heavy metals posed a
toxicity problem for the aerobic biological treatment units. The streams
requiring heavy metal removal have a highly variable flow and include
intermittent contributors such as rainfall, runoff, wash- downs and sump
discharges.
Composite samples of the effluents of the stripper column and the
heavy metal removal system were fed to the activated sludge system. The
composite sample of the heavy metal containing streams also was fed with-
out metal removal to the anaerobic filters with the addition of sufficient
sulfates to encourage the formation of insoluble heavy metal sulfides under
anaerobic conditions.
75
-------
APPENDIX D
AEROBIC TREATABILITY PERFORMANCE DATA
TABLE D-l. COMPOSITION OF BAY CITY "AEROBIC" COMPOSITE
(FREE OF NITRATE- AND ORGANOCHLORINE-CONTAINING WASTES
pH
TC (mg/1)
1C (mg/1)
TOC (mg/1)
COD (mg/1)
Total N (Kjehldahl) (mg/1)
Nitrate N (mg/1)
Ammonia N (mg/1)
Sodium (mg/1)
Drum No. 1
20673-23-1
4.5
7200
<10
7200
21 400
185
50
89
710
Drum No. 2
20673-23-2
4.5
8000
<10
8000
20 200
155
60
86
730
DrumNo. 3
20673-23-3
4.3
10 080
10
10 070
18 200
150
45
105
800
DrumNo. 4
20673-23-4
4.3
9750
5
9780
17 200
155
48
100
760
76
-------
TABLE D-2. ACTIVATED SLUDGE TREATMENT OF BAY CITY EFFLUENT
(First Stage)
Date
20 Mar
21
22
23
24
25
26
27
28
29
30
31
1 Apr
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Feed
COD, Feed,
ml /I ml /day
3383
4034
8950 3790
4006
3881
4085
3954
8750 3756
3807
3750
3920
3881
3722
3443
9500 3699
3545
3443
3523
3454
3352
6450 3334
3352
3295
3290
3210
3188
3193
6400 3045
3057
3073
2955
3335
Detention,
day
1.5
1.5
1.5
1.6
1.6
1.6
1.5
1.7
1.7
1.7
1.7
1.8
1.8
1.8
1.8
1.8
1.9
1.9
2.0
2.0
1.9
1.9
Loading,
Ib/cu ft /day
1.08
0. 38
0. 37
0.35
0. 34
0. 35
.0. 36
0. 32
0. 33
0. 36
0. 36
0. 34
0.34
0. 33
0. 33
0.23
0.22
0.21
0.20
0.20
0.21
0.21
Outfall,
COD
1500
1600
1460
1500
1700
1600
2000
1850
1700
1400
1350
1300
1050
% COD
Removal
83.2
82. 1
83.7
82.9
80.6
81.7
78.9
80.5
82. 1
78. 3
79. 1
79.8
83.7
(continued)
77
-------
TABLE D-2. (CONTINUED)
Date
21 Apr
22
23
24
25
26
27
28
29
30
1 May
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Feed
COD,
ml /I
5600
8750
8750
7750
6050
4750
3700
4400
3600
Feed,
ml /day
2989
4216
4199
4079
4210
4142
4619
3846
4131
4170
4244
4188
2875
4278
4250
4267
4244
4295
4290
4239
4358
4329
4148
4227
4250
4199
4273
4159
4170
4040
4136
4045
4119
4057
4307
4051
4148
4114
Detention,
day lb
2.0
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
2.1
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.5
1.5
1.5
1.5
1.4
1.5
1.5
1.5
1.4
1.5
(continued)
78
Loading,
/cu ft /day
0.20
0.28
0.24
0.24
0.24
0.24
0.24
0.38
0. 38
0.38
0.15
0. 38
0.39
0. 39
0.39
0.39
0. 35
0. 34
0.34
0.26
0.26
0.27
0.21
0.21
0.20
0. 16
0. 16
0. 16
0. 16
0. 19
0. 19
0. 19
0. 15
0. 15
Outfall,
COD
1050
1100
1100
973
1100
1350
1150
1150
1100
860
732
738
660
670
595
600
670
% COD
Removal
83.7
8.29
80.4
82.6
87.4
85.7
86.9
86.9
87.4
88.9
87.9
87.8
86. 1
85.9
87.5
86.4
84.8
-------
TABLE D-2. (CONTINUED)
Date
29 May
30
31
1 Jun
2
3
4
5
6
7
8
9
10
11
Feed
COD, Feed,
ml /I ml /day
3450 4096
4221
4040
4250
4600 4381
4131
4108
6300 4330
3710
4244
3960
6250 4119
4063
4023
Detention,
day
1.5
1.5
1.4
1.4
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
Loading,
Ib/cu ft /day
0. 15
0. 15
0.15
0. 15
0.20
0.20
0.23
0.27
0.27
0.26
0.27
0.27
Outfall,
COD
623
695
550
673
683
685
% COD
Removal
81.9
79.9
88.0
89. 3
89.2
89.0
79
-------
TABLE D-3. ACTIVATED SLUDGE TREATMENT OF BAY CITY EFFLUENT
(Second Stage)
Date
9 May
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
1 Jun
2
3
4
5
6
7
8
9
Feed
COD, Feed,
ml /I ml /day
7.75
398
460
714
665
869
767
1085
4.75 1106
1170
1190
1108
3.70 1364
693
1659
1687
4.40 1761
1722
1744
1733
3.45 1721
1807
1943
4.60 2045
2433
2352
2960
6.30 3181
2682
3273
6.30 2966
3307
Detention,
day Ib
12.4
13. 1
8.6
8. 8
7. 0
7.9
5.5
5. 3
8.5
3.6
3.4
3.5
3.5
3.4
3. 5
3.5
3.4
3.0
2.9
2.5
2.5
2.0
2.0
2. 1
1.9
1.9
1.8
(continued)
Loading, Outfall, % COD
/cu ft/day COD Removal
0.04
0. 04
0.06 170
0. 05
0. 07 125
0. 06
0.09 50
0. 09
< 50
10
0. 03
0. 06
0. 07
0. 07
0. 08
0. 08
0. 08 50
0. 08
0. 08 < 50
0.09
0.09
0. 11 < 50
0. 11
0. 14 < 50
0. 14
0. 19 < 50
0.21
0.20
0.21 245 96.1
80
-------
TABLE D-3. (CONTINUED)
Date
10 Jun
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
1 Jul
2
Feed
COD, Feed,
ml /I ml /day
6.70 ?
3455
3216
3256
3239
3324
7.15 3188
3506
3455
3148
3381
3426
3244
6.75 2915
2830
3244
3449
3347
3511
3358
3369
3278
3403
Detention,
day
1.8
1.9
1.8
1.8
1.9
1.8
1.7
1.8
1.9
1.8
1.8
1.9
1.9
1.8
1.8
1.7
1.8
1.8
1.8
1.8
1.8
1.8
Loading,
Ib/cu ft /day
0.23
0.22
0.23
0.23
0.23
0.23
0.26
0.25
0.24
0.25
0.25
0.24
0.23
0.24
0.23
0.24
0.24
0.24
0.24
0.24
0.23
0.24
Outfall,
COD
350
485
1700
2600
3750
2600
1035
970
1200
% COD
Removal
94.8
92.8
74.6
61.2
47.6
63.6
84.7
85.6
82.2
81
-------
APPENDIX E
DENITRIFICATION IN PACKED BEDS
TABLE E-l. COMPOSITION OF BAY CITY NITRATE-CONTAINING EFFLUENT COMPOSITE
(Analytical Summary)
PH
TC (mg/1)
1C ;mg/l)
TOC (mg/1)
COD (mg/1)
Total N (Kjehldahl) frng/1)
Nitrate N frng/1)
Nitrate N (mg/1)
Copper (mg/1)
20673-6-1
1. 1
1470
200
1270
310o(a)
2650
2400
180
33
20673-6-2
1. 1
1520
200
1320
390o(a)
2800
2600
220
35
20673-46-1
1.45
1640
240
1400
498o(a)
2040
2650
6
26
20673-46-2
1.45
1690
210
1480
498fl(a)
1790
1600
52
29
22754-8-1
1.1
1810
260
1550
6ooo(a)
1550
1500
4
20
(a) Wet chemical analysis (dichromate oxidation).
82
-------
00
16
60
Q Feed, Liter»/Day, Scale A
Q Offgai. Liters/Day. Scale B
A Mixed Liquor pll. Scale
70 7*
M»r i A).r I
Figure E-l. Packed bed denitrification unit.
-------
00
ADC
May 1
Figure E-l. (Continued)
-------
oo
On
0.3
Hydraulic Detention, D»y«, Scale A
COO Loading, Lb« COD/Cu Ft/Day,
20
35
75
Figure E-2. Packed bed denitrification unit.
-------
oo
loo HIS no Hi uo izs nu m
Figure E-2. (Continued)
-------
00
-J
APPENDIX F
ANAEROBIC METHANOGENIC TREATMENT
(13.6 Liter Initial Void Volume, Operating Temperature - 37°C, pH, Hydraulic Detentions, Specific Gas Yield)
y
me, day*. Void-volume of the filter
(at «tartup)/daily feed (normalized I.
Specific gaa yield; daily off gaa volume {read at 23 +
1 "C, not corrected for-variationa in atm. pressure)/
daily feed volume.
4 Solid line « indicate percentages of composite *B"
in the feed.
10 13 20 25 3*> 1*> 40 4^ ?0 ?>' 60 65 TO 75 80
Figure F-l. Operating parameters for backmixed anaerobic filter.
-------
00
oo
G Detention time, days. Void volume of the filter,
at startup)/daily feed (normalized)
A Specific gas yield; daily off gas volume (read at 23 +
1 C, not corrected for variations in atm. pressure)/
daily feed volume.
Solid lines indicate percentages of composite "B
in the feed.
100 105 110 105 120 125 130
160
Figure F-l. (Continued)
-------
00
O Detention time, days. Void volume of the filter,
(at startunV "aily feed (normalized).
; daily off gas volume (read at
rected for variations in atm.
ed volume.
te percentages of composite B
^
ISO Ie5 !70
190 l^S 200 20-;
Figure F-l. (Continued)
230 Z35 24P
-------
O Detention time, days. Void volume of the filter, (at startup}/
daily feed (normalized)
10
A Specific gas yield; daily off gas volume (read at 23 t l'C,not
corrected for variations in atrn, pressure)/ daily feed volume.
Solid line indicates percentages of composite "B" in the feed.
240 245
250 255
260
265 270
Figure F-l. (Continued)
-------
(13.6 Liter Initial Void Volume, Operating Temperature = 37°C, Off-Gas Composition, TOG and COD Removal)
<3> Methane in off gas, percent x.
A Grams COD removed per liter of feed; calculated f»>w. ^pwrtfr ^ frj
where v~ = specific gas yield, CCH* = mole fraction of methane in
off gas, .it.- moles volume at conditions of measurement, and 64 = g.
of oxygen required to oxidize one mole of CH4 to COj * H2O.
©Percent TOO removal efficiency TOC in outfall ^ ^Q
TOC in feed
O Percent COD removal efficiency, from^px CCr^ x 64 ^ ^^
£i x COD in feed
This expression can be > 100, and is valid only under lined-out
conditions.
Solid lines indicate percentages of composite "B" in the feed.
10 B 20
O % methan* in off gas
A P. -OD removed per T. of efficiency
Figure F-2. Operating parameters for backmixed anaerobic filter.
-------
Methane in off gas, percent x.
A Grams COD removed per liter of feed; calculated fromx**xCCH*x64
where'TO = specific gas yield, CCH4 = mole fraction of methane in
off gas,_fL = moles volume at conditions of measurement, and 64 = g.
of oxygen required to oxidize one mole of CH+ to CO2 4 H2O.
O Percent TOC removal efficiency TOC in outfall x
TOC in feed
70
O Percent COD removal efficiency, from,**-, x
x 64
X 100
This expression can be > 100, and is valid only under lined out conditions.
Solid lines indicate percentages of composite "B" in the feed.
vo
ts)
_K
80 85
1°° 105 110 115 120 125 130 135 140 145 150 155 160
Figure F-2. (Continued)
-------
90
60
30
> Methane in off gas, percent x.
i Grams COD removed per liter of feed; calculated from
., TO xCCH4x64 where -O= specific gas yield. CCFi, -
-* mole fraction of methane in off gas, _f";_ =
moles volume at conditions of measurement, and 64 - g.
of oxygen required to oxidize one mole of CH4 to CO2 + H2O,
) Percent TOC removal efficiency TQC in outfall
TOC in feed
_
O Percent COD removal efficiency, from - a 'j
x 64 xlOO
_fL xCOD in feed.
This expression can be > 100, and is valid only under lined out
conditions.
Solid lines indicate perc<
Figure F-2. (Continued)
-------
80
70
60
50
40
30
20
10
<4> Methane in off gas, percent x.
A Grams COD removed per liter of feed; calculated from-TOcCCCH, x 64
SI-
where..^ - = specific gas yield, CCH4 = mole fraction of methane in off gas.-Ti-
moles volume at conditions of measurement, and 64 = g. of oxygen required
to oxidize one mole of CH4 to CO2 + H2O.
(•}Percent TOC removal efficiency TOC in outfall
TOC in feed * 10°
O Percent COD removal efficiency, from,-3—-xCCK^ x 64
jf\~ x COD in feed"
This expression can be > 100, and is valid only under line d out conditions.
Solid lines indicate percentages of composite "B" in the feed.
X 100
240
245
250
255
Figure F-2.
260 265
(Continued)
270
Z75
94
-------
APPENDIX G
ANAEROBIC DENITRIFICATION FLUIDIZED BED
vO
IS 20 25 JO 35 40 45 50
Figure G-l. Anaerobic denitrification fluidized bed.
-------
0.1
0.7
0.6
o.;
l.b
1.4
1.2
1.0
fl.3
O.t
HydraoHe Detention, Day*. Scale A
COD Loading, l.b. COD/Cu Fl/Oay.
OCOD
Sole
1" 15 ZU 25 30 35 40 46 50 55 l-f ' s 70
Figure G-2. Anaerobic denitrification fluidized bed.
-------
VO
O * CH, In Oflgas
Nj in Offgai
"« .N';O in Otlgab
5 10 15 to I', 30 3f. 40
45 50
60 65 70
Figure G-3. Anaerobic denitrification fluidized bed.
-------
OO
A N Accounted for in Offgai, g/1
Nitrite N in Effluent, g/1
10 IS "' " <0
35 40 45 50 S5 60
Figure G-4. Anaerobic denitrification fluidized bed.
-------
VD
2.0
Kinetic COD, g/1. Scale A
N in Offga., R N/l ol Feed. Scale B
IS 20
May 6
A
/un 1
10 IS 40 45
«•
55
60 65
Figure G-5. Anaerobic denitrification fluidized bed.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
PORT NO.
EPA-600/2-79-172
I. TITLE AND SUBTITLE
Biological Treatment of High Strength Petrochemical
Wastewater
7. AUTHOR(S) ~~~ "
William J. Humphrey and Enrique R. Witt
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
August 1979 issuing date
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO
'ERFORMING ORGANIZATION NAME AND ADDRESS
Celanese Chemical Company, Inc.
Technical Center
Corpus Christi, TX 78703
10. PROGRAM ELEMENT NO.
1BB610
11. CONTRACT/GRANT NO.
Grant No. 12020 EPH
iPONSORING AGENCY NAME AND ADDRESS
Robert S. Kerr Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Ada, OK 74820
13. TYPE OF REPORT AND PERIOD COVERED
Final/9-23-79 to 2-20-76
14. SPONSORING AGENCY CODE
EPA 600/15
'LEMENTAF
NOTES
The biological treatment of a complex petrochemical wastewater containing high con-
centrations of organic chlorides, nitrates, and amines was initially studied using
a sequence of anaerobic methanogenesis and oxygen activated sludge. Bench-scale and
pilot-plant treatability studies were conducted using various composite samples and
process wastewater blends. The results of preliminary studies indicated the need
for stream segregation and waste reduction programs at the petrochemical plant.
Segregation of components of the combined plant waste stream was required to elimi-
nate nonbiodegradable materials and pretreatment minimized the concentration of
substances which were toxic or inhibitory to biological treatment. Nitrates inhibi-
ted methanogenesis in the anaerobic system but quantitative removal of nitrates
was accomplished. Only partial removal of chemical oxygen demand (COD) was achieved
during anaerobic denitrification because of the relatively low nitrate/COD ratio.
Anerobic methanogenic treatment also was unsuccessful in reducing the COD concentra-
tion to any great extent, even after pretreatment by anaerobic denitrification. The
activated sludge system was effective in removing the biodegradable portion expressed
as biochemical oxygen demand (BOD) of the pretreated combined wastewater stream;
but the yellow color of the effluent was unacceptably dark. The activated sludge
system performed equally well when high purity oxygen, or air was used for aeration.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Activated Sludge Process
Anaerobic Processes
Industrial Waste Treatment
Pilot Plants
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Denitrification
Petrochemicals Industry
Organic Chlorides
Bench-scale Plants
Oxygen Activated Sludge
68D
DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport!
UNCLASSIFIED
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
21. NO. OF PAGES
112
22. PRICE
EPA Form 2220-1 (9-73)
100
4 U.S. GOVERNMENT PRINTING OFFICE: 1979 -657-060/5396
------- |