EPA-600/2-77-032
June 1977
Environmental Protection Technology Series
DESIGN PROCEDURES FOR
DISSOLVED OXYGEN CONTROL OF
ACTIVATED SLUDGE PROCESSES
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
EPA-600/2-77-032
June 1977
DESIGN PROCEDURES FOR DISSOLVED OXYGEN
CONTROL OF ACTIVATED SLUDGE PROCESSES
by
Michael J. Flanagan
Brian D. Bracken
Brown and Caldwell
Walnut Creek, California 94596
Contract 68-03-2130
Project Officer
Joseph F. Roesler
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal 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.
11
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.
Research and development is the necessary first step in problem solution
and it involves defining the problem, measuring its impact, and searching
for solutions. The Municipal Environmental Research Laboratory develops new
and improved technology and systems for the prevention, treatment, and
management of wastewater and solid and hazardous waste pollutant discharges
from municipal and community sources, for the preservation and treatment of
public drinking water supplies, and to minimize the adverse economic, social,
health, and aesthetic effects of pollution. This publication is one of the
products of that research; a most vital communications link between the
researcher and the user community.
The purpose of this manual is to provide a recommended design procedure
to guide the design engineer in the selection of aeration equipment and
control techniques for achieving optimal dissolved oxgen control of the
activated sludge process.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
111
-------
ABSTRACT
This report presents design procedures and guidelines for the selection
of aeration equipment and dissolved oxygen (DO) control systems for
activated sludge treatment plants. A review of process configurations and
design parameters is made to establish system requirements. Aeration
methods, equipment and application techniques are examined and selection
procedures offered. Various DO control systems are described with recom-
mendations for system applications to various aeration equipment types and
process configurations. Performance, operational and maintenance data for
aeration equipment and DO control systems for twelve activated sludge
plants is presented. This information and other design recommendations in
the report are used to develop automatic DO control systems for various size
hypothetical activated sludge system configurations for an economic analysis
of manual and automatic DO control. The conclusion is drawn that the
capital and operating costs of automatic DO control systems are justified for
activated sludge plants larger'than 1 mgd (44 dm3/s) provided equipment is
selected and applied in accordance with the guidelines of the design manual
and a power cost equal to or greater than the national average power rate is
applicable,
This report was submitted in fulfillment of Contract No. 68-03-2130 by
Brown and Caldwell under the sponsorship of the Environmental Protection
Agency. Work was completed as of October, 1976.
IV
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables xi
Abbreviations and Symbols xiii
Acknowledgment xvi
1. Introduction 1
2 . Conclusions 2
3. Recommendations for Future Research 4
4. Unit Processes 5
5. Aeration Systems 16
6. Design of Air and Pure Oxygen Dissolution Control Systems . . 62
7. Case History Summary of Dissolved Oxygen Control Systems . . 93
8. Cost of Automatic Dissolved Oxygen Control Systems 103
9. Dissolved Oxygen Control System Selection 131
References 137
Appendix
Case histories of dissolved oxygen control system performance,
operational and maintenance data 141
v
-------
FIGURES
Number Page
1 Conventional Activated Sludge Process 7
2 Oxygen Utilization in a Plug Flow System
q
3 Complete Mix Process
4 Oxygen Utilization in Activated Sludge Systems
5 Step Feed Process
i ^
6 Contact Stabilization Process ...............
7 Hatfield and Kraus Processes ........... ....
8 Relationship of Aeration Air Requirements for Oxidation of
Carbonaceous BOD and Nitrogen ............ 20
C
9 Diurnal Variations in Load at the Chapel Hill, N.
Treatment Plant 21
10 Relation between Ammonia Peaking and Hydraulic Peaking
Loads for Treatment Plants with No In-Process Equalization 23
11 Relationship of Maximum/Minimum Nitrogen Load Ratio to
Maximum/Average Flows 24
12 Characteristics of Orifice-Type Coarse Bubble Diffusers. ... 26
13 Characteristics of Valve-Type Coarse Bubble Diffusers .... 27
o p
14 Typical Shear-Type Coarse Bubble Diffuser
29
15 Shallow Submergence Aeration System.
31
16 Spiral and Cross Roll Diffuser Arrangements
VI
-------
FIGURES (cont'd)
Number Page
17 Aeration System - South Eastern Purification Plant
o o
Melbourne, Australia °°
18 Characteristics of Porous Plate Diffuser at the South Eastern
Purification Plant, Melbourne, Australia 34
19 Oxygen Transfer Efficiency of Diffusers at Various Flow
Rates and Submergence 35
20 Details of Walker Process Equipment Company Deep
Submergence Diffusion System 36
21 System Characteristic Curve 39
22 Blower/System Curves 39
23 Blower Characteristic Curve 40
24 Discharge Throttling Control 41
25 Blower Speed Control 42
26 Inlet Guidevane Control for Centrifugal Blower 43
27 Axial Blower Characteristic Curve 44
28 Typical Performance Characteristic Curve for an Axial
Blower with Variable Speed Control 45
29 Typical Performance Characteristic Curve for a Centrifugal
Blower with Variable Speed Control 46
30 Pressure Oriented Surge Control System for Suction
Throttled Blower 47
31 Pressure Oriented Surge Control System for Variable
Speed Blower 48
32 Flow Oriented Surge Control System for Variable Speed
Blower 49
vn
-------
FIGURES (cont'd)
Number Page
33 Surge Control System to Maximize Blower Operating
Range 50
34 Plate Type Aerator 51
35 Updraft Type Aerator 51
36 Combination Type Aerator. 52
37 Brush Type Aerator. 52
38 Submerged Turbine Mixer Oxygen Dissolution System 54
39 Surface Mixer Oxygen Dissolution System 55
40 Open-Loop Control 63
41 Closed-Loop Feedback Control 64
42 Closed-Loop Feedforward Control 65
43 Feedforward-Feedback Control 66
44 Plug-Flow, Single Pass Oxidation Tank - Probe/Receptacle
Locations 70
45 Plug-Flow, Multiple Pass Oxidation Tank - Probe/Receptable
Locations 70
46 Complete Mixed Process Oxidation Tank - Probe/Receptacle
Locations 71
47 Single Aerator Installation - Probe Location 71
48 Multiple Aerator Installation - Oxidation Tank -
Probe/Receptacle Locations 72
49 Multipoint Dissolved Oxygen Monitoring System Using a
Single, Multiplexed Analyzer . . 72
Vlll
-------
FIGURES (cont'd)
Number Page
50 Dissolved Oxygen Probe Assembly 73
51 Diffused Air Aeration Blower Control System 75
52 Typical Centrifugal Blower Starting/Stopping Diagram 78
53 Diffused Air Aeration Dissolved Oxygen Control System. ... 80
54 Feedforward-Feedback Hydraulic Load Following Dissolved
Oxygen Control System 81
55 Feedforward-Feedback Hydraulic and Organic Load
Following Dissolved Oxygen Control System 82
56 Mechanical Aeration - Plug Flow Configuration 85
57 Mechanical Aeration - Completely Mixed Configuration ... 86
58 On-Off Aerator Control System 87
59 Two-Speed Aerator Step Control System 88
60 Aerator Capacity/Time Step Diagram 89
61 Variable Speed Aerator Control System 90
62 Variable Impeller Depth Control System 90
63 Variable Level Control System 91
64 Pure Oxygen Dissolution Control System 92
65 Automatic DO Control System for Mechanically Aerated
44 dm3/s (1 mgd) Activated Sludge Plant 109
66 Automatic DO Control System for Diffused Air 44 dm3/s
(1 mgd) Activated Sludge Plant 110
IX
-------
FIGURES (cont'd)
Number Page
67 Automatic DO Control System for Diffused Air 0.44 m3/s
(10 mgd) Activated Sludge Plant Ill
68 Automatic DO Control System for Diffused Air 2.2 m3/s
(50 mgd) Activated Sludge Plant 112
69 Automatic DO Control System for Diffused Air 4.4 m^/s
(100 mgd) Activated Sludge Plant 113
70 Average U.S. Industrial Power Cost per KWh Compared to
Industrial Rates of Typical Utilities 121
71 Automatic Dissolved Oxygen Control Economics for
Various Size Activated Sludge Plants 125
x
-------
TABLES
Number Page
1 Application Chart for Blowers Used in Sewage Plant
Aeration Systems 38
2 Comparison of Unox Operating Results with a Typical
Conventional Air Aeration Plant 56
3 Relation of Oxygen Transfer Efficiency to Aerator Power
Efficiency 58
4 Summary of Aerator Characteristics Related to Activated
Sludge Aeration 61
5 Case History Plants 96
6 Summary of Plant Loading, Flow and Test Duration Data ... 97
7 Summary of Performance Comparisons Between Manual and
Automatic Dissolved Oxygen Control 98
8 Test Data Averages of Case History Plants for Automatic
Compared to Manual Dissolved Oxygen Control 99
9 Design Data for Typical Activated Sludge Plant Aeration Systems 104
10 Installed Cost of Components Unique to Automatic Dissolved
Oxygen Control Systems 115
11 Typical Frequencies of Dissolved Oxygen Probe
Maintenance Functions 116
12 Estimated Annual Instrumentation Contract Parts and Labor
Cost for Automatic Dissolved Oxygen 117
13 Estimated Incremental Annual Instrumentation Contract
Parts and Labor Cost for Automatic Dissolved Oxygen
Control Equipment 118
xi
-------
TABLES (cont'd)
Number
14 Estimated Annual Labor Credit for Automatic Dissolved
Oxygen Control 119
15 Estimated Annual Plant Power Cost 120
16 Estimated 1975 Operating and Maintenance Costs of
Adding Automatic Dissolved Oxygen Control to
Various Size Activated Sludge Plants 122
17 Economic Analysis of Adding Automatic Dissolved Oxygen
Control to Various Size Activated Sludge Plants in 1975 . 123
18 Priority Ranking of Factors Affecting Choice of Dissolved
Oxygen Control System 131
xii
-------
LIST OF ABBREVIATIONS AND SYMBOLS'
Instrument Identification Letters
AC analysis controller
AE analysis element
AIC analysis indicating controller
AIR analysis indicating recorder
AIT analysis indicating transmitter
AR analysis recorder
AS analysis switch
AY analysis computing relay
FE flow element
FI flow indicator
FIC flow indicating controller
FT flow transmitter
Instrument Line Symbols
^— process flow line
AS;
pneumatic signal
electrical signal
air supply
mechanical linkage
HIK manual control action
HS manual (hand) switch
KG time (step) controller
PIC pressure indicating controller
PIT pressure indicating transmitter
TR temperature recorder
TT temperature transmitter
UC multivariable controller
UY multivariable computing relay
ZIC position indicating controller
ZT position transmitter
ZY position computing relay
Instrument Symbols
C~) local mounted instrument
O rear panel instrument
Q front panel instrument
interlock
Xlll
-------
LIST OF ABBREVIATIONS AND SYMBOLS (cont'd)
Control Valve Body Symbols
M gate valve
[>0 globe valve
101 plug valve
101 ball valve
Control Valve Actuator Symbols
O spring opposed diaphragm
Fn pilot actuated cylinder
~]~ hand actuator
(M) rotary motor
Relay and Controller Function Designations
I-Q on-off control
2 add or totalize
A subtract
A 1-0 differential gap control
-=-.•*•«• bias
|\| butterfly valve
f\| check valve
£><] 3-way valve
[xl 4-way valve
Primary Sensing Element Symbols
—|if— orifice plate
£=C] venturi or flow tube
pitot tube
Xn
f (x)
o/
/o
proportional control
multiply
divide
extract square root
I/p
P/I
d/dt
raise to power
characterize
high select
low select
current to pneumatic
pneumatic to current
integral control
derivative control
XIV
-------
LIST OF ABBREVIATIONS AND SYMBOLS3 (cont'd)
Other Abbreviations and Symbols
AA aeration air SP set point
CXHX hydrocarbons S starter
DO dissolved oxygen TOO total organic carbon
FC fail closed T trap
FO fail open VS variable speed drive
LEL lower explosive limit ^> panel patchboard
O2 oxygen <^> field patchboard
Extracted from Instrument Society of America "Instrument Symbols and Iden-
tification," (ISA-S5.1-1973 or ANSI Y32.20-1975) (13). Any abbreviations
and symbols used in this manual and not listed here can be found in
Reference 13.
UNITS
The International System of Units (SI) is used in this manual in accordance
with American Society of Testing Materials (ASTM) Metric Practice Guide
E 380-72 or American National Standards Institute (ANSI) Z 210.1-1973.
xv
-------
ACKNOWLEDGMENTS
Grateful appreciation is acknowledged to Mr. Valentine H. Lewin of the
Thames Water Authority, Oxford, England for information on the Oxford Sewage
Works and various aspects of dissolved oxygen control.
Mr. Ron Briggs, Head of the Laboratory Services Division of the Water
Research Centre, Stevenage, England provided information on dissolved
oxygen measurement and control systems and the Rye Meads Works.
Sincere thanks is offered to the management and operating staffs of the
treatment plants at which performance tests were run for the purposes of this
report or from which data was obtained from past testing. Particular thanks
and appreciation goes to the following:
1. Mr. Richard Finger of the Renton Wastewater Treatment Plant
2 . Mr. Ron Doty and Mr, Ramon Valdez of the Palo Alto Water
Quality Control Plant
3. Mr. James Johnson and Mr. Bob Swanson of the Valley Community
Services District Wastewater Treatment Plant
4. Mr. James Y. Ihara, City of Sparks, and Mr. Gene Davis of
the Reno-Sparks Joint Water Pollution Control Plant
5. Mr. Edward Walker of the Simi Valley Water Quality Control Plant
6. Mr. Alvin L. Benas and Mr. Eugene L. Mooney of the San
Francisco International Airport Water Quality Control Plant
7. Mr. John Lawton, Mr. Lowell L. Knapp, Jr. and Mr. Bill Roepke
of the St. Regis Paper Company
8. Mr. Carl A. Nagel of the Los Angeles County Sanitation District
9. Mr. Frank Belick, Mr. Edward Becker and Mr. Erwin Jacoby of
the San Jose-Santa Clara Water Pollution Control Plant
Additional information and assistance on the plants selected as case
histories was provided by Systems Control Inc. in Palo Alto on the Palo Alto
Plant, Dr. William K. Genthe of Rexnord on the St. Regis Plant and Mr. John
Kontor from the Environmental Systems Center of Raytheon Company on the
Cranston Water Pollution Control Facility.
xvi
-------
SECTION 1
INTRODUCTION
Control of air or pure oxygen dissolution in the mixed liquor is an impor-
tant parameter in the activated sludge process. Increasingly, there is an
economic incentive to minimize unnecessary oxygenation. The hypothesis
is that a desirable automatic control strategy is to add only sufficient air or
oxygen to meet the time-varying demand of the mixed liquor. Field tests
indicate that the benefits of such a control strategy are improved process
performance and efficiency.
The purpose of this manual is to provide a recommended design procedure
to guide the design engineer in the selection of aeration equipment and control
techniques for achieving optimal dissolved oxygen control of the activated
sludge process. The information included herein on performance, operation
and maintenance of various dissolved oxygen control systems, instrumentation
and mechanical equipment details, and associated capital and operating costs
was obtained from the available literature, reports from operating plants,
progress reports of demonstration studies, and private communications with
investigators actively working in the field. Design guidelines are developed
from these sources.
-------
SECTION 2
CONCLUSIONS
The capital and operating cost of automatic dissolved oxygen (DO) control
systems is justified for activated sludge plants larger than 44 dm^/s (1 mgd),
provided equipment is selected and applied in accordance with the guidelines
of this manual and a power rate equal to or greater than the national average
power rate is applicable.
Automatic DO control may not be warranted in plants with (a) constant
loading conditions (b) insufficient oxidation tank or aerator capacity or
(c) aerator turndown limitations.
Careful attention should be given to blower selection by the designer if
power savings are to be expected using an automatic DO control system.
If aeration blowers are used to supply air to two or more separate pro-
cesses, the control systems which regulate air supply to meet the varying
air demands of each process should be decoupled to prevent interaction under
dynamic operating conditions.
DO probe layouts should be flexible enough to permit reconfiguration
while the plant is on line. Multiple DO probe receptacles should be provided
in each tank to permit probe relocation.
An effective method of activated sludge DO control with centrifugal
blowers is a combination of a discharge pressure control loop for the blowers
and a DO regulated flow control loop for each oxidation tank air feed header.
Centrifugal and axial blowers have operating characteristics that allow
horsepower to be conserved over a variable flow range and are, therefore,
better suited to automatic DO control systems than positive displacement
blowers or mechanical mixers.
DO probe maintenance is a major expense in automatic DO control sys-
tems. Maintenance time can be significantly reduced through proper probe
selection, good maintenance programs and probe installations that facilitate
probe access for cleaning and calibration.
-------
Automatic DO control offers the potential advantage of decreased effluent
quality variability through optimization of the air/oxygen application rate.
Managers of existing activated sludge plants should examine the perfor-
mance of their existing DO control systems in terms of effluent quality and
operation and maintenance costs. Where improvement is possible, modifi-
cations to, or substitution of, the control system in accordance with guide-
lines presented herein should be considered. Projected benefits should be
compared to net annual costs of new equipment, where such comparisons can
be estimated.
Managers of existing plants with automatic control systems should run
long term performance tests of manual compared to automatic DO control to as-
certain how effectively their existing control systems perform. Modifications
to the control system should be considered if results are less than expected.
Designers of activated sludge plants should consider potential power
savings through automatic DO control when selecting aeration equipment. The
performance record of various DO control systems should be examined to
determine which system is desirable for a particular plant configuration.
Automatic DO control should be considered at the onset, rather than as an
afterthought, of design.
-------
SECTION 3
RECOMMENDATIONS FOR FUTURE RESEARCH
An important by-product of this study is the identification of certain
areas in the application of dissolved oxygen control systems that warrant
further research. The itemization below is not exhaustive but does reference
many specific problems observed in the field. Research areas are identified
in terms of application of DO control to the activated sludge process.
1. Evaluate the effect of dissolved oxygen control system set point level on
effluent quality and sludge handling.
2. Determine the optimum dissolved oxygen set point, considering both
performance and economics ;>for various process configurations.
3. Develop guidelines for location of DO probes in oxidation tanks for
different process configurations, such as conventional, step feed,
extended aeration and contact stabilization.
4. Evaluate different control strategies employing DO measurement and
control such as feedforward, feedback, adaptive and optimal control.
5. Recommend maintenance procedures and equipment installation guide-
lines for efficient maintenance of DO probes.
6. Evaluate different methods of conserving power when applying mechani-
cal aeration and positive displacement blowers in automatic DO control
systems.
7. Identify wastewater characteristics that lead to greater frequency of DO
probe cleaning.
8. Evaluate use of cleaner/agitator assemblies compared to careful place-
ment of DO probes in oxidation tanks in terms of resultant probe mainte-
nance requirements and accurate DO readings.
9. Evaluate the capital cost, operating cost, and performance of other
devices compared to DO probes for measuring oxygen demand in oxida-
tion tanks.
10. Evaluate current DO probe design technology for activated sludge DO
control system application. Rank desirable DO probe characteristics in
order of priorities.
11. Conduct additional longer term tests of automatic vs manual DO control
of the activated sludge process at a large number of plants. Evaluate
any resultant power savings and process performance improvements in
terms of possible correlation with specific DO control systems, plant
design or operating characteristics.
-------
SECTION 4
UNIT PROCESSES
The activated sludge process has evolved through the years into a
versatile biological treatment method. It has been adapted to a variety of
biological waste treatment problems, resulting in several modifications to
the basic method. This chapter describes preaeration and postaeration systems
and schematics and process dynamics of various activated sludge systems.
PREAERATION SYSTEMS
Aeration of wastewater prior to primary sedimentation has been practiced
for over 50 years in the United States. It is generally employed for odor
control and to improve treatability of wastewater. Short aeration periods of
15 minutes or less are adequate for these purposes, although longer aeration
periods yield the additional benefits of grease separation and improved solids
flocculation (28, 31). Additionally, aeration-induced agitation improves
separation of organic and inorganic fractions of the solids, which enhances
grit removal and promotes uniform distribution of suspended solids to the
treatment works.
The two basic parameters in preaeration system design are air feed rate
and detention time. Preaeration tank depths are generally about 4.5m
(15 feet) , and air requirements range from 0.75-3 m3/m^ (0.1 to 0 .4 cu ft
per gal) of wastewater (20) . In order to maintain proper agitation, the air
supply system should be capable of providing air at the rate of 0.5-1.9 dm /s
(1.0 to 4.0 cfm) in the wastewater distribution channel per lineal 0.3m (foot)
of channel.
Where chemical addition is combined with the preaeration process, air
requirements may vary from those reported for conventional preaeration systems.
Studies at the Central Contra Costa Sanitary District in Concord, California
have shown that air supply rates for lime flocculation may be lower than the
rates used for conventional preaeration to avoid shearing of floe (12, 15).
Effective preaeration has been achieved with detention times of 45 minutes
and less.
-------
POSTAERATION SYSTEMS
Minimum effluent dissolved oxygen (DO) concentrations are required for
treatment plants in many states to minimize oxygen depletion of receiving
waters. Minimum DO concentrations of 4 to 5 ppm are common criteria for
surface water quality. Yet DO levels in secondary effluents normally range
from 0.5 to 2.0 ppm. Postaeration is a way of increasing effluent dissolved
oxygen to meet discharge requirements.
There are four possible methods of postaerating treatment plant effluent.
Two of these methods, diffused aeration and mechanical aeration, are
essentially the same as those employed in biological treatment processes.
The other methods, called cascade aeration and U-tube aeration,can be
attractive alternatives. In cascade aeration, effluent is discharged over a
series of steps or weirs in thin layers with the objective of maximizing turbu-
lence, thereby increasing oxygen transfer. Cascade aeration is an attractive
postaeration technique where sufficient head is available after secondary
clarification. While U-tube aerators have not yet been employed in post-
aeration, they have been suggested as an attractive alternative to other post-
aeration methods (31).
ACTIVATED SLUDGE SYSTEMS
In the activated sludge process, suspensions of microorganisms stabi-
lize soluble and colloidal organics to carbon dioxide and. water in the presence
of molecular oxygen. In the stabilization process, waste organic material is
used to synthesize new cells, some of which subsequently undergo endo-
genous respiration as substrate concentration decreases. Synthesized cells
are removed from 'treated wastewater in secondary clarifiers. For continuous
operation, a major fraction of the cells must be recycled to the oxidation tank.
Excess sludge is withdrawn from the clarifier underflow for disposal. Oxygen
is required in the process to support oxidation and synthesis reactions.
Due to the adaptability of the process, activated sludge plants have
been designed with a variety of flow plans. Nearly all these modifications
have been outgrowths of the conventional process flow plan and have resulted
from attempts to improve or correct deficiencies in the conventional system.
There are advantages and disadvantages to each modification. Some achieve
better biochemical oxygen demand (BOD) and suspended solids (SS) removals
than others. Some cost less to construct; others cost less to operate. Some
produce less sludge, and some provide better waste removal. Each of these
factors must be considered in. selecting a flow plan for a particular application.
-------
Conventional Process
The conventional activated sludge process is shown schematically in
Figure 1. In this system, primary effluent is mixed with return activated
sludge and passed on to the oxidation tank as mixed liquor. The oxidation
tank is the biochemical reactor in which microorganisms in the mixed liquor
aerobically stabilize organics in the wastewater. Mixed liquor leaving the
oxidation tank is displaced into secondary clarifiers , where suspended solids
are separated from treated wastewater. A large fraction of the settled biomass
is returned to the head of the oxidation tank as returned activated sludge to
sustain the continuing reaction. Excess solids are removed from the system
as waste activated sludge. Clarified secondary effluent continues through
the plant for subsequent treatment and disposal.
PRIMARY
EFFLUENT
EFFLUENT
WAS
Figure 1 . Conventional activated sludge process.
Most conventional activated sludge systems are designed for organic
loadings of 2.3-5.8 mg/s/kg MLVSS (0.2 to 0.5 Ibs BOD/day/lb MLVSS).
Designs are usually based on volumetric loadings of 3.7-7.4 mg BOD/m3/s
(20 to 40 Ibs BOD/day/1000 cu ft) and sludge retention times of 5 to 15 days.
Oxidation tanks commonly provide 6 to 10 hours of aeration time. Sludge
return rate is normally between 25 and 50 percent with minimum and maximum
rates of 15 and 75 percent respectively.
To achieve plug flow, oxidation tanks are designed as long rectangular
basins with length to width ratios ranging from 5:1 to 50:1. Primary effluent
and return activated sludge are fed at one end of the tank and treated mixed
liquor is withdrawn at the opposite end. In addition to supplying oxygen for
stabilization, aeration mixes the tank contents. Conventional air diffusion
equipment is arranged to obtain a "spiral-flow" action perpendicular to the
direction of waste flow.
A disadvantage of the conventional system is that plug flow produces a
continually changing environment within the oxidation tank. The concentration
of substrate continually decreases as the flow proceeds down the tank. Con-
7
-------
sequently, food/microorganism (F/M) ratio is constantly varying, with
higher BOD loadings at the head of the oxidation tank and relatively low
loadings at the effluent end. Since air is supplied evenly throughout the
tank, DO concentrations at the head of the tank are low, while those at the
effluent end are much higher.
The conventional plant can be modified to incorporate a tapered aera-
tion system. With tapered aeration, the air supply system is "tapered" to
match actual oxygen demand in the oxidation tank. More oxygen is supplied
at the front of the tank, where oxygen demand is greatest, decreasing in
proportion to load toward the discharge end of the tank. A well designed
tapered aeration system should have a fairly constant oxygen level throughout
the oxidation tank. In order to balance aeration taper against actual oxygen
load, it is necessary to accurately determine the relationship between oxygen
demand and longitudinal distance along the tank. Figure 2 shows an average
relationship for plug flow systems. Such a curve can be useful for estimating
aeration taper requirements .
0 25 50 75 100
PERCENT OF TOTAL OXIDATION TANK LENGTH
Figure 2. Oxygen Utilization in a plug flow system.
Volumetric loading and food to microorganism ratios for tapered aeration
plants are in the same range as those used in conventional activated sludge
systems. Oxygen requirements average about 1.4 mg/s/g MLVSS (5 mg/l/hr/
1000 mg MLVSS) and detention time varies from one to 12 hours (17).
-------
Complete Mix Process
In the complete mix activated sludge process, effluent wastewater and
return activated sludge are introduced uniformly throughout the oxidation tank,
as shown in Figure 3. This uniform load distribution results in a constant
oxygen demand throughout the tank and in a more homogeneous sludge com-
munity. As a result, the oxidation tank volume in a complete mix system is
used more efficiently than in a conventional plant, permitting design for
higher volumetric loadings. Complete mix plants have been designed for
volumetric loadings of 9.3-22 mg BOD/m3/s (50 to 120 Ib BOD/day/1000 cu ft)
at F/M ratios of 2.3-6.9 mg EOD/s/kg MLVSS (0.2 to 0.6 Ib BOD/day/lb
MLVSS) and hydraulic detention times of 3 to 5 hours (20) . For this type of
system, return sludge ratios should be as high as possible, consistent with
good pumping economics. The recommended range for recycle ratios is 35 to
100 percent (17) .
WASTEWATER
OR
PRIMARY
EFFLUENT
rt 111 tt
OXIDATION TANK
i.
11 <' i' \
v r
EFFLUENT
R AS
WAS
Figure 3. Complete mix process.
In the complete mix system, influent wastewater and return sludge
should be mixed with the entire tank contents almost instantaneously. This
is achieved through both proper initial flow distribution and adequate mixing
in the tank. For rectangular tanks, influent distribution and mixed liquor
withdrawal normally occur along the entire longitudinal tank axis, as shown
in Figure 3. Aeration is normally arranged to mix contents in the longitudinal
direction. This provides greater dispersal of the influent and more rapid mix-
ing within the tank. A round or square aeration tank is more desirable where
mechanical aerators are to be used. Depending on the size and arrangement
of mechanical aerators , single or multiple point feed and discharge can be
provided. Mechanical aerators are especially attractive for use in complete
mix systems because they (a) put more energy into the mixing process, and
(b) mix in all directions (17).
-------
As shown in Figure 4, complete mixing permits use of higher BOD
loadings for a given temperature and oxygen uptake rate. Lesperance (17)
has noted that this increased BOD loading is only an apparent increase. In
the complete-mix system, more of the microorganisms in the mixed liquor
are actively stabilizing waste organics due to process homogeneity. There-
fore, while the biochemical reaction rate does not increase, a higher effec-
tive BOD loading rate results.
CONVENTIONAL
PROCESS
COMPLETE MIX
PROCESS
264-288 K
LOADING, kg BQD/kg MLSS/DAY
NOTE: TF= 1.8 TK - 460
Figure 4, Oxygen utilization in activated sludge systems.
Figure 4 was compiled from full scale and pilot plant operating data on
oxygen uptake rates for a variety of wastes, including sewage, oil refinery
pharmaceutical, chemical, and pulp and paper plant wastes (17). This figure
is intended to show only an operating range for activated sludge processes.
Actual plant performance depends upon the capability and characteristics of
process components, such as oxidation tanks and aeration equipment.
Furthermore, Figure 4 does not account for diurnal flow and load variations,
which can increase oxygen uptake rate 40 to 60 percent above the average
daily design figure for carbonaceous oxidation alone.
10
-------
Step Feed Process
A typical flow diagram for a step feed activated sludge plant is shown
in Figure 5. The influent wastewater is split into several portions, which
are introduced at several points along the oxidation tank. Return sludge is
normally introduced at the head of the oxidation tank. Distribution of in-
fluent flow along the oxidation tank compensates for the high initial oxygen
demand normally experienced in conventional plants .
OXIDATION TANK
PRIMARY
EFFLUENT
EFFLUENT
»•
to. WAS
Figure S. Step feed process.
As in the conventional activated sludge system, air is fed uniformly
along the tank. As a resultof step feeding, however, a plot of dissolved
oxygen versus tank length results in a characteristic saw-tooth curve, which
has given rise to the misnomer "step aeration1'. Regardless of name, the net
result is that partial load equalization produces a more homogeneous system.
Hence, the process can take higher BOD loadings with shorter detention
times and more effective utilization of air.
Step feed plants are usually designed for volumetric loadings of 7.4-
11 mg BOD/s/m3 (40 to 60 Ib BOD/day/1000 cu ft) and F/M ratios of 2.3-
4.6 mg BOD/s/kg MLVSS (0.2 to 0.4 Ib BOD/day/lb MLVSS) . Plant deten-
tion times are characteristically three to five hours and efficiencies range
from 85 to 95 percent BOD removal.
11
-------
Sludge Reaeration Process
The sludge reaeration process is a variation of the step feed process.
It is a technique for improving systems which are not achieving design BOD
reduction efficiency due to insufficient oxidation tank volume. In general,
the size of a step aeration tank with sludge reaeration can be less than half
the size of a conventional tank when both tanks receive the same influent
solids loading (2).
Sludge reaeration can be provided before or after secondary clarifica-
tion. In some cases, sludge reaeration is provided in an activated sludge
system by eliminating wastewater feed to the first pass and directing the
flow to the downstream passes. A sludge reaeration zone is thereby es-
tablished in the first pass (4).
Contact Stabilization Process
As shown in Figure 6, the contact stabilization process is essentially
a two-step activated sludge process. In the first step, wastewater BOD in
the colloidal or insoluble state is rapidly removed from wastewater in a rela-
tively short contact time by the combined mechanism of biological sorption,
synthesis, and flocculation. For domestic wastes, this process proceeds
rapidly, requiring as little as 20 to 40 minutes for removal of most of the
colloidal, finely suspended and dissolved organics . However, while the
BOD is almost completely adsorbed, or physically removed from the liquid in
this time, the microorganisms in the sludge have not had enough time to re-
duce or stabilize the organic matter. In the second stage of the process,
settled sludge is pumped to a stabilization tank where microorganisms are
aerated in the absence of a food supply, metabolically assimilating adsorbed
organics. Microorganisms are retained in the stabilization tank until they
enter endogenous respiration, becoming "lean and hungry" . Thus, when they
are finally discharged to the contact tank, they are capable of rapidly remov-
ing large amounts of BOD.
The contact stabilization process is designed for volumetric loadings
of 11-14 mg BOD/s/m3 (60 to 75 Ib BOD/day/1000 cu ft) with F/M ratios of
2.3-6.9 mg BOD/s/kg MLVSS (0.2 to 0.6 Ib BOD/day/lb MLVSS) . This
process is commonly thought to have a somewhat lower BOD removal effi-
ciency than the conventional or complete-mix processes. However, effi-
ciencies of 90 to 94 percent have been reported for a number of plants in
Texas and New Jersey (31). For domestic wastewater containing normal
amounts of insoluble, colloidal BOD, contact tanks are designed for deten-
tion times of 0.5 to 1 hour at average flow, and stabilization tanks are de-
signed for 3 to 6 hours based on sludge recycle flow. Total air required for
this process is similar to that for conventional activated sludge. Normally,
an equal amount of air is required for the contact and reaeration tanks.
12
-------
PRIMARY
1
EFFLUENT
•»•
STABILIZATION
TANK
CONTACT
TANK
RAS
-*
CLARIFIER
1
t
^mm
EFFLUENT
WAS
Figure 6. Contact stabilization process.
The major advantage of the contact stabilization process is the reduced
oxidation tank volume needed to support the biochemical reaction. However,
this advantage is lost where BOD is largely in the soluble state, as in the
case of many industrial wastes. In addition to a smaller total requirement
for aeration volume, the contact stabilization process has the advantage of
being able to sustain greater shock and toxic loadings than the conventional
process. This resistance to upset is largely due to the biological buffering
capacity of the stabilization tank and to isolation of most of the activated
sludge from the main plant flow stream.
Hatfield and Kraus Processes
Some types of wastes, especially fruit and cannery wastes, are defi-
cient in nitrogen, an essential nutrient for microorganism growth. Uncor-
rected, this deficiency limits waste treatability. The Hatfield and Kraus
processes are variations of the contact stabilization process in which anaer-
obic digester effluent is supplied to the sludge reaeration unit. This fortifies
the largely carbonaceous activated sludge solids with amino acids and other
nitrogenous byproducts of the anaerobic digestion process. As shown in
Figure 7, the essential difference between the Hatfield and Kraus processes
is that in the Kraus process some return sludge bypasses the reaeration unit
and goes directly to the mixed liquor oxidation tank.
These processes are normally designed on the basis of volumetric
loadings of 7.4-18 mg BOD/s/m3 (40 to 100 Ib BOD/day/1000 cu ft) and
F/M ratios of 3.5-9.2 mg BOD/s/kg MLVSS (0.3 to 0.8 Ib BOD/day/lb
MLVSS) with hydraulic detention times of 4 to 8 hours (20) .
13
-------
INFLUENT
SLUDGE
in
OXIDATION
Ji 4 < i
TANK
TTTH
EFFLUENT
WAS
(KRAUS PROCESS)
SLUDGE
REAERATION
DIG. SLUDGE £ SUPERNATANT
WASTE SECONDARY SLUDGE
•RAS
SLUDGE DISPOSAL
Figure 7. Hatfield and Kraus processes.
Short-Term Aeration Proces_s
The short-term aeration process is also known as high rate aeration
and modified aeration. As the names imply, the process is a modification of
the conventional activated sludge system which applies high loading rates
with short aeration times. The major advantage of the process is construc-
tion economy due to the reduced oxidation tank capacities required. How-
ever, due to the relatively small mass of microorganisms in contact with the
wastewater and the short period of contact, a greater amount of unused BOD
remains in the process effluent. Additionally, the process employs high F/M
14
-------
ratios. Net growth of volatile suspended solids is relatively large in these
systems. Incomplete removal of these solids degrades effluent quality;
complete removal presents a problem of sludge disposal.
The short-term aeration process follows the same basic schematic as
the conventional activated sludge process. However, the short-term process
allows volumetric loadings of 18-46 mg BOD/s/m3 (100 to 250 Ib BOD/day/
1000 cu ft) and F/M ratios of 5.8-60 mg BOD/s/kg MLVSS (0.5 to 5 Ib BOD/
day/lb MLVSS) (38). Hydraulic detention times are normally in the range of
0.5 to 2 hours, and the process is reported to have an efficiency of 75 to 90
percent BOD removal (20) .
Extended Aeration Process
The extended aeration process follows the same basic schematic as
the conventional activated sludge process. By virtue of its low loadings and
long aeration times, the extended aeration process operates in the endoge-
nous respiration portion of the growth curve.
As originally conceived, the extended aeration process was designed
to provide continuous sludge return with no sludge wastage. In practice,
this meant that the net growth of suspended solids was wasted from the sys-
tem in the process effluent. As a result, normal domestic flows treated by
the extended aeration process have shown apparent BOD removals of only 75
to 85 percent. However, Lesperance (17) reports that with competent daily
operating control and proper sludge control and wastage, extended aeration
systems should yield a BOD reduction of 98 percent or better.
With the exception of comminution of solids, primary treatment is
normally omitted from the extended aeration flow sheet. Treatment plants are
generally sized for less than 44 dm3/s (1 mgd) capacity, with wastes enter-
ing the plant continuously or intermittently. Plants are designed for volu-
metric loadings of 1.8-4.6 mg BOD/s/m3 (10 to 25 Ib BOD/day/1000 cu ft)
and F/M ratios of 0.58-1.7 mg BOD/s/kg MLVSS (0.05 to 0.15 Ib BOD/day/
Ib MLVSS) . Hydraulic detention times are normally in the range of 18 to 36
hours and mean cell residence times range from 20 to 30 days.
15
-------
SECTION 5
AERATION SYSTEMS
The environmental engineer designing a system which employs aeration
is confronted with a bewildering assortment of aeration equipment and pro-
prietary devices, all purporting to provide the most economical solution to
his design problem. Based on his assessment of economic and functional
considerations, he must select an aeration method and choose the applicable
aeration device. This chapter reviews aeration methods, aeration equipment,
and basic considerations in aeration systems design.
OXYGEN TRANSFER
Aeration devices are used for many purposes in wastewater treatment,
including preaeration, postaeration, grit removal, channel aeration, pond
aeration and activated sludge aeration. The comments in this chapter are
directed primarily to aeration of mixed liquor in the activated sludge treat-
ment process. In that context, aeration has four basic functions, which
are: (a) providing oxygen for biological oxidation of organic substrates
and respiration of the microorganisms in the oxidation tank, (b) agitating
the mixed liquor to prevent sludge from settling in the oxidation tank, (c)
mixing to ensure intimate contact between activated sludge microorganisms
and waste organic matter in the mixed liquor, and (d) removing carbon
dioxide produced in the oxidation of carbonaceous and nitrogenous sub-
stances .
Oxygen Transfer Theory
Gas transfer basically involves exchange of molecules between gases
and liquids. All that is required for exchange to occur is a gas-liquid inter-
face with a disequilibrium in the concentration of a particular type of gaseous
molecule. Hence, transfer can occur with (a) gas bubbles in a predominantly
liquid phase, (b) liquid droplets in a gaseous phase, and (c) relatively large
air-liquid interfaces, such as standing water surfaces.
Lewis and Whitman (18) postulated that two films, one liquid and one
gas, restricted the passage of gas molecules between the liquid and gaseous
phases. Where the solubility of gases is low, such as with CO2 and O2,
the rate of gas transfer may be expressed by Equation 1.
16
-------
f = Kt(°s - °1> «
where dc r .
~ = rate of change of concentration
K = transfer coefficient incorporating the specific area of the
gas/liquid interface
C = saturation concentration of dissolved gas
o
C. = actual concentration of gas dissolved in the liquid
For low solubility gases, the liquid film offers the most resistance to gas
transfer. Stirring or agitation of the liquid reduces thickness of the liquid
film, thus promoting greater gas transfer (i.e. , a larger K ) .
The two common methods of oxygenating activated sludge are (a) diffu-
sion and (b) mechanical agitation. In a diffused air system, oxygen transfer
takes place in three distinct phases: bubble formation, bubble rise, and
bubble collapse. During the time the bubbles are in contact with the liquid,
there is a continuous transfer of oxygen from the air through the interfacial
film to the liquid. Since the coefficient Kt is proportional to the specific
area of the gas/liquid interface, the size and number of bubbles are important
criteria with respect to optimal oxygen transfer; thus, more numerous, smaller
bubbles are inherently more efficient than coarse, large bubbles. In a me-
chanical aeration system, oxygen is transferred to the liquid by exposing the
liquid to the atmosphere through turbulent mixing. Consequently, the effi-
ciency of mechanical aeration is largely dependent upon the amount of turbu-
lence created per unit of power.
Oxygen Requirements
In the aeration of a mixture of primary effluent and return activated
sludge, the amount of oxygen required depends upon the quantity of carbona-
ceous BOD5 to be oxidized, the relative amount of endogenous respiration
taking place, and the required degree of nitrification.
In normal activated sludge treatment, when nitrification is not required,
the amount of oxygen needed to oxidize the BODg can be calculated by the
following equation:
B = X(BOD5) (2)
where B = oxygen required for carbonaceous oxidation, mg/1 or ppm
X = a coefficient
BOD5 = 5-day biochemical oxygen demand
17
-------
The coefficient X relates to the amount of endogenous respiration taking place
and to the type of waste being treated. For normal domestic wastewater, the
X value would range from a low of 0.5-0.7 for high rate activated sludge
systems to a high of 1.5 for extended aeration. For conventional activated
sludge systems, X can be taken as 1.0.
In the case of nitrification, the oxygen requirement for oxidizing ammonia
must be added to the requirement for BOD removal. The coefficient for nitro-
gen to be oxidized can be conservatively taken as 4.6 times the ammonia
content to obtain the nitrogen oxygen demand (NOD), and the value of X in
Equation 2 can be assumed to be approximately 1.0. This yields the follow-
ing oxygen requirement:
W = BOD + NOD (3)
D
where W = the total oxygen demand, ppm, and
NOD = oxygen required to oxidize a unit of ammonia, taken as
4.6 times the total Kjeldahl nitrogen (organic plus ammonia
nitrogen)
Since aeration devices are rated using tap water at standard conditions,
the rated performance of the aerator must be converted to actual process con-
ditions by the application of temperature corrections and by factors which re-
late waste characteristics to tap water characteristics. These factors are:
(T—o n)
1. Temperature corrections, made by the factor 1.024 ' where
T = process temperature in degrees C.
2. The a factor (the ratio of oxygen transfer in wastewater to that in
tap water), represented by:
a =
K (process conditions)
K (standard conditions)
(4)
Values of a can vary widely in industrial waste treatment applica-
tions, but for most municipal plants, it will range from 0.40 to
0.90.
3. The 3 factor (the ratio of oxygen saturation in waste to that in tap
water at the same temperature) . A value of 0.95 is commonly used,
The actual amount of oxygen required (W)*can be determined from the
amount transferred under test conditions (Wj by equation (5), taken from
Reference 1:
18
-------
W = WQ a h.024T~2°) ( Cs " Cl ) (5)
V J \ 9.2 '
where W = oxygen transferred at process conditions, Ib/day
Wn = oxygen transferred at standard conditions, Ib/day
(T = 20C, DO = .01 ppm, tap water)
T = process temperature, degrees C
C = oxygen saturation in water at temperature T, ppm
o
C, = process dissolved oxygen level, ppm
The above result may be converted to kg/s units by multiplying answer by
5.250 x 10~6.
Where nitrification is desired, the process dissolved oxygen level (C,)
must be set high enough to prevent inhibition of nitrification rates (4) . For
this purpose, a minimum value of 2 . 0 ppm is recommended. This value is
also applicable under peak diurnal load conditions, and the practice of
allowing the DO to drop below 2.0 ppm under peak is not recommended.
Using typical values for domestic sewage (a= 0.9, $ = 0.95, C-, = 2.0,
T = 21 and C =9.0), the relationship between oxygen required under test
conditions and that required under actual process conditions is: W = WQ/! . 5
Wn can be converted to hp requirements for m
metric air flow rates for diffused air plants.
latter is accomplished by solving Equation 6.
Wn can be converted to hp requirements for mechanical aerators or to volu-
metric air flow rates for diffused air plants. According to Aberley (1), the
w°
where Q = air flow, cfm (60°F, 14.7 psia)
e = aerator rated oxygen transfer efficiency at
standard conditions , percent
air composition =23 percent oxygen, weight basis
air density = .075 Ib/cf
3
The above result may be converted to m /s by multiplying answer by 4.719
x
ID-4-
19
-------
Applying the above equation to diffusers of various efficiencies produces
air rates from 94-38 m3/kg (1500 to 600 cubic feet per Ib) BOD5 + NOD cor-
responding to diffuser efficiencies of 6 to 15 percent (1) (Figure 8).
20
CO
z
o
r 18
Q
UJ O
14
o
UJ
5* l2
(L CL
111 '
o: z
Q or
10
8
Temperature 293K
(m3/kg)(!6.02) = cf/Ib
TF 1.8 T«- 460
30
60
90
100
AIR REQUIRED , m3/ ( kg BOD5 + NOD)
Figure 8. Relationship of aeration air requirements for oxidation
of carbonaceous BOD and nitrogen (4) .
Diurnal load variations are an important concern in designing air supply
systems. Hanson, et al, (11) reported substantial diurnal variations in
BOD^, total nitrogen, and ammonia nitrogen in the influent to the Mason
Farm Plant in Chapel Hill, North Carolina. Diurnal variations in wastewater
flow and nitrogen load for that plant are shown in Figure 9. Total nitrogen
was found to vary between 26 percent of average and 217 percent of average
in a 24-hour period,and ammonia nitrogen ranged from 30 to 223 percent of
average; BOD loading ranged from 13 to 190 percent of the daily average,
while wastewater flow rate varied between 39 and 144 percent of average
flow.
The North Carolina data do not show the relative fractions of soluble
and insoluble BODr . Hence, it is not possible to accurately estimate the
percentage of influent BOD leaving primary clarification for subsequent
20
-------
3
u.
iu
-------
secondary treatment. On the other hand, nitrogen is largely soluble; thus,
primary influent nitrogen loadings are approximately equal to secondary
treatment process influent nitrogen loadings.
It is important to note in Figure 9 that fluctuations in concentration very
nearly match diurnal variations in waste-water flow. The combined effect of
diurnal variations in flow and concentration is a fairly dramatic fluctuation in
total nitrogen load. Parker, etal.,(4) have shown maximum hourly ammonia
loads nearly 2-1/2 times as great as average daily ammonia loads for plants
with maximum hourly flows of only 1.5 to 2 times the average daily flows, as
shown in Figure 10. The complete range of diurnal load variations can be
quite extreme. In Figure 11, peak to minimum hourly load ratios are plotted
against the flow peaking factor; ratios as high as 10 to 1 have been observed.
The extra aeration capacity and tankage required for handling diurnal varia-
tions, where nitrification is practical, may dictate in-plant flow equalization
in many instances . The reductions in capital and operating cost of aeration
and tankage in aeration facilities must be compared with the cost of flow
equalization to determine applicability to specific cases . Design data for
flow equalization are contained in Chapter III of Reference 31.
To conserve energy, careful consideration must be given to maximizing
oxygen utilization per unit of input power. Under significant load variation,
aeration systems should be designed to match the load variations, while
economizing on power input. Designing an aeration system to provide for
maximum hourly demand twenty-four hours a day, without turndown capability,
results in overaeration most of the time, with wasteful losses of power.
As an example of the importance of large aeration capacity to achieve
nitrification, let us assume a plant having an average flow of 0.44 m3/s
(10 mgd), a raw sewage BOD5 of 250 ppm, an average NHg concentration of
30 ppm, a peak NHg concentration of 60 ppm, a primary BOD,- removal of 35
percent, an aerator efficiency of 10 percent, and an effluent BODr of 20 ppm.
For such conditions, the air delivery capacities are as follows: (a) no nitri-
fication, 4.72 m3/s (10,000 cfm) and (b) complete nitrification, 13.7 m3/s
(29,000 cfm). Thus, the aeration capacity required for complete nitrification
under peak ammonia input is about three times that needed for carbonaceous
oxidation. This is one reason why many aeration plants designed for conven-
tional air rates are unable to nitrify.
AERATION TECHNIQUES
Two basic aeration methods are normally used: (a) diffused aeration,
where the air is introduced into the mixed liquor through a series of diffusion
devices; and (b) mechanical aeration, where air is introduced to the mixed
liquor by exposing the liquid to the air.
22
-------
2.5
<
Q
X
<
1.0
KEY
Plant
Lebanon, Ohio
L ivermore, Ca
CCCSD,Ca
Sacramento, Ca
Blue Plains, DC
Chapel Hill, NC
Canberra, Austra
Weston Creek.
Belconnen
(dm3/s)(2.28 X I0~2) =
ADWF
Sample dm3/s
Primary 50
Roughing Filter 150
Primary 920
Primary 2,000
Primary 12,000
Raw 80
ia -
Raw 40
Raw 440
1.5
2.0
2.5
MAXIMUM HOURLY FLOW, m3/s / AVERAGE DAILY FLOW, m3/s
Figure 10. Relation between ammonia peaking and hydraulic peaking loads
for treatment plants with no in-process equalization (4) .
23
-------
See Figure 10
for symbols
1-0 1.5 2,0 2,5
MAXIMUM HOURLY FLOW, m3/s / AVERAGE DAILY FLOW, m3/s
Figure 11 . Relationship of maximum/minimum nitrogen load ratio to
maximum/average flows (4) .
24
-------
Diffused Aeration
The term diffused aeration is applied to any process where air from an
external source is injected into the oxidation tank through a diffusion device.
The diffusion device can be anything from a perforated pipe to a complex dif-
fuser. Diffused aeration systems are broadly segregated into two classifica-
tions; coarse bubble diffusion and fine bubble diffusion. Each of these main
classifications contains a number of subclassifications .
Coarse Bubble Diffusers---
As the name implies, a coarse bubble diffuser releases relatively large
air bubbles into the mixed liquor. Coarse bubble diffusers fall into four
general categories: orifice, valve, shear, and shallow submergence.
Orifice diffusers--Orifice diffusers have one or more small openings for
the passage of air. Typically, the openings are 3.2-13 mm (1/8 to 1/2 inch)
in diameter. The diffusion units are generally made of molded plastic and
are either screwed or clamped to the air header pipe. Two typical types of
orifice diffusers, together with head loss characteristic curves, are shown
in Figure 12 .
Valve diffusers--Valve type diffusers contain a built-in valve that closes
when air flow is stopped. Two types of valve diffusers are shown in Figure
13. The Eimco diffuser employs a neoprene disc flapper as the valve, while
the PFT unit utilizes a plastic ball. In the latter, the air passage area is the
space between the ball and ball seat. This can be adjusted in size using
different size spacers. The main advantage of valved orifice diffusers is
their ability to prevent backflow of mixed liquor into the air header on loss
of air pressure.
Shear diffusers--Shear diffusers, pictured in Figure 14, use counterflow
of air-liquid streams to shear large air globules into smaller bubbles. Shear
diffusers are difficult to attach to removable pipe headers and are, therefore,
generally limited to-applications where they can be fixed to the bottom of the
tank.
Shallow submergence diffusers—Air diffusers are normally installed at
or near the bottom of the oxidation tank. To overcome the problem of inject-
ing air at those depths, the Inka aeration system, shown in Figure 15, was
developed. This system employs a series of perforated pipes located about
0.9m (three feet) below the water surface. A baffle is installed to force the
desired circulation pattern. The major advantage of this system is the low
air pressure required. Thus, fans may be substituted for more costly blowers
25
-------
AIR HEADER
CHICAGO PUMP COMPANY
"DEFLECTOFUSER"
12
-j 10
O
O
in
O
ki
1
( inches of water ) (0.249) k Pa
(feet) (0.305) metres
(inches) (2.54 X I O"2) metres
(cfm)(4.72X \0
,-2,
m3/s
PLUNGE PIPE
WALKER PROCESS
"MONOSPARJ"
CHICAGO PUMP CO.
"DEFLECTOFUSER"
13' SUBMERGENCE
15/32" ORIFICE
WALKER PROCESS
"MONOSPARJ "
HEAD LOSS LIMITED
TO 5" W.C. BY PLUNGE
PIPE LENGTH
02 4 68 10 12 14 16
AIR FLOW PER DIFFUSER, cfm
Figure 12. Characteristics of orifice-type coarse bubble diffusers (1)
26
-------
NEOPRENE
FLAPPER
PLASTIC
BALL
PLASTIC BODY
SPACER
E I MCO
PFT VALVED
ORIFICE
inches of water) (0.249) - kPa
(cfm )(4.72 X lo" ) m3/s
PFT VALVED
ORIFICE
8 10 12 14 16 18 20
AIR FLOW PER DtFFUSER. cfm
Figure 13. Characteristics of valve-type coarse bubble diffusers (1)
27
-------
CONTROL ORIFICE-]
AIR
/•:•• ••-••;
FLOOR OF OXIDATION TANK
O. 18m (7")
Figure 14. Typical shear-type coarse bubble diffuser (1) .
Fine Bubble Diffusers--
Aeration by fine bubble diffusers is characterized by the production of
very small air bubbles, which impart a gentle movement to the mixed liquor
in lieu of the turbulence associated with coarse bubble diffusers. Diffusion
devices in fine bubble diffuser plants are invariably of the porous ceramic
type, composed of silicon dioxide or aluminum oxide grains held in a porous
mass with a ceramic binder. Porous diffusers are available in a variety of
shapes, with the most common being flat plates, hollow tubes, and domes.
Head loss characteristics of all shapes of porous diffusers are comparable,
but oxygen transfer efficiencies are not necessarily the same due to a num-
ber of variables, particularly bubble coalescence, which is a function of
diffuser shape.
Porous plates are rated on the basis of "permeability", or the volume of
air in cubic feet per minute which is passed through 9.3 dm2 (one square
foot) of diffuser 2.5 cm (one inch) thick when tested dry at a 500 Pa (two-
inch water column) differential pressure under standard conditions of temp-
erature, pressure, and humidity. Permeabilities commonly range from 9,37
dmVminute (20-80 cfm/minute) .
28
-------
r— AIR HEADER
DIFFUSER SUBMERGENCE , 0.9m (3 ft)
FLOW
L— BAFFLE
I— CIRCULATION PATTERN
Figure 15. Shallow submergence aeration system (1)
Other types of fine bubble diffusers include Saran wrapped tubes and
sock-type diffusers. Saran wrapped tubes are 0.61 m (24 inches) long by
76 mm (3 inches) in diameter and are formed by tightly wrapping Saran cord
around a corrugated stainless steel core. An integral control orifice is pro-
vided and the bubble obtained is about the same size as that from a 19 dm^/
minute (40 cfm/minute) permeability ceramic tube. The sock-type diffuser
was developed in an attempt to minimize clogging problems. The sock is
made of plastic fabric and fitted over an opening in an air header, which is
equipped with a control orifice. When air is introduced, the sock fills out.
The flexing action is designed to reduce clogging.
Diffuser Arrangement--
The first question to answer in selecting a good diffuser arrangement is
whether the operator must remove the diffusers without emptying the tank.
All diffusers require maintenance, particularly fine bubble diffusers. If oxi-
dation tanks cannot be taken out of service, the diffusers should be easily
removable for maintenance outside the tank. In large plants with multiple
tanks, fixed diffusers are acceptable where it is possible to drain a tank for
29
-------
maintenance without impairing plant treatment efficiency. However in small
plants, where only one or two tanks are provided, diffusers should be remov-
able while the tanks are in service.
If diffusers must be removable while the tank is in service, flat.porous
plates cannot be used and the diffusers must be installed on pipe headers
arranged in either a spiral-roll or cross-roll pattern (Figure 16) . Either
coarse or fine bubble diffusers can be installed in this manner providing they
can be easily attached to the pipe header. Spiral-roll patterns are less de-
sirable because of short circuiting and inferior oxygen transfer performance.
These problems are partially overcome in the cross-roll configuration, which
provides closely spaced air curtains, through which the mixed liquor must
flow.
If fixed diffusers are used, a large variety of diffuser arrangements are
available. Therefore, traditional configurations, such as spiral or cross-roll
patterns, should be avoided in favor of more efficient arrangements. For ex-
ample, recent experimental work at Milwaukee (15,16) and Melbourne
(Aberley, R. C., Rattray, G. B., and Douglas, P.P., "Design and Perform-
ance Testing of Air Diffuser Units for Melbourne's South Eastern Purification
Plant," unpublished paper, November 1972) has demonstrated that flat porous
plates arranged over the entire tank bottom provide a much higher oxygen
transfer capability than any other arrangement. The Milwaukee experiments
even demonstrated that the old "ridge and furrow" diffuser pattern is more
than twice as efficient as the modern spiral-roll arrangement.
At Milwaukee (15,16), full scale in situ oxygen transfer studies were
made of seven different arrangements of fine bubble diffusers. The results
of these tests are given below:
Oxygen transfer efficiency
Diffuser pattern 95% confidence range, percent
Porous tubes, spiral-roll 3.3 - 9.1
Porous tubes, cross-roll 4.7 - 7.7
Porous plates, ridge and furrow 8.0 - 26.2
Porous plates, crows foot 8.6 - 10.4
Porous plates, modified crows foot 9.3 - 11.5
Porous plates, transverse placement 10.3 - 12.3
Porous plates, longitudinal placement 11.4 - 14.4
This experiment was valuable in pointing out the importance of diffuser place-
ment and it amply demonstrated that the placement of diffusers over the entire
tank floor is the most efficient arrangement from the standpoint of oxygen
transfer.
30
-------
.SWING DIFFUSER ASSEMBLY
SHOWN RAISED FOR MAINTENANCE
SPIRAL ROLL PATTERN
AIR HEADER
CIRCULATION PATTERN
CIRCULATION PATTERN
CROSS ROLL PATTERN
Figure 16. Spiral and cross roll diffuser arrangements (1)
31
-------
The use of flat porous plates over the entire oxidation tank bottom as
shown in Figures 17 and 18 eliminates tank geometry restrictions imposed by
conventional spiral-roll oxidation tanks. Aberley (1) has shown that 0.36 m
(14 inch) diameter porous plates, mounted flush with the tank bottom, give
excellent performance at the South Eastern Purification Plant in Melbourne,
Australia. A full-scale plenum unit was tested in a 4.6 m (15 foot) deep test
tank in tap water with the results shown in Figures 18 and 19. Oxygen trans-
fer efficiencies are corrected to standard conditions and indicate that the
diffuser design is extremely efficient.
Deep Submergence Aeration
Primarily because the need to place diffusers near the tank bottom im-
poses economic limitations on blower selection, oxidation tanks have tradi-
tionally been restricted to a maximum depth of about 4.6m (15 feet) . Aeration
at greater depths is desirable since greater oxygen transfer efficiency is
possible with increased diffuser submergence (see Figure 19) . With this
incentive, Walker Process Equipment Company has developed a system,
based on the airlift principle, to aerate deep tanks with medium pressure
(41 kPa) (6 psi) blowers (30). Conventional circular eductor tubes used for
air or gas lifting applications become increasingly inefficient as their size
increases, because in large (up to 2 . 1 m) (7 foot diameter) tubes, the air-
lifted mass in the center possesses no more energy than that at the lip of the
tube. Consequently, the flow at the center has difficulty leaving the tube,
while at the same time, its entrained air exhausts vertically, leaving the
center portion of the mass virtually as an overburden. Tests have shown that
airlifts remain efficient as long as the outflow is no further than 0.6 m (2 feet)
in any direction. Thus, circular tubes larger than 1.2 m (4 feet) in diameter
are less efficient than smaller tubes. Utilizing that principal, Walker Process
designed a rectangular tube 1.1-1.5 m (3.5 to 5.0 feet) in width which can
be as long as tank geometry requires, without loss of efficiency. Coarse
bubble diffusers are installed at about middepth of the tube. In this manner,
tanks up to 9.1 m (30 feet) deep can be efficiently aerated. The advantages
are in greater oxygen transfer efficiency, reduced tank plan dimensions, and
more economical tank construction. The manufacturer reports oxygen trans-
fer efficiency in tap water at standard conditions to be from 12 to 14 percent.
Figure 20 shows details of a typical deep submergence aeration system.
Aeration Blower Systems
An integral part of aeration systems is the air supply network, consisting
of filters, blowers and the necessary ductwork, valving and controls to effec-
tively distribute compressed air. Filters remove minute dust particles (aero-
sols) from the air flow, mainly to prevent clogging of the diffuser plates.
Blowers compress the air, raising it to sufficient pressure to overcome hydro-
static head in the aeration tanks and losses in the delivery system. Air for
32
-------
15 em (6') MANUALLY
CONTROLLED
BUTTERFLY VALVE
O.6lm (24 )
PRECAST
CONCRETE
BEAM
MULTIPLE PORT SETTLED
SEWAGE GATE
EXPANSION JOINT
ASSEMBLY
TILT-UP
CONCRETE
HULL
SETTLED SEWAGE CHANNEL
TYPICAL PftfCAST
CONCRETE FL£NUH UNIT
CONCRETE
INFILL
LOCATION KEY
Figure 17
Aeration system - South Eastern Purification Plant,
Melbourne, Australia (1).
33
-------
wovr
14" DIAMETER x 1.25" THICK
POROUS PLATE DIFFUSER^
OXIDATION TANK BASE SLAB
TYPICAL PLENUM UNIT
.2
(inches)(2.54X IO )-metres
Ul
CD
ce
Lu
h-
I
CO
CO
O
-j
LU
(inches of water)(O.24
- fcfm )(4.72 X /O'2) - m3/
(feet)(O.3O5) - metres
X
/•^
9) -- kt
'5
x1
°a
(
^— !
—
012345678
AIR FLOW, cfm/PLATE
HEAD LOSS THROUGH DIFFUSER PLATES WITH
DRY PLENUM AT 10 FT DIFFUSER SUBMERGENCE
Figure 18. Characteristics of porous plate diffuser at the South Eastern
Purification Plant, Melbourne, Australia (1).
-------
-2 3
(cfm)(4.72 X IO ) - m /s
(feet)(O.3O5) - metres
5 7.5 IO 12.5
WATER DEPTH, FEET
Figure 19. Oxygen transfer efficiency of diffusers at various flow rates
and submergence (1) .
35
-------
RECTANGULAR OR CIRCULAR TANK
AIR MAIN
AMPLIFICATION BAFFLES (SURFACE)
Figure 20.
— zzr -^^^^
~r
Details of Walker Process
Equipment Company deep
submergence diffusion
system (1) .
J
/
X
f
f
w-
i
o
* •
a"
(
mi1
A
-\
\
3.8 m
(12.*.
? ft)
DE
6.1
(2O-
\
V
"> --
PTH
- 9.1 m
3Oft)
36
-------
the activated sludge process has been furnished by almost every type of
compressor developed to this date (10). In the past, a distinction was made
between blowers and compressors relative to a discharge pressure of 276 kPa
(40 psig) . Modern applications over a wide range of pressures have made
this distinction meaningless (29). Accordingly, the terms blowers and com-
pressors are used interchangeably at the low discharge pressures encountered
in sewage treatment plant aeration systems. However, the Compressed Air
and Gas Institute recognizes compressor as being a more universally applied
term. Three types of blowers or compressors commonly used in activated
sludge plants are (a) positive displacement, (b) centrifugal, and (c) axial.
Table 1 summarizes application information and polytropic efficiencies of
various types of blowers used in sewage plant aeration systems. Polytropic
efficiency indicated in Table 1 is defined as: the ratio of the polytropic
compression energy transferred to the gas to the actual energy transferred
to the gas, where compression energy is expressed in the form PVn = a
constant, and P is pressure, V is volume and n is the polytropic exponent.
The polytropic exponent applies to the actual process, including heat trans-
fer and friction, in contrast to the isentropic exponent, k, which applies to
the ideal, frictionless adiabatic process (29).
In the application of a blower to any practical situation, it is essential
to consider the characteristics of the load, as well as those of the blower.
Pressure required to force a quantity of air through any system is dependent
on two main components: static head and frictional resistance. In the case
of the aerator system, static head is imposed by the depth of mixed liquor
above the diffusers, while frictional resistance is imposed by the length of
pipework, bends, valves, etc., through which the air must pass and also
by the diffusers themselves. Static head is almost constant (although small
changes of liquid depth in the tank will cause minor variations), but friction-
al resistance increases according to the square of air flow. A typical pres-
sure resistance/air flow curve is shown in Figure 21. In this figure, line
C-D represents static head; curve E-F represents frictional resistance; and
the resultant of these two components is represented by curve C-G.
To explain the application of a blower to a given service, the perform-
ance curve A-B is superimposed on the demand or load curve C-G, as illus-
trated in Figure 22. With this combination of blower and load, the capacity
handled will come at the intersection of the two curves, i.e., point H. This
is the only point at which the blower will operate in that system, and this
determines air flow and back pressure.
Positive Displacement Blowers--
The positive displacement blower most frequently used in diffused aera-
tion systems is the rotary, two impeller, positive displacement blower. This
blower is a constant volume, variable pressure machine. It employs two
symmetrical figure eight shaped impellers that rotate in opposite directions
37
-------
TABLE 1 APPLICATION CHART FOR BLOWERS USED IN
SEWAGE PLANT AERATION SYSTEMS3
Volume range in
standard m /s
(scfm)
Up to 7.1
(15,000)
7.1-22
(15,000-47,000)
22-47
(47,000-100,000)
47-71
(100,000-150,000)
47-94
(100 ,000-200,000)
Blower type
Positive displacement
lobe type
Multistage vertically
split centrifugal
Integral-gear, single-
stage centrifugal
Pedestal type, single-
stage centrifugal
Multistage, horizontally
split centrifugal
Pedestal type, single-
stage centrifugal
Multistage, horizontally
split centrifugal
Pedestal type, single-
stage centrifugal
Axial
Polytropic
efficiency
percent
71^78
71-77
76-79
77-80
78-80
77-80
78-80
77-80
82-83
Remarks
Low first cost
Low cost,
minimum
spaces
Intermediate
cost, more
space
Traditional
approach,
costly
Low first cost
Traditional
approach,
costly
Low first cost
High first cost,
high efficiency
Data extracted by permission from Reference 10.
See text for definition of polytropic efficiency.
C q
For low capacities such as 0.9 m/s (2,000 scfm).
in a cylinder. Compression occurs by entrapment of the air by the rotating
lobes. As each lobe of an impeller passes the blower inlet, it traps a volume
of air equal to one-fourth of the total blower displacement. Timing gears
maintain the blower impellers at minute clearances from each other. During
operation a certain amount of air,defined as slip, leaks around the impellers
back to suction. Slip is constant for a given blower at a given discharge
pressure. Thus, the blower should be operated at high speed to maximize
the volumetric efficiency.
The rotary, positive displacement blower can be capacity controlled by
(a) providing a variable speed transmission or driver, (b) using multiple
units, (c) blowoff control, and (d) recirculating discharge to suction. The
blower automatically furnishes variable pressure output since it "floats" on
38
-------
SYSTEM
CURVE
PRESSURE
RESISTANCE
FRICTIONAL
RESISTANCE
AIR FLOW
Figure 21 . System characteristic curve.
PRESSURE
AIR FLOW
Figure 22. Blower/system curves,
39
-------
the line, developing a pressure equal to the back pressure of the system.
Good installation practice calls for use of a relief valve to protect the
blower and motor from damage due to a pipeline restriction or closure (29) .
Centrifugal Blowers —
The characteristic curve of a typical centrifugal blower is illustrated in
Figure 23. This shows how blower output varies according to back pressure
exerted on the discharge side. The curve shown is for a blower driven at
constant speed and without inlet guide-vane throttling. Note that the curve
(A-B) is broken off where it becomes nearly horizontal. This is the point
where unstable "surge" conditions develop.
SURGE LIMIT
BACK
PRESSURE
BLOWER OUTPUT
Figure 23. Blower characteristic curve,
In the example given previously (Figure 22), point H represents the maxi-
mum blower output capability in that system, i.e. , with the blower running
at its rated speed and with no upstream or downstream valves throttled. If it
is wished to reduce blower output (for example, when activated sludge orga-
nisms require less air), three methods of control might be used: discharge
throttling, blower speed variation, or inlet throttling.
Discharge throttling--The effect of discharge throttling on a typical
blower is shown in Figure 24. Curve C-G represents the load curve of the
unthrottled system (as before), with point H representing blower delivery
under these conditions. Curve C-G' shows the modified load curve that
would be obtained by throttling a valve installed in the delivery pipework,
and point H1 represents blower output under these new conditions. Curve
C-G" and point H" show the effect of throttling the delivery somewhat
further.
40
-------
PRESSURE
AIR FLOW
Figure 24. Discharge throttling control,
Although discharge throttling can be used quite effectively to regulate
blower output over a limited range, this control method has two disadvantages
The first is that it is not efficient, since this artificially-created resistance
represents an irrecoverable loss of power (energy being dissipated in the dis-
charge valve), and the second is the risk of creating "surge" conditions.
If the blower's characteristic curve is studied (see Figures 23 and 24),
it will be observed that its slope decreases as back pressure is increased to
reduce blower output until, at point J, the direction of slope reverses, as
indicated in Figure 24. In practice, however, this situation is impossible
to achieve, since the blower would enter an unstable part of its operating
range, known as the "surge" zone.
If the blower is allowed to operate at a point on its curve close to point
J, a very minor variation in back pressure could cause the blower to operate
at two points on its curve—A and K'—i.e. , at two different outputs (see
Figure 24) . If this condition develops, output of the blower will fluctuate
between these limits, causing back pressure oscillations which, in turn,
cause further air flow variations. Output of the blower is virtually uncon-
trollable in this range. If allowed to continue, this surging might result in
damage to the blower.
To avoid the surge zone,, blower delivery must be kept above about 60
percent of rated output. Since discharge throttling would produce large ener-
gy losses, other control methods are usually more attractive for regulating
blower delivery.
41
-------
Blower speed variation—Another method of regulating blower output is
to vary blower speed. Effects of variable speed control on a typical centrif-
ugal blower are illustrated in Figure 25. It should be noted that minor varia-
tions in motor speed such as could be caused by slight fluctuations in volt-
age and frequency of power supply can cause noticeable variations in blower
output.
PRESSURE
100% SPEED
AIR FLOW
Figure 25. Blower speed control,
Inlet throttling—The most versatile method of controlling blower output
over wide ranges of back pressure and flow is to vary the throttling effect of
guide-vanes installed on the inlet side of the blower.
The effects of varying inlet guide-vane angles on output of a typical
centrifugal blower are illustrated in Figure 26. This method of control allows
blower output to be controlled over a much wider range (30 to 110 percent of
rated capacity) without danger of incurring surge. Two points are worth
noting in Figure 26. The first is that blower output can be varied over a wide
range, even with discharge pressure held constant. The second is that
"negative" angle of guide-vane throttling allows an increase of blower out-
put above rated capacity.
The apparent contradiction shown by the second point—that increased
(negative angle) throttling produces greater blower output—is explained by
the fact that negative throttling angles introduce a helpful "preswirl" into
the air stream just before it enters the impeller eye (positive throttling angles
introduce a swirl that is opposite in direction to that of impeller rotation,
thus reducing output). It should be understood, however, that although neg-
ative vane angle throttling allows blower output to be increased above rated
capacity, this is not without an increase in power consumption.
42
-------
D
Q_
UJ
IT
UJ
cc
Q.
UJ
o
cc
<
I
o
CO
70
60
50 -
40
30
20
10
APPROX. SURGE
POINTS (TYP)
4-90°
(kg/s)(2 21) Ib/s
(kPoHO.15) psi
POS ITION OF INLET
GUIOEVANES (TYP)
2.5
7.5 10 12.5
AIR FLOW, kg/S
15
17.5
20
Figure 26. Inlet guidevane control for centrifugal blower.
Axial Blowers —
A multistage axial blower has two or more rows of rotating vanes operat-
ing in series on a single rotor and in a single casing. The casing includes
stationary vanes that direct the gas to each succeeding row of rotating vanes .
As the name implies, air flows through the blower axially. When the
blower is operating, kinetic energy is converted to static energy; hence,
during stable operation, the volume of air handled by the blower at a given
speed is nearly constant, regardless of the discharge pressure. A typical
pressure-output curve for axial blowers is shown in Figure 27. As indicated
from the steepness of the curve, means must be provided for controlling com-
pressor output. Three methods normally used are as follows:
Constant speed with throttled suction—A butterfly valve in the suction
line is position-regulated to control the compressor output.
Constant speed with movable inlet vanes--For large blowers , handling
air or clean gas, regulation is accomplished by a "prerotation" device
43
-------
UJ
CO
en
LJ
cr
0.
o
<
CO
LINES OF
CONSTANT STATOR
POSITION
BLOWER OUTPUT
Figure 27. Axial blower characteristic curve.
consisting essentially of a system of movable vanes at the eye of the com-
pressor impellers. By moving these vanes, the flow of air at the impeller
eye is directed toward the direction of rotation, which has the effect of re-
ducing the pressure rise through the impeller without the losses accompany-
ing throttling of the inlet.
44
-------
Variable speed drive—The range of speed adjustment depends upon the
characteristics of the drive unit.
Surge Control of Dynamic Blowers--
All dynamic blowers (machines in which gas is compressed by the mechan-
ical action of rotating vanes or impellers imparting velocity and pressure to
the flowing medium (29)) have a minimum flow point below which performance
is unstable. If the output of a dynamic blower is decreased to a minimum
point, a condition is reached where two flow rates can be associated with
the same pressure, thereby causing instability. This instability or surge
limit comprises pulsations in pressure and flow which may cause compressor
damage. The surge limit is affected by (a) type of compressor, (b) design
pressure ratio, (c) characteristics of the gas handled, and (d) speed.
Surge is prevented by maintaining a flow greater than the safe minimum.
This can be accomplished by blowing off or recycling excess flow by auto-
matically opening a bypass or recirculation valve. Figures 28 and 29 repre-
sent typical performance characteristic curves for an axial and centrifugal
blower, respectively. These curves demonstrate that, at a given speed, an
axial blower approximates a constant volume, variable pressure characteristic,
while a centrifugal blower approximates a constant pressure, variable volume
characteristic. Thus, axial blower antisurge control should be pressure ori-
ented, while centrifugal blower surge control should be flow oriented (29).
SURGE CONTROLLER
SET POINT
MAX SPEED
FLOW
Figure 28. Typical performance characteristic curve for an axial
blower with variable speed control (29) .
45
-------
SURGE CONTROLLER
Q
FLOW
Figure 29. Typical performance characteristic curve for a
centrifugal blower with variable speed control (29)
A surge controller is employed in surge control systems to limit the dis-
charge pressure of axial blowers and maintain a minimum flow for centrifugal
blowers. The surge controller compares the values of two sides of an equa-
tion which defines the surge limit condition. If the equation is unbalanced
in the "safe" direction, the controller keeps the bypass valve closed. If the
equation approaches balance, i.e., the surge line, the controller must start
to open the bypass valve and also prevent the equation from becoming un-
balanced in the surge direction. The controller throttles the bypass valve to
various positions to maintain the system slightly on the safe side of surge.
An auxiliary device lets the controller catch the system without a temporary
overshoot in surge conditions, despite the presence of the integral control
mode. This overshoot or anticontroller wind-up feature must always be used
with the surge controller and is readily available from all instrument manu-
facturers in both pneumatic and electronic models.
The surge line equation depends on actual blower characteristics and is
quite different for variable speed operation than for inlet throttling. The
surge controller must keep the blower from operating to the left of the surge
line, shown in Figures 28 and 29, regardless of speed, flow, pressure, and
inlet gas temperature. The surge line is parabolic when plotted on a graph
of discharge pressure versus volumetric flow. At constant gas specific
gravity, the surge line is defined by equation (7), taken from Reference 5:
46
-------
f(T)
DP
= KH
1
(7)
where f(T)
DP
H,
1
K
= a computed function of the inlet gas temperature
= the differential pressure across the blower
= the flowmeter differential
= a constant
Rather than trying to calculate the surge line theoretically, it is much
more satisfactory to obtain the actual performance test curves from the manu-
facturer after he has tested the blower. Often, many simplifications can be
made in the instrumentation system to obtain a good approximation to the
surge line.
Typical pressure oriented surge control systems for suction throttled and
variable speed blowers are shown in Figures 30 and 31. In Figure 30, surge
controller PIC maintains the discharge pressure below a set point by opening
the relief valve as required. The surge controller set point is modified
by a signal from the inlet pressure transmitter, PT1 and temperature trans-
mitter TT through characterizing relay FY. In Figure 31, surge controller PIC
functions in a similar manner as in Figure 30, with the exception that the
controller set point is modified by speed transmitter ST as well as suction
temperature and pressure.
RELIEF
VALVE
AIR
Figure 30. Pressure oriented surge control system for suction
throttled blower (29) .
47
-------
RELIEF
VALVE
k VARIABLE
SPEED DRIVE
Figure 31 . Pressure oriented surge control system for variable
speed blower (29) .
Figure 32 shows a basic flow oriented surge control system suitable for
a centrifugal blower. Referring to Figure 29, surge controller FIG will main-
tain the discharge flow above the set point (dotted line) by opening the relief
valve as required. No set point modification with speed is required with dis-
charge flow monitoring since the flow element differential pressure remains
essentially constant for speed variations.
Various surge control system arrangements result in different operating
ranges available to the compressor. Figure 33 illustrates a surge control
system that maximizes the operating range for a suction throttled compressor.
Surge controller UIC maintains a minimum DP/P2, where DP is the differential
pressure across the flow element and P? is the discharge pressure. Relay FY
computes the ratio of DP/P2 and the result is fed to surge controller UIC.
Since a constant value of DP/P2 can be maintained that matches the surge
line, maximum compressor operating range is available. In practice, the set
point is set just to the right of the surge line as shown in Figure 33 (29) .
Mechanical Aeration
Mechanical aerators have gained wide acceptance in the wastewater
treatment industry in the past ten years. They fall into five broad categories:
plate, updraft, downdraft, combination and brush type.
48
-------
AIR
RELIEF
VALVE
VARIABLE
SPEED DRIVE
Figure 32. Flow oriented surge control system for variable speed
blower (29) .
Plate Type—
The plate type aerator (Figure 34) employs a circular plate equipped with
radial blades and creates a large amount of turbulence by causing a peripheral
hydraulic jump. Air is entrained in the mixed liquor by surface turbulence
from the hydraulic jump and by air entrainment in a low pressure region be-
hind the rotor blades. Normally only a surface turbine plate is provided,
but for some applications a second rotor, located at midtank depth, is used.
Updraft Type —
The updraft type aerator (Figure 35) is perhaps the most popular of the
mechanical aerators. It employs a surface impeller which draws liquid up-
ward and violently outward at the surface. The surface turbulence effects
oxygen transfer. Some updraft designs include draft tubes which direct liquid
to the impeller more efficiently.
Downdraft Type--
The downdraft type aerator employs an impeller in a vertical tube to force
liquid from the surface, down through the tube to the bottom of the tank. Air
is entrained in the liquid as it is forced down into the tube.
49
-------
RELIEF
VALVE
AIR
LU
CK
D
to
(/)
UJ
(£
0.
UJ
O
tr
<
I
o
tn
SURGE LINE AND LOCUS
OF CONSTANT DP/P2
SET POINT
FLOW
Figure 33. Surge control system to maximize blower
operating range (29) .
Combination Type—
The combination type aerator (Figure 36) employs both mechanical and
diffused aeration. Usually, two impellers are provided; one at the surface
and one near the tank bottom at the point of diffused air input. Air is intro-
duced through a ring containing orifices and the turbine shears the bubbles
into smaller sizes. The surface impeller provides a high degree of oxygen
transfer by turbulent mixing.
Cage or Brush Type--
The cage or brush type aerator (Figure 37) rotates around a horizontal
shaft equipped with a series of projections. Oxygen transfer is accomplished
by surface turbulence. The brush type aerator is used extensively in oxida-
tion ditch applications because, in addition to its good oxygen transfer char-
acteristics, it is an excellent means of inducing circulation in channel-type
aeration basins .
50
-------
^
-LOWER
PLATE
OPTIONAL
Figure 34. Plate type aerator (1).
Figure 35. Updraft type aerator (1) .
51
-------
*r<-
X
f
•f IA
•o
, ->~ _
VI .-, <<> ) '?'-2
•:vv 6^
a
.• •.'. l.- . . x-s> ,--
^ :.-:;.• .L-;:.^-^ ^ -1
^ ^ ) £^±±\
S
-^
V )
AIR
Figure 36. Combination type aerator (1).
Figures?. Brush type aerator (1)
52
-------
The performance of all fixed mechanical aerators is affected by their
submergence. Therefore, careful control of the oxidation tank water surface
level is an important concern. Oxygen transfer efficiency decreases and
power increases with increasing submergence. Some manufacturers have
taken advantage of this characteristic to optimize aerator power consumption
by means of an automatically adjustable outlet weir. In such systems, the
height of the weir is varied to change the aerator submergence and hold the
mixed liquor dissolved oxygen concentration at the desired level. This has
not always been successful, however, as the aerator tends to "pump" the
fluid over the weir, causing a nonuniform elevation in the tank. Additional
information on types and applications of mechanical aerators can be found in
Reference 2 .
High Purity Oxygen
Oxygenation of mixed liquor with pure oxygen has been the subject of
experimentation for about 20 years. Only recently, however, has an attempt
been made to commercially market such a system. The Union Carbide Corp-
oration and others have been engaged for several years investigating the use
of pure oxygen aeration. The research has largely taken the form of pilot
plant studies and has culminated in the development of patented processes
such as the "Unox" (37) and "Marox" (FMC Corporation) systems. The major
advantage of the Unox system over conventional air aeration systems is an
ability to dissolve larger amounts of oxygen per unit volume of reactor.
Typically, the volume of oxygen used in the Unox system is only about
1-2 percent of the volume of air required by a comparable air aeration plant.
Because that volume is insufficient to ensure mixing and prevent settling of
solids, mechanical mixing must be employed. Oxygen dissolution systems
are designed around surface, submerged turbine or combined mixers. Figure
38 shows a schematic cross section through a typical oxygen oxidation tank
with a submerged turbine design. The tank is covered (usually with a con-
crete slab) and is divided into a number of stages by baffle walls . The liq-
uid and gas flow concurrently through the tank, with the primary effluent,
return sludge and oxygen being introduced together in the first stage. The
oxygen is introduced at a gage pressure of 0.25-1.0 kPa (1-4 inches) of water
column, and small recirculating gas compressors in each stage collect the
gas above the liquid level and discharge it through the hollow shaft of a ro-
tating sparger device, or through a separate pipe, at a rate sufficient to
maintain the required dissolved oxygen concentration.
Gas is recirculated within each stage at varying rates to meet the de-
creasing oxygen demand as the mixed liquor progresses from stage to stage.
The rate of gas flow within each stage is usually greater than the rate from
stage to stage. Effluent mixed liquor from the system is settled in the con-
ventional manner and sludge is returned to the oxidation tank.
53
-------
Cn
OXYGEN
GENERATING
PLANT
CONTROL
VALVE
r- OXIDATION TANK COVER
AGITATOR
GAS RECIRCULATION COMPRESSOR
u
RETURN ACTIVATED SLUDGE
STAGE BAFFLE
VENT GAS
MIXED
LIQUOR
ROTATING
SPARGER
FINAL
CLARIFIER
Figure 38. Submerged turbine mixer oxygen dissolution system.
-------
The entire tank is fitted with a gastight cover to contain the oxygen gas .
A restricted exhaust gas line from the final stage vents the waste gas to the
atmosphere. Due to the net dissolution of gas, the amount of gas vented
represents only 10 to 20 percent of the oxygen gas feed rate and the vented
stream is about 50 percent oxygen.
Alternatively, the mass transfer and mixing in an oxygen dissolution
system may be accomplished with surface aerators as shown in Figure 39.
The surface aerators bring quantities of the wastewater to the surface for
contact with the oxygen-rich atmosphere under the tank cover. Oxygen trans-
fer occurs through direct contact as the wastewater is sprayed through the
oxygen atmosphere and by entrainment as splashing liquid impinges into the
bulk liquid. Lower impellers, as shown in Figure 39, are sometimes required
for operation in deep basins.
OXIDATION
TANK COVER
CONTROL
VALVE -
OXYGEN
FEED GAS
PRIMARY
EFFLUENT
IMPELLER
RECYCLE
SLUDGE
VENT
GAS
MIXED LIQUOR
EFFLUENT TO
CLARIFIER
Figure 39. Surface mixer oxygen dissolution system.
Hybrid or combined surface and submerged turbine oxygen dissolution
systems have been successfully used in a number of installations. For this
design, a surface mixer is placed on a common shaft with a submerged turbine
aerator. Flexibility is achieved through changes in mixer speed and gas flow
rate.
Oxygen is usually supplied from an oxygen generating plant that forms an
essential part of the treatment system. Two types of oxygen generating
55
-------
plants are offered: (a) a cryogenic air separation process for larger plants;
and (b) a package type pressure-swing absorption (PSA) process for smaller
plants.
Operation of oxygen aeration plants is characterized by:
1. High BOD loadings per unit volume of reactor.
2. High mixed liquor volatile suspended solids concentrations.
3. High dissolved oxygen concentrations (6 to 10 ppm).
4. Denser secondary sludge.
5. Low sludge yields.
Table 2 compares typical air aeration characteristics with those of seven
pilot Unox installations (39). As indicated, detention time is 1/2 to 1/4 that
in conventional plants. Short detention time, coupled with high mixed liquor
solids concentration, results in smaller tanks, meaning less capital cost and
less land area. On the other hand, larger secondary clarifiers are normally
required to maintain the higher mixed liquor suspended solids levels in the
oxygen system.
TABLE 2 . COMPARISON OF UNOX OPERATING RESULTS WITH A
TYPICAL CONVENTIONAL AIR AERATION PLANT (39)
Item
Aeration time , hrs
(based on raw
sewage flow)
MLVSS, ppm
Loading, kg BOD5/kg
MLVSS/day
Loading, mg/m /s
RAS solids cone. ,
percent solids
Raw sewage loadings
BODs, PPm
SS, PP™
Final effluent charac-
teristics
BOD5, PPra
SS, PPm
Sludge Volume Index
Solids production,
kg/kg BODj removed
Typical air
aeration plant
4-8
1500-2500
0.3-0.4
4.6-15
0.5-1.0
150-300
200-300
10-25
10-25
100-150
0.5-0.75
Unox Pilot Studies
I
1.5
7000
0.80
39.5
3.0
220
175
22
19
36
.58
II
2.0
7000
0.57
26.9
2.0
195
180
10
10
38
.70
III
1.8
5100
0.75
33.4
1.7
215
185
22
55
62
.30
IV
1.7
4000
0.90
35.2
2.2
200
130
20
16
45
.75
V
1.8
5100
1.34
68.6
1.5
440
200
22
18
80
.48
VI
1.8
5500
0.30
17
1.6
115
100
12
20
42
.39
VII
2.2
5500
0.70
40.8
1.6
340
120
34
36
65
.45
Xote (mg/m J/s 1(5. 39) lb/1000 cf/day
56
-------
With its high oxygen transfer capabilities, the Unox system is able to
operate at high MLSS concentrations . This enables the system to readily
absorb shock organic and toxic loads, much as in a completely mixed acti-
vated sludge system (37) .
The reasons for high solids in the oxygen system are not fully understood,
although work to date indicates the primary causes are: (a) more fine parti-
cles in the oxygen plant sludge due to the higher process rates; and (b) the
high MLSS in the oxygen plant being applied to conventionally sized final
clarifiers. The latter reason is probably the most important of the two. The
present lack of information on the reasons for poorer clarification point to the
need for further research on this aspect of operation to develop more rational
design criteria for final clarifiers .
Optimum nitrification takes place within a fairly narrow pH range of 7.0
to 9.0, with pH 8.5 being the ideal. In conventional plants which are nitri-
fying, the CO2 produced by organic decomposition is continually stripped out
of the mixed liquor by the aeration air. Maintenance of reasonable pH levels
is, therefore, possible. In high purity oxygen systems utilizing closed tanks,
such stripping is not possible, with the result that CCu builds up and the pH
is lowered. In some of the Unox pilot studies, mixed liquor pH values as
low as 6.0 have been reported. Nitrification reactions are severely retarded
at such low pH, although recovery has been noted following acclimation
(Haug, R. T. and McCarty, P. L., "The Effects of High Oxygen Tension and
pH on Nitrification," unpublished paper, November, 1970). Research is be-
ing pursued to solve this problem with two possibilities under investigation.
The most promising appears to be two-stage oxygenation with air stripping of
CO? taking place in a channel connecting the two stages. The second meth-
od involves lime treatment of the raw sewage which not only raises pH, but
results in higher removal of carbonaceous BOD in primary sedimentation tanks
with a corresponding decrease in BOD load to the aeration system.
Comparison of Conventional Systems
The designer is faced with a difficult choice between mechanical and
diffused aeration. Data can be produced to demonstrate the high oxygen
transfer efficiency, the low power requirements, and the benefits to bio-
logical growth afforded by any particular system. Care must be exercised in
evaluating such data; unless one knows the conditions under which tests
were conducted, how calculations are made, and what factors have or have
not been taken into account, he should not select system A over system B
simply because system A claims a higher oxygen transfer efficiency. The
major factors that should be evaluated are: (a) aerator oxygenation efficiency,
(b) overall system economy, (c) mixing and effects on biological growths, and
(d) flexibility of operation under the required range of loadings and modes of
operation. Each of these factors is discussed below.
57
-------
Aerator Oxygenation Efficiency—
Mechanical aerators are commonly rated in terms of pounds of oxygen
transferred at standard conditions per brake hp hour. Diffusers are rated by
their oxygen transfer efficiency expressed as a percentage of oxygen supplied
The two can be approximately compared by the values given in Table 3 (21) .
TABLE 3. RELATION OF OXYGEN TRANSFER EFFICIENCY TO AERATOR
POWER EFFICIENCY (37)
Diffuser oxygen
transfer efficiency,
per cent
10
12
Aerator power
efficiency,
kg O2/kW/hr
0.75
1.12
1.49
1.87
2.25
Note (kg O2/kW/hr)(1.65) = Ib O2/bhp/hr
Overall System Economy--
Economic comparison of two aeration methods must be made on the basis
of overall system economy, taking into account all cost factors including
capital recovery, preventive maintenance requirements, operational attend-
ance, breakdown maintenance requirements, utilities and the like.
Mixing and Effects on Biological Growth--
The degree of mixing is affected by tank geometry, locations of inlets
and outlets, and the type of aeration device employed. Either diffused aera-
tion or mechanical aeration can achieve good mixing, providing the tank de-
sign is correct. The belief persists, however, that mechanical aerators are
superior to diffusers from the standpoint of mixing.
On the other hand, many engineers believe that the turbulent environ-
ment in a mechanical aeration plant is not conducive to balanced biological
growth. In a study recently conducted at a medium sized plant in England,
where half the flow goes to a diffused air plant employing fine bubble, dome-
type diffusers, and the other half goes to a mechanical aeration plant, re-
searchers identified ten types of organisms in the mechanical aeration basins
while more than 60 were found in the diffused aeration tanks. The difference
was attributed to the high degree of turbulence in the mechanical aeration
basins causing floe breakup and discouraging growth of the more fragile types
of protozoa. The ten organisms identified in the mechanical aeration basins
were all of the hardy, stalked, ciliate type.
58
-------
Other researchers have reported that the settling characteristics of
sludge from diffused air plants are superior to those from mechanical aeration
plants. This phenomenom is also attributed to the breakup of activated
sludge floe caused by the violent turbulence associated with mechanical
aeration (22,23,24). However, further work needs to be completed to define
the effects of different aeration methods upon mixing and biological growth.
Flexibility of Operation—
Flexibility of operation implies the ability of the aeration method to
efficiently cope with the entire range of loads and operational modes encoun-
tered in normal operation. For example, tapered aeration is more easily
achieved with a diffused air system than with mechanical aerators. Where
a high degree of nitrification is required, diffused air is again the better
choice because such a system can more easily and economically cope with
the large daily variations in ammonia concentration which are experienced
in most plants .
Of the available aeration devices, mechanical surface aerators are least
well suited to nitrification applications because they are normally designed
to operate at fixed speed and, therefore, must overaerate most of the day to
satisfy peak oxygen demands. Even when designed for variable blade sub-
mergence, the units are limited to matching less than a 2:1 variation in load.
When variable submergence is coupled with a two-speed drive, less than a
3:1 load variation can be matched. Therefore, unless flow equalization is
provided somewhere in the system, mechanical surface aerators cannot
match variations in nitrogen loads without overaerating the mixed liquor dur-
ing a significant portion of the day.
Diffused air aeration presents a different picture. Air rates can be easily
modulated to closely match the load by turning down or shutting off individual
blowers. Thus, the diurnal load variations can be matched without the neces-
sity of overaerating the mixed liquor and wasting power. Fine bubble diffus-
ers can be arranged across the tank floor, allowing fairly even distribution of
energy input. Gentler mixing is provided than with mechanical aeration
plants, providing less tendency for floe breakup.
Submerged turbine aeration systems are intermediate in terms of their
responsiveness to the problem of aeration in nitrification systems. Because
of their capability to vary the air rate to the sparger, they may be designed
to match the load variation in oxygen demand. A drawback, however, is that
the impeller normally operates at fixed speed, imparting no turndown capa-
bility for a significant part of the power draw.
59
-------
Comparison of oxygenation capacity of various devices is possible on a
quantitative basis. Comparison of other aspects of operation, particularly
those related to mixing and the effects on biological growth, is difficult, if
not impossible, because of the subjective nature of such comparisons. Each
case must therefore be decided upon its own merits. In some cases, diffused
air will be an obvious choice, whereas in others, mechanical aeration may
be appropriately selected. As a guide, the broad advantages and disadvan-
tages of the various types of aeration devices are summarized in Table 4 (1).
60
-------
TABLE 4. SUMMARY OF AERATOR CHARACTERISTICS RELATED TO
ACTIVATED SLUDGE AERATION
Aerator Type
Advantages
Disadvantages
Best Use
Coarse bubble diffuser
Low capital cost
Non clogging; air filtration
not required
Low maintenance cost
Flexibility in operation -
can put air where load is .
Turbulent mixing
Central blower station possible
Low oxygen transfer efficiency
Turbulence promotes floe breakup.
Small municipal waste
treatment plants and
areas where dirty air
creates filtration
difficulties
Fine bubble diffuser
CTi
High oxygen efficiency
Flexibility in operation-
can put air where load is .
Flexibility in tank design
Good mixing
Central blower station possible
Gentle stirring promotes floe
formation
Diffuser clogging; requires high
degree of air filtration
High capital cost
Not suitable for completely mixed
systems
Large municipal plants.
Plants requiring low
effluent BOD. Plants
requiring high degree of
nitrification.
Mechanical aerators -
Updraft and Turbine
Types
Low capital cost
High oxygen transfer efficiency
Good mixing
Multiple units require large number
of operating drives and high
maintenance and operation cost
Insufficient oxygenation capacity
for nitrification
Icing problems in cold climates
Turbulence promotes floe breakup
Circulation patterns produce uneven
D.O. distribution
Water level control critical
Small plants where nitri-
fication is not required
Warm climates.
Mechanical aerators
Combination Type
Good mixing
Moderate oxygen transfer effic-
iency
High oxygen input capacity
Wide oxygen input range
Requires both mechanical drives
and central compressor station
High capital cost
High maintenance cost
Turbulence promotes floe breakup
Plants with wide range
of oxygenation require-
ments (trade wastes)
Nitrification tanks
Mechanical aerators
Brush Type
High oxygen transfer efficiency
Low capital cost
Low maintenance cost
Restrictive tank design flexibility
Icing Problems in cold climates
Water level control critical
Insufficient oxygenation capacity
for nitrification
Turbulence promotes floe breakup
Oxidation ditches
-------
SECTION 6
DESIGN OF DISSOLVED OXYGEN CONTROL SYSTEMS
Control of air and oxygen dissolution in the mixed liquor is an important
parameter in the activated sludge process. The desired strategy is to add
sufficient air or oxygen to meet the time-varying oxygen demand of the mixed
liquor. Because electrical energy is one of the major operating costs of the
activated sludge process, there is an economic incentive to minimize un-
necessary oxygenation.
If the DO level drops below approximately 0.5 ppm, oxygen becomes
rate limiting and the aerobic bacteria become inactive. On the other hand,
a DO level that is too high represents wasted power and, according to Ryder
(32), can cause sludge bulking. Nitrification rates are limited at DO levels
of less than 2.0 ppm; thus, DO control should be provided when nitrification
is desired.
At the present time, DO is manually controlled in most activated sludge
processes. The operator may attempt to pace oxygen transfer in proportion
to the oxygen demand, but to ensure adequate oxygenation, the operator
generally maintains an excess level of oxygen. In doing so, he usually pro-
vides more aeration than is required. Power costs for oxygenation can be
minimized if oxygen transfer capacity is automatically paced in proportion to
the time-varying oxygen demand.
Most existing activated sludge plants use air as an oxygen source. The
preferred method of air flow control is to maintain the minimum DO level in
each oxidation tank or tank pass that is necessary for adequate oxygenation.
In the recently developed high purity oxygen process, the oxidation tanks are
covered and, thus, function as on-line respirometers. Oxygen is supplied to
the first stage of each oxidation tank in proportion to the oxygen uptake rate,
as reflected by changes in oxidation tank gas pressure, and is exhausted as
vent gas from the last stage of each tank. The oxygen purity control system
maintains constant gas purity in the gas venting from each oxidation tank by
modulation of the vent gas valve. The system is usually designed to con-
sume 90 percent of the oxygen supplied. However, the pure oxygen systems
do not directly provide DO control of the mixed liquor in the oxidation tanks .
62
-------
PROCESS CONTROL FUNDAMENTALS
There are many ways of implementing control from a hardware viewpoint,
including mechanical, pneumatic, electrical, electronic, analog and digital
techniques. However, regardless of the mechanism used, the basic theory
remains the same.
There are two basic types of control: open-loop and closed-loop. The
two main forms of closed-loop control are feedback and feedforward. A brief
description of each control type is given below.
Open-Loop Control
Figure 40 represents a typical open-loop control system. Open-loop
control involves making an estimate of the form or quantity of action
necessary to accomplish a desired objective. No check is made to determine
whether or not the corrective action taken has accomplished the desired ob-
jective. In Figure 40, the fixed program is a time-related estimate of the
amount of air required to satisfy the oxygen demand of the primary effluent
over a 24-hour period.
PRIMARY
EFFLUENT
ACTIVATED
SLUDGE
PROCESS
MIXED
LIQUOR
FIXED
PROGRAM
A
AIR
Figure 40. Open-loop control
63
-------
Closed-Loop Feedback Control
Figure 41 represents a typical closed-loop feedback control system. The
controlled variable is the DO level in the mixed liquor. A measurement of
the DO level is transmitted to an automatic controller, where it is compared
with a set point or reference point. If a difference, or error, exists between
the actual and desired DO level, the automatic controller adjusts the position
of the air control valve, thereby modifying the effect of the manipulated vari-
able (air) on the controlled variable (DO) and eliminating the error between
the actual and desired DO level.
PRIMARY OXIDATION
EFFLUENT
I
MANIPULATED
TANK
i i— r>nn
0 I Z
CESS Q I 0
f
MIXED
LIQUOR
- CONTROLLED
VARIABLE
A~i
CONTROL f
VALVE —'
A
AIR
FEEDBACK
CONTROLLER
T
DO SET POINT
Figure 41 . Closed-loop feedback control.
Closed-Loop Feedforward Control
Figure 42 represents a typical closed-loop feedforward control system.
One or more measurements related to the oxygen demand of the primary ef-
fluent, together with the DO set point, are used to compute the correct
amount of air to meet the current oxygen demand of the process. Whenever
a change in oxygen demand (disturbance) occurs, corrective action starts
immediately to cancel the disturbance before it affects the controlled vari-
able (DO level in mixed liquor) . Feedforward control is theoretically cap-
able of perfect control; its performance is only limited by the accuracy of
the measurements and computations.
64
-------
PRIMARY
EFFLUENT
i
LL
.1
SIGNAL
TOC
) 1
SIGNAL
1
\
OXIDATION A MIXED
TANK
t
s\ •-
Z LIQUOR
.^FE THAT THE
CONTROLLED VARIABLE,
DO, IS NOT MEASURED.
FEEDFORWARD
CONTROLLER
DO SET POINT
A
AIR
Figure 42. Closed-loop feedforward control
The essential feature of feedforward control is the forward flow of in-
formation. The controlled variable is not used by the system, because this
would constitute feedback; this point is important, because it shows how it
is possible to control a variable influent with a continuous measurement of
it available. A set point is essential, however, because any control system
needs a "command" to give it direction.
Feedforward-Feedback Control
The only serious failing of feedforward control is its dependency on
accuracy. To provide perfect control, a system must model the plant ex-
actly; otherwise, whatever error that may exist in positioning the manipulated
variable causes offset. Therefore, if the controlled variable can be measured,
as is the case in the feedforward-feedback control system shown in Figure 43,
a feedback controller is coupled to the feedforward controller to compensate
for any input changes that have not been taken into consideration in the feed-
forward model and for inaccuracies in the various transmitters and computing
elements in the system. In general, the feedback controller is used to adjust
65
-------
PRIMARY
EFFLUENT ' "
1 1
5L L
|l| °l|
""I*0 lw
1 1
FEEDFORWARD
CONTROLLER
t
1
1
1
^/1
— *
-------
employing pure oxygen for oxygenation of the mixed liquor. The two indirect
methods are (a) influent flow and (b) oxidation-reduction potential (ORP) of
the mixed liquor. Each of the four methods for the continuous measurement or
computation of the dissolved oxygen of mixed liquor is described below.
Electrochemical Dissolved Oxygen Sensors
Dissolved oxygen sensors of the electrochemical type are suitable for
continuous, in situ,DO measurement. All electrochemical DO sensors are
affected by temperature, velocity, ionic strength and other environmental
factors. Currently, three types of DO sensors are commercially available,
and they operate on the following principles:
0 A galvanic sensor in which molecular oxygen diffuses through a
membrane and reacts with a lead/silver electrode system to produce
a current proportional to the DO concentration.
0 A polarographic cell that requires oxygen to diffuse through a
membrane (typically Teflon) and be reduced by a polarizing voltage
across metal electrodes of different nobility. This cell produces a
current proportional to the DO concentration.
0 A thallium cell in which oxygen reacts with thallium metal, producing
thallous ions in proportion to the DO concentration. The potential
developed is a function of thallous ions at the surface of the elec-
trode but needs no membrane. However, caution should be exercised
in the use of this electrode since thallium is injurious to health.
Membraned probes have been developed to avoid the problems encoun-
tered when electrodes contact the sample. For example, the membrane ex-
cludes most materials that could cause an erroneous output since they cannot
diffuse through it. Electrolyte strength is also preserved since it is isolated
from the sample by the membrane.
Oxygen Uptake Rate in a Closed Oxidation Tank
When a closed oxidation tank is used in the oxygen activated sludge
process, it is possible to determine oxygen uptake of the mixed liquor in a
given oxidation tank from the oxygen purity of the gas being vented from the
oxidation tank. An increase in the oxygen uptake causes a reduction in the
oxygen purity by opening the vent gas valve which causes a drop in oxidation
tank gas pressure, which, in turn, causes additional oxygen to be added to
the oxidation tank to restore the gas pressure to the desired level. Thus,
oxygen is added to match the time-varying oxygen demand of the mixed liquor.
67
-------
Influent Wastewater Flow as a Dissolved Oxygen Control Parameter
An indirect method for DO control in the activated sludge process is
based upon maintaining a constant ratio between influent wastewater flow
and air flow. In ratioing air to mixed liquor in the oxidation tank, it is cus-
tomary to maintain the ratio at 4-11 cubic metres of free air per cubic metre
(0.5-1.5 scfm/gallon) of wastewater treated. Present practice in many treat-
ment plants involves manual adjustment of air feed to maintain the desired
ratio. Since the wastewater flow through a treatment plant will undergo large
fluctuations in the course of a 24-hour day, considerable plant operator at-
tention must be devoted to maintaining proper air/wastewater flow ratios.
In plants where 24-hour supervision is impossible, operating procedure
usually consists of increasing the air flow in the morning and decreasing
it at night. Under such an operational procedure, changes in wastewater
flow will result in either over or underaeration, accompanied by improper
treatment. The main disadvantage of flow ratio control is that the oxygen
demand per unit volume of fluid varies considerably in wastewater. Conse-
quently, it is recommended that DO control systems based on influent waste-
water flow be utilized only for wastewaters with a relatively uniform diurnal
oxygen demand.
Oxidation-Reduction Potential as a Dissolved Oxygen Control Parameter
Oxidation-reduction potential (ORP) is the potential between the
oxidants and reductants in a system without regard for the total quantity
of either constituent or their biological activity. Biological activity consid-
erations are important since inert oxidized salts may be recorded as oxidants,
although their effect on biological reactions will be minimal. Therefore, a
high ORP can develop even though the biological reaction is predominantly
reductive in nature. Although it may be possible to correlate the ORP of any
given mixed liquor to its oxygen demand, it is not possible to correlate the
ORP's of mixed liquors in different treatment plants.
INSTALLATION, APPLICATION AND CALIBRATION OF DISSOLVED OXYGEN
PROBES
All of the control systems described in this paper for diffused air and
mechanical aeration systems employ,in situ, electrochemical dissolved
oxygen (DO) probes for the measurement of mixed liquor DO levels. Accord-
ingly, some guidelines concerning the application and installation of DO
probes are given prior to discussing various aeration methods and associated
DO control systems.
68
-------
Installation and Application
The location points for DO probes will vary from plant to plant as well
as for different operating configurations. Since it is difficult for the design
engineer to predetermine the most suitable DO probe locations, any DO
probe layout should be flexible enough to permit reconfiguration while the
plant is operating. Multiple DO probe receptacles should be provided in
each tank to permit probe relocation. Typical DO probe and receptacle loca-
tions for various process configurations are given in Figures 44 through 48.
For reliability, it is recommended that a DO analyzer/transmitter be pro-
vided for each DO probe that is employed for control purposes. However,
for monitoring purposes, it is acceptable to use one analyzer and multiplex
the outputs from two or more DO probes as shown in Figure 49.
DO probes are designed to be placed in sites where the transition time
of the sample to the probe is essentially zero. However, if it is desired to
locate the DO probe remote from the oxidation tank and pump a sample to the
probe, the elapsed sample transfer time must be considered. In particular,
if mixed liquor from an activated sludge process is to be transferred, the up-
take rate of oxygen by the biological mass in the mixed liquor will decrease
the DO level in the sample by the time it reaches the DO probe. For example
assuming an uptake rate of 600 grams of oxygen per cubic metre per hour, the
loss of oxygen would be 10 grams per cubic metre per minute. Thus, it can
be seen that sample transfer should be rapid to minimize the change in DO
concentration. However, sample transfer devices may cause reaeration.
Consequently, unless the exact relationships are known, the value of sample
and transport for DO measurements is questionable. It is recommended that
DO probes be installed directly in the mixed liquor whenever possible.
Mounting
DO probe mounting requires careful consideration. The mounting struc-
ture must be sturdy enough to withstand continuous buffeting resulting from
air agitation of the mixed liquor. In addition, the mounting assembly must
permit easy removal of the probe assembly from the mixed liquor for routine
maintenance and calibration. Figure 50 shows a typical installed dissolved
oxygen probe assembly.
69
-------
-DO PROBE RECEPTACLE(Typ)
INFLUENT
OXIDATION
TANK
DO PROBE
EFFLUENT
Figure 44. Plug flow, single pass oxidation tank - probe/receptacle locations.
INFLUENT
EFFLUENT
DO PROBE (Typ)
PASS I
PASS 2
PASS 3
PASS 4
DO PROBE
RECEPTACLE(Typ)
OXIDATION TANK
Figure 45. Plug flow, multiple pass oxidation tank - probe/receptable locations
70
-------
DO PROBE RECEPTACLE (Typ)
INFLUENT
RAS
J 1 J
DO PROBE
EFFLUENT
Figure 46. Complete mixed process oxidation tank
probe/receptacle locations.
AERATOR
NFLUENT
o
OXIDATION TANK
EFFLUENT
Figure 47. Single aerator installation - probe location.
71
-------
-DO PROBE (Typ)
INFLUENT
DO PROBE RECEPTACLE(Typ)
O
771 i xx Pol V
EFFLUENT
-AERATOR (Typ)
Figure 48. Multiple aerator installation - oxidation tank
probe/receptacle locations.
4-CHANNEL
t
_L
'AIT'
e~-
4
-L
HS
INFLUENT
I I
DO /-k DO /-k DO /-k DO
(AE2J (AE3J (AE4)
OXIDATION TANK
EFFLUENT
Figure 49. Multipoint dissolved oxygen monitoring system using
a single, multiplexed analyzer.
72
-------
BRACKET
RECEPTACLE
CONDUIT TO
ANALYZER
(AIT)
PIPE HANGER
^—SPACER TO CTR ROD
(TYP 2)
1.9cm (3/4 in)
SCH 4OS PIPE
— LOWER PIPE
HANGER
PROVIDE SUFFICIENT
FLEX TO PERMIT
RAISING ELECTRODE
ASSEMBLY TO LOWER
PIPE HANGER
SUBMERSION TYPE
ELECTRODE ASSEMBLY
Figure 50. Dissolved oxygen probe assembly.
73
-------
Calibration
The recommended procedure for calibration of a DO probe is to remove
the probe from the mixed liquor, rinse it off to remove accumulations of sol-
ids and then place it in a bucket of tap water. The DO concentration in the
water is previously determined by Winkler titration. For convenience, the
sample of water should be slightly warmer than ambient temperature and
nearly saturated with oxygen so that the DO level does not change signifi-
cantly during the calibration procedure.
A second calibration procedure called "air calibration" is to remove the
probe from the mixed liquor and permit it to stand exposed to the air for a
period of from 5 to 15 minutes. The probe should be sheltered from the sun or
rain. During the 5-15 minute period, the membrane will become dry and the
probe assembly will approach ambient temperature. Under these conditions,
the output from the probe is equivalent to saturated water at the ambient
temperature since the partial pressure of oxygen in air is equivalent to the
partial pressure of oxygen in saturated water. The ambient temperature can
be read by switching the selector knob to the temperature position. Once
the temperature is determined, the saturation value can be read from a table
of DO values for pure water at saturation, and the calibrated output from the
DO monitoring system can be adjusted accordingly. Alternately, calibration
circuitry may be provided for direct calibration to percent oxygen concentra-
tion in air, avoiding the need for tables.
AIR AND PURE OXYGEN DISSOLUTION CONTROL SYSTEMS
A number of air and pure oxygen dissolution control systems can be
employed for various modifications of the activated sludge process and vari-
ous aeration methods. Three basic aeration systems are described, namely:
diffused air aeration systems, mechanical aeration systems, and pure oxygen
dissolution systems.
Diffused Air Aeration
Diffused air aeration is accomplished by introducing air under pressure
as bubbles into the mixed liquor. Blowers of the positive displacement,
centrifugal or axial type are used for air compression. Table 4 summarizes
information concerning the application of various types of blowers used in
diffused aeration systems .
In a multiple blower installation, a control system is required to start,
stop and sequence the blowers to meet the air demands of the activated
sludge process. A detailed description of the operation of a typical multiple
74
-------
blower installation is given below. Centrifugal blowers are described as
they are most commonly employed for diffused air aeration.
Control of Centrifugal Blowers--
A typical schematic diagram for the control of multiple centrifugal
blowers is shown in Figure 51 (Reference 9). The purpose of a blower control
system is to match blower output to the total oxygen demand at all times . It
should be noted that regulation of blower output does not directly control the
amount of air put into the oxidation tanks or other air-demanding systems,
such as channel aeration. Rather, the oxidation tank air demand is controlled
solely by individual modulation of the oxidation tank air header valves in
response to measured DO levels in the associated pass. Demand, therefore,
is controlled solely by the needs of each system.
AIR
NON-RETURN
•INLET GUIDE VANES VALVE
• CENTRIFUGAL
BLOWER (TYP)
FROM
OTHER
BLOWERS
TO OTHER
OXIDATION TANK
AIR HEADERS
— HEADER
Figure 51. Diffused air aeration blower control system.
Blower volume control--The aeration air header pressure must remain
constant at all times to ensure stable operation of the individual DO control
loops . Typical means of ensuring that blower output matches demand is to regu-
late the speed, inlet guide vanes or suction throttling valves of all running
blowers to maintain a constant preset pressure in the header.
75
-------
As shown in Figure 51, each blower is provided with power-actuated
inlet guide vanes and a low head loss flowmeter (FT) for the control and
measurement of flow, respectively. The temperature and pressure of the air
in the common air header is measured by a temperature transmitter (TT) and a
pressure indicating transmitter (PIT), respectively. The signal from each FT
is fed to a computing relay (UY) which also receives the temperature and
pressure measurements from the TT and PIT. Computing relay UY solves the
following equation and its output is volumetric flow at standard pressure and
temperature, which serves as the controlled variable input to flow controller
FIG.
3 3
The equation for converting actual m /s (acfm) to standard m /s (scfm)
is as follows:
? ^ /i m + P \ f 7 Q 4 \
i i *-> / , i ^ / / -L U 1 ~ JT \ I & 3 '-I \ / n \
standard m /s = actual m /s I—rr;—I ( ~r~ ) (°a)
where P = pressure in kPa
T = temperature in degrees Kelvin
K
3
multiply m /s by 2120 for cfm.
In U. S. units the equation takes the following form:
f f fl±JL±J?\ ( 530 \ ,n,.
scfm = acfm ^ 14>? ) (jg^T^) (8b)
where P = pressure in psig
T = temperature in °F
One pressure controller (PIC) is provided which compares the actual
header pressure to the desired pressure (set point). The output from the
PIC serves as the set point for the FIG associated with each blower. The
output from each FIG is fed to a current-to-pneumatic converter (ZY) which in
turn produces a pneumatic signal that positions the inlet guide vane assembly
on the associated blower.
The control system, which can be described as a pressure-to-flow cas-
cade control system, is employed to ensure a specific discharge header
pressure is constantly maintained and all operating blowers produce the same
output. For example, if supply exceeds demand, the header pressure will
start increasing; the output from controller PIC will then reduce and the inlet
guide vanes will be further throttled until supply again matches the demand.
The reverse is true when demand exceeds supply.
76
-------
Blower starting/stopping control--In addition to modulating blower out-
put to match demand, the control system is designed to start and stop blowers
to match their capacities to the output required. Figure 52 shows a typical
centrifugal blower starting/stopping sequence diagram. Under conditions of
rising air demand, an additional blower is started when the power consump-
tion (measured by wattmeters) or the mass flow (volumetric flow compensated
for temperature and pressure) of the running units indicates they have reached
maximum rated output. Under conditions of falling air demand, the last
started blower is stopped when one less blower will give the output required.
It is always more economical to run the least number of blowers that will
meet demand.
The criterion used by the control system to test if remaining blowers can
satisfy the air demand is: if the average blower discharge is less than
x(n-l)(n ), then the last-started blower is stopped, where: n = number of
operating blowers and x = maximum rated output of one blower. Duty rotation,
i.e. , rotating the assignment of blowers to lead and follow duty, is performed
by the operator via lead selector switch HS shown in Figure 51. Rotation of
blower duty should be practiced to equalize running times of all units .
Parallel operation of centrifugal blowers — If the centrifugal blowers are
operated in parallel, each should be provided with its own surge control
system, and the outputs of all operating blowers should be modulated uni-
formly. It is very difficult to operate one variable speed blower in parallel
with a number of fixed "wide open" blowers which can be cut in or out as
demand varies. To do so, the variable speed units must be of considerably
larger capacity than the others, because of the relatively small flow turn-
down ratio. (In general, the surge line for variable speed operation limits
the useful rangeability of a blower to about a 2:1 flow turndown; the range-
ability of blowers with inlet throttling is usually at least a 3:1 flow turndown
(5).) This would require different blower characteristics and, therefore, an
expensive spare unit. It is preferable to modulate all operating blowers to-
gether and choose unit capacities small enough so two units are required at
minimum flow. This control system can be applied to any activated sludge
process flow scheme because the DO in each tank pass can be independently
controlled.
Process Configurations--
Three basic process configurations applicable to diffused air aeration
systems are described below.
Plug-flow, single pass — Oxidation tanks with a large length to width
ratio are generally designed to pass the mixed liquor through the oxidation
tank on a plug-flow basis . Biological stabilization proceeds more rapidly
near the oxidation tank inlet, where substrate concentration is highest,
77
-------
14
12
10
•o
c
TO
0>
Q
Max Calculated Demand
[-Output of First
Follow Blower
6 8
Blower Output,
10 12
14
Figure 52. Typical centrifugal blower starting/stopping diagram.
-------
thereby implying highest oxygen demand. Since it is desirable to maintain
aerobic conditions at all points in the oxidation tank pass, it is recommended
that DO probe receptacles be provided near the inlet, middle, and end of a
plug-flow oxidation tank.
Plug-flow, multiple-pass — Multiple-pass oxidation tanks may be con-
sidered the equivalent of a folded, single-pass oxidation tank. Therefore,
control logic applicable to multi-pass tanks is essentially the same as in
the single-pass oxidation tank. DO probe receptacles should be provided
near the inlet, middle and end of each pass.
Completely mixed--Completely mixed aeration provides for the equali-
zation of the oxygen demand throughout the oxidation tank and a more uniform
DO concentration. In systems where a completely mixed operating regime is
approximated, probe location in the oxidation tank is less critical. However,
long tanks may not accomplish complete mixing, thereby developing a grada-
tion of oxygen demand from one end of the tank to the other. Flexibility in
DO probe location is therefore recommended.
Typical Diffused Air Aeration Control Systems —
A typical diffused air aeration control system is shown in Figure 53. This
control system can be applied to any of the activated sludge process flow
configurations described in Section 4, because the DO in each tank pass can
be independently controlled. The control system employs a cascaded configu-
ration whereby the primary DO loop adjusts the set point of a secondary flow
control loop. Although it is possible to use a single DO control loop to
position the associated air control valve (particularly if the air header pres-
sure is held constant), the cascade control loop approach is recommended to
obtain good system stability and response.
The output signals from on-line DO probes require conditioning to mini-
mize the effects of (a) DO probe noise, and (b) random noise caused by turbu-
lent mixing conditions in the oxidation tank. Simple time-averaging and low-
pass filters are considered suitable for estimating the true DO value from the
real-time DO probe output signal.
The distribution of air from a tank pass air header is usually not auto-
matically controlled. The DO control system goal is to meet the aggregate
air demand of a given tank pass. Manual control valves are normally pro-
vided on each header downcomer feeding one or more air diffusers to permit
balancing. If the operator wishes to change the air distribution along a given
pass header, e.g., for tapered aeration, the manual control valves are used
for this purpose.
79
-------
FLOW CONTROLLER
D.O. CONTROLLER
1
I/P
FROM
BLOWERS\
1 ^ / i-r- \ *J^*I -
\ *»
to -
TOOTHER
OXIDATION
TANK PASS
AIR HEADERS
(AE>O
i t H i
OXIDATION TANK
PASS (TYP)
Figure 53. Diffused air aeration dissolved oxygen control system
Feedforward-Feedback Dissolved Oxygen Control Systems--
Two feedforward-feedback DO control systems that are currently being
tested are shown in Figures 54 and 55. It should be noted that the limiting
factor in matching the air/oxygen addition to an oxidation tank with a time-
varying organic input load is the aeration equipment. If the aeration equip-
ment does not have sufficient turndown, this deficiency cannot be remedied
by the DO control system. .
The total oxygen demand of the mixed liquor at the inlet end of the oxi-
dation tank is time-varying. Furthermore, the oxygen demand at any point
in the oxidation tank depends on the physical configuration of the oxidation
tank with respect to influent and effluent structures, tank geometry and the
type and location of the aeration equipment. As an example, in a typical
plug-flow type activated sludge process, the oxygen demand is highest at the
inlet end of the tank and gradually decreases to a minimum at the effluent end
of the tank. In contrast, a completely mixed plant has a relatively uniform
oxygen demand throughout the length of the tank.
80
-------
FEEDFORWARD
CONTROLLER
TOOTHER FEEDBACK
DO CONTROLLERS
Y T
I I I
EFFLUENT
CONSTANT
— FEEDBACK
DO CONTROLLER I
H\K\ (TYP)
I /P
V
f f t
DO
t t t
OXIDATION TANK PASS (TYP)
Figure 54. Feedforward-feedback hydraulic load following
dissolved oxygen control system
Feedback DO control automatically compensates for all of the afore-
mentioned load changes that can occur in the activated sludge process.
However, because there is a time lag between a change in oxidation tank
load and a corresponding change in the DO level in a given pass, feedforward
control is recommended to reduce DO level variations in plants that treat
wastewater with a highly variable oxygen demand.
-------
With feedforward control, the basic concept is to measure the dominant
load change and then feed it forward through a control strategy to all related
manipulated variables. The feedforward model for aach variable reflects the
steady-state process relationships between the load change and the controlled
variable. The feedforward DO control system continually balances the oxygen
FEEDFORWARD
CONTROLLER
__ TOOTHER FEEDBACK
F "*" DO CONTROLLERS
EFFLUENT
FEEDBACK
DO CONTROLLER
(TYP)
I /P
,-L SAMPLE
/AY_\& HOLD
( AIT j TOTAL
t t t
t t t
OXIDATION TANK PASS (TYP)
Figure 55. Feedforward-feedback hydraulic and organic
load following dissolved oxygen control system
82
-------
delivered to the oxidation tank against the estimated steady-state oxygen
demand of the mixed liquor.
To correct for errors in the feedforward calculation, a DO feedback con-
troller is linked to the feedforward control system. The difference between
the set point and the output of the feedback controller is the offset that would
have appeared if feedback had not been used. The outputs of both the feed-
forward and feedback control systems are summed, and the resultant signal
is cascaded to the set points of. the individual tank pass air flow corftrollers.
A feedforward-feedback DO control system incorporating hydraulic load
following is shown in Figure 54. The basic feedforward strategy is to vary
the air flow rate in proportion to the primary effluent flow rate and use DO
feedback control to maintain the DO level at the desired value. The primary
effluent flow rate is not usually measured directly; it is computed by summing
the plant influent and recirculation flow rates and subtracting the primary
sludge underflow rate. In the feedforward calculation, the primary effluent
flow signal is multiplied by a constant (based on the average diurnal value of
the mixed liquor oxygen demand). The output of UY1 is the estimated total
air flow rate required.
The output from UYl is divided by "a" , a manual input that represents
the number of oxidation tank air headers in service. The output of UY2 is the
estimated air flow rate for a single air header. If the feedforward controller
is implemented by digital logic, variable dead time compensation is provided
via KY to account for the variable time delay between a load change at the
oxidation tank inlet and an oxygen demand change in the tank. Unless the
process is completely mixed, individual dead time compensation is required
for each tank pass because the load change is reflected sequentially in each
pass.
A DO probe is installed at an intermediate point in each pass. If the DO
level in any pass deviates from 'a preset low value, the associated DO con-
troller adjusts the output of the feedforward controller to correct the error.
The resultant signal serves as the set point to the associated flow controller
which, in turn, regulates the air flow to the particular pass. The flow control-
ler is equipped with a minimum output limiter (HIK) to ensure adequate mixed
liquor mixing at all times.
In Figure 55, a feedforward-feedback DO control system incorporating
hydraulic and organic load following is presented. The oxygen demand of the
83
-------
mixed liquor at the influent end of the oxidation tank is estimated from an on-
line, near real-time, total organic carbon (TOG) measurement and the sum of
the primary effluent and return activated sludge flow rates. The feedforward
air flow estimate using flow and TOG measurements is much more ac-
curate than an estimate based on flow alone. However, the capital and operat-
ing costs of TOG analyzers are such that this technique may only be warranted
for plants experiencing large variations in loading rate. On-line respirometers
can also be employed to estimate the near real-time oxygen demand of the
mixed liquor.
Mechanical Aeration
Mechanical aeration is accomplished by either transferring atmospheric
or pure oxygen to the liquid by surface renewal and interchange or by dispens-
ing compressed air or oxygen fed below the surface to a rotating agitator. As
described in Section 5, mechanical aerators are broadly classified as plate,
updraft, downdraft, combination and brush types. If the impeller diameter
and tank area are fixed, the aeration capacity of mechanical aerators is a
function of speed and immersion depth. Speed regulation, immersion depth
control and intermittent operation modes are available for DO control.
Floating aerators inherently maintain a constant depth of submergence.
However, platform or fixed-mounted aerators pose a problem in DO control
because, unless the overflow or outlet capacity is quite large relative to the
size of the oxidation tank, the water level and submersion will vary with
flow fluctuations through the oxidation tank. Water level control is normally
accomplished by the positioning of butterfly valves on the outlet piping or by
sliding gates.
Process Configurations--
Two basic process configurations applicable to mechanical aeration
systems are described below.
Plug flow—Oxidation tanks with large length to width ratios and con-
taining several aerators, as shown in Figure 56, approach plug-flow condi-
tions. The degree of mixing around the aerator at the influent end of the
tank provides essentially complete mixing in its zone of influence, but the
hydraulic translation from the zone of influence of the first aerator to the
zone of influence of the second aerator progresses slowly, and the situation
is repeated down the tank. As a result, the tank operates essentially like
five smaller tanks in series, and the result approximates plug-flow rather
than complete mixing conditions .
84
-------
MECHANICAL
AERATOR (TYP)
INFLUENT
c
x
• i
1 ^ i
Cx Cx Cx
EFFLUENT
Figure 56. Mechanical aeration - plug flow configuration.
If adjustment of oxygen transfer capacity is applied to the entire tank
(by procedures such as raising and lowering the effluent weir to change sub-
mergence of the aerators), the location of the DO probe should be left to the
discretion of the operator, and DO probe receptacles should be provided so
that the probe may be placed at either the influent end, the effluent end, or
the middle of the tank.
If changes in aeration capacity are to be accomplished by speed changes
of the individual aerators, the most advantageous control logic may not
necessarily be speed changes on all aerators at the same time. For example,
increased concentrations of wastewater influent will mainly affect the oxygen
demand in the vicinity of the aerators at the influent end of the tank. There-
fore, the influent-end aerators can be grouped for control of oxygen transfer
capacity based on demand, and the remaining aerators can be controlled as a
separate group. In general, adequate DO monitoring and control can be ac-
complished by placing the probes in the zone of influence of any aerator in
a given control group.
Rectangular basin configuration—Where mechanical aerators are applied
as shown in Figure 57, the problem of confined mixing around a given aerator
exists in a similar manner to that suggested for the plug-flow configuration.
For control purposes, aerators near the inlet or effluent ends of the tank may
be grouped for collective control. In Figure 57, the five aerators at the in-
fluent end of the tank may be considered as a group,and speed changes for
oxygenation capacity control can be made simultaneously to that group. Like-
wise, the five aerators near the effluent end may be controlled and considered
as a separate group.
-------
MECHANICAL
AERATOR (TYP)
1 t t t
EFFLUENT
fx /
x
INFLUENT
Figure 57. Mechanical aeration - completely mixed configuration.
Other aerator groupings might be considered so that smaller step
changes in aeration capacity can be made. For example, the five aerators
near the influent end of the tank may be subdivided into groups of two and
three aerators. If this approach is taken, the isolation established by the
zone of influence must be recognized. For example, if two-speed aerators
are employed, and two aerators are on high speed while three aerators are on
low speed, there is no single sampling point in the line of five aerators which
represents the average impact of the transfer capacity. The impact can be
averaged by integration of samples from the zone of influence of each aerator,
but integration requires a complicated arrangement of pumped samples. A
practical solution is to carefully select a probe location within the zone of
one aerator so that acceptable information for control purposes is obtained.
In the application of mechanical aeration to rectangular oxidation tanks,
the direction of rotation of the aerators should be considered. The objective
is to select aerator rotations which do not establish a strong flow pattern
along the line between aerators. For example, if the aerators in Figure 57
86
-------
were arranged so that the aerators on-the left rotated clockwise and the
aerators on the right rotated counterclockwise, a strong short-circuiting flow
would be established down the middle of the tank. Therefore, it is suggested
that all aerators rotate in the same direction.
Mechanical Aeration Control Systems--
Five mechanical aeration control systems are described below and in-
clude the following: (a) on-off control, (b) two-speed control, (c) variable
speed control, (d) variable impeller depth, and (e) variable level control.
On-off aerator control system--On-off aerator control is usually limited
to oxidation tanks equipped with a single speed surface aerator(s). Further-
more, on-off operation should only be considered for wastes that do not
exhibit pronounced separation characteristics. Settling tests should be made
to determine whether on-off operation is feasible and, if so, the maximum
"off" time that can be tolerated before liquid-solid separation occurs should
be established. A typical on-off aerator control system is shown in Figure 58.
A I-O
Figure 58. On-off aerator control system.
Two-speed aerator control system—A control system which is designed
to maintain the DO level between preset high and low limits by adjusting the
combined output of two-speed aerators is shown in Figure 59. Control sys-
tems of the step-control variety are often employed with final control ele-
ments that cannot be continuously output-modulated, such as two-speed
87
-------
AT
A 1-0
MOTOR STARTER
CONTROL LOGIC
Figure 59. Two-speed aerator step control system.
mechanical aerators. Step control permits the time constant in the aeration
system (reflected basically as the rate of change of oxygen demand in the
aeration system) to be matched by step changes in aerator capacity taken at
adjustable intervals of time .
The step change in aeration capacity can be accomplished by a prede-
termined change in mechanical aerator speed or any other operating step
which will change the oxygen transfer capacity in the system. For example,
Figure 60 shows a time step diagram for an oxidation tank with four two-
speed aerators, where minimum capacity is defined as all four aerators at
low speed and maximum capacity corresponds to all aerators at high speed.
The time indicated as A T in Figure 60 is adjusted by the operator so the air
supply approximately matches the instantaneous air demand of the aeration
system. Timer KG in Figure 59 is adjusted so that at every preset time in-
terval A T, the value of the DO signal input to control unit AIC is compared
to preset high and low limits. If the DO level is at an intermediate value
between the high and low limits, no adjustment to the aerator capacity is
88
-------
MAX CAPACITY
o
Q.
O
cc
o
I-
QC
LU
<
MIN CAPACITY
STEP CHANGE IN
AERATOR CAPACITY
(e.g. A CHANGE FROM
HIGH TO LOW SPEED)
A T
TIME
Figure 60. Aerator capacity/time step diagram,
made. If the DO level is lower than the preset low limit, the motor starter
control logic steps upward by one capacity increment, and this new level of
aerator capacity is sustained for the remainder of the time interval. At the
end of the AT-interval, timer KG once more activates the control logic, and
the DO level is again compared to the high-low limits which are preset in
controller AIC. If the DO signal is lower than the low limit, another step up-
ward in aerator capacity is made. However, if the DO signal is higher than
the preset high limit, a downward step in aerator capacity is made. Accord-
ing to Woodruff (40), each step adjustment in oxygen transfer capacity should
be at least 10 percent of the total oxygenation capacity of a given oxidation
tank.
Variable speed aerator control—Large mechanical aerators are sometimes
equipped with either variable speed motors or variable speed gear reducers.
A typical variable speed aerator control system is shown in Figure 61. The
speed of the aerator is modulated by controller AIC to maintain a preset DO
level.
Variable impeller depth control system—The oxygenation capacity of a
turbine aerator can be varied by adjusting the depth of submergence of the
impeller blades . The impeller elevation is normally adjusted by means of an
automatically controlled spline-type vertical shaft on the turbine aerator.
The greater the blade submergence, the greater the oxygenation capacity
of the aerator,,and vice versa. A typical variable impeller depth control sys-
tem is shown in Figure 62. A cascade control system is employed, whereby
89
-------
OXIDATION TANK
Figure 61. Variable speed aerator control system,
OXIDATION TANK
Figure 62. Variable impeller depth control system.
90
-------
the output of DO controller AIC serves as the set point to submergence
controller ZIC, which regulates the submergence of the turbine impeller to
maintain the desired DO level in the oxidation tank.
Variable level control system—Another method of varying the oxygenation
capacity of a turbine aerator is to control the liquid level in the oxidation
tank and thereby control the depth of submergence of the turbine impeller
blades. The liquid level can be controlled by raising or lowering a mechani-
cal rotary weir. Weir positioning is accomplished by using a reduction gear
and a motorized speed reducer. A typical variable level control system is
shown in Figure 63. The output of DO controller AIC serves as the set point
to weir position controller ZIC. The output from controller AIC adjusts the
weir position, thereby regulating the liquid level and submergence of the
turbine impeller to maintain the desired DO level in the oxidation tank.
SUBMERGED
TURBINE AERATOR
OXIDATION TANK
EFFLUENT
TROUGH
Figure 63. Variable level control system.
Pure Oxygen Dissolution Control Systems
When covered and staged oxidation tanks are used in the pure oxygen
process to introduce oxygen gas into the mixed liquor, pressure control can
be used to control the oxygen feed rate. A typical pure oxygen dissolution
control system for a submerged turbine mixer and sparger system with recir-
culation compressors is shown in Figure 64 (Reference 9).
91
-------
Effluent
Stage 1
Stage 2 Stog<
OXIDATION TANK (Typ)
'toge 3 Stage 4 T
Process
Effluent
Figure 64. Pure oxygen dissolution control system.
Oxygen is introduced to each of the covered oxidation tanks in the first
stage only and exhausted as vent gas from the finai stage. The oxygen
supplied to the first stage is controlled to maintain a constant gas pressure
in the first stage of the oxidation tank by automatic modulation of a control
valve on the oxygen supply header. The oxygen purity control system is de-
signed to maintain constant oxygen purity in the gas venting from each oxi-
dation tank by modulation of the vent gas valve. Pressure controller PIC and
oxygen purity controller AIC maintain the pressure and oxygen purity set
points, respectively, for each oxidation tank.. The pressure and purity control
systems interact as follows: A slight increase in oxygen demand in any oxi-
dation tank causes a reduction in the oxygen purity of gas venting from the
tank. Oxygen purity controller AIC compensates for this drop in purity by
opening the vent gas valve, which causes a drop in pressure in the tank,
thus causing a higher oxygen flow into the first stage of the tank. The sys-
tem then stabilizes at a higher oxygen flow rate, while maintaining the same
oxygen purity and gas pressure set points. An increase in oxygen purity at
the vent valve causes a reduction in oxygen supply to the tanks. The system
is designed to utilize 90 percent of the oxygen supplied.
92
-------
SECTION 7
CASE HISTORY SUMMARY OF DISSOLVED OXYGEN CONTROL SYSTEMS
Performance , operational data, and maintenance data for aeration equipment
and associated DO control systems in a number of activated sludge treatment
plants, for both manual and automatic DO control, are presented in the
Appendix. Performance comparisons made for each control mode include the
following, where such data were available:
0 BOD removal efficiency
0 Suspended solids removal efficiency
0 Sludge volume index
0 Air supplied per unit quantity of influent
0 Air supplied per unit quantity of BOD removed
0 BOD removed per blower kwh
Considerable difficulty was encountered in obtaining useful DO control
system data from most of the wastewater treatment plants that were contacted.
Many of the plants are designed with automatic DO control systems, but
these systems are not yet installed or are not being utilized. Other plants
originally reported as having automatic DO control had only a remote manual
control system. Typically, remote manual DO control in these plants
involved reading DO levels on indicators in the control center and then
manually initiating some control action, such as increasing mixer speed to
raise the DO concentration to a desired value.
For the purposes of this study, the ideal plant control situation is one in
which parallel oxidation tank units are operated simultaneously with one
tank under manual and the other under automatic DO control. Such a situation
occurred in the Rye Meads and Reno-Sparks studies.
The next best situation for a comparative DO control system study is a
plant that is operated in a manual and an automatic DO control mode at
93
-------
different times over an extended period. This type of operation occurred at
the Renton, Palo Alto and San Francisco International Airport plants.
The remaining plants reported in the Appendix currently operate in a
manual or automatic DO control mode and required a special test for a compar-
ative analysis. Short-term tests were devised to examine plant performance
parameters under both modes of operation. It is recognized that data derived
from such short experiments can be misleading, but it was felt that 48 hours
of testing was all that could reasonably be requested of the plant operating
personnel. In some cases, such as the San Jose-Santa Clara plant and the
Reno/Sparks plant, the personnel expressed considerable interest in the study
and elected to perform an extended test.
During both long- and short-term testing periods at some plants, problems
arose, such as slug loads of jet fuel, storms, power failures, aeration
equipment failures, misinterpretation of data desired or inadequate staffing
for the specified sampling frequency on composite analyses. Data have been
reported as received, with explanations furnished for inconsistencies.
TEST PROCEDURE
Where short-term DO control tests were performed for the purposes of
this study, test data requested for each test were as follows:
1. Total air flow applied to the aeration tank (deduct air applied to
other uses, such as mixed liquor channels).
2 . Total influent to aeration tanks being examined.
3. Average BOD and suspended solids applied to the aeration tanks
(24 hour composite analysis with samples taken every hour and
stored at 3-4° C),
4. Average BOD and suspended solids in the plant effluent before
chlorination (same type composite analysis as for No. 3 above).
5. Sludge Volume Index (SVI) on mixed liquor taken every four hours.
6. DO Concentration plot at the control probe.
7. Power consumption (kWh) of the blowers or mixers.
It was further requested that: days selected for the tests be typical of
average conditions at the plant; DO should be maintained at approximately
the same level in the oxidation tank during each test; manual control actions
94
-------
be performed approximately every four hours, unless the plant normally
controls at some other time interval.
CASE HISTORIES
A total of 12 case histories are presented in the Appendix, covering a
flow capacity range of 44 dm3/s-4.38 m^/s (1-100 mgd). Both municipal and
industrial plants are included, with examples of constant and variable flow
range facilities. All plants discussed use the activated sludge process with
some also having tertiary treatment. The descriptive format for each case
history is as follows:
Description of Aeration and DO Control System
Operation
Performance
Maintenance
Safety and Emergency Procedures
Table 5 lists the case histories studied with some brief descriptive data on
each plant.
Table 6 summarizes plant loading, flow, and test duration data for each
case history plant. Loadings and flows shown pertain only to that portion of
the plant that was tested under manual and automatic DO control. Table 7
summarizes plant performance comparisons of manual and automatic DO control
in terms of selected performance measuring parameters. As shown, data were
not available from some plants for all parameters measured. Additional para-
meters measured and explanations for inconsistancies are in the Appendix.
Table 7 shows substantial improvements in the measured parameters
occurring in a few plants and slight improvement in most others. Averaging
the test data for each plant studied, excepting those noted, yields the results
shown in Table 8. While results did differ between the case history plants
studied, the conclusion is drawn from Tables 7 and 8 that automatic DO
control generally results in improved plant performance and less dissolution
energy consumption than manual DO control.
95
-------
TABLE 5. CASE HISTORY PLANTS
Case
number
1
2
3
4
5
6
7
8
9
10
11
12
Plant
Renton Wastewater
Treatment Plant,
Washington
Palo Alto Water
Quality Control
Plant, California
Rye Meads Sewage
Purification Works,
Hertfordshire, Eng .
City of Oxford Sewage
Works, England
Valley Community
Services District
Wastewater Treat-
ment Plant, California
Reno-Sparks Joint
Water Pollution
Control Plant, Nevada
Simi Valley Water
Quality Control Plant,
California
San Francisco Inter-
national Airport Water
Quality Control Plant,
California
St. Regis Wastewater
Treatment Plant,
Sartell, Minnesota
Long Beach Water
Renovation Plant,
California
San Jose-Santa
Clara Water Pollu-
tion Control Plant,
California
Cranston Water
Pollution Control
Facility, Rhode Island
Flow, rrrVs
(mgd)
1.6
(36)
1.5
(34)
0.69
(15.8)
-
0.2
(4)
0.88
(20)
0.31
(7)
0.04
(1)
0.3
(7)
0 .55
(12.5)
7.01
(160)
0.50
(11.5)
Type
aeration
Diffused
Diffused
Diffused
Diffused
Diffused
Diffused
Diffused
Mechanical
Mechanical
Diffused
Diffused
Diffused
Flow
scheme
Plug flow or
step feed
Plug flow or
reaeration
Plug flow
Step feed
Multi-mode
Multi-mode
Plug flow, step
feed, reaeration
Plug flow, step
feed
Plug flow
Step feed
Plug flow, step
feed
Contact
stabilization
Remarks
Fully nitrified
effluents
Fully nitrified
effluent
Fully nitrified
effluent
Fully nitrified
effluent
Fully nitrified
effluej^ tapered
aeration
96
-------
TABLE 6. SUMMARY OF PLANT LOADING, FLOW AND TEST DURATION DATA
Case _, a
, Plant
number
1 Renton
2 Palo Alto
3 Rye Meads
4 Oxford
5 Valley
6 Reno-Sparks,
System 2
Reno-Sparks ,
System 3
7 Simi Valley
8 San Francisco A .P.
9 St. Regis
10 Long Beach
1 1 San Jose Santa
Clara
12 Cranston
Test data
Test
mode
M
A
M
A
M
A
M
A
M
A
M
A
M
A
M
A
M
A
M
A
M
A
M
A
M
A
A
BOD loading,
mg/m /s
(lb/1000 cf/day)
3.95
5.86
4.52
5.23
4.71
4.56
(21.3)
(31.6)
(24.4)
(28.2)
(25.4)
(24.6)
no data
no data
6.80
5.19
4.86
3.97
3.65
3.97
3.11
3.45
3.89
7.49
6.45
4.89
3.86
3.32
10.1
10.4
9.08
12.2
13.2
(36.7)
(28.0)
(26.2)
(21.4)
(19.6)
(21.4)
(16.8)
(18.6)
(21.0)
(40.4)
(34.8)
(26.4)
(20 .8)
(17.9)
(54.3)
(56.1)
(49.0)
(65.6)
(71.3)
Flow, m /s
(mgd)
1.07
1.19
1.05
1.03
0.11
0.11
(24.5)
(27.1)
(24.0)
(23.6)
( 2.5)
(2.5)
no data
no data
0. 18
0.15
0.28
0 .28
0.28
0.28
0.20
0.21
0.04
0.04
0 .25
0.24
0 .30
0.30
1.97
2.01
0 .16
0.20
0.19
( 4.0)
( 3.5)
( 6.3)
( 6.3)
( 6.3)
(6.3)
(4.6)
(4.7)
( 1.0)
( 0.9)
(5.6)
( 5.5)
( 6.9)
( 6.9)
(45.0)
(45.9)
(3.6)
(4.6)
( 4.2)
Test
duration
3 months
3 months
1 month
1 month
6 months
6 months
Several
months
Several
months
48 hours
48 hours
48 hours
48 hours
48 hours
48 hours
24 hours
24 hours
1 month
1 month
4 days
4 days
24 hours
24 hours
4 days
4 days
24 hours
24 hours
24 hours
b
See Table 5 or Appendix for complete plant name
M = manual; A = automatic
97
-------
TABLE 7 SUMMARY OF PERFORMANCE COMPARISONS BETWEEN MANUAL
AND AUTOMATIC DISSOLVED OXYGEN CONTROL
Case
number
1
2
3
4
5
6
7
8
9
10
11
12
Plant3
Renton
Palo Alto
Rye Meads
Oxford
Valley
Reno-Sparks-2
Reno-Sparks-3
Simi Valley
San Francisco
A. P.
St. Regis
Long Beach
San Jose/
Santa Clara
Cranston
Control mode
and percent
improvement'-'
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic
% improvement
Manual
Automatic 1
% improvement
Automatic 2
% improvement
BOD
removal
efficiency,
percent
85
96
11
84
84
none
97
98
1
94
95
1
92
78
-14
74
85
11
82
81
-1
92
94
2
97
98
1
97
97
None
85
85
None
91
94
3
92
1
Suspended
solids
removal
efficiency,
percent
46
53
7
91
92
1
90
87
-3
83
86
3
84
82
-2
99
97
-2
79
96
17
88
89
1
90
90
None
86
86
None
89
96
7
83
-6
Sludge
volume
index
(SVI)
332
86
74
45
79
-76
112
95
15
113
108
4
115
100
13
130
127
2
92
201
-118
252
201
20
99
94
5
102
101
1
56
60
-7
69
-23
Air supplied
per unit
volume of
influent,
mVm3C
9.3
8.2
12
3.3
3.4
none
14
11
21
24
28
-17
10.5
9.5
10
7.1
5.8
18
19.8
19.3
3
27
25
7
6.7
6.0
10
19
15
21
16
16
Air supplied
per unit
quantity of
BOD removed,
' mVkgd
137
86
37
33
28
15
88
71
19
122
162
-33
110
140
-27
120
72
40
240
220
8
190
180
5
600
520
11
79
58
27
56
29
BOD
removed
per blower
kwh,
kg/kWhe
0.40
0.63
58
1.3
1.6
23
0.38
0.27
-29
0.55
0.29
-47
0.35
0.57
63
0.24
0.27
13
-
See Table 5 or Appendix for complete plant name
b
Percent improvement computed as automatic control over manual control. For BOD and suspended solids, percent
improvement computed by subtraction. For remaining parameters, percent improvement computed by subtraction
of values and division by manual value.
"mVm3 multiplied by 0 .134 = cf/gal
mVkc multiplied by 16.02 cf/lb
""kg/kYv'h multiplied by 2.20 = Ib/kWh
98
-------
TABLE 8. TEST DATA AVERAGES OF CASE HISTORY PLANTS FOR
AUTOMATIC COMPARED TO MANUAL DISSOLVED OXYGEN CONTROL
Parameter
BOD removal efficiency
Suspended solids removal efficiency
Sludge volume index
Air supplied per unit volume of influent
Air supplied per unit quantity of BOD
removed3
BOD removed per blower kWh
Number
of tests
12
10
11
10
9
5
Average percent
improvement
2.8
2.3
2.6
12
21
33
Excepting Valley and Reno-Sparks System 2
Excepting San Francisco Airport
~i
"Excepting Valley
DISSOLVED OXYGEN CONTROL SYSTEM DESIGN ERRORS AND PROBLEMS
While gathering information on activated sludge plants for this manual,
many design errors and problems were discovered in the application of
dissolved oxygen control systems. In some cases, where problems were
relatively simple , such as replacement of corroded aluminum DO probe termi-
nal connections with noncorroding ones, or adding sealer to DO probe assem-
blies to prevent water intrusion, the plant has taken the necessary corrective
steps. However, in many cases, where the designer had neglected to include
a critical controller or had designed a DO control system where other
critical processes were detrimentally affected by blower throttling, the plant
has reverted to manual DO control.
Many plants investigated are designed with automatic DO control systems,
but these systems are not yet installed or are not being utilized. A typical
reason given for this situation is that the effluent meets required standards
and the plant staff is occupied with more pressing matters. Failure of a
critical component of the automatic DO control system, such as a DO probe or
99
-------
mixer variable speed relay, results in reversion to manual DO control. Since
superintendents claim good quality effluent is produced under manual DO con-
trol, repairing the automatic DO control system becomes a low priority item.
Management at one plant waited almost two years after start-up to begin
installing the DO probes necessary for the automatic DO control system.
Unfortunately, many of the probes could not be made functional after that
length of time and new probes had to be ordered.
One industrial plant reported that the DO control system for regulating
surface aerator speed is so complicated that only the supplier can repair it.
Consequently, after a few outages, the system was abandoned and the plant
now operates in a manual DO control mode.
One design error observed was .oversizing the throttling valve for air feed
to oxidation tanks to accommodate future plant expansion. At present air
flow rates, the plant finds no throttling effect is evident until the valve is 90
percent closed, resulting in poor air application control.
One plant studied was designed for multiple modes of activated sludge
operation, but has operated in only the two-pass mode since start-up. When
the superintendent was asked why he did not check on the efficiency of the
other modes, he replied that alternate DO probe receptacles do not exist to
permit effective DO control in any mode other than two-pass-, and no funds
existed in his budget to install them. He also added he had no inclination to
change modes since the plant easily meets all effluent requirements.
Another plant has an automatic DO control system designed to throttle
an air feed header butterfly valve in response to a DO signal received from
the oxidation tanks. However, the slightest alteration in DO results in wide
excursions of the butterfly valve , even with a tuned PID flow controller in
the control loop. It was recognized later by the engineer that a cascade
control system with a DO controller computing the set point of the flow
controller would have been a better design. This plant is currently operated
in a manual mode and produces an outstanding effluent. Thus, plant personnel
have made no effort to rectify the automatic control system problem.
An additional design error discovered is utilization of positive displace-
ment blowers to furnish air to a number of use points, such as aerobic
digesters, channel aeration diffusers, return activated sludge air lift pumps
and oxidation tanks, without any turndown provision. Throttling the air feed
header butterfly valve to the oxidation tanks through a DO control loop results
in more or less air to the other use points. A resultant increase or decrease
100
-------
in the return sludge rate is beyond the operator's control. Consequently, the
plant is run under manual DO control with one blower at maximum capacity con-
tinuously and a second blower added during canning season. With at least one
blower on at full capacity, plugging of the channel aeration headers is avoided.
Another plant is furnished with positive displacement blowers that supply
air to both a contact and a stabilization tank. Air headers to each tank are
throttled by separate DO control loops. After start-up, the operators observed
that throttling the air headers was not very effective, since the blowers had no
turndown capability and excess air could not be released. Accordingly, the
plant installed a pressure reducing valve on the blower discharge manifold
that vents to the atmosphere. Operating history indicates two blowers are
operated continuously with constant venting of some discharge air. Power
savings through DO control is, thus, nonexistent in this plant.
A major design error observed is incorporation of sophisticated automatic
DO control equipment in a plant with a flow equalization basin or constant
loading. The converse erro'r observed was minimal DO control equipment in a
plant that exhibits a daily flow variation of 5 to 1 but has no flow equalization
basin. The first plant easily maintains a relatively constant DO concentration
in the oxidation tanks by manual DO control. The second plant has difficulty
maintaining an adequate DO level with a single DO probe and a mixer control
system without automatic impeller submergence, variable speed, or other such
air application control method .
Additional investigation of operating plants showed that adequate con-
sideration is not given to maintaining mixed liquor solids in suspension in
DO control system design. Operators of many plants with mechanical aera-
tors cited mixing considerations as a major reason for abandoning automatic
DO control in favor of manual control. The problem lies in sizing the mixer.
The mixer should be sized to maintain solids in suspension under existing
and future loadings at low speed, while reserving high speed operation for
additional oxygen as required. Operators of mechanically mixed plants that
reported satisfactory operation at start-up frequently later found they had to
operate the mixers at high speed most of the time to maintain solids in sus-
pension, regardless of DO level
In general, more consideration should be given to proper blower selection,
interdependence of process air systems, flow and loading variability, control
system capabilities, expertise of control system in-plant maintenance person-
nel, control system flexibility, and adaptability to current as well as future
plant capacity. It was evident in our investigation that automatic DO con-
trol is not warranted in every plant. An industrial or municipal sewage
plant in which all blowers must be operated at full capacity on a continuous
101
-------
basis has no need for automatic blower throttling. The use of dual installed
controllers and multiple DO probes for critical DO readings may be suitable
for large scale wastewater treatment plants but may not be in the budget of a
44-438 dm-Vs (1-10 mgd) facility. However, a plant with a successfully
operating automatic DO control system should not be dependent on a single
DO probe with no spares, as was the case in one facility visited.
102
-------
SECTION 8
COST OF AUTOMATIC DISSOLVED OXYGEN CONTROL SYSTEMS
While studying DO control systems at the twelve wastewater treatment
plants, summarized in Section 7, it was apparent to the authors that a con-
siderable variation exists in plant design, performance, aeration and DO
control system design and operational flexibility. Comparing capital and
operating costs of DO control system components in plants that have manual
DO control with the same size plants having automatic DO control is un-
justifiable because of the above mentioned variations. One solution to this
problem is to apply capital, operation and maintenance cost data reported
on specific components and types of control systems for actual plants to
hypothetical plants of various capacities with assumed characteristics. Com-
parisons can then be drawn between the cost of manual and automatic DO
control on a common basis by adding the incremental cost of automatic DO
control to a plant having manual DO control.
DESCRIPTION OF THE HYPOTHETICAL ACTIVATED SLUDGE PLANTS
Five hypothetical plant designs with automatic DO control systems have
been synthesized. These plants cover a plant flow capacity range of from
44 dm3/s to 4.4 m /s (1-100 mgd) and were designed in accordance with (a)
standards and guidelines referenced in this report, (b) information developed
in the earlier chapters, and (c) accepted engineering practice. The designs
are based on the conventional activated sludge process with either single or
multiple-pass oxidation tanks. Table 9 shows the assumed design charac-
teristics of the hypothetical plants .
Loading
BOD loading figures for each size plant were developed assuming com-
bined wastewater collection systems and medium strength wastewater. An
average 5-day BOD load of 200 ppm was used (6,7,20). An average of 5.6
mg/m3 of aerator volume/s (30 Ib BOD/1000 cf/day) is assumed (20,27) .
Considering this loading and an average influent BOD of 200 ppm, approxi-
mately 1.58 dam (55,700 cf) of oxidation tank volume is required per 44
dm3/s (1 mgd) of plant flow.
103
-------
TABLE 9. DESIGN DATA FOR TYPICAL ACTIVATED SLUDGE PLANT
AERATION SYSTEMS
Plant size , dm /s
Aeration type
Aerator type
BOD loading ,
mg/m /s
(lb/1000 of/day)
Air supplied ,
m3/m3
(cf/gal)
Number of oxidation tanks
Number of passes
Oxidation tank dimensions
length x width x water depth,
Metres
(Feet)
Pass width,
Metres
(Feet)
Number of mixers
Number of blowers
Blower capacity , each,
std m3/s
( scfm)
Mixer or blower power,
kW
(hp)
44
Mechanical
Surface
mechanical
aerator
3.58
(19.3)
2s
1
18x18x3.7
(60x60x12)
18
(60)
2
30e
(40)
44
Diffused
Positive
displacement
lobe type
blower
5.6
(30)
1
(7.5)
1
2
26x13x4.6
(84x44x15)
6.7
(22)
2b
0.33
(700)
22f
(30)
440
Diffused
Positive
displacement
lobe type
blower
5.6
(30)
1
(7.5)
2
3
61x28x4.6
(200x90x15)
9.1
(30)
3C
1.65
(3500)
93.29
(125)
2200
Diffused
Single-
stage
centrifugal
blower
5.6
(30)
1
(7.5)
8
4
61x37x4.6
(200x120x15)
9.1
(30)
-
3C
8.49
(18,000)
447h
(600)
4400
Diffused
Single-
stage
centrifugal
blower
5 .6
(30)
1
(7.5)
16
2
120x18x4.6
(400x60x15)
9.1
(30)
3C
16.5
(35,000)
89 5h
(1200)
Each tank designed to treat 2/3 of the load.
Blowers sized to each deliver 100% of air required.
Blowers sized to each deliver 50% of air required.
At 52 kPa (7.5 psig) discharge pressure.
Based on 746 kW/28m3 (1.0 hp/1000 cf) of oxidation tank volume.
Assumed isothermal efficiency of 60%.
Assumed isothermal efficiency of 70%.
h
Assumed adiabatic efficiency of 78'%.
Note:
104
-------
Number of Tanks and Dimensions
Most oxidation tanks in the United States are 4.6m (15 feet) deep and
are spiral flow type (2,34). References 2 and 27 specify a liquid depth of
3-4.6 m (10-15 feet). Each oxidation tank generally has 1 to 4 relatively
narrow channels or passes since multiple passes control longitudinal mixing
and short circuiting (39) . Width to depth ratios for spiral flow tanks general
ly run from 1:1 to 2.2:1 with 1:2 being the limit recommended by Reference 2.
The oxidation tank numbers and dimensions for each plant shown in Table 9
were determined using these guidelines and practical experience. Given the
tank depth, oxidation tank volume required and acceptable width to depth
ratios, tank lengths were determined based on a reasonable number of
passes and accepted design practice.
Mechanical and Diffused Air System Considerations
Although mechanical aerators are used in activated sludge plants of all
sizes, they are usually installed in plants up to 44 dmvs (1 mgd) in size.
Reference 34 quotes another reference as saying the dividing line between
diffused air and mechanical aeration is 44 dmvs (1 mgd) . Another reference
quoted in Reference 34 defines the division as 70 dm /s (1.6 mgd). It is
also reported mechanical mixers are seldom used for plants that serve popu-
lations in excess of 5000 (34). Assuming 6.57 cm /capita/s (150 gpcd) as
the average water supply and a 75 percent return of supply water as sewage
flow, this converts the population figure of 5000 into a plant size of 24 dmvs
(0.56 mgd) . For the purposes of this study, it is assumed that plants of 44
dm /s (1 mgd) capacity would be furnished with either mechanical or diffused
air systems. Larger plants are assumed to have diffused air systems.
Oxidation Tank Mechanical Mixer Selection
Oxidation tanks with mechanical mixers are usually square with 18m x
18m (60 x 60 feet) being common dimensions. Following the Ten States
Standard (27) guidelines of a 3-4.6 m (10-15 feet) water depth, the tanks are
designed to be 3.7 m (12 feet) deep.
Standby capability for diffused air systems is provided by adding an addi-
tional compressor. Standby capability for a mechanical aeration system is
achieved by oversizing each oxidation tank. Design experience indicates
oversizing each of two tanks to take two-thirds of the average plant loading
is a reasonable approach. Accordingly, the BOD loading for the 44 dm3/s
(1 mgd) mechanical mixed plant in Table 9 is shown as only 3.5 kg/m3/s
(19 lb/1000 cf/day) instead of 5.6 kg/m3/s (30 lb/1000 cf/day) as for the
diffused air plants. The mixer horsepower selection is based on providing
26 kW/m3 (1.0 hp/1000 cf) of oxidation tank volume.
105
-------
Diffused Aeration Air Requirements
Aeration air requirements for conventional activated sludge plants are
given in terms of cubic metres of air per cubic metre (cubic feet per gallon) of
influent (3.7-11 m3/m3) (0.5-1.5 cf/gal) (7), cubic metres of air per kilo-
gram (cubic feet per pound) of BOD applied (93. 6 mVkg) , (1500 cf/lb) (27) or
cubic metres of air per kilogram 9£ BOD removed (cubic feet per pound of BOD
removed) (31-44 m3/kg) (500-700 cf/lb) (34). Assuming a plant has an in-
fluent BODr of 200 ppm and 35 percent of the influent BOD is removed in the
primaries, results in 130 ppm of BODs being applied to the oxidation tanks.
Assuming a 90 percent 6005 removal in the secondary treatment stage,
results in 117 ppm of 6005 being removed. Applying the above air flow rate
guides to these figures results in air requirements ranging from 0. 19-0.52
m3/s (400-1100 scfm) of air for a 44 dm /s) (1 mgd) plant.
A more rigorous analysis utilizing equation (11) (taken from Reference 6)
for oxygen required results in 15.3 g/s/kg MLVSS (1320 Ib/day/lb MLVSS)
(6):
Ib O2/day = aLR + b (Ib MLVSS) (11)
where a = 0.45 = fraction of 5-day BOD removed that is used to
provide growth energy (typically 0.35-0.55)
b = 1.4 ug O2/s/g (0.120 lb/O2/day/lb) = endogenous
respiration rate
MLVSS = 2100 ppm (assumed)
JR
L = g BOD5 removed/sec = 5.12 g/s (976 Ib/day) for a 44 dm3/s
(1 mgd) plant with 117 ppm BOD removed
oxidation tank volume = 1.6 dam3 (0.42 million gallons) for 44 dm3/s
(1 mgd) plant
Assuming an oxygen transfer efficiency of 8. 6 percent (6, 7),and using an
air density of 1.20 kg/m3 (0.075 Ib/cf) and an oxygen content of 23.2 percent
by weight results in 0.295 m3/s (625 scfm) being required for a 44 dm3/s
(1 mgd) activated sludge plant with an assumed loading of 200 ppm BODr.
Considering this result, along with other data available, results in an
assumed air flow rate for the 44 dm3/s (1 mgd) plant of 0.33 m3/s (700 scfm)
or about 7.48 m3/m (1 .0 cf/gal) of influent.
Blower Selection
The assumption that the aeration system diffusers are located one foot
above the oxidation tank bottom results in a static head of 42 kPa (14 feet)
for the blowers to develop. A distribution system and diffuser loss of 25
percent of the mixed liquor depth is assumed, or 10.5 kPa (3.50 feet).
106
-------
Summing the static and dynamics heads results in a required discharge
pressure of 52 kPa (7.5 psig) .
Blower types, sizes and number of units for each plant are selected in
accordance with (a) the guidelines of Table 1, (b) field observation of exist-
ing installations and (c) practical experience. A 100 percent capacity stand-
by blower for the 44 dm^/s (1 mgd) plant is selected because it is judged
more economical and practical to provide two complete 100 percent capacity
units for the small size blowers required, rather than three 50 percent
capacity units, as for the other diffused air plants. Since the aeration
system is considered a critical plant process in maintaining required plant
effluent standards, it is assumed the system should be capable of operation
with the largest unit out of service. Loss of a single blower in any of the
plants will, therefore, not degrade process performance.
Theoretical power requirements for the rotary, positive displacement,
lobe type blowers are computed assuming isothermal compression according
to the following formula:
kW = PQ1 ln(Q2/Q) (12a)
where P, - suction pressure , kPa
Q, = actual inlet flow, m /s
2
Q2 = actual outlet flow, m /s
In US units the equation takes the following form:
ghp = (144 PiQi/33,000) ln(Q2/Q1) (12b)
where ghp = gas horsepower
P, = suction pressure , psia
0^ = actual inlet flow, acfm
Q2 = actual outlet flow, acfm
Isothermal efficiencies of 60 percent and 70 percent are utilized in calcu-
lating the actual or brake horsepower requirements of the blowers for the
44 dm /s (1 mgd) and 440 dm^/s (10 mgd) plants respectively.
Theoretical blower horsepower requirements for the centrifugal blowers
are computed assuming isentropic (reversible adiabatic) compression accord-
ing to the following formula;
kW = PQ k/(l-k)r
(k
(13a)
107
-------
where P = suction pressure , kPa
3
Q = actual inlet flow, m /s
k = Cp/Cv = ratio of specific heats = 1.395 for air
r = pressure ratio of discharge in kPa divided by suction in kPa
In US units the equation takes the following form:
ghp = 144 PQ/33000 [k/(l-k)j [r(k"1)/k-l] (13b)
where ghp = gas horsepower
P = suction pressure, psia
Q = actual inlet flow, acfm
k = Cp/Cv = ratio of specific heats = 1.395 for air
r = pressure ratio of discharge in psia divided by suction in
psia
An adiabatic efficiency of 78 percent was assumed for the centrifugal blowers
in calculating the brake horsepower required.
Dissolved Oxygen Control Equipment
Figures 65 through 69 schematically illustrate the oxidation tank confi-
gurations, diffused and mechanical aeration equipment and DO control sys-
tems for the five hypothetical plants. The DO control systems presented
reflect current practice and have a history of successful operation.
As shown in Figures 65 through 69, the DO probes are positioned for a
plug flow operating mode. Alternate DO probe receptacles are provided to
permit flexibility in DO probe placement if the plant operating mode is
changed to step feed, complete mix, or other as provided in the plant design.
Field experience indicates that plant operating modes are changed only once
or twice per year. It is therefore assumed that the operator would recalibrate
and relocate probes at alternate receptacles as necessary when the operating
mode is switched. Since DO probe receptacle groups are wired into a
common transmitter, plugging a DO probe into any of the receptacles in a
given receptacle group automatically connects the probe output to the trans-
mitter. If a DO probe malfunctions, the operator can either switch to another
DO control probe by patching in another transmitter with a connected DO
probe to controller AIC, or the defective DO probe can be replaced.
108
-------
r
PE
1
PE
OX ! DAT ION
ALTERNATE DO PROBE
RECEPTACLE (TYP)
TANKS
@D°
AI-0
^
('A I
30kW(40hp) 2-SPEED
MOTOR (TYP)
~1
1-0
L
Figure 65. Automatic DO control system for mechanically aerated
44 dm^/s H rngd) activated sludge plant.
-------
o
2 - SPEED
MOTOR (TYP)
/0.3m3/s (700SCFM)
POS. DISPL. BLOWER
(TYP)
A1'0
c\ /c\ /c
ML.
i PAS S J 1
fAd00 !
DO
0
PASS , 2
1 / | \ |
T / >——ALTERNATE
OXIDATION TANK-7 I RFrFPTAri
DO PROBE
RECEPTACLE (TYP)
1
Figure 66. Automatic DO control system for diffused air 44 dm3/s
(1 mgd) activated sludge plant.
-------
r
I
FROM (A 1C) FOR 2ND
OXIDATION TANK
I.6m3/s
POS DISPL
BLOWER
!ND
OXIDATION
—I
)
PE
1^_
»•
TAN
P 2)
i
i
i
^
t —
\y /•-
(
V
K^
(Zn)- -l T
1 1
1 1
PASS i @r-
i -, ^
— ^.
PASS 3Y rr r"Xoo
I ^ALTERNATE DO F
jL RECEPTACLE (TYP
' ©
(
— ^
ROBE
)
[AIT)
~\
ML
AIR HEADERS TO OTHER
OXIDATION TANK
NOTE: i m3/s = 2100 CFM
Figure 67. Automatic DO control system for diffused air 0.44 m3/s
(10 mgd) activated sludge plant.
-------
INLET
GUIDE
VANES
8.5m°/s
SINGLE STAGE
CENTRIFUGAL
BLOWER
OTHER TWO
8.5m3/s
SINGLE STAGE
CENTRIF BLOWERS
NOTE: i m3/s = 2100
i
.AIR HEADERS
TO OTHER
OXIDATION TANKS
SP
%-f
V Y
OXIDATION TANK
(TYP OF 8)
. ALTERNATE DO PROBE RECEPTACLE—/ I
I/P I , —
-------
INLET
GUIDE
VANES
I6m3/s
SINGLE STAGE
CENTRIF. BLOWER
OTHER TWO
16 m3/s
SINGLE STAGE
CENTRIF. BLOWERS
NOTE I m3/s = 2100 CFM
I L (_
OXIDATION TANK
(TYP OF 16)
-ALTERNATE DO
PROBE RECEPTACLE(TYP)
Figure 69. Automatic DO control system for diffused air 4.4
(100 mgd) activated sludge plant.
-------
For the purpose of identifying the incremental costs of adding automatic
DO control equipment to manually controlled hypothetical plants, the follow-
ing definitions of manual and automatic control are employed: A manual DO
control system involves the use of a portable DO probe or a laboratory analy-
sis, such as the Winkler method, to obtain DO levels in the mixed liquor.
Aeration system final control elements are manually started, stopped and ad-
justed from a central plant control room. An automatic DO control system
involves the use of in situ DO probes for the automatic, continuous measure-
ment and centralized display of DO levels in the mixed liquor. Automatic
controllers are provided in a central plant control room for starting, stopping,
and adjusting all aeration system final control elements.
Both manual and automatic DO control systems include monitoring and
protective devices for the aeration equipment. Local manual controls are
provided for maintenance purposes.
CAPITAL COSTS
Specific components unique to automatic DO control systems that can be
identified by type and quantity are listed in Table 10. Because the combined
cost of the DO probes and transmitters represents a substantial part of the
total installed cost of automatic DO components for each plant, the oxidation
tank configuration has a significant effect on the capital cost of DO equipment.
Different numbers of tanks and passes and different tank dimensions result in
either more or fewer DO probes and transmitters, thereby affecting the total
capital cost of the DO control system components. Capital costs and operat-
ing and maintenance costs discussed below are based on mid-1975 prices.
OPERATING AND MAINTENANCE COSTS
Operating cost considerations include control and instrumentation equip-
ment maintenance costs, savings in manual labor,and power savings through
DO automatic control. DO probe maintenance typically involves checking,
cleaning, calibration, membrane replacement, and recharging. Usually this
maintenance is performed by the plant staff. More complex instrumentation
maintenance is normally done by an instrumentation maintenance contractor.
The extent of in-plant maintenance required is a function of many vari-
ables, such as type of probe, capacity of electrolyte reservoir, type of
wastewater, probe placement, preventive maintenance schedules, ease of
probe access , and competence of maintenance personnel. DO probe maintenance
schedules and manpower requirements for the case history plants described
in the Appendix were analyzed, and the result reported in terms of man-
hours per probe per year. Man-hours expended range from a low of 7 mh/
probe/year at Oxford and Renton to a high of 64-68 mh/probe/year at the
Valley and Simi Valley plants. Obviously, maintenance requirements vary
1L4
-------
TABLE 10. INSTALLED COST OF COMPONENTS UNIQUE TO AUTOMATIC
DISSOLVED OXYGEN CONTROL SYSTEMS
a
Components
DO probe
DO probe receptacle
DO indicating transmitter
DO recorder
DO indicating controller
Flow indicating controller
I/P converter
DO 2 -speed controller
Total
Unit
installed
cost-$b
1300
200
1400
1200
1600
1600
400
400
Plant size and aeration type
44 dm3/s
mechanical
#
2
4
2
1
2
Installed
cost-$
2,600
800
2,800
1,200
800
$8,200
44 dm3/s
diffused
#
2
3
2
1
1
Installed
cost-$
2,600
600
2,800
1,200
400
$7,600
440 dm3/
diffused
#
6
14
8
2
2
Installed
cost-$
7 ,800
2,800
8,400
2,400
800
$22,400
2200 dm3/s
diffused
#
24
80
24
8
16
16
16
Installed
cost-$
31,200
16,000
33,600
9,600
25,600
25,600
6,400
$148,000
4400 dm /s
diffused
#
48
128
48
16
16
16
16
Installed
cost-$
62,400
25,600
67,200
19,200
25,600
25,600
6,400
$232,000
Remarks
Includes agitator,
receptacle & hardware
Extra junction box
2-mode
2-mode
Differential gap
cn
All instruments electronic unless otherwise indicated.
Installed cost based on October 1975 prices and includes a multiplier of 2 for wiring, calibration and installation.
Note: 1 dm3/s = 2.28xlO"2mgd
-------
drastically from one plant to another. It is, therefore, difficult to assign an
average labor expenditure per probe .
Experience in the United Kingdom at plants such as Basingstoke, where
continuous DO monitoring is performed, has shown probe calibration is
required once every four to six weeks. Probe cleaning is required only prior
to calibration. Membrane and anode replacements are performed every six to
nine months (19).
Considering the eight case history plants in the Appendix on which DO
probe maintenance data is available , the average labor time per probe per
year is 36 man-hours. Table 11 shows typical frequencies of various DO
probe in-plan't maintenance functions reported by staff members of the case
history plants.
TABLE 11 . TYPICAL FREQUENCIES OF DISSOLVED OXYGEN PROBE
MAINTENANCE FUNCTIONS
DO probe maintenance function
Checking and cleaning
Recalibration
Recharging
Membrane replacement
Frequency
2 -
1
6 -
1 -
7 days
4 weeks
8 months
6 months
Dissolved Oxygen Control Instrument Maintenance
Since maintenance of the plant instrumentation systems for the case
histories studied is generally done under an instrumentation contract, esti-
mates for maintaining the components of Table 10 were solicited from instru-
ment vendors specializing in these contracts. Table 12 details annual parts
and labor costs per unit of equipment for typical instrumentation contracts.
Information contained in Tables 10 and 12 was used to develop the total
outside instrumentation contract costs for various plant sizes as shown in
Table 13. Care was taken to ensure no overlap occurred in estimated labor
and materials cost for the DO probes since certain in-plant maintenance
functions are already accounted for in Table 11.
Laboratory Time Credit
It was assumed that a savings in plant laboratory time would be realized
with an automatic DO control system since DO concentrations are available
automatically. However, a check with some of the plants described in the
Appendix, including Renton, indicated that no labor savings exists. Labora-
tory personnel report that additional time required for DO probe calibration
116
-------
offsets the time that would be required for laboratory determination of oxida-
tion tank DO level.
TABLE 12. ESTIMATED ANNUAL INSTRUMENTATION CONTRACT PARTS
AND LABOR COST FOR AUTOMATIC DISSOLVED OXYGEN
CONTROL EQUIPMENT
Component
DO probe and transmitter
Controller - 2 mode
Controller - 2 speed
Recorder
I/P converter
Labor
(man-hours)
7
4
3
4
2
Labor
Aa
cost-$
210
120
90
120
60
Parts
cost-$
85
40
25
100
25
Total parts
and labor
cost-$
295
160
115
220
85
Labor cost @ $30/man-hour.
Manual Labor Credit
Installation of automatic DO control equipment is expected to result in
some savings in manual labor that is normally required in valve throttling and
blower speed changes. Plant experience and the results of the performance
tests reported in the Appendix indicate that without automatic control systems
the valves, blowers or mixers involved in the DO control system would be
checked and altered as required about every four hours. Assuming a checking
and throttling operation requires one man, one minute per device, a count of
the blowers, mixers and header control valves for each plant as shown in
Figures 65 through 69 results in an annual labor credit that is realized by the
automatic DO control system. The estimated labor credit that results from
automation is given in Table 14.
Power Savings Credit
An analysis of the percent improvement in air supplied per unit quantity
of BOD removed indicates that an average air and power savings of 20 per-
cent can be realized with automatic DO control systems incorporating centri-
fugal blowers with adjustable inlet guide vane or suction throttling constant
pressure control. Positive displacement blowers and mechanical mixers
realize a power savings with automatic DO control through dual or multiple
117
-------
TABLE 13. ESTIMATED INCREMENTAL ANNUAL INSTRUMENTATION
CONTRACT PARTS AND LABOR COST FOR AUTOMATIC
DISSOLVED OXYGEN CONTROL EQUIPMENT
a
Component
DO probe & transmitter
Controller - 2 mode
Controller - 2 speed
Recorder
I/P converter
Total
Units
parts
and
labor
cost-$
295
160
115
220
85
Plant size and type aeration
44 dm3/s
mechanical
#C
2
2
1
Cost-$
590
230
220
$1,040
44 drri3/s
diffused
#C
2
1
1
Cost-$
590
115
220
$925
440 dm3/s
diffused
#C
6
2
2
Cost-$
1,770
230
440
$2,440
2200 dm3/s
diffused
#C
24
32
8
16
Cost-$
7,080
5,120
1,760
1,360
$15,320
4400 dm3/s
diffused
#c
48
32
16
16
Cost-$
14, 160
5, 120
3,520
1,360
$24,160
b
Receptacles not included.
Unit parts and labor cost from Table 12 .
"Number of components from Table 10 .
3 -2
Note: 1 dm /s = 2 .28 x 10 mgd.
-------
speed operation. For the purposes of this economic analysis, a 20 percent
power savings is assumed for the positive displacement blowers and mixers
in the hypothetical plants if 2-speed drives are used.
TABLE 14. ESTIMATED ANNUAL LABOR CREDIT FOR AUTOMATIC
DISSOLVED OXYGEN CONTROL
Hypothetical
plant and
aeration type
3
44 dm /s mechanical
3
44 dm /s diffused
440 dm°/s diffused
2200 dm3/s diffused
3
4400 dm /s diffused
Number of
blowers, mixers
or valves
2
2
3C
16
16
Man-hours
per year3
88
88
132
704
704
Annual
costb
$
968
968
1,452
7,744
7,744
Assumes one minute/valve every 4 hours (44 man-hours per unit per year).
Based on a salary of $14,000/year for a plant operator plus 50% allowance
for fringe benefits ($11.00 per hour total) .
i— i
Assumes 1 unit each for blowers and 1/2 unit for each header throttling valve.
Note:
dm3/s = 2.28 x 10~2 mgd
A straight line relationship was assumed between air saved and power
expended. If suction throttling is employed on centrifugal blowers, a power
savings is realized that is almost directly proportional to the reduction in
discharge air flow. A slight reduction in blower efficiency occurs during
suction throttling, but this effect was ignored in the computation of power
savings.
The positive displacement blowers and mechanical mixers of the hypo-
thetical plants realize a power savings under automatic DO control only
through speed reduction. Power consumed by the St. Regis plant (Case
History No. 9) 45 kW (60 hp) mixers was measured and a 30 percent power
decrease was observed in dropping from high to low speed. For the purposes
of this economic analysis, it is assumed a 20 percent power savings can be
realized with the positive displacement blowers and mixers of the hypo-
thetical plants if 2-speed drives are used.
119
-------
The conversion from theoretical blower horsepower required to kWh
consumed is made using an assumed load or demand factor of 1 .0 and induc-
tion motor efficiencies ranging from 89-95 percent, depending on motor size.
Power cost perkWh for each plant is derived from published information on
the average power consumption of various size activated sludge plants (36),
consideration of typical utility rate schedules, and Federal Power Commission
data. Figure 70 is a plot of power costs derived from these sources. The
Federal Power Commission, data reflect average monthly cost of power for
industrial users in. the United States during 1973. Information on the Cincin-
nati Gas and Electric Company rates and the vertical dashed lines represent-
ing average plant power consumption were obtained from a recent EPA report
(36). The Sacramento Utility District costs for selected power consumption
rates were computed from the May 1 , 1974 rate schedule for large industrial
users (33). The synthesized curve of power costs used in this study was
developed from consideration of the rate plots for the two typical utilities and
the more general Federal Power Commission data. Table 15 shows estimated
plant power consumption after Smith (36) and power cost perkWh read from the
synthesized curve. Table 16 summarizes the operating and maintenance costs
for all size plants considered,
TABLE 15. ESTIMATED ANNUAL PLANT POWER COST
Plant size ,
m /s
44
440
2200
4400
Average powe r
consumption
(kilowatt hours/month)
35,000
273,800
1,277,500
2,585,000
Power cost
(cents/kilowatt hour)
3.0
2.4
2.3
2.3
-
Note: 1 dm /s = 2.28 10 mgd .
SUMMARY OF CAPITAL, OPERATING AND MAINTENANCE COSTS
Table 17 is a combination of Tables 10 and 16 and summarizes the econo-
mics of adding automatic DO control to various size activated sludge plants
on an annual cost basis. The annual capital recovery cost was computed
using an interest rate of seven percent and a service life of 12 years for
instruments and controls as recommended in Reference 8. Inflation of electric
power, labor and equipment costs was not considered since the costs involved
120
-------
35
30 h
25 h
CA 20 h-
O
o
O
Q_
15 h
10
NOTE: i MILL= $ .oci I
dm3/s)(2 28X I0"2} = MGD
COMMISSION - 1973
100 1000
PuWER CONSUMPTION IN MWh / MONTH
10,000
Figure 70. Average U.S. industrial power cost per kWh compared to
industrial rates of typical utilities.
-------
TABLE 16. ESTIMATED 1975 OPERATING AND MAINTENANCE COSTS OF
ADDING AUTOMATIC DISSOLVED OXYGEN CONTROL TO VARIOUS
SIZE ACTIVATED SLUDGE PLANTS
Item
DO probe maintenance
@ $11.00/hour and g fa
36 man-hours/probe/year
DO control instrument
maintenance including parts
and labor0
Credit for valve modulation,
blower or mixer speed change
@ $11 .00/hour and 44 man-hours/
unit/year
Credit for power saved
Total annual credit
Annual cost in dollars for various
sized plants and type aeration
44 dm3/s
mechanical
800
1,000
(1,000)
(4,400)
(3,600)
44 dm /s
diffused
800
900
(1,000)
(1,700)
(1,000)
440 dm3/s
diffused
2,400
2,400
(l,500)f
(11,000)
(7,700)
2200 dm3/s
diffused
9,500
15,300
(7,700)
(52,000)
(34,900)
4400 dm3/s
diffused
19,000
24,200
(7,700)
(101,000)
(65,500)
3Based on a basic salary of $14,000/year for an instrument technician or plant operator plus 50%
additional for fringe benefits.
Probe maintenance includes labor for checking, cleaning, calibration, membrane replacement and"
recharging.
Assumed under outside contract at a labor cost, including fringe benefits of $30/hour. Includes
any special DO probe maintenance and all parts, See Tables 12 and 13.
See Table 14.
6Based on a 20% reduction in air required under automatic DO control. See Table 15 for power
rates used.
Air header throttling valves treated as 1/2 unit each.
3 -2
Note: 1 dm /s = 2.28 x 10 mgd.
-------
TABLE 17. ECONOMIC ANALYSIS OF ADDING AUTOMATIC DISSOLVED
OXYGEN CONTROL TO VARIOUS SIZE ACTIVATED SLUDGE
PLANTS IN 1975
Item
Capital cost
Annual capital recovery cost
Annual operation and
maintenance credit0
Total annual savings
Annual savings in dollars for various
sized plants and type aeration
44 dm3/s
mechanical
8,200
1,000
(3,600)
(2,600)
44 dm3/s
diffused
7,600
1,000
(1,000)
0
440 dm3/s
diffused
22,200
2,800
(7,700)
(4,900)
2200 dm3/s
diffused
148,000
18,600
(34,900)
(16,300)
4400 dm3/s
diffused
232,000
29,200
(65,500)
(36,300)
to
to
From Table 10.
Based on 7% interest for a 12 year service life and no salvage value. (Capital Recovery
Factor = 0. 1259) .
"From Table 16.
-2
Note; 1 dm3/s = 2.28 x 10 mgd.
-------
are expected to change through time by approximately the same percentage (8) .
Table 17 indicates that an annual savings can be realized through automatic
DO control for all of the cases shown except for the 44 dm3/s (1 mgd) diffused
air plant.
Figure 71 is plotted from the data on diffused air plants presented in
Table 17 and illustrates that an annual savings in using automatic DO control
can be expected for activated sludge plants larger than 44 dm /s (1 mgd).
The conclusion is drawn that potential savings can be obtained by employing
automatic DO control in activated sludge plants. It is recommended that an
economic study be performed for each case in which automatic DO control is
being considered. The methodology outlined in this manual is one approach
to this analysis.
An intrinsic benefit of automatic DO control not treated in the economic
analysis above is the improvement in plant performance parameters compared
to manual DO control. It is difficult to quantify an improvement of 10-15
percent in suspended solids or BOD removal efficiencies in a plant where a
conversion from manual to automatic DO control is being considered, particu-
larly if the plant is meeting required effluent standards under manual DO
control. On the other hand, an owner faced with more stringent discharge
requirements that may normally necessitate expensive plant modifications
should consider automatic DO control as a possible solution to the problem
of improving the plant performance with a relatively small capital investment.
SIGNIFICANT COST FACTORS
A number of factors enter into the economic analysis of DO control sys-.
terns that have a major impact on estimated costs. The significant cost
factors identified in the preparation of this manual are: (a) blower selection,
(b) number of DO probes, (c) DO probe arrangement for service, and (d) power
cost.
Blower Selection
Blower selection largely determines the type of DO control system em-
ployed as well as the power credit realized under automatic DO control. Since
DO probes and associated transmitters account for a major part of automatic
DO control system capital costs and a substantial portion of the operating
costs, careful consideration should be given to ensure that no more than the
minimum number required are installed.
It has been stressed in this report that blower selection largely determines
the type of DO control system employed as well as the power credit realized
under automatic DO control. In general, dynamic blowers appear better
suited to automatic DO control systems since horsepower can be conserved
124
-------
80,000
60,000
40,000
20,000
10,000
8,000
<£ 6,000
l! 4,000
O
0
2,000
1,000
800
600
400
200
n
NOTE: 43.8dm3/s= 1
ANNUAL CAPITAL
RECOVERY CREDIT \
rf^
j
^
r
^
/
/
^
/
/
(
\ />
/
/
ANNUAL
\s
£
D /
O
SAV
/
%
\
INC
/
/
;s
mgd
/
//
'
-ANNU/
/
^ /
kLO&M
A
P
P
COS!
10
100
1000
10,0000
PLANT SIZE, dm0/*
Figure 71 . Automatic dissolved oxygen control economics for
various size activated sludge plants.
125
-------
over a variable blower operating range. Use of unloader valves or relief
valves to regulate the volumetric output of positive displacement blowers
does not conserve power, as discussed in Section 5. Speed variation, as
observed in the field with saturated variable core reactor drives, wastes any
potential power savings by expending energy through resistors when reducing
blower speed.
Two-speed control systems were designed for the positive displacement
(PD) blowers of the hypothetical plants. It was assumed that 20 percent
power savings could be realized with this system, although not one of the
case history plants had 2-speed PD blower control. Thus, power cost savings
assigned to these systems is subject to question and should be further
explored. One shortcoming of a 2-speed control system is that the blower
output cannot closely follow the dissolved oxygen control level. It is, there-
fore, expected that process performance improvements would not be as pro-
nounced with this control system as with a totally variable air flow output
arrangement. A question also arises as to whether a 2-speed automatic DO
control system with PD blowers will provide equivalent power savings and
performance improvement as would a more sophisticated, variable range,
automatic DO control system with centrifugal blowers.
Number of Dissolved Oxygen Probes
As demonstrated in Table 16, the number of DO probes utilized can make
a substantial difference in the annual cost. Since such a large investment
and operating and maintenance expense is involved with the DO probes and
transmitters, careful consideration should be given to ensure no more than
the minimum number required are installed.
Dissolved Oxygen Probe Arrangement for Service
As discussed above in the section on DO probe maintenance, a consider-
able difference in expended manhours per year per probe was encountered in
the case history plants presented in the Appendix. DO probe in-plant mainte-
nance may require as little as 7 mh/year per probe or as much as 66 mh/year
per probe. Some of the higher and lower estimates were rechecked with the
plant personnel. The Renton plant personnel attribute the relatively low re-
quirement of 7 mh/probe/year to efficiency of maintenance and infrequency
of cleaning required. The DO probes are mounted on long shafts that are
pivoted on the tank railing. By pushing the short end of the shaft above the
pivot, the probe is rotated out of the water parallel to the tank wall. The
shaft is then dropped in a U-pipe holder and the probe is wiped clean with a
rag. Calibration is done by adjusting the probe output to correspond to the
oxygen content of the atmosphere (air calibration) . Careful attention to probe
installation design to facilitate maintenance can result in a significant reduc-
tion in man-hour requirements for keeping the probes in an effective operating
condition.
126
-------
Power Cost
As shown in Table 16, the major saving that can be attributed to auto-
matic DO control is power cost. Consideration of the electric utility billing
practice is an important prerequisite in determining the annual savings to a
plant by installing automatic DO control equipment.
Billing systems based on maximum demand are not as favorable to auto-
matic DO control equipment power savings as are systems based on actual
energy consumed. The computation of a power credit in Table 16 is based on
a utility billing system whereby a customer pays for actual energy consumed.
A 93 kW (125 hp) centrifugal blower may draw 101 kW or 2424 kWh/day with a
load factor of 1.0 and a motor efficiency of 92 percent when the suction
throttling valve is wide open. Under automatic DO control, this same blower
is expected to draw an average of 20 percent less or 1939 kWh/day. But dur-
ing the course of throttling to meet the variable oxygen demand of the process,
the blower will draw a variable amount of power, depending on its operating
point. Under some utility billing systems, the plant is charged at a rate based
on the maximum 15-minute power demand. Thus, if the blower happened to be
drawing full power for 15 minutes during any given month, no amount of suction
throttling to conserve power is going to affect the plant power bill charge for
that blower. On the other hand, a utility billing system based on actual
energy consumed is expected to result in a credit to the plant power bill of
approximately 20 percent of the maximum demand if the blower is operated in
an automatic DO control system. Consideration of the utility service district
billing practice is thus an important prerequisite in determining a net annual
cost to a plant for installing automatic DO control equipment.
The actual cost of power is also a significant factor in the economic
analysis of DO control systems. For example, a 4.38 m3/s (100 mgd) plant
in Sacramento may realize $46,000 a year in savings with automatic DO con-
trol equipment, whereas the same sized plant in New York City may save
$138,000. Since capital and other maintenance costs for the two plants are
assumed to be the same, power cost can drastically affect the decision re-
garding the addition of automatic DO control equipment to a plant.
ADDITIONAL COST FACTORS
Additional cost factors that should be considered in an economic analysis
of dissolved oxygen control systems are (a) process flexibility, (b) type of
wastewater, (c) type of DO probes, (d) actual incremental labor costs, and
(e) equipment sizing and service life.
127
-------
Process Flexibility
A significant cost factor discussed above is number of probes. Number
and location of probes for optional DO control are recommended in Section 6.
An important consideration here is process flexibility. By having a number of
probes at specified locations, the process can be switched from conventional
to step feed or recycled sludge can be introduced at a different point without
worrying about having a suitable DO control probe located at the proper spot
for effective DO control.
Considering the demonstrated capital and maintenance cost intensiveness
of DO probes, a given process application should be examined in terms of
needed process flexibility and frequency of change. It appears more cost
effective to use fewer DO probes and more DO probe receptacles for those
processes where operating mode is changed infrequently. A reduction in DO
probes utilized will have a pronounced effect on the annual cost of an auto-
matic DO control system.
Type of Wastewater
As discussed above in DO probe maintenance, man-hours required per
probe vary considerably from plant to plant. At the Valley plant (Case History
No. 5), 50 percent of the total probe maintenance time is spent on probe
cleaning, but cleaning is done on a daily basis. At the Oxford treatment
plant, cleaning is required only once every two weeks. At the Renton, Long
Beach and Simi Valley plants, cleaning is performed only once or twice per
week. Assuming the probes are all subject to the same water circulation
velocity and are similarly designed, it is apparent the cleaning schedule is
largely a function of wastewater quality. A waste high in grease and fila-
mentous bacteria would be expected to result in greater frequency of probe
cleaning to maintain an accurate DO reading. The effectiveness of agitator
assemblies in mitigating cleaning requirements is questionable since such
assemblies are on all probes at the Valley plant.
Type of DO Probes
Some operators report the DO probe reservoirs are too small. With a
larger reservoir, recharging the probe with electrolyte is required less fre-
quently. Other operators report DO probe drift is excessive. One plant
engineer could not understand why the aeration blowers continued to run at
full capacity on an automatic DO control system, yet the tank DO concentra-
tion would not rise above 2 ppm. A check with a portable DO probe revealed
the actual tank DO was 5 ppm. Some probes are reported to have a much
greater propensity to drift than others, thus necessitating more frequent re-
calibration. The Simi Valley plant staff reports their probes have a tendency
to drift to the low side.
128
-------
Actual Incremental Labor Costs
DO probe in-plant maintenance cost and labor credit for valve modulation
and mixer or blower speed change reported in Table 16 may not be actual
costs or credits incurred on an incremental basis. A DO probe checking and
cleaning operation may require five minutes. Yet with a greater number of
probes in a given facility the labor cost per probe is certainly reduced.
Assuming the operator has some other maintenance function to perform at the
tanks or in the pipe gallery, where a throttle valve is located, it is question-
able whether the calculated costs or credits are true costs or credits for auto-
matic DO control equipment.
The superintendent of one of the plants discussed in the Appendix reported
allocating a certain amount of operator time for checking DO profiles in the
oxidation tanks. When asked how many man-hours per month were spent on
this operation, the superintendent responded that profiles were checked only
if the operator had sufficient time remaining after other duties were performed.
EFFECT OF CURRENT AND PROJECTED AVAILABILITY OF ENERGY ON VARIOUS
DO CONTROL SCHEMES
An examination of Table 16 and consideration of the above section on
power cost points out the significant impact of energy cost on dissolved
oxygen control systems. For the smaller 44 dmvs (1 mgd) plants, the power
savings of installing automatic DO equipment is not substantial. For the
larger plants, power savings becomes so important it can offset all capital
and operating costs of the automatic DO control system. The effect of current
availability of energy on various DO control schemes is largely a function of
the utility rate schedules and billing procedures. It has been demonstrated
that an automatic DO control system is power effective. It is estimated aera-
tion system power requirements can be reduced approximately 20 percent with
automatic DO control when compared to manual DO control.
In the past, power has been abundant and inexpensive. Today and in the
foreseeable future, the use of petroleum distillate or natural gas fuels is
becoming increasingly expensive. Reevaluation of energy consumption and
alternate energy sources are being explored. Solar power, wind power and
power from solid waste are being investigated as being alternate energy
sources. "Higher costs for natural gas and oil have removed these fuels from
consideration for new power plants, leaving primarily nuclear power and coal
to compete" (3). It is estimated that by the year 2000 almost one-half of the
electrical generating capacity of the United States will be nuclear (3).
able
The energy needs of this country will ensure that electric power is avail-
. Domestic energy demand has been increasing by 4-5 percent per year
129
-------
since 1947 (26) . We are not likely to use less, but rather more power in
our technologically oriented labor saving society. The differences will be in
the source and the cost. The source of power does not really affect aeration
air systems, but the cost does. A power cost inflation rate matching the
escalation of capital goods and labor will not affect the economic analysis
of DO control systems presented in this section. (Such an assumption is made
in all wastewater treatment plant economic analyses in accordance with the
guidelines of the Federal Register (8).) But a power cost escalation rate that
outstrips other DO system costs will more greatly support a DO control system
that is power effective .
Equipment Sizing and Service Life
Since the use of automatic DO control results in reduced overall air
requirements for the activated sludge process, equipment sizing and service
life can be reduced. For example, less costly air supply equipment (blowers,
valves, piping, etc.) can be used in new plant construction if an automatic
DO control system is installed. Alternatively, the service life of equipment
in existing plants with manual DO control will be extended if automatic DO
control is employed. As these potential savings were not included in the
economic analysis, the estimated annual savings are considered to be con-
servative .
130
-------
SECTION 9
DISSOLVED OXYGEN CONTROL SYSTEM SELECTION
The selection of a suitable DO control system for a specific plant appli-
cation is a complicated process involving many considerations. The selec-
tion can be facilitated by assigning relative priorities for factors involved
and then developing guidelines to aid in considering these factors.
IDENTIFICATION AND RANKING OF SELECTION FACTORS
A number of factors influencing the selection of a plant DO control sys-
tem are discussed in this report. These factors are identified and ranked by
degree of importance in Table 18. The ranking is offered as a general guide
to the order of priorities recommended for consideration in a systematic selec-
tion of a plant DO control system. Circumstances for a particular plant may
dictate that the priorities be altered.
TABLE 18. PRIORITY RANKING OF FACTORS AFFECTING CHOICE OF
DISSOLVED OXYGEN CONTROL SYSTEM
Priority
Factors
1
2
3
4
5
6
7
Existing equipment or conditions
Capital & O&M cost
Energy cost
Reliability
Plant staff capability
Effect on plant performance
Flexibility
Adaptability to various control modes
131
-------
DESCRIPTION OF FACTORS
The factors listed in Table 18 encompass many subfactors not readily
apparent. An attempt was made to define the factors with a minimal amount
of overlap into other factors; however, some interdependence is unavoid-
able .
Existing Equipment
A plant manager contemplating the addition of dissolved oxygen control
equipment is certainly influenced in his consideration of available control
system options by the equipment and design philosophy already present in
the plant. The presence of existing equipment has the dual effect of limiting
the selection of alternate DO control schemes, and of simultaneously facili-
tating the selection.
For example, the type of blower utilized affects the type of DO control
system which may be employed. A DO control system designed around the
characteristics of centrifugal blowers may not be suitable for positive dis-
placement or axial displacement blowers .
Some plants examined for possible inclusion in the case histories des-
cribed in the Appendix have the aeration air blowers committed to providing
air for other uses, such as air lift return sludge pumps. Since the compressors
are positive displacement machines, an automatic control system installed to
conserve air output would result in the excess air being applied to the air lift
pumps. This equipment design limits the options for other than manual DO
control. Conversion of the compressor drives to 2-speed units may be feas-
ible if it can be demonstrated the resultant power savings can offset the
capital cost of conversion.
Flow equalization basins and other situations resulting in constant flow
and loading can eliminate the need for automatic DO control. Coupled with
existing equipment are existing conditions. For example, the 1.8 m3/s (42
mgd) pulp and paper wastewater treatment plant in Pasadena, Texas, operated
by the Gulf Coast Waste Disposal Authority, has eight 2-speed 112 kW (150
hp) mixers in each of two oxidation tanks. The mixer speeds can be auto-
matically changed by a DO control system similar to that discussed in the
appendix for the St. Regis plant. However, Gulf Coast operators report mixer
speed change is more a function of solids suspension than dissolved oxygen
concentration. Accordingly, mixer speed changes are made by the operators,
as required, to keep the basin solids in suspension.
132
-------
Capital, Operating and Maintenance Costs
Cost of dissolved oxygen control equipment involves capital, operating,
and maintenance considerations. A decision based on capital cost alone is
economically unjustified without proper consideration of all operating and
maintenance costs involved.
As demonstrated in Section 8, cost analysis of DO control systems can
be done on an add-on basis if it is assumed that manual DO components are
common to both manual and automatic DO control systems. All contemplated
automatic DO control systems can then be compared incrementally. Since
service life estimates for cost-effectiveness analysis in the Federal Register
(8) for structures and process equipment are much greater than for instru-
ments and controls, a present worth cost-effective analysis over the recom-
mended planning period of 20 years should include sufficient funds to replace
the DO control equipment once.
Automatic DO control, as practiced in the field, involves selection of
only one or two control probes. However, the remaining probes must still be
maintained. Capital and operating cost could be significantly reduced if a
plant used fewer DO probes and more alternate receptacles.
Energy Cost
The effect of various utility rate schedules on the cost of DO control
systems has been discussed in Section 8. The applicable rate schedule has
a direct effect on the operating cost of a DO control system. One of the first
steps that should be taken in a systematic analysis of a suitable DO control
system is to examine the rate structure for a particular plant in terms of
energy and demand changes and see what effect a 20 percent reduction in air
requirements would have on the power bill.
Figure 71 illustrates a net savings can be expected for activated sludge
plants above 44 dm /s (I mgd) capacity using automatic DO control. Indi-
cations are that power rates less than those shown in the synthesized curve in
Figure 70 will result in a larger plant size being required before it is econom-
ical to install automatic DO control.
Utility rates and billing structures overlap into the cost factor
discussed above. Neglecting process performance improvement, the key
factor in deciding whether or not to install an automatic DO control system
is the projected power savings. The utility should also be consulted with
regard to future rate projections. Rate escalations may not necessarily keep
pace with inflation of capital and maintenance costs of equipment.
133
-------
Reliability
A major concern when considering various DO control scheme options is
system reliability. Many plants are designed for minimal attention using
sophisticated alarm and safety systems. The weakest link in an automatic
DO control system is the DO probe. If sufficient plant maintenance time
cannot be devoted to maintaining the probes, the DO control system will be
unreliable. On the other hand, a good preventive maintenance schedule for
the probes can produce consistently good results.
Plant Staff Capability
Plant staff capability refers to the caliber and training of in-plant person-
nel who will operate and maintain the DO control equipment. It is expected that
larger treatment plants will have specialized staff members trained for effi-
cient instrument maintenance. The prospect of adding a maintenance inten-
sive dissolved oxygen control system to such a plant is not met with as much
concern as in a smaller plant. Efficiency of DO probe maintenance will un-
doubtedly be higher for the larger 2.2-4.4 m3/s (50-100 mgd) plants.
Plant staff at one facility examined during this study reported the DO con-
trol system installed was too complicated to repair without calling in the
manufacturer at great expense. The equipment was, therefore, left in manual
mode when the automatic control circuits failed.
Effect on Plant Performance Parameters
Automatic DO control has been demonstrated to result in an improvement
in the performance parameters typically used to measure treatment efficiency.
A plant staff having difficulty meeting effluent requirements may find this
factor of particular importance as automatic DO control may be an alternative
to more expensive treatment modifications.
Flexibility
Many activated sludge plants are now designed with multiple operating
modes that require a flexible DO control system. Shifting operating modes
normally results in a redistribution of the oxygen demand in the tank. Effec-
tive DO control requires selecting the proper DO control probes . Depending
on the number of operating modes and the frequency of mode shift, additional
DO probes and receptacles may be required.
Adaptability to Various Control Modes
Selection of a DO control system should involve consideration of existing
and planned plant control modes, such as manual, local automatic and cen-
134
-------
tralized automatic. The type of DO control system selected should be suited
to the plant operational environment. Application of a complex DO control
system would be unjustified when proper instrument maintenance is unavail-
able or qualified operators cannot be obtained.
Future expansion plans for the plant may'involve computer control of
the plant processes. It may be justifiable to install computer-auto-manual
(CAM) controllers in the DO control system to permit future computer control
with minimal modifications.
SELECTION GUIDELINES
Existing Activated Sludge Plants
Based on the information in this manual, managers of an existing plant
with a flow capacity greater than 44 dm /s (1 mgd), a power rate equal to or
greater than that shown in Figure 70, and having centrifugal or axial blowers
for the diffused air system should consider adding automatic DO control if it is
not already installed. The recommended DO control system should control
blower output by discharge pressure and control air flow to the oxidation tanks
by DO concentration of the mixed liquor.
o
Managers of an existing plant of 44 dm /s (1 mgd) capacity or greater
and having aeration air furnished by mechanical aerators or positive dis-
placement blowers should also consider adding automatic DO control. How-
ever, economic considerations must be carefully evaluated for plants less
than 44 dm /s (1 mgd) capacity as the annual capital and maintenance cost
of automatic DO control equipment may be greater than the annual savings in
power cost. For mechanical aerators and positive displacement blowers, the
system recommended is two-speed or multiple-speed control based on DO
level, as described in this manual.
Managers of plants smaller than 44 dm3/s (1 mgd) should consider auto-
matic DO control as a means of stabilizing and improving plant performance.
Cost savings may be realized with the addition of automatic DO control equip-
ment, provided an adequate DO probe maintenance program is implemented at
the plant and the probes specified have a history of successful operation in
other plants.
Managers of plants with relatively constant flow and loadings should not
expect to derive any benefit from installing an automatic DO control system.
Also, automatic DO control is not recommended for plants with air require-
ments governed by maintaining solids in suspension rather than DO level,
when such air flows produce a DO level in excess of that required for effi-
cient treatment.
135
-------
New Activated Sludge Plants
Designers of new plants should perform a cost-effective analysis of
installing automatic DO control systems, considering the guidelines pre-
sented in this manual and the selection factors presented in Table 18. Where
possible, dynamic type blowers should be utilized to provide maximum flexi-
bility and range in DO control. Number and placement of DO probes should
influence oxidation tank design to minimize the high capital and operating
cost associated with these devices.
DO probes should be mounted for convenient access by maintenance
personnel. A rotating shaft, pivoted on the tank handrail and having a U-bar
holder similar to the Renton plant design (see Case History 1) is recommended.
Alternate prewired DO probe receptacles should be installed at locations
suitable for all contemplated modes of process control.
When economic or other considerations dictate the use of positive dis-
placement blowers or mechanical aerators, these devices should be equipped
with drives that conserve power at different speeds .
o
Designers of plants smaller than 44 dm /s (1 mgd) should consider not
using automatic DO control. Flow equalization basins or other flow and load-
ing stabilizers should be investigated as alternatives for these smaller plants.
136
-------
REFERENCES
1. Aberley, R. C., "Aeration Innovations", presented at the February 5-9,
1973, Australian Water and Wastewater Association Summer School,
held in Canberra, A.C.T.
2. Aeration in Wastewater Treatment (WPCF Manual of Practice No. 5),
Water Pollution Control Federation, 1971.
3. Booth, H. R. etal., "Nuclear Power Today", Chemical Engineering,
Oct. 13, 1975, pp 102-118.
4. Brown and Caldwell, "Process Design Manual for Nitrogen Control",
U.S. Environmental Protection Agency Technology Transfer, Oct.,
1975.
5. Buzzard, W. S., "Controlling Centrifugal Compressors , "Instrumenta-
tion Technology, Vol. 20, No. 11, Nov., 1973, pp 39-43.
6. Eckenfelder and O'Connor, Biological Waste Treatmeht, Pergamon Press,
New York, N.Y. , 1961.
7. Fair, G. M. et al. , Water and Wastewater Engineering, Vol. 2 , John
Wiley & Sons Inc., New York, N.Y. , 1968.
8. Federal Register, Vol. 39, No. 29, Feb. 11, 1974.
9. Flanagan, M. J., "Automation of the Activated Sludge Process", pro-
ceedings of the September 23-25, 1974, U. S. Environmental Protec-
tion Agency - Clemson University Workshop on Research Needs for
Automation of Wastewater Treatment Systems, held at Clemson , South
Carolina, pp 38-46.
10. Gibbon, D. L., ed., Aeration of Activated Sludge in Sewage Treatment,
Pergamon Press, Elmsford, N.Y., 1974.
11. Hanson, R. L., Walker, W. C. and Brown, J. C., Variations in Char-
acteristics of Wastewater Influent at the Mason Farm Wastewater
Treatment Plant, Chapel Hill, North Carolina, UNC Wastewater Re-
search Center Report No. 13, Dec., 1970.
137
-------
12. Horstkotte, G. A., Niles, D. G., Parker, D. S. andCaldwell, D. H.f
"Full Scale Testing of a Water Reclamation System" , Journal Water Pol-
lution Control Federation, Vol. 46, No. 1, 1974, pp 181-197.
13. Instrument Symbols and Identification, Instrument Society of America,
Pittsburg, Pa., 1973.
14. Lawrence, A. W- and McCarty, P. L., "Unified Basis for Biological
Treatment Design and Operation, "Journal of the Sanitary Engineering
Division. Proceedings of the American Society of Civil Engineers, Vol.
96, No. SA3, June, 1970.
15. Leary, R. O., Ernest, L. A. andKatz, W. J., "Effect of Oxygen Trans-
fer Capabilities on Wastewater Treatment Plant Performance", Journal
Water Pollution Control Federation, Vol. 40, No. 7, 1968.
16. Leary, R. O., Ernest, L. A. and Katz, W. J., "Full Scale Oxygen
Transfer Studies of Seven Diffuser Systems, "Journal Water Pollution
Control Federation, Vol. 41, No. 3, Part 1, 1969.
17. Lesperance, T. W. , "A Generalized Approach to Activated Sludge " ,
Water Works and Wastes Engineering, Vol. 2, 1965.
18. Lewis, W- K. and Whitman, W. C., "Principles of Gas Adsorption",
Industrial Engineering Chemistry, Vol. 16, 1924.
19. Meredith, W. D., "Dissolved Oxygen Control of Activated Sludge
Process", presented at the September 17-20, 1973 International Work-
shop, Instrumentation Control and Automation for Wastewater Treatment
Systems, held at London, England.
20. Metcalf & Eddy, Inc., Wastewater Engineering, McGraw-Hill, New
York, N.Y., 1972.
21. Nogaj, R. J., "Selecting Wastewater Aeration Equipment", Chemical
Engineering, Apr. 17, 1972.
22. Parker, D. S., Kaufman, W. J. and Jenkins, P., Characteristics of
Biological Floes in Turbulent Regimes, Sanitary Engineering Research
Laboratory, University of California, Berkeley, July, 1970.
23. Parker, D. S., Kaufman, W. J. and Jenkins, D., "Floe Breakup in
Turbulent Flocculation Processes", Journal of the Sanitary Engineering
Division, ASCE. Feb., 1972.
138
-------
24. Parker, D. S. , "Effects of Turbulence on Activated Sludge Effluent
Clarity", presented at the October 3, 1970, 12th Annual Northern
Regional Conference of the California Water Pollution Control Associa-
tion, held at Stockton, Calif.
25. Parker, D. S. and Niles, D. G. , "Full-Scale Test Plant at Contra
Costa Turns Out Valuable Data on Advanced Treatment" , Bulletin of
the California Water Pollution Control Association, Vol. 9, No. 1,
1972.
26. Project Independence, A Summary, Federal Energy Administration,
Nov. , 1974, pg 2.
27. Recommended Standards for Sewage Works, Great Lakes-Upper Missi-
ssippi River Board of State Sanitary Engineers, Albany, Health Educa-
tion Service, 1973.
28. Roe, F., "Preaeration and Air Flocculation", Journal Water Pollution
Control Federation, Vol. 23, No. 2, 1951, pp 127-140.
29. Rollins, J. P., ed. , Compressed Air and Gas Handbook, 4th ed.,
Compressed Air and Gas Institute, New York, N.Y., 1973.
30. Rose, W. L. and Gorringe, R. E., "Deep Tank Extended Aeration
System at Atlantic Richfield Company", presented at the May 3, 1972,
Purdue Industrial Waste Conference, held at Lafayette, Indianna.
31. Roy F . Weston, Inc . , Process Design Manual for Upgrading Existing
Wastewater Treatment Plants , U . S . Environmental Protection Agency,
Oct. , 1974.
32. Ryder, R. A. , "Dissolved Oxygen Control of Activated Sludge Aeration",
Water Research, Vol. 6, Pergamon Press, 1972, pp 441-445.
33. Sacramento Municipal Utility Schedule Rate 47, May 1, 1974.
34. Sewage Treatment Plant Design (WPCF Manual of Practice No. 8),
Water Pollution Control Federation, 1959.
35. Shinskey, F. G.. Process Control Systems, McGraw Hill, 1967,
pp 204-229.
36. Smith, R., "Electrical Power Consumption for Municipal Wastewater
Treatment", U. S. Environmental Protection Agency, EPA-R2-73-281,
July, 1973.
139
-------
37. Stamberg, J. B., Bishop, D. F. , Hais, A. B. and Bennett, S. M. ,
"System Alternatives in Oxygen Activated Sludge" presented at the
October, 1972, 45th Annual Conference of the Water Pollution Control
Federation, held at Atlanta, Georgia.
38. Stewart, M. J., "Activated Sludge Process Variations, The Complete
Spectrum", presented at the September 26-27, 1963, Western Canada
Water and Sewage Conference, held at Saskatoon, Saskatchewan.
39. Union Carbide Corporation, "Operating Experience and Design Criteria
for Unox Wastewater Treatment Systems" , presented at the February 29 •
March 1, 1972, Design Seminar for Wastewater Treatment Facilities,
held at New York, N.Y.
40. Woodruff, P. H. , "Dissolved Oxygen Control of the Activated Sludge
Process", Progress in Water Technology, Vol. 6, Pergamon Press, 1974
140
-------
APPENDIX
CASE HISTORIES OF DISSOLVED OXYGEN CONTROL SYSTEM
PERFORMANCE, OPERATIONAL AND MAINTENANCE DATA
141
-------
CONTENTS
Case History 1
Renton Wastewater Treatment Plant, Washington 143
Case History 2
Palo Alto Water Quality Control Plant, California 149
Case History 3
Rye Meads Sewage Purification Work, Hertfordshire, England .... 159
Case History 4
The City of Oxford Sewage Works, England 164
Case History 5
Valley Community Services District Wastewater Treatment Plant,
California 167
Case History 6
Reno-Sparks Joint Water Pollution Control Plant, Sparks, Nevada . . 173
Case History 7
Simi Valley Water Quality Control Plant, California 180
Case History 8
San Francisco International Airport Water Quality Control Plant. ... 188
Case History 9
St. Regis Wastewater Treatment Plant, Sartell, Minnesota 194
Case History 10
Long Beach Water Renovation Plant, Long Beach, California 199
Case History 11
San Jose-Santa Clara Water Pollution Control Plant, San Jose,
California . „ 205
Case History 12
Cranston Water Pollution Control Facility, Cranston, Rhode Island . 211
References for Appendix 216
142
-------
CASE HISTORY 1
RENTON WASTEWATER TREATMENT PLANT, WASHINGTON
Description of Aeration and Dissolved Oxygen Control System
The Renton plant, located near Seattle, Washington, commenced operation
in June 1965 and was enlarged in 1973 to an average dry weather flow (ADWF)
capacity of 1. 6 m^/s (36 mgd). Two oxidation tanks are provided, each with
four passes. Air is supplied by six single stage centrifugal blowers and intro-
duced through two headers in each tank. Each header serves two passes.
Currently, the Renton plant is equipped with three 5.7 m^/s (12,000
scfm) , 370 kW (500 hp) blowers, and three 6.6 m3/s (14,000 scfm) , 450 kW
(600 hp) blowers. All blowers are single stage centrifugal types supplied
with 4160 volt power. The 370 kW (500 hp) blowers were installed in 1963
and are driven by synchronous motors. The 450 kW (600 hp) blowers were
installed in 1973 and are driven by squirrel cage induction motors. All
blowers are located in a temperature and humidity controlled gallery and
utilize finely filtered air.
An automatic dissolved oxygen (DO) control system is provided which
incorporates a pressure control loop in the blower feed manifold and a DO
regulated flow control loop for each of the four tank headers. Three probes
are installed in each of four passes in each oxidation tank for a total of 12
probes per tank. An instrumentation and control diagram of the DO control
system is shown in Figure A-l. Components include the following:
0 Single stage centrifugal blowers with individual suction throttle
valves, and flow regulated surge control systems (6 ea.)
0 Blower discharge manifold pressure control loop with pressure
transmitter and pressure indicating controller (1 ea.)
0 DO analyzers and probes (24 ea.)
0 12 point DO strip chart recorders (one recorder for each tank) (2 ea.)
0 DO probe selector switches (one per air header) (4 ea.)
143
-------
'A IT) (AIT) (AIT) AIT) (AH (AH
AERATION
BLOWER
(Typ)
JICY"- /FY
IT) (AIT)^n(AIT)(An
OTHER 5
BLOWERS
SAME AS
ABOVE
i — ri_i --
HEADERS TO
2 nd TANK
TO OTHER
5 BLOWERS
NOTE: INSTRUMENTATION OF
SECOND TANK SAME
AS FIRST TANK
Figure A-1 . Automatic dissolved oxygen control system - Renton, Washington.
-------
0 DO controllers (4 ea.)
0 Flow control loops for each header, including orifice plate, square
root extractor, recorder, totalizer, indicating controller and piston
operated butterfly valve (4 ea.)
Operation
Currently, the Renton plant receives insufficient loading to warrant use
of both oxidation tanks. However, both tanks are fully instrumented for auto-
matic DO control. Via the appropriate probe selector switch (HS) , the opera-
tor selects one DO probe in each two-pass tank section and uses that probe
to control the air flow rate through the corresponding supply header. All DO
probes are continuously monitored on strip chart recorder AR, and the selected
control probes may be changed at any time . The output of the selected probe
is transmitted to oxygen controller AIC that provides an output to vary the set
point of flow controller FIC as required. The flow controller modulates a
butterfly valve in the air flow header in accordance with the computed set
flow rate .
The DO in the first two passes is typically maintained at 1.5 ppm utiliz-
ing one of the last two probes in the second pass for DO measurement. DO
in the last two passes is usually maintained at 2-2.5 ppm using one of the
last probes in the fourth pass for DO measurement.
A constant pressure of 55 kPa (8 psig) is maintained in the blower dis-
charge manifold by simultaneously throttling all blower suction valves through
a pressure indicating controller (PIC). A discharge flow regulated, surge con-
trol s'ystem with a load control override (load limit controller) is provided for
each blower. If the pressure indicating controller (PIC) calls for the blower
throttling valve to be open for a period of time such that the blower power
draw becomes excessive, the power indicating controller (JIG) will override
the pressure indicating controller and throttle the blower to a lower, less
power consuming discharge rate.
Performance
The Renton plant has continuously operated under automatic DO control
since 1967 with the exception of a 13 month period beginning March 1970,
when the older oxidation tank was modified from a two-pass to a four-pass
structure. During this period, all influent was diverted to the new oxidation
tank. However, dissolved oxygen probes had not yet been installed in the
new tank so DO data collection was performed by the plant laboratory. Set
points on the header FIC's were determined by the laboratory using the copper
sulfate-sulfamic acid flocculation modification to the Winkler test for DO.
The FIC set points were varied once or twice per day according to directions
145
-------
from the laboratory. Other analysis performed by the laboratory included
influent and effluent BOD, COD, SVI, and 30-minute settleability of the
sludge. These analyses were completed once per day from 24-hour composite
wastewater samples. Detailed data has been reported on three months of
manual DO control in 1970 compared to the same three months of automatic
DO control in 1971 (A-6).
Frequency distribution plots of BOD removal efficiency indicate that the
removal efficiency was consistently high under automatic control, while
efficiencies varied considerably under manual control. Data compiled on
BOD removal efficiency and other performance parameters are shown in Table
A-l. As indicated, the arithmetic means for SVI were 332 and 86, respective-
ly, for manual and automatic control. This represents almost a four fold
improvement by automatic DO control. However, bulking problems reported
to have occurred during the manual control period would have a significant
effect on SVI. Plant personnel do not attribute the occurrence of a bulking
problem during the manual DO control period to use of a manual DO control
system; however, the operators claim that under automatic DO control bulking
is not prevalent and sludge settling and handling characteristics are better
than when manual DO control is employed.
It has been reported in the literature that automatic DO control signifi-
cantly reduces the air required for secondary treatment (A-2). This is con-
firmed by the Renton Plant data as shown on Table A-l.
Maintenance
Maintenance of the DO control system with associated blowers has been
judged by the plant maintenance superintendent to require minimal labor and
material costs. However, some problems have been experienced with DO
probe drift and moisture accumulation in the probe plugs. At least two out of
12 probes in the operating oxidation tank have displayed excessive drift
uncorrectable by recalibration. Several probes have been taken back by the
manufacturer to determine the cause of the problem.
The DO probes in the operating oxidation tanks are cleaned and calibrated
once a week and recharged about every eight months. Cleaning and calibra-
tion of 12 probes normally requires 1-1/2 man-hours; recharging 12 probes
normally requires two man-hours . Instrument technicians performing such
maintenance earn $1100-$1300 per month.
The average annual maintenance cost of each blower is approximately
$1000. Of this, $500 per year per blower is spent on parts and materials and
$500 per year on labor. Typical annual maintenance involves changing the oil
and oil filters and inspecting the bearings. Each blower is completely over-
hauled every five years.
146
-------
TABLE A-1 . PERFORMANCE COMPARISON OF MANUAL AND AUTOMATIC
DISSOLVED OXYGEN CONTROL AT THE RENTON WASTEWATER
TREATMENT PLANT3
Parameter
BOD removal efficiency,
percent"
Q
Sludge volume index
Air supplied per unit quantity
of influent,
3/ 3
m /m
(cf/gal)
Air supplied per unit quantity
of BOD removal,
m /kg
(cf/lb)
BOD removed per
blower, kWh,f'h
kg/kWh
(Ib/kWh)
Manual
85
3329
9.3
(1.2)
137
(2190)
0.40
(0.88)
Automatic
96
86
8.2
(l.D
86
(1400)
0.63
(1.4)
Percent
improvement
11
74
12
37
58
Data from October, November, December of 1970 and 1971.
-, f)
Average daily flow - 1.07 m /s (24.5 mgd) .
Average BOD loading - 3.95 mg/m3/s (21.3 lb/1000 cf/day).
c 3
Average daily flow - 1.19 m /s (27.1 mgd).
Average BOD loading - 5.86 mg/m3/s (31.6 lb/1000 cf/day).
d
Geometric mean.
Q
Arithmetic mean.
Average.
Bulking problems occurr.ed.
hBased on 15.2 dm3/s/kW.
147
-------
Blower maintenance is usually performed by a journeyman machinist
assisted by a maintenance mechanic. The plant maintenance supervisor rates
the journeyman machinist as a highly skilled mechanic and the maintenance
mechanic as a semiskilled mechanic. A journeyman machinist earns $1100-
$1200 per month and a maintenance mechanic earns $970-$! 100 per month.
Safety and Emergency Procedures
The aeration system blowers are monitored and protected by an elaborate
instrumentation and control system. The units may be started from the main
plant console only when a permissive start light notifies the operator that all
starting conditions have been met. The switchgear is locked out when any
maintenance is being performed on a blower.
Each blower has a flow controlled surge control system and a load limit
controller. A malfunction of the surge control system causes a blower shut-
down by the vibration or bearing temperature monitoring system. Functioning
of the load limit controller is described in the previous section on operation.
During a plant power failure, air flow to the oxidation tanks ceases as
all blowers coast to a stop. Auxiliary air operated oil lube pumps supply the
blowers with oil while they are slowing down. Air is supplied to the lube pump
receivers through air compressors operated by an auxiliary diesel engine
powered generator station.
Loss of a sensor or other control system component is not serious
because of the redundancy in the design of the aeration equipment. For
example , a drifting DO probe can readily be replaced by another probe for
the control function.
148
-------
CASE HISTORY 2
PALO ALTO WATER QUALITY CONTROL PLANT, CALIFORNIA
Description of Aeration and Dissolved Oxygen Control System
The Palo Alto Water Quality Control Plant is an activated sludge facility
with a current ADWF capacity of 1.5 m3/s (34 mgd) and an average wet weath-
er flow (AWWF) capacity of 2 .3 m3/s (53 mgd). Four oxidation tanks are
provided with piping arranged for plug flow or re aeration modes of operation.
Air is supplied by three positive displacement air blowers and delivered to
each oxidation tank through a sparger ring. A 37 kW (50 hp) fixed speed
mechanical mixer in each tank is used to mix the rising air bubbles with the
mixed liquor.
Each of the three air blowers are 224 kW (300 hp), 550 rpm wound rotor
motor driven units designed to deliver 3.02 m3/s (6400 scfm) of air at 55 kPa
(8 psig). The blowers are positive displacement lobe type units installed in
1972. Two saturable core reactor variable speed drive units are used to vary
the speeds of the three air blower motors. One drive unit is dedicated to one
blower, while the other drive unit can be switched to either of the other
blowers using transfer contactors. The blowers discharge into a common
manifold that delivers air to each oxidation tank through separate 0.36m
(14-inch) risers.
A DO probe is installed in each oxidation tank halfway between the me-
chanical aerator units and the tank dividing wall. A portable DO probe is
also available to measure DO concentrations in the tanks. An instrumentation
and control diagram of the DO control system is shown in Figure A-2. Compo-
nents include the following:
0 Positive displacement air blowers with saturated reactor core variable
speed drive (3 ea.)
0 Manual flow control stations for each tank including orifice plate,
flow transmitter, square root extractor, flow indicator, motor oper-
ated butterfly valve and a remote manually operated valve position
controller (4 ea.)
149
-------
Cn
o
37 kW (50 HP)
MECHANICAL
MIXER (Typ)
SPARGE
RING (Typ)-
iDd
(AE
TANK
DC
TANK
JDO
(AE
TANK
TANK
220 kW(300HP)
BLOWER (Typ)
^® >
IV M~ L
v_y /TTN ^-^
-SATURA-TEO REACTOR CORE
VARIABLE SPEED DRIVE(Typ)
Figure A-2. Dissolved oxygen control system - Palo Alto, California.
-------
° Flow recorder for total flow delivered to all four oxidation tanks
(1 ea.)
0 DO probes, agitators and analyzers (4 ea.)
0 A fixed speed mechanical mixer in each oxidation tank (4 ea.)
0 Single channel DO recorder with a manually operated DO probe
selector switch (1 ea.)
Operation - Remote Manual
Normally, the Palo Alto plant is operated in the conventional activated
sludge mode. Each oxidation tank simultaneously receives primary effluent
and discharges mixed liquor to an associated final tank. Dissolved oxygen
concentration in each oxidation tank is indicated on a control panel in the
plant operations building. A single channel recorder may be switched from
tank to tank to record the DO level.
Based on DO concentration in each tank, the operator modulates the
blower motor speed on the motor operated butterfly valve in the corresponding
tank air feed header. Normally, only two blowers are operating. Primary
control of DO in the tanks is achieved with blower speed modulation, while
secondary control is made with remote operated header feed valves. Blower
speed is typically altered three times per day while the header valve positions
are changed two or three times per day. The valve in the tank header farthest
from the blowers is normally left wide open. Dissolved oxygen concentration
in the oxidation tanks is usually maintained at 0.5-1.0 ppm.
Operation - Remote Semiautomatic
In 1973 , Systems Control Inc. and the Palo Alto Water Quality Control
Plant, under a grant by the EPA and the State of California Water Resources
Control Board, conducted a study to compare manual vs. digital computer
control of the DO control system (A-5). Following four weeks of monitoring
normal control procedures in the manual mode, a digital computer system was
integrated into the DO control system with a programmed DO control algorithm.
The computer received a 4-20 mA DO signal from each probe analyzer, com-
puted any changes required in blower flow rates, and typed out such changes
on a teletype. A change in computed air flow rates of more than 47 dm^/s
(100 scfm) was required before a control change message was typed. Since
the operator was still required to perform the blower speed change requested,
this type of control is most accurately designated as computer-assisted open
loop.
151
-------
The DO control algorithm used in the test was an incremental, propor-
tional plus integral, process control program designed to compute the required
change in air flow rate. On occasion, such change commands required the
operator to place blowers in or out of service.
The testing program at Palo Alto was divided into three stages summarized
in Table A-2. During all three stages of the study, the header feed control
valves were left in a fixed position.
During the first stage of approximately four weeks, the plant was oper-
ated in the remote manual mode as previously described. Blower speed was
changed twice daily.
TABLE A-2.
MANUAL AND SEMIAUTOMATIC DISSOLVED OXYGEN CONTROL
TESTING PROGRAM AT PALO ALTO, 1973
Stage
1
2
3
Phase
-
1
2
1
2
3
Duration
4 weeks
3-1/2 weeks
3 days
4 weeks
3-1/2 weeks
3 days
Remarks
Remote manual mode of
operation.
Infrequent data collection
under remote manual mode.
Frequent data collection
under remote manual mode
for average and extreme
operating conditions..
Process stabilization under
semiautomatic mode.
Infrequent data collection
under semiautomatic mode.
Frequent data collection
under semiautomatic mode
for average and extreme
operating conditions.
During the next stage of the study, from July 9-August 7, 1973, data was
collected on plant performance for both a three and one-half week nonintensive
and a 3-day intensive collection phase. The intensive period of data collec-
tion encompassed the extreme operating conditions of the plant as well as an
average condition. During the intensive period, all relevant data was col-
lected on a two-hour basis.
152
-------
During the final stage of the study, the semiautomatic DO control system,
as previously described, was activated. After a four-week period of process
stabilization to the new control mode, intensive and nonintensive study
phases of process performance were made from September 1 6 to October 11,
1973. Data collection phase durations for the final study step were identical
to those of the previous step.
During both stage 1 and stage 2 tests, air flow applied to the oxidation
tanks was totalized and BODg in the primary effluent was monitored. Samples
were taken from the primary effluent at four-hour intervals and combined for
composite BODr analysis every 24 hours.
Performance
Figures A-3 and A-4 illustrate the difference between the manual and
semiautomatic DO control systems in maintaining a 1.0 ppm DO set point
during the intensive testing periods. Under semiautomatic control, Figure A-4
shows the controller maintained the DO at or near set point, although sub-
stantial changes in air flow occurred due to considerable loading variation.
Under manual control, wide excursions in DO concentration were experienced
as shown in Figure A-3.
Average values of plant operating variables during the first and second
phases of test stages 1 and 2 are shown in Table A-3. Although the percent
improvements for BODs and COD removal were relatively slight, suspended
solids and total organic carbon percentage removal under semiautomatic DO
control show a marked improvement compared to manual DO control.
Assuming a linear relationship between blower output and power consump-
tion and a power cost of $.01/kWh, Systems Control Inc. reports power costs
corresponding to the air required for each mode as shown in Table A-3. The
computed annual cost savings for the automatic as compared to the manual DO
control system amounts to $5400.
Maintenance
DO Probes and Analyzers--
According to the chief operator at the Palo Alto plant, DO probe calibra-
tion is typically performed once per month. During the course of the study,
it was determined that preventive maintenance was required every two weeks
on the probes. One man-hour was required to check all probes. Approximate-
ly once every three months, about four man-hours were required to thoroughly
check, clean and recalibrate all probes. Recharging requires about 2 man-
hours per year for all tour probes.
153
-------
Ln
I2.OOO
IO,OOO
8,000
5
J*-
w
- 6,000
O
u.
cc
- 4,000
^
2,000
-
—
,
0
1
I
i
1
1
1
1
1
1
1
1
\
/
/
1
/
/
/
/
\
\
\
\
\
\
\
\
\
\
\
\
\
.
'\
1 V
/ v^
/ \
/—DISSOLVED OXYGEN
! \
i
i
i
i
\
\
\
\
—
' » ^AIR FLOW
/v
/ \
{ ' \ "
V V
i
i
i
1
(cfm)(472X ICT4) = m3/s
6
1
12 18 0
SUNDAY
8/5/73
i
6
\ /
\ /
\
\
\
\
\
\
\
\
\
\
\
\
\
\
\A ,*
V*' x
\
i i 1
12 18 0
MO NDA Y
8/6/73
^
/
/
/
/
/
/
/
/
/
(
/
1
/
/
/
y
i
6
—
\
\
\
\ —
\
\
\
\
\
\ r
\ / —
I /-"
1 1
a.u
E
4.0 §:
-
O
^
«5
Ct
3.0 ^.
LU
^
O
Cj
LU
2.0 £
o
Q
LU
^j
O
1.0 ^
Q
n
12 18 0
TUESDAY
8/7/73
D47E - TIME OF DAY
Figure A-3. Manual dissolved oxygen control - dissolved oxygen and
air flow as a function of time at Palo Alto (A-5) .
-------
12,000
On
Cn
10,000
8,000 ~
6,000 -
O
-j
k.
4,OOO -
2,000 -
DISSOLVED OXYGEN
cfm)(4.72Xlcf 4 ) = m3/s
0.0
12
MONDAY
9/10/73
DA TE - TIME OF DAY
Figure A-H. Semiautomatic dissolved oxygen control - dissolved oxygen
and air flow as a function of time at Palo Alto (A.-5) .
-------
TABLE A-3 PERFORMANCE COMPARISON OF MANUAL AND REMOTE
SEMIAUTOMATIC DISSOLVED OXYGEN CONTROL AT THE
PALO ALTO WATER QUALITY CONTROL PLANT
Parameter
BOD removal efficiency,
percent
SS removal efficiency, percent
TOG removal efficiency, percent
COD removal efficiency.
percent
Air supplied per unit quantity
of influent,
3, 3
m /m
(cf/gal)
Air supplied per unit quantity
of BOD removed, d
m /kg
(cf/lb)
Q
BOD removed per blower, kWh,
g/kWh
(Ib/kWh)
Manual
84
46
53
63
3.4
(0.45)
33
(520)
1.3
(2.9)
c
Semiautomatic
84
53
60
64
3.4
(0.45)
28
(450)
1.6
(3.4)
Percent
improvement
none
7
7
1
none
15
23
Data from Systems Control Inc. Study, 1973. Operating mode for manual
and semiautomatic was contact stabilization. Test duration was four weeks
for each control mode.
Average daily flow - 1.05 m /s (24.0 mgd) .
Average BOD applied to oxidation tanks - 4.52 mg/mVs (24.4 lb/1000
cf/day).
c 3
Average daily flow - 1.03 m /s (23.6 mgd) .
Average BOD applied to oxidation tanks - 5.23 mg/m3/s (28.2 lb/1000
cf/day).
Computed from total air supplied over testing period, average BOD in
primary effluent and reported BOD removal efficiency-
e 3
Based on 13.5 dm /s/kW (21.3 scfm/hp).
156
-------
The DO analyzers were reported by Systems Control Inc. to produce an
electrically noisy signal with inherent oscillating variation about a specific
value. Electronic filtering was employed to correct this problem.
Blowers --
Blower maintenance, which is done by a mechanic earning $1090 per
month, has been essentially preventive since the units were installed so
recently. Every six months, all three blowers are cleaned and the oil is
o
changed. Approximately 38 dm (10 gallons) of oil are required' for each
blower. One man-day of labor is required for this semiannual service on all
three machines.
Once per month the primary and secondary collector rings on the wound
rotor motors are cleaned to remove brush deposits. Two man-hours per blower
are required for this operation.
Every six months, the drive coupling bearings are cleaned and repacked.
The total time required is two man-hours per blower.
Based on a labor cost of $1090 per month plus 50 percent additional for
fringe benefits, the annual maintenance cost of the three blowers is estimated
at $350. Special maintenance, such as overhaul, would substantially add to
this cost.
Computer--
Although no problems were reported on computer downtime during the
stage 3 test, the computer did fail at other times during other phases of the
testing program. Over a period of about 13 months, approximately 20 compu-
ter failures were reported, of which about 50 percent would have resulted in
loss of the DO controller. If the DO controller had been operating, the data
loss affecting DO control would have been 49 hours. Since the computer
operated continuously for about 14 months, the computer was capable of
effecting automatic DO control 99.5 percent of the time.
Computer system failures occurred predominantly with the peripheral
devices. The disk memory accounted for 70 percent of the total computer
downtime.
Safety and Emergency Procedures
The aeration blowers are furnished with safety switches for high oil tem-
perature , high oil pressure, low oil pressure, excessive current, and low
water pressure. Activation of any of these switches will automatically shut
off the blower and cause an alarm.
157
-------
All blowers must be locally started. Thus, following a shutdown through
an alarm condition or power failure, remote starting is not available. Loss
of blowers through a power failure will effectively stop activated sludge
treatment, as no auxiliary power source exists.
158
-------
CASE HISTORY 3
RYE MEADS SEWAGE PURIFICATION WORKS, HERTFORDSHIRE, ENGLAND
Description of Aeration and Dissolved Oxygen Control System
The Rye Meads Works of the Middle Lee Regional Drainage Scheme,
located in southeast Hertfordshire, is a tertiary treatment plant treating a
dry weather flow of 0.692 m^/s (15.8 mgd). It uses conventional primary
treatment, diffused air activated sludge and sand filters and lagoons for
effluent polishing. Eight oxidation tanks are provided, each with four passes.
Air is supplied by thirteen lobe type, rotary, positive displacement blowers.
Eleven blowers are rated at 1. 6 m^/s (3500 scfm) each, while the remaining
two deliver 0.71 m /s (1500 scfm) each. Air is introduced through 0.18 m
(7-inch) dome type diffusers. The diffusers are uniformly spaced along the
base of seven tanks and distributed in a tapered pattern in the eighth tank.
The blowers discharge air at 46 kPa (6.7 psig) into a common manifold
serving all eight oxidation tanks. A separate header with throttling butterfly
valve is furnished for each tank pass to deliver air from the manifold to the
diffusers.
The secondary process is designed to produce a fully nitrified effluent.
One of the plant discharge requirements is an ammonia nitrogen concentration
of less than 10 ppm at all times.
During the period July-December 1967, the Water Pollution Research
Laboratory (WPRL) conducted a 6-month test of automatic dissolved oxygen
control at the Rye Meads Works. Tanks 5 and 6 were selected as control
units and tank 8 was used as the experimental unit. Two dissolved oxygen
probes were equally divided between tanks 5 and 6.
A breadboard type of DO control system was used that involved sequential
scanning and recording of all 16 DO probe outputs with preselection of four
probes in tank 8 for control. The output of the four control probes was com-
pared to four separate reference voltages corresponding to the desired oxygen
concentrations. Utilizing a twin-coil relay for the butterfly valve in each of
four feed headers, a 3-way, double solenoid valve was used to effect changes
in pilot line pressure to a positioner on each spring opposed diaphragm valve.
The valve was thereby modulated in accordance with the desired DO set point
159
-------
for the corresponding tank pass. An instrumentation and control diagram of
the DO control system is shown in Figure A-5. Components include the fol-
lowing:
0 Aeration blowers (13 ea.)
0 DO electrode assemblies (16 ea.)
0 Multipoint potentiometric recorder fitted with a retransmitting slide
wire driven in synchronism with a uniselector (1 ea.)
0 Twin coil relays (Sea.)
0 Double solenoid 3-way valves (4 ea.)
0 Spring opposed diaphragm operated butterfly valves with positioner
(4 ea.)
Operation
During the experiments at Rye Meads, tanks 5 and 6 were used as com-
parison tanks, and tank 8 was used as the experimental unit. Tank 7 was not
operated during this time, and the remaining four tanks were not involved in
the experiments.
Tank 8 was completely isolated from tanks 5 and 6 by having its own final
clarifier and return sludge system. An unfortunate aspect of the experiment is
that tank 8 was designed for tapered aeration, while the comparison tanks 5
and 6 had conventional aeration. Thus, judgments regarding the improvement
on process performance must consider the effect of tapered aeration as well
as automatic DO control in tank 8.
Performance
Performance of the activated sludge system during the study period under
manual and automatic DO control is compared in Table A-4. The WPRL con-
cluded, "using a 4-point control system, it was possible to maintain a more
constant level of dissolved oxygen in the aeration tank using about 20 per-
cent less air than in the comparison plant while still producing a high quality,
fully nitrified effluent" (A-2).
An experiment was also performed at the Rye Meads Works to assess the
effect of automatic DO control at higher loadings. Influent flow to tanks 5,
6 and 8 was increased 11 percent, thereby reducing-the retention time to 8.5
hours. After seven weeks of operation in this mode, there was a significant
deterioration in nitrification, particularly in tank 8, which was under auto-
160
-------
en
AIR
AIR, l03kPa(!5psi)
(Typ,1
AERATION
BLOWER (Typ)
INPUTS FROM
OTHER 8
PROBES
A-
OXIDAT ION TANK NO. 8
AIR TO OTHER
OXIDATION TANKS
Figure A-5. Automatic dissolved oxygen control system - Rye Meads, England.
-------
TABLE A-4 PERFORMANCE COMPARISON OF MANUAL AND AUTOMATIC
DISSOLVED OXYGEN CONTROL AT THE RYE MEADS SEWAGE
PURIFICATION WORKS3
Parameter
BOD removal efficiency.
percent
Suspended solids removal
efficiency, percent
Sludge volume index
Air supplied per unit quantity
of influent,
3, 3
m /m
(cf/gal)
Air supplied per unit quantity
of BOD removed,
3/
m kg
(cf/lb )
Manual
(tank 6)
97
91
45
14
(1.8)
88
(1400)
c
Automatic
(tank 8)
98
92
79d
11
(1.5)
71
(1100)
Percent
improvement
1
1
-76
21
19
Data from six-month study July - December, 1967.
i Q
Average daily flow - 0.11 m /s (2.58 mgd) .
Average BOD loading - 4.71 mg/m /s (25.4 lb/1000 cf/day) .
c 3
Average daily flow - 0.11 m /s (2.48 mgd) .
Average BOD loading - 4.56 mg/m3/s (24.6 lb/1000 cf/day).
Higher SVI attributed to shearing action of different type return sludge
pumps used on tank 8 compared to tank 6.
162
-------
matic DO control. Nitrification difficulty in all three tanks was attributed to
unidentified inhibitory substances.
Maintenance
Maintenance costs of the DO control system and blowers were not made
available by the Laboratory Services Division of the Water Research Center.
163
-------
CASE HISTORY 4
THE CITY OF OXFORD SEWAGE WORKS, ENGLAND (A-3)
Description of Aeration and Dissolved Oxygen Control System
The diffused air activated sludge plant at Oxford was started up in 1969
and the automatic DO control system was placed in operation in December
1970. Eight parallel oxidation tanks are provided, and the aeration air is
supplied by two 75 kW (100 hp) variable speed blowers and one 74 kW (100 hp)
fixed speed blower. Eight DO probes of the polarographic type are installed
in lanes (passes) 4 and 5. The four probes in each lane are located at equal
intervals.
An automatic DO control system is provided to regulate air addition to
the oxidation tank to maintain a DO level of 1 ppm at the outlet end of lane 4
or 5. An instrumentation and control diagram of the DO control system is
shown in Figure A-6. Components include the following:
0 Variable speed centrifugal blowers (2 ea.)
0 Fixed speed centrifugal blower (1 ea.)
0 DO probes and analyzers (8 ea.)
DO controller (1 ea.)
Operation
The output from each DO probe is fed to an eight-point chart recorder (AR).
A selector switch (HS) is provided to select one of the DO signals as the
controlled variable input to controller AIC. Usually, one of the two probes
at the outlet end of the oxidation tank is used for control purposes. The out-
put from each tank outlet probe is fed to a summing indicator (AI) fitted with
high-low alarm contacts set to detect a difference of ± 0.2 ppm of DO between
the two readings. The occurrence of a DO difference alarm could signify
drifting or membrane failure of one of the DO probes.
The output from controller AIC is fed to the blower variable speed drive
units to regulate the addition of air to the oxidation tanks. The controller is
164
-------
tn
V
I
I
r
_
TO OTHER VAR.j
TSPEED BLOWER!
A IR
AIR
AIT) (AIT) (AIT AIT (AIT) (AIT) (AIT) (AIT
VARIABLE SPEED
BLOWER (Typ. 2)
FIXED SPEED
BLOWER
ADO
OXIDATION TANK
Figure A-6. Automatic dissolved oxygen control system - Oxford, England.
-------
set to maintain a DO level of 1 ppm at the outlet end of the oxidation tanks.
The minimum output signal from controller AIC is provided by regulator HIK to
ensure sufficient air flow for adequate mixing of the mixed liquor even under
conditions of zero loading.
Performance
Comparisons between manual and automatic DO control have been con-
ducted at the Oxford plant. A number of experiments were conducted for peri-
ods ranging from seven days to several months. Care was taken to ensure
that the periods selected for the manual and automatic control experiments
provided similar plant loading conditions to permit meaningful performance
comparisons.
Under manual control, the speed of the operating blowers was preset and
not changed for the duration of each experiment. Under automatic control,
the speed of the operating blowers was modulated to maintain a constant DO
level of 1 ppm in the oxidation tanks. An average reduction in power consump'
tion of 20 percent was achieved with the automatic DO control system.
In the opinion of the manager of the Oxford plant, the savings in power
consumption through the use of automatic DO control was not the most impor-
tant benefit. The main benefit was the maintenance of a healthy biomass
with good settling and dewatering characteristics. However, no data were
presented to support this claim.
Meredith (A-4) reports further work at Oxford established DO can be
maintained at a desired level over three-quarters of a pass (lane) length if
the DO probe is located one metre from the effluent weir and one metre below
the water surface.
Maintenance
The DO probes are removed from the mixed liquor once every two weeks
and washed with a jet of water from a portable wash bottle. Each probe is
recalibrated every month. The maximum membrane life of the DO probes has
been 18 months. The average membrane life has been six months.
166
-------
CASE HISTORY 5
VALLEY COMMUNITY SERVICES DISTRICT
WASTEWATER TREATMENT PLANT CALIFORNIA
Description of Aeration and Dissolved Oxygen Control System
The Valley Community Services plant has an average dry weather flow
(ADWF) capacity of 0.2 m^/s (4 mgd) and has been in operation since 1961.
The plant was upgraded in 1971 to provide tertiary treatment and incorporates
a flow equalization basin for hydraulic load balancing. One oxidation tank
is provided with several operational modes available. The operator may
select conventional plug flow, step feed, contact stabilization, tapered aera-
tion, two-pass plug flow or a combination of these activated sludge modes.
The plant is normally operated in a two-pass mode with primary effluent
introduced at three separate points and partial reaeration of return sludge in
the head end of the first pass.
Air is supplied by three 110 kW (150 hp) and one 150 kW (200 hp) centri-
fugal blower, all sized to deliver 1.4 m^/s (2900 scfm) each at a discharge
pressure of 52 kPa (7.5 psig). Under actual operating conditions, the plant
instruments indicate the blowers deliver approximately 1.9 m^/s (4000 scfm)
each at a discharge pressure of 31-34 kPa (4.5-5.0 psig).
Air is delivered to both oxidation tanks by a single air header, located in
a trench between the tanks. Twelve separate take-offs from the header supply
air to the diffusers for each tank. Each take-off has a manually controlled
butterfly valve permitting adjustments for tapered aeration.
An automatic DO control system is provided which incorporates a pressure
control loop in the blower feed manifold and a DO/flow cascade control loop
for the oxidation tank air header. Three DO probes are provided with agitator
assemblies. They are located along the outside wall of the second pass at
the influent, effluent and mid-point of the tank. An instrumentation and con-
trol diagram of the DO control system is shown in Figure A-7. Components
include the following:
Centrifugal blowers with individual suction throttle valves (4 ea.)
167
-------
TO OTHER
BLOWERS
r"
a
AIR
AERATION
BLOWER(Typ)
FROM
OTHER I
THREE "S '
BLOWERS
PRIMARY
EFFLUEMT
/\
RETURNED
ACTIVATED
SLUDGE
DO
(AE)
1 t
~L
~i
t t t
OX 1 DAT ION TANK
ML
—^
Figure A-7 . Automatic dissolved oxygen control system - Valley Community
Services District, California.
168
-------
0 Blower discharge manifold pressure control loop with pressure trans-
mitter and pressure indicating controller (1 ea.)
0 DO probes and analyzers (3 ea.)
0 Air pass header flow transmitter with flow control valve and flow
controller (1 ea.)
0 DO controller (1 ea.)
Operation
The operator selects one of the three DO probes as the control probe via
selector switch HS. The signal from the selected probe is fed to DO control-
ler AIC; the DO set point is varied according to the mode of operation of the
activated sludge process. The output of controller AIC is cascaded to the
set point of flow controller FIC. The total air flow to the oxidation tank is
measured by flowmeter FIT; the output of FIT is linearized by square root
extractor FY and fed to the FIC as the controlled variable. The output from
the FIC positions a motor-operated butterfly valve which regulates air flow to
the oxidation tank.
When the plant is operated in a two-pass mode, the DO in pass. No. 2
is typically maintained at 0. 6 ppm for the influent probe, 1.0 ppm for the
midpoint and 2 .5 ppm for the effluent probe. The oxygen concentration is
maintained at this level to ensure complete nitrification. Current plant
effluent requirements are 10 ppm of BOD and suspended solids, less than 1.0
ppm of ammonia and zero residual chlorine. A constant pressure of 31 kPa
(4.5 psig) is maintained in the blower discharge manifold by the simultaneous
throttling of all blower suction valves via pressure controller PIC. Because
the hydraulic load on the plant is uniform through flow equalization, the vari-
ation in air demand by the activated sludge process is not sufficient to war-
rant automatic starting and stopping of the aeration blowers. Likewise,
automatic surge control is not provided. The only time surge control is
required is during blower startup and shutdown. Thus, only manual control
is employed for blower starting, stopping, sequencing and surge control.
Performance
On June 10-11, 1975, the Valley Plant staff agreed to participate in two
24-hour tests devised to compare manual vs. automatic DO control. During
both tests, the plant was operated in its normal two pass mode on days
selected for typical flow and loading conditions. DO control was performed
using the probe located in the midpoint of the second pass since the plant
had operated for an extended period of time using this probe prior to the tests.
Normally, control from the last DO probe (effluent probe) is desired but this
169
-------
probe had been under repair for such a long time that the plant personnel had
become accustomed to operating off the midpoint probe.
Under manual control, the air feed header butterfly valve was modulated
as required based on hourly checks of DO concentration in the second pass.
Blowers were started and stopped as required to maintain the desired DO of
2-3 ppm in the tank..
Under automatic control, the air feed header butterfly valve was auto-
matically modulated by the DO controller (AIC) in accordance with a DO set
point of 2-3 ppm for the midpoint probe. Blowers were started and stopped as
for the manual test to ensure that an adequate amount of air was available.
The tests were subject to a constraint that a maximum of three blowers would
be permitted to run simultaneously due to power consumption considerations.
Table A-5 illustrates the results of the dissolved oxygen control study
test performed on June 10-11, 1975. As indicated by the data, automatic
control of DO resulted in a deterioration of virtually all parameters selected
for evaluating performance. A recheck of the data verified that the calcula-
tions were correct, although the deterioration was inconsistent with test
results at other plants presented in this report.
A factor that could have contributed to the performance deterioration of
automatic compared to manual DO contact is blower turndown. Normally
2-1/2 blowers are needed to furnish sufficient air for the oxidation tanks, but
the blowers can be throttled to only 80 percent of capacity. Plant experience
has demonstrated that throttling below 80 percent results in "free wheeling"
with consequent heat rise in the casing to the point where a unit drops off
line. The superintendent has installed a stop on the suction throttle valve of
each blower to prevent throttling below 80 percent. Thus, the automatic DO
control system "was constrained by blower throttling capability.
Since the plant operates with a flow equalization basin and experiences
little flow or loading rate variation, it is questionable whether an automatic
DO control system is necessary. Currently, the plant operates under manual
DO control and the superintendent is very pleased with the quality of the
effluent.
Maintenance
The centrifugal blowers at the Valley Plant have a preventive maintenance
schedule as shown in Table A-6.
Plant operating personnel have experienced eonsiderable difficulty with
the DO probe connections. Apparently, these connectors are aluminum and
rapidly corrode when submerged in the oxidation tank. Connector replacement
170
-------
TABLE A-5. PERFORMANCE COMPARISON OF MANUAL AND AUTOMATIC
DISSOLVED OXYGEN CONTROL OF THE VALLEY COMMUNITY
SERVICES DISTRICT WASTEWATER TREATMENT PLANT3
Parameter
BOD removal efficiency,
percent
Suspended solids removal
efficiency, percent
Sludge volume index
Air supplied per unit quantity
of influent,
3, 3
m /m
(cf/gal)
Air supplied per unit quantity
of BOD removed, d
m /kg
(cf/lb)
e
BOD removed per blower, kWh,
kg/kWh
(Ib/kWh)
,b
Manual
94
90
112
24
(3.2)
122
(1960)
0.38
(0.85)
c
Automatic
95
87
95
28
(3.7)
162
(2590)
0.27
(0.60)
Percent
improvement
1
-3
15
-17
-33
-29
Data from 24-hour tests on June 10-11, 1975.
i o
Average daily flow - 0.176 m /s (4.02 mgd) .
o
Average BOD applied to oxidation tanks - 6.80 mg/m /s (36.7 lb/1000
cf/day).
c 3
Average daily flow - 0.154 m /s (3.52 mgd) .
2
Average BOD applied to oxidation tanks - 5.19 mg/m /s (28.0 lb/1000
(cf/day) .
Computed from total air supplied over testing period and 24-hour composite
BOD of primary and secondary effluent.
Q
Based on actual power consumption for blowers 1, 2 and 3 and estimate
for blower 4.
171
-------
involved the entire connecting lead at a cost of $50/probe. The problem was
solved by fusing the connectors to the probe terminals and encasing the
assembly in plastic. Probe removal from the tank now involves bringing 4.6-
6.1m (15-20 feet) of connector line back to the lab as well, but the terminal
corrosion problem has been solved.
TABLE A-6. BLOWER PREVENTIVE MAINTENANCE SCHEDULE AT
THE VALLEY PLANT
Item
Lube blower bearings
Lube motor bearings
Check blast gate setting
Repack shaft coupling
Clean unit and retouch paint
Flush and lube blower and motor bearings
Turn lube cups one turn
Relube turn cups and turn 4 turns with
caps removed
Remove, check and repack bearings
a
Frequency
Q
A
S
A
A
2 yr
M
S
S
Labor time
15 minutes
15 minutes
20 minutes
30 minutes
2 hours
1 hour
5 minutes
10 minutes
2 hours
0 = quarterly, S = semiannually, A = annually, M = monthly
Initial experience with the DO probes indicated rubber bands and other
debris consistently entered the DO probe assembly and fouled the agitator
assembly. Under a recommendation from the supplier, the operating personnel
placed a 3 mm (1/8 inch) mesh screen around the probe openings and effective-
ly eliminated the fouling problem.
Probe membranes have been found to last 30-60 days. Evidence of mem-
brane failure is usually indicated by a pronounced drift in the DO output.
Membrane replacement requires less than one man-hour. DO probes are
cleaned daily by washing with a hose. Approximately twice per week one man-
hour is expended thoroughly cleaning and recalibrating all probes.
172
-------
CASE HISTORY 6
RENO-SPARKS JOINT WATER POLLUTION
CONTROL PLANT, SPARKS, NEVADA
Description of Aeration and Dissolved Oxygen Control System
The Reno-Sparks Joint Water Pollution Control Plant began operation in
1967 with an average dry weather design flow of 0.88 m^/s (20 mgd). Cur-
rently the plant is treating an average dry weather flow of 0.83 m3/s (19 mgd).
The plant is a diffused air, activated sludge treatment facility composed of
three separate treatment systems of equal capacity. Each system includes
a primary settling tank, oxidation tank, final settling tank and postoxidation
tank. Wastewater may be treated using conventional, three-pass, step feed,
tapered aeration and contact stabilization. Digested sludge or supernatant
may also be introduced into the oxidation tanks as a further process modifi-
cation. Normally the oxidation tanks are operated in a three-pass mode with
primary effluent and return sludge step fed in the first pass.
Diffused air for the oxidation and postoxidation tanks is furnished by
three 520 kW (700 hp) , 8.5 m3/s (18,000 scfm) , single stage, pedestal bear-
ing mounted, centrifugal blowers at a discharge pressure of approximately
50 kPa (7 psig). The blowers were installed in 1966. Each blower is furnished
with a diaphragm operated, flow controlled relief valve, a piston operated
butterfly check valve , and a manually operated butterfly suction valve.
The blowers discharge into a common manifold with four headers; one to
each of three oxidation tanks and one to the single postoxidation tank. Each
oxidation tank header delivers air to an oxidation tank through two smaller
headers; one for the first pass and one for both the second and third passes.
Air is introduced into the mixed liquor through diffusers mounted on swing arm
assemblies fitted with manually operated butterfly valves.
Air flow delivered to each oxidation tank is throttled by a flow control
loop receiving a flow set point from a DO controller. Throttling the tank air
feed header butterfly valves results in throttling the constant speed blowers,
as the blowers react to the change in downstream pressure by re stabilizing
at another discharge flow rate to maintain the same discharge pressure. A
flow regulated surge control system prevents blower surging by releasing the
blower discharge to atmosphere when flow reaches a limiting value. An
173
-------
instrumentation and control diagram of the DO control system is shown in
Figure A-8. Components include the following:
0 Single stage, constant speed, centrifugal blowers with flow regu-
lated surge control system (3 ea.)
o Oxidation tank and postoxidation tank supply header air flow con-
trol systems with flow tube, flow indicating transmitter, square root
extractor, totalizer, flow indicating controller and piston operated
butterfly valve (3 ea.)
o Dissolved oxygen probes with analyzer/transmitter (3 ea.)
o Dissolved oxygen recorders (3 ea.)
o Dissolved oxygen indicating controllers (3 ea.)
Operation
Although three blowers are provided at the Reno-Sparks plant, control
circuitry is designed to permit simultaneous operation of only two blowers,
thus reserving standby capacity and minimizing power costs. All blowers
must be locally started. Starting and stopping a blower requires manual throt-
tling of the blower suction valve . Blowers are normally started or stopped
once or twice per day. Starter interlock circuits prevent starting a blower
unless the discharge check valve is closed and the relief valve is open. Once
on line, the blower suction valves are in a full open position and blower
throttling is effected by automatic throttling of the piston operated butterfly
valves in each air header through separate cascade flow control loops as
illustrated in Figure A-8.
A DO probe (AE) is located at the end of the third pass in each oxidation
tank and at the effluent end of the postoxidation tank. The output of trans-
mitter AIT is recorded on single point recorder AR and transmitted to a DO
controller (AIC). The AIC compares the input DO signal to the desired set
point and decrements or increments the set point of flow controller FIG as
required through a current to pneumatic converter. The FIG receives a header
flow signal from flow tube transmitter FIT and throttles the header butterfly
valve as required to maintain the flow set point received from the AIC . If
blower discharge falls below about 3.4 mvs (7200 scfm) on a single blower,
the surge control system will automatically bypass the flow to atmosphere.
Dissolved oxygen concentration is normally maintained at 0.4 ppm at the
DO probe (AE) location in each oxidation tank. The postaeration tank DO
concentration is typically maintained at 6 ppm or higher. A drawback of the
automatic DO control system design is that only one throttling valve is fur-
174
-------
RELIEF
VALVE
AIR
520kW(70OHP)
| CENTRIFUGAL
BLOWER ( Typ)
I
OTHER TWO
520KW (700HP)
BLOWERS
AIR TO POST OXIDATION TANK
*" )AIR TO
OTHER TWO
OXIDATION
TANKS
TANK NO. I AIR HEADER
PRIMARY
EFFLUENT'
DIFFUSED AIR HEADER (Typ)
\
RAS RAS
PE, * PASS PE
PE PE
PASS
PE
PASS
^—OXIDATION TANK
(Typ of 3)
ML
Figure A-8. Automatic dissolved oxygen control system - Reno/ Sparks
Joint Water Pollution Control Plant, Nevada.
-------
nished for each system. A throttling valve and DO control loop for each
header would permit close control of DO concentration in different tank
passes and probably reduce air requirements. Another problem associated
with the DO control system is that the System 2 influent channel is lined up
with the primary effluent channel directly across the distribution channel,
while Systems 1 and 3 receive primary effluent from either end of the distri-
bution channel. As a result, the solids loading on System 2 is invariably
higher than that on Systems 1 and 3 .
Performance
For the purposes of this report, the City of Sparks Division Engineer
agreed to run comparative tests of manual and automatic DO control on two
of the treatment plant systems. Systems 2 and 3 were selected for the tests
since an ongoing study was being performed in System 1 . It was initially
planned to run Systems 2 and 3 in parallel, with one system under manual
and the other under automatic. However, due to significant differences in
bacterial concentrations in each system, it was decided to run each system
in alternate DO control modes so that two sets of comparative data would be
available. Accordingly, two 24-hour tests were performed on each plant sys-
tem August 27-28, 1975.
Under manual DO control, the butterfly valve in a system's air feed
header (see Figure A-8) was manually modulated approximately every four
hours to maintain a desired DO in the effluent of the third tank pass. Under
automatic DO control, each system was operated as described above in the
section on operation. Dissolved oxygen was maintained at approximately
0.4 ppm for both modes of operation in both systems. Table A-7 illustrates
the results of the tests.
Table A-7 shows that System 3 demonstrated considerable improvement
under automatic DO control for all performance parameters measured except
for suspended solids removal. However, due essentially to a poor BOD
removal result under automatic DO control, System 2 demonstrated marked
deterioration in air supplied per unit quantity of BOD removed and BOD re-
moved per blower kWh. This result was surprising, considering that other
plants tested invariably showed improvements in the measured performance
parameters under automatic DO control.
Since the reported BOD removal efficiency for System 2 under automatic
DO control is so poor compared to manual DO control, it can be assumed the
test was in error. Standard Methods for the Examination of Water and Waste-
water reports a coefficient of variation of 17 percent in BOD data for glucose-
glutamic acid mixtures. If it is assumed the BOD removal efficiency of
System 2 under automatic DO control was the same as under manual DO
control, applying the same plant flow rate results in the same amount of BOD
176
-------
TABLE A-7.
PERFORMANCE COMPARISON OF MANUAL AND AUTOMATIC
DISSOLVED OXYGEN CONTROL AT THE RENO-SPARKS JOINT
WATER POLLUTION CONTROL PLANT3
Parameter
BOD removal efficiency,
percent
Suspended solids removal
efficiency, percent
Sludge volume index
Air supplied per unit
quantity of influent,
m3/m3
(cf/gal)
Air supplied per unit quantity
of BOD removed1?
m3/kg
(cf/lb)
BOD removed per blower kWh
kg/kWh
(Ib/kWh)
System 2
Manual
92
83
113
10.5
(1.41)
110
(1700)
0.55
(1.20)
Autod
78
86
108
9.5
(1.3)
140
(2300)
0.29
(0.63)
Percent
improvement
-14
3
4
10
-27
-47
System 3b
g
Manual
74
84
115
7.1
(0.96)
120
(1900)
0.35
(0.79)
Auto
85
82
100
5.8
(0.78)
72
(1200)
0.57
(1.2)
Percent
improvement
11
-2
13
18
40
63
Data from 4ea 24 hour tests performed on two independent activated sludge systems in same plant on
August 27-28, 1975.
Average daily flow -0.28 m3/s (6 .3 mgd) .
CAverage BOD applied to oxidation tanks - 4.86 mg/m /s (26.2 lb/1000 cf/day).
Average BOD applied to oxidation tanks - 3.97 mg/m3/s (21.4 lb/1000 cf/day).
P o
Average BOD applied to oxidation tanks 3.65 mg/m /s (19.6 lb/1000 cf day).
Average BOD applied to oxidation tanks - 3.97 mg/m /s (21.4 lb/1000 cf day).
Computed from total air supplied over testing period and 24-hour composite BOD of primary and
secondary effluent.
hBased on 16.3 dm3/s/kW (25.7 scfm/hp).
177
-------
removed in each case. Considering air supplied under automatic DO control
was about 9 percent less than manual, results in a 9 percent improvement in
air supplied per unit quantity of BOD removed and a 10 percent improvement
in BOD removed per blower kWh.
Maintenance
The dissolved oxygen control system components are maintained through
a plant instrumentation contract that currently costs $ 19 , 270/year. The plant
superintendent estimates about one-third of the contract time and expense is
required for the DO control system, or approximately $6400 per year.
The dissolved oxygen probes, transmitters and analyzers are of the same
manufacturer. Approximately 10-12 man hours per month outside of the instru-
mentation contract are spent cleaning and recalibrating the probes by plant
personnel. Included in this maintenance is replacement of probe membranes
about twice per month.
Blower maintenance, including cleaning, oil changes, lubrication, and
electrical repair is estimated by the plant superintendent to require about 50
man-hours per year for all three blowers. Shortly after installation, a pro-
blem in the surge control system developed with one blower due to a faulty
instrument that caused damage to the rotor, causing considerable maintenance
time for repair. However, all blowers are functioning normally now, and the
plant superintendent reports minimal problems with blower instrumentation
systems .
Safety and Emergency Procedures
Interlock circuits prevent starting a blower unless its check valve is
closed and its relief valve is open. Both valves may be operated manually as
required. After a blower starts, it will shut down if the check valve fails to
open within a set period of time (approximately 45 seconds) or if the bearing
lubrication oil overheats or loses pressure.
Electrical interlocks also require that one blower be placed in a standby
condition before either of the other blowers can be started. A blower must be
stopped by normal shutdown procedure before it can be placed in a standby
status.
An auxiliary oil lubrication pump is furnished for each blower to provide
sufficient oil pressure during start-up and shutdown and to back up the main
oil pump. Failure of the main, shaft driven oil lubrication pump during
blower operation will cause automatic start-up of the auxiliary pump.
178
-------
Each aeration flow controller (FIG) can be operated in manual, automatic
and ESP (automatic external set point modes). The DO controllers (AIC) can
be operated in manual or automatic control modes. If the automatic mode of
DO controller AIC is malfunctioning or out of service, the manual mode can
be used together with the ESP mode of aeration air flow controller FIG to
maintain a relatively constant ratio between the aeration air flow rate and the
sewage flow rate. If the ESP mode is out of service, the automatic mode of
aeration air flow controller AIC can be used to maintain a desired aeration air
flow rate. If the automatic mode is malfunctioning or out of service, the
manual mode of aeration air flow controller AIC can be used to establish a
desired flow rate. Under normal conditions DO/air flow cascade control is
used as described in the previous section on operation.
The plant operating manual advises setting air flow indicating controller
FIG to manual mode before removing any portion of the DO control system for
maintenance or repairs. Thus, an approximate desired air flow rate can be
maintained when an item such as the air flow transmitter (FIT) is taken out of
service.
179
-------
CASE HISTORY 7
SIMI VALLEY WATER QUALITY CONTROL PLANT, CALIFORNIA
Description of Aeration and Dissolved Oxygen Control System
The Simi Valley plant has an ADWF capacity of 0.31 m3/s (7 mgd) and
has been in operation since 1974. One oxidation tank with three passes is
provided for carbonaceous oxidation and nitrification. The design permits
plug flow, step feed and reaeration operating modes. Air is supplied by five
centrifugal blowers; three modulating blowers, each with a maximum output
of 1.79 m3/s (3800 scfm) , and two 0.94 m3/s (2000 scfm) nonmodulating
blowers.
All blowers were furnished by the same manufacturer. Two of the 112 kW
(150 hp) units were initially installed in 1964 and rebuilt about seven years
later, while the third was installed in 1974 under a plant expansion contract.
The 56 kW (75 hp) blowers were installed in 1964. All plant blowers are driv-
en by constant speed induction motors.
An automatic DO control system is provided which incorporates a pressure
control loop to modulate the suction valves on the 1.8 m3/s (3800 scfm)
blowers to maintain a constant pressure of 41 kPa (6.0 psig) in the blower
discharge manifold. Two DO probes and analyzers are provided. An instru-
mentation and control diagram of the DO control system is shown in Figure
A-9. Components include the following:
0 Centrifugal blowers with individual suction throttling valves (3 ea.)
° Nonmodulating centrifugal blowers (2 ea.)
0 Blower discharge manifold pressure control loop with pressure trans-
mitter and a pressure indicating controller (1 ea.)
0 DO probes, analyzers and controllers (2 ea.)
0 Flowmeter (FIT) for measurement of total aeration air flow (1 ea.)
0 Air header flow control valves (2 ea.)
180
-------
DISCHARGE
CHECK VALVE
AIR
SUCTION
VALVE
H2k,W.(l50HP)
AERATION BLOWER
(Typ of 3)
OTHER 2
112 kW
(150 HP)
BLOWERS
56kW(75HP)
BLOWERS
r-
l/P
PASS r
P4SS 2
PASS 3
^-O
XIDATION TANK
1 /AlcV!
Figure A-9. Automatic dissolved oxygen control system - Simi Valley,
California.
-------
Operation
The complete blower system includes five blowers. Three are 110 kW
(150 hp) and are automatically inlet choke controlled to maintain a 40 kPa
(6 psig) outlet pressure, regardless of the volume requirements. The remain-
ing two blowers are 56 kW (75 hp) and are manually operated to increase the
total air capacity under peak loading conditions. The starting sequence of
the 110 kW (150 hp) blowers is switch selectable on the control board in the
plant control center.
Each of the 110 kW (150 hp) blowers (numbers 1,2,3) has a butterfly
suction valve. All three of these valves are pneumatically actuated from a
single pressure indicating controller (PIC) on the control board, based upon
a desired outlet pressure in the common air header. The butterfly valves all
have limit switches at both the fully open and fully closed valve positions,
which are used to actuate the blower sequence control system. When the
first inlet valve is fully open, the limit switch activates the motor starter
relay on the second blower to bring it on line . When enough capacity is
needed that both inlet valves are fully open, the third 110 kW (150 hp)
blower is started automatically.
It is imperative that all blowers be operated within nominal operating
ranges to eliminate the possibility of blower "surge". This condition occurs
when a blower is operated at less than approximately 30 percent of full load.
Due to this requirement, the starting of the manual blowers can be quite
critical. If they are not started at the correct time, as indicated on Figure
A-10, severe underloading of the other online blowers can occur. For example,
the blower sequence diagrams indicate that blower number 4 (75 hp-manual) ,
can either be started at 1. 6 m3/s (3500 cfm) loading or at 3.3 m3/s (7000 cfm)
loading, but should not be started at any other loading. Blower number 5
should be started at 4.0 m3/s (8500 cfm), or higher. If this sequence is not
followed, surging of online blowers is likely.
When a blower is not in service, the blower bypass valve is always open.
This allows the blower to start up by recycling air back to its inlet. When the
blower motor starts, it triggers a time delay relay which shuts the bypass
valve after 15 seconds of blower operation. This forces open the check valve
into the main air header. When power to a blower is stopped, the bypass
valve immediately opens, allowing the blower to bypass air while slowing to
a stop.
Total air flow is monitored in the combined air header with flow tube
transmitter (FIT) . The output from the FIT is linearized by square root extrac-
tor FY mounted behind the control board. Total air flow is then indicated with
a 500 cm3/s-7.1 m3/s (1-15,000 cfm) range and totalized, with 1 count =
0.5 rnVs (1000 cfm) , on the control board. Recorder outputs are available.
182
-------
12
tr 10
o
o
o
O
o 6
0 -
o
<
o
o
14
12
£ 10
o
o
o
O o
0 -i
I
t
7
Jf2\
4 6 8 10 12
TOTAL LOAD 1000 CFM
4 6 8 10 12 14
TOTAL LOAH 1000 CFM
14
NOTES: 1. Diagrams taken from plant operating manual
2. (cfm) (4.72 x ± '
Figure A-10. Blower sequence diagrams -Simi Valley Water Quality
Control Plant.
183
-------
Aeration air control is divided into two separate and identical systems,
with each controlling the dissolved oxygen levels in a specific region of the
3-pass tank. One system controls pass 1, and the other controls passes 2
and 3.
The sensors used for dissolved oxygen measurement have a 0-5 ppm DO
range. The two DO control loops can be separately controlled, with both set
points individually adjustable from the control board. The reaeration DO
control loop (pass 1) set point is 1.5 ppm DO, and the mixed liquor DO
control loop (passes 2 and 3) set point is 3.0 ppm DO. Recorder outputs are
included for both control loops.
Performance
On July 8-9, 1975, a 48-hour test was run at the Simi Valley plant to
evaluate the performance of the automatic DO control system. These days
were selected as average days for plant flow and loading. The test was
divided into two periods of 24 hours each.
During the first 24 hours, the operators to the two automatic feed control
valves in the two air headers to the oxidation tanks were disconnected. The
DO control system was operated in a manual mode by reading the DO concen-
tration on the remote indicators on an hourly basis and manually modulating
the air feed header butterfly valves as required. Approximately 12 adjust-
ments were made to the valves over the 24-hour test period.
During the second period of 24 hours, the butterfly valve operators were
reconnected, and the DO control system was operated in an automatic mode
as illustrated in Figure A-9. All instruments functioned normally during the
test and the controllers were considered tuned.
During the tests, blowers were throttled, started or stopped as,required
to maintain the required manifold discharge pressure. Only two 110 kW
(150 hp) blowers were used since the third unit was out of service. The
56 kW (75 hp) blowers were used to a greater extent in the manual DO control
test than in the automatic test.
Dissolved oxygen was maintained close to 0.5 ppm at the end of pass 1
and close to 3.0 ppm at the end of pass 3 for both tests. The DO probes
required cleaning once during each test. Results of the performance tests
are presented in Table A-8. Although air supplied per unit quality of influent
showed a relatively slight improvement under automatic DO control mode ,
other more significant parameters relative to the quantity of BOD removed
demonstrated a greater improvement.
184
-------
TABLE A-8. PERFORMANCE COMPARISON OF MANUAL AND AUTOMATIC
DISSOLVED OXYGEN CONTROL AT THE SIMI VALLEY WATER
QUALITY CONTROL PLANT
Parameter
BOD removal
efficiency, percent
COD removal
efficiency, percent
Suspended solids
removal efficiency, percent
Sludge volume index
Air supplied per unit quality
of influent,
m-Vm^
(cf/gal)
Air supplied per unit quantity
of BOD removed, d
m^/g
(cf/lb)
Q
BOD removed per blower kWh,
kg/kWh
(Ib/kWh)
Manual
82
91
99
130
19.8
(2.64)
240
(3900)
0.24
(0.53)
Q
Automatic
*
81
92
97
127
19.3
(2.57)
220
(3500)
0.27
(0.59)
Percent
improvement
-1
1
-2
2
3
8
13
Data from 24-hour tests on June 8-10, 1975.
b ?
Average daily flow - 0 .204 m°/s (4.65 mgd) .
Average BOD applied to oxidation tanks - 3.11 mg/m /s (16.8 lb/1000
cf/day.
°Average daily flow - 0.208 m3/s (4.74 mgd) 3
Average BOD applied to oxidation tanks - 3.45 mg/m /s (18.6 lb/1000
cf/day).
Computed from total air supplied over testing period and 24-hour composite
BOD of primary effluent and secondary effluent.
Q
Based on blower operating times.
185
-------
Maintenance
It is estimated by the plant superintendent that blower maintenance
requires an annual average of two man-hours per day, at a labor rate of $10/
hr. , including fringe benefits. Included in this labor estimate are blower
overhauls that are normally performed once every five years on each machine.
The dissolved oxygen probes are normally cleaned and recharged, if
required, once or twice per week for a total labor time for both probes of
about one man-hour per week. Approximately two man-hours per month is
spent replacing DO probe membranes. Recalibration is required every two
weeks for a labor time of two man-hours for two probes.
Recently, plant personnel have reported considerable difficulty with DO
probe drift. Recalibration has become necessary every week. Approximately
four man-hours per week are required for recalibration of both probes. The
project engineer for the Simi Valley plant reports one instance where the DO
control system had the blower suction throttling valves wide open to maintain
the 3 ppm set point in the third pass. A test with a portable DO analyzer
revealed the actual tank DO was 6 ppm, yet the probes had been recently
recalibrated.
Virtually all instrumentation maintenance is performed under an outside
instrumentation contract that costs $2200 per year and involves one routine
service call per month. Additional labor required for special calls typically
costs $700 per year for a total of $2900 per year. This contract does not
include parts. No estimate was available for parts cost since the plant was
only recently placed in operation.
Safety and Emergency Procedures
Each of the 110 kW (150 hp) blowers has a motor overtemperature protec-
tion system. If a blower motor overheats, it will automatically shut down the
blower, and the next blower in sequence is automatically started. In addition,
an interlocking circuit is included to provide starting of a backup blower if an
on-line blower fails. A vibration switch is also provided on each blower that
will shut down that blower in the event of excessively high vibration.
The 110 kW (150 hp) blowers are designed for automatic remote start and
remote shutdown. Currently, the units may be remotely started or stopped,
but the automatic remote start circuits have not been implemented. The 56 kW
(75 hp) blowers can also be remotely started or stopped, but will not be auto-
matically controlled.
Two separate feeders from different substations supply the plant with
power. Loss of power through one feeder will result in automatic switchover
to the alternate power source.
186
-------
Critical components of the dissolved oxygen control system may be re-
placed from the plant spare parts inventory. During repair or replacement of
certain components, the plant may be operated under manual DO control.
187
-------
CASE HISTORY 8
SAN FRANCISCO INTERNATIONAL AIRPORT
WATER QUALITY CONTROL PLANT
Description of Aeration and Dissolved Oxygen Control System
The San Francisco International Airport plant started operation in 1971.
It is designed for an average daily flow of 0.10 m3/s (2.2 mgd) and currently
treats an average dry weather flow of about 44 dm^/s (1.0 mgd). The plant is
a secondary treatment facility installed to treat the domestic sewage discharge
from the airport. Two oxidation tanks are provided, with design permitting
step feed or plug flow modes of operation.
Each oxidation tank is 18 m (60 feet) square with approximately a 3.7 m
(12 foot) water depth. A fixed mechanical aerator is installed in the center
of each basin. Each mixer is driven by a 30 kW (40 hp) , 1730 rpm, 460 v
wound rotor motor. Reduction gearing is provided, so that the mixer will
rotate at a maximum speed of 47 rpm and a minimum speed of 25 rpm. The
speed of each mixer can be varied by a remotely located saturated core re-
actor variable speed drive unit. A separate drive control unit is furnished
for each mixer.
A dissolved oxygen probe is located near the influent end of the second
tank, so that mixed liquor passing from the first tank to the second flows past
the probe. An alternate receptacle exists for installation of the DO probe in
the first tank. A DO analyzer and DO controller are located in the plant
operations building proximate to the mixer variable speed drive control units.
An instrumentation and control diagram of the DO control system is shown in
Figure A-11 . Components include the following:
0 Fixed position, variable speed drive mechanical mixers (2 ea.)
0 Saturated reactor core variable speed drive units for mixers (2 ea.)
0 Dissolved oxygen probe (1 ea.)
0 Dissolved oxygen analyzer/transmitter (1 ea.)
188
-------
SATURATED
REACTOR CORE
VARIABLE SPEED
DRIVE (Typ)
30 kW
(40 HP)
MECHANICAL
MIXER (Typ)
SPEED ,
REDUCTION I—
GEAR (Typ)
Figure A-11 .
Automatic dissolved oxygen control system - San Francisco
International Airport, California.
189
-------
0 Dissolved oxygen controller (1 ea.)
0 Dissolved oxygen recorder (1 ea.)
Operation
Plant operating experience indicates the two-pass mode produces a
better quality effluent than operating the oxidation tanks in parallel or in
series. Under parallel operation, primary effluent is not equally distributed
to each tank. Under series operation, with return sludge applied to the first
tank, nitrification occurs, resulting in excessive chlorine demand. Accord-
ingly, the oxidation tanks have been operated in a two-pass mode since 1972,
Under the two-pass system, primary effluent is aerated in the first tank,
then applied to the second tank through a sluice gate in the common wall.
The mixed liquor is then aerated with the second mixer and combined with
return sludge from the final sedimentation tank.
The automatic DO control system is designed to modulate the speed of
both mixers to maintain the desired DO concentration in the oxidation tanks.
DO is sensed in tank 2 , and the signal is transmitted to the DO analyzer
(AIT) located in the rear of a remote control panel. The analyzer transmits
the linearized DO level signal to controller AIC and recorder AR. The con-
troller set point is typically set at 1.5-2.0 ppm. An error signal correspond-
ing to the deviation above or below set point is transmitted by controller AIC
to each mixer variable speed drive control unit. The drive control unit will
cause the corresponding mixer to increase or decrease speed according to
the error signal received from controller AIC.
Normally, the drive control unit for the mixer in tank 1 is placed in man-
ual mode , and the mixer speed is set at 40 rpm by the manual control station
(HIK). The manual control station for the second oxidation tank is normally
set in automatic mode, allowing the DO controller (AIC) to vary the mixer
speed from 25-47 rpm as required. Between 10 pm and 7 am, when the plant
flow drops to about 30 dm3/s (0.6 mgd), both mixers are operated in manual
mode, and set at about 30 rpm. Mixers are operated at low speed at night to
prevent poor mixing and bearing wear resulting from alternate revving and
braking of the blades under the low water level condition. The above des-
cribed "normal operation" occurred during the automatic mode reported in
Table A-9.
In mid-1975, the plant was operating without automatic DO control due
to failure of the same circuit relay in both saturated reactor core drive units.
Under manual control, both mixers are maintained at 30 rpm from about llpm
to 8 am. In the morning, the mixer in the first tank is set at 40-43 rpm and
that in the second tank set at 47 rpm.
190
-------
TABLE A-9 PERFORMANCE COMPARISON OF MANUAL AND AUTOMATIC
DISSOLVED OXYGEN CONTROL AT THE SAN FRANCISCO
INTERNATIONAL AIRPORT WATER QUALITY CONTROL PLANT'
Parameter
BOD removal
efficiency, percent
COD removal
efficiency, percent
Suspended solids
removal efficiency, percent
Sludge volume index
:b
Manual
92
70
79
92
c
Automatic
94
84
96
201d
Percent
improvement
2
14
17
-118
3Data from plant records for April 1975 (automatic) and June 1975 (manual).
bAverage daily flow - 42.9 dm3/s (0.980 mgd).
Average BOD loading- 3.89 mg/m3/s (21.0 lb/1000 cf/day) .
CAverage daily flow - 38.0 dm3/s (0.867 mgd).
Average BOD loading - 7.49 mg/m3/s (40.4 lb/1000 cf day).
Jet fuel slug from airport received at plant.
191
-------
Performance
Data collected from the San Francisco International Airport plant under
manual and automatic modes of operation are presented in Table A-9. As pre-
viously explained, the automatic control mode was not operating between
about 10 pm and 8 am every day. Operating experience shows that when the
flowrate is 30 rfm3/s (0.6 mgd) or less , the oxidation tank water level drops too
low for effective mixing by the aerators at high rpm. Thus, the mixer speed is
lowered to 30 rpm in both tanks until plant flow picks up again about 8 am.
Table A-9 was compiled from data recorded over the months of April, 1975
and June, 1975 for respective operating' modes of automatic and manual DO
control. Improvement in plant performance is demonstrated for automatic
compared to manual DO control, although the plant was in automatic mode
only during the day. The high sludge volume index of 201 under automatic
mode is attributed to a slug of jet fuel received at the plant. Plant records
show the average COD load applied to the oxidation tanks for April, 1975 was
664 ppm, while that for June, 1975 was 323 ppm. COD and BOD removal under
these conditions was surpirsingly good, even with such a high SVI.
Maintenance
Maintenance requirements of the DO control system and mechanical mix-
ers have been relatively minor, according to the plant senior stationary engi-
neer. However, certain parts of the saturated reactor core variable speed
drive units appear difficult to obtain since in 1975 the units were out of ser-
vice almost two months awaiting relays .
Preventive maintenance on the mixers involves changing or cleaning the
motor brushes and changing the drive oil. Motor maintenance requires 4-6
man-hours per year per unit at a labor rate of $10.65/hour, including fringe
benefits. The mixer gear drive unit oil is changed every six months. Approx-
imately four man-hours per year are required for oil changes at a labor rate of
$8.76/hour including fringe benefits. Each mixer requires 60 dm3 (16 gallons)
of new oil per year.
Other maintenance reported on the mixers has involved changing a motor
bearing by the contractor soon after installation. This operation resulted in
mixer shutdown for almost three days.
The saturated core reactor variable speed drives for the aerators require
occasional adjustment. Plant records show 20-40 man-hours per year are
expended adjusting and maintaining both of these units.
The dissolved oxygen probe is calibrated daily and cleaned twice per
week by the plant chemist. Calibration requires less than one man-hour,
192
-------
while cleaning usually takes five minutes. Approximately every three months,
one man-hour is spent changing the probe membrane and recharging the probe.
All probe maintenance is performed by the plant chemist.
Safety and Emergency Procedures
In March, 1974, the plant was equipped with a 400 kW standby diesel
generator, sized to provide full operating power for the facility. Loss of
plant power causes a switchover to the auxiliary power source in less than
three seconds.
The aerators are remotely started from the plant operations building.
Each unit is furnished with a local lockout switch to prevent remote start
when the mixer is being serviced. The drive motors are thermally protected
against overload.
Presently, the plant stocks a limited number of spare parts for the auto-
matic DO control system. Critical parts of the DO controller are on hand and
may be readily installed. A spare DO probe was not available at the time of
this writing in mid-1975.
Each saturated core reactor variable speed drive is electrically committed
to a single aerator. Thus, loss of the variable speed drive control unit to a
particular aerator would put the corresponding tank out of service. Plant oper-
ating personnel make hourly checks of DO on the remote control panel and
adjust the mixer speed in tank 2 accordingly.
193
-------
CASE HISTORY 9
ST. REGIS WASTEWATER TREATMENT PLANT,
SARTELL, MINNESOTA
Description of Aeration and Dissolved Oxygen Control System
The St. Regis Wastewater Treatment Plant treats pulp and paper waste
produced during the manufacture of paper, using the conventional activated
sludge process. The plant, which went on-line in April, 1973 , is designed
to treat an average flow of 0.30 m3/s (7 mgd) but currently receives about
0.20 m3/s (5 mgd). Since paper is manufactured continuously at the paper
plant, wastewater is produced 24 hours per day. Influent flow typically
varies from 0.20-0.30 m3/s (4.5-6 mgd). Two parallel oxidation tanks are
provided, each with two 45 kW (60 hp) two-speed mechanical mixers driven
by 460 volt induction motors. The tanks are each 30 m (100 ft) long, 18m
(60 ft) wide with a 3.7 m (12 ft) water depth. A dissolved oxygen probe with
an agitator assembly is mounted near the effluent end of each tank, equidis-
tant from the mixers. An instrumentation and control diagram of the automatic
DO control system is shown in Figure A-12. Components include the follow-
ing:
0 Mechanical mixers (4 ea.)
0 Dissolved oxygen probes with associated analyzer and transmitter
(2 ea.)
0 Dissolved oxygen controllers (2 ea.)
0 Dissolved oxygen recorders (2 ea.)
Operation
The St. Regis plant is normally operated in an automatic DO control mode.
Dissolved oxygen is sensed by a DO probe (AE) in each oxidation tank, ana-
lyzed by a panel mounted analyzer (AIT) and transmitted to DO controller (AIC)
and recorder (AR) . The DO controller (AIC) effects a speed change in both
mixers through each motor starter in accordance with the dissolved oxygen
control level.
194
-------
r
1-0
ML
ML
X
M
DO
I
-------
The DO controller is designed to output a control signal if the DO level
deviates beyond a permissible control band. Since the mixers are 2-speed
units, the control output causes the mixer to change from one speed to a-
another, depending on the oxidation tank DO level.
Performance
From August 10 through August 22, 1975, personnel of the St. Regis
Paper Company conducted a performance test for the benefit of this report.
Both oxidation tanks were operated in an automatic control mode August 10-17
and August 22 and in a manual control mode August 18-21. Under manual
operation, the mixer speeds were varied as required to maintain 1-2 ppm DO
in each oxidation tank. The average frequency of mixer speed change adjust-
ment was once every four hours.
Under the automatic control mode, mixer speed was automatically varied
to maintain a DO concentration of 1-2 ppm. Test records indicated that the
mixer speed was automatically changed about every half hour. One mixer was
lost during the test due to bearing failure. Accordingly, data obtained on
August 15 and 16 is considered to poorly represent the manual control mode
under which the mixers were operating. Certain erratic readings of BOD and
suspended solids occurred on other dates that St. Regis Paper Company
personnel attributed to boilouts, washouts and sampler line plugging. Thus,
selection of suitable time periods for comparison of the manual and automatic
DO control modes was difficult due to the testing problems that occurred.
August 10-13 and August 18-21 were chosen as the automatic and manual
DO control periods, respectively.
The results of the DO control study test are presented in Table A-10 and
indicate that automatic DO control provided no significant performance im-
provement over manual DO control. This result is to be expected since
St. Regis Paper Company produces paper 24 hours per day, and the waste
stream is, thus, reasonably constant in flow and loading.
Maintenance
According to the St. Regis Paper Company technical assistant responsible
for wastewater treatment plant operations, the automatic DO control system
has presented no significant maintenance problems since startup in April, 1973.
The operating and maintenance staff is generally pleased with the DO control
system and offered no complaints.
The DO probes are checked once per week and cleaned and calibrated if
required. It is estimated that five man-hours per month is required for DO
probe maintenance. Maintenance staff members earn $4 . 50-4 . 75/hour. Cali-
bration of probes is performed in the company instrument shop. Calibration
time required is estimated as minimal.
196
-------
TABLE A-10.
PERFORMANCE COMPARISON OF MANUAL AND AUTOMATIC
DISSOLVED OXYGEN CONTROL AT THE ST. REGIS PAPER
COMPANY WASTEWATER TREATMENT PLANT3
Parameter
BOD removal efficiency,
percent
Suspended solids removal
efficiency, percent
Sludge volume index
Manual
97
88
252
Q
Automatic
98
89
201
Percent
improvement
1
1
20
Data obtained from two four-day test periods conducted August 10-22, 1975.
Average daily flow - 0.246 m3/s (5.62 mgd). Average BOD loading - 6 45
mg/mVs (34.8 lb/1000 cf/day).
Q
Average daily flow - 0.241 m3/s (5.50 mgd). Average BOD loading - 4 89
mg/mVs (26.4 lb/1000 cf/day).
H ^
Average BOD loading - 4.89 mg/m /s (26.4 lb/1000 cf/day).
The mixers are normally lubricated every three weeks. Lubrication time
is estimated at one man-hour per month. The mixer gear drive oil is changed
twice per year at a labor requirement of one man-hour per unit or four man-
hours total. An additional five man-hours per month is expended in checking
the mixers as a preventive maintenance measure. The only significant break-
downs to date have been attributed to failure of motor bearings, motor insula-
tion and 460 volt feeder line insulation.
Safety and Emergency Procedures
Power to the waste treatment plant and to the paper processing plant is
normally provided by a local utility company. However, on-site steam and
hydroelectric power generation equipment permits operation of critical paper
company equipment as well as the waste treatment plant during a utility
company power failure.
197
-------
The mixer motors are all equipped with combination starters that contain
heaters for overcurrent and fuses for short circuits. Overcurrent or a short
circuit will thus cause automatic shutdown of a mixer. All mixers may be
started, stopped or have speed changed from a remote control panel. Inter-
locks are furnished at each motor to prevent remote start when a unit is being
serviced.
Spare parts for the mixers and other DO control components are minimal.
However, the plant can be operated briefly with one mixer out of service.
This situation occurred during the DO control study, resulting in all other
mixers being operated at high speed until the fourth mixer was repaired.
198
-------
CASE HISTORY 10
LONG BEACH WATER RENOVATION PLANT
LONG BEACH, CALIFORNIA
Description of Aeration and Dissolved Oxygen Control System
The Long Beach Water Renovation Plant, located in the southeast portion
of Los Angeles County, California, was placed in operation in January, 1973,
with a designed capacity of 0.548 m3/s (12.5 mgd). It is a secondary acti-
vated sludge plant with complete nitrification. Flows presently being
treated are approximately 0.31 m3/s (7.0 mgd). One oxidation tank with four
passes is provided.
Primary effluent is distributed to the oxidation tank through step feed
channels and step feed flow meters. Under current operation, primary efflu-
ent is divided equally between the first and second passes through two step
feed gates in each pass.
Air is provided by two 10.6 m3/s (22,500 scfm), single stage, pedestal
bearing mounted, centrifugal blowers. Both blowers are supplied with 4160
volt power and driven by 600 kW (800 hp) induction motors. An automatically
controlled common bypass system is provided to vent excess air from the
common blower discharge line back to suction. For plant odor control, the
primary sedimentation tanks, primary effluent channel, and aerator step feed
channels are covered. Supply air for the air compressors is withdrawn from
under these covers. Air is supplied to the aeration system through coarse
bubble diffusers. Air is discharged from the blowers into a common manifold
and distributed through two headers to the oxidation tank and through one
header to the channel aeration system. Each header is furnished with a
flow tube and a control throttling valve. The channel aeration and pass 1-2
header control valve is manually controlled, and the pass 3-4 header control
valve is automatically controlled by the DO control system.
Four DO probes are provided with DO probe receptacles located in the
middle and effluent end of each pass. Air flow to passes 3 and 4 is throttled
by a DO control loop to maintain a desired DO concentration at the selected
control probe. Air flow to passes 1 and 2 is indirectly throttled by simulta-
199
-------
neously modulating the blower inlet guide vanes and the blower bypass valve
through a split range cascade DO control loop to maintain a desired DO con-
centration at the selected control probe. An instrumentation and control dia-
gram of the DO control system is shown in Figure A-13.
Operation
The process air blowers may be locally or remotely started. However,
the plant operators report remote starting is not used because they prefer to
observe and listen to each machine during start-up.
Based on the DO output on recorder (AR) and operating experience, the
operator selects a DO control probe in pass 3 or pass 4 through selector
switch (HS) . The output of the probe is transmitted to a DO controller (AIC)
that modulates a cylinder operated butterfly valve in the pass 3-4 air feed
header to maintain a set point DO concentration in the tank at the DO control
probe.
The operator also selects a DO control probe in pass 1 or pass 2 in the
same manner as above via a selector switch (HS). The control probe output
is transmitted to another dissolved oxygen indicating controller (AIC) that
outputs a flow set point to a flow indicating controller (FIC) to maintain the
desired tank DO at the control probe. The flow indicating controller (FIC)
outputs a 4-20 ma control signal to two relays (FY). One relay accepts any
signal between 12-20 ma and modulates the inlet guide vanes of the blowers
to deliver the set point flow of the flow indicating controller (FIC). Flow
from each blower is metered by flow tubes, transmitted to a summing relay (FY)
and input to the flow indicating controller (FIC). The second relay accepts all
control signals from the flow controller (FIC) between 4-12 ma and modulates
a blower bypass valve to return sufficient discharge air to suction to maintain
the blower flow rate above surge condition. Under present operation, the
desired DO level is set to be maintained at 1.5 ppm at the middle probe loca-
tion in the second and fourth passes which are used as the control points.
A problem associated with the above control system observed by the Los
Angeles Sanitation District (LACSD) staff with this loop within a loop control
scheme is that the DO control loop for passes 3 and 4 may be calling for more
air, while the DO control loop for passes 1 and 2 may be calling for less air.
Consequently, the throttling valve on the pass 3-4 air feed header may be
wide open, and insufficient air may be available for throttling due to DO
requirements of the other control loop.
The LACSD only recently implemented the above control scheme as a
modification and improvement of the previous DO control system. The District
considers it an experiment and is actively seeking alternate DO control con-
figurations to find the best DO control solution for the Long Beach plant.
200
-------
CO
o
CENTRIFUGAL
BLOWER (Typ)
I \ ' — ' "• " ^ |
I I
~I
•~I
AIT) AIT)
OXIDATION
1fc
^
PE
3fc
*
PE
£5)°°
CP4SS 4
PASS 3
(Ai)2-0-
PE PE©^0"!
CP4SS 2
PASS 7 | ^
^.
•^*/
PE
ML
^A^l
TO
.CHANNEL (AlT) (AIT)
AERATION
HEADER
I
I I
I
Figure A-13. Automatic dissolved oxygen control system -
Long Beach Water Renovation Plant, California.
-------
Performance
For the purposes of this report and an evaluation of the Long Beach Water
Renovation Plant DO control system, the LACSD agreed to run a manual vs
automatic DO control performance test on August 20-22, 1975. Under manual
DO control, air to the third and fourth passes of the oxidation tank was
adjusted by manually throttling the header valve. Air to the first and second
passes was controlled by manual adjustment of the blower inlet guide vanes.
All adjustments were made at approximately one-hour intervals. Under
automatic control, the DO control system was operated as described above in
the section on operation. A DO concentration of approximately 1.5 ppm was
maintained at DO control probes located in the middle of the second and fourth
passes for both manual and automatic DO control tests.
As indicated in Table A-ll, there was a noticeable improvement under
automatic DO control for most of the performance parameters measured. It
is expected that improvements would have been greater if manual adjustments
of the pass 3-4 butterfly control valve and the blower inlet guide vanes were
done less frequently. Adjustments at hourly intervals is considered a fair
approximation of automatic DO control.
Maintenance
Maintenance of the DO control system has required minimal labor and
material costs. Minor problems have been experienced with DO probe drift.
The probes are checked daily by comparing with a portable DO meter. Clean-
ing and recalibration, if necessary, occurs approximately once per week and
requires one man-hour per probe. Recharging and membrane replacement is
necessary about once every six to eight months and requires about 4 man-
hours per probe.
Maintenance costs of the blowers are also minimal. Typical annual main-
tenance involves replacing oil filters, conducting vibration analysis for deter-
mination of bearing conditions and cleaning of the impellers of volutes.
Replacement of oil is conducted only when analysis of the oil indicates dirt
or degradation. Since being placed in operation in January of 1972, it has
not been necessary to replace the oil or bearings, or make any other type of
compressor repair. Annual maintenance labor on the blowers is estimated at
approximately 40 man-hours per year per blower.
Safety and Emergency Procedures
The aeration system blowers are monitored and protected by an elaborate
instrumentation and control system. Each blower is prevented from operating
in a surge condidion by a signal limiting device in the control system which
prevents closure of the inlet guide vanes before surge point is reached. In
202
-------
TABLE A-11.
PERFORMANCE COMPARISON OF MANUAL AND AUTOMATIC
DISSOLVED OXYGEN CONTROL AT THE LONG BEACH WATER
RENOVATION PLANT
Parameter
BOD removal efficiency,
percent
COD removal efficiency,
percent
Suspended solids removal
efficiency, percent
Sludge volume index
Air supplied per unit quantity
of influent,
3, 3
m /m
(cf/gal)
Air supplied per unit quantity
of BOD removed, d
m /kg
(cf/lb)
Air supplied per unit quantity
of COD removed,6
m /kg
(cf/lb)
Manual
97
90
90
99
27
(3.6)
190
(3100)
92
(1500)
c
Automatic
97
90
90
94
25
(3.4)
180
(3000)
85
(1400)
Percent
improvement
none
none
none
5
7
5
8
Data from 24-hour tests on August 20-22, 1975.
Average daily flow - 0.302 mVs (6.90 mgd).
O
Average BOD applied to oxidation tanks - 3.86 mg/m /s (20.8 lb/1000
cf/day).
CAverage daily flow- 0.302 m3/s (6.90 mgd).
2
Average BOD applied to oxidation tanks - 3.32 mg/m /s (17.9 lb/1000
cf/day).
Computed from total air supplied over testing period and 24-hour composites
of primary effluent BOD minus secondary effluent BOD.
Q
Computed from total air supplied over testing period and 24-hour composites
of primary effluent total COD minus secondary effluent soluble COD.
203
-------
case of failure of this system, vibration switches cause blower shutdown if
a surge condition occurs. Low oil pressure, high bearing temperature, high
oil temperature, and high motor winding temperature are also monitored and
cause blower shutdown if normal operating ranges are exceeded.
Under a plant power failure, air flow to the oxidation tanks ceases , and
the operating blowers coast to a stop. An auxiliary turbine generator auto-
matically starts and supplies 480 volt power to all other plant processes as
well as supplying power to the air compressor lube pumps to supply lubrica-
tion while the blowers are coasting to a stop.
204
-------
CASE HISTORY 11
SAN JOSE-SANTA CLARA WATER POLLUTION CONTROL PLANT,
SAN JOSE, CALIFORNIA
Description of Aeration and Dissolved Oxygen Control System
The San Jose-Santa Clara Water Pollution Control Plant began operation
in 1964 with an average dry weather flow design of 4.1 m3/s (94 mgd). In
1973, the plant was expanded to a dry weather flow capacity of 7.01 m^/s
(160 mgd). Currently, the plant is treating an average flow of 3.9 m /s
(90 mgd). The plant is an activated sludge treatment facility that employs the
Kraus Nitrified Sludge Interchange Process. Two tank batteries are provided,
each composed of six two-pass oxidation tanks and two two-pass nitrification
tanks. However, piping for each battery is arranged to permit conversion of
two more oxidation tanks to nitrification tanks if desired. Wastewater may be
treated using plug flow, step feed or tapered aeration activated sludge oper-
ating modes. Normally, the oxidation and nitrification tanks are operated in
the plug flow mode.
The Kraus process utilized involves mixing a portion of the activated
sludge with supernatant and digested sludge from the anaerobic digesters and
aerating the combination for about 24 hours in the nitrification tanks. The
mixture is then pumped to the oxidation tanks for further aeration with the
primary effluent.
Two air systems are provided to deliver air to the secondary process at
different pressures. The high pressure, or diffused air system, delivers 55
kPa (8 psig) air to twelve oxidation and four nitrification tanks at a level
two feet above the tank bottom. Air is introduced from one side of each tank
pass through diffusers that produce minute air bubbles. The low pressure, or
distributed air system, delivers 28 kPa (4 psig) air to the twelve oxidation
tanks at a level five feet below the tank surface. This system introduces
air into each tank pass opposite the fine bubble diffusers and produces a
much larger bubble diameter.
Six engine driven, single stage, centrifugal blowers supply air for the
high and low pressure air systems. Four blowers are furnished for the high
pressure and two for the low pressure system. In addition, four additional
motor driven rotary, lobe type, positive displacement blowers were installed
205
-------
in 1970 to augment the high pressure air system. These blowers are driven
by 298 kW (400 hp) motors and are designed to each deliver 4.7 m^/s (10,000
cfm) at 55 kPa (8 psig). The diffused or high pressure system engine driven
blowers are each designed to deliver 28 m3/s (60,000 cfm) of air at a pressure
of 55 kPa (8 psig) and are driven by 1.8 MW (2400 hp) engines. The distri-
buted, or low pressure system engine driven blowers are each designed to
deliver 40 m3/s (85,000 cfm) at 28 kPa (4 psig) and are driven by 1.4 MW
(1850 hp) engines. The engines are tri-fuel units that can operate on (a) a
blend of digester gas and natural gas, (b) a blend of digester gas, natural gas
and diesel fuel, and (c) die sal fuel.
The low pressure air system blowers are throttled by varying engine speed
through a flow control loop that senses blower manifold discharge flow. The
high pressure air system blowers are throttled by varying engine speed through
a pressure control loop that senses blower manifold discharge pressure. The
set point for each 28 kPa (4 psig) and 55 kPa (8 psig) header was manually
derived from plant flow (in mid-1975) and experience but will soon optionally
originate from a DO or ORP probe located near the second pass end of each
tank. The set point for the main 28 kPa (4 psig) header air flow controller is
set by operating experience. Each tank header control valve is throttled by
a cascade flow control loop to maintain 28 kPa (4 psig) in the low pressure
manifold.
During the 1970-73 plant expansion, the plant diffused and distributed
air systems were placed under direct digital control using a dual computer
system. The previously installed pneumatic control systems remain intact
and functional, but the computer was interfaced directly with the primary and
final control elements. Control functions previously accomplished by pneu-
matic analog systems are now effected by either plant computer, utilizing
suitable control algorithms analogous to the pneumatic analog control func-
tions . Four nitrification tanks were added to the original twelve tanks during
the expansion. The new tanks have all electric instrumentation, thereby
eliminating the need for P/I and I/P converters. An instrumentation and con-
trol diagram of the DO control system is shown in Figure A-14. Components
include the following:
0 Single stage centrifugal engine driven blowers with flow regulated
surge control system, current transmitter and high and low speed
alarms (6 ea.)
0 Rotary, lobe type positive displacement, motor driven blowers (4 ea.)
0 Low pressure blower discharge flow control system with pitot tube ,
square root extractor, flow transmitter and flow controller (1 ea.)
206
-------
»-To other 13 F I C 's
To other II OXIDATION TANKS
1 »•
1 55 kPa __
. 28 kPo
^—— — to
PASS 1 — N
A
(AE) PASS 2 •*"^
^-p
I/p
-^-»
OXIDATION TANK
(Typ. of 12)
SOFTWARE
ALGORITHM (Typ.)
OTHER 3- ISOOkW
ENGINE DRIVEN
CENTRIFUGAL
BLOWERS
To other II OXIDATION TANKS
SOOkW
POS. DISPL.
BLOWER
OTHER 3 MOTOR-
DRIVEN, POSITI VE
D ISPLACEMENT
BLOWERS
55KPO
PASS 1
PASS 2
To other 3
N ITRIFICATION
TANKS
NITRIFICATION TANK
(Typ of 4 )
NOTE I' BLOWER SPEED CONTROL LOOPS AND OXIDATION TANK HEADER
FLOW CONTROL LOOPS CAN PRESENTLY BE COMPUTER CONTROLLED
NOTE 2 (kW)(l.34) = hp
NOTE3. (kPo)(O.I45)= psl
Figure A-14. Automatic dissolved oxygen control system - San Jose/Santa Clara
Water Pollution Control Plant, California.
-------
0 High pressure blower discharge manifold pressure control system
with pressure sensor, pressure transmitter and pressure controller
(1 ea.)
Low pressure header butterfly throttling valves and flow control
systems (14 ea.)
High pressure header butterfly throttling valves and flow control
systems (12 ea.)
Dissolved oxygen probes with analyzer/transmitter (5 ea.) + 5 future
Oxidation reduction potential probes with analyzer/transmitter
(10 future)
Computers (2 ea.)
Operation
All six blowers at the San Jose-Santa Clara plant are currently computer
controlled to deliver a specific flow rate for the low pressure system and a
specific pressure for the high pressure system. However, starting and stop-
ping blowers must be done manually from local control panels. The computer
receives a linearized flow signal from the FIT connected to the square root
extractor (FY) for the pitot tube (FE) in the low pressure discharge manifold.
Using a propoportional and integral flow control algorithm, the computer
decrements or increments an output signal that controls the fuel supply to the
blower engine drives. The high pressure blower speed is controlled in a
similar manner with the exception that the computer receives a pressure signal
from the PIT connected to a pressure sensor in the high pressure discharge
manifold, and a pressure control algorithm is used.
The low pressure or distributed air supply headers are each flow controlled
by the computer. The computer receives a signal from the FIT connected to the
header flow tube , inputs it to a flow control algorithm with a preset set point
and outputs a control signal directly to the butterfly positioner in the header.
The high pressure or diffused air supply to each oxidation tank and nitrifi-
cation tank is also flow controlled by the computer. Currently, the computer
receives a flow input from the FIT in each high pressure air feed header, com-
pares it to a set point using a proportional and integral control algorithm and
outputs valve position changes as required to the butterfly valve positioners in
each air header.
In mid-1975, the plant installed dissolved oxygen probes in the effluent
end of five of the oxidation tanks in one battery. At the computer console,
208
-------
the operator will be able to select DO, plant flow or other variable as a con-
trol reference for computer computation of a flow control set point for the flow
control algorithm in each high pressure air feed header.
Performance
For the purposes of this study and for a performance analysis of the new-
ly installed dissolved oxygen probes, the San Jose/Santa Clara plant staff
agreed to run a DO control study test. A DO probe was installed in the efflu-
ent end of four oxidation tanks in Battery B and the DO output wired to the
computer. A program called for printout of each DO probe reading at 15-minute
intervals. Testing began on October 21 , 1975, and ran for a total of eight
days. Each oxidation tank was operated under manual DO control October 21,
23, 25 and 27 and under automatic DO control on October 22, 24, 26 and 28.
Under manual DO control, the header air feed valve on the 55 kPa (8 psig)
header to each tank was manually modulated approximately every four hours.
The amount of valve position change required was estimated based on the
computer printout of DO in the respective tank.
Under automatic DO control, the computer modulated the air header feed
valves as required to maintain a DO set point of 2.5 ppm in each tank. The
control algorithm included proportional and integral control modes.
Blower control systems under manual and automatic DO control were
operated as described above in the section on performance. DO in each tank
was maintained at approximately 2.5 ppm throughout the testing program.
Results of the performance tests are shown in Table A-12.
Table A-12 shows a general improvement in almost all performance para-
meters under automatic DO control. In particular, the air supplied per unit
quantity of BOD removed improved over 12 percent. Improvement of this para-
meter would have been about 16 percent if data obtained on October 26 is
neglected. For some unexplained reason, the air supplied per amount of BOD
removed was about 28 percent higher in this day than any other day on auto-
matic DO control mode.
Maintenance
Since the DO control system was only recently installed, maintenance
data is not available.
209
-------
TABLE A-12 . PERFORMANCE COMPARISON OF MANUAL AND AUTOMATIC
DISSOLVED OXYGEN CONTROL AT THE SAN JOSE/SANTA CLARA
WATER POLLUTION CONTROL PLANT
Parameter
BOD removal efficiency,
percent
Suspended solids removal
efficiency, percent
Sludge volume index
Air supplied per unit quantity
of influent,
3/3
m /m
(cf/gal)
Air supplied per unit quantity
of BOD removed, d
m /kg
(cf/lb)
Manual
85
86
102
6.7
(0.89)
37
(600)
. c
Automatic
85
86
101
6.0
(0.80)
33
(520)
Percent
improvement
none
none
1
10
11
Data from continuous tests on October 21-28, 1975; automatic control on
alternate days .
Average daily flow - 1.97 m /s (45.0 mgd) .
3
Average BOD applied to oxidation tanks - 10.1 mg/m /s (54.3 lb/1000
cf/day) .
c 3
Average daily flow - 2.01 m /s (45 .9 mgd) .
Average BOD applied to oxidation tanks - 10.4 mg/m /s (56.1 lb/1000
cf/day) .
d
Computed from total air supplied over testing period and 24-hour composites
of primary effluent BOD minus secondary effluent BOD.
210
-------
CASE HISTORY 12
CRANSTON WATER POLLUTION CONTROL FACILITY,
CRANSTON, RHODE ISLAND
Description of Aeration System
The City of Cranston Water Pollution Control Facility was built in 1942
with an average dry weather flow design of 0.20 m^/s (5 mgd). During 1964,
it was expanded to its present average flow capacity of 0.50 m^/s (11.4 mgd).
The plant is an activated sludge facility using the contact stabilization pro-
cess. Aeration facilities include four oxidation and four contact stabilization
tanks. However, one oxidation tank and associated contact tank is not cur-
rently being used. Return sludge can be applied directly to each oxidation
tank inlet, or first to the contact stabilization tank and then to the oxidation
tank at any desired ratio. Presently, the plant operates with a 30 percent
direct return and 70 percent diversion to contact stabilization.
The aeration system for the plant expansion (tanks 3 and 4 and associated
contact tanks 3 and 4) is made up of two independent parts called the upper
and lower aeration systems. Aeration air is provided by lobe type, rotary,
positive displacement blowers, driven by 220 kW (300 hp) wound rotor motors.
Each blower may be remotely controlled to operate at 1/2, 3/4 or full speed.
Upper aeration air is delivered to all oxidation and contact stabilization
tanks through a single manifold and applied to each tank via a separate feed
header. It is introduced through diffusers at a high level in the tank at about
34 kPa (5 psi) .
Lower aeration air is provided to oxidation tanks 1 and 2 and contact
tanks 1 and 2 by the old blower system and to oxidation tanks 3 and 4 and
contact tanks 3 and 4 by the new blower system. Delivery pressure is approx-
imately 48 kPa (7 psi). Lower aeration air is introduced to each new oxida-
tion tank and associated contact tank by a single header.
Description and Operation of the Dissolved Oxygen Control System
In late 1975, a demonstration project was underway at the plant in which
the Environmental System group of the Raytheon Company intended to show the
economic benefits and improvements in plant operation realized by micro-
computer control of the activated sludge process. DO control will be an in-
211
-------
tegral part of the overall process control system. Four DO probes have been
located in oxidation tank 3.
Since the micro processer and other control components are not presently
installed, an automatic DO control system, as illustrated in Figure A-15, was
temporarily arranged for the purpose of this study. In Figure A-15, air flow
to oxidation tank 3 is modulated by throttling a motor operated butterfly valve
in the air feed header to tank 4 through analog flow controller (FIC). The FIG
set point is provided by DO controller (AIC), which receives a process input
from the DO probe (AE) located ten feet from the tank effluent weir. At a
given blower speed, a constant amount of air is discharged from the positive
displacement blower. By varying the resistance in the air feed header to
tank 4 through the butterfly valve, more or less air is diverted to tank 3.
Performance
For this study, the Cranston plant staff and personnel from Raytheon
agreed to perform 24-hour DO control tests. One manual and two automatic
DO control tests were run during the period November 20-26, 1975.
Under manual DO control, air flow to basin 3 was adjusted every four
hours to maintain 1 ppm of DO. Under automatic DO control, the set point
on the AIC was set at 1 ppm and the control system was operated as described
above. Test procedures were performed as described in the first part of
Section VII of this manual. Manual DO control was effected by manually
setting the FIC set point every four hours rather than having it set by the AIC.
Blower speed in both tests was manually varied in accordance with the system
requirements.
During the first automatic DO control test, a thunderstorm caused a one-
hour power outage. Accordingly, the automatic test was rerun. Table A-13
shows results of all three DO control tests.
In general, automatic DO control provided an increase in BOD removal
and a decrease in air supplied. Results for suspended solids removal are
inconclusive since removal increased in the first automatic test and decreased
in the second automatic test.
SVI data reported in Table A-13 is an average of six samples taken during
each test. SVI's at the plant generally run about 50. A graph of SVI data
shows 70 percent of all samples fell within a range of 40-60. The remaining
samples were all higher than 60 with several samples during the automatic
tests being above 100. Insufficient data exists to draw any conclusions with
regard to SVI readings and DO control.
212
-------
t\J>
I—'
LO
-*- TO OXIDATION TANK No. 4
I
% s
'F I C'\ ! S P Ml CN
TGI
AIT
PE
RAS
ML
S 1 S I I ! 1 I
3-SPEED
POSITIVE DISPLACEMENT
BLOWER
OXIDAT ION TANK No. 3
Figure A-15. Automatic dissolved oxygen control system - Cranston
Water Pollution Control Facility, Rhode Island.
-------
TABLE A-13. PERFORMANCE COMPARISON OF MANUAL AND AUTOMATIC
DISSOLVED OXYGEN CONTROL AT THE CRANSTON WATER
POLLUTION CONTROL FACILITY
Parameter
BOD removal efficiency,
percent
Suspended solids removal
efficiency, percent
Sludge Volume Index
Air supplied per unit quantity
of influent,
m3/m3
(cf/gal)
Air supplied per unit quantity
of BOD removed,
m3/ks
(cf/lb)
Manual
91
89
56
19
(2.5)
79
(1300)
Automatic0
t 1
94
96
60
15
(2.0)
58
(940)
Percent
improvement
3
7
-7
21
27
Automatic
#2
92
83
69
16
(2.2)
56
(900)
Percent
improvement
1
-6
-23
16
29
Data from 24 hour tests November 20-26, 1975.
b -i
Average flow to oxidation tank 3 -0.16 m /s (3 .60 mgd).
Average BOD applied to total volume of oxidation tank 3 & contact tank 3 9.08 mg/m3/s
(49.0 lb/1000 cf/day) .
Average flow to oxidation tank 3 0.20 rn3/s (4.59 mgd).
Average BOD applied to total volume of oxidation tank 3 & contact tank 3 - 12.2 mg/m3/s
(65.6 lb/1000 cf/day).
Average flow to oxidation tank 3 0.19 m /s (4.23 mgd).
Average BOD applied to total volume of oxidation tank 3 & contact tank 3 13.2 mg/m3/s
(71.3 lb/1000 cf/day).
214
-------
Maintenance
Since installation in August, 1975, until the DO control testing period
described above, DO probe maintenance has consisted of one cleaning and
one membrane replacement for each probe. DO probe readings have been
found to consistently agree well with DO analysis by the Winkler method.
215
-------
REFERENCES FOR APPENDIX3
A-l Belick, M. and Van Kirk, F. N., "California Plant Gets Straight A's
in Computer Control", Water and Wastes Engineering, Mar., 1975,
pp 20-24.
A-2 Jones, K., Briggs, R.; Carr, J. G. and Potten, A, H., "Automatic
Control of Aeration in a Fully Nitrifying Activated Sludge Plant",
Institute of Public Health Engineering Journal, Vol. 58, Oct., 1969,
pp 271-295.
A-3 Lewin, V.. H. and Henley, J. R., "Automation of an Activated Sludge
Plant", Process Biochemistry, Feb., 1972, pp 17-20.
A-4 Meredith, W. D. , "Dissolved Oxygen Control of Activated Sludge
Process", presented at the September 17-20, 1972, International Work-
shop, Instrumentation Control and Automation for Wastewater Treatment
Systems, held at London, England.
A-5 Petersack, J. F. and Smith, R. G., "Full-Scale Demonstration of
Advanced Automatic Control Strategies for the Activated Sludge Treat-
ment Process", Environmental Protection Technology Series, EPA-
670/2-75-039. May, 1975.
A-6 Roesler, F., "Plant Performance Using Dissolved Oxygen Control",
Journal of the Environmental Engineering Division. ASCEVol. 100,
No. EE5, Oct., 1974.
A-7 "San Jose-Santa Clara Expand Water Pollution Control Plant", Power/
Compression News, Spring, 1965, pg 8.
Much of the information contained in the Appendix was obtained through
private communications between Brown and Caldwell staff members and
responsible operating and management personnel at the various plants.
Other information was extracted from Brown and Caldwell job files.
216
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-77-032
2.
3. RECIPIENT'S ACCESSI OP* NO.
4. TITLE AND SUBTITLE
DESIGN PROCEDURES FOR DISSOLVED OXYGEN CONTROL
OF ACTIVATED SLUDGE PROCESSES
5. REPORT DATE
June 1977 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Michael J. Flanagan
Brian D. Bracken
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Brown and Caldwell Consulting Engineers
1501 North Broadway
Walnut Creek, California 94596
10. PROGRAM ELEMENT NO.
1BC611
11. CONTRACT/GRANT NO.
68-03-2130
12. SPONSORING AGENCY NAME AND ADDRESS
OH
Municipal Environmental Research Laboratory—Cin.,
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final June, 1974 to Oct. 1976
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report presents design procedures and guidelines for the selection of
aeration equipment and dissolved (DO) control systems for activated sludge
treatment plants. A review of process configurations and design parameters
is made to establish system requirements. Aeration methods, equipment and
application techniques are examined and selection procedures offered. Various
DO control systems are described with recommendations for system applications
to various aeration equipment types and process configurations. Performance5
operational and maintenance data for aeration equipment and DO control systems
for twelve activated sludge plants is presented. This information and other
design recommendations in the report are used to develop automatic DO control
systems for various size hypothetical activated sludge system configurations
for an economic analysis of manual and automatic DO control. The conclusion
is drawn that the capital and operating costs of automatic DO control systems
are justified for activated sludge plants larger than 1 mgd (44 dm /s)
provided equipment is selected and applied in accordance with the guidelines
of the design manual and a power cost equal to or greater than the national
average power rate is applicable.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Activated Sludge Process, Dissolved
Gases, Oxygen, Automation, Automatic
Control, Cost Effectiveness,
Instruments, Process Control, Sewage
Treatment, Waste Treatment, Control
Equipment
Dissolved Oxygen Control,
Monitoring
13B
3. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
233
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
217
ftUS GOVERNMENT PRINTING OFFICE 1977—757-056/6417
------- |