EPA-600/2-76-195
December 1976
Environmental Protection Technology Series
EVALUATION AND UPGRADING OF A
MULTI-STAGE TRICKLING FILTER FACILITY
industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental 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-600/2-76-195
December 1976
EVALUATION AND UPGRADING OF A MULTI-
STAGE TRICKLING FILTER FACILITY
by
John H. Koon
Robert F. Curran
Carl E. Adams, Jr.
W. Wesley Eckenfelder, Jr,
AWARE, Inc.
Nashville, Tennessee
and
CIBA-GEIGY CORPORATION
Cranston, Rhode Island
Project 12020 FOH
Project Officer
David H. Stonefield
Surveillance and Analysis Division
Region I
Needham Heights, MA 02194
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
Cincinnati, Ohio 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental
Research Laboratory, 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 recommendation for use.
ii
-------
FOREWORD
When energy and material resources are extracted, processed,
converted, and used, the related pollutional impacts on our
environment and even on our health often require that new and
increasingly more efficient pollution control methods be used.
The Industrial Environmental Research Laboratory - Cincinnati
(IERL-CI) assists in developing and demonstrating new and
improved methodologies that will meet these needs both effi-
ciently and economically.
This report evaluates the applicability of several
alternate treatment methods for complex organic waste waters and
documents the results of an in-depth study of a large scale,
multi-stage trickling filter system. From the study, it was
concluded that the anticipated acclimation of the microorganisms
and enhanced BOD removal were not achievable but rather, that
undesirable air stripping of volatile organic constituents con-
tributed very significantly to apparent pollutant removal.
This study is one of several undertaken by IERL-CI to
evaluate alternate systems for treatment of complex wastes and
reduction of'discharges to our Nation's waterways. The report
alerts industrial waste system designers, state officials, and
EPA officials responsible for standard-setting to some of the
problems which must be anticipated even with relatively con-
ventional technology.
For further information on this subject, contact the
Industrial Pollution Control Division, Organic Chemicals and
Products Branch.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
ill
-------
ABSTRACT
The applicability of a full-scale, six-stage trickling filter plant
was investigated for the treatment of waste from a multiproduct
organic chemical plant. Reductions of BOD per unit volume of packing
indicated that the series design of the system did not lead to stage-
wise acclimation of the microorganisms or enhanced BOD removals. Tests
to determine BOD removal mechanisms in the system indicated that air
stripping and biological mechanisms both contributed significantly to
the total observed reduction.
Effluent recycle considerably improved filter performance. However,
600 percent recycle was required for an approximate 90 percent re-
duction of BOD at a hydraulic loading of 2 gpm/sq ft (0.08 cu m/min-
sq m).
Bench-scale activated sludge investigations showed that this process
could be used successfully for upgrading the trickling filter system.
Kinetic parameters necessary for the design of activated sludge systems
were determined. Comparative studies of air and oxygen aeration indi-
cated that the treatment of wastes containing volatile substances might
be difficult in a closed oxygen system.
Activated carbon adsorption,also tested in small-scale systems, was
most effective as a means of upgrading the trickling filter system
when applied following activated sludge treatment. Activated carbon
tests indicated that some refractory organics and color-producing
substances could be effectively removed by adsorption; however, a
significant nonadsorbable organic fraction remained.
This report was submitted by AWARE, Inc. in fulfillment of U. S.
Environmental Protection Agency Grant 12020 FOH to CIBA-GEIGY Corporation,
iv
-------
CONTENTS
Page
Foreword i i i
Abstract iv
Table of Contents v
List of Figures vi
List of Tables x
Sections
I Conclusions I
II Recommendations 3
III Introduction 4
IV Description of Treatment Facility 8
V Experimental Procedures 27
VI Small-Scale Trickling Filter Investigations 44
VII Full-Scale Trickling Filter Investigations 59
VIII Automatic Monitoring 78
IX Activated Sludge Investigations 82
X Activated Carbon Investigations 99
XI Removal of Organic Constituents by Air Stripping 110
XII Acknowledgements 125
XIII References 126
XIV List of Symbols and Abbreviations 128
XV Appendices 130
-------
FIGURES
No. Page
1 Correlation of Measured BOD to BOD 12
Attributable to Solvents
2 Correlation of Measured TSC to TOC 13
Attributable to Solvents
3 Wastewater Treatment Plant, CIBA-GEIGY 15
Corporation Cranston, Rhode Island
4 Preliminary Treatment Facilities 17
5 Trickling Filter Medium 19
6 Schematic Illustration of Original 21
Trickling Filter Design
7 Trickling Filter Recycle System 24
8 Wastewater Monitoring System 26
9 Process Flow Diagram-Existing Full-Scale 28
and Experimental Treatment Sequences
10 Schematic Illustration of Pilot Trickling 31
Filter
11 Schematic of the Microtower Arrangement 33
12 Schematic Illustration of Activated Sludge 35
Units
13 Schematic Illustration of Oxygen Activated 37
Sludge Units
14 Schematic Illustration of Activated Carbon 39
Columns
15 Aerated Equalization and Stripping Basin 41
16 Variation of Removal Effluent with Applied 46
COD Load
vl
-------
FIGURES (Cont'd)
No. Page
17 Removal Efficiency Versus Applied Organic 47
Loading in the Recirculation Study
18 BOD and COD Removal Efficiency Versus Recycle 48
Ratio
19 Effects of Organic Loading on Oxygen Uptake Rate 49
20 Oxygen Uptake Versus Recirculation Ratio 51
21 Oxygen Utilization in the Microtower 52
22 Fraction of COD Removal Due to Oxygen Uptake 54
in the Microtower
23 Effect of Recycle on Pilot Trickling Filter 57
Performance
24 Variation of Pilot Trickling Filter Performance 58
and Organic Loading
25 Chronological Variation of BOD Removal 61
26 Relationship of Trickling Filter BOD Loading 63
and Removal
27 Effect of pH Fluctuation on Trickling Filter 65
Performance
28 Temperature Dependence of the BOD Removal Rate 67
Constant
29 Variation of Solvent Concentration through the 69
Trickling Filter System
30 Oxygen Uptake In, Full-Scale Filters 71
31 BOD Removal Rate Constant-Phases I and II 84
32 Temperature Effect on Activated Sludge 86
Performance
Vll
-------
FIGURES (Cont'd)
No^ Page
33 Determination of Oxygen Utilization 87
Coefficients, BOD Basis-Phases I and II
34 Excess Sludge Production, BOD Basis- 89
Phases I and II
35 BOD Removal Rate Constant, Unstripped 91
36 BOD Removal Rate Constant, Stripped Waste - 91
Phase III
37 Performance of Oxygen - Acclimated Sludge 93
Subjected to Air Aeration
38 Oxygen Utilization for Unstripped Waste, 94
BOD Basis-Phase III
39 Oxygen Utilization for Stripped Waste, 94
BOD Basis - Phase III
40 Summary of BOD Removal Relationships for 97
Activated Sludge Systems
41 Influence of pH on Carbon Adsorption Capacity 101
42 Breakthrough Curve - Hydrodarco 4000 Carbon 103
43 Breakthrough Curve - Westvaco 8 x 30 Carbon 103
44 Breakthrough Curve - Filtrasorb 400 Carbon 103
45 Color Removal Breakthrough Curves for Hydro- 105
darco 4000 and Westvaco 8 x 30
46 Color Removal Breakthrough Curve for Filtra- 105
sorb 400
47 TSC Breakthrough for Regenerated Hydrodarco 106
4000
48 COD Breakthrough for Filtrasorb 400 Using 109
Trickling Filter Effluent
Vlll
-------
FIGURES (Cont'd)
No. Page
49 TSC Removal in Batch Stripping Tests 111
50 Variation of Stripping Removal Coefficient 113
with Temperature
51 Packed Tower Air Stripping Results, BOD Basis 115
52 Linearized Data for Packed Tower Stripping - 116
TSC Basis
53 Linearized Data for Packed Tower Stripping - 117
BOD Basis
54 TSC Removal in Aerated Equalization Basin-D.T.= 118
1.5 Days
55 TSC Removal in Aerated Equalization Basin-D.T.= 120
3 Days
56 Organic Removals in the Aerated Basin 122
-------
TABLES
No. Page
1 Wastewater Characterization 10
2 Solvent Determinations for Raw Equalized Waste 11
3 Stripping Results From Pilot Trickling Filter 56
4 Monthly Trickling Filter Performance 62
5 BOD Removal Attributable to Oxygen Uptake 72
6 Effect of Tower Blowers on Dissolved Oxygen 74
Level and Total Carbon Removal
7 Recycle Data Summary 76
8 Comparative Dissolved Oxygen Measurements 79
9 Summary of Activated Sludge Design Parameters 96
10 Carbon Isotherm Results Using Activated 100
Sludge Effluent
11 Comparison of Results for Activated Carbon 107
Adsorption of Activated Sludge Effluent
12 Summary for Stripping in Aerated Basin 119
13 Solvents Measured in Stripping Investigations 123
Al Treatment Facility Capital Cost Summary 131
A2 Treatment Facility Operating Cost Summary 132
Bl Trickling Filter Weekly Data Summary 133
B2 Full-Scale Trickling Filter Recycle Data 134
Cl Summary of Activated Sludge Results-Phase I 135
C2 Summary of Activated Sludge Results-Phase II 135
X
-------
TABLES (Cont'd)
No. Page
C3 Summary of Activated Sludge Results, 137
Varied Temperature Operation - Phase II
C4 Summary of Activated Sludge Results - Phase III 138
xi
-------
SECTION I
CONCLUSIONS
1. Over the period of this investigation, the six-stage trick-
ling filter system removed only 32 percent of the effluent
BOD load. The effluent BOD averaged 2,180 mg/1. Due to ex-
treme organic overloading in the system, no stage-wise
acclimation of microorganisms was observed. The quantity
of BOD removed in the system ranged from approximately 100
to 200 lb/1,000 cu ft-day (1.5 to 3 kg/cu m). Since removals
of this magnitude are not uncommon for single-stage, plastic
medium trickling filter systems, it cannot be concluded that
multiple staging resulted in extraordinarily high removals/
unit volume for this waste.
2. Recycle investigations in three stages of the full-scale
system demonstrated that approximately 50 percent BOD re-
moval could be achieved at 20°C using a recirculation ratio
of 2.3. However, while recirculation improved filter per-
formance, these results and small-scale trickling filter
tests indicated that BOD reductions higher than approximately
90 percent could not be achieved at recycle ratios as high
as 800 percent. For this high strength waste (initial BOD of
3,200 mg/1), this level of treatment would probably be
inadequate in most instances.
3. The six-stage trickling filter system, even when modified
to include effluent recycle, failed to produce acceptable
results as the sole treatment method for the wastewater.
Due to the failure of the system to provide extraordinarily
high organic removals, their multi-stage design can be used
only as an initial treatment step.
4. Investigations made using the small-scale trickling filter
to determine the mechanism of BOD removal indicated the
principal removal mechanism at low recycle ratios was by
air stripping, but that the fraction of BOD removed by air
stripping decreased significantly as the recycle rate was
increased.
-------
The annual treatment cost for the 1.44 mgd facility including
neutralization, equalization, trickling filtration in a six-
stage system, and clarification totaled $690,000/yr including
capital and operating and maintenance costs. This was equiva-
lent to $1.58/1,000 gal and $0.19/lb BOD removed.
Bench-scale activated sludge investigations showed that this
process could be used to upgrade the trickling filter effluent.
Treatment of the waste was characterized by a relatively high
oxygen demand per unit of BOD removed, a relatively low sludge
yield, and poor oxygen transfer characteristics. The effluent
was characterized by high effluent suspended solids concentra-
tions ranging from 90 to 175 mg/1 and a high refractory effluent
COD of approximately 450 mg/1.
Performance of the activated sludge process was significantly
impaired using pure oxygen aeration in a closed atmosphere
system treating the unstripped raw wastewater. However,
performance using aeration treating a stripped waste was equal
to or better than that obtained with air aeration.
Activated carbon adsorption was capable of removing refractory
organic substances from the activated sludge effluent. Results
indicated that treatment of the trickling filter effluent using
this process would be feasible following removal of high con-
centrations of short-chained solvents.
Results of air stripping investigations showed that a signi-
ficant fraction of the raw waste organics could be removed in
an aerated equalization basin with surface aerators and in
packed towers such as trickling filter systems. In order to
achieve 60 percent BOD reductions, an air flow of 500 cu ft/gal
was required in the packed tower, while a 3-day detention time
using 150 hp/mil gal aeration capacity was required in the
basin. These results indicate that the effects of stripping
especially as they relate to scale-up from small-scale systems
can be significant in designing biological systems to treat
wastes containing volatile substances.
-------
SECTION II
RECOMMENDATIONS
1. The results of this investigation indicated that waste pro-
duced in the manufacture of miscellaneous organic chemicals
can be treated using both biological and physical-chemical
processes. However, the complexity of the waste and large
fluctuations in waste composition caused by changes in pro-
duction schedules dictate that long-term responses of ap-
plicable treatment systems be thoroughly considered prior
to preparing a final design. Primary consideration should
be given to designs in which fluctuations in concentration
and composition have a minimum effect on system performance.
2. In designing plastic medium trickling filters it is impor-
tant that consideration be given to the structural stability
of the packing and additional requirements for vertical and
lateral supports in the tower superstructure. Packing such
as the one used in this design has a minimum ability to
support vertical and lateral loads and, therefore, must be
strengthened by more complex supporting structures than
would be required for other commercially available packing
designs.
3. Effluent recirculation should be included in trickling fil-
ters designed for treatment of high-strength industrial
wastes.
4. In designing multi-stage trickling filters, care should be
taken not to overload the first stages of a system to the
extent that failure of the entire system will be experienced.
-------
SECTION III
INTRODUCTION
Wastes from many organic chemical plants typically contain sig-
nificant concentrations of short-chained organics used as sol-
vents and lesser concentrations of more complex organic substances.
In comparison to municipal sewage, these wastes are highly vari-
able in composition depending on plant production schedules, are
composed principally of soluble organic substances which cannot
be removed by precipitation or adsorption onto biological floes,
and may contain exotic chemical species which would interfere
with conventional treatment processes. When treating wastes of
this type, biological processes would be most effective in re-
moving short-chained and other degradable organic constituents,
while physical and chemical processes, such as adsorption using
activated carbon, would be required to remove substances refractory
to biological treatment.
The purpose of this project was to evaluate the use of multi-
stage trickling filters and other methods for treating wastes
produced in a diversified organic chemicals plant. The project
involved demonstration of a full-scale, six-stage plastic medium
trickling filter for reduction of organic constituents in the
wastestream and the investigation of other potential treatment
methods. Advantages which might be attributed to the multi-stage
trickling filter include the sequential development of micro-
organisms in the system acclimated to the different organic con-
stituents in the waste. If this were the case, greater removals
of organic constituents per unit volume might be experienced com-
pared to conventional trickling filter designs. Small-scale
tests using the activated sludge and activated carbon processes
were performed to determine the applicability of these methods
for further treatment of the waste required to meet anticipated
best practicable and best available treatment technology.
BACKGROUND INFORMATION
The CIBA-GEIGY plant located at Cranston, Rhode Island is a multi-
product organic chemicals plant in which over 500 different raw
materials are used to make more than 200 different chemical pro-
-------
ducts. Products manufactured in the plant include plastic additives,
Pharmaceuticals, dyestuffs and intermediates, and miscellaneous
chemicals. The products produced here are destined for use largely
in the plastics, leather, textiles and paper industries. Because
of the diversity of products produced in the plant, it is difficult
to precisely relate the plant production to the nature of the waste-
waters produced. During 1973, 30,000,000 Ib of products were manu-
factured in the plant. The production was distributed among various
product categories as follows: Pharmaceuticals, 3 percent; plastic
additives such as antioxidants, 59 percent; chemical products (mostly
fluorescent whitening agents and textile auxiliaries), 26 percent;
tanning agents, 11 percent; and printing inks, 1 percent.
Situated on the Pawtuxet River approximately one mile upstream from
Narragansett Bay, the plant is known for its modern facilities and
versatility. Due to the extensive development and pilot plant
facilities located at this site and the diverse nature of products
manufactured in the plant, it has become known within CIBA-GEIGY
Corporation as the "pilot plant of the world."
Originally, wastes generated in the plant were discharged through
numerous outfalls to the Pawtuxet River. When it was recognized
that direct disposal of untreated wastes to the river was unaccepta-
ble, a long-term wastewater management program was implemented in
the plant. In 1965 cooling water used in the plant was separated
from process discharges and combined in a recirculated cooling
system. Approximately 7 mgd (26,000 cu m/day) of cooling water
is recirculated through the plant.
In order to reduce process waste discharges, certain solvents and
acids were segregated and either cleaned, concentrated and recycled
or transported away from the plant for proper disposal. Having
isolated most materials which lent themselves to this type of
treatment, it was realized that additional treatment of wastes
would be required. Investigations were initiated in July, 1965
to evaluate alternative means for treating and disposing of wastes
in the plant. After considerable study a decision was made to
construct a multi-stage plastic medium trickling filter system.
This system was designed and constructed by Biodize Systems,
Enviro-Chem Division of Monsanto Company and was placed in opera-
tion during the spring of 1972.
In 1970 CIBA-GEIGY Corporation received a research and development
grant from the Environmental Protection Agency to evaluate the
multi-stage trickling filter system over a two-and-one-half year
period. Following start-up of the system, CIBA-GEIGY Corporation
retained Associated Water and Air Resources Engineers, Inc. (AWARE)
to perform the on-site investigations associated with the demon-
stration grant. In November, 1972 investigations were initiated.
-------
THE APPLICATION OF MULTI-STAGE TRICKLING FILTERS
Although most work relating to pilot- and full-scale trickling
filter installations has been concerned with plants in which only
a single stage is employed, two- and three-stage systems are not
uncommon. Advantages of multi-stage systems which have been
advanced in the past include: greater flexibility in treating
high-strength wastes to levels acceptable for discharge and the
development of varied microbial populations from one stage to
the next which might provide greater organic removals compared
to a single stage having an equivalent volume. The primary dis-
advantage of multi-stage filters is the tendency to organically
overload the first stages of the system.
Good performance of plastic medium, multiple-stage trickling fil-
ters has been reported for numerous industrial effluents (1, 2).
Food processing industries routinely use multiple-stage plastic
filters in the treatment of wastes from dairies, meat packing
plants, fruit and vegetable processing plants, potato processing
plants, and breweries (3, 4). In one case, a four-stage system
was designed to treat a very strong pharmaceutical effluent (3, 5).
The design included a series of four plastic filters, each with
individual recycle and an individual settling tank (although inter-
stage clarification is not considered essential) followed by a
fifth stage consisting of a six-foot granite polishing filter.
Recirculation ranging from 100 to 200 percent was practiced within
each stage.
The effect of effluent recirculation cannot be predicted a priori
in any single case. In certain cases, negative effects have been
reported. For the case of concentrated industrial effluents,
however, most researchers have reported beneficial effects of
recirculation (6, 7, 8). This is due to enrichment of specific
enzymes, aeration, adsorption of inhibitory compounds, dilution,
and dampening of short-term raw waste load variability.
Equally opposed opinions have been expressed concerning the exist-
ence of a limiting organic loading. The limiting loading is
defined as a load beyond which the BOD removal decreases. The
existence of such a loading has been considered by numerous in-
vestigators (9-13). Other authors, working primarily with plastic
media, reported no limiting loading (14, 15).
-------
OBJECTIVES
The principal objectives of this investigation were to evaluate a full
scale multi-stage plastic medium trickling filter system for the
treatment of waste from a diversified organic chemical plant
and to determine the applicability of the activated sludge and
activated carbon processes for further treatment of the waste
as judged by tests in small-scale systems. Specific objectives
outlined in the grant application were modified during the course
of the investigation by joint agreement between CIBA-GEIGY
Corporation and the U.S. EPA. The restructured objectives of
the investigation were as follows:
1. To optimize operation of present treatment facilities.
2. To examine operation of the multi-stage trickling filter
system at various organic loads, temperature conditions,
and other operating variables.
3. To investigate BOD removal mechanisms in the trickling filter
system and to determine whether stage-wise acclimation of
microorganisms occurred.
4. To evaluate the feasibility of using automatic monitoring
equipment for the control of trickling filter systems.
5. To examine the activated sludge process as a means of further
treating the waste and to determine the kinetic parameters
required for a process design.
6. To examine activated carbon adsorption as a means of reducing
BOD and removing refractory organic materials.
Personnel from AWARE, Inc. were responsible for sampling, operation
and care of experimental treatment units and compilation and pre-
sentation of all data obtained from the project. CIBA-GEIGY per-
sonnel were responsible for the conduct of all analytical work, the
operation and maintenance of automatic monitoring equipment and
the conduct of effluent recycle investigations in the full-scale
trickling filter system.
-------
SECTION IV
DESCRIPTION OF TREATMENT FACILITY
Although this project was concerned primarily with the perfor-
mance of the multi-stage trickling filter system, several other
unit operations were required for treatment of the complex
chemical waste discharged. The several unit operations included
in the entire treatment system and characteristics of the influ-
ent waste are discussed in this chapter.
SOURCES OF WASTEWATER
The complexity of waste treatment problems at the Cranston plant
is indicated by the large number of products produced in the
facility and by the variety of manufacturing processes employed.
Production of the great number of products required from this
plant necessitates the use of batch processing methods which
usually generate greater quantities of waste than comparable
continuous-flow production methods. Since each reactor or
"kettle" is used for producing a large number of products, segre-
gation of wastes would be impossible. It is often common for
products to be manufactured during "campaigns" lasting several
weeks during which several months' supply of a particular product
is produced. Therefore, in addition to daily fluctuations in
wastewater strength and composition, significant weekly and
longer-term variations are also evident.
The manufacturing plant consists of six buildings containing
over 100 independent reactor units varying in size up to
3,000 gal (12,000 liters). In addition to agitated vessels,
reaction units include appropriate auxiliary equipment such as
distillation columns, condensers, receiving vessels, many types
of filters and dryers and associated material transfer equipment.
Because of the many different reaction cycle times, no fixed
number of batches per day can be quoted. However, the plant
operates on a 24 hr/day basis and material transfer operations
are in progress simultaneously on anywhere from 5 to 25 percent
of the equipment.
Sources of wastewater include: 1) detergents, solvents, and
other materials used to clean kettles, 2) solvents and other
-------
materials used to wash and purify products, 3) filtrates and
centrates from product dewatering and drying, 4) water from vacuum
jets, 5) blowdown from a recirculating cooling system, 6) still
bottoms, 7) materials used to clean floors in production areas,
and 8) stormwater run-off from the production area. Wastewater
from manufacturing areas is collected in the plant sewer system
and conveyed to the main pumping station. Based on the use of
city water recorded in the production area and the quantity of
wastewater received at the treatment plant, there is very little
water loss in the plant due to evaporation or packaging with pro-
ducts. Comparison of these flows over the duration of the study
showed that recordings from the magnetic flow meter in the treat-
ment plant were approximately 15 percent greater than water meter
readings. Some difference would be expected due to the intro-
duction of stormwater flows to the plant sewer.
During most of the study period, work in the manufacturing areas
proceeded 24-hr/day, 6 days/wk. However, at times the work
schedule was decreased to a 5-day week or increased to 7 days/wk
depending upon inventory backlogs. During periods when the plant
was not operating, the waste flow averaged approximately 200 gpm
(0.8 cu m/min) due to water used for vacuum jets and cooling
systems.
WASTEWATER CHARACTERIZATION
Twenty-four hour composite samples of wastewater were collected
over the nine month period from November 1972 through July 1973
and analyzed for BOD, COD, TSC (total carbon test performed on
filtered wastewater, total soluble carbon), suspended solids,
volatile suspended solids, color and temperature. The results of
these and other tests performed on the wastewater are summarized
in Table 1. The waste contained a high concentration of organic
material, demonstrated by a total influent BOD of 3,190 mg/1 and
a total influent COD of 4,676 mg/1. The BOD/COD ratio of the raw
waste was 0.68. At median concentrations approximately 90 per-
cent of the total BOD and COD was in the soluble fraction, due
largely to solvents present in the wastewater. Weekly grab sam-
ples were analyzed to determine the concentrations of various
solvents. The ranges and average values for these solvents are
shown in Table 2.
In order to determine the approximate fraction of the total organic
waste load due to solvents, the BOD attributable to each solvent
-------
TABLE 1
WASTEWATER CHARACTERIZATION
Parameter
Flow3 (mgd), all days
Weekdays
BOD (mg/1) Total
Soluble
COD (mg/1 ) Total
Soluble
TSC (mg/1)
Suspended solids (mg/1)
Volatile suspended solids (mg/1)
Total kjeldahl nitrogen (mg/1)
Total phosphorus (mg/1 )
Alkalinity (mg/1 as CaCOo to pH 3.7)
Total solids (mg/1)
Color (APHA units)
Temperature (°F) Influent
Effluent
Average
1.26
1.48
3190
3090
4676
4100
1200
151
131
37
6
227
2725
1800
—
—
Percenti le Values
10%
0.48
1.12
1400
1200
2400
2100
620
48
44
2.5
2.2
155b
1500°
400°
57
52
50%
1.42
1.54
3000
2600
4650
4150
1220
108
100
30
3.8
—
—
—
67
62
90%
1.64
1.68
5050
4750
6900
6200
1830
288
250
94
13.
282°
4860°
4000D
77
72
Flow in cu m/day = mgd x 3790-
Values are minimum and maximum observed.
10
-------
TABLE 2
SOLVENT DETERMINATIONS FOR RAW EQUALIZED WASTE
Solvent
Acetone
Methyl ethyl ketone
Methanol
Isopropyl alcohol
Toluene
Concentration (mg/1 )
Average
280
20
220
1,260
120
Range
30-600
10-100
0-700
500-3,100
0-300
was calculated and correlated to the total BOD of the influent
waste as shown in Figure 1. This correlation indicates that
approximately 61 percent of the BOD was due to solvents. In a
similar manner total carbon values were calculated for each sol-
vent analysis. The results shown in Figure 2 indicate that
practically all of the total carbon may be attributed to the
solvent load. However, while total carbon values for the sol-
vents were theoretical, less than 100 percent total carbon can
be measured by any laboratory procedure used for the influent
waste. Relationships were developed between total BOD and COD
and between soluble BOD and TSC, resulting in the following
equations:
BODT = 0.788 (CODT) - 420
BOD$ = 2.65 (TSC) - 382
The suspended solids content of the waste averaged 151 mg/1;
however, a significant fraction of the solids was very highly
dispersed and non-settleable. Volatile suspended solids com-
prised 87 percent of the total suspended solids. Nitrogen and
phosphorus concentrations of the raw wastewater were determined
from composite samples collected from mid-November to mid-
December of 1972. The quantities present would be insufficient
11
-------
5000 r
4000
z
u
>
8 3000
o
h-
LU
_J
< 2000
m
ir
1000
Q
O
CD
0
I Slope = 0.61
0
1000 2000 3000 4000 5000
INFLUENT WASTE BOD (mg/l)
6000
FIG. I . CORRELATION OF MEASURED BOD TO BOD ATTRIBUTABLE
TO SOLVENTS
-------
2500
~ 2000
V)
8 1500
UJ
_j
CD
ID
go
oc
<
o
o
1000
500
0
Regression Line
y= 0.818 xt 203
r=0.7l
Visual best fit
through origin
I i
I
0
500
1000
RAW WASTE TSC (mg/l)
1500
2000
FIG. 2 . CORRELATION OF MEASURED TSC TO TOC ATTRIBUTABLE
TO SOLVENTS
-------
to maintain a high level of biological treatment; therefore,
nutrients must be added to this wastewater prior to biological treat-
ment. Total solids, determined from samples composited over one-
week periods, averaged 2,725 mg/1 in the influent waste and 3,800
mg/1 in the trickling filter effluent. The alkalinity determinations
made from July 13 to July 20, are significantly greater than would
be expected from the low inorganic carbon concentration and might
have been due to organic constituents present in the waste,
PLANT DESIGN
Haste Collection and Transport
All process wastes from the manufacturing area in the plant flow
through a central sewer system and are subsequently collected
in a raw wast0 pumping station. All sanitary wastes from the
analytical and development laboratories are discharged separately
to the municipal sewer system. In addition to process waste-
waters, stormwater runoff from some material storage areas is
collected in the pumping station. Because of the highly variable
pH of process waste discharges, the raw waste collection sump is
lined with an acid-alkali resistant brick. Waste is pumped from
the sump to the treatment plant using two 1,250 gpm (5 cu m/min),
acid resistant pumps constructed of a high nickel stainless
steel. Each pump is capable of handling the average plant waste
flow of 1,000 gpm (4 cu m/min). To insure that raw waste is not
discharged during power failures, a back-up power supply consist-
ing of gas-operated generators is provided for operation of the
raw waste and equalization basin influent pumps. Prior to enter-
ing the sump, all waste is screened through a mechanically-cleaned
bar rack in order to remove extraneous materials discharged to the
plant floor drains.
From the pumping station waste is transported through a polyester-
fiberglass pipe approximately one-quarter mile to the treatment
plant site. This particular construction was required in order
to provide adequate resistance to acid, alkali, and solvents
which might be present in high concentrations prior to equaliza-
tion and neutralization. A photograph of the treatment facility
is shown in Figure 3.
Preliminary Treatment
Preliminary treatment processes include solvent separation, flow
measurement, neutralization, equalization, and nutrient addition.
14
-------
FIG. 3 .WASTEWATER TREATMENT PLANT, CIBA-GEIGY CORPORATION
CRANSTON, RHODE ISLAND
-------
A schematic illustration of pretreatment facilities is shown in
Figure 4.
Due to the possibility that large spills of solvents might occur
in the plant, facilities are provided to separate immiscible sol-
vent loads from the wastestream. This is accomplished in a
45,000 gal (180 cu m) cylindrical tank. Waste enters the tank
through a tangential inlet in order to maintain most solids in
suspension and yet allow separation of any immiscible solvents.
In addition to serving as a solvent separator, the tank provides
some equalization of pH and flow prior to the neutralization
system. Immiscible solvent layers are withdrawn through a tan-
gential outlet into a 5,000 gal (20 cu m) solvent holding tank.
Collected solvents may be trucked away for ultimate disposal or
transported to the manufacturing area for reclamation and reuse.
In order to provide collection of large solids which settle in
the solvent separation tank, the tank bottom slopes to one side
which is equipped with a valved outlet suitable for wasting
sludge. Sludge is usually collected in 55-gal (200 liter) drums
as necessary and taken for disposal with other concentrated
wastes from the manufacturing area.
Flow Measurement and Neutralization. Waste is pumped from the
solvent separation tank into the control building where the flow
is metered and the waste neutralized. The waste flow is meas-
ured using a magnetic flow meter attached to the raw waste force
main. Neutralization is accomplished in a one-stage, tubular
reactor which was constructed by baffling an 8-ft (2.4 m) section
of pipe in the raw waste force main. Probes are located at the
downstream end of the reactor to monitor and control pH. Con-
trollers are coupled to transmitters and automatic valves which
control acid and caustic additions to the wastestreams. Concen-
trated sulfuric acid and 50 percent sodium hydroxide are used
to neutralize the waste.
Equalization. Following neutralization the waste is stored in
two, 1-mil gal (4,000 cu m) equalization tanks. These cylindri-
cal steel tanks are 100 ft (30 m) in diameter with a depth of
approximately 20 ft (6 m) and are located above grade. Together
they provide a holding time of approximately 1.5 days for the
plant design flow of 1.4 mgd (5,600 cu m/day). Each tank is
lined with a phenolic material to prevent corrosion.
16
-------
BAR
SCREEN
TO
TRICKLING
FILTERS
RAW
WASTE
PUMPING
SOLVENT
SEPARATION
TO
TRICKLING
FILTERS
§-a
8--Q
SOLVENT
STORAGE
FLOW
MEASUREMENT
NUTRIENT
ADDITION *
NEUTRALIZATION
EQUALIZATION
Nutrient addition shown as modified in 1973.
FIG. 4 PRELIMINARY TREATMENT FACILITIES
-------
Flow enters the tanks through diffusers which were constructed
across the bottom of each tank approximately 2 ft (0.6 m) from
the bottom. Two 1,000 gpm (4 cu m/min) pumps convey the waste
from the equalization tanks to the trickling filter system.
It was originally planned that the inlet diffuser system would
provide sufficient mixing for equalization and would prevent
solids from settling in the basin. However, a significant
quantity of sludge accumulated during the first year of opera-
tion which dictated the need for mechanical mixers. Subse-
quently, a 40-hp (30 kw) mixer equipped with a marine-type
propeller was installed in the side of each tank. Since that
time no problems have been encountered due to the accumulation
of solids in the tank or to poor equalization because of
stratification.
Nutrient Addition. According to the original plant design, an-
hydrous ammonia and 75 percent phosphoric acid were added to
the wastestream as it was pumped from the equalization basin
to the trickling filter system. However, several problems
were encountered in obtaining adequate control over nutrient
addition which resulted in severe pH shocks to the biological
system. Subsequently, piping modifications were made to
permit nutrient addition prior to equalization.
Trickling Filtration
The trickling filtration system consists of six, 20-ft (6 m)
deep trickling filters designed to operate in series. The fil-
ters were constructed in two rectangular banks with each bank
containing three stages. Each stage consists of a 20 ft x 25
ft (6 m x 8 m) surface area packed with 10,000 cu ft (300 cu m)
of plastic packing. The packing is supported by metal gratings
in four, 5-ft (1.5 m) sections. Waste is distributed to each
stage through fixed nozzle distributors.
The plastic packing is illustrated in Figure 5. The material
is constructed from polyvinyl chloride and resembles a corru-
gated expanded metal grating. The corrugations are approxi-
mately 1 5/8 in. (4 cm) high. The packing was placed horizon-
tally in each trickling filter tower with corrugations placed
at right angles with each other in alternate layers.
Waste is collected beneath each tower in a sump having a capa-
city of approximately 15,000 gal (60 cu m). Since the bottom
18
-------
FIG. 5 . TRICKLING FILTER
MEDIUM
-------
of each sump is flat, sludge which accumulates in corners of the
basin sometimes produces objectionable odors during summer months.
A schematic illustration of the trickling filter system is shown
in Figure 6.
Piping provided in each tower permits the following flow arrangements;
1. Under normal operation, waste is pumped from the equalization
tank or from the preceding column up the riser to the distribu-
tion system at the top of the subsequent tower.
2. In order to provide for repairs, flow may be bypassed around
one, or any number, of towers.
3. Flow may be recycled in any particular tower.
In addition, provisions were included to pump waste directly to the
sump of the first trickling filter, thereby bypassing equalization
facilities, or to recycle flow from Tower 6 to Tower 1 in order to
provide flow through the system during periods when no raw waste is
available. Each stage is provided with a 1,000-gpm (4 cu m/min)
pump driven by a 25-hp (19 kw) motor. Pumping from each sump is
controlled by level indicators in the sumps which are connected to
control valves located on the outlet side of each pump.
Ventilation in the system is provided by a 10,000 cfm (28 cu m/min)
blower mounted at the base of each tower which is driven by a 7.5-
hp (5.6 kw) motor. Air is forced up through the trickling filter
and exists through one 4 ft x 4 ft (1.2 m x 1.2 m) port located
in each trickling filter bank.
Final clarification is provided in a 10-ft (3 m) deep clarifier
having a diameter of 50 ft (15 m). The overflow rate for design
conditions is 730 gpd/sq ft (30 cu m/day-sq m). Sludge is dis-
charged to the city sewer and subsequently handled in the Cranston,
Rhode Island municipal wastewater treatment plant.
Both the trickling filter system and the final clarifier were con-
structed above grade. Due to a high groundwater level in the treat-
ment plant area, above grade construction was employed. Since the
plant is located adjacent to a residential area, it was necessary
to enclose the trickling filter structures, except for one air
discharge port in each bank, in order to minimize the escape
of biological odors to the surrounding community.
20
-------
N5
From
Equalization
By pass from
Neutralization
( ;»
j
\ _.
r
£>
t>
°x
a
3. y
MI >ii
TOWER
!
D
,
/^v
-o
J
*
D
i
y - —
^ v
ill III
TOWER
2
y —
iS y
!l^ III
TOWER
4
£_i
*-
t
A
TRICKLING FILTRATION
CITY SEWER
i
RWTUXET
RIVER
.CLARIFICATION
FIG.6.SCHEMATIC ILLUSTRATION OF ORIGINAL TRICKLING FILTER DESIGN
-------
DESIGN EVALUATION
Several problems were encountered with the original design. Two
of these have been mentioned in the previous discussion: the
addition of nutrients downstream from neutralization and equali-
zation facilities and the absence of mixers in the equalization
basins. Additional problems which have been encountered are
discussed below.
The neutralization system, originally designed as a tubular re-
actor, provided poor pH control. As a short-term solution to
this problem, waste was neutralized on a batch basis in the
equalization tanks during most of this investigation. Using a
procedure developed by the plant operator after the plant was
in service, waste was introduced into one tank until approxi-
mately 80 percent of the total capacity was reached. Subse-
quently, a sample was taken and titrated to neutrality in the
plant laboratory. The quantity of acid or caustic required to
neutralize the waste was calculated from the titration and
metered into the tank with the influent flow. Following a 15
min mixing period, another waste sample was taken from the tank
and checked to insure that the pH was within acceptable limits
for feeding to the trickling filters. As one tank of neutralized
waste was fed to the trickling filters, raw waste was collected
in the other equalization basin. Although it has been proven
that this method can reliably be used by plant operators for
controlling the wastewater pH, the capacity of the equalization
system to absorb fluctuations in waste strength was sacrificed.
A more suitable neutralization system is presently being designed.
The adequacy of the equalization system was not specifically
evaluated during this project. While it appears that the system
is probably adequate to absorb short-term fluctuations in waste-
water strength and flow when the plant is operated 7 days/wk,
the equalizing capacity of the system is somewhat impaired by
the use of two parallel basins. Considering the difficulties in
operating two parallel basins in such a manner to maximize equali-
zation of the wastewater, the design of one larger basin would
have provided more control over equalization of the waste.
Structurally, the trickling filters are supported on an external
galvanized steel framework. Each 5-ft section of packing is
supported on a structural steel framework extending throughout
each tower. The system is designed to support an interior load
22
-------
of 80 Ib/cu ft. Originally, the towers were wood impregnated
with a decay-resistant chemical and were designed to support
an interior load of 20 Ib/cu ft. They suffered a complete
structural failure soon after startup.
Failure of the tower system was enhanced by the lack of lateral
support for the packing. Compression of the corregated-type
packing when loaded with the slippery, biological slime resulted
in lateral force within the columns in addition to the vertical
load of the wet biota. Visual inspection of the packing, shown
in Figure 5, indicates that this packing would have significantly
less strength when subjected to both lateral and compression
loads than compared to other commerically available plastic
packing.
The capability to recycle effluent and mix it with the raw waste
was not included in the original design. The advantages of efflu-
ent recirculation have been disputed in many applications. However,
recirculation should be considered in treating all high-strength
industrial wastes since it provides recycling of nutrients and
active microorganisms to the head of the system, dilution of the
influent organic concentration, and dampening of fluctuations in
the influent waste load.
During the fall and winter of 1973, the system was modified to
provide continuous effluent recycle. A schematic diagram showing
piping modifications for recycle is shown in Figure 7. The six-
stage system was modified to allow operation as two parallel
units, each containing three filters in series. Flow pumped from
the equalization tanks is split to Towers 1 and 4 using orifice
plates and pressure-actuated control valves. Raw waste is com-
bined with recycle flow at the top of Towers 1 and 4 and flows
sequentially through three towers. Effluent from Towers 3 and 6
is split using sump level controllers (not shown in Figure 7)
and an orifice plate meter, which is coupled to another control
valve. A preset flow is returned to the head of the system for
recycle. The balance of the flow, as sensed by the sump level
monitors, is pumped to the final clarifier.
TREATMENT COSTS
A summary of treatment costs incurred for this plant are summarized
in Appendix A. Table A-l shows capital costs for the plant as
estimated immediately following construction. Since the plant
was built under a turnkey contract, it was not possible to obtain
23
-------
to
From
Equalization
Tank
Existing Piping
•New Piping
FIG. 7. TRICKLING FILTER RECYCLE SYSTEM
-------
exact costs in all cases. In particular, no delineation of
major electrical costs or piping costs (included in "Site Work")
were available. It should be noted that the cost given for the
trickling filters reflects the original wooden superstructure
and not the steel and fiberglass structure constructed after
the collapse of the original system.
The operating costs listed in Table A-2 are for 1973 which was
the first year of operation following startup of the reconstructed
plant. Part of the cost given for "Supervisory and Administrative
Labor" included salary and other costs for the plant environmental
supervisor whose duties were not completely related to treat-
ment plant operation. While "Laboratory and Analytical" costs
were somewhat high compared to other industrial plants, this level
of testing will probably be required to adequately monitor a plant
treating such a complex waste to insure compliance with permit
limitations. Maintenance costs reflected startup problems which
occurred during the year.
In order to calculate treatment costs on a volume and performance
basis, it was assumed that the plant would generate 1.4 mgd of
waste for 310 days/yr (6 day/wk operation) and that 11,900 Ib/day
of BOD would be removed for each plant operating day (equivalent
to 32 percent removal from an influent BOD of 3,200 mg/1 observed
during this study). In order to obtain an annual cost for the
capital cost of the system, an interest rate of 7 percent and a
life of 10 years was assumed. Thus, the amortized cost was
$263,000/yr. When combined with the annual operating cost of
$327,000/yr, the total annual cost was $690,000/yr. This is
equivalent to costs of $1.58/1,000 gal of waste treated and
$0.18/lb BOD removed.
INSTRUMENTATION SYSTEMS
In order to determine the applicability of this system to auto-
mated control, an Ionics on-line total carbon monitoring system
and a Weston and Stack dissolved oxygen system were installed in
the treatment plant.
A schematic illustration of monitoring points in the system is
shown in Figure 8. The total carbon analyzer and all recorders
were installed in an instrument building located adjacent to
Tower 3 Dissolved oxygen probes were located to monitor the
contents of the effluent from each tower and the final clanfier.
25
-------
r
FROM
EQUALIZATION
TANKS
I
CLARIFIER
DO-Dissolved Oxygen
TC -Total Carbon
FIG.8.WASTEWATER MONITORING SYSTEM
-------
SECTION V
EXPERIMENTAL PROCEDURES
Due to the diversity of the tests conducted and the processes
evaluated in this investigation, the experimental equipment
utilized during the course of the study varied depending on
the specific phase of the project. Trickling filter studies
were carried out using a single-stage microtower which was
small enough to be operated indoors. A two-stage pilot trick-
ling filter was constructed on-site at the CIBA-GEIGY plant
for use in other parts of the study. Activated sludge inves-
tigations using air aeration were performed in 40-gal (150
liter) pilot units which were later sealed for use in the
oxygen aeration investigation. Activated carbon tests were
made using 1-in. (2.5 cm) and 4-in. (10 cm) column units.
In this chapter the equipment and operating procedures, analyti-
cal methods, and data correlation procedures employed during the
study are described.
CONDUCT OF THE INVESTIGATION
A schematic drawing of the present treatment facilities showing
points at which feed was taken for experimental units is shown
in Figure 9. Activated sludge investigations were performed
using raw equalized waste, unclarified trickling filter efflu-
ent, and clarified trickling filter effluent in order to deter-
mine any differences in the response of these wastes to activated
sludge treatment. Granular activated carbon was tested using
both the clarified trickling filter effluent and the activated
sludge effluent. All trickling filter and air stripping inves-
tigations were performed using raw equalized waste.
Because the principal objective of the pilot trickling filter
and microtower studies was to determine the effect of various
operating modifications on the full-scale treatment system, all
tests were conducted using equalized raw waste. A continuous
feed of raw waste from the equalization tanks was pumped to the
pilot trickling system, while feed for the microtower was taken
from the equalized raw waste sampler at 24-hr intervals. There-
fore, the feed to the microtower was subjected to fewer concen-
tration fluctuations than was the pilot trickling filter system.
27
-------
AfTTrn—.
SCREENING SOLVENT NEUTRALIZATION EQUALIZATION
SEPARATION
NJ
00
PILOT TRICKLING
FILTRATION
MICROTOWER
A.S = Act'vated Sludge
A.C - Aclivaltd Carbon
AERATED BASIN
STRIPPING
FIG. 9 .PROCESS FLOW DIAGRAM-EXISTING FULL-SCALE AND EXPERIMENTAL TREATMENT SEQUENCES
-------
The testing program with the microtower included the use of: 1)
raw waste, 2) raw waste diluted with tap water and, 3) effluent
recycle. All tests in the pilot trickling filter used raw waste as
feed to the system and effluent recycle.
Activated sludge investigations were divided into three phases.
Phase I was conducted during January and February, 1973 and had
the objective of determining the feasibility of activated sludge
treatment for this waste. During Phase I investigations, tests
were conducted using raw wastewater, unclarified trickling filter
effluent and clarified trickling filter effluent. Phase II investi-
gations were conducted during the months of March, April and May
and had the objective of determining the response of the activated
sludge system to cold temperatures (during March and April) and
of determining the longer term response of activated sludge
treatment to this waste. Based on favorable results from Phase
I, unclarified trickling filter effluent was used for the Phase
II investigation. The objective of Phase III activated sludge
investigations was to compare air and oxygen aeration for the
treatment of this wastewater. During this portion of the investi-
gation, air and oxygen activated sludge units were fed unstripped
and stripped equalized waste.
During the Phase I activated sludge investigation, units were
operated at food-to-microorganism (F/M) ratios of 0.1, 0.3, and
0.5 g BOD/day-g MLVSS. In the remaining portions of the study,
the loading of all units was maintained relatively constant at an
F/M of approximately 0.3.
Activated carbon tests were conducted to determine the feasibility
of adsorption for removing color or gross organics from both the
trickling filter and activated sludge effluents. Isotherm tests
were initially made using eight different varieties of carbon.
Based upon the results of these tests, three carbons were chosen
for further study in column contacting systems.
Air stripping investigations were conducted to determine the remov-
als of organic constituents which could be achieved in an aerated
basin or a packed tower. A 13,500-gal (51,000 liter) basin was
constructed to simulate an aerated equalization basin, while one
pilot trickling filter was used to simulate air stripping in a
packed tower. Raw equalized waste was used as a feed for both
experimental investigations. By conducting Phase III activated
sludge investigations using both stripped and unstripped waste,
effects of stripping on the response of the waste to activated
sludge treatment were determined.
29
-------
DATA CORRELATION
Data collected in this investigation were generally correlated
using established mathematical models. In trickling filter
investigations, the model which most closely represented experi-
mental findings was a first-order model used by Eckenfelder (6,
16). Activated sludge data were correlated using the models
which most closely conformed to those proposed by Eckenfelder
(17, 18, 19) and Pearson (20). The effects of temperature were
corrected using the Streeter-Phelps model (21).
EXPERIMENTAL EQUIPMENT
Pilot Trickling Filter
A schematic diagram of the pilot trickling filter used in this
study is shown in Figure 10. This system consisted of 2 ft x
2 ft (60 cm x 60 cm) towers containing 20 ft (6 m) of the
Monsanto Biodize trickling filter medium. The plastic medium
was supported in four, 5-ft (1.5 m) sections on expanded metal
grating. Each tower was supported above a reservoir with a
capacity of approximately 500 gal (2,000 liters) and was equipped
with a forced draft ventilation system adjusted to deliver 100
cfm (2.8 cu m/min) of air through the tower packing. Spray
nozzles were placed above the media in each tower at a distance
to provide proper wetting of the media.
Provision was made for piping raw equalized waste to a holding
reservoir which was used as the influent for the system. Raw
waste feed to the first column was controlled using a Fisher and
Porter flow controller. A level controller on the reservoir
insured that an adequate supply of feed was present at all times.
Pumping was provided for the influent to the first tower, from
the first to the second tower, from the second tower to the
clarifier, and for recycle of effluent from the second tower to
the top of the first tower. It was possible to adjust all pumps
to deliver flows ranging from 1 to 10 gpm (4 to 40 1 iters/min).
Level controllers were also installed in the reservoir beneath
each tower.
The system was also provided with a 5-ft (1.5 m) clarifier.
Sludge wasting from the clarifier was controlled by automatic
timers.
30
-------
Level
Controller
Flow
Controller
Raw
Waste
Overflow
HOLDING TANK
— 1
1
/K
' i \
i
-y
0
Level
Controller
TOWER I
Each tower packed
with 4,5- ft. sections
of Biodize medium
Level
Controller
TOWER 2
•Effluent
Sludge
CLARIFIER
FIG. 10. SCHEMATIC ILLUSTRATION OF RLOT TRICKLING FILTER
(m=ftx0.3)
-------
Although nutrients were being added in the full-scale system,
it was necessary to add more in the experimental systems due
to the greater quantities of BOD being removed. Nutrient feed
to the tower system was provided by continuously pumping a
mixture of (NH.kSO. and Na3PO. to the reservoir beneath the
first tower. Because effluent recycle was practiced in all
tests, an adequate supply of nutrients was available to biota
in the first tower.
Hicrotower Apparatus
The microtower used in these investigations was a laboratory-
scale trickling filter constructed in two, 10-ft (3 m) sections.
which simulated a 20-ft (6 m) plastic medium trickling filter.
The medium in the microtower consisted of inclined PVC sheets
which were designed by the manufacturer to simulate B. F.
Goodrich Koroseal media. The microtower assembly along with
auxiliary pumps and holding tanks is illustrated in Figure 11.
Equalized raw waste was fed daily into an influent tank which
was continuously mixed to prevent settling of suspended solids.
Following the adjustment of pH, when necessary, and the addition
of nutrients, this waste was fed to the microtower and was pumped
from the reservoir on the second tower to an effluent holding
tank. During effluent recycle tests, effluent was pumped from
this barrel to the top of the first microtower section. During
dilution studies, the necessary dilutions were made in the influ-
ent feed tank. Supplementary inorganic salts including NaCl
and Na2$04 were added to produce a mixed TDS concentration of
2,000 mg/1 to simulate the salts and ionic strength of the raw
wastewater.
Daily maintenance of the microtower included adjustment of flows,
measurement of effluent volumes, correction of uneven flow dis-
tribution pattern through the filter, and cleaning of the influ-
ent and effluent drums. Routine analyses included pH and TSC
daily on influent and effluent samples. When the tower was
acclimated to a particular operating condition as judged by
consistent TSC measurements, additional samples were taken.
Under stabilized operating conditions for each run, the fol-
lowing data were collected at five points through the depth of
the microtower: pH, temperature, TSC, COD, BOD, and dissolved
oxygen. In addition, oxygen uptake rates of the slime were
determined at these points in the filter. The oxygen uptake
procedure used for this investigation is described in a separate
section below.
32
-------
Mixing Unit
to
EQUALIZATION
INFLUENT
TANK
MICROTOWER
Flow
Distributor
Media Segment Approx.
2 ft long
Bi
-------
Activated Sludge Units
Continuous-flow activated sludge units were fabricated from 55-
gal (200 liter) drums and had a 40-gal (151 liter) capacity. A
schematic illustration of these units is shown in Figure 12.
Units were constructed by welding a baffle into the side of each
drum which served as a clarification section. Three-quarter inch
(2 cm) nipples welded through the side of each barrel permitted
the maintenance of a constant level in the aeration basin and
allowed effluent to overflow into containers provided for this
purpose. Diffused air aeration was supplied through 12-in. (30
cm) porous stone diffusers. These diffusers were suspended in
the tank to provide adequate mixing of the mixed liquor solids,
to maintain desired oxygen levels and to prevent turbulence in
the clarifier section.
The feed to the units was provided through a variable-speed tube
pump. Effluent, collected from various points in the treatment
plant, was collected daily and added to feed drums. Effluent
was collected in containers provided for each unit and the efflu-
ent volume was measured daily. Nitrogen and phosphorus were
added daily to the waste feed tanks in order to provide a BOD:
N:P ratio of 100:5:1. When required, the pH of the influent
waste was adjusted to approximately pH 7. When necessary, mixed
liquor was wasted from the units in order to provide an operating
MLVSS level of 2,500 mg/1. After sludge was wasted, effluent
collected during the previous 24-hr period was added to make up
the lost volume in the aeration basin.
Operation of the activated sludge units was initiated by using an
equal mixture of sludge obtained from the Cranston, Rhode Island
municipal activated sludge plant and clarifier sludge from the
CIBA-GEIGY trickling filter system. Operation of the units was
begun the first of December, 1972. Approximately five weeks were
required for complete acclimation of the biological solids to
this waste.
Flow to the units was adjusted daily in order to maintain the
desired F/M ratios for each unit. Upon occasion it was necessary
to adjust the pH of the aeration basin by adding 50 percent NaOH.
In addition it was frequently necessary to add a silicone defoam-
ing agent to the aeration basin to suppress frothing.
The influent, effluent, and contents of the aeration basin were
sampled daily. Analyses which were made on the influent and
34
-------
UJ
Waste ^-^—
Feed —(- )
Drum X-A
Compressed
Air Supply
Air
Stone
TC,
\
Baffle
55-gal. drum
\
Overflow
Pipe
Effluent
Drum
Activated
Sludge Unit
FIG. 12. SCHEMATIC ILLUSTRATION OF ACTIVATED SLUDGE UNITS
(Liters = gal x 3.8)
-------
effluent samples daily included pH, TSC, and suspended and volatile
suspended solids. These samples were analyzed three times per week
for total and soluble BOD and COD.
Contents of the aeration basin were analyzed daily for total and vola-
tile mixed liquor suspended solids and pH. These samples were analy-
zed four times per week for oxygen uptake rate, zone settling velo-
city, and sludge volume index. Weekly samples were taken for total
kjeldahl nitrogen and total phosphorus measurements. Occasionally,
samples were analyzed for solvents.
During Phase I investigations, activated sludge units were operated
indoors at a waste temperature of approximately 17°C. During Phase
II investigations, one unit was moved outside in order to determine
the effect of temperature on activated sludge performance. In Phase
III investigations, all units were moved outdoors where the tempera-
ture varied from approximately 20 to 24°C.
During the Phase III activated sludge investigations, two units were
modified for use as oxygen activated sludge units. A schematic
illustration of these units is shown in Figure 13. Modification of
these units essentially consisted of adding a sealed cover to each
unit so that an oxygen-rich atmosphere could be maintained above
the liquid. Additional piping in the covers for these units was
provided to allow for the introduction of oxygen, withdrawal of off-
gases, recirculation of gas to the mixed liquor, and introduction of
feed to the unit. A gas compressor was piped into each unit to pro-
vide a constant diffused air supply for the mixed liquor. These
compressors had a capacity of approximately 1 scfm (0.03 cu m/min)
and were adequate to maintain proper mixing and dissolved oxygen
levels in the mixed liquor. Additional piping was added to allow
the withdrawal and introduction of mixed liquor samples for daily
testing.
Oxygen feed to the units was supplied from gas cylinders. Influent
oxygen flow was controlled using a stainless steel metering valve
and was monitored using gas rotometers. Off-gas measurement was
accomplished using wet test meters. Operational procedures for the
oxygen activated sludge units were essentially the same as those
followed for air activated sludge.
Activated Carbon Tests
Carbon Isotherms. Batch adsorption tests were conducted by con-
tacting waste samples and granular activated carbon in a wrist-action
36
-------
Waste
Feed
Drum
Oxygen
Supply
Sample
Port
Activated
Sludge Unit
FIG.I3. SCHEMATIC ILLUSTRATION OF OXYGEN ACTIVATED SLUDGE
UNITS
(Liters = gal x 3.8)
-------
laboratory shaker for a 24-hr period. Although powdered carbon
is frequently used for isotherm tests, results using the granular
carbon are more representative since waste constituents must
diffuse through pores in the carbon particles before being ad-
sorbed onto the carbon. Granular carbon was sieved to obtain
the desired range of particle sizes, soaked in distilled water
for 24 hr, oven dried at 103°C for 24 hr, and cooled in a dessi-
cator. Varying amounts of carbon were added to each test flask,
and then contacted for 24 hr with 200 ml of waste which had been
previously filtered through Reeve Angel Grade 802 filter paper.
Usually carbon was added to obtain COD to carbon ratios ranging
from 0.05 to 5.0 g/g.
After 24 hr the waste-carbon mixture was filtered through GF/C
glass fiber filter paper to separate the carbon from the waste.
A control sample was subjected to the same test procedure as
each carbon sample. Initial tests conducted using varying con-
tact times determined that the 24-hr period would be sufficient
for all practical purposes to obtain equilibrium between the
carbon and wastewater.
Column Units. One carbon column unit was fabricated from 1-in.
(2.54 cm) Pyrex glass tubes and is illustrated in Figure 14. Feed
to these columns was provided using a tube pump. Columns were
operated upflow with each column containing approximately 3 cm
of glass beads supporting 60 cm of activated carbon. The columns
were operated at a flow rate of 2.5 gpm/sq ft (0.1 cu m/min-sq m).
Clamps were provided at the top of each column so that each carbon
bed could be backwashed individually. Solids accumulation made it
necessary to backwash each column daily. Grab samples were col-
lected daily from the influent and effluent from each column.
Daily analyses were made for TSC, suspended solids, turbidity, and
color. In addition, samples were analyzed twice per week for BOD
and COD.
A set of four, 4-in. plexiglass columns were used for testing ad-
sorption of the trickling filter effluent. Feed for these columns
was provided by constantly pumping clarifier effluent from the
full-scale treatment system into the control building. Waste was
pumped through the columns at a rate of 5 gpm/sq ft. Grab samples
of the influent waste and the effluent from each column were taken
periodically and analyzed for COD, BOD, TSC, color and suspended
solids. Columns were backwashed as necessary to prevent excessive
headloss.
38
-------
u>
FULL SCALE
TRICKLING
FILTERS
PILOT ACTIVATED
SLUDGE UNITS
F/M • 0.3 -0.5
Screw
Clamps
2.5 cm Pyrex
pipe
170 g
Carbon
Glass
Beads
PUMP
to
-14-
V
HOLDING TANK SAMPLING POINTS EFFLUENT TANK
FIG. 14. SCHEMATIC ILLUSTRATION OF ACTIVATED CARBON COLUMNS
-------
Packed Stripping Tower
The removal of organic constituents which could be achieved by air
stripping in a packed tower was investigated by converting the
first pilot trickling filter for this purpose. Necessary modifi-
cations included the installation of a 2,500 cfm (70 cu m/min)
blower into the side of the reservoir of the tower and the addition
of a 40-gpm (151 liter/min) pump to supply influent to the top of
the tower. The influent pump was piped to receive the equalized
raw waste flow in the influent reservoir for the pilot trickling
filter system.
Influent and effluent samples were collected during each run and
analyzed for BOD, COD, and suspended solids. In addition, occa-
sional samples were analyzed for solvent concentrations.
Aerated Equalization Basin
Because of the possibility of adding additional equalization capa-
city to the treatment plant, the feasibility of achieving removal
of organic constituents by air stripping in an aerated basin was
investigated. For this purpose a 24-ft (7.3 m) diameter, 5-ft
(1.5 m) deep swimming pool was constructed at the treatment fa-
cility. The basin was designed for a water depth of 4 ft (1.3 m)
and had a resulting capacity of 13,500 gal (51,000 liters).
Aeration was provided by a 2-hp (1.5 kw) aerator (Mixing Equipment
Comi.,-ny, Rochester, N. Y.). This corresponded to a power level of
150 hp/mil gal (0.03 kw/cu m). Influent to the basin was provided
from the equalization basin. An inverted U-tube located at the
outlet of the basin was used to maintain a liquid level of 4 ft
(1.2 m). Samples of the basin influent and effluent were taken
daily and analyzed for BOD, COD, and suspended solids. An illus-
tration of this basin is shown in Figure 15.
SAMPLING AND ANALYTICAL METHODS
Analytical methods used during the course of this investigation
conformed to those prescribed in Standard Methods for the Examina-
tion of Hater and Mastewater (13th edition).The graduate dilution
method was used for BOD analyses except in a few cases when time
dictated that the pipetting method be used. Soluble BOD and COD
determinations were made by filtering samples through glass fiber
(GF/C) filter paper.
40
-------
Effluent
Influent
Woste
C_—- 2 h.p. Aerator
24 ft dia. x 4 ft. deep
basin - Vol. = 13,500 gal
TOP VIEW
Steel
Support
Concrete
Foundation
SIDE VIEW
FIG. 15. AERATED EQUALIZATION AND STRIPPING
BASIN
(m = ft x0.3,kw = hpx 0.77, cu m =galx 0.004)
41
-------
Total carbon analyses were initially made using an Ionics total
carbon analyzer and later using a Beckman Model 915 analyzer.
Tests were made to insure that identical results were obtained
using both instruments. Inorganic carbon analyses determined
from raw waste samples at the beginning of the investigation
showed that an average of approximately 5 mg/1 were present.
Therefore, in subsequent analyses, except as noted, only total
carbon determinations were made. Based on recommendations of
the manufacturer of the total carbon analyzer, all samples were
filtered through Whatman No. 2 paper prior to analysis. For
consistency total carbon values determined in this investiga-
tion are referred to as total soluble carbon, TSC. For this
waste the variation between TSC and total organic carbon (TOC)
was approximately 0.5 percent for the untreated wastewater.
Solvent concentrations were determined using gas chromatographic
procedures employed in the CIBA-GEIGY analytical laboratory.
Because the oxygen uptake and zone settling velocity tests are
not described in Standard Methods, these tests are described in
detail separately.
Oxygen Uptake Test
The oxygen uptake rate was measured by placing mixed liquor from
the aeration chamber in a BOD bottle. A dissolved oxygen probe
was inserted, insuring that no air bubbles were trapped inside,
and the mixer turned on to maintain the contents of the bottle
in suspension.
Oxygen uptake measurements in the trickling filter systems were
made by taking plastic medium patches and immersing them in
waste samples taken at the same depth in the filter system.
The decrease of oxygen in the waste was recorded using a dis-
solved oxygen probe and chart recorder. The results were re-
corded as milligrams of oxygen used per square centimeter of
plastic medium per hour. For the microtower system, patches
cut from the medium were placed in the microtower at 4, 8, 14,
and 16-ft (1.2, 2.4, 4.3, 4.9 m) depths. In order to determine
the oxygen uptake rates of the biological slime, the patch was
removed from the tower and scraped free of siime except for a
3-in. x 3-in. (7.6 cm x 7.6 cm) area of active slime. Subse-
quently, the patch was placed in a closed 1-liter container
42
-------
filled with waste and the oxygen depletion recorded with time.
Patches were located in the microtower so that they could be
replaced after each determination.
To make oxygen uptake determinations in the full-scale trick-
ling filter system, a 5-in. x 5 3/4-in. (13 cm x 15 cm)
piece of plastic medium was cut from a larger section located
at mid-depth in each tower. Care was taken to obtain sample
patches located far enough into the tower to receive a repre-
sentative portion of waste flow.
Zone Settling Velocity
The zone settling velocity test (ZSV) expresses the rate of
interface settling of particular sludqe in ft/hr (m/hr). The
ZSV test is conducted by filling a 1,000-ml cylinder with mixed
liquor suspended solids from the aeration chamber. The con-
tents were slowly stirred with a mechanical device at a speed
of 4-6 revolutions/hr. This stirring was used to break up the
bridging of the sludge in the small-diameter cylinders. As
the sludge settled, the interface was recorded at selected
time intervals. The interface level was plotted against set-
tling time and the zone settling velocity calculated from the
slope of the straight line portion of the curve.
43
-------
SECTION VI
SMALL-SCALE TRICKLING FILTER INVESTIGATIONS
Investigations in small-scale trickling filter systems were under-
taken to examine and verify BOD removal mechanisms which were
occurring in the full-scale filter system, and to determine the
effects of recycle on filter performance. Tests were made both
with a single-stage microtower operated in the laboratory and
a two-stage plastia,medium pilot trickling filter. Investiga-
tions were conducted using the microtowers in order to gain
increased operating flexibility compared to the pilot system.
MICROTOWER TESTS
Equalized raw wastewater, collected continuously from the influ-
ent to the full-scale system, was fed daily to the microtower.
Investigations in the microtower system consisted of three parts:
1) tests using a raw waste feed, 2) tests using a raw waste feed
diluted with 0.5, 1, or 2 parts of tap water, and 3) tests in
which filter effluent was recycled along with raw waste. Remov-
able sections of plastic packing were placed into the microtower
at several depths in order that oxygen uptake determinations
could be easily made.
Effect of Variations in Organic Loading
A range of raw waste flows varying from 0.66 to 3.10 gpm/sq ft
(0.03 to 0.13 cu m/min-sq m) was applied to the filter in order
to study its response to varying concentrations and organic load-
ings. The influent COD concentration for these tests varied
between approximately 4,000 and 6,600 mg/1. The corresponding
organic loading varied between approximately 1,000 and 13,000
Ib COD/day-1,000 cu ft (16 and 200 kg COD/day-cu m). In dilution
studies, raw waste feed was diluted with 0.5 to 2.0 parts of tap
water. Applied organic loadings in these tests varied from
approximately 1,000 to 6,000 Ib COD/day-1,000 cu ft (16 to 96
kg COD/day-cu m) with applied COD concentrations ranging from
approximately 750 to 2,400 mg/1 COD. Hydraulic application rates
for the dilution tests were either 1.4 or 2.1 gpm/sq ft (0.06 or
0.09 cu m/min-sq m). Thus, in these tests the applied COD con-
centration for corresponding loading rates (Ib COD applied/day-
1,000 cu ft or kg COD applied/day-cu m) was significantly lower
44
-------
compared to tests in which undiluted raw waste was applied to the
filter. The results of these tests, shown in Figure 16, reveal
that significantly greater percentage COD removals were obtained
at corresponding organic loadings when the undiluted raw waste
was applied to the filter. Differences in performance might be
due to either differences in hydraulic loading or to the concen-
tration differences. It is felt that concentration differences
largely explain the difference in performance due to reduced
removals by air stripping at the lower applied concentrations.
In the effluent recycle studies, hydraulic loadings applied to
the filter ranged from 1.9 to 2.8 gpm/sq ft (0.08 to 0.11 cu
m/min-sq m). The COD removal efficiency for these tests, shown
as a function of applied organic loading, is illustrated in
Figure 17. In this case, the removal was calculated based on
the organic concentration of the raw wastestream. It can be
seen that organic removals obtained using effluent recycle are
much greater for low organic loadings than was obtained in other
tests. The higher removals obtained at low organic loadings
corresponded to the highest recycle ratios.
These results show that optimum conditions for biological
activity in the trickling filter may be produced by recycling
effluent. On the other hand, changes in operational conditions
to optimize biological activity in the filters also result in
a decrease in the effectiveness of the other principal BOD
removal mechanism, i.e. air stripping. The results in Figure
18 show that increasing the recirculation ratio to approx-
mately six greatly increased removals which can be achieved
in the trickling filter. However, increasing the recycle
ratios above six had very little additional effect on the
improvement of removal efficiency.
Oxygen Uptake Measurements
The oxygen uptake rate is illustrated as a function of applied
organic load in Figure 19. These data show that oxygen use in
the raw waste tests increased slightly with applied organic load,
but that the absolute values were significantly lower than in the
dilution and recycle studies. Oxygen utilization in dilution
tests was relatively constant over the range of loadings applied.
In effluent recycle tests, oxygen utilization was highest at low
45
-------
60
40
LJ
tr
§ 20
0
0
-Raw Waste Tests
Dilution Tests
2000 4000 6000 8000 10,000 12,000 14,000
ORGANIC LOADING (Ib COD/day-IOOO cu ft)
FIG. 16 .VARIATION OF REMOVAL EFFLUENT WITH APPLIED COD
LOAD
(kg/cu m-day =lb/IOOO cu ft-day x 0.016)
-------
100
80
60
5 40
O 20
LU
CE
Q
O
O
CC
O
Q
O
CD
0
0
100
80
60
40
20
0
COD data
BOD data
1000 2000 3000 4000 5000 6000 7000
0 1000 2000 3000 4000 5000 6000
APPLIED ORGANIC LOAD (Ib/day-IOOO cuft)
FIG. 17 .REMOVAL EFFICIENCY VERSUS APPLIED ORGANIC LOADING
IN THE RECIRCULATION STUDY
(kg/cu m-day = lb/IOOO cu ft-day x 0.016)
-------
a>
CO
,co
CO
o
LJ
tr
o
o
0
o:
o
Q
o
CD
BOD data
Raw Waste Flow = 0.3-1.7
gpm/sq fl
0
0
3456
RECYCLE RATIO, N
FIG. 18 .BOD AND COD REMOVAL EFFICIENCY VERSUS RECYCLE
RATIO
-------
2000
4000 6000 8000 IQOOO
ORGANIC LOADING (Ib COD/day-1000 cu ft)
12,000
14,000
FIG. 19 .EFFECTS OF ORGANIC LOADING ON OXYGEN UPTAKE RATE
(kg/cu m-day = lb/IOOO cu ft-day x 0.016)
-------
organic loadings due to increased biological activity. Total
oxygen uptake is shown as a function of the recycle ratio in
Figure 20. These data show that oxygen use increased with
increasing dilution of the influent waste, but decreased at
very high recycle ratios probably due to the significant de-
crease in organic loading.
Oxygen utilization shown as a function of the quantity of COD
removed in Figure 21 was used to determine oxygen utilization
coefficients. The data in this figure were divided into
three separate parts: 1) raw waste tests, 2) dilution tests
having a dilution ratio of 0.5 to 2 and recycle tests having
recycle ratios ranging from 0.2 to 1.9, and 3) recycle tests
having recycle ratios ranging from 2.8 to 8.4. The slope of
the line defined by each set of data yields the oxygen utili-
zation per gram of COD removed. The data show that signifi-
cantly higher quantities of oxygen were used per gram of COD
removed as the dilution of the influent waste increased. The
value for a' of 0.6 g 02/g COD removed in the recycle studies
was almost identical to the oxygen utilization value of 0.67
g 02/g COD removed measured in activated sludge investigations
using air aeration. The ordinate intercept of each line in
this figure defines the endogenous oxygen requirement for
microorganisms under each operating condition. These data
indicate that the endogenous activity of the microorganisms is
inhibited by high organic loading rates. This phenomenon is
probably the result of the high concentrations or organic con-
stituents at low waste dilution ratios.
Using the oxygen utilization coefficient obtained in Figure 21,
it was possible to estimate the quantity of COD removed biologi-
cally from oxygen uptake data. Removals of COD from oxygen uptake
data were calculated according to the following relationship:
(SJ = (A0? - b')/a'
r B c
where (S ) = COD removed biologically (g/sq cm-day)
r B
A02 = total oxygen uptake (g 0?/sq cm-day)
b' = auto-oxidation uptake (g 02/sq cm-day)
a' = oxygen utilization coefficient (g 0,,/g
COD removed) "
-------
0.40
I 0.30
CT
(0
£
LJ
0.20
Q_
LJ
O
X
o
0.10
0
0
8
I 234567
RECYCLE RATIO, N
FIG.20.OXYGEN UPTAKE VERSUS RECIRCULATION RATIO
10
-------
Cn
K5
0.40
7: o.3o
CM
o
o
CT
to
UJ
CO
ID
UJ
CD
X
O
0.20
0.10
0
Recycle N=2.8-8.44
• Raw Waste Tests
A Dilution Tests
O Recycle Tests
Dilution N = 0.5-2.0
Recycle N = 0.2 - 1.93
Autooxidation
Uptake
0
Raw Waste
i.o 2.0
COD REMOVED (g CODr/sq cm-hr)
3.0
FIG.21 .OXYGEN UTILIZATION IN THE MICROTOWER
-------
The calculated COD removed according to oxygen uptake data (SrOn
is compared to the measured COD removal, (Sr)T, as a function B'
of the applied COD concentration, Sa, in Figure 22. These data
show that the fraction of COD removal accounted for by biological
mechanisms increased significantly as the applied COD concentration
to the filter decreased. At high applied COD concentrations these
data indicate that only approximately 10 percent COD removal
was due to biological mechanisms. For high influent COD
concentrations, this value was significantly lower than that esti-
mated for the full-scale treatment system. However, the results
for the microtower were obtained from a single-stage trickling
filter system, while oxygen uptake values discussed for full-scale
system were total values for all six stages.
PILOT TRICKLING FILTER INVESTIGATIONS
A two-stage pilot trickling filter consisting of two 20-ft (6 m)
towers having cross-sectional areas of 4 sq ft (1.4 sq m) were
constructed to determine the effect of certain operational changes
on the full-scale trickling filter system. The objective of these
investigations was to determine the operating characteristics of
the pilot system at various recirculation ratios and to determine
the degree of BOD removal by air stripping occurring in the first
part of the full-scale system. The pilot trickling filter was
constructed so that equalized raw waste from the riser to the
first tower of the full-scale system could be used as feed for the
pilot system. Considerable difficulty was experienced in generat-
ing an adequate biological slime in these towers. Growth in the
pilot system was very similar to the small amount of growth ob-
served in the full-scale system. However, no difficulty was en-
countered in obtaining an adequate growth of microorganisms in
the microtower. The difference might have been due either to the
lesser degree of equalization provided in the pilot and full-scale
systems or to differences in the characteristics of the plastic
medium used. Pieces of wood placed at mid-depth in the second
pilot tower showed a somewhat greater growth of slime than did
the plastic medium, but the growth was still less luxuriant than
would be expected for the organic loading rates which were used
in this system. During the months of July and August, it appeared
that the maximum amount of growth obtainable had been generated
into pilot systems. Therefore, biological studies were initiated
at that time. Prior to growing a biological slime in the pilot
system, removals across the two pilot towers were monitored
53
-------
1.0
CL
or
to
p
LJ
m
0.8
0.6
§
LJ
a:
o
o
0.4
0.2
0
0
• Raw Waste Tests
O Recycle Tests
A Dilution Tests
1000 2000 3000 4000 5000 6000
APPLIED COD CONCENTRATION, Sa (mg/l)
7000
FIG.22. FRACTION OF COD REMOVAL DUE TO OXYGEN UPTAKE IN
THE MICROTOWER
-------
in order to determine the amount of stripping occurring in the
first part of the full-scale system. The results of these investi-
gations are shown in Table 3. The average removal achieved over this
period was significantly greater for Tower 1 than Tower 2 indicating
that most volatile substances were removed through the first tower,
while additional volatile materials were removed in the second
tower, but to a lesser degree. Total carbon removals averaged
9.6 and 4.7 percent for Towers 1 and 2, respectively. Corresponding
total carbon removals for days on which data were available for
both the pilot and full-scale systems were 6.8 and 7.1 percent for
the pilot system and 6.8 and 7.7 percent for the first two towers
in the full-scale system. These data indicate that practically
all removal of COD in the first two towers was due to the air
stripping mechanisms.
Recycle investigations in the single-stage microtower indicated
that approximately 85 to 90 percent BOD removal could be achieved
at recycle ratios greater than 6.0. While this degree of recycle
would not be possible in the full-scale system, further tests
in the pilot towers were undertaken to determine removals which
could be obtained with recycle ratios ranging between 1.0 and 3.0.
The effect of recycle on COD removal in the pilot filter system
is shown in Figure 23. The recycle data indicate that the per-
formance of the full-scale trickling filter could be improved by
adding recycle capabilities. Organic loading rates to the first
tower for data shown in this figure were 915, 984, 770, and 152
Ib COD/day-1,000 cu ft (1.47, 1.58, 1.23 and 0.24 kg COD/day-cu m)
for recycle ratios of 1.0, 1.4, 1.67, and 3.0, respectively.
Because the strength of waste was somewhat lower than average
during the period of these studies, organic loads applied to the
filters were less than would be anticipated for the full-scale
treatment facility. The variation of COD removal with organic
loading is presented in Figure 24. These results show that the
major fraction of COD removal occurred through the first tower
indicating that the rate of degradation will decrease through the
system. This might be due to stripping effects or to the removal
of more readily degradable substances in the first tower.
55
-------
TABLE 3
STRIPPING RESULTS FROM PILOT TRICKLING FILTER
Date
(1973)
3/12
3/14
3/15
3/16
3/20
3/21
3/23
3/24
3/29
3/3,
4/2
14/3
Average
COD (niq/1)
Influent
-
6720
-
5920
-
4200
3460
-
5040
2940
Tower 1 Eff.
-
5720
-
5670
-
3400
3840
-
4640
2240
Tower 2 Eff.
-
5080
-
4800
-
3800
5080
-
4520
2280
TSC (ma/1)
Influent
1450
1850
1640
1530
875
1115
-
1325
1140
1160
750
975
Tower 1 Eff.
1300
1690
1420
1600
450
1025
-
1160
1140
1090
650
950
Tower 2 Fff
1230
1500
1420
1295
470
1090
-
1185
1115
1115
600
875
11.9* 6.4* 9.6% 4.7%
56
-------
100
80
< 60
o
UJ
40
Q
O
o
20
0
0
A I Tower
• 2 Towers in series
1.0 2.0
RECYCLE RATIO, N
3.0
FIG.23.EFFECT OF RECYCLE ON PILOT TRICKLING FILTER
PERFORMANCE
-------
Ul
00
100
80
S60
o
LJ
CC
Q
O
O
20
0
0
500 1,000
COD LOADING (
1,500
2,000
Ib
1,000 cu f t-day
FIG.24. VARIATION OF PILOT TRICKLING FILTER PERFORMANCE
AND ORGANIC LOADING
(kg/cu m-day = lb/1000 cu ft-day x 0.016)
-------
SECTION VII
FULL-SCALE TRICKLING FILTER INVESTIGATIONS
Performance of the present trickling filter facility was monitored
for a nine-month period beginning in November, 1972. Following
modification of the system to provide effluent recycle, additional
tests were made during April and May, 1974. In addition to daily
measures of plant operations and treatment efficiency, data were
collected to determine organic carbon removals and oxygen availa-
bility in each of the six trickling filter stages. Additional
data taken to provide further insight regarding system behavior
included oxygen uptake rates using biota from the trickling filter
and analysis for individual substances present in the wastestream.
From November, 1972 through May, 1973 all raw equalized waste and
trickling filter effluent samples were composited continuously over
24-hr periods. During June'and July, 1973 increased biological
activity in the waste due to higher temperatures made it advisable
to implement another sampling procedure. Over this time, four
grab samples were taken during each 24-hr period and composited for
one daily sample. While some samples of waste from individual
towers were taken continuously over a 24-hr period, most analyses
were made from grab samples taken on the morning of the day on
which analyses were performed.
Due to several upsets in the neutralization system during the months
of December and January, the trickling filters were reseeded twice
during February using mixed liquor from a nearby municipal activated
sludge plant. This procedure was undertaken to insure as uniform
a growth of biota as possible for subsequent studies. However, after
reseeding the system, no definite beneficial effects were observed.
In February, 1973, a batch equalization procedure was implemented to
assure greater control over the influent waste pH. Beginning in
March, 1973, only one pH upset occurred during the subsequent five-
month period.
This chapter contains all data concerning the operation of the full-
scale treatment facility along with an interpretation of the findings
as they pertain to the upgrading of the system to meet anticipated
regulatory requirements. A summary of operational data is presented
in Appendix B.
59
-------
ORGANIC REMOVAL PERFORMANCE
Variation of Organic Loading and Removal Efficiency
Due to variation in waste organic strength, applied flow rate, and
seasonal temperatures, several fluctuations in trickling filter
performance were observed during the course of the investigation.
The chronological variation of BOD removal in the treatment system
over this period is presented in Figure 25. Performance of the
system averaged for monthly periods beginning in November, 1972
are shown in Table 4. While it appeared during March and April
that a significant improvement in trickling filter performance
would be realized in warmer weather, the results for May and June
show that another factor significantly reduced removals for these
two months. Data presented in Figure 25 indicate that the in-
creased BOD reductions (judged on a percentage basis) observed
during March and April were due to the decreased influent waste
load, since the quantity of BOD removed remained at approximately
100 lb/1,000 cu ft-day for January through June. Following that
time, 200 gpm (0.75 cu m/min) of cooling water was removed from
the plant process wastestream resulting in increased BOD concen-
trations beginning the last half of June. Organic loadings ap-
plied to the filter system (Ib BOD/day-1,000 cu ft or kg BOD/
day-cu m) were higher than average during June and July due to
intensive production schedules.
During February, March, and April, investigations were also under-
taken to determine the performance of the system at reduced loading
rates. Therefore, during March and April, changes were made in
several variables which are expected to have a positive effect on
treatment plant performance. These included: 1) a decrease in
the influent BOD concentration, 2) a decrease in the applied hy-
draulic and organic loadings, and 3) an increase in temperature.
While it is not possible to precisely identify the effect of these
changes on system performance, it appeared that decreases in hy-
draulic loading did not produce significant improvements in system
performance.
The relationship of applied BOD load to the quantity removed at
various hydraulic loading rates is illustrated in Figure 26. A
line which best fits these data indicates that an average of 33
percent BOD was removed in the system. These data also indicate
that increasing quantities of BOD were removed at organic loading
60
-------
100
~ 80
_j
^ 60
UJ
oc
40
20
0
i i
1 1 r
IN 12-3 12-31 1-28 2-24 3-25 4-22 5-20 6-17 7-15 7-29
TIME (weeks)
l-l 12-3 12-31 1-28 2-24 3-25 4-22 5-20 6-17 7-15 7-29
TIME (weeks)
FIG. 25 .CHRONOLOGICAL VARIATION OF BOD REMOVAL
(kg/cum-day = lb/IOOOcu ft-day x 0.016)
61
-------
TABLE 4
MONTHLY TRICKLING FILTER PERFORMANCE
Date
(1972-73)
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
BOD (mg/1 }
Inf
Total
3300
4600
3204
3574
3002
1577
1735
3059
4345
Eff
Total
-
3538
2557
2642
1731
805
1147
2365
2537
SoT
-
Rema
(i)
-
3310 '23.1
2098
2548
1566
619
1054
2728
2287
20.2
26.1
42.3
49.0
33.9
22.7
41.6
COD (mg/1 )
Inf
Total
4279
5240
5316
5012
4449
3002
2869
5029
6333
Eff
Total
2844
4278
4142
3637
2741
1695
2219
3870
3881
Sol
2841
3549
3764
3279
2301
1368
2026
3576
3578
Rema
(*)
33.5
18.4
22.T
27.4
38.4
43.5
22.6
23.0
38.7
TSC (mg/1)
Inf
Total
1115
1586
1402
1407
1257
828
800
1225
1376
Eff
Total
-
-
-
-
-
-
-
-
_
Sol
853
1244
1143
1186
810
493
663
948
1011
I
Rem
(*)
23.5
21.6
18.5
15.7
35.6
40.5
17.1
22.6
26.5
Temperature
(#)
Inf
63
62
58
59
62
65
70
88
73
Eff
59
58
55
54
60
63
64
71
81
Air
41
34
31
31
43
49
57
68
78
Avg
Flow .
(mad)0'
1.13
1.05
1 .19
0.87
0.56
0.89
0.99
0.95
1 .0
1
System Loading1" i
(Ib organic/
1000 cu'ft-day)
BOD
518
671
530
432
223
195
239
404
604
COD
627 1
765
879
606
346
371
395
664
880
Based on total influent and effluent values.
Flow in cu m/day = mgd x 3.7.
cLoading to first filter is 6 times these values: loading in kg/cu m-day = lb/1000 cu ft-day x 0.016.
-------
400
CD
3300
•200 -
§
o
UJ
-------
rates as high as approximately 750 Ib BOD/day-1,000 cu ft (12
kg BOD/day-cu m). While the BOD load applied to these filters
ranged higher than usually employed, the quantity of BOD re-
moved was significantly lower than expected. For readily de-
gradable wastes, approximately 80 to 90 percent BOD removal
would be expected for loadings ranging from 100 to 200 Ib BOD/
day-1,000 cu ft (1.6 to 3.2 kg BOD/day-cu m). Removals observed
in this range shown in Figure 26 were only approximately 30
to 70 percent.
Effect of pH
During the months of November, December, and January, problems
were encountered in adequately neutralizing the feed to the
trickling filter system. In order to demonstrate the adverse
effect of these pH fluctuations, COD removal is shown as a
function of the pH range during a particular 24-hr period in
Figure 27. Reductions in COD removal efficiency were most
noted for pH fluctuations of 3 or more pH units. The pH range
identified in this figure refers to the difference in the
minimum and maximum pH values observed during a 24-hr period
which was sustained for at least 30 min.
BOD Removal Relationships
A first-order relationship was used to describe BOD removal in
the trickling filter system:
c -KD/Q
(1 + N) - N
S = influent organic concentration, mg/1
S = effluent organic concentration, mg/1
D = filter depth, ft or m
Q^ = hydraulic loading, gpm/sq ft or cu m/min-sq m
K = substrate removal rate constant, (gpm/sq ft)n/ft or
(cu m/min-sq m) /m
n = flow exponent
N = recirculation ratio, recycled flow/raw waste flow
64
-------
400
D
TJ
3 300
o
.o
~ 200
UJ
cc
o
o
o
100
0
Data from
11/18/72 to 1/16/73
I I
J i
02^68
pH FLUCTUATION DURING 24-HR PERIOD(max.pH-min.pH)(pH Units)
FIG. 27. EFFECT OF pH FLUCTUATION ON TRICKLING FILTER
PERFORMANCE
(kg/cu m-day= Ib/IOOO cu ft-day x 0.016)
-------
Preliminary data collected during this investigation indicated
that the exponent, n, had a value of 1.0. Variation of the sub-
strate removal rate constant, K, was described using the Streeter-
Phelps relationship:
KT = KT 0T1"T2
'l 2
Variation of the value of K with waste and air temperature is
illustrated in Figure 28. Data used in this figure were compiled
for one-week periods throughout the investigation.
From these data, the temperature coefficient, 6, was determined
to be 1.11 based on air temperature and 1.27 based on waste tem-
perature. This is considerably higher than the values which have
been reported in treating other wastes. However, there is some
fundamental basis for the observation of higher 6 values for com-
plex organic wastewaters such as this one.
The variation of the data shown in Figure 28 was most likely due
to changes in waste composition that might have occurred during
the nine-month period of observation. The line of best fit in
this figure was drawn including all points obtained from November,
1972 through May 26, 1973. Data obtained during the months of
June and July are also shown in Figure 28, but there was a signi-
ficant change in waste treatability during this period which can-
not be attributed solely to temperature differences. Points for
June and July are those having air temperatures greater than 19°C
and waste temperatures greater than 20°C.
While temperature effects on biological processes are usually
correlated to waste temperature, it can be argued that the tem-
perature effect in trickling filters is primarily a function of
air temperature. This is especially true of filter media which
is not always covered by a laminar film of wastewater.
In this case, the organisms are constantly exposed to the atmos-
phere and not submerged in waste. For the data collected in this
investigation, the correlation using waste temperature gave a
value of 6 which is extraordinarily high, while the correlation
with air temperature resulted in a value of 6 very close to that
obtained in activated sludge investigations discussed subsequently
in this report. At this time, no definitive conclusion may be
reached because of the possible influence of other variables.
66
-------
\J\JC.
| o.o,
I 0.008
i
j< Q006
0
(i! 0.004
—
H-
1
-S 001
5 0.008
1-
(/)
g 0.006
o
5 0.004
or
_i
5
o
^ 0.002
0.001
/
/
/
/
/
**
/ •
i* • / *
. /f 1 y '0 046=11!
x •
/
/ • •
••^ •
/ * *
4 • *
•
1 1 1 1 1 1 — —-
-10 -5 0 5 10 15 20 25 30
B. CORRELATION WITH AIR TEMPERATURE
FIG.28.TEMPERATURE DEPENDENCE OF THE BOD REMOVAL RATE
CONSTANT
(cu m/min-cu m=qpm/cu ftx 0.13)
67
-------
Fate of Solvents In the Trickling Filter System
Samples were taken each week to determine the change in concen-
tration of solvents in the treatment system. Correlation of the
concentrations of these solvents to BOD and TSC measurements made
for the waste indicated that the solvents accounted for a larqe
portion of the organic constituents present in the wastewater.
Variations in the concentrations of these substances through the
depth of the trickling filter system are shown in Figure 29.
Values shown in this figure represent averages for all measure-
ments made between mid-December, 1972 and July, 1973. In all
cases, methyl ethyl ketone and toluene were not observed after
the first and second towers, respectively. Detection of odors
at the tower outlets indicated that the primary removal mecha-
nism for these substances was by air stripping. The decrease
in the isopropanol concentration through the system was accompa-
nied by an increase in the acetone concentration. This suggested
that the disappearance of IPA was partially a result of chemi-
cal oxidation to acetone. However, acetone concentrations in-
creased by approximately 400 mg/1 through the system, while
IPA concentrations decreased by approximately 700 mg/1. The
increase of acetone through the system is not completely under-
stood; however, it is believed that this is due either to the
formation of by-products from biological oxidation, or to the
auto-oxidation of isopropanol through the system.
The stability of methanol concentrations through the towers is
of significance because of the highly degradable nature of this
substance. In a healthy biological system, it would be expected
that methanol would be one of the first organic substances to
be degraded. However, decreasing concentrations of isopropanol
through the system indicate that the decrease of this readily
degradable substance (IPA) might have been due to biodegradation.
Because of the heavy organic loading and consequent poor BOD
removal in the system the stability of methanol might simply be
a result of the preference of the microorganisms for isopropanol.
Generally, the data strongly indicate that some removal of organic
constituents was accomplished by air stripping (e.g., toluene)
and that, while there was evidence of biological removal of
organic compounds, the extent to which the compounds were being
removed was not as great as would be expected in a healthy biologi-
cal system under normal loading conditions.
63
-------
1500
t 1000
o
<
QC
UJ
O
O
u
500-
0 20 O 60 80 100 120
DEPTH (ft)
FIG. 29 .VARIATION OF SOLVENT CONCENTRATION THROUGH THE
TRICKLING FILTER SYSTEM
(m-ftx 3.3)
-------
OXYGEN UTILIZATION
Oxygen uptake measurements in the full-scale trickling filter
system were made to determine the effect of various operating
variables on the oxygen uptake rate of the biota and as a means
of elucidating the mechanism of BOD removal in the tower system.
Oxygen uptake rates for trickling filters are conventionally
expressed as grams of oxygen used per sq cm of media per hour.
In this application, the oxygen uptake rate will be a function
of the quantity of biota on the media sample, the organic load
to which the biota are exposed, and the presence of any toxic
or inhibitory substances in the wastestream. The variation of
oxygen uptake rates as measured through the filter system are
shown in Figure 30. The date on which the measurement was made
and the waste flow to the system are indicated for each measure-
ment. For most tests, uptake rates were relatively low through
the first two filters followed by significantly increasing values
through the rest of the system. These results indicate inhibi-
tion of the biota in the first two towers followed by decreasing
inhibition at greater tower depths. This inhibition might have
been due either to pH and organic loading shocks felt most sig-
nificantly in the first two towers or to the presence of inhibi-
tory or organic compounds which are removed through the system
either by air stripping or by biological oxidation.
While these results indicate significantly increased oxygen up-
take levels at reduced waste application rates, it should be
noted that these tests were made during the months of January,
February, and March when flows were 1,000, 500, and 350 gpm
(3.8, 1.9, 1.3 cu m/min), respectively. Thus, differences in
temperature may have had some influence on these results.
Additional oxygen uptake values determined during the months of
April and June are also shown in Figure 30. The uptake relation-
ship measured on April 26 showed one of the highest levels of
activity measured during the course of this investigation and
corresponded to the time period in which the organic applied to
the towers was the lowest during the 9-mo study. Oxygen uptake
rates measured June 14, when BOD removal performance of the fil-
ters was generally poor, showed significantly less activity.
Examination of biota samples used for oxygen uptake analyses
showed a significant increase of slime thickness through the
treatment system during the spring months. Slime thickness in
Towers 1 and 2 were approximately 1 mm or less with thickness
70
-------
O.I I
0.10
0.09
I 0.08
D>
e
LU
0.07
0.06
0.05
o 0.04
i
0.03
0.02
0.01
0
0
20
900 gpm
T=25°C
40 60 80
DEPTH (ft)
100
FIG.30. OXYGEN UPTAKE IN, FULL-SCALE
FILTERS 3
(cu m/min = gpm x 3.79 x 10 ; m = ft x 3.3)
120
71
-------
increasing in subsequent towers to approximately 2 mm in the last
three towers. During the months of May and June, it appeared
that a significantly smaller quantity of slime was present in the
system compared to the spring months. Periodic microscopic examina-
tion of slime samples indicated that a significant growth of
filamentous organisms was present in the system. Some higher
forms of life were present, principally stalked and free-swimming
ciliates. Round worms, which are usually found in trickling fil-
ters, were observed only on one occasion.
In order to determine the relative significance of the air strip-
ping and biological methods as BOD removal mechanisms, estimates
were made from oxygen uptake measurements of the quantity of BOD
removed biologically. For days on which oxygen uptake measure-
ments were made, estimates of the quantity of BOD removed biologic-
ally were obtained by assuming an oxygen utilization of 1 g/g BOD
removed. The quantity of BOD removal attributable to biological
mechanisms for each test is shown in Table 5. On most occasions,
TABLE 5
BOD REMOVAL ATTRIBUTABLE TO OXYGEN UPTAKE
Date
(1973)
1/16
2/28
3/28
4/18
4/26
6/1
6/14
6/22
BOD
applied
(lb/day)a
35,340
24,020
18,310
23,320
12,610
21 ,350
30,080
52,780
BOD removed
(lb/day)c
7,070
1,800
6,000
7,140
8,080
6,960
3,370
18,580
(%)
20
7.5
32.8
30.6
64.1
32.6
11.2
35.2
BOD Removal Due to Qo Uptake
(lb/day)a
1,490
1,780
3,970
5,750
3,920
6,980
3,250
8,700
Fraction
of BOD
applied
(%)
4.2
7.4
21.7
24.6
31.1
32.7
10.8
16.5
Fraction
of all BOD
removed
(%)
21
99
66
80
48
100
96
47
'(kg/day = Ib/day x 0.45)
72
-------
the oxygen uptake rate for the waste alone was determined at two
or three depths in the tower system. For measurements made through
the month of May, the uptake of the control waste sample was very"
small and did not significantly affect the value measured for the
uptake of the organisms attached to the filter medium. For measure-
ments made during June and July, it was necessary to determine the
oxygen uptake rate of the waste separately and correct the value
obtained for the waste plus the medium sample for the uptake of
the waste alone. Visual observations indicated that a biologi-
cal growth established itself in the sumps beneath each tower
during the summer months. Because these organisms also contri-
buted to the biological removal of organic constituents, an
estimate of the removals accomplished by these dispersed organisms
were included in estimates of biological BOD removal.
These results show that significantly greater BOD removals were
attributable to biological mechanisms during the month of April.
Smaller biological removals of BOD during the months of January
and February were probably due to lower temperatures and inade-
quate waste neutralization. The relatively poor performance
during the months of June and July cannot be explained. While
the quantity of BOD removed biologically varied considerably
during the 6-mo period of observation, it appeared that biological
mechanisms account for approximately 50 percent of the total BOD
removed in the trickling filter system.
EFFECT OF FORCED DRAFT VENTILATION ON BOD REMOVAL
As a further indication of the relative significance of air strip-
ping and biological mechanisms of BOD removal, tests were made to
determine the effect of forced draft ventilation on removal effi-
ciencies. For a one-week period during the month of May, dissolved
oxygen and total carbon values were measured for the effluent from
each tower with and without the forced air ventilation system run-
ning. After the fans had run throughout the night, samples were
taken with all fans running each morning. Following the collection
of samples, the blowers were turned off for several hours to allow
the system to equilibrate to the new conditions; subsequently,
samples were taken with the blower system not running.
The results of these tests are summarized in Table 6. The results
show that while dissolved oxygen levels were lower in each tower
when the fans were off, this level was not too low for aerobic
biological activity. Total carbon removals for this system for
the two conditions exhibited significant differences. For the
73
-------
TABLE 6
EFFECT OF TOWER BLOWERS ON DISSOLVED OXYGEN
LEVEL AND TOTAL CARBON REMOVAL3
1 Sample
Influent
Tower 1
Tower 2
Tower 3
Tower 4
Tower 5
Tower 6
Dissolved Oxygen (mq/1)
Fans on
3.1
6.7
7.0
7.0
6.9
7.0
7.3
Fans off
3.2
5.9
5.7
5.2
4.8
4.6
5.0
1
Temperature (°C)
Fans on
19.8
19.8
19.7
19.5
19.2
19.2
19.2
Fans off
21.0
21.2
21.3
21.6
21.7
Total soluble carbon
Fans onb
908
846
788
741
687
22.3 649
22.3 611
j
i
Fans offc
985
949
913
882
861
832
810
Data taken daily over period from May 14 to May 19, 1973.
}Removal averaqed 32.7 percent.
'Removal averaged 17.8 percent.
-------
one-week period, total carbon removals during periods with forced
draft ventilation averaged 32.7 percent, while the removal without
forced ventilation was only 17.8 percent. Because the dissolved
oxygen measurements indicated that D.O. was not limiting the bio-
logical performance of the system, it was concluded that the dif-
ferences in removal observed with and without the blowers running
can be attributed to air stripping of volatile components from
the wastestream. While some air stripping probably occurred when
the blowers were not running, it is felt that these reductions
were minimal due to the great reduction of air flow through the
system. While strictly quantitative judgments cannot be made on
the basis of these data, it appears that biological and air strip-
ping mechanisms were responsible for approximately the same degree
of BOD removal, i.e., 50 percent of the total removal.
FULL-SCALE EFFLUENT RECYCLE INVESTIGATIONS
Results of pilot-scale trickling filter investigations conducted
earlier in the investigation showed that effluent recycle was
capable of improving treatment performance. Based on these favor-
able results, the trickling filter system was modified to provide
continuous effluent recycle. In making this modification to the
system, the six trickling filters were divided into two parallel
systems, each containing three filters in series. Flow recorders
and controllers were added so that both the raw feed rate and the
recycle flow could be adjusted. Recycle flow consisted of unclari-
fied effluent from each of the parallel systems.
Bank A (originally Towers 1, 2, and 3) was fed 700 gpm (2.6 cu m/min)
of raw waste and 300 gpm (1.1 cu m/min) of recycle waste, while Bank B
(originally Towers 4, 5, and 6) were fed 300 gpm (1.1 cu m/min) of raw
waste and 700 gpm (2.6 cu m/min) recycle flow. Following stabilization
of this period on April 1, 1974, data were collected to demonstrate
the performance of this modified system.
A summary of data obtained for a 7-wk period is shown in Table 7.
Weekly data are presented in Table B2 of Appendix B. During the
recycle investigations, Bank A, operated at a recycle ratio of
0.43, achieved a BOD reduction of 22 percent which was somewhat
less than the 1973 average of 32 percent for six filters in series.
Bank B, operated at a recycle ratio of 2.33, achieved approximately
45 percent BOD reduction. For operation at identical hydraulic
loading rates, it would be expected that the three filters in
Bank A would achieve less BOD removal than six filters in series.
The lower organic concentration in the first filter would probably
reduce air stripping of organics but might not be significant
75
-------
TABLE 7
RECYCLE DATA SUMMARY8
Bank
A
B
Recycle
0.43
2.33
Air
temp .
(°c)
11.4
11.4
BOD results
Inf. Eff.
(mg/1)
2,203
2,203
1,711
1,216
Reduction
(*)
22.3
44.8
Applied load
(lb/1000 cu ft-day)D
820
605
COD results
Inf. Eff.
(mg/1 )
3,925
3,925
3,330
2,683
Reduction
U)
15.1
31.6
Applied load
(lb/1000 cu ft-day)
1,500
1,220
Data summarized for operation during 7-wk period from April 1 to May 19, 1974.
^Loading in kg/cu in-day = lb/1 ,000 cu ft-day x 0.016.
-------
enough to enhance BOD removal. On the other hand, reductions in
Bank B, operated at a recycle ratio of 2,33, were significantly
greater than the 1973 average. This was perhaps due to the dilu-
tion of the influent organic concentrations and dampening of flue-
tuations in the organic load to the filter.
On a BOD basis, the total load applied to Bank A (including the
BOD content of the recycle stream) was approximately 820 Ib BOD/
1,000 cu ft-day (13 kg/cu m-day), while that applied to Bank B
was approximately 600 Ib BOD/1,000 cu ft-day (9.6 kg/cu m-day).
This is somewhat higher than the total load applied to a system
without recycle. However, the load applied to the first tower
in Bank B (approximately 1,800 Ib BOD/1,000 cu ft-day or 28.8
kg/cu m-day) was significantly less than the load of approximately
3,000 Ib BOD/1,000 cu ft-day (48 kg/cu m-day) applied to the
first tower when six filters were operated in series.
It should be mentioned that the biota present in Bank B at the
beginning of these tests was probably better developed than that
present in Bank A. Operation of the recycle system over a longer
period of time might result in greater BOD reductions at a re-
cycle ratio of approximately 0.4 than those experienced in this
investigation.
77
-------
SECTION VIII
AUTOMATIC MONITORING
One objective of this grant was to evaluate the potential for
controlling the multi-stage trickling filter system using auto-
matic monitoring equipment. The methods to be evaluated in-
cluded control of the raw waste feed rate using automatic total
carbon measurements and control of the blower speed in the forced
air ventilation system using dissolved oxygen measurements.
Tests conducted during this investigation had the objective of
demonstrating the applicability of automatic monitoring equip-
ment in the trickling filter system. The investigation did not
include control of the feed rate or blower speeds; however, a
control system could be designed once the stability of monitoring
systems has been established. During the fall and winter of 1973-
1974, dissolved oxygen and total carbon monitoring equipment was
installed to measure the contents of each sump and the clarifier
effluent. A more complete description of these systems is pre-
sented in Section IV.
DISSOLVED OXYGEN MONITORING
Comparative dissolved oxygen analyses were made of measurements
from the on-line system and a laboratory dissolved oxygen meter.
For each comparison, grab samples were taken from the line feeding
each D. 0. probe. Simultaneous readings were made from the on-line
D. 0. indicator. The portable laboratory dissolved oxygen meter
was taken to the field to provide a comparison for values measured
using the on-stream system. This portable meter was calibrated in
the laboratory prior to each use. A least squares analysis was
performed on comparative analysis at each sampling point over a
one-month period. Results of these analyses and average U. 0. con-
centrations observed for both the laboratory and on-stream systems
are summarized in Table 8. During this period routine maintenance
in the system consisted of cleaning the probe assemblies at least
once each week and recalibration of each probe approximately two
times per week in order to maintain good agreement between the on-
stream and laboratory analyzers.
The data presented in Table 8 show that the average D. 0. differ-
ences between the two measurements ranged from 0.0 to 0.4 mg/1 for
78
-------
TABLE 8
COMPARATIVE DISSOLVED OXYGEN MEASUREMENTS
Sample
location
Tower 1
Tower 2
Tower 3
Tower 4
Tower 5
Tower 6
Clarifier
Average D. 0.
concentration3 (mq/1)
lab
meter
9.0
8.5
8.9
8.4
8.4
8.9
7.0
on-stream
system
8.9
8.2
8.9
8.3
8.2
8.5
6.0
Averages are from 14 comparative measure-
ments made during a one-month period.
each of the towers and 1.0 mg/1 for the clarifier probe. Consider-
able difficulty was encountered in obtaining proper operations of
the clarifier probe. Shortly before investigations with the auto-
matic monitoring system were initiated, chlorination of the efflu-
ent was begun by adding approximately 50 mg/1 chlorine to the
clarifier influent. As long as the clarifier probe remained
immersed in the waste, readings recorded in the automatic moni-
toring system for this probe continually drifted downward. Upon
changing the membrane of this probe and placing it in contact with
tap water, readings made over a one-week period indicated that
stable operation could be maintained with tap water. When the
probe was returned to the wastewater system, the same lack of
stability was noted and readings again drifted downward.
The least squares analysis of data for each sampling point indi-
cated that somewhat lower readings were obtained in the on-stream
system compared to laboratory measurements. A regression analysis
79
-------
of all data collected had a slope of 1.00 and y-intercept of -0.4
mg/1. This indicated that measurements using the laboratory meter
were approximately 0.4 mg/1 greater than measurements in the con-
tinuous-flow system.
Additional measurements were taken to determine the operation of
the system when calibrated less frequently. While collecting data
summarized in Table 8, each probe was recalibrated approximately
two times per week. During a subsequent period, probes were cali-
brated approximately once each two weeks. Data collected with
less frequent calibration showed that readings with the monitoring
systems tended to drift downward. Comparative analyses made
during this period showed that on-stream measurements averaged
approximately 1.1 mg/1 lower than equivalent measurements made
using the laboratory analyzer for each of the tower probes. Dis-
solved oxygen concentrations indicated with the clarifier probe
averaged approximately 3 mg/1 less than comparative analysis made
with the laboratory meter.
Observations made during this period showed the need for calibra-
tion of the system approximately twice each week. Experience
indicated that if the probes were not cleaned or calibrated for a
period of one week, the dissolved oxygen concentrations were
approximately 20 percent lower than measurements made using the
laboratory meter. Operationally, the system was well designed to
allow for routine maintenance. The only maintenance required
included cleaning and recalibration of each probe assembly peri-
odically. To disassemble and clean the probe assembly, change
the membrane, air calibrate the instrument, and reassemble the
probe unit required approximately two hours per probe. Thus, ap-
proximately two days were required to clean and recalibrate the
seven probes included in this system.
TOTAL CARBON MONITORING SYSTEM
During the initial one-month period of operation, distinct sample
peaks were obtained only five days. Of the other days during this
period, the instrument was not operational due to plugging of the
sample prefiltration system (1 day), a leaking sample injection
valve (16 days), and a plugged injection tube (2 days). During
this test period the analyzer was not operational due to mechanical
malfunctions 75 percent of the time. During the remaining 25 per-
cent of the test period, values were recorded; however, most of
these values did not agree well with laboratory measurements.
30
-------
Required maintenance included replacement of the sample injection
valve, packing of the reaction chamber with a new catalyst, re-
placement of the oven and mounting plate, cleaning of the flow
regulator, and complete calibration of the instrument. The ori-
ginal catalyst used in the reaction chamber became poisoned soon
after the instrument was placed into operation. Although a
second type of catalyst was evaluated, the same problem recurred.
It appeared that problems with plugging of the sample injection
tube, poisoning of the catalyst, and leakage in a block unit used
to meter samples into the reaction chamber were all caused by
tar-like and other substances in this particular wastewater.
During the five days in which distinct sample peaks were obtained
from the instrument, total carbon concentrations ranged from 50
to 65 percent of values measured using the laboratory organic
carbon analyzer. Because of the more consistent operation of the
laboratory instrument, it was apparent that values determined
using the on-stream analyzer were incorrect. Laboratory opera-
tion of the total carbon analyzer prior to its installation in
the continuous-flow system indicated that several injections
(approximately 3 to 6) were necessary to obtain consistent read-
ings from each sample. Additionally, it was found that the injec-
tion of 5 to 8 distilled water samples were necessary to purge
the reaction chamber when changing from one sample to another.
Since the multi-stream selector was designed to make only one
injection of each waste sample alternated with one injection of
distilled water, consistent operation of the instrument was not
obtained.
During a two-month test period, operation of the total carbon
analyzer was unsatisfactory. Even with a high degree of mainte-
nance, the instrument was incapable of properly measuring and
recording the total carbon of waste samples from various points
in the treatment system. It is believed that problems encountered
in using this instrument were due to the type of materials which
are not uncommon in wastewaters produced in most diversified
organic chemical plants.
81
-------
SECTION IX
ACTIVATED SLUDGE INVESTIGATIONS
Activated sludge investigations were conducted in three phases.
Phase I was initially conducted to determine the feasibility of
activated sludge treatment for this waste. Phase II investiga-
tions were continued from the Phase I investigation to determine
longer range performance of these units and to determine the
effects of temperature on removal efficiencies. Phase III tests
were conducted in order to compare differences in the response
of the waste to treatment using air and oxygen aeration. The
results of these investigations are described in this chapter.
AIR AERATION TESTS
During Phase I investigations, activated sludge units were fed
raw equalized waste, unclarified trickling filter effluent, and
clarified trickling filter effluent. The objectives of this por-
tion of the investigation were to determine the removals of organ-
ic constituents possible using activated sludge treatment, to
determine kinetic parameters required for process design, and to
determine the effect of pretreatment in the present trickling
filter system on the effectiveness of activated sludge treatment.
Five 40-gal (151 liter) activated sludge units were operated
during this portion of the study.
After a five-week stabilization period, the units were judged to
be acclimated to the waste and the comprehensive testing program
was begun. Data for these tests were collected over a seven-
week period from January 11 to March 3, 1973.
Phase II investigations were initiated on March 10 and continued
until May 26, 1973. One unit was placed outside and allowed to
operate at the ambient temperature. Two other units were operated
indoors at a temperature of approximately 18°C.
Substrate Removal
A summary of operating results for Phase I is shown in Table Cl of
Appendix C. For units receiving the unclarified trickling filter
effluent feed, the influent total BOD concentration averaged 2,270
82
-------
mg/1. Total effluent BOD concentrations averaged 15, 53, and
, ,
0 pd p f°r UnitS Operated at F/M ratios of 0.08,
0 23 and 0.40. Percentage removal efficiency ranged from 92
w Tonf* /huS' While the waste can be treated to low
effluent BOD values, these values can be achieved only at low
F/M ratios. J
Table C2 contains data for all units operated at temperatures
ranging from 14 to 18 C during Phase II. Table C3 contains data
for a three-week period in which the temperature of the outside
units averaged 6 C. Effluent BOD values were 26 and 71 mg/1
respectively, for units operated at 17 and 6°C, while the in-
fluent BOD was 1 ,100 mg/1 .
The variation of soluble effluent BOD with the substrate removal
rate for Phase I and Phase II is shown in Figure 31. These data
show that the BOD removal rate constant, k, was 0.0045 1/mg-day.
The average basin temperature during this study was 18°C.
From basic principles describing the activated sludge process, a
linear relationship is expected between soluble effluent BOD and
the substrate removal rate. Greater slopes from this relation-
ship indicate greater amenability to activated sludge treatment.
Deviation from this linear relationship is usually found at high
substrate removal rates.
Because of the significant weekly variability of this waste, a
statistical distribution of substrate removal rate constants was
developed from data collected during this part of the investiga-
tion. On a BOD basis, the substrate removal rate constant, k,
varied from approximately 0.0013 to 0.017 1/mg-day, with higher
substrate removal rate constants concurrent with periods of
clearer effluent. Because of the variability of manufacturing
operations in the plant, the long-term observation of treatability
characteristics of this waste is important for the proper design
of an activated sludge system. For operation at constant F/M
values, the effluent BOD will vary in proportion to the value of
the substrate removal rate constant.
The principal effect of temperature changes on the activated
sludge process is manifested in the effect on the substrate re-
moval rate constant, k. A decrease in temperature will decrease
the rate of biological metabolism which reduces the observed k
value. For a system treating a given volume of waste, operation
at reduced values of k will result in an increased BOD concen-
tration. To determine the effects of low temperatures on the
no
OJ
-------
Q
§
cc.
0.7
0.6
i
0.5
»
0.4
0.3
LU
a:
LU
DQ
O.i
O.O
1 1 1 r
-i r
PHASE I
A Raw Waste
• Unclarified Effluent
• Clarified Effluent
PHASE
oAD Unclarified Effluent
k = 0.0045 liter/mg-day
T= I8°C
0
20 40 60 80 100 120 140
SOLUBLE EFFLUENT BOD, Se (mg/l)
160 180
FIG. 3 I . BOD REMOVAL RATE CONSTANT - PHASES I
AND I
-------
biological process, all three units in Phase II were operated in
parallel receiving the unclarified trickling filter effluent as
feed. The relationship of soluble effluent BOD to the substrate
removal rate at different temperatures is shown in Figure 32.
The relationship used to relate the change of the substrate
removal rate constant with temperature is expressed as:
= kn
JrT2
The symbol, 8, is the temperature coefficient for the system.
From the BOD removal rate constants determined in Figure 32, the
value of 9 was determined to be 1.09. Due to the short sampling
period the collection of additional data would be advisable to
confirm this value.
Oxygen Requirements
Oxygen requirements may be calculated from a correlation of the
substrate removal rate with the oxygen utilization rate expressed
as grams oxygen utilized per gram MLVSS per day. The slope of
the line defined by this relationship represents the assimilative
oxygen utilization coefficient, a1, while the intercept on the y-
axis represents the endogenous oxygen utilization coefficient, b'.
The data illustrated in Figure 33 show that oxygen requirements
for this waste are 1.12 g Og/g BOD removed and 0.03 g 02/g MLVSS-
day. The oxygen utilization coefficient is somewhat higher than
usually encountered for aerobic biological treatment systems,
while the endogenous utilization coefficient is somewhat lower
than usually measured.
Sludge Production
The production of biological sludge is directly proportional to
the quantity of organic matter removed. The net sludge growth is
the sum of the sludge growth resulting from the removal of organic
constituents and the decrease in sludge mass due to endogenous
respiration of the microorganisms.
Since sludge production is directly related to the organic loading
rate, the net sludge production rate is correlated to the substrate
35
-------
00
ON
Q
O
CD
$0.3
o>
UJ 0.2
<
a:
UJ O.I
cr
LU
cc
CO
CD
CO
0
/
* /3k = 0.00!
• ' T=I7°C
0.0050 l/mg-day
C
Jk=0.002 l/mg-day
T=6°C
0 20 40 60 80 100
SOLUBLE EFFLUENT BOD, Se (mg/l)
FIG.32. TEMPERATURE EFFECT ON ACTIVATED SLUDGE
PERFORMANCE
-------
0.8
CD
-J
0.7
t
§ 0.6
I
9! 0.5
LJ 0.4
oc
0.3
I
5 0.2
X
o
O.I
0.0
PHASE I
A RAW WASTE
O UNCLARIFIED EFFLUENT
ti CLARIFIED EFFLUENT
PHASE IE
A UNIT C
• UNIT D
• UNIT F
'» Q03 90,/g MLVSS- doy
0.0 O.I 0.2 0.3 0.4
SUBSTRATE REMOVAL RATE,
0.5
0.6
0.7
FIG.33.DETERMINATION OF OXYGEN UTILIZATION COEFFICIENTS,
BOD BASIS- PHASES I AND H
-------
removal rate. The slope of the line represented from this rela-
tionship is the sludge yield coefficient, a, and the y-intercept
is the endogenous decay rate, b. The data correlated in Figure 34
define a sludge yield coefficient of 0.37 g VSS/g BOD removed
and an endogenous decay rate of 0.02 g VSS destroyed/g MLVSS-day.
In order to determine the total sludge solids produced, the vola-
tile content of the wasted sludge must also be known.. Measurements
made over the duration of this study showed that the volatile fraction
of mixed liquor solids ranged from 74 to 83 percent for the five units
operated.
Sludge Settleability
Sludge settling, as measured by the sludge volume index and the
zone velocity tests, was determined during the course of the in-
vestigation. Because sludge settling in small-scale units does
not adequately simulate the performance which can be achieved in
prototype clarifiers, additional tests were made to determine the
probable concentration of effluent solids from full-scale activated
sludge treatment. For most organic loadings the sludge volume
index was less than 60 ml/g. While the data indicate that the best
sludge settleability is obtained at higher organic loadings, efflu-
ent suspended solids data show that a higher concentration of
solids is present in the effluent at high loadings. Average values
for the zone settling velocity ranged from 13.6 to 24.1 ft/hr
(4.15 - 7.34 m/hr). The highest zone settling velocity was ob-
tained at an F/M of 0.23/day using the unclarified trickling filter
feed. Compared to values usually obtained for these measures of
sludge settleability, this sludge settled very well.
AIR AND OXYGEN AERATION COMPARATIVE TESTS
For the Phase III investigations two of the 55-gal (208 liter)
activated sludge units were modified to be operated using oxygen
aeration. In order to determine the effect of pretreatment by air
stripping on the treatability of the waste, one air and one oxygen
unit received a stripped waste feed. Two other units received a
feed consisting of raw equalized waste. A summary of the operating
data for this portion of the investigation is shown in Table B4 of'
Appendix B. The quantity of oxygen fed to the oxygen activated
88
-------
00
VQ
Q.
CO
o»
0.3
CO
CO
0.2
o»
O.I
o:
CD
iu 0.0
CD
Q
CO
PHASE I
A RAW WASTE
O UNCLARIFIED EFFLUENT
Q CLARIFIED EFFLUENT
PHASE H
UNIT C
UNIT D
• UNIT F
0.0
J_ b=0.02/day
O.I
0.2
0.3
SUBSTRATE REMOVAL RATE
0.4 0.5
Sr gBODr
Xvt CgMLVSS-day
0.6
0.7
FIG.34. EXCESS SLUDGE PRODUCTION, BOD BASIS-
PHASES IANDH
-------
sludge systems receiving unstripped and stripped waste required approxi-
mately 7 cu ft/day (0.2 cu m/day) to maintain a dissolved
oxygen in the mixed liquor of 5 mg/1 or higher and to maintain the
oxygen content in the off-gas less than or equal to 50 percent.
The average volumes of off-gas measured were greater than would
be expected, indicating the presence of some leaks in the system.
However, the volume of gas flowing through these systems was con-
siderably less than the approximately 1,800 cu ft/day (51 cu m/day)
which passed through the air activated sludge systems.
Mhile oxygen aeration is usually technically feasible in cases
where air activated sludge treatment may successfully be employed,
several considerations warranted a feasibility study regarding the
use of oxygen in this case. Because problems were encountered in
Phase I tests due to the tendency of the aeration basin pH to
decline below 6.0, there was concern that pH stabilization would
prove even more difficult in oxygen systems containing a high
partial pressure of C02 in the gas phase above the mixed liquor;-
An additional indication of the need for a feasibility study was
the apparent contribution of air stripping as a significant BOD
removal mechanism in the air aerated systems. Because there was
reason to believe that inhibitory constituents in the waste were
being removed by air stripping, there was the possibility that
these substances might accumulate in an oxygen aerated system and
be detrimental to treatment performance.
Substrate Removal Relationship
The BOD removal rate constants for the unstripped waste are defined
in Figure 35. The value of k for the air system was 0.0034 1/mg-day
which was somewhat less than that found in previous investigations.
This is more significant when it is realized that the average basin
temperature for these tests was 23°C compared to a value of 18°C for
previous investigations. The difference in these values is perhaps
a result of the withdrawal of some cooling flow from the plant
wastestream which might have resulted in the increase in concentra-
tions of some inhibitory substances. The BOD removal rate constant
for the oxygen system was 0.00083 1/mg-day, a value significantly
lower than that measured for the air system.
Similar BOD removal relationships for units receiving the stripped
waste feed are shown in Figure 3:6. BOD removal rate constants meas-
ured for these units were 0.0041 1/mg-day for the air unit and 0.0076
90
-------
100
700
200 300 400 500 600
SOLUBLE EFFLUENT BOO, Se (mg/l)
FK3.35.BOD REMOVAL RATE CONSTANT, UNSTRIPPED WASTE-PHASE III
0.6|—i 1 1 1 1 r i 1 1 1 r- "
? n.
to
I
o
0.5
0.4
a 0.3
_i
o
I 0.2
UJ
8 oj
Oxygen - Unit D
k=OD04l
l/mg-doy
T=23"C
Air-Unit C
0 20 40 60 80 100 120
SOLUBLE EFFLUENT BOD, Se (mg/l)
FIG. 36.BOD REMOVAL RATE CONSTANT, STRIPPED WASTE-PHASE TH
91
-------
1/mg-day for the oxygen unit. Although these data exhibited signi-
ficant scatter, it appeared that the BOD removal relationship for
the oxygen unit was somewhat better than for the air system. These
data indicate that, in fact, some volatile constituent present in
the raw wastestream was purged from the waste by the air strip-
ping pretreatment and through aeration in the air activated sludge
system. Further, it is evident that this substance was not removed,
or not nearly to such a great degree, in the oxygen system receiving
the unstripped waste feed and that this substance resulted in a
significant deterioration of performance in the closed-atmosphere
oxygen system.
Following the conclusion of-this investigation on July 28, Units
III-A and III-B were kept in operation. The operation of Unit III-B
was changed from oxygen to air aeration in order to determine
whether or not the BOD removal performance of this unit would ap-
proach that of Unit III-A under similar operating conditions. The
chronological variations of control parameters for these units are
shown in Figure 37. These data show that these units were operated
under essentially identical conditions for a four-week period and
that the effluent BOD and COD concentrations of Unit III-B decreased
and became identical to effluent values from Unit III-A during this
period. These data give further substantiation to the hypothesis
that some volatile constituents accumulated in the mixed liquor of
the oxygen-aerated units receiving the raw waste feed.
It is also significant to note that performance in the air system
was somewhat improved following the air stripping pretreatment.
This is evidenced by the k values of 0.0034 and 0.0041 1/mg-day for
units receiving the unstripped and stripped waste feed, respectively.
No adequate explanation for the superior performance of the oxygen
units compared to the air system receiving the stripped waste feed
can be offered which is consistent with other data observed for air
and oxygen activated sludge treatment.
Oxygen Utilization
Oxygen utilization data for units receiving the unstripped waste
feed are shown in Figure 38. From these data, values of the assimi-
lative oxygen utilization coefficient, a', were determined to be
1.14 and 0.81 g Oz/g BOD removed for oxygen and air aeration,
respectively. Values for units receiving the stripped waste feed
(Figure 39) were 0.89 and 0.68 g 02/g BOD removed for oxygen and air
aeration, respectively. In both cases the values for oxygen units
92
-------
6000
10 15
AUGUST
FIG.37. PERFORMANCE OF OXYGEN-ACCLIMATED SLUDGE
SUBJECTED TO AIR AERATION
93
-------
D
T3
0.7
UJ
cr
Z
UJ
&
s
0.6
0.5
0.4
0.3
0.2
O.I
t i i
Oxyge
I b' = O.07 g
g 02/gMLVSS-doy
0.6
0.7
0.8
0 O.I 0.2 0.3 0.4 0.5
SUBSTRATE REMOVAL RATE, ^
FIG.38. OXYGEN UTILIZATION FOR UNSTRIPPED WASTE, BOD BASIS-
PHASE m
o
0 O.I 0.2 0.3
SUBSTRATE REMOVAL RATE, ^r f-
Xvt \ i
0.4 0.5 0.6 07
.' g MLVSS -day j
FIG.39. OXYGEN UTILIZATION FOR STRIPPED WASTE, BOD BASIS
PHASE m
94
-------
were significantly higher than for air aeration indicating BOD
removal by air stripping.
Sludge Settleability and Effluent Solids Concentration
Data summarized in Table B4 of Appendix B show that sludge settle-
ability as measured by the sludge volume index and zone settling
velocity tests was very good for all units. However, the units
operated using oxygen aeration did exhibit lower SVI values of 37
and 48 ml/g (unstripped and stripped waste) compared to 55 and 61
ml/g for air aerated units. Values for the zone settling velocity
were not as conclusive. While sludge settling measured for all
units was very good, effluent suspended solids concentrations were
high.
SUMMARY
Activated sludge investigations showed that the waste is amenable
to treatment using this process. However, extremely low organic
loadings (F/M ratios) were required to achieve effluent BOD values
less than 30 mg/1. Process performance was significantly impaired
using oxygen aeration in a closed atmosphere system treating the
unstripped waste. Performance using oxygen aeration treating a
stripped waste was equal to or better than that obtained with air
aeration. However, the failure of the unstripped waste oxygen
unit to produce acceptable BOD values raised questions regarding
the feasibility of a staged-oxygen process treating a stripped
waste.
Values for kinetic parameters determined in this investigation are
presented in Table 9. Data correlated on a COD basis are also
presented for comparison with BOD data. Some variation of data
was apparent during the Phase III investigation when the plant
was operating on a heavy production schedule. Sludge production
data for Phase III were scattered; however, it appeared that the
production was not significantly different from that observed in
Phases I and II.
The BOD removal rate constants corrected to 20°C for different
phases of the investigation are shown in Figure 40. Data points
shown in this figure are averages for entire portions of the inves-
tigation. These data indicate the poor performance of oxygen
95
-------
TABLE 9
SUMMARY OF ACTIVATED SLUDGE DESIGN PARAMETERS
VO
Parameter
Aeration
Temperature
BOD Basis
BOD Removal Rate Constant3, k (1/mg-day)
Sludge Yield Coefficient, a (g VSS produced/g BODr)
Endogenous Respiration Rate, b (g VSS destroyed/day-
g MLVSS)
Assimilative Oxygen Utilization Coefficient, a1 (g 02/g BODr)
Endogenous Oxygen Utilization Coefficient, b1
(g 02/g HLVSS-day)
COD Basis
COD Removal Rate Constant, k (1/mg-day)
Sludge Yield Coefficient, a (g VSS produced/g CODr)
Endogenous Respiration Rate, b (g VSS destroyed/day-
g MLVSS}
Assimilative Oxygen Utilization Coefficient, a'
(g 02/g CODr)
Endogenous Oxygen Utilization Coefficient, b'
(g 02/g MLVSS-day)
Phases I and II
Air
17
0.0048
0.38
0.02
1.11
0.03
0.0025
0.18
0.02
0.67
0.03
Air
18
0.0042
0.37
0.02
1.14
0.02
0.0024
0.25
0.02
0.79
0.02
Phas
Unstripped Waste
Air
- 23
0.0034
b
c
0.81
0.07
0.0016
d
e
0.62
0.055
Oxygen
25
0.00083
b
c
1.14
0.07
0.00097
d
e
0.78
0.055
e }II
Stripped Waste
Air
23
0.0041
b
c
0.68
0.06
0.0048
d
e
0.43
0.055
Oxygen
25
0.0076
b
c
0.89
0.06
0.010
d
e
0.57
0.055
aValues given for temperatures at which Investigations were made.
^Average value of the sludge yield coefficient was 0.30 g VSS produced/g BODr. However, data were not sufficiently accurate
to permit determination of values for each unit.
cAccurate determination of b could not be made from data available. However, value was not significantly different from 0.02/day
determined In earlier tests-.
^Average value for all units was 0.18 g VSS produced/g CODr.
eAverage value appeared to be 0.02/day.
-------
0.5
VD
June
June-July
All values corrected to 20 °C
using 0 = 1.09
50
100 150 200 250 ' 500
SOLUBLE EFFLUENT BOD, Se (mg/l)
550
600
FIG.40.SUMMARY OF BOD REMOVAL RELATIONSHIPS FOR
ACTIVATED SLUDGE SYSTEMS
-------
aeration for the unstripped waste and the beneficial effect which
air stripping pretreatment had on the performance of both air and
oxygen aeration. The great weekly variations of k for Phases I
and II must be taken into account in the final process design.
Oxygen transfer characteristics for the activated sludge effluent
were determined several times during the investigation. The mini-
mun alpha value was 0.45. Repetition of the tests in warm weather
generally confirmed earlier data and demonstrated that a value of
approximately 0.4 should be used for design.
In general, activated sludge treatment was characterized by high
effluent suspended solids (90 - 175 mg/1), high refractory effluent
COD (275 - 500 mg/1), a high temperature dependence (6 = 1.09),
low oxygen transfer characteristics (a = 0.45), and a low BOD re-
moval rate constant, k. Sludge settleability was good throughout
most of the investigation.
98
-------
SECTION X
ACTIVATED CARBON INVESTIGATIONS
Activated carbon investigations were performed to determine the
feasibility of this process for achieving additional removal of
organic substances from the biological effluent. Initially,
batch isotherm tests were conducted to determine the amenability
of various carbons for use in treating this waste and to deter-
mine the effect of pH on adsorption efficiency. Subsequent
tests were performed in column systems to determine breakthrough
characteristics for the most promising carbons identified in the
batch tests. While most column tests were conducted using acti-
vated sludge effluent to determine removals of refractory organic
substances, some column tests were conducted using the more concen-
trated trickling filter effluent. Results of both batch and column
tests are summarized in this section.
ISOTHERM TESTS
Batch isotherm tests were conducted at room temperature (approxi-
mately 20°C) by contacting the carbon with activated sludge efflu-
ent for a 24-hr period in flasks attached to a wrist-action shaker.
In most cases, several runs were made for each individual carbon
in order to minimize the effect of variations in wastewater compo-
sition. It was found that while the adsorption capacities ob-
tained for individual runs might vary significantly with wastewater
composition, the relative ranking of various carbons by capacity
remained basically unchanged for most of the runs.
Results from the isotherm tests for COD removal were correlated
using the Freundlich isotherm. However, color removal data did
not follow this isotherm and was correlated using the relationship:
y - A (xk)
where y is the equilibrium in color concentration in APHA units,
X is the carbon dose in mg/1, A is the ordinate intercept in APHA
units, and k is a constant determined from the slope of the line
99
-------
obtained plotting equilibrium color concentration versus carbon
dose. Using this relationship lower values of A and increasing
values of the absolute k indicate increasing affinity for color-
producing substances.
Carbon isotherms were measured with Hydrodarco 4,000, Filtrasorb
400, and Westvaco Aqua Nuchar 8 x 30 carbons. Although nine
different carbons were tested, these carbons showed the best per-
formance for COD removal in initial isotherm screening tests,
color removal, and combined COD and color removal characteristics.
A summary of the data obtained from these isotherms is presented
in Table 10. Calgon Filtrasorb 400 and Westvaco 8 x 30 showed the
highest affinity for COD removal, while the Hydrodarco and Westvaco
carbons showed the greatest affinity for color removal.
TABLE 10
CARBON ISOTHERM RESULTS
USING ACTIVATED SLUDGE EFFLUENT
Carbon
Hydrodarco 4000
Filtrasorb 400
Westvaco aqua
nuchar 8 x 30
COD isotherm
extrapolated
capacity, X/M
(g COD/
g carbon)
0.6
1.3
4.8
1
n
0.86
0.50
1.80
Color isotherm
k
APHA units
(mg/l)
0.650
0.264
0.585
A
(APHA units)
67
220
120
Additional batch tests were made to determine the effect of pH on
adsorption performance. Tests were run as described previously
except that the pH of the wastewater sample was adjusted prior to
contacting with the carbon. Following the 24-hr equilibration
period, the sample was filtered and the pH adjusted back to 7.0
prior to analysis for residual COD and color. The results of
these tests are presented in Figure 41. Because it was necessary
100
-------
200O
1500
1000
500
1
5 o
< 2000
Q.
1500
1000
500
0
2000
1500
1000
500
. HLTRASORB 400
HYDRODARCO 4
A COD (Blank 310 mg/l)
• Color (Bionk 600 mg/l)
WESTVACO NUCHAR 8 x 30
\
COD (Blank 690 mg/l -
Color (Blank ITOOmg/l.
kCOD (Blank 690 mg/l).
»Color (Blank 1750 mg/ll
800
600
400
200
0
800
600
400 §
200
0
800
H600
400
Q
O)
LU
300
0
6 8
pH
10
12
14
R6.4I .INFLUENCE OF pH ON CARBON
ADSORPTION CAPACITY
101
-------
to run these tests on different days, the influent COD and color
concentrations were different for the tests with Filtrasorb 400
than with the other two carbons. Some precipitation occurred in
tests with Hydrodarco 4000 and Westvaco Aqua Nuchar 8 x 30 at
pH values greater than 10.0. The data shown in Figure 41 indi-
cate that this resulted in additional COD removal for the sample
adjusted to pH 12.0. These data show that optimum conditions for
the removal of both COD and color occur in the region of pH 4
to 7.
COLUMN INVESTIGATIONS
Tests Using Activated Sludge Effluent
Activated carbon column tests using activated sludge effluent
were conducted in 1-in. (2.54 cm) glass columns. Activated
sludge effluent from Unit IE and IID operated at an organic
loading of 0.4 g BOD applied/g MLVSS-day was used as the feed
for these tests. Four columns were operated in series at a
flow rate of 2.5 gpm/sq ft (102 liters/min-sq m). The system
was charged with 170 g of carbon at the beginning of each run.
Carbon was initially contacted with distilled water for 24
hr to expel air from the pores, then charged to each column
and backwashed with tap water until a clear effluent was ob-
tained. High suspended solids concentrations made it necessary
to backwash these columns daily. Prior to backwashing, samples
were taken from the influent and from the effluent of each of
the four columns. Waste was applied to the columns upflow.
Hydrodarco 4000, Filtrasorb 400 and Westvaco Aqua Nuchar 8 x 30
were all tested in the column investigations.
Normalized breakthrough data for these runs on a TSC basis are
shown in Figures 42 - 44. Immediate TSC breakthrough for the
Hydrodarco 400 run was approximately 15 to 20 percent, while
values for the Westvaco Aqua Nuchar and Filtrasorb 400 runs
ranged from approximately 30 to 40 percent. The high immediate
breakthrough for the Westvaco run might be due to high influent
total carbon values (approximately 230 mg/1) during the first
of this run, while average influent TSC values for the other
two runs were approximately 120 mg/1. While the immediate
breakthrough for the Filtrasorb 400 run was high, the break-
through did not increase above approximately 40 percent on an
average basis for the first 800 BV (bed volumes) throughput.
However, as will be discussed in the next section, several
cycles of regeneration in actual use might reduce this initial
advantage.
102
-------
2 468 10 12 14 16 18 20 22 24 26 28 30
^^^^^^^ TIME (Coys)
0 200 400 60O BOO OOO
THROUGHPUT(8V)
FIG42.BREAKTHROUGH CURVE - HYDRODARCO 4000 CARBON
«0.04)
2 4 6 8 10 12 14 16 8 20 22 24 26 28 30
TIME (Days)
' '
_ _ _ -
O 200 400 6OO 6OO 1000
THROUGHPUT (BV)
FIG, 43. BREAKTHROUGH CURVE - WESTVACO 8 x 30 CARBON
20 X
TiME (flays)
*-
6OO 800 1000 I2OO 1400 1600
THROUGHPUT (BV)
FIG. 44. BREAKTHROUGH CURVE - FILTRASORB 400 CARBON
103
-------
Normalized color breakthrough curves for these same tests are shown
in Figures 45 and 46. Although no entirely acceptable routine
method exists for measuring color in industrial wastes, measure-
ments were made using APHA units since the waste color closely
matched these standards. Color breakthrough relationships show
that the best color removal was obtained with H.ydrodarco 4000
with progressively less capacity for Westvaco Aqua Nuchar and
Filtrasorb 400, respectively. However, it should be pointed out
that while color breakthrough for the Filtrasorb 400 carbon rose
above 30 percent after approximately 75 BV throughput, the color
breakthrough did not rise above 50 percent for any significant
period during the remainder of the test run.
In order to determine any significant loss of capacity on regenera-
tion, the spent Hydrodarco 4000 was returned to the manufacturer
for regeneration. Subsequently, another test run was made in order
to compare the performance of the carbon before and after one re-
generation cycle. The TSC breakthrough curve for the regenerated
carbon is shown in Figure 47. For this test, the adsorption ki-
netius were very favorable as evidenced by the sharp breakthrough
approximately nine days after the run was started.
A comparison of the results is presented in Table 11. These re-
•sults show that Hydrodarco 4000 exhibited the best removal charac-
teristics for both COD and color based on a 30 percent breakthrough
of the influent concentration. However, the total COD exhaustion
capacity to approximately 90 percent breakthrough was greater for
the Westvaco and Filtrasorb carbon. Reference to Figure 44 shows
that while Filtrasorb 400 reached a TSC breakthrough of approxi-
mately 40 percent very quickly, the breakthrough did not increase
above 40 percent for an extended period of time. While the 30
percent breakthrough capacity for COD was almost identical for the
virgin and regenerated Hydrodarco 4000 carbon, the 90 percent COD
breakthrough was significantly greater for the regenerated carbon.
Because these tests were run at different times, changes in waste-
water composition are probably responsible for these differences.
However, it is possible that enlargement of small pores in the
carbon upon regeneration resulted in an increase in the COD capa-
city. This might be expected in this case, since organic constitu-
ents contained in the activated sludge effluent were quite proba-
bly very large molecules. While the 30 percent breakthrough
capacity for color was significantly less for the regenerated
carbon, this might be due to the particular influent color varia-
tion at the time of the two tests. The exhaustion color break-
through capacities for the two tests were more similar in value.
104
-------
100
Influent Color = 400-1200
APHA Units
Carton = 4 columns,
170 a/column
10 12
14 16 8 20 22
TIME (days)
24 26 28 30
0
200
800
1000
400 600
THROUGHPUT (BV)
FIG.45.COLOR REMOVAL BREAKTHROUGH CURVES FOR HYDRO-
DARCO 4000 AND WESTVACO 8 X30
(cumAnin-sqmaqpm/8qft xO.04)
100
80
60
40
20
Influent Color =200-800 APHA Units
Surface Loading = 25 gpm/sqft
Carbon = l24 a/column, 4columns
r
-•-•-**
/ I
Jl
;, r
JL
11
8 12 16 20 24 28
TIME (days)
32 36 40 44
j—u
0 ' 200 ' 400 600 800 1000 1200 1400 1600 1800 2000 220O
THROUGHPUT (BV)
FIG.46.COLOR REMOVAL BREAKTHROUGH CURVE FOR FILTRA-
SORB 400
(eumAnln-sqm=qpm/sq ft xO.04)
105
-------
120
100
80
o° 60
40
20
Influent TSC= 75-235 mg/l
Surface Loading = 2.5 gpm/sq ft
Carbon = 124 g/Column, 4 Columns
0
10 15 20 25 30
TIME (days)
35
40
45
50
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600
THROUGHPUT (BV)
FIG.47.TSC BREAKTHROUGH FOR REGENERATED HYDRODARCO 4000
(cum/min-sqm = qpm/sqft xO.04)
-------
TABLE 11
COMPARISON OF RESULTS FOR
ACTIVATED CARBON ADSORPTION
OF ACTIVATED SLUDGE EFFLUENT
Parameter
30% Breakthrough capacity
(g COD/g carbon)
90% Breakthrough capacity
(g COD/g carbon)
30% Color breakthrough capacity
(APHA units x liters throughput]
(g carbon) (1000)
Exhaustion color
breakthrough capacity
(APHA units x liters throughput'
(g carbon) (1000)
Color breakthrough at end of
run (%)
Hydrodarco 4,000
• Virgin
0.27
0.44
1.12
1.39
50
Regen-
erated
0.26
0.60
0.40
1.25
75
Westvaco
8 x 30
Oa
0.60
0.34
1.58
100
Filtra-
sorb 400
0.01
0.85
0.13
0.95
100
Immediate breakthrough to greater than 30 percent.
107
-------
Column Tests Using Trickling Filter Effluent
Additional column tests were conducted in 4-in. (10 cm) carbon col-
umns to examine treatment of the trickling filter effluent using
activated carbon. The results of these tests are shown in Figure 48
While a longer contact time might have resulted in additional
COD removal, these results might be expected since the trickling
filter effluent contained considerable concentrations of unde-
graded low molecular weight solvents which are not readily sorbable
on activated carbon.
Greater degrees of color removal were achieved in these tests
showing the greater sorbability of color molecules compared to
other organic substances in the waste. Based on color analyses
using the APHA standards, the Filtrasorb and Hydrodarco 3000
carbons averaged approximately 50 percent color removal through-
out most of this run. Color removal through the Hydrodarco 3000
was more erratic, though. These tests indicate that further re-
ductions of gross organics would be necessary in the trickling
filter system before this combination of treatment processes
would accomplish much COD removal. However, if color removal
were a major treatment objective, such an approach would be more
attractive.
108
-------
~ 1-0
o
o
o
0.8
C 0.6
cr
3 °4
O
I
0.2
o
o
o
,8
0.8
0.6
8 04
0.2
0
0
Surface Loadings 6 gpm/sq ft
Carbon Depth= 3 ft in 4,4 in columns
2000 4000 6000 8000
. . THROUGHfftJT (liters)
10,000 12,000
0
100
300
400
200
THROUGHPUT (Bv)
RG.48.COD BREAKTHROUGH FOR FILTRASORB 400 USING TRICKLING
FILTER EFFLUENT
(cu m/min- sq m= qpm /sq ft x 0.04)
-------
SECTION XI
REMOVAL OF ORGANIC CONSTITUENTS BY AIR STRIPPING
Due to the large proportion of organic solvents in this wastestream,
the removal of organic substances by air stripping was investigated
to determine both the incidental removal by this mechanism in bio-
logical treatment processes and removals which could be obtained
using air stripping as a unit treatment process. Preliminary labo-
ratory experiments in which raw waste was stripped using diffused
air aeration showed that approximately 75 percent removal of COD
could be achieved in three days and approximately 92 percent could
be removed in ten days. Because no biological inhibitory agent was
added to the waste during these tests, it is probable that some re-
moval of organic constituents was achieved by biological mechanisms.
However, it seemed unlikely that the major part of the organics
could be removed biologically during this period in the unseeded
raw waste. Based on these results, additional tests were under-
taken both in the laboratory and on a pilot-scale.
BATCH LABORATORY-SCALE TESTS
Initially, batch stripping tests were performed in the laboratory to
determine the degree of organics removal which could be achieved by
diffused air stripping and to determine the effect of temperature
on these removals. Tests were performed in 2-liter graduated cyl-
inders initially containing 1,000 ml of raw equalized waste. The
laboratory compressed air was passed through an oil trap and dif-
fused through the sample using a porous stone. Air flows were
measured using a rotometer. In order to measure the effect of
temperature on organics removal, the 2-liter cylinder was partially
immersed in a constant-temperature bath and the temperature adjusted
using ice and hot water.
The results of batch experiments run at temperatures of 5, 17, and
38°C are shown in Figure 49. While first-order removal of organic
constituents would be anticipated for individual organic constit-
uents, the removal of organic carbon in a multi-component system
would be the summation of removals for individual organic constit-
uents occurring at different rates. The semi-logarithmic plot of
the data in Figure 49 corresponds to these mechanisms, although it
does not confirm the hypothesis. Although all tests were not run
for an extended period of time, it appears from available data that
110
-------
101
0
Initial Volume = 1000 ml
Air Flow = 1.2 scfrn
10
20
2468
TIME (hr)
FIG.49.TSC REMOVAL IN BATCH STRIPPING TESTS
(std. cu m/min = s cfm x 0.028)
-------
the residual organic concentration at all temperatures tended toward
a common value indicating the presence of a non-strippable portion
of organics in the waste. The greatest removal of total carbon was
achieved in the 17°C test which was run over a 20-hr period. The
total carbon removal in this case was approximately 80 percent.
The determination of BOD in this testing indicated that 96 percent
of the BOD was stripped from the waste in 20 hr. The residual BOD
concentration was 130 mg/1.
The effect of temperature on the rate of stripping was correlated
to the relationship:
where ks is the stripping waste constant in hr and T and T1
are temperatures in °C. The rate constant, ks, was calculated as
follows:
, % TSC remaining at t-,
Ks L
where t] is the time from the beginning of the test in hours. The
data correlated according to this relationship is shown in Figure 50.
These data show that the temperature coefficient, 9, decreased
with time from a value of 1.043 for the first hour of these tests
to 1.023 for a 4-hr period. The value during the first hour of
1.043 was almost identical to the value of 1.044 derived from data
obtained from stripping tests performed with 1-component systems
of acetone, alcohol, and methyl ethyl keytone found in the litera-
ture. The decrease of the temperature coefficient with time ob-
served in this investigation would be expected for data obtained
from batch experiments in which the residual total carbon concen-
tration after an extended aeration time would be approximately the
same for a range of test temperatures.
CONTINUOUS-FLOW STRIPPING INVESTIGATIONS
Packed Tower Stripping
Stripping was investigated in a packed tower by adding a 2,500 cfm
(71 cu m/min ) blower to the first pilot trickling filter. Tests
112
-------
10
20
60
70
30 40 50
TEMPERATURE (°C)
FIG.50.VARIATION OF STRIPPING REMOVAL COEFFICIENT WITH
TEMPERATURE
-------
were conducted at surface loadings varying from 1 - 10 gpm/sq ft
(41 - 410 1/min-sq m) and air flows ranging from 300 - 2,500 cfm
(8.5 - 71 cu m/min). Results obtained from this investigation on
a BOD basis are shown as a function of the gas-liquid ratio in
Figure 51. The highest BOD removal observed in these tests was
approximately 60 percent and occurred most consistently at a gas-
liquid ratio of 500 cu ft/gal (3.75 cu m/liter). Assuming that
the stripping action in a continuous-flow system may be approxi-
mated by a first-order relationship, the data from these investi-
gations is shown in linearized form in Figures 52 and 53. The
mathematical relationships defining these data may be used to
estimate BOD and TSC removals for full-scale systems having a
depth of 20 ft (6.1 m).
Stripping in an Aerated Basin
Additional air stripping investigations were conducted in a 24-ft
(7.3 m) diameter plastic swimming pool containing a 1.5 hp surface
aerator. The volume of the basin was approximately 13,500 gal
(51,000 liters) and the applied power level corresponded to approxi-
mately 150 hp/mil gal (29.5 kw/mi1 liter). Tests were conducted at
detention times of 1.5 and 3 days. Results of these tests for the
1.5-day detention time on a chronological basis are shown in Figure
54. Total carbon removals of this period averaged approximately
40 percent. As an indication of biological activity which occurred
in this basin, oxygen uptake tests for the influent and effluent of
the basin were made periodically. The results of these tests shown
in Figure 54 indicate significant biological activity in the waste.
From these data it was estimated that approximately 30 percent of
the BOD removed in this basin was due to biological mechanisms.
A summary of results is shown in Table 12.
Chronological data for a 3-day detention time are shown in Figure 55.
Because of the relatively high oxygen uptake rates observed in the
basin, an additional test was run using the 3-day detention time in
which 20 mg/1 Cu was added to the basin influent in order to inhibit
biological growth.
Only one oxygen uptake measurement was made during the period
June 15 - 30. This measurement, made near the end of this run on
June 28, was 52 mg/l-hr. Oxygen uptake measurements made during
theperiodin which copper was added to the basin were 1.0, 6.0,
114
-------
UJ
cr
o
CD
100
80
60
40
20
0
Surface Loading
V 10 gpm/sq ft
• 5
A 2.5
100
200 300 400 500
GAS- LIQUID RATIO, G/L (cu ft/gal)
600
700
FIG.5I .PACKED TOWER AIR STRIPPING RESULTS, BOD BASIS
(cu m/min - sq m = gpm/sq ft x O.O4; cu m/cu m = cu ft/gal x 748)
-------
h-1
I—1
ON
0.5
0.4
o
CO
co 0.3
o o>
CO (CO
CD 0.2
o
O.I
0
0
= 0.894S0IO-(a00056f>
100
200 300 400
GAS-LIQUID RATIO
G. /cu ft N
1 L \ gal I
500 600 700
FIG.52. LINEARIZED DATA FOR PACKED TOWER STRIPPING
BASIS
(cu m/cu m = cu ft/gal x 748)
-TSC
-------
Q
O
CD
0.7
0.6
0.5
0.4
§ 0-3
0.2
0.1
0
• Total Inf. - Total Eff.
A Tota Inf. - Sol. Eff.
Se=Q833 SQIO-(000064
0 100 200 300 400 500 600 700
GAS-LIQUID RATIO, ~- ( c" |f
FIG.53.LINEARIZED DATA FOR PACKED TOWER
STRIPPING-BOD BASIS
(cu m/cu m = cu ft/gal x 7.48)
117
-------
I
LL)
QC
O
2000
0>
O
1000
LJ
CVl
O
July 2-26, 1973
D.T = 1.5 days
10
TIME (days)
15
FIG.54.TSC REMOVAL IN AERATED EQUALIZATION
BASIN -D.T. = 1.5 DAYS
113
-------
I-1
vo
TABLE 12
SUMMARY FOR STRIPPING IN AERATED BASIN
Detention
time (days
3.1
3.0C
1 .5
Temp
Air Basin
23
25
26
23
24
25
Flow
(gpm)<
3
3
12
TSC(mq/l)
Inf. Eff.
1579
1264
1327
548
522
804
TSC | BOD (mg/1)
Removed1? Total Soluble
(%)
65.3
58.7
39.4
Inf.
6000d
3565
4467
Eff.
2110d
617
2009
Inf.
4758
3685
4097
Eff.
1453
1435
1680
BOD
removed"
(X)
59.8
82.7
55.0
COD (mg/1)
Total Soluble
Inf.
6361
6240C
5630
Fff.
1990
2029
3128
Inf.
6100
--
4853
Eff.
1400
1740
2473
COD
removed'3
(SO
68.7
53.2
44.4
SS (mg/1)
Inf.
211
228
168
Eff.
409
344
321
dFlow in cu/min = gpm 3.79 x 10
Percent removals based on total influent and effluent values.
GDuring this period, 20 mg/1 Cu was added to basin influent.
Based on one value.
-------
o
60
40-
20
0
1800
|>!200
o
H 600h
0
June 15-30, 1973
D.T = 3 day*
July 21-28,1973
D.T. = 3 days
20 mg/l Cu added
H 1-
1 1 1 h
10
15
TIME (days)
FIG.55TSC REMOVAL IN AERATED EQUALIZATION
BASIN-D.T. = 3 DAYS
120
-------
and 24 mg/l-hr for this 7-day test period. These measurements in-
dicated that biological activity was significantly lower during
this test period, although the value of 24 mg/l-hr indicates that
complete inhibition of microorganisms was not achieved. Corres-
ponding total carbon removal during these two tests were 59 and
65 percent, respectively, for tests with and without the copper
addition indicating that only slightly reduced removals resulted
during the tests in which copper was added to the system. This-
might be due to a complete inhibition of the microorganisms or
to competition between the stripping and biological removal mecha-
nisms. In the latter case, a greater removal of organic constit-
uents due to air stripping might be realized under conditions where
there is no removal by biological mechanisms.
Removals of BOD, COD, and TSC observed in these investigations is
shown in Figure 56. The BOD removal for the 3-day detention time
was significantly less than that which would be predicted from
corresponding removals of COD and TSC. The expected removals, of
BOD have also been indicated in Figure 56. Because only one
valid BOD determination was obtained from this test, it is felt
that BOD removals of approximately 80 percent would be observed
on an average basis under these conditions.
Removal of Specific Organic Constituents
Periodically during the continuous-flow stripping investigations,
samples were taken and analyzed for solvents. All samples taken
for solvent analysis were made on a grab basis. A summary of
these results appears in Table 13. These data show very signi-
ficant removals of toluene and isopropanol in both the aerated
basin and the packed tower. Toluene was completely removed in
the stripping processes for each observation.
Acetone concentrations increased in the aerated basin except for
the sample taken on July 26 during the test in which copper was
added to the basin influent. Decreases in acetone concentration
were observed in the packed tower on all occasions except for the
the sample taken on July 9. Data taken for the packed stripping
tower followed by a biological trickling filter on July 26 showed
that acetone decreased through the stripping tower but increased
to its original level in the trickling filter. These data tend
to corroborate an observation made in the trickling filter system
concerning increases in acetone concentration. Acetone concen-
trations decreased in the packed tower and in the aerated basin
121
-------
100
N>
N5
I 2 3
DETENTION TIME (days)
FIG. 56 .ORGANIC REMOVALS IN THE AERATED BASIN
-------
TABLE 13
SOLVENTS MEASURED IN STRIPPING INVESTIGATIONS
Process
Aerated basin
Packed tower
Trickling filter
Following packed
tower
Date
(July 1973)
9
12
23
26
9
12
23
26
26
Influent (mg/1)
Acetone
300
200
200
400
300
200
200
400
300
Methanol
100
-
Tr
400
100
-
Tr
400
200
Isoproponal
1900
1500
1300
1600
1800
1500
1300
1600
800
Toluene
100
300
Tr
200
100
Tr
200
-
Effluent (mg/1)
Acetone
700
500
400
200
700
100
100
300
400
Methanol
Tr6
-
200
-
100
-
-
200
200
Isoproponal
100
300
100
200
200
700
400
800
600
Toluene
-
-
-
_
-
-
-
r-o
uo
Tr = Trace.
-------
tests in which biological growths were inhibited. On the other
hand, acetone concentrations increased in the aerated basin before
copper was added and in the trickling filter following the
packed stripping tower. This indicated that increases in ace-
tone concentrations are due to by-products of biological degra-
dation and not to autooxidation of isopropanol in the presence
of air.
124
-------
SECTION XII
ACKNOWLEDGEMENTS
This investigation was conducted as a joint effort between CIBA-
GEIGY Corporation and Associated Water and Air Resources Engineers,
Inc. (AWARE). The project was sponsored in part by Grant No. 12020
FOB from the U. S. Environmental Protection Agency and also by
CIBA-GEIGY Corporation. Mr. Robert F. Curran, Director of Environ-
mental Technology for CIBA-GEIGY Corporation served as Project
Director. Technical aspects of the project were directed by Dr.
John H. Koon with the consultation of Dr. Carl E. Adams, Jr. and
Professor W. Wesley Eckenfelder of Associated Water and.Air
Resources Engineers, Inc. Technical assistance was also provided
by Dr. Jan A. Oleszkiewicz. Mr. David H. Stonefield served as
Project Officer for the Environmental Protection Agency. Techni-
cal evaluation from CIBA-GEIGY Corporation was provided by manage-
ment and technical personnel of the Cranston Plant.
125
-------
SECTION XIII
REFERENCES
1. Chipperfield, P. N. J. Performance of Plastic Filter Media in
Industrial and Domestic Waste Treatment. J. Water Pollution
Control Fed. 39^ (11): 1860^1874, 1967.
2. Richard, J. G. "Design Considerations for Plastic Media Biologi-
cal Towers," presented at the W. Va. Water Pollution Control
Assoc 1973 Meeting, April 26, 1973.
3. Chipperfield, P. N. J., and M. Ashew. Multiple Stage Plastic
Media Proving Effective in Treating Biodegradable Industrial
Wastes. Ind. Water Eng, August 1970.
4. Allen, T. S. Water Reuse in the Food Processing Industry. Paper
72^ - PIDrll, ASME, March 1972.
5. Chipperfield, P. N., M. W. Ashew, and J. H. Benton. Multiple
Stage Plastic Media Treatment Plants. J. Water Pollution Con-
trol Fed. 44 (10):1955-1967, 1972.
6. Oleszkiewicz, J. A., and W. W. Eckenfelder, Jr. Mechanism of
Substrate Removal in High Rate Plastic Media Trickling Filter.
Vanderbilt Univ., Tech. Report, No. 33, 1974.
7. Bryan, E. H. and D. H. Moeller. Aerobic Biological Oxidation
Using Dowpack. Advances in Biological Waste Treatment. New
York. Pergamon Press, 1963.
8. Gulp, G. L. Direct Recirculation of Trickling Filter Effluent,
J. Water Pollution Control Fed. 35 (6):742-746, 1963.
9- Weston, R. F. Fundamentals of Aerobic Biological Treatment of
Wastewaters. Public Works 1^:77-83, 1963.
10. Velz, C. J. A Basic Law for the Performance of Biological Fil-
ters, Sew Wks J. 20_:607-617, 1948.
11. Sorrels, J. H., and P. J. Zeller. Heavy Loading on Trickling
Filters. J. Water Pollution Control Fed. 3J5 (9) :1184-1194,
1963.
126
-------
13. Ganczarczyk, J. J. Grabowska. Limiting Loadings of Biological
Filters (Polish). Z. N. Politechn Slaskiej, Inz. Sanit. 2 1:
20, 1960.
14. Rincke, G. Technology of Plastic Medium Trickling Filters.
Proc. 5th Int'l Conf. on Hater Pollution Research. San Francisco,
TOT:—
15. Germain, J. E. Economical Treatment of Domestic Waste Using
Plastic Media Trickling Filters. J. Water Pollution Control
Fed. 38^ (2):192, 1966.
16. Eckenfelder, W. W. Trickling Filter Design and Performance. J.
Sanitary Eng Div ASCE. 87^ (SA4):33-45, 1961.
17. Eckenfelder, W. W., Jr. Water Quality Engineering for Practicing
Engineers. New York, Barnes and Noble, 1970.
18. Thackston, E. L., and W. W. Eckenfelder, Jr. Process Design in
Water Quality Engineering New Concepts and Development?, New York
Jenkins Publishing Co., 1972.
19. Eckenfelder, W. W., Jr., and Carl E. Adams, Jr. Process Design
Techniques for Industrial Waste Treatment, Nashville, Tennessee,
Enviro-Press, Inc., 1974.
20. Pearson, E. A. Kinetics of Biological Treatment. Advances in
Water Quality Improvement, Vol. ]_, ed. by E. F. Gloyna and W. W.
Eckenfelder, Jr., Austin, University of Texas Press, 1968.
21. Benedict, A. H. and D. A. Carlson. The Real Nature of the
Streeter - Phelps Temperature Coefficient. Water and Sewage
Works. 117 (2):54-57, February 1970.
127
-------
SECTION XIV
LIST OF SYMBOLS AND ABBREVIATIONS
a Sludge yield coefficient (g VSS produced/g substrate
removed)
a' Assimilative oxygen utilization coefficient (g Op/g
substrate removed)
b Endogenous respiration rate (g VSS destroyed/day-g MLVSS)
b' Endogenous oxygen utilization coefficient (g 09/g MLVSS-
day) ^
BV Bed volume; equal to the gross volume occupied by a
material
C Concentration (mg/1)
C Influent concentration (mg/1)
D Filter depth (ft)
D.O. Dissolved oxygen
F/M Organic loading, food to microorganism ratio (g substrate
applied/g MLVSS-day)
G Gas (air) flow (cfm) (cu m/min)
k Activated sludge substrate removal rate constant (1/mg-day)
K Trickling filter substrate removal rate constant
([gpm/sq ft] /ft) (cu m/min-sq m) /m)
L Liquid (waste) flow (gpm) (cu m/min)
MLSS Mixed liquor suspended solids
MLVSS Mixed liquor volatile suspended solids
n Flow exponent
128
-------
N Recycle ratio, N = R/Q
Q Flowrate (liters/day or mgd) (cu m/day)
Q» Applied hydraulic loading (gpm/sq ft) (cu m/min-sq m)
R Recycle flow (liters/day or mgd) (cu m/day)
Rr Oxygen utilization rate (g CL/day-1)
SS Suspended solids
S Effluent substrate concentration (mg/1)
S Influent substrate concentration (mg/1)
Sr Substrate removed, S = S -S (mg/1)
t Aeration time (days)
T,T' Temperature (°C)
IDS Total dissolved solids (mg/1)
TSC Total soluble carbon (mg/1)
V Volume (liters or mil gal)
VSS Volatile suspended solids
X Mixed liquor volatile suspended solids (mg/1)
AX Excess volatile suspended solids (Ib/day) (g/day)
129
-------
SECTION XV
APPENDICES
Page
A. Treatment Facility Costs 131
B. Trickling Filter Data Summary 133
C. Activated Sludge Data 135
130
-------
APPENDIX A
TREATMENT FACILITY COSTS
TABLE Al
TREATMENT FACILITY CAPITAL COST SUMMARY3
Item
Cost
Raw waste sump pumps, motors, and bar'screen
Solvent separation tank (50,000 gal, steel)
Neutralization (vessel, acid and caustic metering
pumps, acid and caustic storage tanks)
Nutrient feed (phosphoric acid and anhydrous
ammonia tank)
Equalization tanks, two (100 ft dia, 20 ft
SWD, steel)
Pumping through preliminary treatment (pumps
and motor)
Trickling filters, six (including packing)
Trickling filter fans and motors, s.ix
Trickling filter pumps and motors, six
Clarifier Shell (50 ft dia, 10 ft SWD, steel)
Clarifier Rake
Sludge Pump
Instruments (including control panel)
Control building (three offices, conference room,
laboratory, maintenance area, motor control
room, equipment and chemical storage room)
Site work (including excavation, soil testing,
and piping)
Engineering
Total
$ 48,800
42,900
17,400
4,900
416,000
26,400
540,000
6,600
9,600
52,000
39,000
1,000
32,900
165,000 '
170,000
277,500
$1,850,000
aCosts are at 1970 levels.
131
-------
TABLE A2
TREATMENT FACILITY OPERATING COST SUMMARY"
Item
Supervisory and administrative labor
(including engineering support)
Operating labor
Electricity
Chemicals
Laboratory and analytical costs
Maintenance (labor and materials)
Yardwork and custodial costs
Water
Total
Cost
$ 26,000
67,000
17,000
83,000
26,000
83,000
24,000
1,000
$327,000/yr
Costs are for calendar 1973.
132
-------
APPENDIX B
TRICKLING FILTER DATA SUMMARY
TABLE B1
TRICKLING FILTER WEEKLY DATA SUMMARY
Date
ii/i
1 1 / 1
11/12
11/19
11/26
12/3
12/10
12/17
12/24
12/31
1/7
1/14
1/21
1/28
2/4
2/11
2/18
2/25
3/4
3/11
3/18
3/25
4/1
4/8
4/15
4/22
4/29
5/6
5/13
5/20
5/27
6/3
6/10
6/17
6/24
7/1
7/8
7/15
7/22
BOD I nig/ 1)
Inf
Total
„
—
3300
4313
4097
5985
4545
4820
3620
3460
2570
2470
2997
4337
3080
3558
2672
4052
2273
3035
1987
1258
1793
1164
1870
1315
1750
1747
1995
2398
2598
4750
5400
5227
3963
4037
4060
Eff
Total
,.
—
—
2848
3177
5060
3940
3280
2630
3027
1797
2335
2480
2893
2280
2625
1850
2347
1281
1447
1179
629
888
464
1323
611
1031
1223
1810
1753
2292
3700
—
3153
2140
2563
2170
Sol
„
-- •
--
2737
3253
4200
3365
2770
1750
2297
1607
2155
2225
2620
1950
2805
1707
2450
1118
1141
1056
414
582
355
1150
569
1186
1160
1510
1823
2263
3800
5600
2930
2030
1760
2500
''-
REM
__
—
—
34.0
--
15.4
13.3
32.0
27.3
12.5
30.1
5.5
17.2
33.3
26.0
26.2
30.8
42.1
43.6
52.3
40.7
50.0
50.5
50.1
29.2
53.5
41.1
30.0
9.3
26.9
11.8
22.1
39.7
46.0
36.5
46.6
COD (mq/1)
Inf
Total
A nun
HUJU
5048
3930
3737
6030
Eff
Total Sol
-.
2680
2953
4237
6100 4603
5773 4730
4957 3175
5970
6753
5233
4390
4453
4707
5443
4680
5107
4440
5773
3507
4080
3813
2647
3347
2200
2600
2960
2947
2960
2880
3253
3371
6253
8240
7147
6267
6160
5760
4300
5480
4000
3347
3637
3120
3910
3540
3947
2580
3503
2387
2547
2360
1485
1747
1187
1928
2053
2320
2507
2320
2680
3133
4067
5600
4707
3573
3893
3353
once
cyoo
3322
1947
2737
3538
4153
3493
2750
3805
4587
3653
2930
3240
2727
3470
2980
3840
1825
3147
2126
2267
2200
1127
1360
787
1784
1720
2107
2403
2160
2367
2963
3813
5160
4320
3367
3520
3107
%
REM
__
31.8
21.0
29.7
24.5
18.1
35.9
28.0
18.8
23.6
23.8
18.3
35.7
28.2
24.4
22.7
41.9
39.3
31.9
37.6
38.1
43.9
47.8
46.0
25.8
30.6
21.3
15.3
19.4
17.6
7.1
35.0
32.0
34.1
43.0
36.8
4KB
TSC(mg/1)
Inf
Total
-«.
1050
1151
1716
1763
1732
1016
1584
1732
1543
1180
976
1122
1348
2950
1495
1337
1614
1099
1037
888
807
892
727
748
940
842
857
659
786
800
1312
1865
1644
1365
1285
1253
tff
Sol
__
570
934
1291
1513
1390
680
1233
1495
1155
898
916
912
1018
2750
1338
847
1013
744
668
629
500
547
296
545
695
691
729
660
689
785
841
1408
1161
1007
991
912
% .
REM
45.7
18.9
24.8
14.2
19.2
33.1
22.2
13.7
25.1
23.9
6.1
18.7
24.5
6.8
10.5
36.6
37.2.
32.3
35.6
29.2
38.0
38.7
59.3
27.1 ,
26.1
17.9
14.9
0
12.3
1.9
35.9
24.5
29.4
26.2
22.9
27.2
Temperature
:°F}*
Inf
66
63
63
63
63
60
61
58
56
59
60
59
61
57
59
59
61
62
61
63
63
64
67
67
68
69
70
70
71
75
78
81
80
73
75
72
74
Eff
52
61
58
58
58
58
57
56
50
57
57
56
55
52
54
53
59
62
57
63
60
59
65
66
63
63
62
63
69
72
77
73
75
82
81
82
81
Air
46
40
41
36
33
33
34
37
14
40
38
26
31
30
35
30
45
49
36
41
44
39
58
57
54
54
54
55
67
70
77
70
70
78
79
77
78
Flow
(mgd)lJ
1 99
I . uC.
1.28
0.96
1.07
1.28
1.18
0.85
0.90
1.05
1.15
1.28
1.25
1.21
1.01
1.01
0.72
0.72
0.50
0.50
0.50
0.72
0.72
0.72
1.08
1.03
0.99
0.94
0.99
1.10
0.91
1.15
0.91
0.81
0.94
0.97
1.13
1.10
0.79
Loading (lb/
1000 cu ft-
dav)c
BOD
_„-
—
491
767
672
707
569
703
579
616
446
415
421
609
308
356
186
230
158
304
199
126
269
167
142
172
241
267
252
383
329
534
705
704
622
617
446
COD
CQA
OUH
898
524
556
734
1000
682
620
871
1079
931
763
749
661
764
468
511
309
401
244
409
382
265
502
315
358
387
405
453
364-
520
426
869
1145
993
871
856
801
Total
Solids
(mq/1)
Inf
OCl A
CO 1H
3114
2843
328
—
2540
2650
3680
2130
2125
3230
—
4540
2210
2970
2950
3012
2078
3279
2639
—
4860
2430
2390
2750
2520
2388
2032
2436
—
2144
1817
1497
3694
—
3590
3020
3420
Eff
?K7T
t3/ 0
2828
2943
694
_.
3000
2710
2360
2260
£113
S/828
—
2373
2490
2730
2750
2114
1884
3645
2431
2225
2560
2380
2780
2930
£140
2133
2652
2743
—
2244
2297
2013
3505
—
3650
2770
3670
SS
(mg/1)
Inf
i -an
1 JU
302
104
96
80
187
167
225
98
497
262
141
119
100
104
159
103
95
101
72
170
124
108
126
78
78
89
111
98
127
220
193
205
239
179
213
235
235
Eff
t\?
u£
37
104
71
99
97
126
69
83
272
143
129
120
159
113
105
91
84
205
134
96
137
126
187
232
129
105
132
89
99
152
155
121
122
156
143
164
145
VSS
(mg/ )
Inf
-,-
73
86
63
160
145
182
89
486
245
116
111
86
93
106
90
74
86
69
111
102
88
99
75
66
80
105
92
124
196
150
187
'160
135
194
200
188
Eff
cc
«JO
32
63
63
80
79
110
67
79
225
130
116
109
130
97
.79
75
71
181
124
86
106
104
163
209
110
94
121
86
97
142
131
107
89
137
131
143
121
OJ
temperature in °C = 5/9 (F-32).
bFlow in cu m/day = mgd x 4,000.
°Loading in kg/cu m-day = lb/1,000 cu ft-day x 0.016..
-------
TABLE B2
FULL-SCALE TRICKLING FILTER RECYCLE DATA
Date
(1974)
Apr. 1
Apr. 8
Apr. 15
Apr. 22
Apr. 29
May 6
May 13
Avg. Air Temp.
(°C)
9.4
6.7
11.7
11.1
13.3
10.6
17.8
Inf. BOD
(mg/1)
1,546
1,492
1,522
2,532
2,351
2,851
2,720
Bank Aa
Eff. BOD
(mg/1)
1,435
890
1,453
1,810
1,836
2,555
2,097
Removal
(*)
7.2
40.3
4.5
28.5
21.9
. 10.4
22.9
Bank Bb
Eff. BOD
(mg/1)
892
630
1,030
1,537
1,382
1,696
1,472
Removal
(X)
42.3
57.8
32.3
39.3
41.2
40.5
45.9
Operated at a recycle ratio of 0.43.
Operated at a recycle ratio of 2.33.
134
-------
APPENDIX C
ACTIVATED SLUDGE DATA
TABLE Cl
SUMMARY OF ACTIVATED SLUDGE RESULTS-PHASE I
JNIT
OPERATIONAL PARAMETERS
F/M (g BOD/day-g MLVSS)
Detention time (days)
MLVSS (mg/1)
Volume (1)
D.O. (mg/1)
Oxygen uptake (g 0£/day-g MLVSS)
SVI (ml/g)
ZSV (ft/hr) (m/m)
Temperature (°C)
Sludge age (days)
INFLUENT QUALITY PARAMETERS
BOD (mg/1)
Total
Soluble
COD (mg/1)
Total
Soluble
pH (median value)
EFFLUENT QUALITY PARAMETERS
BOD (mg/1)
Total
Soluble
COD (mg/1)
Total
Soluble
pH (median value)
SS (mg/1, median value)
ORGANIC REMOVAL EFFICIENCIES
BOD (%)
COD (55)
BOD removal rate
(g BODr/day-g MLVSS)
COD removal rate
(g CODr/day-g MLVSS)
IAa
0.22
5.3
2620
151
6.2
0.22
36
14.7
17
15.5
3070
2910
4760
3980
7.3
77
54
723
421
6.1
300
98.2
91.2
0.218
0.314
IBb
0.084
11.3
2410
151
6.7
0.11
48
14.4
17
32.9
2270
2200
3860
3500
7.3
15
14
416
336
7.2
100
99.4
91.3
0.083
0.130
icb
0.23
4.1
2440
151
5.9
0.23
39
24.1
17
15.3
2270
2200
3860
3500
7.3
53
47
603
384
6.6
214
97.3
90.0
0.221
0.347
ID15
0.40
2.1
2700
151
4.9
0.45
33
19.5
17
7.5
2270
2200
3860
3500
7.3
233
184
825
519
5.9
302
91.9
86.5
0.36
0.581
IEC
0.24
3.6
2600
151
5.8
0.26
40
13.6
17
13.5
2220
2160
3900
3510
7.1
39
39
547
379
6.3
182
98.2
90.3
0.232
0.376
?Unit fed raw equalized waste.
Unit fed unclarified trickling filter effluent.
cUnit fed clarified trickling filter effluent.
135
-------
TABLE C2
SUMMARY OF ACTIVATED SLUDGE RESULTS - PHASE II'
Influent
Unit
OPERATIONAL PARAMETERS
F/M (g BOD/day-g MLVSS)
Detention time (days)
MLVSS (mg/1)
Reactor volume (1 )
D. 0. (mg/1)
Oxygen uptake (g 02/g MLVSS-day)
Sludge age (days)
Temp (°C)
INFLUENT QUALITY PARAMETERS
BOD, Total (mg/1)
COD, Total (mg/1)
Soluble (mg/1)
pH (median value)
EFFLUENT QUALITY. PARAMETERS
BOD (mg/1)
Soluble
COD (mg/1)
Total
Soluble
pH (median value)
ORGANIC REMOVAL EFFICIENCIES
BOD (%)
COD (%}
BOD removal rate
(g BOD /g MLVSS-day)
COD removal rate
(g CODr/g MLVSS-day)
Unclarified Trickling Filter Effluent
IIC
0.23
2.18
3150
151
6.9
0.28
16.7
18
1559
2096
1610
6.9
47
365
261
7.0
97.8
87.5
0.22
0.27
IID
0.43
1.19
3020
151
6.4
0.45
11.4
18
1559
2096
1630
6.9
111
516
350
6.9
94.4
83.3
0.40
0.48
IIP
0.23
2.48
2760
151
6.7
0.31
22.7
14
1559
2172
-
7.0
44
354
268
7.0
96.8
87.7
0.22
0.28
Includes data collected from March 10 to May 26, 1973 except for the
cold temperature operating period from March 24 through April 13 excluded
from Unit II F results.
136
-------
TABLE C3
SUMMARY OF ACTIVATED SLUDGE RESULTS,
VARIED TEMPERATURE OPERATION - PHASE II
Influent
Unit
OPERATIONAL PARAMETERS
F/M (g BOD/day-g MLVSS)
Detention time {days)
MLVSS (mg/1)
Reactor volume (1 )
D. 0. (mg/1)
Oxygen uptake (g 02/g MLVSSOday)
Sludge age (days)
Temp (°C)
INFLUENT QUALITY PARAMETERS
BOD, Total (mg/1)
COD, Total (mg/1)
Soluble (mg/1)
pH (median value)
EFFLUENT QUALITY PARAMETERS
BOD (mg/1)
Soluble
COD (mg/1)
Total
Soluble
pH (median value)
ORGANIC REMOVAL EFFICIENCIES
BOD (%)
COD (%)
BOD removal rate
(g BODr/g MLVSS-day)
COD removal rate
(g CODr/g MLVSS-day)
Unclarif led
Tricklinci Filter Effluent
IIC
0.130
2.50
3260
151
6.7
0.28
32.3
17
1107
2546
2165
6.8
26
416
273
7.0
97.7
89.3
0.128
0.268
IID
0.240
1.50
3070
151
6.6
0.31
35.7
17
1107
2546
2161
6.8
49
666
375
7.0
95.6
85.3
0.230
0.471
IIP
0.142
4.44
1750
151
9.3
0.51
48.8
6
1107
2450
2027
6.8
71
406
301
6.8
93.6
87.7
0.133
0.277
'includes data collected from March 24 to April 13, 1973.
137
-------
TABLE C4
SUMMARY OF ACTIVATED SLUDGE RESULTS-PHASE IIIa
Feed*>
IAe.ratina Gas
UNIT
OPERATIONAL PARAMETERS
F/M (g BOD/day-g MLVSS)
Detention time (days)
MLVSS (mg/1)
Volume (1)
D.O. (mg/1)
Oxygen uptake (g 02/day-g MLVSS)
SVI (ml/g)
ZSV (m/hr)
Temperature (°C)
Sludge age (days)
INFLUENT QUALITY PARAMETERS
BOD (mg/1)
Total
Soluble
COD (mg/1)
Total
Soluble
pH (median value)
EFFLUENT QUALITY PARAMETERS
BOD (mg/1)
Total
Soluble
COD (mg/1)
Total
Soluble
pH (median value)
SS (ing/1, median value)
ORGANIC REMOVAL EFFICIENCIES
BOD (%)
COD (%)
BOD removal rate
(g BODr/day-g MLVSS)
COD removal rate
(g CODr/day-g MLVSS)
Unstripped
Air Oxvaen
III A
0.31
5.2
2540
151
4.1
0.35
55
17.1
23
12.4
•
4060
3720
5737
4972
8.2
134
99
836
432
7.8
240
97.6
92.5
0.30
0.40
III B
0.34
3.43
3500
151
8.8
0.46
37
15.3
25
14.1
4030
3390
5440
4806
8.3
521
366
989
546
7.1
280
89.5
90.0
0.31
0.41
Stripped
Ai r Oxvaen
III C
0.29
2.4
2470
151
4.1
0.25
61
14.1
23
17.1
1710
1440
2484
2016
8.3
82
72
373
320
7.9
130
96.3
87.1
0.28
0.37
III D
0.32
2.3
2210
151
8.6
0.37
47
18.0
25
28.0
1620
1430
2410
193S
8.3
51
47
32]
276
7.7
115
97.4
88.5
0.31
0.42
blncludes data from June 18 to July 28, 1973.
All units fed raw equalized waste with or without prestripping as indicated.
First week of data not included in averages.
138
-------
TECHNICAL REPORT DATA
(I'liuse read iHiilruciions on the reverse before completing)
, REPORT NO.
3. RECIPIENT'S ACCESSION-NO.
.ITLE AND SUBTITI. E
Evaluation and Upgrading of a Multi-Stage Trickling
Filter Facility
!3. REPORT DATE
December 1976 (Issuing date)
6. PERFORMING ORGANIZATION CODE
.AUTHOR(S)
Koon, John H.; Curran, Robert F.; Adams, Carl E. Jr.;
and Eckenfelder, W. Wesley, Jr.
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORG \NIZA I ION NAME AND ADDRESS
AWARE under contract to:
CIBA-GEIGY Corporation
Cranston, RI
12. SPONSORING AGENCY NAIWf AND ADDRESS
Industrial Environmental Research Laboratory - Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
10. PROGRAM ELEMENT NO.
1BB036; Roap 21AZO
11. iSfflCKiXdiJXr/GHANT NO.
12020 FOH
13. TLYPE Qf; REPORT AND PERIOD COVERED
inal
hi
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOT IS
16. ABSTRACT
The applicability of a full-scale, six-stage trickling filter plant was investigated
for the treatment of waste from a multiproduct organic chemical plant. Reductions
of BOD per unit volume of nacking indicated that the series design of the system
did not lead to stage-wise acclimation of the microorganisms or enhanced BOD removals.
Tests to determine BOD removal mechanisms in the system indicated that air stripping
and biological mechanisms both contributed significantly to the total observed re-
duction. Effluent recycle considerably improved filter performance. However, 600
percent recycle was required for an approximate 90 percent reduction of BOD at a
hydraulic loading of 2 gpm/sq ft (0.08 cu m/min-sq m). Bench-scale activated sludge
investigations showed that this process could be used successfully for upgrading
the trickling filter system. Kinetic parameters necessary for the design of activated
sludge systems were determined. Comparative studies of air and oxygen aeration in-
dicated that the treatment of wastes containing volatile substances might be diffi-
cult in a closed oxygen system. Activated carbon adsorption, also tested in small-
scale systems, was most effective as a means of upgrading the trickling filter
system when applied following activated sludge treatment. Activated carbon tests
indicated that some refractory organics and color-producing substances could be
effectively removed by adsorption; however, a significant nonadsorbably organic
fraction remained.
17.
a.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
*TnckTTn'g Filters, *BioTogicaT Treatment,
*Activated Sludge, *Chemical Wastes,
Chemical Treatment, Adsorption, Industrial
Wastes, Monitoring, Waste Water Treatment
13. DISTRIBUTION STATEMENT
Release to public
b. IDENTIFIERS/OPEN ENDED TERMS
*Air Stripping, *0rganic
Chemical Industry
19. SECURITY CLASS /Tills Rejxin/
Unclassified
"20. SECURITY CLACB (This pdge)
Unclassified
COSATl Field/Group
13/B
21. NO. Of PAGES
151
CPA Form 2220-1 (9-V3)
•fr U.S. GOVERNMENT PRINTING OFFICE: 1977-757-056/5525 Region No. 5-H
139
------- |