United States
Environmental Protection
Agency
Office of Research and Development
Office of Water
Washington, DC 20460
EPA/625/R-93/010
September 1993
Manual
Nitrogen Control
-------
-------
EPA/625/R-93/010
September 1993
Manual
Nitrogen Control
U.S. Environmental Protection Agency
Office of Research and Development
Center for Environmental Research Information
Risk Reduction Engineering Laboratory
Cincinnati, Ohio
Office of Water
Office of Wastewater Enforcement and Compliance
Washington, DC
Printed on Recycled Paper
-------
Notice
This document has been reviewed in accordance with the U.S. Environmental Protection Agency's peer and
administrative review policies and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
-------
Contents
Page
Chapter 1 Introduction
1.1 Background and Purpose 1
1.2 Scope of the Manual 2
1.3 How to Use This Manual 3
1.4 Nitrogen in the Environment 4
1.5 Effects of Nitrogen in Discharges from Wastewater Treatment Plants 12
1.6 Establishing Nitrogen Limits for Wastewater Discharges 19
Chapter 2 Total System Design with Nitrogen Control
2.1 Introduction 23
2.2 Summary Checks for Process Selection and Design 23
2.3 Fundamental Treatment Technology Concepts 29
2.4 Lower Technology, Transitional, and Natural System Approaches 38
2.5 Physical/Chemical Treatment Processes 44
2.6 Key Fundamental Process Selection and Design Issues 49
2.7 Frequently Encountered Linked Design Issues 54
2.8 Nitrogen Control Troubleshooting and Problem-Solving 59
2.9 The Design Examples 64
2.10 References 85
Chapter 3 Process Chemistry and Kinetics of Biological Nitrification
3.1 Introduction 87
3.2 Fundamentals of Nitrification 87
3.3 Nitrification Kinetics 88
3.4 Attached Growth Kinetic Considerations 96
3.5 References 97
Chapter 4 Process Chemistry and Kinetics of Biological Denitrification
4.1 Introduction 101
4.2 Fundamentals of Denitrification 101
4.3 Kinetics of Denitrification 104
4.4 References 110
-------
Contents (continued)
Page
Chapters Mathematical Modeling of Nitrification and Denitrification
5.1 Introduction ... 111
5.2 Modeling Approaches 111
5.3 Model Development 113
5.4 Model Applications , : 114
5.5 Available Models 116
5.6 Example: Analysis of a Single-Sludge Nitrification/Denitrification System. 122
5.7 References 126
Chapter 6 Design Considerations for Biological Nitrification Processes
6.1 Introduction 129
6.2 Classification of Nitrification Processes 129
6.3 Comparison of Nitrification Systems at Higher and Lower Carbonaceous
Feed Concentration 129
6.4 Nitrification in Suspended Growth Reactors :....... 132
6.5 Nitrification in Attached Growth Reactors , 164
6.6 Combined or Coupled Suspended and Attached Growth Systems 197
6.7 References 203
Chapter 7 Design Considerations for Denitrification Processes with Supplemental
Substrate Addition
7.1 Introduction 211
7.2 Suspended Growth Systems 212
7.3 Downfiow Packed-Bed Systems 217
7.4 Upflow Fluidized-Bed Systems 232
7.5 Methanol Handling, Storage, Feed Control, and Excess Methanol Removal 242
7.6 References 247
Chapter 8 Design Considerations for Single-Sludge Nitrification-Denitrification
Processes
8.1 Introduction 249
8.2 Classification of Single-Sludge Processes 250
8.3 Process Selection Considerations 283
8.4 Design Considerations 289
8.5 Process Design Examples .. 291
8.6 References '. 310
8.7 Bibliography .311
IV
-------
List of Figures
Figure Page
1-1 The nitrogen cycle 5
1-2 The nitrogen cycle in surface water. 7
1-3 The nitrogen cycle in soil and ground water 8
1-4 Time history data analysis for main stream of Patuxent River (station: PXT0708
State of Maryland) 15
1-5 Effect of temperature and pH on un-ionized ammonia. 16
1-6 EPA chronic criteria for ammonia (salmonids absent) ...-,. 17
1-7 Depth profiles of NH3, FCV, and 0.5 x FAX in Onondaga Lake for June 19, 1988 18
1-8 EPA water quality standards: Review and revision process 20
1-9 Major elements of the water quality-based standards to permits process. 21
2-1 pH dependency of selected reactions of interest 45
2-2 Typical breakpoint chlorination curve and stoichiometric reactions for an advanced
treatment effluent ..;..... 47
2-3 Schematic of how carbon-to-nitrogen ratio influences development of a nitrogen
control strategy along with the design response. 52
2-4 Basic schematics of "simple" and "complex" wastewater treatment facilities with a
typical flow range (4 to 440 L/s). 65
2-5 Detailed schematic of "simple" wastewater treatment facility (Plant A) 66
2-6 Detailed schematic of "complex" wastewater treatment facility (Plant B) 67
2-7 Relationship between ammonia and hydraulic peaking loads for treatment plants
with no in-process equalization 70
2-8 Likely soluble CBOD5 and COD removals as a function of mean solids retention
time in a biological reactor(s) 72
2-9 Characterization of biological reactor substrate distribution 74
2-10 Estimates of volatile solids production 75
3-1 Effect of temperature on oxidation of ammonium by Nitrosomonas.. 91
3-2 Effect of reactor pH conditions on rate of nitrification 93
3-3 Schematic of conceptual biofilm model 97
4-1 Effect of temperature on denitrification rates 108
4-2 Effect of pH on denitrification rates 109
5-1 Conceptual nitrification model 111
5-2 Stepwise approach to model development 113
5-3 Division of organic matter in Activated Sludge Model No. 1 116
6-1 A listing of the majority of reactor configurations available for nitrification.. ,. 130
6-2 Suspended growth reactor configurations 133
6-3 Variation in observed nitrification rates 136
6-4 Simplified schematic for Design Example No. 1 employing a complete mix
suspended growth reactor for nitrification 137
6-5 Effect of design factor on steady state effluent ammonia levels in complete mix
and plug flow suspended growth reactors 144
6-6 DO and ammonium-nitrogen profile in a plug flow system 145
6-7 Oxidation ditch system 151
6-8 Covered high-purity oxygen reactor with three stages and mechanical aerators. ....... 155
6-9 Single-tank SBR system operating steps 157
6-10 Powdered activated carbon activated sludge system. 159
-------
Figures (continued)
Figure Page
6-11 Hydraulic flow patterns in vertical and cross-flow trickling filter media 167
6-12 Effect of BOD5 volumetric loading on nitrification performance of rock trickling
filters 168
6-13 Effect of BOD5 surface loading on nitrification efficiency of rock and plastic
media trickling filters 168
6-14 Comparison of performance of various trickling filter media 169
6-15 Correlation between TKN surface removal rate and wastewater BOD5:TKN
proposed by American Surfpac Corporation 171
6-16 Common process variation of the trickling filter solids contact process 172
6-17 Effect of BOD5 surface loading on nitrification performance. 173
6-18 Ammonium surface loading versus removal rate 175
6-19 Effect of bulk liquid DO conditions on surface loading versus removal rate
correlation 175
6-20 Effect of temperature on nitrification rates in trickling filter systems 176
6-21 Nitrification performance on trickling filters receiving low enfluent CBOD5—
Plants A, B, and C 179
6-22 Nitrification performance of trickling filters receiving low influent CBOD5—
Plants D and E 179
6-23 Typical application of rotating biological contactors to municipal wastewater
treatment 183
6-24 RBC design curves developed from Equation 6-33 184
6-25 Temperature correction factors for required RBC area 185
6-26 Effect of organic substrate loading on the rate of nitrification 186
6-27 Full-scale RBC nitrification rates at a wastewater temperature of 13°C (55°F) 186
6-28 Typical RBC design curve 188
6-29 Nitrification design relationships 189
6-30 RBC configuration for Design Example 4 192
6-31 Biocarbone BAF system example flowsheet 194
6-32 Effect of COD volumetric loading on Biocarbone BAF nitrification performance in
the treatment of primary effluent 195
6-33 Effect of COD volumetric loading on nitrification of primary treated wastewater in
a Biocarbone BAF pilot unit 196
6-34 ABF/AS process schematic 198
6-35 Effect of upstream nitrification on the effluent NHJ-N from downstream
suspended growth unit 199
6-36 Comparison of predicted and actual monthly average effluent NHJ-N for Duck
Creek 200
6-37 Nitrogen control schematic for Stow, Massachusetts 202
7-1 Schematic of suspended growth system 212
7-2 Design example schematic of suspended growth system 217
7-3 Observed denitrification rates for suspended growth systems using methanol 218
7-4 Schematic of River Oaks Advanced Wastewater Treatment Plant. 218
7-5 Linear schematic of downflow packed-bed system 219
7-6 Cross-section schematic of downflow packed-bed system 220
7-7 Typical design curves for empty-bed contact time.. 222
7-8 Effect of nitrate concentrations on loading rates in downflow packed-bed systems 223
7-9 Downflow packed-bed denitrification performance 224
7-10 Alternative filter underdrain systems 225
7-11 Design example schematic of downflow packed-bed system 226
7-12 Schematic of Hookers Point Advanced Wastewater Treatment Plant 232
7-13 Schematic of Dale Mabry Wastewater Treatment Plant 232
7-14 Schematic of upflow fluidized-bed system 234
VI
-------
Figures (continued)
Figure Page
7-15 Temperature vs. loading rate for upflow fluidized-bed system 242
7-16 Design example schematic of upflow fluidized-bed system 242
7-17 Schematic of Reno-Sparks Wastewater Treatment Plant 243
7-18 Schematic of automatic feed forward control system for methanol pacing 246
8-1 Wuhrmann process 251
8-2 Ludzack-Ettinger process 251
8-3 Modified Ludzack-Ettinger process 253
8-4 A2/O process with nitrification-denitrification 253
8-5 University of Capetown (UCT) process 253
8-6 MLE mass balance schematic. 255
8-7 Theoretical oxidized TKN removal performance for a single anoxic zone process
as a function of total recycle rate ,.. 257
8-8 Theoretical oxidized TKN removal performance for a single anoxic zone process
as a function of internal recycle rate 257
8-9 Bardenpho process 259
8-10 Modified UCT process 260
8-11 Multi-anoxic zone with step feed 260
8-12 Bardenpho mass balance schematic 263
8-13 Bardenpho process nitrate removal as a function of recycle rates and
denitrification performance 264
8-14 Modified UCT mass balance schematic 264
8-15 Modified UCT process nitrate removal as a function of internal and return sludge
recycle rates 265
8-16 Multi-anoxic zone step-feed process 266
8-17 Theoretical percent nitrogen removal as a function of COD:TKN for a triple
anoxic zone process with step feed 266
8-18 Town of Owego, NY, water pollution control plant 268
8-19 Schreiber process 268
8-20 Vienna-Blumenthal Wastewater Treatment Plant 271
8-21 Orbal oxidation ditch 271 ,
8-22 Carrousel oxidation ditch 272
8-23 Orbal Sim-Pre process 273
8-24 Kruger BioDenitro process (Type DE) 273
8-25 Kruger BioDenitro process (Type T) 274
8-26 Sequencing batch reactor 278
8-27 Cyclical Activated Sludge System 279
8-28 Intermittent Cycle Extended Aeration System 280
8-29 Suggested operating strategies for SBR systems 281
8-30 Denitrication rate as a function of anoxic F/M 297
8-31 Design Example No. 1—Single anoxic zone system. 300
8-32 Design Example No. 2—Dual anoxic zone system 304
8-33 Design Example No. 3—CNR process. 308
8-34 Design Example No. 4—Dual anoxic zone system 309
VII
-------
List of Tables
Table Page
1-1 Major Direct Sources of Nitrogen 9
1-2 Atmospheric Nitrogen: Representative Concentrations and Unit Areal Loadings 11
1-3 Representative Distribution of Sources of Nitrogen to Chesapeake Bay, Long
Island Sound, and Swedish Coastal Areas 13
1-4 Effect of Ammonium Oxidation on Total Oxygen Demand of Treated Wastewater
Discharge 14
1-5 Examples of DO Improvement to River Segments Because of Treatment Plant
Upgrades to Nitrification 15
2-1 Survey of 150 Small Plants with Debilitating Problems 25
2-2 Wastewater Characterization and Solids Management Checks for the Design of
Municipal Wastewater Treatment Facilities with Nitrogen Control Technology 26
2-3 Stoichiometry of Nitrogen Control and Other, Often Related Reactions 28
2-4 The Three Most Important Review Checks for Nitrogen Control Facilities 30
2-5 Classification of Wastewater Treatment.Processes ; 33
2-6 1990 Status of Nitrogen Control Technologies in Municipal Wastewater Treatment
Applications 35.
2-7 Lower Technology, Transitional, and Natural System Screening Criteria 39
2-8 Comparative Assessment of Suspended and Attached Growth Technologies
Against Key Process Selection and Design Issues 55
2-9 Step-by-Step Recommendations for Wastewater Treatment Plant Problem-Solving 61
2-10 Effluent Objectives of the Design Examples 69
2-11 Design Examples: Average Day Raw Influent Wastewater Characterization 71
2-12 Design Examples: Influent Wastewater Peaking Factors 71
2-13 Volatile Solids Production Considerations and Assumptions 73;
2-14 Remaining Assumptions for Example Mass Balances 77
2-15 Mass Balance for Plant A 79
2-16 Mass Balance for Plant B 80
2-17 More Complex Plant B Solids Processing Recycle Impacts on Main Processing
Stream 83
2-18 Summary of Controlling Design Conditions for Biological Reactor with Design
Examples 84
3-1 Oxygen Utilization, Biomass Yield, and Alkalinity Destruction Coefficients
Acceptable for Design of Nitrification Systems 88
3-2 Maximum Specific Growth Rates and Half-Saturation Coefficient Values for
Nitrosomonas at Constant Temperature (20°C) 90
3-3 Maximum Specific Growth Rate Values for Nitrosomonas as a Function of '.
Temperature 90
3-4 Industrially Significant Organic Compounds Inhibiting Nitrification 94
3-5 Metals and Inorganic Compounds Identified as Potential Nitrification Inhibitors 94 •
3-6 Calculated Threshold Values of Ammonia Plus Ammonium-Nitrogen and Nitrite
Plus Nitrous Acid-Nitrogen Where Nitrification Inhibition May Begin 94
4-1 Values for Denitrification Yield and Decay Coefficients for Various Investigations
Using Methanol 106
4-2 Temperature Correction Coefficients for Modeling Denitrification (Endogenous
Rate) 107
viii ,
-------
Tables (continued)
Table Page
5-1 Activated Sludge Model No. 1: System Components 117
5-2 Activated Sludge Model No. 1: Kinetic and Stoichiometric Parameters 118
5-3 Activated Sludge Model No. 1: Kinetic Expressions 118
5-4 Activated Sludge Model No. 1:.Values of Stoichiometric and Kinetic Parameters 119
5-5 Activated Sludge Model No. 1 Presented in Matrix Format 120
5-6 Examples: Use of IAWPRC Activated Sludge Model No. 1 121
5-7 Example Nitrification and Denitrification Mathematical Modeling Microcomputer
Packages 122
5-8 SSSP Input Parameters and Output for Example Problem 124
5-9 Comparison of Process Designs for Complex and Simple Wastewater Treatment
Plants and Effluent Total Nitrogen of 10 and 5 mg-N/L 125
6-1 Classification of Nitrification Facilities 130
6-2 Design Conditions for Example 1: Plant B in a Complete Mix Configuration with
Higher Carbonaceous Feed and More Stringent Effluent Requirements 137
6-3 Summary of Example 1 Design Results: Plant B at Higher Carbonaceous Feed
Concentration and More Stringent Effluent Requirements 143
6-4 Design Information for Nitrification of a Low Carbonaceous Feed Concentration 146
6-5 River Oaks Advanced Wastewater Treatment Plant—Design Information for
Carbonaceous Oxidation/Nitrification System Following Primary Clarification 149
6-6 River Oaks Advanced Wastewater Treatment Plant—Carbonaceous
Oxidation/Nitrification System Operating Conditions August 1988 to July 1989 149
6-7 River Oaks Advanced Wastewater Treatment Plant—Carbonaceous
Oxidation/Nitrification System Performance, August 1988 to July 1989 150
6-8 Jackson, Michigan, Wastewater Treatment Plant Nitrification System Operation
Conditions and Performance Results, 1990 150
6-9 Frederick, Maryland, Wastewater Treatment Plant Design Information and
Operating Performance Results 152
6-10 Average Nitrification Performance at Flint, Michigan, for Eight Months 154
6-11 Effect of Temperature and Solids Residence Time on Nitrification Efficiency at
Flint, Michigan 154
6-12 Town of Amherst, New York, Wastewater Treatment Plant Carbonaceous
Oxidation-Nitrification System Design Information Following Primary Clarification ...... 156
6-13 Town of Amherst, New York, Wastewater Treatment Plant Carbonaceous
Oxidation-Nitrification System Operating Conditions and Performance Results,
October 1990 to September 1991 157
6-14 Typical Cycle for a Single Tank in a Dual Tank SBR System Designed for
Nitrification 158
6-15 Nitrification Performance Information for SBR Operating Plants 159
6-16 Summary of Municipal PAC/WAR Facilities Reviewed 160
6-17 Comparative Physical Properties of Example Synthetic Trickling Filter Media
Suitable for Nitrification of Municipal Wastewaters 166
6-18 Amherst, Ohio, Wastewater Treatment Plant Carbonaceous Oxidation-
Nitrification System Operating Conditions and Performance Results,
February 1989 to January 1990 172
6-19 Wauconda, Illinois, Wastewater Treatment Plant Carbonaceous
Oxidation-Nitrification System Operating Conditions and Performance Results,
January 1989 to December 1989 : 173
6-20 Calculated Trickling Filter Nitrification Model Parameters from Pilot Plant Studies 178
6-21 Comparison of Measured and Predicted Nitrification Rates 178
6-22 Annual Operating Information from Five Nitrifying Trickling Filters Receiving Low
Influent CBOD : 179
ix
-------
Tables (continued)
Table Page
6-23 Design Information for Nitrification of a Low Carbonaceous Feed Concentration
and Less Stringent Effluent Requirements 180
6-24 Manufacturer Recommendations for RBC Staging 187
6-25 Design Conditions for Example No. 4 190
6-26 Trickling Filter Removals at Livermore, California 199
6-27 Nitrification Performance from ABS/AS Pilot Studies 201
6-28 Average Influent and Effluent Data for Stow, Massachusetts 203
7-1 Design Example: More Complex Plant B with Year-Round Effluent Limits-
Suspended Growth System 213
7-2 River Oaks Wastewater Treatment Plant: Monthly Performance Data 219
7-3 Packed-Bed Denitrification (Partial Installation List) ... 221 i
7-4 Selected Downflow Packed-Bed Application Rates 223
7-5 Typical Design Criteria for Downflow Packed-Bed System 225
7-6 Influent Characteristics and Design Effluent Limits for Denitrification Design
Examples: Downflow Packed-Bed System 226
7-7 Operating Data for Selected Downflow Packed-Bed Systems 231
7-8 City of Tampa, Hookers Point Wastewater Treatment Plant: Monthly Performance
Data 233
7-9 Dale Mabry Wastewater Treatment Plant: Monthly Performance Data 234
7-10 Full-Scale Applications of Upflow Fluidized-Bed Technology for Nitrogen Control 235
7-11 Process Design Information for Upflow Fluidized-Bed Systems 236
7-12 Types of Fluidized-Bed Denitrification 236
7-13 Selected Upflow Fluidized-Bed Loading Rates -. 237
7-14 Upflow Fluidized-Bed System Design Examples—Influent Characteristics and
Effluent Limits 238
7-15 Reno-Sparks Wastewater Treatment Plant: Monthly Performance Data 243
7-16 Key Properties of Methanol 244
8-1 Typical Design Criteria for Single Anoxic Zone Predenitrification Systems 254
8-2 Performance Summary of Single Anoxic Zone Processes 256
8-3 Monitoring Requirements and Rationale for Single Anoxic Zone Reactors 258
8-4 Typical Design Criteria for Dual Anoxic Zone Systems 261
8-5 Summary of Bardenpho Plant Operating Data 262
8-6 Monitoring Requirements and Rationale for Bardenpho Reactors 267
8-7 Cyclical Aeration Design Criteria 269
8-8 Cyclical Aeration Operating Results 269
8-9 Operating Parameters at Various Oxidation Ditches 275
8-10 Design Parameters for Orbal and Orbal Sim-Pre Process 276
8-11 Nitrogen Removal Performance for Various Oxidation Ditch-Type Plants 276
8-12 Typical Design Criteria for Sequencing Batch Reactors 281 |
8-13 Summary of SBR Plant Operating Data , 282
8-14 Design Example Effluent Limits 292
-------
Acknowledgments
This manual is an update of the EPA Nitrogen Design Manual released in 1975. In a number of
instances, information contained in that manual was incorporated directly into the current revision.
Also material from an unpublished manual revision prepared in 1985 was selectively incorporated into
the current revision.
This manual resulted from the contributions of numerous individuals active in the field of wastewater
treatment and nitrogen control. The manual content and focus reflect the consensus achieved at a
planning meeting to determine the manual scope held in December, 1990. Persons participating in
this meeting were O. K. Scheible (HydroQual, Inc.), D. Parker (Brown and Caldwell Consultants), G.
Daigger (CH2M Hill), M. Mulbarger (Paladin Enterprises), A. Condren (James M. Montgomery, Inc.),
E. J. Middlebrooks (Univ. of Tulsa), E. Barth (BarthTec), D. Schwinn (Stearns and Wheler), and several
EPA personnel.
Authors of the manual were: O. K. Scheible, Chapter 1; M. Mulbarger, Chapter 2; P. Sutton (P. M.
Sutton and Assoc.), Chapter 3; T. Simpkin (CH2M Hill), Chapter 4; G. Daigger, Chapter 5; P. Sutton
with assistance from O. K. Scheible, M. Mulbarger, and J. Heidman (EPA), Chapter 6; M. Yoder
(CH2M Hill), Chapter 7; and D. Schwinn and D. Starrier (Stearns and Wheler), Chapter 8.
Official EPA peer review was provided by K. Deeny (Junkins Engineering), T. Wilson (Greeley and
Hansen) and D. Stensel (University of Washington). In addition, peer review comments were also
provided by E. Barth, D. S. Parker, S. Gilbert (Parametrix), C. Pycha (EPA), J. Wheeler (EPA), and
Norbert Huang (EPA).
O. K. Scheible and J. Heidman had the responsibility to edit the various chapters and synthesize them
into a coherent document. Overall contract management was the responsibility of A. Condren and R.
Stevenson (James M. Montgomery Consulting Engineers) and H. Schultz (Eastern Research Group,
Inc.).
In an undertaking of this magnitude and for a topic as diverse and complex as that addressed, it was
not possible to reach unanimous agreement among all authors and reviewers on each issue. The
reader should be aware that approaches or opinions different from those expressed herein may be
equally applicable to a given design situation.
Funding for this manual was provided by the Risk Reduction Engineering Laboratory (RREL), the
Office of Wastewater Enforcement and Compliance (OWEC), and the Center for Environmental Re-
search Information (CERI). The RREL Project Officer was J. Heidman and the CERI Project Officer
was R. Revetta.
XI
-------
-------
Chapter 1
Introduction
1.1 Background and Purpose
The growing demand for water resources has generated
an equivalent need for effective water and wastewater
management strategies. This need is heightened by an
awareness that a sound hydrosphere is fundamental to
the world's ecology. In the United States these driving
forces have found support in the regulations that mandate
the proper handling and treatment of wastewaters dis-
charged to the aquatic environment.
Application of processes to control wastewater dis-
charges to the aquatic environment began around the
turn of the century, generally addressing the removal of
particulates and oxygen-demanding carbonaceous mate-
rials. The removal of other constituents did not receive
substantive attention until the latter part of this century.
The control of nitrogen has subsequently been identified
as an important environmental activity, demonstrated by
the adverse effects that excess levels of different forms
of nitrogen have had on aquatic systems. Ammonium-
nitrogen oxidation processes were fairly widely applied in
England by mid-century. Such processes were not imple-
mented on a significant scale in the United States until
the late 1960s, although early activated sludge and trick-
ling filter plants often did nitrify, even if not explicitly de-
signed to do so. The first full-scale application of a
nitrogen removal process took place in South Lake Ta-
hoe, California, in 1969, although, as a relatively new
technology, it experienced problems.
Research and development activities through the mid-
1970s served as the basis for the U.S. Environmental
Protection Agency's (EPA's) first comprehensive manual
of design for various nitrogen control technologies (1).
The Process Design Manual for Nitrogen Control, pub-
lished in 1975, covered a broad spectrum of processes,
reflecting the diverse approaches being evaluated and
applied at the time. Its intent was'to present design infor-
mation for technologies that appeared to have a viable,
practical application to nitrogen control. Such processes
can be divided to two broad categories. The first group
provides for the conversion of organic and ammonium
nitrogen by oxidation to nitrate nitrogen. These are bio-
logical processes and are generally termed nitrification.
The second group accomplishes the actual removal of
nitrogen from the wastewater, not simply the conversion
from one form to another. This is also typically done by
biological means, coupling an anoxic denitrification step
with nitrification. Physical/chemical processes were also
presented for nitrogen removal, including ion exchange,
ammonia stripping, and breakpoint chlorination.
Since the first manual's publication, the trend in nitrogen
control technology applications has been overwhelmingly
in favor of biological processes, with only a few in-
stances in which physical/chemical processes have
been implemented. Virtually all of the greater than 15,000
publicly owned treatment works (POTWs) in the United
States use biological processes to remove organics.
While the total number of facilities is projected to rise by
approximately 30 percent to meet the demands of the
full U.S. population, those systems identified specifically
for biological nitrogen control are expected to at least
double in number (2,3). Biological processes are proven
and well demonstrated and lend themselves most effi-
ciently to expansion or upgrade for biological nitrification
or total nitrogen removal. Biological processes also mini-
mize the use of often expensive arid sometimes environ-
mentally incompatible chemicals and will generally
achieve residual ammonium levels that are lower than
can be effectively accomplished by most physical/chemi-'
cal processes.
Breakpoint chlorination, ion exchange, and air stripping
processes received considerable attention in the 1975
manual, but have been employed in a very limited number
of POTW applications over the last 15 years. This is not
to say that these technologies are not viable. Practical,
applications for these nonbiological controls exist in some
POTW settings, such as very cold climates or for indus-
trial pretreatment. Where particularly high concentrations
of ammonium nitrogen exist (e.g., landfill leachates), both
physical/chemical and biological processes are utilized.
Conversely, significant attention is being given to natural
processes for nitrogen control, particularly for application
to small systems. These configurations, including natural
and constructed marshes and wetlands are generally
considered to be in an emerging, developmental state at
-------
this point, although with potentially significant, wide-
spread applications.
This manual is an update and revision of the original 1975
edition. It strives to maintain the high technical quality and
generous provision of reference materials provided by the
1975 edition, although it also represents a significant shift
in overall content. Given the experience of the past 18
years, the focus of this second edition is directed to those
biological/mechanical systems that have found wide-
spread use for nitrification and nitrogen removal. Design
criteria for physical/chemical systems are not provided;
however, there is a brief discussion of such processes in
Chapter 2, in which their applicability under specific site
conditions and wastewater applications is addressed. The
design of natural systems also is not considered within
the context of this manual, except in the planning and
development of alternatives for technology selection, a
point of discussion within Chapter 2. Adequate references
are given in Chapter 2 to assist the reader in seeking
design information on both natural systems and physi-
cal/chemical processes.
The primary audience is the designer of small- to me-
dium-sized facilities, although the application of the man-
ual is not limited to any range of plant sizes. The intent
is to provide a manual that can be readily used by those
who do not necessarily specialize in the design of nitro-
gen control facilities. Detailed theoretical discussions are
not provided. Rather, the manual focuses on the major
process and design aspects considered in the develop-
ment of an effective design. It begins with process basics
and proceeds to the presentation of detailed design cri-
teria and the development of process designs, using ex-
amples to demonstrate calculation sequences. In
addition, the manual is organized to help the designer in
the planning stages of a facility, highlighting important
process and operation and maintenance (O&M) consid-
erations. The intention is to give direction before plans
advance to a level at which redesign would be costly. For
design purposes this manual is most prudently used in
its entirety.
The manual also is assembled for use as a desk refer-
ence, or handbook. The table of contents is detailed
enough for the user to be able to find specific topics
quickly, and an effort has been made to present informa-
tion as often as possible in the form of charts, tables, and
figures. In addition to aiding designers, an objective of
the manual is to serve as a source for reviewers, opera-
tors, regulators, and manufacturers.
1.2 Scope of the Manual
The manual is constructed to progress from a broad dis-
cussion of nitrogen in the environment, to the concepts
of using biological processes to control or remove nitro-
gen, and finally to the details of designing specific sys-
tems. This first chapter describes the relationships of ni-
trogen in the environment. The fundamental purpose of
the manual, implementation of nitrogen controls in mu-
nicipal systems, is brought into focus in Chapter 2 by
outlining design principles. An array of issues are pre-
sented that enter into the designer's strategy. The chapter
discusses the relative importance of each issue in order
to help the designer avoid pitfalls, before they are com-
pounded by the growing detail of design. Chapters 3 and
4 give the theoretical foundations of the nitrification and
denitrification processes by drawing on concepts; of mi-
crobiology, chemistry, and kinetics. Although the discus-
sions of mathematical and computer modeling in Chapter
5 are intended to stand independently from the other
chapters, they provide a useful bridge between Chapters
3 and 4 and Chapters 6, 7, and 8.
In the latter three chapters, the conceptual bases of nitri-
fication and denitrification are developed to design crite-
ria, and design examples are presented to assist in
producing a specific configuration that will meet perform-
ance objectives. Chapter 6 addresses suspended growth
and attached growth configurations for biological nitrifica-
tion. Denitrification processes are addressed in Chapter
7, but only as applied in a separate stage using a sup-
plemental carbon source. The current trend has been to
accomplish nitrification and denitrification in single-sludge
systems, using wastewater carbon for the denitrification
step. Single-sludge systems, which are addressed in
Chapter 8, have seen increasing application in lieu of
the alternative two- or three-sludge systems for nitrogen
removal.
A significant portion of the information in this manual has
been carried forward from the 1975 document. Substan-
tial modifications have been made, however, in directing
the emphasis toward bjological/mechanical systems.
Contributions to this work were developed from the fol-
lowing sources: 1) the experience of the individuals in-
volved in the preparation of the manual; 2) the EPA
research, development, and demonstration program; 3)
existing literature; 4) design/performance and operating
experience at existing nitrogen control facilities; 5) pro-
gress reports regarding ongoing projects; 6) private com-
munication with investigators active in the field; and from
7) operating personnel at existing wastewater treatment
plants.
The material presented is a distillation of knowledge re-
flecting relatively few generations of full-scale experience
with nitrogen control technologies. As such, it represents
a perspective of the present state-of-the-art, and not nec-
essarily a complete understanding of the technology. En-
hanced understanding and knowledgeable application of
current demonstrated approaches, coupled with; new
technologies or approaches to join those that are emerg-
ing, represent the expectation and challenge of the future.
-------
1.3 How to Use This Manual
A perusal of the Table of Contents gives the reader an
overview of the subject matter contained in this manual.
All users should read Chapter 2, which serves to set the
background and protocol for effective design. It is a lead-
in to the design chapters (6, 7, and 8) and establishes
the design examples used in these chapters. The follow-
ing chapter-by-chapter description provides a more de-
tailed presentation of the contents and the objectives of
each chapter.
Chapter 1, Introduction, first gives the reader a broad view
of the manual, a perspective on how designers can use
the manual to serve their needs, and an explanation of
the manual's limitations. The remaining sections give a
general overview of the cause-and-effect relationships of
nitrogen in the environment and why nitrogen removal is
important. Sources of nitrogen are identified and the
movement and transformation of nitrogen's various forms
are presented. The effects of nitrogen within the aquatic
environment are then discussed, giving the designer a
perspective on how a treated wastewater discharge af-
fects the local environment. From this, a brief discussion
is presented of the regulatory structure that is in place to
evaluate and set discharge limits with respect to nitrogen
(and other constituents) and to ensure the quality and
beneficial use of the receiving water.
Chapter 2, Total System Design with Nitrogen Control,
presents the groundwork necessary to initiate the design
process. An overview of technology concepts, process
selection considerations, and special issues that require
a multidimensional approach are discussed. Physio-
chemical and natural systems approaches are treated
from the perspective of appropriate technology selection.
Chapter 2 also introduces the design examples, which
provide a common basis for illustrative designs developed
in Chapters 6 through 8.
Chapter 2 is an important element of the manual as a
whole and should be read by all who intend to use the
manual in support of a facility design or review. The chap-
ter is designed to give an overview to the considerations
and calculations that are addressed in the early stages
of facility design and offers mass-balancing methods
and common design information from which checks can
be made of an ongoing design process. The chapter
addresses key considerations when defining site require-
ments and the bases for selecting alternative technolo-
gies. Careful attention to such issues during the early
stages of a project will minimize the need for redirection
or redesign and help to avoid the common pitfalls encoun-
tered in the design of a facility.
Chapter 3, Process Chemistry and Kinetics of Biological
Nitrification, offers the designer the fundamentals of the
microbiology of nitrifying organisms, their means of me-
tabolism, the biochemical pathways these organisms use
to nitrify, and the stoichiometry useful in materials balanc-
ing. The relevant kinetic rate expressions are given, fo-
cusing on the intrinsic and controlling environmental
factors associated with the process. Chapter 3 serves as
the basis for design concepts and criteria presented in
Chapters 6 and portions of 8.
Chapter 4, Process Chemistry and Kinetics of Biological
Denitrification, presents the microbiology, the metabolism,
and the stoichiometry of denitrifying organisms in a format
similar to that used in Chapter 3. Rate expressions are
also similar, but the specifics of the controlling environ-
mental factors, such as pH, dissolved oxygen (DO), tem-
perature, and inhibitors, differ substantially. Effective use
and understanding of Chapters 7 and 8 follow from the
basic factors given in Chapter 4.
Chapter 5, Mathematical Modeling of Nitrification and De-
nitrification, establishes the conceptual and mathematical
frameworks that integrate the theoretical concepts of
Chapters 3 and 4 with quantified design experience. The
objective is to make the designer aware of how models
can serve as effective tools for formulating a design, while
at the same time warning of the hazards of using such
models without a detailed understanding of the processes
and an adequate data base. This chapter outlines the
phases of model development and the various uses of
models. A number of different types of models are pre-
sented and one, the IAWPRC model, is analyzed as an
example. This chapter need not be read to understand
the design material in Chapters 6, 7, and 8.
Chapter 6, Design Considerations for Biological Nitrifica-
tion Processes, is a sequel to Chapter 3. Topics relevant
to the development of design criteria for nitrification sys-
tems are divided between those applicable to suspended
growth systems and those applicable to attached growth
systems. The suspended growth section emphasizes the
application of kinetic theory and highlights both solids
residence time and specific rate approaches. Designs for
various completely mixed and plug flow configurations are
analyzed. The attached growth design criteria are devel-
oped more from empirical approaches and less from ki-
netic theory, drawing from an accumulation of design
experience. Trickling filters, rotating biological contactors,
and other configurations of attached growth systems are
presented. The objective of this chapter is to direct the
designer in the specifics of selecting, linking, and sizing
the various unit operations for nitrification.
Chapter 7, Design Considerations for Denitrification Proc-
esses with Supplemental Substrate Addition. Today,
separate-stage denitrification processes with a supple-
mental substrate are generally applied using suspended
growth, packed bed, or fluidized bed systems. Design
considerations and case studies are presented for each
of these three system configurations. The requirements
surrounding the handling, storage, feed, and control of
methanol are also presented, since methanol is.essen-
-------
tially the only supplemental substrate used in current
separate-stage denitrification systems. The treatment for
removal of excess methanol, if needed, by post-aerobic
treatment is also discussed in Chapter 7.
Chapter 8, Design Considerations for Single-Sludge Ni-
trification-Denitrification Processes, classifies the various
single-sludge nitrogen removal technologies into five
categories according to their number of stages, phases,
and anoxic zones. The more common proprietary and
nonproprietary single-sludge process configurations for
nitrogen removal are reviewed by comparing their design
criteria and expected process performance against their
unique design features. Process design scenarios are
described and a sample configuration for each scenario
Is selected from among the classifications. These repre-
sentative examples demonstrate controlling design fac-
tors, such as recycle and aeration.
1.4 Nitrogen in the Environment
The total mass of Earth's nitrogen circulates within the
biosphere among four main banks: the atmosphere, the
hydrosphere, the Earth's crust, and the tissues of living
and dead organisms. Each bank contains inventories of
nitrogen in its various forms. Although the total is un-
changing, the inventories of the various forms are in con-
stant flux. In a natural world, relative to one influenced
by the activities of people, this flux could be considered
approximately steady-state. However, there are two ac-
tivities that cause significant non-steady-state shifts in the
baseline nitrogen balance: 1) the mining and use of ni-
trogen-containing minerals and fossil fuels that have long
been out of active nitrogen circulation and 2) a positive
net fixation of nitrogen gas attributable to chemical manu-
facturing and the cultivation of leguminous, nitrogen-fixing
plants. Unfortunately, the hydrosphere has become a sink
for much of the excess nitrogen that has been mobilized
as a result of human commerce. Consequently, some
significant undesirable effects result from the accumula-
tion of nitrogen within niches of the aquatic environment.
The quality of water may be affected and the potential
beneficial uses of natural water systems may be dimin-
ished. Nitrogen, in its various forms, can deplete DO
levels In receiving waters, stimulate aquatic growth, ex-
hibit toxlcity toward aquatic life, present a public health
hazard, and affect the suitability of wastewater for reuse.
This manual presents the design of processes and tech-
nological configurations that will control and/or remove
nitrogen from wastewaters discharged to natural waters.
It Is appropriate, therefore, to first review the nature of
nitrogen and its presence in the environment. The remain-
der of this chapter 1) reviews how the various nitrogen
forms are physically transported and chemically changed
in what is known as the nitrogen cycle (Section 1.4.1);
2) surveys the major sources of both natural and human-
caused nitrogen loading (Section 1.4.2); and 3) discusses
some, of the effects that nitrogen compounds have in
altering the environment (Section 1.5). Finally, a closing
section (Section 1.6) presents a brief discussion of the
regulatory structure by which nitrogen control require-
ments and performance levels are implemented for
wastewater treatment plants in the United States.
1.4.1 The Nitrogen Cycle
Nitrogen exists in many forms in the environment. The
movement and transformation of these nitrogen com-
pounds through the biosphere is characterized by the
nitrogen cycle, a general depiction of which is shown in
Figure 1-1. The atmosphere serves as a reservoir of ni-
trogen in the form of nitrogen gas. Although virtually in-
exhaustible (the atmosphere is 79 percent nitrogen), the
nitrogen must be combined with hydrogen or oxygen be-
fore it can be assimilated by higher plants; the plants, in
turn, are consumed by animals. Human intervention
through industrial nitrogen fixation processes and the
large-scale cultivation of nitrogen-fixing legumes has
played a significant role in altering the historical nitrogen
cycle. The amount of nitrogen fixed annually by these two
mechanisms now exceeds by as much as 10 percent the
amount of nitrogen fixed by terrestrial ecosystems before
the advent of agriculture (4).
Nitrogen can form a variety of compounds because of the
different oxidation states it can assume. In the environ-
ment, most changes from one oxidation state to another
are brought about biologically. Consider the nitrogen
forms that are of interest in the soil/water environment:
Nitrogen
Compound
Ammonia
Ammonium ion
Nitrogen gas
Nitrite ion
Nitrate ion
Formula
NH3
NHJ
N2
NOi
NO;
Oxidation
State
-3
-3 ,
0
+3 :
+5
The un-ionized, molecular ammonia exists in equilibrium
with the ammonium ion, the distribution of which is de-
pendent upon system pH and temperature; in fact, very
little ammonia exists at pH levels less than neutral. This
is an important relationship and is discussed in greater
detail in Section 1.5.3 and in Chapters 3 and 6.
Transformation of these nitrogen compounds can occur
through several mechanisms. Those of importance in-
clude fixation, ammonification, synthesis, nitrification, and
denitrification. Each can be carried out by particular mi-
croorganisms with either a net gain or loss of energy;
energy considerations often play an important role in de-
termining which reactions occur.
-------
Atmo
Denitrification
t.
Nitrous
Oxide
Plant and Animal Wastes
Dead Organisms
Figure 1-1. The nitrogen cycle.
Fixation of nitrogen means the incorporation of inert,
gaseous nitrogen into a chemical compound such that it
can be used by plants and animals. Fixation of nitrogen
from N2 gas to organic nitrogen is predominantly accom-
plished biologically by specialized microorganisms and
the associations between such microorganisms and
plants (5). Atmospheric fixation by lightning and industrial
fixation processes (fertilizer and other chemicals) plays a
smaller, but significant, role as a fixation method.
Fixation Process
-» biological
N2 gas -» lightning
-» industrial
Product
organic nitrogen compounds
nitrate
ammonium, nitrate
Ammonification is the change from organic nitrogen to
the ammonium form. An important hydrolysis reaction in-
volves urea, a nitrogen compound found in urine:
Synthesis, or assimilation, is a biochemical mechanism
that uses ammonium or nitrate compounds to form plant
protein and other nitrogen-containing compounds:
+ CO2 + green plants + sunlight -» protein
l + CO2 + green plants + sunlight -> protein
Animals require protein from plants and other animals.
With certain exceptions, they are not capable of trans-
forming inorganic nitrogen into an organic nitrogen form.
Nitrification is the biological oxidation of ammonium. This
is done in two steps, first to the nitrite form, then to the
nitrate form. Two specific chemoautotrophic bacterial gen-
era are involved, using inorganic carbon as their source
of cellular carbon:
Nitrosomonas
H2NCONH2 + 2H2O
urea
(NH4)2C03
ammonium carbonate
bacteria
>
Nitrobacter
bacteria
ammonium
nitrite
->NOi
nitrate
In general, ammonification occurs during decomposition
of animal and plant tissue and animal fecal matter:
organic nitrogen + microorganisms -> NHs/NHj
(protein, amino acids, etc.)
The transformation reactions are generally coupled and
proceed rapidly to the nitrate form; nitrite levels at a given
time are relatively low. The nitrate formed may be used
in synthesis to promote plant growth, or it may be
subsequently reduced by denitrification, as suggested by
Figure 1-1.
-------
Denitrification is the biological reduction of nitrate to ni-
trogen gas. It can proceed through several steps in the
biochemical pathway, with the ultimate production of ni-
trogen gas. A fairly broad range of heterotrophic bacteria
are involved in the process, requiring an organic carbon
source for energy:
NOI+organic -» NOa +organic -> N2 + carbon + water
carbon carbon dioxide
nitrate
nitrite
nitrogen
gas
It is important to note that if both oxygen and nitrate are
present the bacteria will typically preferentially use oxy-
gen in the oxidation of the organic matter because it
yields more energy. Thus, for denitrification to proceed,
anoxic conditions must usually exist, although this is not
strictly the case for all bacteria.
The ammonification, synthesis, nitrification, and denitrifi-
cation reactions are the primary mechanisms employed
in the treatment of wastewaters for nitrogen control and/or
removal. Detailed discussions of their microbiological,
chemical, and kinetic aspects may be found in Chapters
3 (nitrification) and 4 (denitrification).
The transport mechanisms primarily responsible for the
movement of nitrogen through the environment are pre-
cipitation, dustfall, sedimentation in water systems, wind,
ground-water movement, stream flow, overland runoff,
and volatilization. Although these are not mechanisms by
which transformations take place, they can cause a
change in the environment whereby conditions will force
change and transformations will occur. Environmental
conditions that affect the behavior of reactions include
temperature, pH, microbiology, oxidation/reduction poten-
tial, and the availability of substrate, nutrients, and oxy-
gen. Although transport and transformation mechanisms
are described as individual processes, it is useful to un-
derstand that these comprise a dynamic continuum and
there may be no distinct boundary governing the trans-
formation of specific forms of nitrogen.
Since this manual's purpose is to address wastewater
treatment for nitrogen control, it is of interest to differen-
tiate between the surface water and sediment environ-
ment and the soil-ground-water environment of the
nitrogen cycle, which are directly affected by wastewater
treatment practice. This aids in understanding the roles
that nitrogenous compounds play in each and the prob-
lems that can be encountered.
1.4.1.1 The Nitrogen Cycle in Surface Waters and
Sediments
A representation of the nitrogen cycle that is applicable
to the surface water environment is presented in Figure
1-2 (6). As shown, nitrogen can be added by atmospheric
deposition through precipitation and dustfall; surface run-
off; subsurface ground-water entry; and direct discharge
of wastewater effluent. In addition, nitrogen gas from the
atmosphere can be fixed by certain photosynthetic blue-
green algae and some bacterial species.
Ammonification, nitrification, synthesis, and denitrification
can occur within the aquatic environment. Ammonification
of organic matter is carried out by microorganisms. The
ammonium thus formed, along with nitrate, can be
assimilated by algae and aquatic plants for synthesis.
If excessive, such growths may create water quality
problems. •
Biological nitrification of ammonium can occur; if signifi-
cant, it can result in depletion of the DO content of the
water (the oxidation of 1.0 mg/L of ammonium-nitrogen
will consume 4.6 mg/L of oxygen). Denitrification pro-
duces nitrogen gas, which may escape to the atmos-
phere. Because anoxic conditions are required for
denitrification, the oxygen-deficient hypolimnion (or lower
layer) of lakes and the sediment zone of streams and
lakes are important zones of denitrification activity (7).
1.4.1.2 The Nitrogen Cycle in Soil and Ground Water
Figure 1-3 shows the major aspects of the nitrogen cycle
associated with the soil-ground-water environment (8).
Nitrogen can enter the soil from the application of waste-
water or sewage treatment plant effluent, artificial fertiliz-
ers, plant and animal matter, precipitation, and dustfall.
In addition, nitrogen-fixing bacteria in the soil convert ni-
trogen gas into forms available to plant life. Humans have
increased the amount of nitrogen fixed biologically by
cultivation of leguminous crops (e.g., peas and beans).
Usually more than 90 percent of the nitrogen present in
soil is organic, either in living plants and animals or in
humus originating from decomposition of plant and animal
residues. Nitrate content is generally low because it is
taken up for synthesis, it is leached by water percolating
through the soil, and because of denitrification activity
below the aerobic top layer of soil. Synthesis and denitri-
fication rarely remove all nitrates added to the soil from
fertilizers and nitrified wastewater effluents: As such, ele-
vated ground-water nitrates leached from soil are a major
ground-water quality problem in many areas.
1.4.2 Sources of Nitrogen
An understanding of the various sources of nitrogenous
materials offers a perspective on the relative contribution
that municipal wastewater treatment plants make to the
cumulative effects of nitrogen in the environment. This, in
turn, can influence decisions regarding the level and type
of treatment that is most suited for a particular location.
In analyzing a nitrogen pollution problem, care must be
taken to ensure that all possible sources of pollution are
investigated and that the amount to be expected from
each is accurately estimated. Once estimates are made
and used in concert with water quality studies, nitrogen
-------
Wastewater
Effluent
Precipitation and
Dustfall
Runoff
Water
Column
Organic N
i
r
'
Sedimentation
Sediment • j j
Layer
tionj
Figure 1-2. The nitrogen cycle in surface water.
-------
Precipitation and Wastewater and Mineral
Dustfall Wastewater Effluent Fertilizers
Plant and Animal
Residue Compost Atmosphere
Ground-Water Level
Figure 1-3. The nitrogen cycle in soil and ground water.
-------
control measures can be oriented to the more significant
sources, and specifically to the type of impacts unique to
the individual sources. As an example, an estuary with
excessive aquatic growth may receive nitrogen-contain-
ing effluent from a sewage treatment plant, urban runoff,
and runoff from animal feedlots. The nitrogen from the
treatment plant is shown to be a small fraction of the
contribution from the other two sources, yet further study
of the geometry and flushing action of the estuary shows
significant local toxicity impacts from the plant's dis-
charge. Since the treatment plant plays no significant role
in the total nitrogen load to the estuary and its consequent
biostimulatory effects, it should not be required to remove
nitrogen from its discharge. However, nitrification of the
plant discharge may be needed to eliminate localized
ammonia toxicity effects.
Nitrogenous materials can enter the aquatic environment
from either natural or human-caused sources. The proper
delineation of these sources can often be clouded, since
the apparent quantities from natural sources can include
nitrogen generated from human activity. For example,
while nitrogen fixed by lightning may be expected in rain-
fall, the combustion of fossil fuels or the application of
liquid ammonia agricultural fertilizers, with subsequent
release to the air through volatilization, can increase rain-
fall concentrations of nitrogen substantially. The perva-
siveness of human impact on the environment limits the
certainty with which naturally occurring nitrogen and pol-
lution source nitrogen can be differentiated quantitatively.
Measurements of naturally occurring baseline nitrogen
levels are best made in more remote, undeveloped and
pristine areas; yet interpretation of these data should still
be considered uncertain to some degree.
The several direct sources from which the bulk of nitrogen
enters the aquatic environment are listed in Table 1-1,
along with the principal transport mechanisms responsi-
ble for delivering that nitrogen to natural receiving water
systems. Nitrogen originates from a source in soluble
and/or particulate form; it may also change form en route
to the receiving waters. Soluble nitrogen is primarily in
the inorganic form of ammonia, ammonium, nitrite, and
nitrate. Lesser amounts of reduced soluble nitrogen are
in the form of low molecular weight organic compounds
such as urea a'nd proteins. Particulate nitrogen can also
be either organic or inorganic in nature. Particulate or-
ganic nitrogen can consist of insoluble synthetic com-
pounds, flocculated one-celled microbes, and particles of
detritus from animal and plant tissues and animal wastes.
Inorganic particulate nitrogen typically consists of inor-
ganic species adsorbed or ionically exchanged onto sedi-
ment that becomes suspended in water.
1.4.2.1 Natural Sources
Natural sources or transport mechanisms of nitrogen sub-
stances include atmospheric precipitation, dustfall, nonur-
ban and nonagricultural runoff, and biological fixation.
Table 1-1. Major Direct Sources of Nitrogen
Source
Principal Transport to Natural
Water Systems
Untreated sewage
POTW effluent
POTW waste solids
Industrial sources
*•
Volcanoes and other
earth emissions
Fertilized farms
Animal wastes
Decaying plants and
animal tissue
Septic tanks and
leaching fields
Ship/other vessels
Urban surfaces
Fossil fuels
Nitrogen-fixing
organisms
Disturbed earth
Landfill leachate
Direct discharge
Direct discharge, land application
Direct discharge,.land application
Direct discharge, ground-water
movement, precipitation
Precipitation, wind, and
gravitational settling
Surface runoff, ground-water
movement
Volatilization/precipitation, surface
runoff, ground-water movement
Surface runoff, ground-water
movement
Ground-water movement
Direct discharge ,
Direct discharge, surface runoff
Precipitation, wind and gravitational
settling
In situ
Wind and gravitational settling
Ground-water movement
Nitrogen measured in precipitation is most often a result
of both soluble and particulate nitrogen forms scrubbed
from the atmosphere. Natural components would include
nitrogen oxides fixed by lightning and emitted from vol-
canic eruptions, wind-blown dust originating from natural
areas, and ammonia released from decaying animal and
plant matter. As examples, total nitrogen in rainfall in
Sweden was cited as 0.2 ppm, while the average con-
centration of nitrogen in western U.S. snows, mainly in
the Sierra Nevada^Mounfains, was 0.15 ppm ammonium-
nitrogen, 0.01 ppm nitrite-nitrogen, and 0.02 ppm nitrate-
nitrogen (5). Again, the degree to which such values are
representative of "natural" conditions cannot be deter-
mined with any certainty.
The quantities of nitrogen in nonurban runoff from non-
fertilized land may be expected to vary greatly, depending
on the erosive characteristics of the soil and the migration
of dustfall and precipitation originating from proximate
urban and/or industrial areas. As an example, runoff from
forested land in Washington contained 0.13 mg/L nitrate-
nitrogen and 0.20 mg/L total nitrogen (7). A survey of
surface runoff from 90 percent forested land in the east-
-------
ern, central, and western United States yielded mean
total nitrogen concentrations of 0.19, 0.06, and 0.07
mg/L, respectively. Concentrations for 50 percent for-
ested lands of the same regions are approximately dou-
ble these values at 0.34, 0.25, and 0.18 mg/L,
respectively (9).
Biological fixation may add nitrogen to surface water di-
rectly or to ground water via leaching from soil. Of par-
ticular interest is the role of fixation in the eutrophication
of lakes. Certain photosynthetic blue-green algae, such
as the species of Nostic, Anabaena, Gleotrichia, and Ca-
lathrix, are common nitrogen fixers. As an example, as
much as 14 percent of the total nitrogen entering eutro-
phlc Lake Mendota in Wisconsin was added by fixation
(6). Finally, lightning contributes notably to the mass
quantity of fixated nitrogen. One estimate reports that
approximately 15 percent of all naturally fixed nitrogen
results from lightning (4).
1.4.2.2 Human-Caused Sources
Sources of nitrogen related to human activity include un-
treated and treated domestic sewage and industrial
wastes, leachates, atmospheric deposition, and surface
runoff. These general areas are presented in the following
discussions.
Domestic Sewage. Untreated sewage flowing from mu-
nicipal collection systems typically contains 20-85 mg/L
total nitrogen. Flow from residences, a major component
of domestic wastewater, has been observed to contain
30-80 mg/L TKN (total Kjeldahl nitrogen). The total nitro-
gen in domestic sewage comprises approximately 60 per-
cent ammonium nitrogen, 40 percent organic nitrogen,
and very small quantities of nitrates. Unit loading from
residences has been estimated at an average of 0.012
kg N/capita/d (0.027 Ib/capita/d). Septage from rural sep-
tic system tanks is often collected and mixed with un-
treated sewage at the head of treatment works. Septage
nitrogen content is 100-1,600 mg/L TKN with 700 mg/L
TKN being a typical value (9). The volume of septage is
generally low relative to wastewater plant flow, although
In smaller systems it can be a significant input. The ni-
trogen content of combined sewer overflows (CSO) is
often calculated from the weighted averaging of storm
water and domestic sewage concentrations. Nitrogen
concentrations of CSO vary radically over a rainfall event,
particularly as an effect of the first wash of drainage area
surfaces and the first flush of settled solids in the drainage
collection grid.
Treated domestic sewage will exhibit a variety of nitrogen
levels, depending on the type of treatment applied. A
typical activated sludge facility reduces total nitrogen con-
tent by virtue of cell synthesis and solids removal. Most
ammonia, however, passes through unless there is a spe-
cific treatment provision for nitrification. Reductions in to-
tal nitrogen beyond 20-30 percent would require
denitrification. Conventional activated sludge treatment
generally yields effluent nitrogen levels of 15-35 mg/L
total nitrogen. Advanced biological nitrification-denitrifica-
tion treatment can generally yield an effluent quality of
2-10 mg/L total nitrogen (9). Sludges generated from the
treatment of wastewaters will also contain nitrogen and
will serve as a nitrogen source, depending on the method
of disposal. Section 2.4.1, provides a discussion of ex-
pected effluent nitrogen levels for various technology ap-
plications to municipal wastewaters.
!
Industrial Wastewaters. Industry adds nitrogen into aque-
ous waste streams as a result of the use of water in
processing and the secondary use of water to scrub gase-
ous waste streams. Some industries that yield more ele-
vated levels of nitrogen in their primary or process water
include chemical (fertilizer and nitrogenous compound
manufacturing); paper and pulp (natural products, ammo-
nia-based pulping); mining and metals (ore processing,
nitric acid pickling); and food processing (protein-enriched
wash and cooking waters). As an example, estimates
from Sweden suggest that these primary industrial cate-
gories account for 95 percent of the country's industrial
aqueous nitrogen discharge. Textiles and leather manu-
facturing account for the remaining 5 percent (10). Indus-
trial and power generation facilities use water for
scrubbing flue gases, which creates secondary nitrogen-
containing waste streams. Nitrogen oxides, created dur-
ing combustion of fossil fuels, hydrolyze to nitrate when
solubilized in water.
Landfill Leachates. Leachate from municipal solid waste
landfills (MSWLs) is characterized as a relatively low-vol-
ume, high-strength wastewater. A survey of leachate
characterization studies for many MSWLs shows ammo-
nium values of 0-1,160 mg/L and nitrate plus nitrite ni-
trogen of 0.2-10.3 mg/L. Treatment design has ,to be
flexible to allow for the typical high degree of variability
in leachate strength observed with landfill age (11).
Atmospheric Deposition. Atmospheric nitrogen generally
enters an aquatic environment in one of two forms: inor-
ganic nitrogen that is solubilized in rainwater or paniculate
organic and mineralized nitrogen that either settles by
gravity or is scrubbed by rain onto the receiving water
surface. Atmospheric deposition over the land area of a
watershed is often accounted for in surface runoff loads;
therefore, atmospheric contributions to a waterbody's ni-
trogen loading is usually attributed only to precipitation
that falls directly over the water surface. Most soluble,
inorganic nitrogen originates from volatilization of ammo-
nia-nitrogen and combustion of fossil fuels. Combustion
oxidizes nitrogen contained in oil and coal and (at higher
temperatures) free atmospheric nitrogen. Nitric oxide is
the principal product, which rapidly oxidizes to nitrogen
dioxide. Nitrous oxide can also be formed, chiefly as a
product of burning wood and other biomass. Nitrogen
dioxide is hydrolyzed to nitric acid (12). Ammonia can
10
-------
of animal wastes and plant and animal tissues, applica-
tion of liquid ammonia fertilizers to farmland, composting
and decay of sewage sludge, and industrial processes.
Paniculate atmospheric nitrogen results primarily from
suspension by wind. Soil disturbance resulting from min-
ing, agriculture, foresting, and construction will contribute
significantly to this process.
Most atmospheric deposition is reported in the literature
as either total continuous deposition (dust and precipita-
tion) or partial, intermittently collected deposition (precipi-
tation only). Concentrations and unit area! loading rates
of various regions are given as representative values in
Table 1-2. Variations in the parameters result from vari-
ations in both natural background and human activity
within the area of meteorological interest.
Surface Runoff. Surface runoff in the urban environment
can contain significant levels of nitrogen. Draining of wet-
lands for land development removes a natural sink for
nitrogen (which occurs through the settling of organic
solids and denitrification in its sediments). Increased irri-
, permeable surfaces ensure rapid conduction of nitrogen
to receiving waters, bypassing natural assimilation. Con-
struction and other disturbances create increased quan-
tities of suspended solids. (SS) in runoff. These solids
generally have a significant particulate organic nitrogen
component. Because soil loss from construction can
reach over 100 metric tons/ha/yr (45 tons/ac/yr) (19), the
portion of area undergoing construction within a water-
shed need not be great in order to contribute significant
nitrogen loading. Urban runoff also generally includes any
atmospheric deposition that occurs over its area.
• • " -.'»-"
A study of urban sources of nitrogen to Narragansett Bay,
Rhode Island, shows typical concentrations from light in-
dustrial runoff, roof runoff, and general stormwater runoff
to be 0.2-1.1, 0.5-4, and 3-10 mg/L total nitrogen, re-
spectively. Construction site runoff measured in this area
yielded 10,000-40,000 mg/L of SS (19). Typical construc-
tion site runoff during median flows has been estimated
at 1-30 mg/L total nitrogen (20). Other average urban
runoff nitrogen concentrations reported from earlier stud-
ies are 2.7 mg/L total nitrogen in Cincinnati, Ohio (14);
2.1 mg/L total nitrogen in Washington, DC (21); 2,5 mg/L
total nitrogen in Ann Arbor, Michigan (22); and 0.85 mg/L
organic nitrogen in Tulsa, Oklahoma (23).
Leakage from failing sanitary sewers, industrial tankage,
and septic systems, as.well as from illegal hookups and
Table 1-2. Atmospheric Nitrogen: Representative Concentrations and Unit Area! Loadings
Location Nitrogen Form Sampling , .Measurement
Reference
Representative Concentrations, mg N/L
Long Island Sound ammonia
Geneva, NY
Ottawa, ON
Cincinnati, OH
nitrate and nitrite
ammonia plus
nitrate-nitrogen
inorganic nitrogen
ammonia
nitrate
total nitrogen
inorganic nitrogen
Precipitation
Precipitation
Precipitation
Snow
Rain
Rain
Precipitation
Precipitation
0.13
0.32
1.1
0.85
1.8
0.35
1.27
0.69'
13
13
5
5
5
14
14
Coshocton, OH
total nitrogen
inorganic nitrogen
Representative Areal Loadings, kg/ha/yr*
Precipitation
Precipitation
1.17
0.80
* 1 kg/ha/yr = 0.8922 Ib/ac/yr
14
14
Potomac River
Lake Huron (northwest)
Sweden (average)
Sweden (south)
Central Europe
Hamilton, ON
Seattle, WA
total nitrogen
total nitrogen
total nitrogen
total nitrogen
total nitrogen
total nitrogen
total. nitrogen
nitrate
. Precipitation and dust
Precipitation and dust
Precipitation and dust
Precipitation and dust
•Precipitation and dust
Precipitafon and dust
Dust , .''•••
Dust
' 18.6 : •' •:
11.0 ' "
1*0.0. '
15.0-25.0 '
•'20,0-30.0'
6.5 '• : '
• 2.6
: o>i
15 , • - .
>6 '
17
• '• 17
17
5
"• :• 5
18
11
-------
discharges, can collectively account for a significant por-
tion of nitrogen loading in a stormwater collection system
(19). Default values used for areal nitrogen loading from
low-, medium-, and high-density residential neighbor-
hoods, as given by STORM (the Urban Runoff Model of
the U.S. Army Corps of Engineers), are 0.008,0.031, and
0.028 kg/ha/d (0.007, 0.028, and 0.025 Ib/ac/d), respec-
tively. Commercial and industrial area runoff loads are
given as 0.237 and 0.234 kg/ha/d (0.211 and 0.209
Ib/ac/d), respectively (9). Land uses described as low-to-
medium density residential, high-density residential and
commercial, and medium- and high-intensity industrial
yield average unit nitrogen (25) loadings to storm sewers
of 9, 11.2, and 7.8 kg/ha/hr (8, 10, and 7 Ib/ac/hr), re-
spectively.
Application of fertilizer to farmlands creates significant
nitrogen loadings to ground and surface waters. Control-
ling factors include application rate of fertilizer, type of
fertilizer, irrigation rate, soil drainage, type of plant cover
and its nitrogen uptake rate, and degree of tillage. Ob-
served values of nitrogen in runoff and ground water often
have a high degree of correlation with these factors
(26,27). A survey of the eastern United States charac-
terized stream flow as a function of percent of agricultural
land use within each watershed. Mean total nitrogen con-
centrations for streams with 50, 75, and 90 percent agri-
cultural watershed were 1.08, 1.82, and 5.04 mg/L total
nitrogen, respectively (24). A sampling of receiving
streams and ground waters from 268 agricultural sites in
southeast Nebraska ranged from less than 0.1 to 233
mcj/L total nitrogen. At 37 percent of these sites, nitrate-
nitrogen levels exceeded 10 mg/L and were often 20-40
mg/L, well above the maximum drinking water limit of 10
mg/L (26). One area in southeast Ireland has five farming
districts that have mean artificial nitrogen fertilizer appli-
cation rates of 47.6-68.2 kg N/ha/yr (42.5-60.9 Ib/ac/yr).
The mean nitrogen loss to ground water and runoff is
4.1-25.5 kg N/ha/yr (3.6-22.8 Ib/ac/yr). The resulting
maximum river nitrate concentrations correlate strongly
with the corresponding fertilizer application rates. Percent
loss values for each farming district also correlate strongly
with percent of land area ploughed (27).
Feedlot runoff constitutes a source of nitrogen that has
become significant as a result of the increased number
of concentrated, centralized feedlots. Ammonium result-
ing from urea hydrolysis is a major constituent of feedlot
waste. Ammonium-nitrogen runoff concentrations may
reach 300 mg/L (6,28,29) and organic nitrogen concen-
trations of up to 600 mg/L (28,29) have been reported.
The growing trend is toward feedlot operations as com-
pared to small-farm livestock production. The centralized,
more contained nature of feedlots lends itself favorably
to collection and treatment, allowing for significant im-
provements in this area of nitrogen control.
Septic fields are responsible for a significant fraction of
the nitrogen load to U.S. ground water. Approximately 25
percent of the population is served by individual home
sewage disposal systems (30). Effluent from a typical
septic system has a total nitrogen content of 25-60 mg/L.
Of this, 20-60 mg/L exists as ammonia and less than 1
mg/L exists as nitrate (9). Another study has charac-
terized a typical septic effluent as containing approxi-
mately 7 mg/L organic nitrogen, 25 mg/L ammonium
nitrogen, and 0.3 mg/L nitrate-nitrogen (31). A survey of
septic fields indicates that rapid nitrification of ammonium-
nitrogen takes place under aerobic conditions within the
leach field (9). Ammonium-nitrogen is easily exchanged
in many soils below a leach field, whereas nitrate remains
soluble and is easily lost to ground water. If exchange
sites become saturated, as in sandy soils, ammonium
breaks through to ground water before it nitrifies. When
septic fields dry out in the summer, or are abandoned,
much adsorbed ammonium is converted to nitrate, and
eventually lost to leaching (31). Natural tertiary POTW
treatment systems utilizing soil infiltration or overland flow
can typically produce a nitrogen loading to ground Water
in a manner similar to septic fields. A typical final effluent
of this type has a total nitrogen of 3-10 mg/L (9).
The fractional contribution from each of the categories
discussed above varies primarily according to the geo-
graphical location of the receiving water body, the type
and intensity of the development within the region influ-
encing the study site, the population density, and the type
of original natural habitat. A number of major studies ex-
amine in part the various sources of nitrogen loading to
water bodies in the United States. Results of studies of
the Long Island Sound, Chesapeake Bay, and coastal
Waters in Sweden are summarized in Table 1-3
(13,32,33).
1.5 Effects of Nitrogen in Discharges from
Wastewater Treatment Plants
Excessive accumulation of various forms of nitrogen in
surface and ground waters can lead to adverse ecological
and human health effects. This section gives an overview
of several effects attributable to nitrogen that can origi-
nate from municipal wastewater discharges. One of the
major effects has been the direct and indirect depletion
of DO in receiving waters. In-stream nitrification directly
consumes oxygen, while biostimulation of aquatic plant
growth lowers oxygen indirectly when the plant growth
dies and undergoes bacterial decomposition. Other im-
pacts can be of major importance in particular situations.
These include ammonia toxicity to aquatic animal life,
adverse public health effects, and a reduction in the suit- -
ability of water for reuse.
12
-------
Table 1-3. Representative Distribution of Sources of Nitrogen to Chesapeake Bay (31), Long Island Sound (13), and
Swedish Coastal Areas (32)
Chesapeake Bay
Long Island
Sound
Swedish Coastal Areas
Point sources
Animal wastes
Atmospheric ammonium
Atmospheric nitrate
Fertilizers
Total
23
4
14
25
34
100
STPs
Industry
Atmospheric
Coastal runoff
CSO
Tributaries
Total
43.6
1.8
11.8
6.2
1.2
35.3
100
Municipal sewage
Atmospheric deposits
on sea surface
Industry
Agriculture
Forest and forestry
Wetlands
Other land uses
Total
19,1
9.8
3.8
26.3
23.3
8.4
6.3
100
Note: Totals have been reached through rounding. CSO = combined sewer overflows; STRs = sewage treatment plants.
1.5.1 Biostimulation of Plant and Algal Growth in
Surface Waters
A major problem in the field of water pollution is eutrophi-
cation, which is defined as excessive plant growth and/or
algae "blooms" resulting from over-fertilization of rivers,
lakes, and estuaries. Eutrophication can result in a dete-
rioration in the appearance of previously clear waters,
odor problems from decomposing plant growth, and a
lower DO level, which can adversely affect the respiration
of fish, benthic aquatic animals, and attached bottom
plant growth.
Primary conditions that are required for plant or algal
growth are adequate macronutrients in the form of nitro-
gen and phosphorus, sufficient carbon dioxide, and light
energy; the absence of any one will limit growth. In spe-
cial cases, trace micronutrients such as cobalt, iron, mo-
lybdenum, and manganese may be limiting factors under
natural conditions. Carbon dioxide is very seldom a factor
in growth-limiting situations. Sunlight becomes growth
limiting in deeper waters because of light extinction or in
quiescent or stratified water where excessive algal growth
at the surface shields light from the lower levels. Since
carbon dioxide and sunlight are virtually impossible to
control, their manipulation is not considered a realistic
means of limiting excessive photosynthesis.
Nitrogen and phosphorous are typically the two key tar^
gets for the control of eutrophication problems. After de-
termining which nutrient, if either, is growth limiting, one
must determine if and how the amount of the limiting
substance entering the receiving water can be controlled.
Under some circumstances, removal of both nitrogen and
phosphorus may be undertaken to limit algal growth.
Eutrophication is of greatest concern in lakes because
nutrients that enter the water body tend to be recycled
within the lake and build up over a period of time (34). A
river, by contrast, is a flowing system in which nutrients
are always entering or leaving at any given section. Ac-
cumulations tend to occur only in sediment or in slack
water, and the effects of these accumulations are nor-
mally moderated by the periodic flushing action of floods.
In estuaries and oceans, nitrogen compounds are often
present in very low concentrations and may limit the total
biomass and the types of species present (34). Thus,
upwelling, which brings nutrient-rich waters to the sur-
face, may result in periodic blooms of algae or other
aquatic plant life. In some estuaries, discharges from
wastewater treatment plants may increase nitrogen con-
centrations to the level where blooms occur. However, the
high dilutions provided by a direct ocean discharge prob-
ably eliminates the danger of algae blooms caused by
such discharges.
Biostimulation caused by excessive nitrogen loading is
considered to be the single largest cause of hypoxia in
the Long Island Sound (35). In 1990 the base loading of
nitrogen from sewage treatment plants into the Sound
was estimated to be approximately 63,600 kg (140,000
Ib/d), which accounted for approximately 43.6 percent of
the total nitrogen load contributed from all sources. If
maximum biological nutrient removal were accomplished
in all sewage treatment plants, the resulting maximum
effluent nitrogen concentration of 4 mg/L (and effluent
BOD5 (5-day biochemical oxygen demand) of no more
than 10 mg/L) would result in a new sewage treatment
plant nitrogen loading of 17,770 kg (39,130 Ib/d). This
would increase the average minimum DO concentration
from 1.8 mg/L to approximately 3.3 mg/L in the critical
Western Narrows region of the Sound. A major nitrogen
removal initiative is being undertaken to upgrade a ma-
jority of the sewage treatment plants discharging to the
Sound.
While nitrogen in wastewater treatment plant effluents
can sometimes contribute to undesirable aquatic growths,
13
-------
determination of the limiting constituent should be made
before the decision is made to require nitrogen removal
in the municipal treatment works. It is important to under-
stand that eutrophication is typically a basin-wide issue,
particularly in confined water systems. The sources of
nitrogen can be widespread, including atmospheric depo-
sition and surface runoff in addition to specific point
sources. Thus, it is critical to understand a wastewater
treatment plant's contribution to the overall nitrogen load
and its significance before considering the imposition of
specific nitrogen controls.
1.5.2 Depletion of DO Caused by Nitrification
Just as bacterial decomposition of the carbonaceous or-
ganic components of wastewaters depletes DO in the
receiving waters, in-stream nitrification of ammonium-
nitrogen creates an additional oxygen demand. In specific
cases where it is shown that the wastewater ammonium-
nitrogen content is a significant factor in the depletion of
the receiving water's DO, it is best to accomplish nitrifi-
cation before discharge.
Table 1-4 gives a simplified example of the impact of
providing treatment for nitrification. If conventional biologi-
cal treatment is utilized to provide 90 percent BODS re-
moval, but no ammonium oxidation (secondary
treatment), the effluent will still contain over 100 mg/L of
total oxygen demand. This high demand may cause sig-
nificant oxygen depletion in the receiving water if there is
insufficient dilution and the environmental conditions are
conducive to nitrification (e.g., a sufficient nitrifying popu-
lation or time to grow one, pH, temperature, buffering).
Although having little effect on organic oxygen demand
removal efficiency, accomplishing nitrification before dis-
charge will greatly reduce the residual total oxygen de-
mand discharged to the receiving water.
An analysis of in-stream conditions before and after the
implementation of point source treatment upgrades pro-
vides examples of the impact of nitrification on in-stream
DO levels (37). Table 1-5 excerpts data regarding three
receiving waters that were selected because the treat-
ment plants accounted for the major, or only, point source
discharges to the river segment and the specific upgrade
encompassed wastewater nitrification.
Figure 1-4 presents DO data from a Patuxent River moni-
toring station (river mile 70.8) near the sag point down-
stream of the Laurel Parkway Treatment Plant. The
upgraded plant was brought on-line in 1974 with secon-
dary treatment and nitrification. These data, collected at
flows near the 7 consecutive day, 10-year low flow
(7Q10), indicate an average 1978 summer DO cpncen-
tration of about 7.6 mg/L in comparison to average 1966
and 1967 concentrations of about 3.7 mg/L (37).
1.5.3 Ammonia Toxicity
Ammonia-nitrogen is the nitrogenous compound most re-
sponsible for toxicity effects in fish and other aquatic life.
Ammonia dissolved in water exists as an equilibrium of
molecular ammonia (NH3) and ionized ammonium (NHJ).
Toxicity of aqueous ammonia to aquatic organisms is
primarily attributable to the un-ionized, molecular free am-
monia form, with NHJ species being relatively less toxic.
The most important conditions that control this equilib-
rium, and hence the percentage of available molecular
free ammonia, are pH and temperature. Figure 1-5 pre-
sents the fractions of total ammonia available in the un-
ionized form as a function of pH and temperature (38).
These are presented within the ranges of pH and tem-
perature normally encountered in continental receiving
waters.
Ionic strength of a solution also has a noticeable, but less
significant, effect on the percent of un-ionized NH3. The
fraction of ammonia in the un-ionized form decreases with
increasing ionic strength in hard water and saline water.
In most natural freshwater systems the reduction of per-
cent un-ionized ammonia attributable to dissolved solids
is negligible. In saline or very hard waters there will be a
Table 1-4. Effect of Ammonium Oxidation on Total Oxygen Demand of Treated Wastewater Discharge (after 36)
Parameter Raw Wastewater Secondary Treatment With Nitrification
Organic matter, mg BODg/L
Organic oxygen demand, mg BOD/L
Organic and ammonia nitrogen, mg TKN/L
Nitrogenous oxygen demand, mg NOD/L
Total oxygen demand, mg TOD/L
Percent TOD due to nitrogen
Percent organic BOD removed
Percent TOD removed
250
375a
25
115b
490
23.5
—
—
25
37a
20
92b
129
71.3
90
73.7
20
30a
1.5
7"
37
18.9
92
92.5
"Takon as 1.5 times organic matter.
"Taken as 4.6 times the TKN level.
14
-------
Table 1-5. Examples of DO Improvement to River Segments Because of Treatment Plant Upgrades to Nitrification (after 37)
Plant:
Receiving Water:
POTW effluent
Total of all sources
Stream
a Before = study period
b Before = study period
c Before = study period
Laurel Parkway3
Patuxent River
Before
CBODg, (kg/d)/(mg/L) 159/9.5
NH3-N, (kg/d)/(mg/L) 128/7,7
CBOD5, kg/d 176
NHg-N, kg/d 135
Average DO, mg/L 5.5
Minimum DO, mg/L 3.8
Maximum CBODS, 18.0
mg/L
Maximum NH3-N, 2.2
mg/L
before 1968; After = study period after 1978
before 1972; After = study period after 1981
before 1968; After = study period after 1979
After
17.3/0.45
5/0.14
35.5
32.3
7.9
7.6
<1.0
0.1
Hurricaneb
Hurricane Creek
Before
121/50
38/15.5
121
37.7
5.6
2.6
5.6
5.4
After
11.4/2.1
3.4/0.64
11.4
3.4
6.3
4.5
5.1
1.3
Springfieldc
Wilsons
Before
738/9.8
718/9.5
742
718
6.4
1.4
26.0
22.2 •
Creek
After
337/1.6
140/0.68
337
140
8.2
6.5
5.0
<1.0
14 _
Jan. Feb. Mar.
May June July Aug. Sept.
June, July, August & September
Oct.
Nov. Dec.
Mean, mg/L
1966-67 1978
% Change
DO
3.7
7.6
108%
Figure 1-4. Time history data analysis for main stream of Patuxent River (station: PXT0708 State of Maryland)
(from Reference 37).
15
-------
u.
•o
S
c
c
0.1
0.01
=i 0.001
.a
0.0001
= I r
i i i r
PH
I
I
I
J_
0.0 5.0 10.0 15.0 20.0 25.0
Temperature, °C
30.0
35.0
40.0
Figure 1-5. Effect of temperature and pH on un-ionized ammonia (from Reference 38).
small but noticeable reduction in un-ionized ammonia
fraction (38).
EPA has assembled extensive research to support its
ambient water quality criteria for ammonia in freshwaters
(38) and saltwaters (39). For both fresh and salt water
the majority of research studies show acute toxicity ef-
fects for salmonid and nonsalmonid fish species between
0.1 and 10 mg/L un-ionized ammonia. The EPA criteria
for ambient water quality, as well as modified-state crite-
ria, give both maximum total and un-ionized ammonia
levels as a function of pH and temperature. The maximal
one-hour average in-stream concentrations of un-ionized
ammonia permissible in a three-year period are all under
1 mg/L. The maximal four-day average concentrations for
the same are all under 0.1 mg/L. Figure 1-6 is a graphical
display of EPA chronic toxicity criteria for ammonia, with
salmonids absent. Supporting research is reported also
for aquatic invertebrates and plants (38).
Free un-ionized ammonia toxicity effects as a function of
temperature and pH are observed in eutrophic lakes.
Lake Onondaga, in New York, is a representative exam-
ple (40). With warm-weather stratification of lake waters,
the DO content of the bottom waters, or hypolimnion,
decreases with bacterial respiration and the lack of circu-
lation with aerated surface waters. Total ammonia, from
the decomposition of decaying organic sediment, accu-
mulates as nitrification decreases under growing anoxic
conditions. A warm-weather vertical profile of a typical
lake shows high total ammonia/low nitrate concentrations
in the hypolimnion and low total ammonia/high nitrate
concentrations near the surface.
Compared with other systems, Lake Onondaga main-
tained relatively high total ammonia concentrations in the
epilimnion, or surface waters, throughout the spring-to-fall
study period in 1988. The lake experienced large fluctua-
tions in algal biomass, reflecting strong variations in the
net algal growth rate. Periods of oxygen supersatura-
tion and elevated solution pH occurred as a result of
photosynthetic oxygen production and carbon dioxide
consumption. Because of these elevated pH (and tem-
perature) levels at the surface, the distributions of free
un-ionized ammonia contrasted strongly with those for
total ammonia. Maximum free un-ionized ammonia con-
centrations were noted in the surface waters, despite the
fact that the highest total ammonia concentrations oc-
curred in the bottom waters. A vertical profile of the ob-
served free un-ionized ammonia with depth is shown on
Figure 1-7, excerpted from the Lake Onondaga study. The
figure also presents the computed final chronic values
(FCV) and one half the final acute values (0.5 FAV) from
the EPA Water Quality Criteria as a function of pH and
temperature (note that 0.5 FAV is shown because the
criterion states that average one-hour un-ionized ammo-
nia concentrations must be less than one-half the FAV).
The elevated free un-ionized ammonia levels resulted in
continuous contravention of the chronic ammonia toxicity
criteria for nonsalmonid fish and less frequent contraven-
tion of the acute toxicity criterion for nonsalmonids.
16
-------
0.1
z
3
0.01
•a 0.001
8
0.0001
Temp., °C E
Z'
<0
.sf
I
i
§
0.1
0.01
0.001
6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
PH
0.0001
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.C
Temperature, °C
-I 1
of
i
E 30
~ Temp
I
1 i
0.05 rill' I ' ' ' ' ' ' - '• i I i ~1 0.05
6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 0.0 5.0 10.0 15-0 20.0 25.0 30.0 35.0 40.0
pH Temperature, °C
i
i
Figure 1-6. EPA chronic criteria for ammonia (salmonids absent) (from Reference 38).
1.5.4 Public Health
Nitrate and nitrite nitrogen constitute a public health con-
cern, related primarily to methemoglobinemia (infantile
cyanosis) and carcinogenesis. Methemoglobinemia is a
disease primarily affecting infants and is often described
by the lay term blue babies. The acute toxicity of nitrate
occurs as a result of its reduction to nitrite, a process that
can occur under specific conditions in the stomach and
saliva. The nitrite ion formed oxidizes iron in the hemo-
globin molecules from the ferrous to the ferric state. The
resulting methemoglobin is incapable of exchanging oxy-
gen, and anoxia or death may occur if the condition is
left untreated. Suffocation is often accompanied by a blu-
ish tinge to the skin. Most reported water-related cases
of infantile methemoglobinemia have been associated
with the use of water containing more than 10 mg/L ni-
trate-N. Thus, drinking water guidelines and standards
recommend that the concentration of nitrate in drinking
water should not exceed 10 mg/L (28,41,42). This stand-
ard is exceeded most often in shallow wells in nonsew-
ered rural areas where septic systems are the dominant
method for sewage disposal.
Carcinogenesis (gastric cancer, in particular) has been
associated with the ingestion of N-nitroso compounds.
Nitrites (indirectly from nitrates) can react with amines
and amides to form nitrosamines and nitrosamides. The
epidemiological evidence suggests that high nitrate in-
gestion may be a contributing factor in gastric cancer.
There appears to be little information available to draw
any conclusions about any other human cancer and high
levels of nitrates in drinking water (41-43).
1.5.5 Reuse of Wastewater
Wastewater reclamation has traditionally been practiced
for pollution abatement, although the benefit derived from
supplementing available water resources has become
increasingly important. The uses of reclaimed water in-
clude indirect potable reuse, agricultural reuse, urban
landscape irrigation, industrial reuse, ground-water re-
charge, and potable reuse. The primary obstacle to rec-
lamation is the removal of pathogens. Nitrogen removal
becomes more of a concern when reclaimed water is
ultimately intended for supplementing municipal drinking
water supplies. Although biological processes are gener-
ally recommended for nitrogen removal from waste-
waters, multistage tertiary treatment often reduces solids
and carbonaceous substrate to an extent where physio-
chemical processing is preferred for nitrogen polishing.
The 44-L/s (1 mgd) Demonstration Plant operated by the
Metropolitan Denver Sewage Disposal District No. 1 re-
17
-------
Free Ammonia, mg N/L
0.0 0.1 0.2 0.3 0.4
0.5
Non-Salmonid, FCV
Observed
_ Non-Salmonid,
0.5 x FAV
18
20
Figure .1*7. Depth profiles of NH3> FCV, and 0.5 x FAV in Onondaga Lake for June 19,1988 (adapted from Reference 40).
moves nitrogen in the form of ammonium by passing the
water through a naturally occurring zeolite media (clinop-
tilolite). Sodium chloride is used to regenerate the zeolite
media. The ammonia is recovered from the regenerant
solution in the form of ammonium sulfate and sold as a
commercial-grade fertilizer (44).
Potable reuse is considered indirect when reclaimed ef-
fluent is discharged to a surface water supply. The Upper
Occoquan Sewage Authority (UOSA) removes ammo-
nium by ion exchange. The ion exchange regenerant
stream undergoes regeneration in a closed-loop ammo-
nia stripping and adsorption process that yields a 40 per-
cent ammonium sulfate solution for resale as an
agricultural fertilizer (44). Ammonia removal for the pur-
pose of ground-water recharge is accomplished at the
Orange County Sanitation District reclamation plant
known as Water Factory 21. Tertiary reclamation proc-
esses reduce the secondary effluent from 45 mg NHJ-N/L
to 4.7 mg/L After blending with lower concentration
sources, the final product contains an average of 0.9
mg NHJ-N/L to satisfy an injection requirement of 1.0
mg/L (45).
Nitrogen limits on water for reuse in agricultural and urban
landscape irrigation are not a factor, providing that there
are no secondary concerns for augmenting the nitrogen
load to ground water. On the contrary, there may be a
benefit in having residual ammonium and nitrate in water
used for irrigation. Nitrogen requirements for industrial
reuse vary widely according to the particular process for
which it is used. A considerable amount of water in this
category is used for cooling purposes, in which case the
concern is primarily total dissolved solids, of which ni-
trates are typically a small component.
18
-------
1.6 Establishing Nitrogen Limits for
Wastewater Discharges
The Federal Water Pollution Control Act of 1965 began
a series of environmental legislative reforms that initiated
a consistent approach to pollution control based on water
quality and beneficial use goals. National technology-
based standards were established, moving all wastewater
treatment facilities to secondary levels, at minimum. Re-
quirements for advanced treatment, such as ammonia
removal or total nitrogen removal, reflected specific water
quality needs for the given receiving water. National water
quality goals have influenced the development of ad-
vanced treatment technologies, especially in the area of
nutrient control.
Various beneficial uses of a water body include potable
water supply, support and propagation of fish and wildlife,
recreational activities, agricultural irrigation, industrial
supply, navigation, and shipping. Excess nitrogen in its
various forms will have adverse effects on a receiving
water (as discussed earlier in Section 1.5), anyone of
which can impact the desired use of the system. Where
appropriate, limits are imposed on the discharge of nitro-
gen (in addition to conventional parameters such as BOD
and SS) from point sources such as wastewater treatment
plants as a means of controlling the water quality and
ensuring the long-term beneficial use of a receiving water.
These "wasteload allocations" are generally based on
water quality criteria or standards established for the spe-
cific receiving water.
The 1987 re-enactment of the Federal Clean Water Act
allows the states to designate water use and to establish
water quality standards (WQS). These generally follow or
are more restrictive than the guidelines promulgated by
EPA; although they are defined based on protocols set
by EPA. Through research both in and outside of EPA,
dataware compiled from which recommendations are
made regarding the biological and chemical conditions
necessary to sustain or achieve a water body's desig-
nated use. In the cases of ammonia toxicity to aquatic
biota and nitrate effects on human health, the mecha-
nisms are well quantified and understood and generally
nonspecific with regard to receiving waters. As such, cri-
teria are established and broadly implemented (these are
discussed in Sections 1.5.3 and 1-.5.4). The impacts re-
lating to in-stream nitrification and biostimulation tend to
be more site-specific. Determination of their causes and
the implementation of controls often require water quality
studies and modeling evaluations specific to the water
body.
Technology-based standards do not exist for nitrogen, as
they do for SS and BOD, because of the varied nature
of nitrogen's effect on receiving waters. As such, state
agencies, with guidance from EPA, will generally use
water quality modeling approaches to determine appro-
priate point source wasteload allocations that will sustain
ambient water quality standards. If nitrogen control (nitri-
fication or nitrogen removal) is needed within a water
system, then a quantitative analysis of all point and non-
point sources is required to determine the appropriate
treatment needs for the municipal discharge.
To ensure that current discharge limits and waste load
allocations applied to each water body effectively main-
tain the water quality standards set for that body, states
periodically assess their condition. This is generally done
every three years, as mandated by the Clean Water Act.
States usually prioritize their waters for evaluation and
assess existing data for each. If data are lacking and the
waterway is a priority, a Water Body Survey and Assess-
ment is conducted. If it is found that existing discharge
limits are not effective in maintaining the beneficial use
of the water, improvements in treatment and controls are
generally required. The state may also propose a down-
grading of the designated use of the receiving water if it
finds that natural conditions and economic constraints
make maintaining the beneficial use impractical. This,
however, is less common and protocols for changing to
•a less^protective use require a rigorous evaluation that
includes public participation. The process of water quality
standards review and revision is outlined in Figure 1-8.
In summary, state-set ambient water quality standards for
nitrogen are translated via wasteload allocation modeling
methods into water quality-based permits for POTWs.
These permits serve as the basis for water quality-based
construction funding decisions. Figure 1-9 outlines the
major elements of the water quality-based standards-to-
permits process.
19
-------
Maintain No
WQS «-
List of Rivers, Streams,
Lakes, Coastal Areas
Not Meeting WQS
x
Are These Water
Quality Limited
Segments?
* X
Issue Technology-
Based Permits
— No
Stream Not
Selected for
Immediate
Detailed Review
Yes
Do These WQ Limited
Requirements Have
Permit and AT
Decisions Pending
Toxic/Human
Health Problems/or
Uses Not Consistent
with101(a)(2)?
Y|s
Select Priority Stream
Segments for Detailed
Review
t
Are Existing Data
Adequate?
No
(Conduct a ~"\
Waterbody Survey and ]
Assessment J
Yes
Physical .
Condition
>^
.«W
^
Are Designated Uses
Appropriate?
t°
Why Are
Designated Uses
Inappropriate?
Designate
Appropriate Uses
\
Section 304(a) «-
^v
Set Appropriate
Criteria
Perform Water Quality
Analysis and
Calculate Preliminary
Limits
I
/*
-
^f
Natural or
Irretrievable
Chemical
Conditions
Site-Specific
Criteria
-
Maintain
• Standard
\
Standards
to Permit
Process
RA
Approves
WQS
1
Standards
to Permit
Process
t
State
Adopts
Revisions
to WQS
and RA
Approves
No
«-
'
Are Economic or
Social Impacts
Widespread and
Substantial?
1
,Yes
f Provide Analyses to i
I Public J
*—
'
'
Hold Public
Hearing
|
State Adopts
Revision to WQS
•
'
States Submit
WQS to RA for
Review,
i
RA Disapproves
WQS; Notifies Required
State of Changes
^
!
Federal WQS
Promulgated in
Federal Register
+
Standards to
Permit Process
State Does
Not Adopt
+ Required
Changes
Figure 1-8. EPA water quality standards: Review and revision process (from Reference 46).
20
-------
I. Identify Water Quality-Limited Segments and
Set Control Priorities; Implement Local
Monitoring Program, If Necessary
II. Review and Revise (or Reaffirm) Water
Quality Standards
III. Develop Water Quality-Based Control
Requirements
IV. Incorporate Identified WQL Segments, Priorities,
Revised/Reaffirmed Standards, TMDLs, Effluent
Limits, and Feasible Nonpoint Source Controls
into Updated WQM Plans
V. Issue Water Quality-Based Permits;
Implement Nonpoint Source Controls
VI. Monitor Municipal and Industrial Sources for
Compliance; Perform Ambient Monitoring to
Document Protection of Designated Uses
Figure 1-9. Major elements of the water quality-based
standards to permits process (from Reference 46).
1.7 References
When an NTIS number is cited in a reference, that docu-
ment is available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
703-487-4650
1. U.S. EPA. 1975. Process design manual for nitrogen
control. EPA/625/1 -77/007 (NTIS PB-259149). Wash-
ington, DC.
2. U.S. EPA. 1984. Needs survey report to congress—
assessment of needed publicly owned wastewater
treatment facilities in the United States. EPA/430/09-
84/011. Washington, DC.
3. U.S. EPA. 1988. Needs survey report to congress—
assessment of needed publicly owned wastewater
treatment facilities in the United States. EPA/430/09-
89/001 . Washington, DC.
4. Delwiche, C.C. 1970. The nitrogen cycle. Scientific
Am. 223(3): 137-146.
5. McCarty, P.L., et al. 1967. Sources of nitrogen and
phosphorus in water supplies. JAWWA 59:344.
6. Martin, D.M., and D.R. Goff. 1972. The role of nitro-
gen in the aquatic environment. Rep. no. 2. Academy
of Natural Sciences of Philadelphia, Department of
Limnology.
7. Sylvester, R.O. 1963. Nutrient content of drainage
water for forested, urban, and agricultural areas. Al-
gae and metropolitan wastes. Tech. Rep. W61-3.
Robert A. Taft Sanitary Engineering Center.
8. Sepp, E. 1970. Nitrogen cycle in ground water. Cali-
fornia Department of Public Health, Bureau of Sani-
tary Engineering.
9. Metcalf and Eddy, Inc. 1990. Wastewater Engineer-
ing. 3d ed. New York: McGraw-Hill.
10. Landner, L. 1977. Sources of nitrogen as a water
pollutant: industrial wastewater. Prog, in Water Tech.
8(4/5):55-65. Pergamon Press. Printed in Great Brit-
ain.
11. Zolten, N.G. 1991. Leachate treatment in landfills.
Water Environ, and Tech. 3(5):63-66 (May).
12. Aniansson, B. 1990. Nitrogen—a vital element or a
threat to life. Acid Magazine—a special issue on ni-
trogen. Swedish EPA. Rep. no. 9 (June).
13. HydroQual, Inc. 1991. Water quality modeling analy-
sis of hypoxia in Long Island Sound. Prepared for the
Management Committee of the Long Island Sound
Estuary Study and the New England Interstate Water
Pollution Control Commission.
14. Weibel, S.R., et al. 1966. Pesticides and other con-
taminants in rainfall and runoff. JAWWA 58:1,075.
15. Metropolitan Washington Council of Governments,
Department of Environmental Programs. 1987. Poto-
mac River water quality—1985. Prepared for Wash-
ington Metropolitan Water Resources Planning
Board.
16. Manny, B.A., and R.W. Owens. 1983. Additions of
nutrients and major ions by the atmosphere and tribu-
taries to nearshore waters of northwestern Lake
Huron. Great Lakes Res. 9(3).
17. Hanneberg, P. 1990. Editorial comment to nitrogen:
vital element or threat to life. Acid Magazine—a spe-
cial issue on nitrogen. Swedish EPA. Rep. no. 9
(June).
1.8. Johnson, R.E., AT. Rossano, Jr., and R.O. Sylvester.
1966. Dustfall as a source of water quality impair-
ment. ASCE. JSED 92(SA1):145.
21
-------
19. Novotny, V. 1991. Urban diffuse pollution: sources
and abatement. Water Environ, and Tech. 3(12):60-
65 (December).
20. Novotny, V., and G. Bendoricchio. 1989. Linking non-
point pollution and deterioration. Water Environ, and
Tech. 1 (3):400-407 (November).
21. American Public Works Association. 1969. Water pol-
lution aspects of urban runoff. FWPCA Rep. no. WP-
20-15 (January).
22. Burn, R.J., D.R Krawezyk, and G.T. Harlow. 1968.
Chemical and physical comparison of combined and
separated sewer discharges. JWPCF 40:112.
23. Avco Economic Systems Corp. 1970. Storm water
pollution from urban land activity. EPA/110/34-
FKLO/770 (NTIS PB-195281). Washington, DC.
24. Haith, D.A., and LL Shoemaker. 1987. Generalized
watershed loading functions for stream flow nutrients.
Water Resources Bull. 23(3) (paper no. 86072).
25. Novotny, V. 1992. Unit pollutant loads. Water Environ.
and Tech. 4(1):40-43 (January).
26. Exner, M.E., and R.F. Spalding. 1985. Ground-water
and well construction in southeast Nebraska. Ground
Water 23(1):26-34 (January).
27. Neill, M. 1989. Nitrate concentrations in river waters
in the south-east of Ireland and their relationship with
agricultural practice. Water Resources 23(11): 1,339-
1,355. Printed in Great Britain.
28. Kaufman, W.J. 1974. Chemical pollution of ground
waters. JAWWA66(3):152-159.
29. Reeves, T. G. 1972. Nirogen removal: a literature
review. JWPCF 44(10): 1896-1908.
30. Water Environment Federation. 1992. Design of mu-
nicipal wastewater treatment plants. Manual of prac-
tice no. 8.
31. Brown, K.W., K.C. Donnelly, J.C. Thomas, and J.F.
Slowey. 1984. The movement of nitrogen species
through three soils below septic fields. J. Environ.
Quality 13(3):460-465.
32. U.S. EPA. 1991. The Chesapeake Bay: a progress
report—1990-1991. Prepared for the Chesapeake
Executive Council by the Chesapeake Bay Program
Office, EPA Region III. Annapolis, MD (August).
33. Ehell, M. 1990. The impact on water quality of nitro-
gen losses from agriculture: recommendations for im-
provements. Acid Magazine—a special issue on ni-
trogen. Swedish EPA. Rep. no. 9 (June).
34. U.S. EPA. 1973. Nitrogenous compounds in the en-
vironment. EPA/ASB-73/001. Washington, DC.
35. U.S. EPA. 1990. Long Island Sound study:i status
report and interim actions for hypoxia management.
Prepared by the staff of the Long Island Sound Study,
representatives from the states of New York and Con-
necticut, and representatives from EPA Regions I and
11. EPA Contract 68-C8-0105.
36. Ehreth, D.J., and E. Barth. 1972. Control of nitrogen
in wastewater effluents. Prepared for EPA Technol-
ogy Transfer Seminars (March).
37. U.S. EPA. 1984. Before and after case studies: com-
parisons of water quality following municipal treat-
ment plant improvements. EPA/430/9-83/007.
Prepared by HydroQual, Inc. Washington, DC.
38. U.S. EPA. 1985. Ambient water quality criteria for
ammonia—1984. EPA/440/5-85/001. Washington,
DC.
39. U.S. EPA. 1989. Ambient water quality criteria for
ammonia (saltwater)—1989. EPA/440/5-88/004.
Washington, DC.
40. Effler, S.W., C.M. Brooks, Martin T. Aver, Susan M.
Doerr. 1990. Free ammonia and toxicity criteria in a
polluted urban lake. Res. JWPCF 62(6).
41. Viraraghaven, T. 1988. Nitrogen budget and septic
tanks systems—an appraisal. Water Supply 6:89-91.
Brussels: Pergamon Journals.
42. Shuval, H.I., and N. Gruener. 1977. Infant methemo-
globinemia and other health effects of nitrates in
drinking water. Prog. Water Tech. 8(4/5): 183-193.
Pergamon Press. Printed in Great Britain.
43. Mirvish, S.S. 1977. N-nitroso compounds, nitrite and
nitrate: possible implications for the causation of hu-
man cancer. Prog. Water Tech. 8(4/5):195-207. Per-
gamon Press. Printed in Great Britain.
44. Miller, K.J. 1990. U.S. water reuse: current status and
future trends. Water Environ, and Tech. 2(11):83-89
(November).
45. Crook, J., T. Asano, and M. Nellor. 1990. Groundwa-
ter recharge with reclaimed water in California. Water
Environ, and Tech. 2(8):42-49 (August).
46. U.S. EPA. 1983. Water quality standards handbook.
(NTIS PB92-231851). Washington, DC.
22
-------
Chapter 2
Total System Design with Nitrogen Control
2.1 Introduction
For many readers, this is one of the more important chap-
ters in this manual. It is directed to all decision-makers
involved in the assessment, selection, and design of a
nitrogen control strategy at a municipal wastewater treat-
ment facility.
The overall intent of this chapter is to convey the mes-
sage that there is no universal response to a nitrogen
control need. The right unit process selection, rather than
standing alone, is influenced by all things that precede
and follow in the overall scheme of the treatment works.
Process selection must respond to the facility's waste-
water and residuals management objectives, under the
constraints imposed by both the natural and social
environments.
This chapter:
• provides summary oversight and detail at a level and
in a form that are convenient for frequent reference,
• introduces and compare the various nitrogen control
technologies and fundamental aspects of their per-
formance, and
• describes sound practices and linked processing con-
siderations that will assist the user in the application
of nitrogen control technology and with related problem
solving within the context of the integrated treatment
works.
The information presented in this chapter focuses on
avoiding the fundamental mistakes often encountered in
the first 5 to 10 percent of the design, which, if continued
through the remainder of the project, can result in a facility
that either fails to meet its design intent or is grossly
oversized.
The interested reader will be best served by reading this
chapter for overall understanding before referring to the
more specific and detailed nitrogen control technology
material contained in the balance of the manual. After
reading the more detailed design chapters, rereading this
chapter may be appropriate before entering into detailed
process design. After design completion, the material
contained in Section 2.2 can be used to check some of
the design decisions.
Those readers desiring more information than found in
this manual are referred to the recently revised Water
Environment Federation's Manual of Practice (MOP) No.
8, Design of Municipal Wastewater Treatment Plants (1).
The three introductory chapters of that publication provide
greater understanding of design approaches and issues,
wastewater characteristics, and the decision-making that
precedes detailed design.
2.2 Summary Checks for Process Selection
and Design
•
This section summarizes a variety of material and will
serve as a convenient reference. The section begins by
focusing on the most frequently encountered problems at
small wastewater treatment plants. Then, with an empha-
sis on nitrogen control aspects of the subject, the reader
is provided with a brief narrative and tabular summary of
general guidance checks for wastewater charac-
terizations and waste solids management schemes, the
stoichiometry of the frequently encountered reactions of
interest, and an oversight review of the proposed process
design.
2.2.1 Overview
Successful wastewater treatment is dependent on opera-
tor understanding, responsible administration, and sound
design. Failure of any agent of successful wastewater
treatment (i.e., operator, administrator, or designer) to re-
spond adequately to his charge inevitably results in proc-
ess upset and eventual failure.
The designer can mitigate the failure opportunities at all
treatment facilities by:
• selecting tolerant, flexible treatment processes with
conservatively designed, responsive sludge process-
ing and disposal schemes,
• urging and providing for operator training, and
• working with the administrator to ensure adequate sup-
port for the needs of the facility.
23
-------
Small wastewater treatment plants are more likely than
larger plants to be underfunded, understaffed, and not
always reached by the professional networks that can
offer troubleshooting advice. Table 2-1, taken from the
recently revised MOP 8 (1), summarizes the findings of
a recent EPA survey of 150 small plants (<44 Us [1 mgd])
with debilitating problems (2).
Table 2-1 reveals that the three agents responsible for
successful wastewater treatment plant operation also
share blame. All would do well to remember the seven
major conclusions and recommendations derived from
the EPA study.
2.2.2 Wastewater Characterization and Waste Solids
Management
The three biggest errors in design of nitrogen control
municipal wastewater treatment facilities are the same as
those encountered in any treatment application:
• improper characterization of the influent wastewater,
• an inadequate and/or improper solids processing and
disposal scheme, and
• failure to consider transient sludge processing recycles
in relation to the buffering capacity of the wastewater
and the processes found in the liquid processing train.
Sound design uses a feed forward (wastewater charac-
teristics), feed backward (solids disposal and effluent limi-
tations) approach to develop the liquid and solids
processing trains that are compatible with the anticipated
staffing plan and the constraints or opportunities that are
established in the natural and institutional environments.
The proper design provides the least expensive, imple-
mentable solution that results in balanced media (i.e., air,
land, and water) impacts and that complies with the cur-
rent and anticipated institutional needs of the public, the
owner, and the regulatory bodies.
Common sense and detailed process understanding go
hand in hand in soundly based process applications. Ta-
ble 2-2 lists some of the wastewater characterization and
solids management checks that may be employed by
both the process designer and the reviewer to ensure a
successful application of nitrogen control technology in
municipal applications.
Two-thirds of Table 2-2 is devoted to influent wastewater
characterizations. Characterization errors may be inad-
vertent (e.g., didn't know any better or trusted an invalid
data base) or conscious (e.g., overquantification of the
influent flow and pollutant load to safeguard against failure).
The effluent standards of the plant (pollutant limit and time
interval) dictate the needed influent characterizations.
These characterizations then feed forward into the design
of those processes needed to meet the effluent stand-
ards. If the plant has a maximum month and week per
year effluent pollutant restriction (corresponding to a de-
sign reliability of 92 and 98 percent, respectively), influent
wastewater pollutants should also be characterized at
these conditions to permit the establishment of the con-
trolling condition for facility design.
However, even the best liquid processing design is
meaningless if the solids processing train is inappropri-
ate, improperly sized, or unreliable. The balance of Ta-
ble 2-2 is devoted to the issues that are found therein
and the recycles that are generated by the solids proc-
essing train.
Solids management decisions, beginning with the ulti-
mate disposal concept, feed backward into the facility
design and, practically, can have much to do with the
sound selection of the liquid processing technology. This
is especially the case when nitrogen control is a process-
ing objective. Failure by the designer and owner to ap-
praise realistically and provide for adequate ultimate
solids disposal places the entire treatment complex at
risk.
The recycles returned to the liquid processing train, and
the waste solids from it, integrate the treatment works so
that "everything is connected to everything else." Too high
a return of SS from the solids processing train can cause
a plant's solids residence time (sometimes referred to as
the mean cell residence time), as determined by dividing
the biomass in the reactor by the mass of SS removed
(wasted and lost in the effluent) per day, to drop below
the acceptable value for retention of the nitrifying organ-
isms. An additional result is the spin (continuous recircu-
lation) of previously generated solids, which taxes the
processing capability of all processes up to and including
the one that was the source of the recycles. The idiscon-
tinuous return of a soluble nitrogenous recycle imposes
a special load that must be anticipated in design and
buffered by the processes and alkalinity of the wastewater
to avoid undesirable nitrogen excursions in the plant's
effluent.
2.2.3 Stoichiometry
Wastewater treatment plant design in general and nitro-
gen control concepts specifically involve a variety of spe-
cific stoichiometric reactions and more uncertain process
assumptions. Table 2-3 summarizes the stoichiometric
reactions that are frequently employed and encountered
in the design of nitrogen control systems. Additional in-
formation regarding these approaches, and all of the
stoichiometric reactions, are provided in the referenced
sections of the manual. The stoichiometric constants are
described to two significant figures; greater precision is
unnecessary.
Table 2-3 shows that the nitrogen control reactions can
deplete and add alkalinity. Alkalinity control is important
in the process design to avoid low pH attenuation of the
nitrification reaction. The alkalinity provided from a variety
24
-------
Table 2-1. Survey of 150 Small Plants with Debilitating Problems (Adapted from References 1 and 2)
MOST FREQUENTLY OCCURRING PERFORMANCE LIMITING FACTORS (PLFs)
(From Survey Summary of Five Most Frequently Occurring PLFs by 10 Regions)
Likely Principal Causative Agent
No. of Regions Performance Limiting Factors Operator Administrator Designer
10
9
8
6
6
4
4
3
Poor Operator Understanding/Application of Process Control X
Solids Handling and Sludge Disposal
Infiltration/Inflow
Staffing (e.g., too few staff, low pay, turnover)
Laboratory Capability for Process/NPDES Testing
Process Design Errors (e.g., clarifiers, aerators, disinfection)
Municipal Support (administrative and technical)
Preventive Maintenance Program
Uncertain
Uncertain
X
Uncertain
X
X
Uncertain
SURVEY CONCLUSIONS AND RECOMMENDATIONS
• Conclusion No. 1 -Activated sludge may not be a good design choice for many small plants.
- Give fuller consideration to simpler, more tolerant treatment processes (e.g., fixed media and natural systems) that are less
dependent on highly skilled operators.
- Select a treatment technology based on realistic appraisal of all costs (including conservative estimates of sludge quantity
and concentration for sludge treatment and disposal, staff salary, recruitment and training, equipment maintenance and
replacement, and administrative costs).
• Conclusion No. 2 -Plant inflexibility undermines operability.
— Designers should conscientiously build flexibility into systems (e.g., piping configurations, redundant unit processes, variable
speed pumps for wasting and return, aeration equipment, and equalization tanks for I/I extremes and/or bleed-back of
discontinuous sludge processing recycles to the liquid processing train).
• Conclusion No. 3 -Small plants have front- and back-end problems with process design.
— Pumps, piping, and aeration systems should be designed to accommodate increased solids and rags in the system when
primary treatment is not provided. '
— Operators should be made aware of the need to remove floating debris that passes primary screening.
- Designers should consider finer bar screens, especially when primary sedimentation is not provided and, once screenings
and floatable material are removed from the liquid processing train, provide practical facilities to facilitate their permanent
removal instead of their internal recycle (and buildup).
- Solids handling facilities should have the capability of properly stabilized liquid sludge removal and disposal.
• Conclusion No. 4 -Heavy loads can confound both skilled and unskilled operators.
— Community administrators and design engineers should frankly discuss and agree on realistic loadings for the facility in the
planning process (a conservative, if not skeptical, design approach should be taken to accommodate I/I and industrial loadings).
• Conclusion No. 5 -Staffing difficulties aggravate poor performance.
— Administrators should seek to attract and maintain a better staff through increased operator status and visibility using at
least one (or nearly so) full-time position with a salary comparable to other critical municipal functions (e.g., the police chief)
and with reasonable authority for budgeting, purchasing, hiring and firing. Administrators should also provide reasonable
opportunities for training and certification.
• Conclusion No. 6 -Plant budgets and user charges may be too low.
— Better fiscal management must start with a separate budget for the treatment plant that includes a sinking fund to cover
replacement of major equipment, and that supports adequate staff salaries as well as training and required certification courses.
• Conclusion No. 7 -Municipal support is a subtle but vital need.
— Outreach and information transfer must be applied to increase community support; consider making the treatment plant into
a multi-use facility that accommodates recreational facilities and shares offices and building space with other community
agencies and organizations.
25
-------
Table 2-2. Wastewater Characterization and Solids Management Checks for the Design of Municipal Wastewater
Treatment Facilities with Nitrogen Control Technology
RAW INFLUENT WASTEWATER CHARACTERISTICS
• Per Capita Pollutant Generation Rates
- BODS and SS = 0.08-0.11 kg/capita/d (0.18-0.25 Ib/capita/d)
• Low values are favored with strictly domestic service and absence of household garbage grinders.
• Quantification lower than cited value should be confirmed (may be encountered with lower income communities), and
may be symptomatic of upsystem diversions and/or poor sampling or analysis.
• Quantification higher than cited value should be confirmed and may be symptomatic of unqualified plant recycles,
unregulated septage dumping, sampling programs that exclude weekends, poor sampling or analysis, incorrect flow
identification (e.g., using influent and recycle flows with raw Wastewater characteristics), industrial or commercial .
discharges, and significant daily transient population. If confirmed as industrial or institutional, 'determine seasonal,
weekly, and daily waste discharge characteristics and plan for the future, as appropriate. Watch out for sampling
schedules at small plants that may be limited to only one to five days per week with sampling only during staffed
hours. Manual composite samples may not be true composites. Remember that industrial and commercial releases are
typically less on weekends and holidays (i.e., after the end of the Friday shift cleanup).
• Perform long-term inert SS mass balance check around whole plant (influent against effluent SS and net waste solids).
See also waste solids production.
* Sustained low-flow period may give more reasonable characterization than high-flow period given sewer system
overflows and losses. However, high-flow period may experience load from previously deposited, now resuspended
solids. See also maximum and minimum characterizations.
• Pollutant Relationships
- BODj/COD > 0.45-0.55
Lower values may indicate a fair degree of stabilization occurring in the sewers (enhanced by steep slopes, aerobic
conditions, higher temperatures), or high levels of I/I (influx of more refractory organics); or attributable to nonacclimated
seed or poorly biodegradable industrial waste. Higher values may indicate fermentation in long residence time anaerobic
sewers, false BODS positives due to sulfide, presence of nitrifiers due to recycles, or high levels of soluble biodegradable
industrial waste.
~VSS/SS = 0.7-0.8
Higher values are favored with domestic wastes. Lower values are often encountered in combined sewer areas.
VSS-to-SS ratios less than 0.7 should be confirmed and are indicative of the routine receipt of partially stabilized wastes
(septage), water plant sludges, industrial wastes or pronounced precipitation induced inflow.
-SS/BOD5S0.8-1.2
See preceding paragraph.
-Soluble BODs/Total BODS = 0.35-0.45
See preceding paragraph. Higher values may be indicative of industrial waste. High values dictate special concern with
filamentous bulking because of high immediate stabilization and DO stresses in suspended growth system.
~ Particulate BODs/Particulate COD < Total BOD/Ibtal COD
Soluble phase is typically the most biodegradable.
-TKN/BODsss 0.1-0.2
Higher values may be indicative of industrial waste or the presence of ammonium from solids processing recycles
introduced before sampling the influent. Lower values may be indicative of nutrient-deficient industrial waste.
Maximum And Minimum Characterizations
- Flows
Guard against the inadvertent inclusion of plant recycles quantified with the raw sewage determination because of location
of the flow meter. Maximum flows may be constrained by sewer system or headworks diversions before the flow meter.
Snow melt can give high flows. Minimum flows may be limited by seasonal infiltration. :
- Pollutants
Guard against poor sampling and/or analysis; this becomes progressively more important as maximums and minimums
are Identified. Look for similar trends of wastewater constituent ratios as a form of data validation. Validate observations
by determining if operating solids levels and waste solids production values follow reported pollutant peaks. Delete any
26
-------
Table 2-2. Wastewater Characterization and Solids Management Checks (continued)
erroneous result from average determination. Analyze several years of data to confirm. Common errors are failure to
completely filter SS samples and overflowing sampling bottles, which yield high SS. Samples with high settleable SS may
yield low BOD5 values because settled solids are not suspended throughout the duration of the test. Believe COD values
more than BOD5 determinations. Make sure that calculated masses are determined with concentration and flow of event.
-Alkalinity
Strongly influenced by native water and water treatment practices, as well as by magnitude of inflow and infiltration. May
be influenced by sea water intrusion in low-lying coastal areas. Note: Inexperienced designers often ignore alkalinity when
designing nitrogen control facilities, especially when planning on the discontinuous return of unoxidized soluble
nitrogen-laden recycles from the solids processing train.
WASTE SOLIDS PRODUCTION, PROCESSING, AND DISPOSAL
• Net Waste and Effluent SS
For most municipal wastewaters with and without primary treatment but without enhanced biological or chemical phosphorus
removal, where applied (BODs + SS) -*• 2 = A, the combined net waste and effluent SS are typically on the order of the following
"rule of thumb" estimates below. (Additive inert solids for enhanced biological or chemical phosphorus removal is 0.02 to 0.04 times
AP, where AP is the enhanced P removal in mg/L)
Multiplier of A
Carbonaceous Stabilization (Oxidation) 0.9-1.1
with aerobic or anaerobic digestion of total sludge mass 0.7-0.9
Carbonaceous and Nitrogenous Oxidation 0.8-1.0
with aerobic or anaerobic digestion of total sludge mass 0.7-0.9
• Ultimate Disposal Scheme
If disposal outlets are not in the owner's total control, or processed sludge storage is inadequate for at least several months of
storage prior to ultimate disposal, the residuals management scheme is at risk and the liquid processing scheme may be subject
to excessive solids-laden reycles; this may upset the performance of one or more unit processes of the liquid processing .train.
• Solids Processing
- Nitrogen Recycles
Soluble nitrogen causes the most severe nitrogen impact. AH solids, whether in process or in storage, undergo hydrolysis
(liquification) and contribute soluble nitrogen. Nitrogen solubilization is most severe in thermal sludge conditioning and
digestion (aerobic, anaerobic, and composting); neither contribute a favorable biodegradable carbon-to-nitrogen ratio such
that the recycled nitrogen is completely removed by biomass synthesis after the flow is returned to the aeration basin.
Aerobic digestion can also contribute ammonium-nitrogen, when nitrogen oxidation in the aerobic digester is limited by the
low pH (typically from 5 to 6) resulting from an unfavorable soluble nitrogen-to-alkalinity ratio. Discontinuous solids
processing subsequent to the process that causes solubilization exacerbates the stress caused in the liquid processing
train by the soluble nitrogen. The more discontinuous the solids processing/dewatering process, the greater the buffer or
treatment reserve needed in the liquid process train. Note: Soluble refractory organic nitrogen, such as encountered with
discontinuous sludge processing or septage, may seriously affect the ability of a plant to achieve stringent total nitrogen
effluent requirements.
- SS Recycles
Sedimentation tanks in the liquid processing train are best used for clarification, not thickening. Most optimum solids
processing is achieved by attempting to concentrate the waste solids in processes that are least sensitive to flow (e.g.,
gravity thickening and dissolved air flotation) before using other more volume-dependent processes (e.g., digestion,
centrifugation, and belt thickening and dewatering). This avoids SS washout and excessive SS recycle as well as
inadequate mainstream capacity resulting from the failure to achieve the anticipated solids concentration in the liquid
processing train. Care should be exercised in assuming that gravity thickeners have too much storage (i.e., undersizing)
as SS captures can change from greater than 95 percent to disastrously low values in a matter of hours.
27
-------
Table 2-3. Stolchiometry of Nitrogen Control and Other, Often Related Fteactions
MANUAL SECTION
CONSIDERATION: NITROGEN CONTROL REACTIONS
Biochemical Nitrification 3.2.1
• 4.6 mg oxygen required/mg nitrogen oxidized
• 7.1 mg CaCOa alkalinity depleted/mg nitrogen oxidized
• 0.10-0.15 mg net volatile solids/mg nitrogen oxidized
Biochemical Denitrification 4.2.2.1, 4.2.2.2
• 2.9 mg oxygen liberated/mg nitrate nitrogen reduced; there are 1.5 mg COD/mg '
methanol (CHaOH)
Note: Sufficient substrate (COD) must be added to satisfy nitrogen reduction and
synthesis needs, typically about 1.5 times stoichiometric predictions.
• 3.6 mg CaCOa alkalinity is recovered/mg nitrate nitrogen reduced
• Same to slightly lower net volatile solids/COD removed as with any biological
system (methanol is preferentially oxidized as opposed to synthesized); yields can
be lower if an aerobic stabilization step is provided
Note: COD removed is the total amount of COD oxidized and synthesized.
Breakpoint Chlorination 2.5.2.3
• 7.6 mg chlorine/mg ammonium nitrogen
Note: Practice typically requires values 1.1-1.3 times stoichiometric predictions
when treating biologically stabilized wastewaters. Values may be appreciably
greater when processing raw or partially stabilized wastewaters because of ;
competing side reactions with the SS, organics, and nitrite nitrogen. ;
• 1.4 mg CaCOa alkalinity loss/mg chlorine added (1 mg dissolved solids added/mg
chlorine added)
• NaOCI is buffered (with caustic) and will not cause an alkalinity depletion (1.7 mg
• dissolved solids added/1 mg chlorine [in NaOCI] added)
CONSIDERATION: OTHER, OFTEN RELATED REACTIONS
Chlorine Demand Due to Incomplete Nitrification or Denitrification 2.8.5
• 5.1 mg chlorine/mg nitrite nitrogen, yielding nitrate nitrogen ;
Dechlorination
• 0.9 mg SOz/mg chlorine (expressed as Cla)
Atote: Practice requires values 1.1-1.2 times stoichiometric predictions. (2.4 mg ;
CaCOa alkalinity depleted/mg SOa added)
• 1.5 mg NaHSOa/mg chlorine (expressed as C\z)
Note: Practice requires values 1.1-1.2 times stoichiometric predictions (1.4 mg
CaCOa alkalinity depleted/mg NaHSOa added)
Alkalinity Sources
• 1.8 mg CaCOa alkalinity added/mg CaO (quicklime) added
• 1.4 mg CaCOa alkalinity added/mg Ca(OH)2 (slaked lime) added
• 1.2 mg CaCOa alkalinity added/mg NaOH (caustic) added
• 0.9 mg CaCOa alkalinity added/mg NaaCOa (soda) added
272
Phosphorus Removal
• Metal Salts
Typically need about 1.25-1.75 moles of metal/mole phosphorus remaining after
background removals to achieve low soluble phosphorus residuals of less than
1 mg/L This results in the following general representations:
Alum (0.87 mg Al/mg P at 1 mole/mole):
28
-------
Table 2-3. Stoichiometry of Nitrogen Control and Other, Often Related Reactions (continued)
- 5.6 mg CaCO3 alkalinity loss/mg Al added
- 3.9 mg AIPO4 formed/nig P removed
- 2.9 mg AI(OH)3 formed/mg Al added
» 6.1 mg AI2PO4(OH)3 formed/mg P removed (assuming 80% P removal) at
1.5 mole Al/mole P remaining for chemical removal
Ferric Chloride (1.8 mg Fe/mg P at 1 mole/mole):
-2.7 mg CaCO3 alkalinity loss/mg Fe added
- 4.9 mg FePO4 formed/mg P removed
- 1.9 mg Fe(OH)3 formed/mg Fe added
«• 7.9 mg Fe2PO4(OH)3 formed/mg P removed (assuming 80% P removal) at
' 1.5 mole Fe/mole P remaining for chemical removal
• Lime
- 5.4 mg CasOH(PO4)3/mg P removed plus
- Precipitated CaCO3, which depends on lime dose and final pH (which will cause
the precipitation of mg(OH)2 at pH >. 11); lime close is influenced by wastewater
alkalinity
• Enhanced Biological Activity
- 3.6 mg cellular storage products/mg P removed beyond normal background
conditions
of external additives is included in the tabulation of
information.
A listing of phosphorus removal reactions is also provided
in Table 2-3 since current practice often finds treatment
standards that specifically limit both nitrogen and phos-
phorus. As shown, phosphorus removal can be achieved
with a variety of well-demonstrated approaches. The met-
al salts that are routinely added for phosphorus removal
also cause an alkalinity depletion. The metal salt alkalinity
depletion associated with phosphorus removal (or if such
depletion occurs as a result of other uses, such as de-
watering) must be considered when designing for control
of nitrogen and phosphorus.
2.2.4 Process Design Review
Table 2-4 summarizes the three most important review
checks for nitrogen control facilities. The first two checks
apply to any wastewater treatment plant design. The last
check specifically emphasizes the details of process de-
sign with a progressively narrower focus on nitrogen con-
trol technologies. The basis for these recommendations
flows from the subsequent discussions contained in this
chapter, and in the remaining chapters of this manual.
The process design details described in Table 2-4 are not
exhaustive and should be used only as general guidance.
The guidance is best applied as "flags" for further explo-
ration; other considerations may prove equally important
in the successful design of any nitrogen control facility.
2.3 Fundamental Treatment Technology
Concepts
The purpose of this section is to introduce the reader to the
fundamentals of wastewater treatment, the multiplicity of
options, and the generally available body of knowledge for
nitrogen control at the time of this manual's preparation.
2.3.1 Wastewater Treatment
2.3.1.1 Overview
In municipal wastewater treatment, the designer and
owner should lean toward low-maintenance, tolerant and
ample capacity facilities to reflect the uncertainty of staff-
ing, maintenance, and remedial action in a public market-
place where funding of major capital improvements is
uncertain and achieved only by public indebtedness with
political and public oversight. Both must walk a careful
line between providing tolerant facilities that can respond
to a multitude of future uncertainties and a blatant over-
design that results in the misuse of public monies for
clearly superfluous facilities. When in doubt, trust experi-
ence, which strongly suggests that simplicity and har-
mony with naturally occurring reactions are likely to serve
better than a multitude of unit operations for an optimized
desktop objective and/or the temptations of an unproven
form of high technology.
With the exception of disinfection, all wastewater treat-
ment processes are directed toward separating pollutants
into innocuous gaseous or concentrated end products
from the liquid flow stream. Each unit process or unit
29
-------
Table 2-4. The Three Most Important Review Checks for Nitrogen Control Facilities
1. RESIDUALS MANAGEMENT: If You Remember Nothing Else, Remember This
• Does the facility have a reliable means of solids handling and disposal, conservatively designed to meet the actual needs of
the intended operation, with some backup should the preferred mode of operation go down or fail to be realized as
anticipated?
2. FUNDAMENTAL FLAWS: If the Assumptions Are Wrong, the Conclusions Probably Are Too
• Do the per capita wastewater characterizations make sense for the service area?
• Were maximum and minimum design wastewater characteristics logically defined and used in the design, and are they ,
consistent with the compliance interval defined in the plant's discharge limitations?
• Can the performance of the plant meet the controlling compliance interval and pollutant restriction; is the anticipated average
day performance value of the technology at least one-half of the allowable monthly maximum?
3. PROCESS DESIGN DETAILS: Last, but Often Not the Least
• Mass Balance
— Was a mass balance for all pollutants of concern prepared? ' !
• Waste Solids
— Does the design fairly anticipate the waste solids gain due to the application of external additives and the recycles from the
solids processing train?
• Waste Solids Recycles-
— Does the design anticipate the recycles as they will be experienced?
— How were the soluble nitrogen recycles addressed?
• Systems Intended to Maximize Natural Biological Denitrification and Enhanced Biological Phosphorus Removal
— Has the design been checked for weaker wastewaters than anticipated in the basis of design?
— Does it have backup strategies, embodied by external additions (substrate and metal salt), should the system not perform
as anticipated?
• Final Sedimentation with Effluent Filtration
— Is shrinking the final sedimentation system because of a following effluent filter likely to yield an effluent filter that cannot
respond to the high flow conditions and needs?
• Alkalinity Check
— Were the nitrification, metal salt addition, and other chemical demands determined?
— Was credit taken for alkalinity return derived from planned biological denitrification?
— Are external sources of alkalinity needed?
— What about the alkalinity demand for the discontinuous recycle of ammonium-laden returns back to the main flow stream?
• External Additions of Foreign Substrates (e.g., methanol) for Denitrification
— Will the biological system automatically respond to the addition of foreign substrates (e.g., nonfood wastes such as
methane!)? The acclimation response is both compound form and mass specific. Unsatisfactory denitrification
performance, even with proper driving substrate additions, and foreign substrate pass-through will occur until the
acclimation response is complete.
• Suspended Growth Nitrogen Control System Specifics
— Is the solids residence time, 9C, of the nitrogen oxidation culture approximately 7 to 10 days when wastewater temperatures
are 15°C (59°F) or lower; is the CBOD5 loading 0.1-0.15 g CBODg/g MLVSS/d?
— Is the design MLSS or the return sludge concentration too high for the anticipated flow regimes and sedimentation system?
MLSS and return sludge concentrations, respectively, of 2,000 and 7,000 mg/L are likely safe under all flow regimes for
surface overflow rates (SORs) of up to 70 m3/m2/d (1,700 gpd/sq ft); values 3,000 and 10,000 mg/L respectively, may be
unsafe under elevated flow regimes for most sedimentation system designs with SORs of 50 m3/m2/d (1,200 gpd/sq ft) or
more).
— Is the return rate consistent with the return sludge concentration under the flow regime of concern (e.g., the maximum flow
week or day)?
— Has a means for bulking sludge control been provided?
Return sludge chlorination is safest even with selector technology.
30
-------
Table 2-4. The Three Most Important Review Checks for Nitrogen Control Facilities (continued)
3. PROCESS DESIGN DETAILS: Last, but Often Not the Least (continued)
- Is the oxygen transfer assumption appropriate?
Using one-half of the clean water transfer rate for design is generally protective unless dealing with fine pore diffuser
designs at the head end of a plug flow reactor; 75 percent of the clean water transfer rate may be unrealistically high for
design purposes without specific justification.
- Has the oxygen supply been checked under all carbonaceous and nitrogenous loading conditions and does it allow the
plant to respond to all reasonably anticipated conditions throughout its design life?
— Are design nitrification kinetic rates based on the correct percentage of nitrifiers in the MLVSS?
The number of nitrifiers is determined by the available nitrogen which changes less than the applied carbon with broad
applications of treatment technology. The biomass formed by the carbonaceous substrate acts to dilute the number of
nitrifiers with increasing carbonaceous loadings. This causes changes in nitrification rates when measured per unit of
'total MLVSS as discussed in Section 3.3.6. Use of the solids residence time design procedure (Section 6.4.1.1) is an
alternate approach.
- For seasonal permits, is the month before the coldest temperature month that the nitrification effluent standard is applicable
also considered when selecting the design temperature?
Permit conditions must be satisfied for the entire month. This can only be ensured by designs that allow for attainment of
the operating conditions for the required level of nitrification in the month that precedes the month that the nitrogen control
effluent standards apply.
- Are the denitrification rates and extent of denitrification reasonable?
High-level denitrification can only be ensured with ample available soluble substrate. The reduction of 1 mg NOa-N will satisfy
a wastewater oxygenation requirement of 2.9 mg COD. A conservative denitrification design would assume that only 50
percent of the applied soluble substrate is directly available for the denitrification reaction; the balance of the denitrification
reaction achieved is by bacterial stabilization of trapped applied wastewater particulates and cellular respiration.
The endogenous denitrification rate can be no greater than the equivalency of a reasonably anticipated endogenous •
solids destruction or respiration rate. If the overall endogenous solids destruction rate is 0.02-0.10 g VSS/g MLVSS/d,
the matching stoichiometric denitrification rate is 0.01-0.05 g NOg-N reduced/g MLVSS/d. Wuhrman and others have
shown that the endogenous oxygen consumption rate with nitrate as the electron donor is about one-half of that using
oxygen; so in actual practice the solids destruction rate of 0.02-0.1 g VSS/g MLVSS/d would normally correspond to
observed rates of about 0.005-0.03 g NOa-N reduced/g MLVSS/d.
• Attached Growth Nitrogen Control System Specifics
- Were design strategies for "as required" operator implementation provided to counter unique SS control issues for trickling
filter and rotating biological contactors?
Sloughing should be considered. Means of downstream capture should be considered. Many studies have shown that in
separate-stage nitrifying trickling filters the effluent SS equal the influent SS; therefore, permit requirements will dictate the
need for downstream clarification (or filtration) in these cases.
Attached growth systems produce a difficult to capture colloidal suspension. Use of suspended growth solids contactor
concepts with combined carbon and nitrogen oxidation technologies, with design MLSS concentrations of 500-1,500
mg/L, will produce a more visually attractive, lower SS effluent quality.
- Is the nitrification technology correctly applied?
Municipal wastewaters have trash that can clog the media. Upstream clarification or fine screening is usually necessary
to avoid media blinding and clogging.
- Is the target NH^-N concentration reliably achievable?
Unlike suspended growth systems, attached growth system nitrification performance may be more limited by oxygen
transfer than temperature. Additionally, nitrification technology with attached growth systems in either a combined or
tertiary application becomes progressively uncertain as the desired effluent ammonium levels become very stringent.
- Does the proposed nitrification system have the needed auxiliaries?
A recirculation capability should be provided for all attached growth systems unless there are consistent provisions for low
BOD in the applied wastewater or other design provisions are made. Recirculation inherently sends nitrifiers to the front
of the system, provides favorable dilution of soluble biodegradable carbonaceous compounds, and is the means of
providing more oxygen to the applied load for trickling filters.
Trickling filters should have the ability to control the instantaneous application rate independently of flow. Care should be
exercised in the election of some cross-flow media with combined carbon and nitrogen oxidation systems. Vertical flow
media is a safer design choice if there are clogging concerns.
31
-------
Table 2-4. The Three Most Important Review Checks for Nitrogen Control Facilities (continued)
3. PROCESS DESIGN DETAILS: Last, but Often Not the Least (continued)
Rotating biological contactors should have the ability to periodically clean (air scour) the carbonaceous stages of excess
blogrowth and remove settled solids from the bottom of each stage.
Tertiary applications of either technology may benefit from the ability to receive upstream diversions of clarified raw or
partially stabilized wastewaters.
— Is the denitrification system correctly sized for its application?
Successful denitrification in tertiary applications requires bacterial acclimation to the foreign substrate and sufficient .
organism numbers to achieve the desire temperature-dependent reaction in the available residence time. A dirty bed
operation is mandatory when denitrification is incorporated into effluent filtration, and hydraulic application rates may be
appreciably lower than routinely encountered with conventional effluent filtration applications. Overall system performance
will be influenced by the time in service of all parallel reactors. Limiting the number of parallel reactors may significantly
compromise overall performance as units are removed for cleaning and acclimate on return to service. The need for
periodic rigorous cleaning must be anticipated. ' '.
operation is preparatory for the next. Those processes
that result in a gaseous product are normally followed by
a subsequent solids concentration step until the liquid or
solids residual is suitable for return to the environment.
"Everything has to go somewhere."
Acceptable ultimate release to the environment is by one
of two ways—dilution or concentration. Management
strategies and treatment technologies that promise the
maximum dilution are likely to be superior to those that
promise the maximum concentration, and should be se-
lected first if the opportunity presents itself. Similarly,
treatment technologies with the fewest moving parts are
likely to be superior to those with many and should be
selected first when the opportunity presents itself.
The foregoing paragraphs provide a generalized set of
Ideals that can be used to guide project decision-making.
Practicalities, of course, ultimately govern. Project deci-
sion-making must ultimately balance the human and en-
vironmental resources of time, space, energy, and capital.
Indeed, most key project decision-making becomes more a
matter of "what you can't do" than "what ydu can do," and
some unit process selections become the only option avail-
able for achieving compliance with the needs of the regu-
lator and public in the time available for implementation.
2.3.1.2 Technologies
Table 2-5 provides a categorical description of the proc-
esses typically encountered in municipal wastewater
treatment. The intent of this discussion is twofold. First,
it is to establish the fundamentals of wastewater treat-
ment from several broadly based perspectives that may
well serve the designer and reviewer in their considera-
tion of the overall process scheme and specific unit proc-
ess options. Second, it is to provide the reminder that
overly optimistic design assumptions elsewhere in the
process design may doom to failure even the most con-
servatively designed nitrogen control facility within the
integrated works of the treatment facility. For more infor-
mation than is provided here, the reader is referred to
References 1 and 3.
A wide variety of unit processes and processing options
are available to the designer. Procedurally, they consist
of a series of reactors and separators. Reactors create
gaseous end products, oxidize, reduce, solubilize, immo-
bilize, or physically condition. Separators result in low and
high solids product streams. Reactor and separator proc-
esses can be "passive" or "reactive," with their design and
performance generally influenced by some key depend-
ency on hydraulics, pollutant concentration, or pollutant
mass. The majority of the unit processes are passive and
largely sized on hydraulic considerations.
The performance of "passive" unit processes is not
strongly subject to operational manipulation. Here, the
designer has the responsibility to err on the conservative
side, since "not enough" would mean "provide more" be-
cause few, if any, remedial operational strategies are
available to counteract the undersized units. Well-designed,
passive unit processes are preferred for plant owners/op-
erators with a low personnel commitment and a desire to
avoid frequent attention. Such plants largely run them-
selves or, at least, require less operator attention. They
are often the most expensive in capital investment and/or
expansive in areal commitment. Some are unresponsive
to new or unanticipated treatment requirements, and,
once upset, may take the longest to recover.
The performance of an "active" unit process is the oppo-
site of the passive process. It is easier to upset, and to
turn around from an upset condition, than in a passive
process. The simplest active process is a chemical mix
tank receiving some additive. Process "activity" is also
promoted by the provision of some recycle to manipulate
the responsiveness of the reactor. Active processes .allow
field optimization and usually present an opportunity to
derive first and operating cost savings. Some are attrac-
tive for seasonal operation, or when flows and/or loads
32
-------
Table 2-5. Classification of Wastewater Treatment Processes
Processing Objective
Representative Technologies
Process
Format3
Key Design Criteria
• Preliminary Treatment—Solids
separation and conditioning for
easier downstream treatment
• Primary Treatment—Solids
removal for more economical
downstream processing
• Biological Treatment—Solids and
oxygen demanding substrate
removal by oxidation, synthesis,
and conditioning
• Disinfection and
Dechlorination—Control infectious
agents and residual chlorine
toxicant
• Integrated Processes for
Enhanced Biological Nutrient
Control—Use reducing anoxic
processing environments for N
and P removal beyond
background synthesis values
Screening and degritting separation Passive
Add-on Processes for Advanced
Waste Treatment—Removal of P,
N, SS, soluble biologically
resistant organics, and dissolved
inorganic solids from liquid
processing train
Sludge Conditioning—Facilitate
subsequent handling or ultimate
disposal
Thickening and Dewatering
Separator/Concentrators—Facilitate
subsequent processing and/or
ultimate disposal
Sludge Storage—Facilitate
subsequent processing or
ultimate disposal
Sludge Stabilization—Further
conditioning to control a sludge's
putrescible fraction and other
beneficial results
Sedimentation tank separation
Suspended and/or attached growth
reactor
Passive
Passive to
active
Chlorination oxidation and sulfite
reduction, ozonation
UV radiation
See Biological Treatment: multistage
processes can be mainstream,
sidestream, and incorporated
offstream; suspended growth process
mandatory for present
understandings of biologically
enhanced phosphorus removal; high-
level nitrogen elimination may only
be ensured with supplemental
nitrogen-free substrate additions
(e.g., methanol)
Depends on application: most
common
Nitrogen Control—see Biological
Treatment
SS—filtration
P control—Fe and AI salts; lime
Chemical addition
and/or grinding
•Gravity thickening
Beds
Other thickening and dewatering
Tanks or piles
Chemical addition (see above)
Digestion and physical pasteurization13
Composting and other autothermal
processesb
Active
Passive to
in-between
Passive
Active
In-between
Active
Passive
Passive
Passive
In-between
to active
Passive
Passive to
active
In-between
Hydraulics
Hydraulics
Varies per technology:
hydraulics, pollutant
concentration and mass,
solids residence time,
returns, operating solids
levels, character and specific
surface of media, and nature
and variability of'wastewater
Hydraulics, dose, and mixing
Hydraulics and dose
Hydraulics and SS
Dose
Dose and hydraulics
Solids mass
Hydraulics
Hydraulics '
Hydraulics and seasonal
exposure
Hydraulics, solids mass,
flotation air supply and
pressure, and chemicals
Hydraulics and solids
concentration
Hydraulics
Solids mass, biodegradability
and concentration
33
-------
Table 2-5. Classification of Wastewater Treatment Processes (continued)
Processing Objective Representative Technologies
Process
Format3
Key Design Criteria
• Thermal Processing—Facilitate
downstream processing or
disposal by solids conditioning,
water evaporation, and/or
destruction
• Ultimate Disposal—Environmentally
sound reuse and/or disposal
Thermal conditioning (and oxidation)15
Drying
Incineration
Landfill
Beneficial reuse
Passive Hydraulics (and solids
concentration and volatility)
In-between Hydraulics and solids >
concentration
In-between Same as above and volatile
to active solids
Passive Hydraulics, solids mass and
character
Active Same as above and reuse
strategy
* "Passive" process performance is not strongly dependent on operator manipulation. "Active" process performance is strongly dependent on operator
manipulation.
6 Those processes necessarily solubilize pollutants in the course of their reaction and, to varying degrees, create soluble pollutant recycles that
should be anticipated In the liquid processing stream. The impact of these recycles can be especially pronounced If sludge processing is
discontinuous and mainstream processes have low hydraulic detention times and/or depend on, or provide, plug flow reactor configurations.
are expected to change markedly over the effective life
of the project.
A process's activity should not be confused with its reli-
ability. The most demanding (and least reliable) unit proc-
esses are those that have many support systems and
moving parts, and operate under elevated temperatures
and/or pressures. Most often, these processes are found
in the solids processing train of the plant.
Special care is required for the subsequent solids han-
dling processes where large volumes of SS are encoun-
tered; examples are a mainstream activated sludge
system or sludge digester. Wasting solids from the liquid
processing train is the principal means that an operator
has for responding to an effluent that is too high in SS.
Poor settling and/or excessive recycles may dictate the
sudden need for elevated wastage rates of low-SS con-
centration streams from the liquid processing train. Overly
optimistic settled digested solids concentrations and/or
separated supernatant quality design assumptions have
resulted in disastrous consequences at many wastewater
treatment installations.
The designer will be best served by remembering that the
purpose of the mainstream clarifiers is to clarify, not
thicken; separate waste solids thickening should be pro-
vided prior to subsequent processing. Thickening proc-
esses that are most tolerant to widely varying flows and
solids concentrations (e.g., gravity and flotation) may be
preferable to those that are not (centrifugation and belt).
If a gravity thickener is elected, it should be remembered
that overflow SS levels can change rapidly from accept-
able to disastrous if unrealistic storage needs have also
been imposed on the thickener without attention to main-
tenance of adequate liquid levels above the clarified liq-
uid/thickened sludge interface.
The anaerobic or aerobic digestion process contains an
appreciable reservoir of SS. The designer should remem-
ber that the ultimate success of the plant depends on the
ability to remove sludge permanently from the facility. A
single or undersized sludge dewatering unit has a high
probability of failure, as does a residuals management
plan without a firm outlet. The safest design approach is
to assume that the digester is down for cleaning or repairs
or that it may not get its anticipated degree of stabilization
or underflow concentration when sizing the plant's dewa-
tering or downstream systems. A capability to add lime to
the plant's final sludge product ensures greater flexibility
and confidence in responding to newly promulgated
sludge management regulations and their pathogen, vec-
tor,, and nuisance control requirements (4,5).
2.3.2 Nitrogen Control Technologies
Table 2-6 summarizes the status of the nitrogen control
alternatives in municipal wastewater applications after
some 20 years of practice in the United States. The first
efforts saw an equal interest in biological and physi-
cal/chemical approaches. The fundamental difference be-
tween the two is that the former can be designed for either
nitrogen oxidation or removal, whereas the latter only
provides for nitrogen removal. More subtle differences
between the two are found with biological processes of-
fering the conversion of biodegradable organic nitrogen
to ammonium, while the physical/chemical approaches
leave the dissolved organic nitrogen essentially un-
touched.
The original biological design approaches were conser-
vative because of uncertain kinetics. These early designs
emphasized the better understood suspended growth
technologies and often elected isolated cultures (staged
activated sludge systems) for the specific processing ob-
jectives of carbon oxidation, nitrogen oxidation, and nitro-
gen reduction. The physical/chemical approaches ;were
34
-------
Table 2-6. 1990 Status of Nitrogen Control Technologies in Municipal Wastewater Treatment Applications
Knowledge3
High
Technology
Low
Well Limited
Demon- Appli- Found Emerg-
strated cationb Lacking0 ingd
BIOLOGICAL TREATMENT
Higher Technology, Mechanical Plant Approach (see Chapters 6 through 8)
Suspended Growth
Single Sludge
Multiphased
Aerator and/or Aeration Basin Cycling O,R
Sequential Batch Reactor O R
Multistaged (e.g., serial application of processes) O,R
Multizone (e.g., ditches) O,R
Two Sludge O,R
Three Sludge O,R
Attached Growth, Single- or Multiphased, and/or Staged Applications
Submerged Media
Fluidized Bed < O,R
Packed Bed
Downflow R R O
Upflow O,R
Nonsubmerged Media
Stationary (e.g., trickling filter) O O R
Rotating (e.g., rotating biological contactor) O O R
Combination Processes
Any of the Above in Serial Application O,R O,R , O,R
Submerged Stationary Media (Vertical Plates or Media) O,R
Nonsubmerged
Stationary Media with Solids Recycle O O,R
(e.g., Activated Biofilter)
Rotating Media in Solids Suspension O,R O,R
Specific Surface Additives to Suspended Growth System
Concurrent Additive Management (e.g., powdered activated carbon) O O,R
Separate Additive Separation, Processing, and Return . O,R
(e.g., Linpor, Captor)
Lower Technology, Transitional and Natural Systems Approach
(see Section 2.4 for more detail)
Transitional
Aerated Lagoons (suspended growth) O O R
Intermittent and/or Recirculating Sand Filtration (attached growth) O O,R O,R
Aquatic-Based
Lagoons (suspended growth) O,R O,R
Facultative (N stripping) . R
Algae Harvesting (N removal by stripping & synthesis)8 R
Natural and Constructed Wetlands (attached growth with
N removal by synthesis)6
35
-------
Tablo 2-6. 1990 Status of Nitrogen Control Technologies in Municipal Wastewater Treatment Applications (continued)
Knowledge3
High
Low
Well Limited
Demon- Appli- Found Emerg-
strated cationb Lacking0 ingd
BIOLOGICAL TREATMENT (continued)
Surface Flow (floating and rooted aquatic plants)
Subsurface Flow (with rooted aquatic plants)
Land-Based (attached growth treatment with and without
N removal by synthesis)0
Slow Rate Infiltration
Rapid Rate Infiltration
Overland Flow
Subsurface Infiltration
R
R
0,R
0,R
O
O
O,R°
O,Re
R
R
PHYSICAL/CHEMICAL TREATMENT (see Section 2.5 for more detail)
Ion Specific
Ammonia Stripping (NH3-N) '
Ion Exchange (NHJ-N) (NOg-N)
Breakpoint Chlorination (NHJ-N)
Non-Ion Specific
Reverse Osmosis
R
R
R
AVOIDANCE OF NITROGEN CONTROL TECHNOLOGIES THROUGH
BENEFICIAL REUSE (not relevant to this manual)
Irrigation
Selected Industrial Reuse (e.g., cooling water)
R
* O * Nitrogen oxidation; R = Nitrogen removal by biological denitrification unless otherwise noted. Classification can vary depending on particular
application.
b Knowledge of performance capabilities is high but process has been used only on a limited basis.
0 Knowledge of performance capabilities is high but process capabilities or economics have been found to be poor, based on limited application.
d Knowledge of process performance capabilities is low because of infrequent or recently emerging application.
•All systems that rely on synthesis for nitrogen removal ultimately must plan for harvest and disposal of the resultant biomass (10-20 times the
synthesized nitrogen is a likely rule of thumb).
attractive to some early designers because of low energy,
fear of industrial wastes toxicity, and the determination
that the reaction kinetics should be more certain.
Practice soon revealed that the physical/chemical proc-
esses had a multitude of problems in full-scale applica-
tion. Some problems were associated with the absence
of knowledge about how to design successfully the physi-
cal features of the desired works. Other problems were
associated with side reactions and precipitates with the
wastewater's dissolved and paniculate pollutants and the
unforgiving nature of rapidly occurring physical/chemical
reactions. The physical/chemical processes soon fell into
disfavor because of high operating costs, the need for
highly skilled operation, and frequent and expensive
maintenance, coupled with greater understanding, accep-
tance, and demonstration of the potential biological ap-
proaches.
Today, for all practical purposes, physical/chemical nitro-
gen control approaches have been abandoned for
municipal wastewater treatment except for polishing ap-
plications, where further nitrogen control may be a par-
ticular design objective (leading to the selection of ion
exchange) or is a natural result of the use of a non-
ion-specific technology (e.g., reverse osmosis).
The application of attached growth technologies for nitro-
gen control soon followed the suspended growth sys-
tems. Concurrently, and progressively thereafter, greater
understanding of the suspended growth systems and the
economic issues associated with culture isolation (sedi-
mentation tanks to serve each culture) led to the integra-
36
-------
tion of the various cultures and processing objectives into
single-sludge systems. Today, single-sludge nitrogen con-
trol systems are found in a variety of multipurpose appli-
cations. As in the past, the evolution and application of
attached growth and combination processes for similar
objectives continue to lag the suspended growth systems.
Familiarity with a technology leads to a greater under-
standing of its proper application. Over the last decade,
the field has begun to recognize that election of a higher
technology, mechanical plant approach may be a poor
choice for smaller, land-rich communities when clima-
tological conditions will allow a lower technology, transi-
tional or natural system approach. Unfortunately, little is
known about the variability of performance in terms of the
worst month and week per year compliance intervals that
are encountered in the permitted effluent standards. It is
reasonable to expect that performance variabilities be-
come appreciably higher as greater reliance is placed on
the biota and media encountered in a natural or man-
made setting that is highly subject to the whims of nature.
This consideration leads to the inevitable concern that the
risk of noncompliance may be a problem with acceptance
of the lower technologies. Additional data are needed to
confirm or mitigate this concern. However, even if true,
an occasional excursion is far more acceptable than
chronic noncompliance through the use of inappropriately
complex technologies that cannot be properly operated
and maintained under local circumstances.
2.3.3 Processes of Principal Focus for This Manual
The foregoing perspective explains the categorical deter-
mination of the knowledge found with the nitrogen control
technologies (Table 2-6) and leads to the focus of this
manual. The categorization varies for some of the treat-
ment systems depending on the particular application un-
der consideration. The processes of principal interest to
this manual are those that are well demonstrated and that
have proven successful with the widest possible applica-
tion in municipal wastewater treatment. These technolo-
gies almost exclusively lie in the higher technology,
mechanical plant approaches and presently emphasize
suspended growth more so than attached growth systems
because of natural evolutionary process development.
Table 2-6 indicates that none of the higher technology,
mechanical plant approaches have been found to be lack-
ing, with the possible exception of powdered activated
carbon (PAC) addition. Here, the problem is only with the
wet oxidation regeneration step needed for economical
application of the technology (PAC regeneration and
reuse). To date, separation of the regenerated activated
carbon fully from the naturally occurring background
wastewater inerts and the associated economic issues
have served to limit the routine acceptance of this
technology.
As shown in Table 2-6, questions regarding the universal
applicability of some of the attached growth systems are
encountered because of the uncertainty of the technology
in submerged, packed bed and nonsubmerged configu-
rations when high-level nitrogen removal requirements
are imposed in colder climates or low-level effluent
ammonium-nitrogen requirements are encountered. This
is attributable to their inherent process inflexibility (pas-
sivity) and the plug flow nature of their performance. Sec-
tion 2.6 provides greater fundamental understanding of
these constraints, as do Chapters 6 and 7.
2.3.4 Approaches That Receive Limited or No
Attention in This Manual
The nitrogen control approaches receiving limited or no
attention in this manual are the concepts embodied in
lower technology, transitional and natural systems most
applicable to small communities; physical/chemical nitro-
gen control strategies; and the beneficial reuse waste-
water management practices that allow avoidance of
nitrogen control systems altogether. The following para-
graphs provide additional understanding of these
concepts.
The lower technology approaches are at the same or
lower point in understanding than the higher technology,
mechanical biological treatment plants and physi-
cal/chemical nitrogen control processes were 20 to 30
years ago. Today, they can be considered only an emerg-
ing technology. Broadly based acceptance and under-
standing of technology limits will come with further
investigative knowledge. The only certainty is that they
are highly attractive conceptual alternatives for rural
and/or land-rich communities that should work, in some
fashion and to some performance level, under the appli-
cable constraints imposed by the natural and/or struc-
tured changes to the environment. Section 2.4 provides
additional detail regarding the lower technology, transi-
tional and natural system approaches to nitrogen control.
It is provided more to stimulate, characterize, and offer
alternate references for additional investigation, than to
offer finite design information.
The physical/chemical processes for nitrogen control are
at the opposite end of the spectrum from lower technology
approaches. Although receiving only limited application,
there is enough knowledge to determine that they have
limited or no potential for most municipal applications. Ion
exchange (NHJ-N type) may be lacking in terms of suit-
ability for the primary N-removal system in municipal
wastewater treatment, but may have use as a tertiary
polishing unit (NO^-N type) to meet demanding effluent
requirements. Ammonia stripping may have limited rough-
ing applications, most attractively applied in warm tem-
perature pond settings. Breakpoint chlorination can only
be recommended for backup polishing applications to en-
sure extremely low or nonexistent ammonium residuals.
Reverse osmosis cannot be recommended specifically for
37
-------
nitrogen control, but invariably achieves its removal by
the very nature of the process itself. The physical/chemi-
cal processes are briefly discussed in Section 2.5, more
in the interest of completeness and to point out the prob-
lems of the past in order to avoid their repetition rather
than to recommend their use.
Actual avoidance of treatment, although not relevant to
this manual, can be a more appropriate nitrogen control
alternative for some applications. The owner and de-
signer should explore the opportunities for beneficial
reuse and injection at the beginning of the project rather
than at its end, or not at all. The City of St. Petersburg,
Florida, for example, has successfully incorporated such
strategies (i.e., cooling water reuse and irrigation for rec-
reational areas, with the balance injected) for all of its
2,200-2,400 Us (50-55 mgd) of wastewater, while mini-
mizing the need for further development of raw potable
water supplies and avoiding expensive nutrient removal
wastewater treatment strategies at its four wastewater
treatment plants. Similar approaches are under active
consideration by other Florida communities facing strin-
gent total nitrogen effluent limits (equal to or less than 3
mg/L). These approaches promise to be more frequently
encountered at coastal and/or arid regions of the country.
2.4 Lower Technology, Transitional, and
Natural System Approaches
This chapter opens with the results of a survey of 150
small plants with debilitating problems (Table 2-1). The
number one problem has been poor operator under-
standing and application of process control. The number
one conclusion has been that activated sludge (sus-
pended growth) treatment may not be a good choice for
many small plants and the consequent recommendation
was to give fuller consideration to simpler, more tolerant
treatment processes (e.g., fixed media and natural sys-
tems) that are less dependent on skilled operators. Un-
doubtedly, in some manner, the number one problem was
linked to the second cited problem area, solids handling
and disposal (too high an operating solids concentration
for the elevated flow regimes associated with the number
three problem, infiltration and inflow, or "couldn't" or
"didn't" remove sufficient solids from the plant). Regard-
less, the conclusion is inevitable. When a choice can be
made, beware of the temptations of elaborate technology.
In the preceding section, lower technology approaches
for nitrogen control were identified as emerging technolo-
gies. In areas rich with land, their acceptance has been
hampered by one or more of the following: failure even
to consider (business as usual); institutional reluctance (it
Is more convenient and often more economical to procure
a packaged system that promises to meet needs; sound
design of any kind takes time and uses skilled, expensive
people); zealous single-issue advocates (anything natural
is better, resulting In failure related to inappropriate appli-
cations); imperfect perceptions (not in my backyard); his-
toric practices (the failing septic tank, honey wagons and
farms, the stinking raw sewage lagoon); implementability
(the required acquisition of large tracts of land); technical
uncertainty (design criteria, performance variability,
clearly incomplete reporting of costs and operating is-
sues); and potentially overly restrictive discharge stand-
ards (which impose unacceptable noncompliance risk).
Lower technology approaches are not new, nor are they
uncommon (1). Even today more than 50 million U.S.
residences (25 percent of alf single-family dwellings) have
remained unsewered, mostly served by septic tanks or
soil absorption systems for their wastes. Stabilization
ponds (or lagoons) number well over 5,000. Specifically
designed land treatment systems number over 1,000.
Constructed wetlands number over 100, along with a few
aquatic plant systems. Nor are land-based systems al-
ways small—Orlando, Florida, uses a rapid infiltration
land treatment system (ground-water recharge) with a
2,200-L/s (50-mgd) design capacity.
Serious consideration should be given to lower technol-
ogy, natural system approaches. Advocates of both higher
and lower forms of technologies should be open to each
other. The best approach for a midsize, land-rich munici-
pal system may not be one or the other, but may be a
blend of both.
The purpose of this section is to introduce the reader to
this technology as it applies to nitrogen control applica-
tions. The balance of the section is devoted to providing
additional understanding of the fundamentals and of the
treatment technology found in broadly based generic
classifications. Thereafter, additional understanding is
best derived from the many readily available publications
that consider this technology in a summarial fashion (1),
comprehensively (6,7), and with technology-specific de-
tail (8-12). The Manual: Wastewater Treatment/Disposal
for Small Communities should be consulted for guidance
in any application specifically dealing with a limited popu-
lation base (13).
2.4.1 Fundamentals
Table 2-7 compares both the higher and lower forms of
technology using readily understandable screening crite-
ria: the land required for the process and the ability of the
technology to meet the nitrogen control objectives. The
table also includes three transitional technologies that
effectively blend the mechanical plant with the natural
system approach.
An inspection of Table 2-7 shows the attractiveness of
the higher technology, mechanical plant approach: sig-
nificantly less land required and the ability to somewhat
easily respond to all of the target effluent objectives ex-
cept the most stringent total nitrogen standard. If all of
the lower technology, natural system approaches were
placed in a similar grouping, the only fair conclusion
38
-------
Table 2-7. Lower Technology, Transitional, and Natural System Screening Criteria
Estimated General Applicability of Technology for Various Levels of
Nitrogen Control
Nitrogen Oxidation,
Effluent NHj-N Nitrogen Removal, Effluent Total N
Technology
Land Need Detention Inter- Inter-
acres/mgd ADFa Time mediate High mediate High
(o—io (2.5—5.0 stringent (lu— lo (o— 10 stringent
Siteb Process0 Days mg/L) mg/L) (<2.5 mg/L) mg/L) mg/L) (<5 mg/L)
HIGHER TECHNOLOGY, MECHANICAL PLANT APPROACH,d Suspended and Attached Growth Technologies (provided for reference)
0.5 0.2 +/- 1.0 +/- Easy Easy Easy Easy Easy Possible
I Varies with technology (see Tables 2-6 and 2-8) I
LOWER TECHNOLOGY, TRANSITIONAL APPROACH,6 Suspended and Attached Growth Technologies, Oxygen Supply by Mechanical Means
or Resting, Respectively. (Note: Imhoff or septic tank preferred for pretreatment for primary settleable solids and floatables capture for the latter two
technologies.)
Aerated Lagoons — 3-30 10-100 Possible Possible Doubtful Possible Possible Doubtful
Intermittent Sand Filtration — 2-10 — Easy Possible Doubtful Doubtful Doubtful Doubtful
Recirculating Sand — 8-10 1-2 Easy Possible Doubtful Easy Possible Doubtful
Filtration with Anoxic
Rock Filter
Enhancement
LOWER TECHNOLOGY, NATURAL SYSTEMS APPROACH*9 (Note: Imhoff or septic tank preferred for pretreatment for primary settleable solids
and floatables capture for submerged flow wetlands and subsurface infiltration; lagoons for other land-based systems.)
Aquatic-Based Technologies
Lagoons, Suspended Growth Technology with N Removal by Stripping (small), Nitrification and Denitrification, and/or Synthesis; Oxygen Supply
Largely by Photosynthesis8
Facultative — 4—20 20-100 Easy Possible Possible Possible Possible Doubtful
Algal Harvesting — 10+ 10+ Easy Easy Possible Easy Possible Doubtful
Constructed Wetlands, Attached Growth Technology with N Removal by Synthesis, Oxygen Supply Largely by Photosynthesis*'9
Free Water Surface — 25-30 7-10 Possible Possible Doubtful Possible Possible Doubtful
Submerged Flow — 5-10 1-2 Doubtful Doubtful Doubtful Doubtful Doubtful Doubtful
Land-Based Technologies, Attached Growth Technology with and without Nitrogen Removal by Synthesis, Oxygen Supply by Resting''11
Slow Rate (Moderate) —
Rapid Infiltration (High) —
Overland Flow (Low) —
Subsurface Infiltration
(Mod - High)
50-700 —
3-60 —
15-110 —
8-80 —
Easy
Easy
Easy
Easy
Easy
Possible
Possible
Possible
Possible
Doubtful
Doubtful
Possible
Easy
Possible
Possible
Possible
Easy
Possible
Doubtful
Possible
Possible
Doubtful
Doubtful
Doubtful
a Average daily design flow.
b Site needs are highly variable. Likely minimum listed for higher technology, mechanical plant approach. Rule of thumb for lower technology, natural
system approach would be twice as large as the process requirement, with no site-specific information to the contrary.
c Process needs limited to biological reactor and clarifier (if appropriate).
d Avoid rigid thinking. Best process selection may be a blend of technologies, both coupled and uncoupled, using each for what it does best. Natural
systems are readily adaptable in add-on tertiary, polishing, or seasonal applications.
a All systems that rely on synthesis for nitrogen removal must provide for harvest and disposal of the resultant biomass (10-20 times the synthesized
nitrogen is a likely rule of thumb).
' Readily available information is not clear as to original design criteria (e.g., ac/mgd); generally, existing conditions reported. Additionally, lowest
process acreage may be associated with treatment strategies not necessarily directed to nitrogen control. Often systems have not been in operation
for several years. Algal harvesting with a view toward maximizing N removal approaches has only been demonstrated at prototype installations
in the United States. Some constructed wetlands systems are used in tertiary applications, others provide secondary and tertiary treatment. Often
performance data are limited to active growing season.
9 Inherent anaerobic conditions make subsurface constructed wetland systems attractive for nitrogen removal by denitrification when applied in a
staged manner or on receipt of a nitrified feed stream. Nitrogen removal characterizations, if not already noted, rise to "possible" when utilized in
this manner.
'Typically encountered soil permeabilities are listed parenthetically after each technology. The technologies are generally applied with the following
soil cm/hour permeabilities: Slow Rate = >0.15 - <5, Overland Flow = <0.5, Rapid Infiltration = >5.
39
-------
would be that they too could respond to most of the target
effluent objectives, even for the most stringent total nitro-
gen standard (more stringent standards result in an in-
crease in biomass production, land required, or seasonal
restrictions).
Clearly, the lower technology approaches, whether tran-
sitional or natural, offer a rich mine of possibilities that
have been little explored or optimized to date. With a
regulatory and owner attitude willing to accept the risk of
the unproven, the opportunities for enhancement and re-
finement seem almost endless. Apparently all that is
needed is a greater appreciation and a sound application
of fundamentals. Section 2.6 of this chapter provides dis-
cussion of the fundamentals needed for sound application
of a lower technology approach: the solids residence
time, the carbon-to-nitrogen (C:N) ratio, the differences
between suspended and attached growth reactors and
their responsiveness, an understanding of oxygen trans-
fer and alkalinity/pH inhibition of the obligate aerobic ni-
trifying autotrophs, the need for driving substrate for the
denitrification reaction, and the mass balance, based on
the total oxygen demanding materials (Section 2.9).
To assist in the visualization of community applications of
lower technology nitrogen control systems, Table 2-7 also
includes a fundamental characterization of the suspended
or attached growth type of biological reactor and, because
of the importance of the oxygen supply in nitrification (and
any desired aerobic reaction), the principal source of oxy-
gen for each system. When the technology relies on syn-
thesis (or plant uptake) or some degree of ammonia
stripping for nitrogen removal, this too is listed.
The mass balances for these systems, with the exception
of the intermittent and recirculating sand filters, must be
expanded to include hydrological considerations (includ-
ing the wettest and driest year expected in the life of the
facility) and, when applicable, the biomass produced by
photosynthetic oxygenation (light is the energy source,
temperature influences algal activity) and the resultant
biomass planned for harvest or accumulation. If accumu-
lation (no harvest) is planned, care must be taken that it
does not impair the performance of the system. The wis-
est course of action for the aquatic systems that use the
least amount of land per unit of flow processed, is to plan
on the eventual need for vegetative debris (solids) re-
moval and disposal (a reliable residuals management
scheme).
More detailed inspection of Table 2-7 shows some as-
pects of the mechanical plant approach to be more preva-
lent in lower technology applications. In general, primary
treatment is recommended for improved downstream
process stability, and freedom from nuisance or operating
problems. Undoubtedly, today's failure to accept a lower
technology is due in part to the absence of primary treat-
ment in many of the historic applications. Significantly, the
success of the downstream process is dependent on what
goes on before it, and no system is completely free of
the need for regularly scheduled maintenance and knowl-
edgeable operation.
True natural systems always bring with them valid issues
of transient and resident vectors, and their control needs.
Control of the human population (via site access barriers)
may also be perceived as important and, depending on
the technology, it may be. However, before the need for
isolation from the human population is accepted as a
foregone conclusion, serious consideration should be
given to passive recreational activities (e.g., trail use and
bird watching) to increase local acceptance and value.
All natural systems that do not make use of artificial bar-
riers (liners) must be evaluated and monitored for ground-
water impacts. Ground-water discharge permits are now
required in several states and are inevitable in all others.
Up- and downgradient ground-water monitoring :wells
must be included in the design.
Finally, the intrinsic values and liabilities of the lower tech-
nology approach beyond the obvious operational and me-
chanical simplicity and land requirement must be
considered. The intrinsic value is the tolerance of time
(mitigation of processing flow and pollutant peaks in the
applied wastewater and daily attentive operation and
maintenance) and dilution (through the commitment of
expansive soil and/or liquid volumes for beneficial treat-
ment). The intrinsic liability is the greater unpredictability
of relying on some harmony of the processes found in
the natural environment with the often unpredictable cli-
matological conditions and media change with time. Ap-
proaches to overcoming the liabilities embody the same
principles often used in mechanical plants: nonaggres-
sive design, fuller use of recycles or returns, load-splitting
capabilities, and low cost backup strategies (flexibility)
whenever possible. The remaining subsections address
the lower technologies in additional detail.
2.4.2 Transitional Approach
2.4.2.1 Aerated Lagoons
Aerated lagoons have potential in nitrogen oxidation
applications. Performance is limited by deposition of vi-
able organisms. Shorter detention times will reduce the
impact of suspended algae. The 7- or 10-day detention
aerated lagoon is normally not well enough mixed to sus-
tain the nitrifiers, and Parker reports finding no aerated
lagoons with a significant amount of nitrification. In the
absence of adequate mixing, the nitrifiers must reside on
a surface and there is not enough surface to sustain
them (14).
If sufficient mixing was provided to keep the viable or-
ganisms in suspension, use of aeration systems and/or
aeration basin configurations that provide mixing inde-
pendently of the oxygen supply could provide detention
times that should closely approach the same solids resi-
40
-------
dence time for nitrification as found in the mechanical
plants. Although back-mixing undoubtedly occurs, recycle
of the aerated effluent would allow a continuous reseed
of an acclimated nitrifying culture and may achieve a
measure of natural denitrification. When settling is de-
sired, depression traps can be used. If desired, floating
sludge collectors can be used to remove the deposited
solids. Designers should consider some form of aerated
lagoon for intermediate to high nitrification applications in
land-rich situations.
2.4.2.2 Intermittent and Recirculating Sand Filtration
Once totaling nearly 500 U.S. installations, the once-
through intermittent sand filter was the process of choice
for the highest known performance for wastewater treat-
ment plants into the 1950s. It is commonly described in
great detail in all regulatory guidelines of that vintage. The
1957 edition of MOP 8 describes the effluent quality as
so sparkling clear and odorless that remote discharges
had to be posted to warn people against use for drinking
(15). Biochemical oxygen demand (BOD) and SS were
reportedly routinely below 5 mg/L and, by inference from
the reported nitrate data for the technology, the effluent
ammonium nitrogen was likely to be routinely less than
10 mg/L.
Recirculating sand filtration was the predecessor of the
classical contact bed and trickling filter technologies (the
latter adopted out of lack of concern for ammonium nitro-
gen). The intermittent sand filter predates the rapid infil-
tration basin of today's land treatment strategies. Similar
to a rapid infiltration basin, it was operated in a cyclic
manner, with one bed operating, one bed resting, and one
bed in cleaning (raking); often a fourth bed was main-
tained for emergency use under high-flow conditions.
Consisting of 1 m (3 ft) of graded sand and 30 cm (12
in) or less of supportive gravel, the intermittent sand filter
was used in raw, secondary, and tertiary applications.
Most often it was preceded by an Imhoff tank (still a good
decision in order to avoid daily or more frequent removals
of primary sludge); operating problems and bed cleaning
were clearly minimized with this operation. There is noth-
ing fundamentally wrong with this technology, and it
should be rediscovered and considered for intermediate
to high nitrogen oxidation applications.
Recirculation provides a means of introducing oxygen-
ated water, nitrate nitrogen, and seed organisms to the
applied wastewater. Recirculation rates depend on the
objectives and the quality of the wastewater applied to
the system. With the recycle of nitrate nitrogen, higher
nitrogen removals than achieved with the nonrecirculated
system will occur.
The full denitrification potential of the recirculating sand
filter system is achieved with exactly the same concepts
as typically reported for the higher technology ap-
proaches (i.e., the use of a front-end anoxic reactor with
nitrified effluent recirculation ratios [R/Q] of 3 to 5, and/or
an anoxic reactor receiving the nitrified effluent along with
a septic or Imhoff effluent as the internal carbon source
to drive the denitrification reaction). In these configura-
tions, the anoxic reactors are typically upflow rock
(2-5 cm [1-2 in] in diameter) filters. Reported anoxic re-
actor empty bed contact times are unchanged for pre-
and post-sand filter applications and, based on Q, are 4
to 72 hr (13). The range reflects differences in seasonal
site-specific temperatures, applied wastewater and per-
formance needs (presently only described for septic tank
effluents), designer certainty, and the need for a highly
optimized system. Similarly, the reported physical and
operating needs of the sand filter are unchanged in these
denitrification applications.
2.4.3 Natural System Approach: Aquatic-Based
Technologies
2.4.3.1 Facultative Lagoons and Algal Harvesting
Simple regression-type ammonium and nitrogen removal
models of facultative lagoons, have been developed and
reported with some suggestion of validation (16,17).
These identify pH to be of primary importance, based on
an ammonia stripping assumption. ApH rise occurs in the
pond because carbon dioxide (CO2) is the carbon source
for the algae, which photosynthetically produce biomass
and oxygen. The CO2 source is largely from the aerobic
(surface layers) and anaerobic stabilization (bottom lay-
ers and deposits) in the lagoon. With insufficient CO2, the
bicarbonate alkalinity will serve as the CO2 source, and
a significant pH rise can be experienced. Significant am-
monia stripping does occurs at a pH of greater than 8.5
(see Section 2.5.2.1 for further understanding).
The reported dependency of ammonia removal on pH
could also be a surrogate parameter for an active algal
biomass, and the actual ammonium and total nitrogen
removals could reflect natural nitrification (using the pho-
tosynthetically produced oxygen), denitrification (bacterial
use of the dormant algal biomass as the driving substrate
during the nighttime hours), and algal synthesis during
the daylight hours.
Facultative ponds should be designed to embrace and
enhance the anaerobic reactions that produce CO2 and,
most important, methane (CH4), occurring in the bottom
of the pond. Failure to do so will likely result in problems
and, inevitably, the progressive buildup of solids and
pass-through to the plant effluent. Many past problems
with this technology were associated with this considera-
tion. The designer would be well served by consulting the
more fundamental publications regarding this technology
(18,19).
Facultative ponds have the potential to achieve nitrogen
oxidation down to the most stringent levels; their natural
daytime to nighttime cycling of photosynthetic activity and
aerobic to anoxic bacterial response provides a possible
41
-------
mechanism of nitrogen removal. Their liability: what to
do with the algal biomass once generated. Procedures
start with submerged drawoff outlet designs and consid-
eration of chemical coagulation and/or filtration for tertiary
algae removal (11). Regulatory standards may allow for
a higher effluent SS. Pumped or submerged outlet re-
moval and the sloped sidewalls of the lagoon allow for
considerable flow equalization.
Facilities with an algal harvest approach (maximizing ni-
trogen removal by synthesis) can be designed to incor-
porate a number of concepts. The large lagoons at
Sunnyvale and Stockton, California, return the sub-
sequently removed algae to lagoons with adequate depth
to ensure anaerobic activity. The systems have operated
since the late seventies with no sludge removal. Alterna-
tively, the pond design could be as shallow as is reason-
able and well mixed, with the objective of maximizing light
penetration. Algae-removal concepts abound (6,7,11) but
are often unused on a sustained basis because of the
uncertainty (and now a liability) concerning use or dis-
posal of the harvested algae.
2.4.3.2 Constructed Wetlands
Constructed wetlands are classified as a function of water
flow: surface (also termed free water surface, FWS) and
subsurface. When simply expressed, constructed wetland
treatment technology makes an artificial receiving water
and its vegetation part of the treatment process. In com-
parison to algae, the higher forms of plant life—floating
(duckweed, water hyacinths), submerged, and emergent
(cattails, rushes, and reeds)—perform less efficiently per
unit weight of biomass.
FWS constructed wetland treatment conceptually relies
on attached growth bacterial performance, receiving oxy-
gen from the evapotranspiration response of the aquatic
vegetation. Practically, the dominant bacterial action is
anaerobic. The ammonium and nitrogen removal mecha-
nisms are a combination of aerobic oxidation, particulate
removal, and synthesis of new plant protoplasm.
An FWS wetland is nothing more than a lagoon, except
that a far greater expanse is needed to maximize the
productivity per unit area. In practice, very large systems
may achieve significant, if not complete, nitrogen oxida-
tion, with surface reaeration contributing to the oxygen
supply. Some nitrification and denitrification undoubtedly
occurs in all systems.
If it is assumed that the wetland vegetation will not be
harvested, as is the case with natural wetland systems,
its capacity for nitrogen control is finite, reflecting the
site-specific vegetation and the ability to expand in the
available space. Thus, the bigger the natural wetland that
is called part of the process, the better, since there is
dilution of the wastewater to the point that it is no longer
significant in comparison to the naturally occurring back-
ground flow and water quality.
Constructed FWS wetlands yield a managed vegetative
habitat that becomes an aquaculture system. Examina-
tion of the evolution of this technology shows the emer-
gence of concepts that include organic load distribution
or artificial aeration to avoid aesthetic nuisances, and
emphasis on plants that grow the fastest (20). Duckweed
and water hyacinth systems (classified as "aquaculture")
have been reported to achieve long-term total nitrogen
residuals of less than 10 mg/L and may be manageable,
with harvesting and sensitive operation, to values of less
than 3 mg/L on a seasonal, if not sustained, basis (20).
Submerged-flow constructed wetlands are simply
horizontal-flow gravel filters with the added component of
emergent plants within the media. They have been clas-
sically used for BOD removal following sedimentation
and/or additional BOD and SS removal from lagoon ef-
fluents as with FWS approaches. This technology has the
potential for high-level denitrification when a nitrified
wastewater is applied; the naturally occurring environ-
ment promotes anoxic (denitrification) pathways for oxi-
dized nitrogen elimination. Unfortunately, application of
this approach to nitrogen removal is only in the research
stage (13).
Ultimately, the success or failure of the wetland approach
for nitrogen control may rest with the harvest of the vege-
tation (how to remove it and what to do with it), the need
for backup (so that areas under harvest have the backup
of areas in active growth), and often natural seasonal
growth and decay cycles. There are no good answers,
and the managed protoplasm production systems using
constructed wetlands or algal production lagoon systems
are the only pure, lower technology, aquatic-based ap-
proaches that have the potential to achieve significant
levels of ammonium and total nitrogen control on a sea-
sonal and/or year-round basis.
If biomass production is an unacceptable goal, the de-
signer should think of a more tolerant mixed vegetation
system that minimizes the need to harvest the accumu-
lated vegetation and maximizes the promotion of concur-
rent or staged nitrification and denitrification in some
fashion. Conceptually, the optimization has to begin with
promotion of nitrogen oxidation systems that may be shal-
low (better aeration for attached and suspended bacterial
growth) with vegetation that minimizes light penetration
and avoids as much algal growth as possible. Cyclic stag-
ing, recycle, forced aeration, and/or mixing represent
some of the enhancements that naturally follow.
2.4.4 Natural System Approach: Land-Based
Technologies
The land-based technologies, although not originally de-
veloped for nitrogen control, have been in use since the
beginning of civilization. Their greater value may be the
use of the wastewater for beneficial return (agricultural
and recharge) in water-poor areas, rather than mere ni-
42
-------
trogen control benefits. If nitrogen control benefits are
desired, some key issues arise concerning the type of
plant crop with its growing and harvesting needs and/or
the cycling of the water application and restorative oxy-
genation resting periods. Native soils and climate add the
remaining variables.
Generally, the wastewater applications are cyclic in land-
based technologies, making some form of storage or land
rotation mandatory to ensure the restorative oxygenation
derived from the resting period. Surface wastewater ap-
plications allow additional beneficial soil aeration (plow-
ing, tilling, and raking), which can become mandatory for
the heavily loaded systems after an elapsed season, or
number of loading cycles. Actual surface cleaning pro-
grams, to remove the plastic, rubber, and other debris
found in pretreated municipal wastewaters, also may be
necessary, although not at the frequency used for bene-
ficial soil aeration.
The following paragraphs provide additional conceptual
information on the four most common land-based tech-
nologies. Subsurface, slow, and rapid infiltration systems
do not discharge to surface waters and conceptually may
allow a more relaxed nitrogen control standard, depend-
ing on local ground-water regulations.
2.4.4.1 Subsurface Infiltration
Subsurface infiltration systems are capable of producing
a high degree of treatment; with proper design, they can
provide a nitrified effluent, and denitrification can be
achieved under certain circumstances. Keys to their suc-
cess are the adequacy of the initial gravel infiltration zone
for solids capture and the following unsaturated zone of
native or foreign soils.
Failure to provide an oxygenated environment by either
resting or conservative loadings can lead to failure. De-
nitrification under gravity loading is likely to be small, but
may be improved through pressure/gravity dosing con-
cepts of liquid application to the trenches.
Subsurface infiltration wastewater management practices
are embodied in the horizontal leach fields that routinely
serve the more than 20 million septic tanks of individual
nonsewered establishments and homeowners. In recent
years, they have also been advanced for collective serv-
ice in small isolated communities.
2.4.4.2 Slow Rate Infiltration
Slow rate land treatment represents the predominant mu-
nicipal land treatment practice in the United States. The
wastewater is applied to a vegetative land surface using
sprinkler or surface irrigation techniques and percolates
(after losses for evapotranspiration) through the plant/soil
matrix. This technology is capable of achieving the high-
est degree of nitrogen removal.
Typically, assumptions for nitrogen losses due to denitri-
fication (15-25 percent), ammonia volatilization (0-10
percent) and soil immobilization (0-25 percent) supple-
ment the primary nitrogen removal mechanism by the
crop. The balance of the nitrogen passes to the percolate.
Typical design standards require preservation of control-
ling depths to ground water and establishing nitrogen
limits in either the percolate or ground water as it leaves
the property site. Nitrogen loading to the ground water is
often the controlling consideration in the design.
2.4.4.3 Rapid Rate Infiltration
Rapid rate infiltration systems require relatively perme-
able, sandy to loamy soils. Vegetation is typically not used
for nitrogen control purposes but may have value for
stabilization and maintenance of percolation rates. The
application of algae-laden wastewater to rapid infiltration
systems is not recommended because of clogging con-
siderations but could be considered with attendant addi-
tional tolerance for surface maintenance, drying and soil
aeration needs.
Designs can be developed that foster only nitrification, or
nitrification and denitrification. Nitrification is promoted by
low hydraulic loadings and short application periods (1 to
2 days) followed by long drying periods (10 to 16 days).
Denitrification can vary from 0 to 80 percent. For signify
cant denitrification, the application period must be long
enough to ensure depletion of the soil (and nitrate nitro-
gen) oxygen. Higher denitrification values predictably
track higher BOD:nitrogen ratios. Enhancement may be
promoted by recycling or by adding an external driving
substrate (methanol). Nitrogen elimination strategies also
may reduce the drying period by about half to yield lower
overall nitrogen residuals with higher ammonium-nitrogen
concentrations.
2.4.4.4 Overland Flow
Overland flow involves the application of wastewater to
the upper section of a gentle, sloping grassland. The thin
film of applied wastewater is then collected in runoff
ditches for subsequent discharge or further processing. It
is designed to be used for relatively impermeable soils or
subsoils to avoid infiltration to the ground water.
Little attempt has been made to design optimized over-
land flow systems with a specific objective of nitrogen
control. Their performance depends on the same funda-
mental issues: nitrification-denitrification, ammonia vola-
tilization, and harvesting of crops. When measured,
overland flow systems designed for secondary treatment
often reveal less than 10 mg/L total nitrogen.
43
-------
2.5 Physical/Chemical Treatment
Processes
Several physical/chemical nitrogen control treatment
processes have been advanced and tried in municipal
wastewater treatment applications. Only two remain in
routine service. Physical/chemical treatment, except in
highly specialized situations, is the process of last resort,
especially at small plants.
The reader interested in more detail than provided here
is referred to earlier editions of the MOP 8 (21) and of
this manual (22). The failures and the reasons for the
failures are not well reported. However, these failures
stand as vivid testimony to the risk that is involved in
out-of-context extrapolation, perhaps superficial interpre-
tation of the results from limited-duration laboratory, pilot,
and demonstration studies, or the failure to fully appreci-
ate the attendant routine demands of normal operation
and maintenance with an emerging technology. Whatever
the reasons, the proof of any concept lies in its applied
practice with the people that use it on a daily basis.
The purpose of this section is to give a token introduction
to the nitrogen control physical/chemical treatment
processes in the hope of providing a theoretical under-
standing, along with a listing of some of the issues en-
countered with their application. Some processes remain
applicable in practice or in concept; these may be the
processes of choice in some municipal applications. How-
ever, care is strongly suggested, except in polishing
(breakpoint chlorination or ion exchange) or roughing
(ammonia stripping) applications.
2.5.1 Fundamentals
When working with physical/chemical treatment technolo-
gies, it is important to remember that inorganic ions may
be added and background organic and/or inorganic spe-
cies may enter into the reaction. What happens may or
may not be pH dependent; some additives and their re-
action may influence the pH. Often, unanticipated severe
or not so severe side reactions occur, which only seem
obvious with "after the fact" analyses.
Figure 2-1 was prepared to illustrate the pH dependency
of several reactions of interest. Some reactions have al-
ready been discussed (e.g., the dependency of photosyn-
thesis on CO2 extraction from the wastewater and its
effect on pH with the facultative lagoon concepts dis-
cussed in Section 2.4.3.1).
The information presented in Figure 2-1 also includes the
pH dependency of hydrogen sulfide (H2S) dissassocia-
tion. As shown, once the pH exceeds 8.3, for all practical
purposes it is impossible to have any odors associated
with hydrogen sulfide. The pH of 8.3 is of particular inter-
est since it is also the point of equilibrium with the CO2
in the atmosphere. Given sufficient alkalinity, all liquids,
when stripped (or aerated) in any fashion, will approach
this value. If the pH is below this value, CO2 will be
removed. If the pH is greater than this value, CO2 will be
added. The overall alkalinity does not change, only its
species distribution.
Figure 2-1 will be used, as appropriate, in the subsequent
discussions of physical/chemical treatment technologies.
The source of the CO2, carbonate, bicarbonate, and hy-
droxide relationships are described in a widely available
textbook (23). The lime dose and soluble calcium and
magnesium distributions in Figure 2-1 were derived for a
hypothetical wastewater by use of the theoretical predic-
tions found with Caldwell-Lawrence diagrams (24).
2.5.2 Ion-Specific Technologies
The ion-specific technologies work only on ammonium or
ammonia nitrogen. They do not work on organic nitrogen
or nitrate and nitrite nitrogen. Raw wastewaters need
sufficient bacterial treatment to convert the organic nitro-
gen to ammonium, but not to a level that would result in
nitrite or nitrate production.
2.5.2.1 Ammonia Stripping (NH3-N)
Ammonia stripping was the conceptual add-on process
of choice with lime treatment for phosphorus removal in
the early years of physical/chemical nitrogen control tech-
nology. Air stripping consists of raising the pH to within
the range of 10.5 to 11.5 (see Figure 2-1), a point that
achieves very low phosphorus residuals, and of providing
sufficient air to strip out the ammonia. The lime (or hy-
droxide) dosage is very alkalinity dependent (25).
The amount of air needed for stripping can be predicted
by the equilibrium relationships expressed with Henry's
Law constant. Design airflow rates are typically twice the
theoretical prediction. Efficiencies decrease with decreas-
ing temperature.
In the early 1970s, the newly installed South Lake Tahoe
Advanced Wastewater Treatment Facility reported ammo-
nia nitrogen removals with its 160 L/s (3.75 mgd), 7.3 m
(24 ft) high, packed (redwood media) tower of about 90
percent (effluent NHJ-N of 1.4 mg/L). Initial pH values
were 10.7; hydraulic loadings were 70 m3/m2/d (1,200
gpd/sq ft); the air-to-water volume ratio was about 100:1;
and air and water temperatures were about 4 and 12°C
(39 and 54°F), respectively (26). The cold weather am-
monia-nitrogen removal efficiency declined to 60 percent
at an air-to-water ratio of 50:1. Warmer temperature, sum-
mer performance was in excess of 90-percent ammonia-
nitrogen removal for these loading conditions and a 50:1
air to water ratio.
Ninety-percent ammonia-nitrogen removal at South Ta-
hoe in winter was accompanied by a 8°C wastewater
temperature decline through the tower. Cooling towers
typically operate with an air-to-liquid volume ratio of 10:1
to 30:1. Thus, to avoid ice formation, operations of strip-
ping towers are effectively limited to wastewater tempera-
44
-------
pH
9
10
11
12
100
S.
100
Dissassociation
100
50
S.
HC03-
100
50
pH, CC>2, CO0, HCOg, and OH" Relationships
400
200
I I
Initial Alkalinity = 200 Mg/L as CaCOg
400
200
o
8
CO
g
PH
Figure 2-1. pH dependency of selected reactions of interest.
tures of about 10°C (50°F) and above. More important,
the 90-percent ammonia-nitrogen removal in winter op-
eration was also accompanied by a 0.4 pH unit decline;
this led to the plate-out of CaCO3 scale, causing serious
difficulties.
A survey in the early 1980s of the only two constructed
towers found that South Tahoe tried a variety of other
packing (e.g., plastic mesh and plastic tubing) and
cleaning procedures in the intervening years in an attempt
to counter the scaling problems (27). None was success-
ful, all were badly scaled, and, at best the towers could
achieve 40-percent ammonia removal (scale formation
deteriorated performance). The towers were torn down
after seven years, with the conclusion that experience
10
11
12
with no packing, or stripping in ponds, could do just as
"good" with much less complexity. Other noted problems
were misting (and resultant CaCO3 deposits) and, with
cold weather, ice formation for approximately 60 m (200
ft) around the tower. Other than ice formation, similar
findings were encountered with the full-scale stripping
tower at Water Factory 21 in Orange County, California.
CaCO3 (calcium carbonate) scale formation problems also
continue to plague the downstream pH adjustment step.
The place for ammonia stripping appears to be in warm-
climate ponds. It also is used for leachate pretreatment
prior to discharge to a POTW. A recent publication reports
50-percent ammonia-nitrogen removal in unaerated
ponds in five days with a pH of 10.5, and the same
45
-------
removal with aerated ponds in about a half day (28).
Likely air-to-liquid volume ratios are on the order of 50 to
1. Collection of the treated water and delivery to a follow-
ing pond, if possible, will achieve the best performance
with the least amount of air usage. Surface sprays or
mechanical aeration are also applicable.
2.5.2.2 Ion Exchange (NH$-N)
Ion exchange technology involves passing a liquid
through a column or bed of specific natural or synthetic
zeolite resin and the exchange of one ion for another. The
column is run until unacceptable breakthrough of the
ion(s) of concern is achieved (reflecting the exhaustion
of the exchange sites for this point of equilibrium). A highly
concentrated regenerant is then passed through the col-
umn to displace the removed ions from the exchange
sites. The regenerant can be processed further for recov-
ery and reuse with a blowdown of a concentrated waste
or recyclable product, or passed to waste.
Clinoptilolite, a natural occurring product, is the zeolite of
choice for ammonium-nitrogen removal applications. Its
order of major ion selectivity is reported as (29):
K*
NH|
Na+
Total exchange capacities are reported around 2 mil-
liequivalents/L (meq/L), but the capacity in wastewater
applications for ammonium appears to range from 0.2 to
0.5 meq/L. Clinoptilolite has a specific gravity of 1 .6 and
a bulk density of about 0.75.
In wastewater treatment applications, prior filtration must
be used to avoid blinding the fine 20 x 50 (0.84 x 0.33
mrn) mesh media. Hydraulic loading rates range from 5
to 10 bed volumes per hour. On startup, the first 5 to 10
volumes will generally show the dragout of the ammonium
left in the column after regeneration. Thereafter, the
effluent ammonium nitrogen concentrations slowly in-
crease from about 0.5 to 1.0 mg/L until about 130 bed
volumes is reached and then progressively deteriorate to
about 5 mg/L at 170 bed volumes. The column may be
removed from service after final effluent quality becomes
unacceptable.
Regeneration is preceded by two to three bed volumes
of backwash to flush the accumulated solids out, and is
accomplished by a high pH salt solution using flow rates
of about 10 bed volumes per hour and 10 to 20 bed
volumes of regenerant. Regeneration is followed by a
rinse cycle of 2 to 3 bed volumes to minimize high pH as
well as salt and ammonia concentrations in the product
effluent on startup.
The two wastewater treatment plants that have Clinop-
tilolite ion exchange use a closed-loop ammonia stripping
process to recover the regenerant. The off-gases are ab-
sorbed In a sulfuric acid solution to form ammonium sul-
fate, which has some fertilizer value. In this case, the pH
is further adjusted to greater than 11 by caustic, and the
resultant Mg (OH)2 is allowed to settle out before entering
the ammonia stripping cycle. Makeup salt is added to the
regenerant after the ammonia is removed.
The 220-L/s (5 mgd) North Tahoe-Truckee Wastewater
Treatment Plant has routinely used its Clinoptilolite system
since the late seventies. Presently, the plant's average
daily flow rate is about 70 percent of its rated capacity.
The 660-L/s (15 mgd) Upper Occoquan Treatment Facility
in Virginia, constructed before the North Tahoe-Truckee
plant became operational, found operation of its Clinop-
tilolite system to be unnecessary because effluent stand-
ards were revised before it began operation. The Upper
Occoquan installation was briefly operated for startup and
operability checks in the early eighties, found to achieve
its 2 mg/L ammonium-nitrogen target, and placed in its
present reserve status.
The North Tahoe-Truckee facility historically achieved
about 2.5 mg/L of ammonium nitrogen in its ion exchange
effluent (because of regenerant dragout and end-of-run
deterioration), which was further reduced to about 0.3
mg/L by breakpoint chlorination. The plant's total effluent
nitrogen standard was then 2 mg/L. Together, the ion
exchange and breakpoint chlorination processes add
about 140 mg/L of chloride and 250 mg/L of dissolved
solids to the plant's effluent.
Most recently, with relaxation of its treatment needs due
to a revised ground-water discharge, North Tahoe-
Truckee has discontinued breakpoint chlorination. The
plant now averages an ammonium-nitrogen release of 5
to 6 mg/L, with the effluent from any one of the three
operating columns allowed to increase to 12 mg/L. This
operation is made possible since the columns (four pro-
vided) are at various states of ammonium breakthrough
(breakthrough occurs rapidly after the operating column
effluent exceeds 5 to 6 mg/L ammonium nitrogen).
The view obtained from the North Tahoe-Truckee plant is
that Clinoptilolite ion exchange is a workable but demand-
ing operation. Maintenance requires frequent hydrochloric
acid washing of the closed loop stripper and absorber
media to remove the scale. About 20 percent of the Cli-
noptilolite is replaced per year. The caustic, acid, and salt
handled in the media regeneration and ammonia recovery
process present a corrosive environment, dictating spe-
cial safety concerns and equipment (e.g., pipes, pumps,
valves, fittings, and instruments) which is two to three
times more expensive than standard equipment, and dif-
ficult to maintain, repair, and replace. However, the plant
management is pleased with the process after 12 years
of operation. :
Operational care is particularly important in terms of pre-
venting ammonium sulfate crystal formation. The formed
ammonium sulfate is contaminated with sodium. Once
given away to a potato farming operation in Nevada, the
46
-------
waste ammonium sulfate is now used as a nitrogen sup-
plement in the commercial production of organic fertilizers
by others.
The ion exchange discussion above has been limited to
large NHJ removal systems. The reader should be aware
that a new generation of small (one to a few) home sys-
tems is being proposed on a throwaway or commercial
regeneration service basis that may have promise. Also,
advances in synthetic ion exchange resins may permit
their use for polishing of effluents that exceed stringent
permit requirements.
2.5.2.3 Breakpoint Chlorination (Nhfi-N)
Breakpoint chlorination for nitrogen control originated
from the water treatment industry, where it was once used
to achieve a free available chlorine residual because of
disinfection concerns with chloramines. Figure 2-2 illus-
trates a typical breakpoint chlorination curve for an ad-
vanced treatment effluent.
Figure 2-2 also summarizes the reaction stoichiometry.
The reader should come away with the view that many
reactions in addition to breakpoint can occur; the process
results in the unavoidable introduction of dissolved solids,
and it can deplete the alkalinity if a buffer is not added
or contained with the chlorine source for this purpose
(which may also add dissolved solids). Dechlorination is
likely to be necessary for today's municipal wastewater
practice and, if activated carbon is not used (which is not
recommended because of the expense), this followup
process will also add dissolved solids and deplete
alkalinity.
The reactions are dependent on pH (around 7 to 7.5 is
the preferred target for the fastest reaction rates, while
o
Dichloramine Formation
with Ammonium Oxidation
Irreducible
Minimum Chlorine
Residual
&
1=
o
z
5 7.6
CI2 :NHj -N Weight Ratio
Theoretical Breakpoint Curve
Description
Reaction Stoichiometry*
Breakpoint reaction
NCI3 formation
Nitrate formation
1-from ammonia
2-from nitrite
1.5 HOCI •
HOCI .
4CI"
+ Ci"
NH4 + 4HOCI —+• NO3
NO2+ HOCI —*- NO^ + H"*
Other inorganic reactions (e.g., H2 S) H 2 S + HOCI —»- s | + HCI + H2 O
Other organic reactions Organics + HOCI —>• Oxidation and chlorinated end products
'Percent distribution of HOCI and OCI" with pH
PH
678
HOCI 97 75 49
OCI" 3 25 51
Figure 2-2. Typical breakpoint chlorination curve and stoichiometric reactions for an advanced treatment effluent
(from Reference 30).
47
-------
avoidance of values below 6 is necessary to minimize the
formation of toxic, pungent nitrogen trichloride); the chlo-
rine dose; the magnitude of the ammonium nitrogen; and
the competition of the various side reactions with nitrite,
inorganics (sulfide is presented for reference in Figure
2-2), and organics (trihalomethane residuals are formed
once the breakpoint reaction occurs). These reactions
affect and extend the breakpoint curve.
Stoichiometric predictions of 7.6 mg of chlorine per mg
of ammonium nitrogen are never seen in practice. Even
the cleanest processing streams will experience weight-
to-weight ratios of 9 or 10:1. Processing streams with high
SS and incompletely biologically stabilized soluble residu-
als can result in weight-to-weight dosages of 20 or more:1.
The mid-seventies breakpoint chlorination experience
with a biologically stabilized effluent produced the follow-
ing end products and chlorine usages (30). The effluent
contained a nominal ammonium nitrogen of 20 mg/L; a
chlorine to ammonium nitrogen dosage of 10 to 1 was
used at a pH point between 7 and 8:
End Products and Chlorine Usages
NHJ toN2
NHJ toNOg
NHJ toNCI3b
NOg to NOJ
Remaining Residual
Unaccounted
Pathway of
Applied
NHj-N,a %
90
5
2
—
3
—
Pathway of
Applied CI2,
%
70
13
5
1
7
4
* Organic nitrogen was reportedly unchanged (likely applied concentra-
tion = 0.5 to 1.0 mg/L).
* Reverts to NHJ-N on dechlorinatfon with sulfite.
The lack of other reported data prevents any conclusion
as to the representative nature of the preceding charac-
terization. However, it is likely that 90-percent removal of
the applied ammonium nitrogen is achieved once the
breakpoint reaction occurs.
The only known operating facility where breakpoint chlori-
nation is the principal nitrogen control strategy is at Sug-
arbush, Vermont. Here, average daily flows range from
about a quarter to one-half of its rated 7 Us (0.16 mgd)
capacity. Ammonium-nitrogen concentrations in the raw
sewage reach up to 30 mg/L in the winter, but typically
average approximately 15 mg/L. The breakpoint chlorina-
tion system is preceded by 7,570 m3 (2 Mgal) of aerated
pond storage, alum clarification of the pond effluent, and
dual media filters. It is followed by 36,000 m3 (9.5 Mgal)
of storage and a 0.4-ha (1-ac) leach field. Chlorine to
ammonium-nitrogen weight dosages average 18:1. Nitri-
fication in the aerated raw sewage storage tank during
the summer allows a reduction of the chlorine usage, but
only with the accompanying result of higher total nitrogen
in the processing effluent. The plant flow release is sea-
sonally restricted. Reported November 1990 through Oc-
tober 1991 performance (based on weekly, eight-hour
composites) is described below. Note that the average
flow was 2.5 L/s (40 gpm) for this period:
Breakpoint Chlorination Performance
Org-N NHj-N NOi-N NO5-N Total N TRC°
Annual Average, mg/L
Influent 9.3
Filter Effluent3 2.3
Breakpoint Effluent —
To Field" 1.0
14.3
10.5
0.1
0.2
0.3
Worst Month Quality, mg/L
To Field" 1.0 0.1
0.6
1.7
3.7
3.8
7.2
24.2+ —
14.8 —
— 6.2
5.0 —
8.3 —
a35 days later
"Q = 0.0029m3/s; 140 + 35 = 175 days later
°TRC = total residual chlorine
The utilities director's recommendation for others consid-
ering full nitrogen control by breakpoint chlorination can
be summarized in one word—"don't."
It is recommended that breakpoint chlorination be rou-
tinely considered only for polishing applications, such as
was used at the previously described North Tahoe-
Truckee Plant, where a low total or unoxidized nitrogen
residual is mandatory. (Note that Tahoe-Truckee operates
a granular activated carbon tertiary polishing system for
organic nitrogen and carbon removal prior to its ion ex-
change/chlorination processes.) Use of sodium hypo-
chlorite is preferred for safety reasons and because of
the attendant buffer found with the caustic carrier (NaQH).
Sodium hypochlorite adds about 1.7 mg of dissolved sol-
ids/mg of chlorine added.
One should anticipate subsequent residual chlorine re-
moval when breakpoint chlorination is practiced. Acti-
vated carbon polishing for chlorine removal is complex
and expensive and is not recommended over dechlorina-
tion by sulfite addition at larger facilities. Liquid sulfite
addition (or possibly activated carbon) is recommended
for small plants because of ease of handling and safely
issues. Larger plants could use SO2 cylinders; SO2 hy-
drolyses in water to form HSOi. The reaction of the sulfite
with free and combined chlorine residuals is essentially
instantaneous. Dosages are only 10 to 30 percent higher
than Stoichiometric predictions. The Stoichiometric weight
ratio of sulfur dioxide to chlorine (CI2) is 0.9:1. One mg
of SO2 will deplete 2.4 mg CaCO3 alkalinity.
48
-------
2.5.3 Non-Ion-Specific Technologies
The non-ion-specific processes are mentioned only in the
interest of completeness. Here, all nitrogen forms are
removed with other dissolved constituents with varying
degrees of selectivity by the demineralization process of
choice.
2.5.3.1 Reverse Osmosis
Reverse osmosis was selected as a representative
demineralization technology. Six months of experience in
the late seventies with reverse osmosis at Orange
County's Water Factory No. 21, in California, suggested
the following macrocontaminant removals (31):
Macro-
contaminant
Sulfate
Ammonium-N
COD
Sodium
TOG
Nitrate
Influent,
mg/L
338
22.5
12.5
158
6.0
2.9
%
Removal
gg
g4
94
B9
82
55
Residual,
mg/L
3
1.4
0.8
17
1.1
1.3
The reverse osmosis system was preceded by the follow-
ing processes: biological stabilization, high pH lime clari-
fication, stripping (no aeration, just allowed to fall through
the packing of the 24-ft tower), recarbonation and clarifi-
cation, mixed media filtration, and activated carbon
adsorption.
It is difficult to interpret the reported performance data for
the reverse osmosis system other than to observe that
nitrate appears more difficult to remove than ammonium
and that the sum of these two nitrogenous compounds is
near 3-mg/L after all of the preceding treatment proc-
esses. (Activated carbon would have achieved some
removal of the soluble organic nitrogen.) Characteriza-
tions such as this lead to the conclusion that the only
likely, and certainly the easiest, ways to achieve low-level
ammonium or total nitrogen residuals are with biological
processes. This is particularly the case if control of the
macrobioavailable pollutants is the primary processing
objective.
The 44-L/s (1 mgd) Denver, Colorado, secondary effluent
to drinking-water quality demonstration pilot plant origi-
nally employed clinoptilolite to achieve its objective of less
than 1 mg/L ammonium nitrogen. Clinoptilolite was re-
ported to be unacceptably expensive ($0.40/1,000 gal)
and management subsequently incorporated reverse os-
mosis for ammonium nitrogen control. Present findings
suggest reverse osmosis ammonium-nitrogen removals
on the order of 85 percent (32).
2.6 Key Fundamental Process Selection
and Design Issues
Section 2.3.3 described the processes that are the prin-
ciple focus of this manual. They embody, alone or in
combination, suspended and attached growth treatment
concepts in a higher technology, mechanical plant ap-
proach. This next section addresses key issues that in-
fluence the selection of the unit processes for the nitrogen
control objectives.
2.6.1 Fundamentals and Empirical Factors
2.6.1.1 Comparative Differences and Needs
The purpose of this section is to provide a fundamental
appreciation of the differences between suspended and
attached growth technologies, and the ramifications of
such in practice. Emphasis is placed on nitrification since
it is the most frequently encountered design objective and
without effective nitrification the best designed denitrifica-
tion system has no chance for success. Finally, in terms
of attached growth systems, emphasis is also placed on
approaches found with trickling filter (TF) and rotating
biological contactor (RBC) systems.
With respect to the system biomass, suspended and at-
tached growth reactors are complete mix and plug flow
systems, respectively. The suspended growth system's
complete mix definition is derived from the return of set-
tled biomass. Thus, save wastage, all of the organisms
grown from treating the wastewaters are returned to con-
tact the newly applied wastewater. Their relative activity
at any point, however, will be controlled by the environ-
mental conditions (e.g., DO), the applied substrate form
and mass, and their history of exposure to the newly
received substrate mass.
Conversely, in an attached film system without a return
of the reseeding biomass or fluidization for mixing, the
organisms found at any point in the vertical or horizontal
processing train will reflect only the wastewater charac-
teristics and environmental conditions that are routinely
encountered at that point. Conceptually, it is difficult to
visualize that the return of reseeding biomass to the fixed
film reactor will impart significant benefit, since changes
in the overall system solids residence time will be small.
For any system that is required to remove appreciable
amounts of SBOD5, this small return may be accompa-
nied by increased front-end DO stresses, with an accom-
panying risk of biomass accumulation and odor problems.
However, if the system is called upon to nitrify, the recycle
of the small amounts of nitrifiers encountered in the ef-
fluent from an attached growth reactor, or a fraction of
the settled sludge from a downstream nitrifying activated
sludge system, may improve nitrification performance.
Some denitrification may also be achieved in a combined
carbonaceous and nitrogenous oxidation system.
49
-------
Aerobic metabolism in either attached or suspended
growth systems is controlled by oxygen transfer to the
liquid and, in turn, DO and substrate penetration into the
bacterial film or floe. In suspended growth systems, the
oxygen supply can be adjusted to satisfy the demand
which, of course, is influenced not only by the applied
substrate but also by the concentration of operating sol-
ids. The large volume (or long detention times) of sus-
pended growth systems also serves to dampen the peaks
and valleys of this demand, contributing to an inherent
stability.
In a classical TF or RBC attached growth system, the
oxygen supply potential is generally fixed by the static
aerator nature of the media and transfer is only achieved
by the magnitude of flow that passes over the media, or
by the speed of rotation of the media through the fluid.
The demand is experienced where it is experienced and
is not subject to control by changing levels of operating
solids concentrations. The limited volume of liquid (i.e.,
short detention time) encountered in an attached growth
system contributes to its inherent conceptual instability.
The literature is mixed on the value of recycle to TF and
RBC attached growth systems. Undoubtedly, that value
depends on the system's application. With highly loaded
systems, low SS recycle offers conceptual value in the
transfer of more oxygen per unit mass of applied biode-
gradable substrate, less oxygen demand per unit volume
of applied wastewater per unit time (giving load migration
farther down the system), and additional scour of accu-
mulated biomass (yielding more active surfaces). With
low-loaded trickling filtration systems, low SS recycle of-
fers some ability to mitigate such nuisance organisms as
filter flies and snails as well as the potential of more
effective exposure of the wastewater to all of the available
surfaces through increased media wetting.
A flexible TF and RBC design approach would incorporate
a capability for hydraulic recycle of low SS liquids to
provide some ability for operating manipulation. An attrac-
tive operating feature for trickling filtration systems is the
ability to control the localized hydraulic application rate
independently of flow for nuisance control (i.e., excess
biomass, organism, and odors) and improved load distri-
bution (forcing it downstream). Similar approaches for
RBCs would make use of an air scour for nuisance control
and have the ability to cyclically reverse and step feed
the flow through the process for improved load distribution
and/or attached culture development.
Conceptually, the TF processing configuration is inher-
ently inferior to the RBC because of its physical inability
to distribute and cycle loads (except by serial application
and change of the lead filter). A forced ventilation capa-
bility is highly desirable for a trickling filter system and is
a recommended design approach.
Recent evolution of RBC technology shows that continu-
ously operating supplemental air supplies are beneficial
for heavily loaded systems. Submerged media, packed-
bed attached growth systems directly use an air supply
to respond to the system's oxygen transfer needs. At-
tached growth technologies with direct control of the air
supply and certain usage of effective surfaces, not unlike
suspended growth systems, yield the greatest certainty
of predictable performance.
Once SBOD5 loadings decline below those that yield
readily observable nuisance conditions in classical TF
and RBC systems, the controlling design considerations
for carbonaceous stabilization are substrate transfer and
oxygen transfer. Initially, in combined carbonaceous-nitri-
fication systems, the dominance of heterotrophic flora and
remaining carbonaceous substrate will control and use
the available oxygen at levels that suppress any inroads
by autotrophic nitrifiers. It should be remembered that
nitrifier replication, in comparison'to that of the organic
stabilizing heterotrophs, is much more sensitive to DO
concentrations.
The onset of nitrification is determined by heterotro-
phic/nitrifier competition and oxygen transfer considera-
tions. For combined carbon and nitrogen oxidation
trickling filters, experience suggests the following sequen-
tial activity through the system (33,34):
• Nitrifier growth is no longer limited by competition with
heterotrophic organisms when soluble five-day carbo-
naceous biochemical oxygen demand (CBOD5) values
fall below about 20 mg/L.
• Once nitrification does begin, the nitrifying population
builds to its maximum as determined by the oxygen
transfer system, the declining significance of the stabi-
lized wastewater's carbonaceous oxygen demand, sol-
uble carbonaceous and nitrogen opportunity for diffusion
into the biofilm, and the oxygen uptake of the combined
aerobic heterotrophic and autotrophic biofilm population.
• When and where the maximum nitrifying population
first occurs is difficult to identify beyond the certainty
that it is probably accompanied by a soluble CBOD5
on the order 10 mg/L.
• Thereafter, the maximum nitrifying population is con-
trolled by the oxygen transfer characteristics of the
reactor until the residual ammonium nitrogen drops
below 5 mg/L. *
• The nitrifying population then declines in proportion to
the ammonium-nitrogen levels that are routinely expe-
rienced in the downstream reaches of the processing
train and in relation to the ability to achieve sufficient
biofilm for the population's retention.
Trickling filter systems will achieve consistently low am-
monia objectives when properly designed and oper-
ated (35). The foregoing observations can explain why
50
-------
classical attached growth system nitrification perform-
ance shows little sensitivity to temperature over a range
of 12-20°C (36,37) and why combined nitrogen-carbon
oxidation and tertiary nitrifying attached growth systems
have shown variability at low-level ammonium objectives
and can have difficulty achieving consistent low-level am-
monium residuals near 2 mg/L or lower.
The interrelationships, of all factors point to the apparent
complexity of fixed film process design. Seemingly unim-
portant or inevitable plant practices, coupled with the nor-
mal variability.of incoming wastewater flows and pollutant
concentrations, may have important consequences.
Higher nitrogen concentrations (due to discontinuous
sludge processing recycles and/or daily raw wastewater
transient peaks) can result in unfavorable alkalinity to
ammonium-nitrogen ratios or lower DO concentrations
(associated with the competing spikes of carbonaceous
material and the dominant heterotrophic population),
which can suppress nitrifier activity at points where or-
ganism concentration is highest. This forces the unoxi-
dized nitrogen to points downstream in the processing
train where the nitrifying population may be lower. Lower
nitrifying populations, regardless of more favorable DO
characteristics, will yield higher unoxidized nitrogen levels
possibly beyond the target concentration. Poor develop-
ment of biofilm and predation in the latter stages of the
attached growth system, during average and low loading
periods, further influences the ability of the system to
perform under infrequent stress periods.
2.6.1.2 Coupled or Combined Processes
Blending attached growth trickling filter and suspended
growth activated sludge technologies offers the designer
the opportunity to take advantage of what each technol-
ogy does best. Application of a suspended growth system
with its returned solids after the trickling filter allows the
designer to achieve superior flocculation, which, funda-
mentally, is a function of the number of solids, time of
contact, and mixing intensity. This allows trickling filter
systems to compete with today's needs for rigid SS
control. The trickling filter can also provide a roughing
operation (or selector), with polishing provided by the
downstream, now very stable, activated sludge system.
Recent or newly proposed advances in attached and sus-
pended growth technologies continue this theme. The
evaluator of coupled or combined new systems is best
served by application of fundamentals, such as those
described in the preceding subsection, and an awareness
of past experiences with the operation of the individual
processes. When compared to the limited understanding
often encountered with emerging technologies, it is useful
to remember that older systems and design approaches
have been tempered by experience gained through years
of practice.
2.6.1.3 Carbon-to-Nitrogen Ratios
Beyond the solids residence time concept, there is no
more fundamental consideration with nitrogen control
systems than what is termed "the carbon-to-nitrogen ra-
tio." The purpose of this subsection is to introduce the
reader to this important concept.
Figure 2-3 was prepared to provide a fundamental con-
ceptualization of how the carbon to nitrogen ratio influ-
ences the development of the nitrogen control strategy
and its design response. Understanding begins with the
characterization of the influent wastewater. As shown, raw
wastewaters are relatively high in organic carbonaceous
material and relatively low in unoxidized nitrogen; typi-
cally, a COD:TKN of 10 to 15:1 is presented.
The biodegradability of the carbonaceous material in the
raw wastewater is influenced by a host of variables, in-
cluding the nature of the collection system, which also
influences the inert SS in the influent wastewater. One
certainty is that biodegradability will increase with dimin-
ishing particle size. Those particles that escape primary
sedimentation have the highest immediate biodegradabil-
ity, while the retained particles must be hydrolyzed (solu-
bilized) to simpler compounds before active cellular
oxidation and synthesis can begin. Gray to black waste-
waters signify that solubilization in the collection system
is under way. Remote pumping stations promote active
degradation by retention in the wet wells and particle
breakup in the pumps. Combined sewers and major in-
terceptors promote deposition and active decay of depos-
its until they are resuspended under rain-swollen events.
These are accelerated by warmer ambient and wastewa-
ter temperatures.
Heavier, larger particles may or may not be separated
prior to the biological reactor. If not, the carbon-to-nitro-
gen ratio will be much higher than if they were. Once the
substrate is applied to the biological reactor, the magni-
tude of the resultant mixed liquor volatile suspended sol-
ids (MLVSS) will be largely influenced by the applied
refractory (or nonbiodegradable) particulate volatile mat-
ter and the cell residence time of the heterotrophic micro-
bial population.
The acclimated heterotrophs in the biological reactor will
respond immediately to the soluble organics and yield
both oxidation and new biomass end products. The
MLVSS is composed of the applied refractory particulate
organics, the slowly biodegradable particles derived from
the applied wastewater, and newly formed cellular volatile
matter. The slowly biodegradable particulate matter will
also be eventually oxidized and synthesized into new
cellular volatile matter. The new cellular matter will also
be oxidized and resynthesized, leading to an accumula-
tion of refractory cellular end products, the extent of which
is dependent on the solids residence time of the system.
51
-------
Applied
Substrate
(Raw Wastewater)
-j- Typical Refractory or
Nonbiodegradable
Fraction (Not
Shown in Nitrogen
Distribution for
1 VlabiB
Organic
Stabilizing
Hotorotrophs
"> MCRT Dependent •
/ VSS Synthesis, Respiration^
I and Resynthesis with
I Refractory End Products
\
Nitrification
Limiting
MCRT
Net Initial Weight Gain
:Due to Oxidation and
Synthesis of Soluble
Substrate
\ -«*.•!*-•*,*. •«'• .«i
"- Organic Stabilizing:;
-°* Heterotrophs-'j;.
r Organic N
jr r Particulate
Net
Nitrifying
Autotrophs
Net
VSS
Production
Net
Carbonaceous
Substrate
Removal by
Oxidation
Soluble Available and
Soluble
(NH4-N)
Refractory Nitrogen
Remaining Viable
Nitrifying Autotrophic Biomass
Remaining Viable
,. Heterotrophic Biomass
Heterotrophic Refractory
; r End Products
Refractory
Material from
Original Substrate
MCRT, Days
•BODg is not presented because it is an
incomplete measure of the substrate and
its biodegradability.
TKN (total Kjeldahl nitrogen) measures
the total unoxidized nitrogen.
Nitrification Issues
Likely Net Nitrogen Removal Due to
—I Refractory and Synthesis
I Removals
Stabilized,
Oxidized
Substrate
No Remaining Soluble Substrate Blodegradabilitv
Figure 2-3. Schematic of how carbon-to-nitrogen ratio influences development of a nitrogen control strategy along with
the design response.
For all intents and purposes, appreciable concentrations
of nitrifying autotrophs are not encountered until the sat-
isfaction of some controlling solids residence time (60)
reflecting the concurrent environmental conditions that
allow their replication and retention within the system.
Once the limiting 60 and environmental constraints are
satisfied in a suspended growth reactor, the nitrifier popu-
lation becomes dependent on the residual nitrogen that
is available for nitrification (nitrogen available after cor-
rection for the amount removed by cellular synthesis,
refractory particulate entrapment, and the soluble refrac-
tory organic nitrogen). In an attached growth system, as
explained in Section 2.6.1.1, at any point in the reactor
the nitrifier population will be influenced by competition
between autotrophic and heterotrophic biomass.
This complex series of events is portrayed above the
Nitrification Issues heading in Figure 2-3. As can be seen
for the portrayed raw wastewater situation, the net viable
autotrophic nitrifier population is small in comparison to
the total net volatile suspended solids (VSS) production,
as is the net nitrogen removed by synthesis. Clearly, to
achieve a higher percentage of nitrifiers in the biomass,
either the applied substrate carbon must be decreased
or the applied available nitrogen must be increased.
Decreasing the applied substrate carbon can be readily
accomplished, if desired (e.g., with primary clarifiers), and
will lead to smaller reactors (for the same 9C) and less
overall oxygen demand. The applied C:N can also be
manipulated in plug flow attached growth systems to af-
fect a greater concentration of nitrifiers in the zone of
ammonium concentration limiting performance (Section
2.6.1.1). Here, by cyclic switching of the lead trickling filter
tower (if available), or by staging and flow reversals with
an RBC system, higher overall nitrifier populations and
lower ammonium nitrogen residuals may be promoted.
The requirements for achieving effective denitrification
are the reverse of those encountered with nitrification.
Again, the carbon-to-nitrogen ratio comes into play. Here,
higher values of biodegradable carbon to nitrate nitrogen
are sought to drive the denitrification reaction, and the
viability of the heterotrophic population in the total
biomass is a paramount concern for small reactor sizing.
52
-------
The denitrification reaction can be positioned at the head,
interstage, or end of a nitrification reactor or system. All
the reaction requires is a source of nitrate nitrogen (which
can be delivered by the return of a nitrified liquid), a
heterotrophic .biomass, and a carbonaceous substrate.
The speed of the reaction and the extent of nitrate-nitro-
gen reduction are influenced by the magnitude and form
of the carbonaceous substrate and its ratio to the oxidized
nitrogen available for denitrification.
The complex series of events and processing alternatives
accompanying the denitrification objective is schemati-
cally portrayed in Figure 2-3 over the Denitrification Is-
sues heading. Unlike nitrification, the election of the
denitrification process involves balancing a multitude of
tradeoffs. The following paragraphs provide additional
fundamental understanding.
The ideal substrate for denitrification is a readily biode-
gradable soluble carbon source, free of nitrogen, that is
preferentially oxidized rather than synthesized. Methanol
is such a substrate, but it is foreign to municipal waste-
waters. As such, the organisms must be acclimated to it
before it can be degraded. The acclimation response is
dependent on the solids residence time and the methanol
concentration. Culture isolation, whether it occurs in a
suspended or attached growth reactor, yields the highest
concentration of viable heterotrophs acclimated to metha-
nol; but methanol leakage from the system still will occur
if application rates suddenly exceed the mass to which
the organisms have become accustomed.
When using foreign substrates to meet stringent, low-
level nitrate-nitrogen goals, dosages must meet the high-
est applied nitrate-nitrogen mass and ensure that
adequate substrate is present for all processing require-
ments. In order to avoid soluble substrate breakthrough,
post-aeration of the mixed liquor suspended solids
(MLSS) is used in suspended growth systems to com-
pensate for overdosages, particularly when the available
nitrate nitrogen is less than needed to satisfy the applied
substrate. Soluble substrate breakthrough is more likely
to occur in attached growth systems.
Practically, the most available substrate for denitrification
at a wastewater treatment plant is the raw or settled raw
wastewater. In this case, it should be remembered that
all substrates are both oxidized and synthesized, and only
that fraction oxidized will enter immediately into the deni-
trification reaction. The most suitable and immediately
usable substrate in the denitrification reaction is the sol-
uble biodegradable carbon. Thereafter, the denitrification
rate proceeds as a function of the concentration and vi-
ability of the heterotrophs in the biomass and the biode-
gradability of the applied particulate carbon and the
synthesized cellular products. Younger cultures promote
a higher viability.
More slow to degrade, biodegradable particulate carbo-
naceous waste can be added through application of raw,
as opposed to settled, wastewater. But this will lower the
percentage of viable heterotrophic organisms in the total
MLVSS and will increase the need for more oxygenation
energy for carbonaceous stabilization, these two com-
promises may yield no net change in (or an even lower)
reaction rate (e.g., mg NOj-N reduced/mg MLVSS/d).
Moreover, where high-level denitrification is" required, it
may be more cost effective to incrementally add a sub-
strate such as methanol than to satisfy the additional
carbonaceous oxygen demand associated with the.raw
wastewater.
Recent advances in combined nitrogen and phosphorus
removal systems have sought to catalyze the biologically
enhanced phosphorus removal process through the an-
aerobic formation of readily biodegradable volatile fatty
acids. Volatile fatty acids formation, which is the first
mechanistic step in anaerobic digestion, can naturally
occur in collection systems, sometimes to a significant
degree. Design approaches promote the formation of
these acids through either a front-end anaerobic process
or a sidestream fermenter.
With or without enhanced biological phosphorus removal
strategies, the volatile acids generated from particulates
also facilitate mainstream denitrification by increasing the
C:N ratio. The sidestream fermenter approach, and sub-
sequent liquid recycle, has particular conceptual attrac-
tiveness in that it minimizes the introduction of non- and
slowly biodegradable organics (and inert SS) to the reac-
tor and their consequent impact on reactor sizing. The
negative aspects of the sidestream fermenter approach
include additional unit processing commitments, a poorly
stabilized sludge end product (needing further stabiliza-
tion), lower potential for methane production, and the
attendant nitrogen solubilization of anaerobic digestion,
yielding the possibility of higher overall system oxygen
demands.
The foregoing discussion reveals a multitude of complex
considerations influenced by the carbon-to-nitrogen ratio.
Fundamental understanding is only derived through the
realization that it is both the magnitude and form of the
carbon that count in nitrogen control systems. Differences
in practice, as influenced by the collection system, and
other unsteady environmental conditions and discontinu-
ous plant practices account for the often significant vari-
ance observed in nitrogen control plant performance,
even in seemingly similar situations.
2.6.1.4 Temperature
The impacts of temperature on process performance can
be, and often are, of prime importance. Care should be
taken when applying the results observed at one facility
to another without understanding the temperature condi-
tions. Care should also be exercised when applying tem-
53
-------
perature adjustments in process kinetic characterizations,
although it should be said that there is generally good
confidence in process kinetic temperature descriptions of
nitrifier growth rates.
The points made in the C:N ratio discussions should be
recalled when addressing carbonaceous biomass predic-
tions. The great bulk of the mixed liquor volatile matter is
biologically inert and is not subject to change with chang-
ing temperatures. This explains why the reported net sol-
ids production remains largely constant with wide
temperature changes and why for long-sludge-age sys-
tems it is not strongly sensitive to the solids residence
time differences that are routinely encountered in prac-
tice.
2.6.2 Process Selection and Assessment of Design
Issues
The material in the preceding subsection and the balance
of this manual provides the basis for the qualitative com-
parative assessment of the suspended growth and at-
tached growth technologies provided in Table 2-8. In
using this table, the reader should bear in mind that it is
only intended to provide general guidance. The often criti-
cal variable of site-specific considerations, as well as
greater understanding and evolution of nitrogen control
technologies with time, may make some of the furnished
direction invalid.
Overall, the relative observations that can be made from
Table 2-8 are summarized as follows:
• Suspended growth technologies are:
- better understood,
- less influenced by other processing considerations
at the plant,
- Inherently more flexible and adaptable to a variety
of processing objectives,
- possibly a safer fundamental choice when high to
stringent effluent standards are encountered.
• Attached growth technologies are:
- often simpler to operate and maintain,
- probably best applied in polishing or roughing roles
with suspended growth technologies, or when the
effluent objectives are not overly stringent.
2.7 Frequently Encountered Linked Design
Issues
The beginning to this chapter introduces the concept that
the appropriate unit process selection (or design) does
not stand alone. It is influenced by all things that precede
and follow in the overall scheme of the treatment works.
The purpose of this section is to provide a fundamental
and practical discussion of one certain and one likely
design issue both of which are linked to the selection,
and often the sizing, of the nitrogen control process.
These are solids-liquid separation and phosphorus re-
moval, respectively.
2.7.1 Solids Liquid Separation
2.7.1.1 Suspended Growth Systems
A limiting consideration for sizing or operating a suspended
growth reactor is its final sedimentation system. Often up-
sets occur because of one or more of the following:
• poor settling characteristics of the applied MLSS,
• too high design or operating MLSS loadings to the
clarifier for the experienced flow regime, and/or
• inadequate or too low a return sludge rate for the
applied MLSS concentration and its settling charac-
teristics, leading to expansion and loss of the sludge
blanket.
MLSS that demonstrate poor settling characteristics are
commonly encountered in air activated sludge systems
on a seasonal, if not endemic, basis. In municipal treat-
ment, the major cause is the presence of filamentous
forms of bacteria, which can flourish in the biomass be-
cause of low to trace DO levels in the mixed liquor or the
sludge floe itself, or other factors. The older (or higher
00 cultures, common in single-sludge continuous flow ni-
trogen control systems, inherently promote the retention
of these filamentous forms. Other systems, such as the
oxidation ditch, by the very nature of the isolated point of
air (oxygen) application create reactor conditions that pro-
mote natural denitrification and the growth of filamentous
forms. Similar situations may be encountered with me-
chanical aeration systems.
Although mitigated to a degree by raw sewage applica-
tions (because of its heavier organic and inorganic parti-
cles), settling problems abound in the historic application
of both plug flow and complete mix variants of the acti-
vated sludge technology. Historically in these systems,
the only available operating strategy was to reduce the
operating MLSS concentration while trying to identify the
cause of the problem. Reducing the MLSS to comply with
the effluent SS objective surely compromises the reactor
performance for denitrification, and may reduce the aero-
bic solids residence time to the point that seasonal am-
monium or total nitrogen limits are violated.
Over the last 20 years, pure oxygen systems have been
developed that can often reduce the variability and en-
hance the settleability and densification characteristics of
MLSS. In this timeframe, and more specifically in the last
decade, the technical literature has seen progressive in-
terest and success in applying front-end selectors (aero-
bic, anoxic, and anaerobic) for improved MLSS settling
and densification characteristics. For nitrification applica-
tions, it is worth considering pre-anoxic zones for their
sludge volume index (SVI) and stability benefits. Pure
54
-------
Table 2-8. Comparative Assessment of Suspended and Attached Growth Technologies Against Key Process Selection
and Design Issues
Technologies*
Issue
Suspended Growth
Attached Growth
COMPATIBILITY: PREREQUISITE PROCESSES,
• Prerequisite Processes
• Nitrogen Oxidation, Effluent NHj-N
-Intermediate, 5-10 mg/L
- High, 2.5-5.0 mg/L
- Stringent, <2.5 mg/L
- Variability
• Nitrogen Removal, Effluent Total N
— Intermediate, 10-15 mg/L
-High, 5-10 mg/L
- Stringent, <5 mg/L
- Variability
• Effluent SS Considerations
— Clarifier for capture
— Effluent SS consistently <10-15 mg/L
• Vulnerability to Intermittent Sludge Processing
EFFLUENT STANDARDS, AND SOLIDS PROCESSING TRAIN
None beyond preliminary treatment Needs prior gross solids removal
when applied to raw wastewaters
INHERENT PROCESS FLEXIBILITY
• Adaptability to Maximize Internal Denitrification
• Adaptability to Variety of Operating Control
Strategies
• Adaptability to Biologically Enhanced
Phosphorus Removal
• Wastewater Temperature Influences
Principal Controlling Process Variables
Easily achievable, difficult to
control
(consider split treatment)
Easily achievable
Good choice
Lower
Easily achievable w/o external
substrate
Achievable (normally w/o external
substrate additions)
Achievable (may require substrate
additions)
Easily achievable
Achievable with good design
May be poor choice without
enhancements
Likely higher
All need external substrate
addition; most installations in
polishing applications.
Easily achievable in most
climates
Likely achievable in most climates
Only fluidized bed and deep
downflow bed demonstrated in
cold climates
Mixed, because of the upstream need for successful nitrification, often
reliance on natural denitrification, and final variable of external substrate
addition
Mandatory
Filter becomes increasingly
mandatory
Lower because of longer hydraulic
detention time
Excellent
High
High
Nitrification and denitrification
kinetics
Reactor volume, operating solids
concentration, oxygen transfer
Depends on application and
technology
Depends on application and
technology
Higher because of shorter
hydraulic detention time
(consider equalization and
bleedback, especially with
anaerobic digestion)
Poor to nonexistent as presently
demonstrated
Low
Not directly
Often not controlling for
nitrification with some technology
applications, applicable to
denitrification kinetics
Reactor volume (media-specific
surface), hydraulic loading, and
oxygen transfer
55
-------
Table 2-8. Comparative Assessment of Suspended and Attached Growth Technologies (continued)
Technologies*
Suspended Growth
Attached Growth
SENsmvny TO NEEDS OF OPERATION
• Operating Simplicity
• Ease of Maintenance
• Vulnerability to Upset as a Result of
- High Flow
- High Raw Pollutant Mass
— High Raw Pollutant Concentrations
- Deterioration of Recycles
— Toxic Compounds
• Recovery from Upset
• Ease of Process Troubleshooting
• If Plant Is Noncompliant, Is the Fix:
— Solvable by Operations?
- Solvable by:
Minor Capital Improvements?
Major Capital Improvements?
• Aesthetics
-Sight
-Sound
-Odors
— Nuisance Organisms
Inferior
Inferior
Higher to same
Lower
Lower
Yes, if sustained
Varies
Mixed
Better
At times
Often
Not as common
Generally superior
Potentially inferior
Generally superior
Nocardia scum or float
ADDITIONAL THOUGHTS ON PROCESS SELECTION
• Present Applied Understanding of Nitrogen
Control Technology
- Oxidation
— Removal
Plant Site
Generally superior
Generally superior
All sites
Superior to same
Superior to worse
Lower
Higher
Much higher
Much higher
Varies
Mixed
Worse
Less frequently
Not as frequently
Often
Generally inferior
Generally superior
Potentially inferior
Flies, snails, and worms
Generally inferior
Generally inferior, except for
warm weather
Often attractive with high natural
gradients
* Avoid rigid thinking. Best process selection may be a blend of technologies, both coupled and uncoupled, using each for what it does best.
oxygen and selector technologies are largely directed to
mitigation of the immediate soluble substrate impact on
the MLSS, in terms of natural selection of filamentous
forms at low DO concentrations. Process development
and proprietary technology considerations with nitrogen
control applications undoubtedly served as a catalyst for
many packaged technologies offered by various suppli-
ers. In some processes, such as the sequential batch
reactor, it is relatively simple to establish the selector
concept when choosing the operating sequence.
Today, the following guidance seems appropriate for con-
trol of poorly settling and densifying activated sludges:
• Beware of reliance on high MLSS designs (see next
paragraph).
• Provide some form of selector technology.
• Still assume that uncontrollable bulking may occur and
provide the ability to add chlorine or hydrogen peroxide
to the return sludge, so that the operator has an im-
mediately implementable external control procedure.
The design MLSS issue is, ultimately, a function of the
applied solids concentration and mass loading to the clari-
fier, and in this respect there are no hard and fast rules,
only general guidance. Ideally, the MLSS concentration
should be set to encourage nonhindered, as opposed to
56
-------
hindered, settling solids characteristics, a breakpoint that
often occurs at about 2,000 mg/L for air activated sludges
and at times some 50 percent higher for pure oxygen
activated sludges or sludges encountered with selector
technologies. The breakpoint is specific to the charac-
teristics of each sludge. Generally, settling velocities with
MLSS concentrations below these values show little gain,
while settling velocities associated with MLSS concentra-
tions above these values rapidly deteriorate.
Care should be exercised in assuming that acceptable
solids loads are constant (since they decrease with in-
creasing MLSS values once hindered solids settling con-
ditions are reached) and that the clarifier solids flux
models are complete (since they ignore effluent SS, their
probable statistical variation and the sludge blankets that
must be achieved to match the desired return sludge
concentrations). Simple solids flux models (when applied
to circular units with suction pickup) do work productively,
as several field scale trials have proven. If scraper
mechanisms are used, solids flux models break down as
the effective thickening area is reduced. Without up-
stream flow equalization or diversion, or unless the design
is for relatively low peak overflow rates, conventional air
activated sludge system designs that anticipate MLSS
concentrations more than 3,000 mg/L under the elevated
flow regimes are at risk unless firm supportive data sug-
gest otherwise; in no case should designs with MLSS
concentrations in excess of 4,000 mg/L be judged accept-
able unless the clarifier has been sized specifically to
handle these high values. With anoxic-aerobic systems*
good SVIs of 80 to 120 ml_/g have been consistently
demonstrated; this easily allows MLSS concentrations of
3,000 to 3,500 mg/L with reasonable clarifier designs.
Similar criteria for pure oxygen (and possibly selector
technologies) are advocated by some to be 30 to 50
percent higher, but this is not generally agreed upon.
The return sludge rate and assumed return solids con-
centration are probably two of the most generally abused
design and operating parameters found in today's prac-
tice. Rates must be set to meet needs and do not simply
reflect an arbitrary percentage of flow. Conceptual iden-
tification of needs starts with the understanding that, in
the ideal situation, there is no sludge blanket. With air
activated sludges, best operation (lowest effluent SS) is
typically achieved with a return sludge solids concentra-
tion of around 5,000 mg/L, Without flow equalization, de-
signs that rely on a return sludge concentration in excess
of 7,500 mg/L for a sustained period of time should be
used with caution. Designs that need a return sludge
concentration in excess of 10,000 mg/L for more than a
day should be judged generally unacceptable. Again,
similar criteria for pure oxygen (and possibly selector
technologies) are sometimes 30 to 50 percent higher, but
even these systems can have problems with poor settling
sludge.
The guidance in the preceding paragraphs will serve the
user well in general practice. Unacceptable excursion of
elevated MLSS concentrations will be avoided by the
provision and use of a reliable, tolerant sludge handling
system and disposal outlet. This guidance inherently as-
sumes that the purpose of the clarifier is to clarify not
thicken. Sidestream thickening, exclusive of storage of
waste activated sludge, yields the greatest assurance of
mainstream processing success.
2.7.1,2 Attached Growth Systems
Attached growth systems also have their solids-liquid
separation problems. These problems are found with dif-
ficult to settle colloidal suspensions and sudden, sus-
tained sloughing from the reactor. The following
paragraphs address these issues.
Colloidal suspension problems are best addressed by
remembering the principles of flocculation. Effective floc-
culation is sensitive to several variables: solids contact
opportunity, time, nature of the material to be flocculated,
and mixing energy. The opportunity for solids contact is
most important and, with trickling filter and rotating bio-
logical contactors, can be achieved through a flocculator
that is external or internal to the clarifier, with a settled
solids return. The design range of operating solids in
these solids contact systems should be more than 500
mg/L and no more than 1,500 mg/L if there is no other
reason to maintain a higher operating level. Contact times
should probably be no less than five minutes at the con-
trolling maximum flow regime. The solids retention time
is also important. One day is typical for TF (plastic media)
systems; longer aerated solids contact detention times or
the use of reaeration tanks may be required. Because
the floe is weak, a flocculator center well is often used
to compensate for breakup in the transfer line. Full-
scale studies have shown improvements in super^
natant SS of 5-15 mg/L by use of the flocculator center
well (38). Energy values should be on the order of 50-100
m/s-m (fps/ft).
The uncontrolled sloughing problem that is frequently en-
countered with trickling filters (more so than with rotating
biological contactors) can be related back to organic load-
ing conditions and inadequate flushing rates. High or-
ganic loads promote the formation of dense growths of
organisms in the lead media sections of the treatment
unit. These growths can build up to either block the media
openings or tear from the media and completely or frac-
tionally end up in the system's effluent. Excessive slough-
ing can quickly overwhelm a downstream reactor;
problems at the clarifier are less certain. If the down-
stream reactor is called upon to nitrify, performance may
be compromised as a result of oxygen transfer limitations
or lower nitrifying solids residence times induced by the
influx of solids.
57
-------
Avoiding sudden sloughing of the biomass is best done.
by promotion of a more continuous, controlled sloughing.
One approach is to increase the recycle rate. As de-
scribed in Section 2.6.1.1, if this is an inadequate or
unavailable option, a flow distribution system can be con-
sidered that is independent of processed flow for TFs and
RBCs. With an RBC system, an air scour capability can
be incorporated.
2.7.2 Phosphorus Removal
The following subsections highlight considerations that
may be encountered when applying phosphorus removal
in concert with nitrogen control technology. The reader
desiring more information is referred to the MOP 8 (1),
MOP FD-7 (39), and the Design Manual for Phosphorus
Removal (25).
2.7.2.1 Metal Salt Addition
Metal salt addition for phosphorus removal is commonly
encountered throughout the United States at facilities with
and without nitrogen control. Its impacts on the design of
nitrogen control systems are the result of additional solids
production and alkalinity demand. Aluminum salts make
less sludge than iron salts, but often do not dewater as
well; the alkalinity demand for both is the same on a molar
basis. Table 2-3 provides stoichiometric information relat-
ing to these two processes.
The designer and user are reminded that phosphorus
immobilization (whether by metal salts or biological en-
hancement) causes the plant's SS phosphorus content to
increase (often double) and that excessive dosages, be-
yond 1.5 to 2.0 moles of metal ion per mole of available
phosphorus (less background and synthesis removals),
do little to enhance soluble phase phosphorus removal if
there is no accompanying decline in the pH. Often, plants,
especially the small ones, add more metal salt than nec-
essary for reliable phosphorus removal.
In alkalinity-poor wastewater, metal salts are best added
to the effluent from the nitrification reactor. Here, target
alkalinities of 20-30 mg/L as CaCO3 are acceptable, and,
if approached, may yield phosphorus removal benefits
due to the declining pH. Otherwise, the target residual
alkalinity from the nitrifying reactor should be at least 50
mg/L as CaCO3.
The addition of iron salts results in more precipitated
material than aluminum salts. The chemical feed system
should be sized to deliver both. The ideal point of addition
is to the biological reactor effluent before clarification.
Split treatment, (i.e., using multiple dosing points located
at unit processes designed for solids-liquid separation)
often yields lower overall use and higher performance.
Such strategies may be - appropriate with suspended
growth systems (applications to the primary clarifier influ-
ent and MLSS effluent) to minimize the inert solids that
are carried in the biological reactor.
The appropriateness of metal salt applications to a pri-
mary clarifier that is followed by an attached growth re-
actor is conceptually questionable, since the formed inert
solids that carry over from the primary clarifier may mask
active media surface areas. However, such concerns may
have no practical significance given the abundance of
more significant variables that influence process perform-
ance.
Polishing dosages of metal salt (alum more often than
ferric) can also be added ahead of any final effluent filter
to achieve both capture of colloidal fines that would oth-
erwise escape the filter and additional phosphorus re-
moval benefits. Metal salts and polymer provide an
excellent means of achieving a low SS effluent.
With suspended growth systems, combined or separate
metal salt and polymer application strategies with primary
treatment are always worthy of consideration when cold-
weather nitrification requirements can result in large
downstream reactors because of the high biomass that
must be carried to meet the longer solids residence time
requirements. This approach is especially attractive if de-
sign flows are uncertain and are substantially in excess
of present flow conditions, or if an existing plant's non-
compliance problems are restricted to colder months.
2.7.2.2 Biologically Enhanced Systems
Biologically enhanced phosphorus removal has advanced
through the last decade to the point that in all cases it is
"worthy of consideration for design," which means that
the technology must stand or fall on its own merits. Ulti-
mately, it may be substrate limited (both for forcing the
reaction and for waste solids production) and nearly, if
not completely, dependent on how the waste solids are
handled in the solids processing train.
Biologically enhanced phosphorus removal needs the for-
mation of fatty acids. These can be best generated in an
advance anaerobic contactor or a sidestream anaerobic
fermentor. Fatty acid generation also can occur in the
denitrifying anoxic reactor (paralleling anaerobic reactions
within the interior of the sludge floe or biofilm), but only
with greater, albeit acceptable, conceptual risk. The wis-
dom of an anaerobic fermenter in terms of nitrogen re-
moval considerations is discussed in Section 2.6.1.3.
Biologically enhanced phosphorus removal technologies
fit naturally with suspended growth single-sludge systems
designed for carbonaceous removal; but they are also
compatible with nitrifying systems. Their concepts are
also used with anaerobic and anoxic selectors to avoid
bulking sludges. The process's adaptability to attached
growth systems is just emerging, but would appear to
necessitate some period of SS contacting, and likely use
of an off-stream acid fermenter.
Low-level influent phosphorus concentrations may make
either biological enhancement or metal salt addition the
58
-------
technology of choice. A high-level phosphorus influent
may make metal salt additions mandatory whether or not
enhanced biological phosphorus removal strategies are
incorporated.
With biological enhancement, the incremental phospho-
rus waste solids production numbers are some 55 and
75 percent of the values predicted for iron and aluminum
metal salts, respectively. These waste solids are inher-
ently more unstable than the precipitated metal hydroxy-
phosphate solids, since the additional phosphorus
removal is a stress-induced, temporary cellular storage
product. Some of the enhanced biologically removed
phosphorus returns to solution under extended periods of
aeration (mainstream and digestion), and short (hydroly-
sis) and long periods (stabilization) of anaerobic treat-
ment. Ultimately, the success of the technology is clearly
a function of how the biologically enriched, phosphorus-
laden solids are processed and removed from the system.
The ideal, generalized processing train would incorporate
direct wastage to a dissolved air flotation thickener, fol-
lowed immediately by solids dewatering and stabilization
of the concentrated product. All other approaches, in
some way or another, present the opportunity for addi-
tional compromise. Sometimes the beneficial formation of
magnesium ammonium phosphate will occur during an-
aerobic digestion. This, however, may replace the solubi-
lization problem with a scaling precipitate, a problem that
appears to be most commonly associated with extremely
long digestion detention times and may correlate with the
mineral hardness of the native water.
The present level of knowledge about this technology
suggests that, although solubilization does occur, it is rare
that the phosphorus removal efficiency would be reduced
to levels equivalent to background synthesis values
(about 0.02 mg P/mg VSS). Given the present level of
uncertainty found in practice, a metal salt addition capa-
bility would be recommended for backup. The safest de-
sign of the biological reactor for nitrogen,control would
downrate, if not ignore, the phosphorus removal credit
assigned to biologically enhanced phosphorus removal.
One would then assume that the required degree of phos-
phorus removal might still have to be achieved by metal
salt addition and account for the resultant inert SS in the
reactor. Alternatively, if there is competition for the avail-
able wastewater carbon between biological phosphorus
removal and biological nitrogen removal, direct methanol
addition should be considered to satisfy the carbon de-
mand for denitrification. This would minimize (or elimi-
nate) the need for chemical addition (and its associated
sludges) in conjunction with phosphorus control.
2.8 Nitrogen Control Troubleshooting and
Problem-Solving
Throughout this chapter, an attempt has been made to
illustrate the concept that "everything is connected to
everything else." In nitrogen control system troubleshoot-
ing and problem-solving, the investigator is equally
served by remembering that "all things ultimately have to
be what they are, not what they are assumed to be." In
the real world (40):
• Each treatment plant is absolutely unique.
The variables are almost limitless: e.g., customers (yearly,
seasonally, weekly, daily, hourly variability), collection sys-
tem (flat, steep, expansive, small), climate (ground water,
precipitation, temperature), plant processes (mainstream
and sidestream, hydraulic and solids detention times,
overload, underload), auxiliaries, pumps and blowers
(use and nonuse, down for repair, out of service), sound
and well-intended operating strategies (continuous and
discontinuous, attended and nonattended), recycles and
solids disposal.
• Process models often are regression (empirical) mod-
els of no mechanistic import and no general validity.
Empirical is not a dirty word. It is the translation of the
experience of practice, or observation of the end result,
to describe or predict complex things. They are site spe-
cific and are as good as the extent and quality of data
used to develop them.
• Mechanistic models are valuable teaching tools but
usually cannot be verified and the cause-and-effect
relationships underlying the observed phenomena may
not be known.
Be aware of their guidance but beware of the "apparent"
validation. Remember that each treatment plant is unique;
model constants are derived from empirical observations
and refined through the implicit regression of experience.
If a real nitrogen control problem is encountered, reread
Tables 2-2 and 2-4 and Sections 2.6.1 and 2.7 (and Sec-
tions 2.9.3.3 and 2.9.3.4 may also be helpful). Then, de-
pending on the problem, read Chapters 3 (nitrification)
and/or 4 (denitrification). The applicable experience re-
ported in Chapters 6 through 8 may also be helpful. Then
consider reading Chapter 5. While being mindful of the
introduction to this section, nitrogen control or related
troubleshooting and problem-solving can begin.
Beginning with some general guidance for problem-solv-
ing, the following subsections describe some of the most
commonly encountered problems with nitrogen control
technology. All but one are associated with nitrogen oxi-
dation, since satisfaction of this objective is key to nitro-
gen removal by denitrification, and the majority of the
design applications exclude nitrogen elimination. Often
the problems can be anticipated and mitigated in design.
59
-------
2.8.1 General Guidance
Troubleshooting and problem-solving start with the acqui-
sition of knowledge. This knowledge is used to under-
stand and define the problem and its cause (or causes).
Solutions readily flow once the problem and the cause
are Identified.
Regardless of the particular situation, problem-solving fol-
lows from a readily identifiable sequence of activities.
Table 2-9 provides step-by-step guidance for wastewater
treatment plant problem-solving. Its applicability is broad.
The five offered steps are always necessary. Thereafter,
the questions and comments may not always be appro-
priate because they depend on the problem at hand. The
questions and comments under each step are the ones
that have been generally found to be appropriate. They
are presented with a view to stimulate. It is recognized
that completeness is an unobtainable goal.
2.8.2 Microbtal Inhibition and Pollutant Breakthrough
2.8.2.1 Toxicity
Microbial toxicity due to inhibitory compounds is probably
the most discussed but least real problem encountered
in municipal wastewater treatment. Section 3.3.5 of this
manual provides specific information on a limited number
of inhibitory compounds and their effects on nitrification
kinetics. Reference 1 provides additional information. In-
dustrial pretreatment programs can, and are legally re-
quired to, address this problem when one exists. The
community that consciously fails to do so is as legally
liable as the industry.
This is not to say that microbial inhibition does not occur,
but only to point out that the cause more often than not
Is environmental (e.g., pH and DO), not inhibitory wastes.
If gross microbial toxicity is suspected, a quick check of
the operating DO level and BODS:COD of the raw waste-
water and that applied to the reactor may determine if it
is a valid postulate. Under acute toxic conditions, the
reactor's operating DO level could suddenly go up (dou-
bling or more) and the BOD5:COD could suddenly go
down (50 percent or more) from background levels. If both
are not observed, acute microbial toxicity is unlikely. Mi-
croscopic examination of the mixed liquor could also be
helpful.
The BODS:COD check may not apply for toxic upset of
nitrifying organisms, which are generally recognized
as the most sensitive of the flora routinely encountered
In wastewater treatment. However, toxic microbial inhibi-
tion could still be demonstrated by a sudden rise in the
reactor's DO with an accompanying significant increase
and decline in the plant's normally occurring ammonium
and nitrate-nitrogen levels, respectively. This sensitivity of
nitrifying organisms to toxics, in toxic amounts, holds con-
ceptual promise as a continuous bioassay or biomoni-
toring device. Chronic toxicity, such as experienced by
heavy metal accumulation, will develop slowly and may
not be fully realized prior to 2 to 3 times the solids resi-
dence time.
2.8.2.2 Soluble BODs Breakthrough
Two causes that yield a soluble BOD5 breakthrough are
inadequate oxygen supply and inadequate biological
treatment for the received biodegradable substrate.
The oxygen supply is rarely the cause of soluble BOD5
breakthrough, especially in suspended growth systems.
Oxygen transfer efficiencies increase with declining DO
values. The biomass has some sorptive affinity for soluble
substrates and heterotrophic organisms retain a soluble
substrate removal capability in the absence of measur-
able DO levels, through either anoxic (e.g., denitrification)
or anaerobic pathways. This leaves inadequate biological
treatment as the most likely culprit for soluble BOD5
breakthrough.
Inadequate biological treatment for the received biode-
gradable substrate starts with the solids residence time
of the reactor and ends with the substrate and its activity
and sources in the collection system. The reactor's re-
sponse to any unique foreign substrate, like methanol
(see Section 2.6.1.3), is only as good as the time and
dosage (mass) provided for acclimation. A discontinuous
release of a unique foreign biodegradable substrate (e.g.,
as with an industrial processing run or an end-of-the-week
cleanup) can mean the absence of an acclimated culture
that may have responded in some fashion to the last
discontinuous release.
2.8.2.3 Ammonium Breakthrough
Ammonium breakthrough from a plant that appears to be
safely designed on the basis of solids residence time (or
rate) is generally associated with one or more of the
following: inadequate oxygen transfer, inadequate alka-
linity, too high an ammonia concentration for the accli-
mated biomass, solids washout, and toxicity.
Inadequate oxygen transfer is more a problem with at-
tached rather than suspended growth reactors, but can
be experienced in both. For problem-solving, it is well to
remember that nitrification cannot occur without DO and
that nitrifier activity can be highly variable when measured
operating DO levels decline below 2 mg/L. The gross DO
measured in the fluid is only an indirect measure of what
is actually available in the interior of the biological floe or
film. Organic load spikes or diurnal swings act to sup-
press and delay nitrifier activity when the DO is depleted
by the more dominant response of the viable heterotrophs
(see Sections 2.6.1.1,2.6.1.3, and 3.3.3). Fundamentally,
the solids residence time of the system is of no conse-
quence to the nitrifiers if a sufficient period of proper
oxygenation is not available to ensure their replication.
Inadequate alkalinity ultimately leads to a pH depression.
Nitrifiers can and do acclimate to lower pH conditions, but
60
-------
Table 2-9. Step-by-Step Recommendations for Wastewater Treatment Plant Problem-Solving
WHAT IS THE BASIS FOR SAYING THERE IS A PROBLEM?
Avoid Stupid Mistakes—Check analytical procedures and quality assurance/quality control.
Analytical kit procedures often differ from more rigorous analyses. Are checks by an independent laboratory appropriate? Does
an independent laboratory know what it is doing? Does the problem coincide with a change in analyst or analysis? Does the
problem lie within the variability of the test? Is there a need for triplicates? Do operations set up or conduct tests on weekends,
while the laboratory people do this on weekdays? Are BODs performed with inhibition of nitrification? Are BODs reseeded after
chlorination?
Is the Overall Problem Understated or Overstated?—Check sampling.
Often a problem is understated. Weekends are often the worst time [Friday (end of the week) industrial cleanups, sampling
and operation most relaxed; worst impact from solids processing train, since what enters the plant on Friday, leaves the plant
processes on Saturday (or Sunday)]. Twenty-four-hour composites mask the severity of breakthrough and dumps. Are samples
manually or automatically collected? Single grabs or composite grabs? What is the compositing period? When are samples
collected? What are the procedures? Are reported samples and flows what they are identified to be (reported raw flows often
include recycles, reported raw wastewater often includes recycles)? Remember nitrogen transformations (organic nitrogen
hydrolizes to ammonium; nitrite-nitrogen is encountered in both oxidative and reducing systems). Do automatically collected
sample containers overflow? Are final effluent samples collected after chlorination? Guard against preferential or lax sampling
and failure to report weekend conditions.
WHEN WAS THE PROBLEM FIRST OBSERVED, AND WHAT WAS DONE ABOUT IT?
Often the problem was not observed when it first occurred, or the reasons for the problem have something to do with something
that happened in advance of the problem. Look for subclinical signs, several months before the problem was first reported.
Remember the great buffer found in the liquid returns and recycles, hydraulic volumes, and solids (the process and plant sludge
age) found at the plant. Remember that mixtures take about 2.2 residence times to come to 90-percent equilibrium with new
stimulants or operations.
Determine remedial cures that have been tried and the results of these tries. Believe the data more than the personal
interpretations. Understand what the words mean (e.g., turned on, speeded up, faster, tried). Remember people present
.themselves in the most favorable light and sometimes, unknowingly, suppress valuable information until the right question is
asked. '
DO PLANT OPERATIONS CONTRIBUTE TO THE PROBLEM?
Look particularly at all discontinuous operations and the reality of all recycles as they are experienced in the main flow stream.
Check solids mass out the door (effluent and to disposal), against solids mass in (raw wastewater and inert solids formed from
added chemicals). Don't overreact to sidestream pollutant concentrations until their mass significance is known. Prepare
approximate mass balance for guidance. Perform inert solids balance. Understand what the processes see (pollutant mass
and concentration) as they actually see it, and how the process responds to it (e.g., complete mix or plug flow). Know what
equipment was down and why. Know if artificial constraints were imposed (e.g., air supply, pumping, chemical feeds). Determine
factors about operation that make equipment operate the way it does.
IDENTIFYING THE CAUSE OF THE PROBLEM: RESPECT EVERYONE'S OBSERVATIONS, MISTRUST EVERYONE'S
INTERPRETATIONS
Remember that the right solution already may have been tried, just not long enough, and that people are smarter and less
smart than they appear. Often the correct solution to the problem has been identified by someone; often the person doesn't
know it. Talk to everyone, ask what they observe, and see what they think. Avoid preconceived ideas. Be open to all. Try to
develop a scenario that matches all observations. Do not eliminate any observation that doesn't fit the scenario until you can
confidently do so. With elimination of irrelevant observations and retention of all others, the remaining scenario probably correctly
identifies the cause(s) of the problem. Expect that it is related to more than one thing, and catalyzed (or most observable) by
only one to a few.
Mistrust yourself. Remember the uncertainties and check concepts by the total wealth of experience. Say when you don't (and
the field doesn't) know. Get help if still confused, or if it is logical to expect that somebody knows more about a critical something
than you do. It is illogical to expect anyone to know everything.
SOLUTIONS TO THE PROBLEM: EASY TO IDENTIFY, OFTEN HARD TO IMPLEMENT
Until demonstrated for a year or more, all solutions have risk, some more than others. Try to develop solutions that don't have
the potential of making anything worse. Remember that bigger is not necessarily better. Seek a rifle, not a shotgun/Consider
everything that can be done in the collection system (operationally), then everything that can be done structurally. Prioritize in
terms of certain benefit. Define risks of each in some qualitative way. Think of ways to minimize risk. Develop fallback procedures
before they are needed. Think of costs as if your money were at stake.
Share this entire thought process with decision-makers. Identify key issues and concerns. Determine what results constitute a
fully acceptable solution. Identify options to achieve desired results. Agree on a remedial program implementation strategy.
Take remedial action steps with a clearly defined flexible plan of action, with known fallback positions and points of possible
irretrievable misuse or loss of resources that can serve as new decision points.
61
-------
not excessively so (usually less than 6.0 to 6.5). When
low pH conditions are encountered, the nitrifiers are in-
hibited in the rate of their reaction. If inhibited too fre-
quently, the total nitrifier population will decline because
of washout, regardless of the operating DO. In diagnosing
pH problems, use grab samples before flows drop over
weirs (this fall can cause the pH to rise, as can the routine
laboratory procedure of shaking the sample thoroughly
before performing any analysis, due to CO2 stripping; see
Section 2.5.1 for more understanding). Places to look for
alkalinity deficient, low pH problems are discontinuous
recycles and acid dumps in the collection system. Routine
monitoring of the effluent alkalinity often provides valu-
able insight as to irregular occurrences.
Closed carbonaceous oxidation systems, such as en-
countered with pure oxygen technology, can contribute to
low pH problems since CO2 stripping cannot occur. Simi-
larly, fine-pore aeration systems may have a lower pH
than coarse bubble systems, and rotating biological con-
tactors may have a lower pH than trickling filters. Ulti-
mately, the pH of any system with adequate alkalinity for
nitrification will reflect its equilibrium with the CO2 pro-
duced and the atmosphere, as determined by the physical
opportunities for CO2 stripping.
2.8.2.4 Oxidized Nitrogen Breakthrough
The anoxic denitrification reaction is accomplished
through the use of driving substrates (externally added
substances or untreated wastewaters) and/or endo-
genous respiration of cellular material. Oxidized nitrogen
breakthrough is generally associated with one or more of
the following: particularly low applications of the driving
substrate, failure to acclimate to the dosage of the driving
external substrate, competition for the driving substrate,
lower rates than expected, and physical limits of the proc-
ess. Responses are self-evident and essentially translate
to provide more substrate, more time for acclimation,
spread the substrate out to more places of application,
and use more solids and/or reactors (see Section
2.6.1.3). Tracking of the soluble COD, organism oxygen
uptake rates, and oxidized nitrogen forms into, through,
and out of the reactors will provide valuable insight for
problem-solving.
2.8.3 Suspended Growth Systems
2.8.3.1 Unwanted Seasonal Nitrification (and Bulking)
The elimination of nitrification can be just as important as
its promotion for cost-effective treatment. Strategies for
its elimination are the reverse of its promotion. They begin
with the direction to drop the solids residence time to a
point that is acceptable to satisfy the applicable process-
ing objectives but is insufficient to sustain nitrification.
Waste solids production levels may marginally increase,
but it is nearly always cheaper to process and dispose of
the extra sludge than to use oxidation energy for its de-
struction.
Methods for dropping the solids residence time include
dropping the MLSS. Do not be afraid of lower values than
those found in the textbooks; but for most sludges, go no
lower than 500 mg/L because of flocculation concerns.
For sludges with high SVIs, even values below 500 mg/L
can work well with a goal of maintaining an initial settling
velocity no higher than about 3.6 m/hr (12 ft/hr). Other
strategies involve taking reactors out of service, provided
oxygen transfer capacity is adequate in the remaining
reactors. The best design is the one that allows this to
be done without loss of the downstream clarification
capacity.
Do not over aerate and be careful if the solids residence
time drops below three days since bulking sludges might
result. Consider the use of return sludge chlorination for
specific nitrifier kill. At low MLSS concentrations, remain
sensitive to the possibility of detergent foam and provide
(or use) the spray water capability.
Most municipal wastewater treatment plants probably op-
erate at a 9C beyond their needs; especially those that
are designed for secondary treatment equivalency and
are significantly below their design load. Often, inroads
of nitrifiers occur in the spring, acting to depress the DO,
and result in seasonal bulking. If this happens, remember
why. Do not unnecessarily add more air; instead, reduce
the MLSS and solids residence time. Find a safe reason-
able minimum for the plant, its service area, and the
effluent objectives.
2.8.3.2 Scum (Floatables) Management
Progressive use of high 6C, nitrifying single-culture acti-
vated sludges over the last two decades has also brought
the coincidental observation of Nocardia froth. Perhaps
coincidentally, this time period has also seen the manda-
tory regulatory requirement of secondary clarifier baffling
and scum capture.
The designer should anticipate Nocardia froth, or process-
generated floatables, with combined carbonaceous oxi-
dation/nitrification activated sludge systems. It is not
generally observed with low-sludge-age air systems, nor
isolated nitrification or denitrification suspended growth
systems. It may be endemic with pure oxygen systems
and their submerged reactor effluent drawbffs. Excessive
froth formation is often encountered in the combined
sludge anoxic reactors with overzealous use of baffling.
Scum (or floatable) problems with attached growth sys-
tems or coupled applications of an attached growth sys-
tem followed by suspended growth reactors have not
been reported to date.
Process floatables are best managed by the philosophy
that seeks to avoid their return to the processing train by
separate processing through the point of ultimate dis-
posal. If this is not possible, then they should only be
slowly introduced, after concentration, to the sludges de-
livered to the dewatering process. In the liquid processing
62
-------
train, the obvious key is to avoid froth entrapment until
planned for. The first likely place for entrapment is at the
mixed liquor effluent channel. Here, consider using its
natural flotation to assist in its removal. Downstream, the
sedimentation tank should have full-width tilting troughs.
Elsewhere, at all other upstream and downstream points
of entrapment, consider covers so scum can be seen or
allowed to escape, and the use of sprays where it might
be observed. A number of facilities report some success
in chlorinating the spray water.
2.8.3.3 Fine-Pore Diffuser Clogging
The rising popularity of fine-pore diffusers may bring an
accompanying problem with clogging. Cost analyses gen-
erally always show that very frequent cleaning can be
justified (41).
The clogging problem appears to be most severe where
the applied load is the highest and the DO levels the
lowest. The problem may be exaggerated with combined
carbonaceous and nitrogenous oxidation systems be-
cause of the collective mixed liquor oxygen demand of
the combined heterotrophic and autotrophic bacteria.
Clogging further exacerbates the problem at the front end
of the reactor and the inadequate DO levels transfer it to
the later stages of the system.
Beyond cleaning, the following mitigation measures are
possible: drop the operating solids level, increase the air
supply, use selector technology, use coarser diffusers or
alternate oxygen dissolution equipment where the oxygen
demand is greater, or spread the load across the length
of the aeration tank. Designs should anticipate the pos-
sible need for frequent cleaning of the front-end diffusers
in plug flow reactors and always provide for convenient
access and rapid drainage of aeration tanks.
2.8.4 Attached Growth Systems
2.8.4.1 Oxygen Transfer
Understanding attached growth systems starts with un-
derstanding oxygen transfer potential and experienced
oxygen demand characteristics (see Section 2.6.1). The
oxygen supply is largely fixed by the reactor and its me-
dia. The dissolution is determined by the fluid applied to
it, the fall, the driving deficit, and, in the case of RBCs,
by the speed of revolution of the contactor. The uptake
is influenced by the localized concentration of biodegrad-
able substrate (both nitrogen and carbon) and localized
concentration of heterotrophs and nitrifying autotrophs.
The microbial population at any point in the train will
reflect more the average conditions than discontinuous
peaks, but will respond to any transient condition.
Heterotrophic reactions, if possible, will dominate nitrify-
ing reactions because of less sensitivity to DO, faster
growth rates, and their greater number. Temperature im-
pacts on nitrifiers will only be experienced after the
achievement of nonlimiting DO conditions. Heterotrophic
activity will be catalyzed at a faster rate in response to
temperature impacts since heterotrophs can perform
aerobic, anoxic, and anaerobic reactions as required and
will be made possible by the depth of growth and the
biodegradability of the material within the growth.
The oxygen demand imposed by the film will vary accord-
ing to current and historical conditions. Throughout the
year, the entire system will be in a state of flux responding
to the localized concentration of pollutants and other en-
vironmental conditions such as temperature, DO, and pH.
Often the lowest seasonal flows will coincide with the
warmest temperatures, yielding concurrence for the most
limiting reactor conditions (highest localized reactor oxy-
gen demands per unit time) with the highest nitrogen
concentration, at a time when effluent standards are the
most stringent.
\
Nitrifier activity will follow wherever the heterotrophic ac-
tivity will allow. The exertion of the nitrogenous oxygen
demand will be controlled by the number of nitrifiers en-
countered in the processing stages. The population will
be limited first by concurrent heterotrophic oxidation, then
by the oxygen transfer of the system, and then by the
available ammonium nitrogen.
Although heterotrophic biomass is not essential for nitri-
fier attachment, the heterotrophs form biogrowth to which
the nitrifiers adhere. Biogrowth formation and attachment
will become progressively more fragile and spotty as the
available carbonaceous substrate and heterotrophs de-
cline. Biogrowth formation in the earlier stages of a com-
bined carbon oxidation/nitrification process will build, and
will either fall off and/or block void space used for oxygen
transfer. This has the potential for permanently or tempo-
rarily blocking flow and air passages, creating new ran-
dom flow patterns, and can serve to create another
complex set of conditions influencing the performance of
the entire system.
The overall "simple to operate and maintain" attached
growth system belies the complexity of the reactions that
occur within it. Precise attempts to predict performance
may be futile. Remedial strategies for the noncompliant
system will surely encompass one or more of the follow-
ing:
• equalization of loads (upstream or as applied to the
reactor),
• split treatment to increase substrate to starved het-
erotrophs in order to form more biogrowth that the
nitrifiers can adhere to,
• instream and/or isolated sidestream cycling of reactors
to force more ammonium to where the biomass sees
little (consider controlled digester supernatant as a
source of ammonium to force more growth of nitrifiers;
be careful of alkalinity needs),
63
-------
• cycling load reversals where possible to maximize ni-
trifier growth,
• higher recycle flows for dilution of the soluble substrate
and higher oxygen transfer, and/or
• forced ventilation.
2.8.4.2 Biogrowth Control
Technology exists to control excessive biogrowth in com-
bined carbon oxidation/nitrification systems. For a trick-
ling filter, the ideal situation may be the ability to control
the localized application rate independent of the flow and
recycle through an electric drive on the distributor arm. If
not available, the next best approach may be to slow
down the arm by reversal of the distribution ports. Con-
ceptually, this approach may bring maintenance needs in
conflict with the retention time needed for performance.
For a rotating biological contactor, maintenance needs
are easily satisfied with air scour.
Both approaches reflect a common desire to scour off
excess biogrowth before it becomes a problem, thus
achieving a constant controlled sloughing, rather than
living with the unpredictable whims of the system and the
unpredictable impacts on the biological culture and the
physical media.
Insufficient biogrowth can only be resolved by the intro-
duction of additional substrate. Controlled additions
should assist in the formation of suitable conditions for
retention of the autotrophic nitrifiers. Strategies include
higher localized flow rates, split treatment, and isolation
and submergence.
2.8.4.3 Nuisance Organisms
A high degree of nuisance organism control is achieved
by the scouring procedures described in the previous
subsection. A capability to chlorinate the recycle water
also seems appropriate.
Some systems, particularly nitrifying systems, seem to be
bothered by snails. The degree of control that can be
achieved by scouring procedures or media selection is
uncertain. Peoria, Illinois, has had some success in con-
trolling snails in its nitrifying RBCs by bypassing a portion
of the primary effluent directly to the affected RBC units.
The best strategy is to assume that snails will occur, and
to provide a baffled spot with a depression and sump in
the line immediately after the attached growth system for
their capture and accumulation. The simplest manage-
ment scheme for the captured snails may be to deliver
them to the head of the plant's degritting operation.
2.8.5 Transient Chlorination Demands
A common problem encountered at many wastewater
treatment plants is a sudden loss of the chlorine residual
and a dramatic increase in the chlorine required to
achieve either a regulatory set or an operationally set
residual. This demand is associated with both chlorine
oxidation of nitrite nitrogen and breakpoint chlorination
with low ammonia levels.
The nitrite-nitrogen transient chlorine demand is typically
encountered at plants that chlorinate year-round and
move from non-nitrifying to nitrifying conditions. It is also
encountered in nitrifying plant startup. During these con-
ditions, the nitrite nitrogen is present only briefly (as much
as a week or two) as an intermediate point (or temporary
operating condition) in establishing a stable nitrifying
population. Appropriate operating strategies range from
living with it (i.e., placing the chlorination system on man-
ual and not worrying about the chlorine residual as long
as the required level of disinfection is met), to getting out
of the condition as fast as possible (i.e., going to a
younger solids residence time operation to eliminate ni-
trification, or dropping back on the sludge wastage as
much as possible to promote the rapid creation of a stable
nitrifying population).
Nitrite-nitrogen transient chlorine demand problems
promise to increase in frequency as more polishing deni-
trification technologies are applied to wastewater treat-
ment plants, since nitrite-nitrogen is an intermediate step
in the denitrification reaction. Its presence here is symp-
tomatic of too little carbonaceous material to drive the
denitrification reaction to completion, or inadequate accli-
mation or reactor contact time. In this application, sudden
excursions of the chlorine demand can be used as an
immediate alarm that something may be wrong at the
denitrification step.
At disinfection doses of about 10 mg/L of chlorine, a plant
averaging 1 mg/L of ammonia nitrogen will go into break-
point chlorination during significant portions of the day
and experience periods with no residual chlorine and no
disinfection. Some plants have had to overdose chlorine
(30 mg/L) to overcome the problem which is more com-
mon than the nitrite problem.
2.9 The Design Examples
2.9.1 Introduction
Sound process design concepts and considerations are
best illustrated through specific examples. This section
was prepared with this objective in mind. Its purpose is
to illustrate the design activities that precede the detailed
evaluation of any unit process and to provide a common
design condition for the unit process nitrogen control de-
sign examples in Chapters 5 through 8.
2.9.2 Treatment Facilities for the Design Examples
The great majority of U.S. wastewater treatment plants
now (and anticipated in the future) range from 4 to 440
Us (0.1 to 10 mgd), with about 65 percent of the plants
treating about 30 percent of the flows. Figure 2-4 sche-
matically characterizes "simple" and "more complex" ge-
64
-------
Plant A (Simple)
| Flow Equalization \-ff\
Plant B (Complex)
Figure 2-4.
440 Us).
Land
Application
Basic schematics of "simple" and "complex" wastewater treatment facilities with a typical flow range (4 to
neric wastewater treatment facilities that may be typically
encountered in this flow range. Other than the possible
provision for alkaline stabilization of the dewatered
sludges to address newly promulgated sludge disposal
criteria, the processing schemes are believed to illustrate
the technologies that are routinely incorporated into
plants of this size for levels of performance equivalent to
secondary. These facilities (designated as Plants A and
B, respectively) provide the basis for design examples
found in this manual. Detailed process flowsheets are
presented in Figures 2-5 and 2-6 for the "simple" and
"more complex" process trains, respectively.
2.9.2.1 Commonalities
Inspection of the process flowsheets provided in Figures
2-5 and 2-6 shows a common approach for pretreatment
and minor residuals management (screenings, grit, and
floatable removal and processing), the possibility of flow
equalization, disinfection and detoxification (dechlorina-
tion), ample solids storage (solids equalization), waste
activated sludge thickening, and dewatering and ultimate
disposal with the firm backup of landfill disposal. The
principles demonstrated in these schematics, with respect
to these considerations, are as follows:
• Although small in average quantity (typically, no more
than 3-6 g dry solids/m3 [25-50 Ib/Mgal] raw waste-
water), minor residuals are highly variable, large in
significance, and often a troublesome and demanding
consideration for design and operation. After their re-
moval, the designer and the operator are challenged
as to what to do with them. The designer should plan
for their subsequent handling and processing with the
same care as the major residuals found at the treat-
ment facility. As shown, adding lime will prevent nui-
sance conditions and provide control of infectious
agents. If at all possible, design and operation should
attempt to blend these materials into the ultimate dis-
posal plan for the major residuals generated at the
65
-------
Influent
"Effluent
Similar Separate
Processing
or I—
Seasonal Agricultural
Application of
Product Sludge
Note: Circled numbers indicate process points for mass balance calculations. Note also that schematic is
not applicable to attached growth systems, because they lack all forms of primary treatment.
Rgure 2-5. Detailed schematic of "simple" wastewater treatment facility (Plant A) (see Table 2-15 for mass balance data).
plant site. This is a possibility for the grit. If this is not
possible, then there is no other recourse than to land-
fill; often onsite burial is used.
• Particular care should be taken in managing floatables
derived from the processing train. Process-generated
scum is often particularly troublesome with suspended
growth systems. Raw scum can be readily concen-
trated on standing, or by screens, to concentrations in
excess of 20 percent solids. Process scum also has a
natural flotation to yield solids similar to what would be
expected from a dissolved air flotation thickener (and
this is an excellent point of reintroduction to the sludge
train if the thickened sludge is immediately dewatered).
In anticipation of the accumulation of troublesome
floating residuals, suspended growth liquid and solids
process reactor design should avoid submerged points
of withdrawal. Raw wastewater floatables and those
generated in the biological processes (e.g., Nocardia
scum) are best managed by separate processing up
to the point of ultimate disposal. This avoids their re-
introduction into the processing train. If this is not pos-
sible,, they should only be slowly introduced, after
thickening, to the sludges delivered to the dewatering
process. When ultimate disposal of the waste solids
for beneficial use is anticipated, both the waste sludges
and the floatables should be ground to eliminate the
visual observation of unsightly rubber and plastic arti-
cles common to municipal wastewaters.
• With so much of the liquid processing train design
ultimately dependent on hydraulic considerations, flow
equalization is an attractive concept for process stabil-
ity, especially with smaller plants and those that con-
template suspended growth treatment technology. The
small plant designer and owner should seriously evalu-
ate this consideration.
• The choice of sodium hypochlorite and sodium bisulfite
for disinfection and dechlorination reflects the desire
for safety (easily handled liquids) and minimal risk in
what may be unattended operations (or a poorly
trained and/or supervised staff) at small plants. Addi-
tionally, where nitrification is required to protect the
oxygen resources (and biota) of the receiving stream,
the attendant need for dechlorination is likely. In prac-
tice, use of CI2 and SO2 is also commonly observed.
Ultraviolet (UV) radiation disinfection has become a
broadly accepted alternative to chlorination, particu-
larly when dechlorination would be required. It is well
demonstrated, does not require handling of chemicals,
and leaves no residual. The designer should seriously
consider UV in lieu of chlorination/dechlorination in ad-
vanced plants (nitrification or greater) and/or plants
that incorporate filtration.
66
-------
Flow
Equalize
Trea
nina
tme
1
Screenings
G
1 ' '
Landfill |" '
4
tion
c
ry •*
nt
it
Separate or Comb ned Processing
1 | 1 | 1
I T CaO MeOH
fe © tfih&>
T T T i i
Separation Bjo|ogjca| [
Floatables — i ,-
V.
Separation
I Reactor ' — '
Single to Multiple Objectives
Internal or External , ,
(9) Returns or Recy
T , N Solids and Flu'
* ^ ">
J^
?.
\£/ Polymer |
-~
f^\ 1 1 ^-^
T ^^
Backup
cles *" w —
ds , ^
i
*] Thickening
'I
Solids
Stabilization
Solids
Concentration
and/or Storage
fcU
1 ^-1
Dewatering
•• Storage
H NaOCL |
as Needed
Backwash
Surge Control
^
^4)
/~* SimiU
^5 Sepa
_ Proce
1
k — — t
Disinfection and
Dechlorination
w
rate
ssing
V^ 0,.
Applic,
asonal Agricultural
at on of Product Sludge
Effluent
Note: Circled numbers indicate process points for mass balance calculations.
Figure 2-6.
data).
Detailed schematic of "complex" wastewater treatment facility (Plant B) (see Table 2-16 for mass balance
Often a point of plant failure results from attempting to
achieve too high a return sludge concentration and the
resultant elevated sludge blanket found in the clarifier.
Both the simple and more complex plants show the
use of a separate thickening process for the waste
activated sludge to highlight the importance of side-
stream concentration of this process stream. Attempt-
ing to concentrate the waste activated sludge in the
primary clarifier can often lead to washout of these
lighter solids into the secondary system during higher
flows. This consideration is particularly important with
nitrogen control facilities where sudden influxes of sol-
ids from the primaries may cause a washout of the
nitrifiers due to effluent losses or the need for exces-
sive wastage. If concentration in the primaries is at-
tempted, lower surface overflow rates than routinely
allowed by regulatory guidelines may be appropriate.
With multiple primary installations, alternate strategies
may include flow peak control for select primaries that
receive the waste activated sludge.
Solids storage systems provide the same service as
flow equalization in the liquid processing train as well
as providing strategic responsiveness to uncertainty.
In terms of equalization, the solids storage system
must be matched with the planned operation of down-
stream operations. In small plants, the frequency of
dewatering operations may range from monthly or sea-
sonal intervals to a fraction of an eight-hour shift per
week. Where agricultural applications are anticipated,
solids applications are limited to preparation for the
one or more crops that are harvested per year. Lime
stabilization of the dewatered solids product and wind-
row curing for a year or more solids production may
be reasonable with some beneficial recycle objectives.
The designer should pay particular attention to odors
if lime stabilization is selected. Ammonium will disso-
ciate to ammonia at elevated pH conditions. This is
particularly important when processing raw primary or
anaerobically digested waste solids.
It is important to note the backup of landfill disposal in
the processing schematic. Sound plant design and op-
eration must provide the assurance that residuals de-
rived from wastewater treatment can be eliminated.
67
-------
This is best done by not assuming that the preferred
way of ultimate disposal will be fully, realized.
2.9.2.2 Differences Between the Two Plant Scenarios
Practically speaking, the principal difference between the
two plants is only the presence of the primary solids
separation step (gravity sedimentation) in Plant B. The
introduction of primary sedimentation allows a more op-
timized approach in terms of tankage commitment (and
lower possible construction costs) but only for a more
demanding operation (and higher possible operating
costs). This, of course, is the crux of wastewater treat-
ment plant design: unit process optimization versus op-
erating convenience and cost minimization.
• Primary sedimentation is often linked with an isolated
anaerobic solids stabilization (digestion) system. The
solids stabilization process shown in Figure 2-6 is as-
sumed to be anaerobic digestion in order to illustrate
the often troublesome impact of digester recycles. Sol-
ids stabilization in the simpler Plant A occurs in the
mainstream reactor by aerobic processes, reflecting a
solids residence time which would yield a well-
stabilized sludge (although still restricted in beneficial
use scenarios).
• The reader should note that the Plant A schematic
specifically excludes attached growth system applica-
tions because of the absence of primary clarification,
which is necessary to avoid fouling and possible clog-
ging of the media. In the more complex Plant B, the
required level of protection is classically achieved by
gravity sedimentation, although some form of finer
screening than found in the preliminary treatment
phase of the process may also be worthy of consid-
eration. Fine screening may present a more demand-
ing operation and, at the time of this manual's
preparation, is rarely encountered in U.S. municipal
wastewater treatment applications.
• The mainstream biological reactor in Plant B can ac-
commodate more varied configurations than the sim-
pler Plant A, including either a suspended growth or
fixed film process. It can contain one or more sludge
cultures or mixed applications of suspended and at-
tached growth technologies, in a coupled or uncoupled
manner. Filtration is also shown in the Plant B process
schematic, addressing final effluent polishing. Consid-
eration of the filter backwash is worthy of further
discussion.
• When backwashing is discontinuous, it can represent
a significant bump in the plant's flow; the processing
schematic in Figure 2-6 illustrates the use of a surge
tank for equalization, as needed. The question of
where to return the filter backwash is a design deci-
sion. The processing schematic shows the filter back-
wash returned to the head of the plant for convenience,
but, practically, it could just as easily be returned to
the upstream secondary clarifier (this would avoid the
hydraulic surge through all of the plant but lose the
peak mitigation which is inherent in the reactors and
conduits of the liquid processing train). However, the
filter backwash is also a source of seed organisms
from the upstream processes. In the case of a nitrifying
system, it may be appropriate to return it to the lead
carbonaceous oxidation reactor, and provide an inocu-
lum of nitrifying bacteria throughout the processing
train. Such practices, using filter backwash and/or
waste sludges, are commonly encountered at multi-
culture treatment plants.
• Finally, detailed inspection of the Plant B schematic
shows internal processing stream callouts for the bio-
logical reactor (Callout 4) and the filter (Gallout 7). The
purpose of Callout 4 is to allow a fuller understanding
of the internal workings of the biological reactor(s) as
they may be applied to meet either internal or overall
processing objectives. The purpose of Callout 7 is to
demonstrate the process response to methanol addi-
tions for denitrification within this unit process.
2.9.3 The First Design Steps
Process design follows a series of logical steps. The first
design steps are described in the following paragraphs.
The introductory chapters of the recently revised MOP 8
provide further information and detail for the interested
reader (1).
2.9.3.1 Understand Processing Objectives
The first step in any process design is to understand the
processing objectives. These objectives, from an overall
plant perspective, include the:
• effluent limitations,
• needs (or standards) for residual solids beneficial use
. or disposal, and
• various considerations that influence the design, in-
cluding:
- future planning and service,
- plant and adjacent area aesthetics (sight, sound,
and smell), and
- O&M expectations and realities.
Collectively, the understandings associated with these
processing objectives feed backward into the design of
the integrated facilities and the specific unit processes
found at the treatment facility.
The processing objectives of the design examples are
provided in Table 2-10. Two different effluents are pre-
sented, in both cases, process calculations are to be
preformed for controlling wastewater temperatures of
10°C (50°F), 15°C (59°F), and 20°C (68°F). The first set
of effluent limits establish typical secondary effluent cri-
teria for BOD5 and TSS, whereby the maximum 30-day
68
-------
Table 2-10. Effluent Objectives of the Design Examples
Effluent Limits
(
-------
likely to be some 10 to 25 percent greater than the ratios
of the dry weather flow peaks to average day conditions;
similarly, minimum nitrogen mass loads are likely to be
some 90 to 75 percent of the corresponding dry weather
flow minima. The relationship between ammonia peaking
and hydraulic peaking loads for several cities is presented
in Figure 2-7. Seasonal infiltration and event-specific in-
flow would be expected to have little influence on the
mass of nitrogen received at the plant (except that the
latter may bring in previously deposited suspended matter
in various degrees of stabilization from the collection
system).
Tables 2-11 and 2-12 summarize the influent wastewater
characterizations elected for the design examples. When
developing such characterizations do not allow precision
to overwhelm common sense. Nothing in municipal
wastewater treatment is ever known with greater validity
than within 10 percent of the correct answer, and variabil-
ity is the norm.
The average day characterization is presented in Table
2-11. The values presented are believed to be repre-
sentative of a typical municipal wastewater in the United
States without any unique industrial, commercial, or insti-
tutional contribution. The serviced population is estimated
at about 6,250 individuals per 44 Us (1 mgd) of average
contribution. The average per capita flow reported irr
ERA'S 1986 Needs Survey was about 0.6 m3/person-day
(160 gpcd) (43). Average annual daily per capita SS and
BOD5 emissions in the example are about 0.09 kg (0.20
Ib) and 0.10 kg (0.23 Ib), respectively. Note that U.S.
domestic per capita releases are on the order of 0.26
m3/person-day (70 gpcd); smaller communities tend to be
less subject to infiltration and inflow and, if free from
significant industrial contributions, may reveal signifi-
cantly lower overall per capita flow rates and slightly lower
per capita pollutant emissions.
The volatility of the influent SS reflects the assumption
that the influent BOD5 is lower than the SS, and the
assumed infiltration condition derived from ground and
storm waters in the collection system. Higher influent SS
volatilities would be expected with tighter collection sys-
tems and a strongly domestic service. Higher wastewater
alkalinities are often encountered when the community's
raw water supply is from ground-water sources. Refer-
ences 1 and 39 provide additional background material
for wastewater characteristics and nutrient contributions,
respectively.
u
3
CO
'c
1
03
3
•i
I
1.0
Y = 1.457 X-0.217
r = 0.823
Key
ADWF
Symbol Plant
Sample
L/s
mgd
Primary 53 •) .2
Roughing Filter 149 3.4
Primary 920 21
Primary 1,972 45
Primary 12,010 274
Raw 79 1.8
Raw 44 1.0
Canberra, Australia
Belconnen, paw 440 10
Canberra, Australia
Lebanon, OH
Livermore, CA
CCCSD, CA
Sacramento, CA
Blue Plains, DC
Chapel Hill, NC
Weston Creek,
1.5 2.0
Maximum Hourly Flow/Average Daily Flow
2.5
Figure 2-7. Relationship between ammonia and hydraulic peaking loads for treatment plants with no in-process equali-
zation (from Reference 22).
70
-------
Table 2-11. Design Examples: Average Day Raw Influent Wastewater Characterization
Parameter, mg/L Inert Volatile* Total Soluble
Particulate*
Total
SS 52
% 30
CBOD5
COD
Total Nitrogen
Ammonium Nitrogen
Organic Nitrogen
Total PO4-P
Alkalinity, as CaCO3
123 175
70 100
50
100
24
20
4
5
120
100
180
6
—
6
1
— -
150
280
30
20
1o
6
120
* With the exception of the BOD5, 30 percent of the volatile SS and the related participate pollutants (COD, TKN, and PO4-P) are assumed as
nonbiodegradable.
Table 2-12. Design Examples: Influent Wastewater Peaking Factors
Ratio of Noted Condition
to Average Day Pollutant Mass
Condition
Minimum Month
Average Day
Maximum Month
Maximum Week
Maximum Day
Maximum Hour
Percent of Time
Conditions*
7.7
50.0
92.3
98.1
99.7
99.99
Flow
0.7
1.0
1.5
1.9
2.5
3.0
SS and
Organics
0.8
1.0
1.3
1.6
2.1
Total P
andN
0.8
1.0
1.2
1.4
1.7
Matching
Alkalinity
0.9
1.0
1.1
1.3
1.5
* Equivalent percent of time conditions are less than or equal to stated values.
Table 2-12 provides a reasonable characterization of the
peaking factors needed to complete the process design.
For the purposes of design, it is assumed that the maxi-
mum and minimum conditions can happen at any time
during the year, and that minimum and maximum pollut-
ant masses may or may not coincide with minimum and
maximum flow regimes. Proper design should identify and
evaluate the controlling condition (e.g., maximum load
with minimum or maximum flows).
2.9.3.3 Understand Performance Characteristics of
Existing Operations
Knowledge of the processing objectives and the influent
wastewater characteristics converge at the operations of
the existing or planned plant. With an existing plant, the
designer should make sure that full understanding of all
operations is achieved, including such fundamentals as:
• the sampling and monitoring program used to describe
the plant (including the presence or absence of any
recycle at any point of sampling or measurement),
• the actual and desired duration of intermittent opera-
tion such as sludge wasting, processing, and
disposal,
• the performance and quality characteristics of the sol-
ids handling processes in terms of their feeds, product
solids, and, if possible, product liquids, and
• validation of any perceived bottlenecks and limitations
which may be described by operations and/or the ex-
perience of the field.
The above information should be used to describe
the performance of the plant under existing condi-
tions (and those in the future) and help identify com-
pliance strategies and likely attendant improvements
(including those which offer more beneficial or opti-
mal use of the existing works) to meet the processing
needs of the future while, if not overcoming the prob-
lems of today, at least not contributing to their unac-
ceptable exacerbation.
71
-------
2.9.3.4 Define Process Constants and Variables
The preceding analyses allow the designer to address
areas of uncertainty and adequacy. Decisions are needed
with regard to the following:
• the adequacy of the existing works and appurtenant
equipment under the present and future operating
strategies,
• the attractiveness of alternate operating strategies to
realize improved process performance, flexibility,
and/or additional processing capacity,
• whether or not alternate unit processes for the same
processing objective need be evaluated, and
• the important design constants and variables for each
process found in the possible integrated processing
trains.
The decisions that are reached at this point guide the
preparation of the mass balances described in the next
subsection. As can be inferred, the nitrogen control sys-
tem decision is only one of many, and not necessarily the
most important, to be reached by the design engineer.
One of the more critical variables needing definition is the
biological reactor solids production. This characterization
begins with an understanding of the carbonaceous re-
movals that can be anticipated under normal operating
conditions. Given freedom from significant dumps of
unique biodegradable or nonbiodegradable industrial
wastes, as associated with discontinuous industrial op-
erations and end-of-the-week cleanups, Figure 2-8 pro-
vides a reasonable characterization of the likely soluble
CBOD5 and soluble COD removals as a function of the
mean solids retention time, 90 (or mean cell residence
time, MCRT) in the biological reactor. Each plant is unique
as represented by the ranges shown in Figure 2-8. Under
non-oxygen limiting conditions, Tampa achieves over 95-
percent SBOD5 removal at 0C near one day. Jenkins and
Garrison (44) note that it is very difficult to detect variation
in soluble degradable effluent COD for plants operating
below substate removal rates of 3 kg COD removed/kg
VSS/d. The likely 9C for designs which provide for secon-
dary treatment equivalency (the maximum month SS and
CBOD5 objectives of 30 mg/L, without nitrification as a
design objective) and for nitrification are shown for refer-
ence purposes.
As shown in Figure 2-8, soluble COD removals parallel
but lie below the soluble CBOD5 removal percentages
since the COD test is a measure of the oxygen demand
associated with all the carbonaceous compounds found
in the wastewater and is indiscriminate as to their biode-
gradability. However, in the solids residence time operat-
ing range for equivalent secondary treatment, there is
little biodegradable material left, and, once 00's of from 3
100
CBODK
COD
Typical Range of Design and Operations for Secondary Treatment
Equivalency, Degree of Nitrification Can Vary
from Trace to Significant
Nitrifying Systems Are Typically Designed and Operated
at 60 > 7 d
J_
I
10 15 20 25
Mean Solids Residence Time (ec), d
30
Figure 2-8. Likely soluble CBODs and COD removals as a function of mean solids retention time in a biological reactor(s).
72
-------
to 7 days are encountered, little significant change in the
soluble residuals can be expected.
Equipped with the knowledge of the likely soluble sub-
strate (CBOD5 or COD) removals for any solids residence
time condition, the next task for the designer is to esti-
mate the solids production derived from the biological
reactor. This solids production estimate is partially deter-
mined from the inert solids that are applied to and, if
applicable, formed in the system (e.g., through biologi-
cally enhanced or chemically induced phosphorus remov-
als) and the coefficients that are used to approximate the
complex response of the biological reactor as it removes
the applied biodegradable and nonbiodegradable sub-
strate. Figure 2-3 and Section 2.6.1.3 fundamentally por-
tray and describe this response.
Table 2-13 presents the basic relationships and provides
the coefficients that have been used to describe the solids
production characteristics of the biological reactor. This
is shown both in terms of the simplistic characterizations
generally used in the past, and the more complex rela-
tionships that have recently emerged and gained some
acceptance in the field, principally because of the devel-
opment of readily available, conveniently packaged, non-
proprietary software and the emergence of powerful
personal computers (45,46). (Chapter 5 provides a more
in-depth review of the International Association on Water
Pollution Research and Control (IAWPRC) model for the
interested reader.) The following paragraphs provide a
broadly based discussion of these relationships for pur-
poses of the design examples, and allow the reader to
formulate some understanding of the significance of any
predicted result.
Table 2-13. Volatile Solids Production Considerations and Assumptions
Classical Approach
IAWPRC Approach
BOD5 Basis
COD Basis
COD Basis
Basic
Relationships
v —
I nAt —
Y(Asubstrate)
(1 + b90)
Parameters
Ynet
Y
A substrate
Net VSS production
Overall Yield Coefficient, VSS/substrate removed
Total in minus soluble out
Decay coefficient, VSS destroyed/d/reactor VSS
•
Solids residence time, days
Nonbiodegradable applied VSS, implicitly assumed
in definition of coefficients for Y and b
Biodegradable fraction of active biomass, implicitly
assumed in definition of coefficients for Y and b
Net VSS production
Overall True Yield Coefficient,
VSS/biodegradable substrate removed
Total in minus soluble out, biodegradable
substrate-only
Decay coefficient, VSS destroyed/d/reactor
VSS (biodegradable substrate, active biomass
basis)
Solids residence time, days
Nonbiodegradable applied VSS
Biodegradable fraction of active biomass
Typical Coefficients
("Constants") Used for
Municipal Wastewaters
Y
b
X0
fb
0.7
0.06
0.5
0.1
0.45
0.25
30% of raw VSS*
0.8
'Should consistently apply the same assumption to the paniculate phase COD, TKN, and P. X0 widely varies in reported literature, undoubtedly
because of the presence or absence of recycles and collection system influences on the native raw wastewater characteristics. Thirty percent
was elected for use in the design example as generally representative of raw municipal wastewaters. Remember that inert nonvolatile solids must
be added to compute total solids production.
73
-------
Biodegradable Substrate Sorbed and/or Converted to
Biodegradable Cell Mass
Biodegradable Substrate Converted to
/ Nonbiodegradable Cell Mass
Substrate
Removed by
Biological
Reactor
Nonbiodegradable Applied Substrate
Direct Substrate Oxidation (Cellular Energy)
Biomass (Indirect Substrate) Oxidation
Net Substrate in Biomass
(Solids Production)
Soluble
Effluent
Substrate
(Largely Nonbiodegradable)
Log Scale
ec,d
(and Contact Time with Reactor Biomass)
Waste VSS
Effluent VSS
COD (Substrate) Mass Balance:
Total COD
(Substrate)
Influent
Soluble
Effluent
COD
(Substrate)
COD of
Biomass
Solids
Production
Net COD of
Substrate
Lost by
Oxidation
Note : Theoretical COD of cells is classically taken at 1.42 VSS. Credit for nonbiodegradable
applied substrate will cause the COD of the biomass to increase (likely to 1.5 VSS).
Alternatively, a COD/VSS of 1.4 could be used to develop a safety factor for determination
of the oxygen supply (and realization that the COD test may not fully measure all of the
carbonaceous oxygen demand) or the safety factor could be applied to the end result.
Figure 2-9. Characterization of biological reactor substrate distribution.
Figure 2-9, a simpler version of Figure 2-3, characterizes
the distribution of the substrate (COD) upon application
to a biological reactor. As explained earlier with Figure
2-8, the soluble substrate is initially removed at a rapid
rate and then progressively attenuates to some stable
residual which is dependent on the relative biodegradabil-
ity of the individual organic compounds that compose trie
substrate and the biological reactor's solids inventory.
Older solids residence times allow more acclimation and
more soluble substrate removal. The removed particulate
substrate sorbs on, and is hydrolyzed prior to transport
within the biomass.
Biodegradable material is metabolized (oxidized while
providing the energy for cellular replication) and synthe-
sized (into new biomass). The reactions for the applied
biodegradable particulate matter are slower than for the
soluble substrate since these must be preceded by hy-
drolysis to reduce the particle size for ready cellular utili-
zation. Nonbiodegradable material, whether inert or
volatile, merely comes along for the ride (sometimes
masking potentially active sites of microbial stabilization).
The formed or stored cellular products consist of both
degradable and nonbiodegradable constituents. The de-
gradable products are oxidized and resynthesized in a
never-ending chain of events, leading to the progressive
accumulation of nonbiodegradable cellular mass residue.
The net solids production of the biological reactor is both
substrate and solids residence time dependent. Older
solids retention times yield the highest degree of stabili-
zation and the least amount of waste solids for sub-
sequent processing.
With the foregoing understandings, Figure 2-10 was pre-
pared to illustrate the use of the coefficients defined in
Table 2-13 with the design example average day raw and
settled wastewater characteristics described in Tables 2-
11 and 2-14, respectively. The settled wastewater char-
acteristics are based on an assumed 65-percent removal
of the raw wastewater SS in the primary clarifiers and
80-percent removal of its nonbiodegradable component.
These assumptions along with the assumed nonbiode-
gradable volatile particulate fraction of 30 percent in the
raw wastewater (i.e., nondegradable VSS of 36.9 mg/L)
yield a 17-percent nonbiodegradable volatile particulate
component (i.e., 7.4 mg/L) in the settled wastewater
which contains a total VSS of 43 mg/L; this is consistent
with the knowledge that poorer biodegradability tracks
larger sized particles. Soluble substrate removals follow
74
-------
0.7
0.6
0.5
T3
s
£
§1 °'4
Ss
II
it
0.3
o.
CO
0.2
0.1
Note: Normal Expectations Are
± 5-10 Percent of Estimate
Typical Range of Design and Operations for Secondary
Treatment Equivalency, Degree of Nitrification Can
Vary from Trace to Significant
Classical BOD5 Basis
Oxygen Demand (COD)
To Be Satisfied by
Oxygen Transfer
Classical BODs Basis
Per COD Removals
Classical COD Basis t
Nitrification Systems Are Typically Designed and
" Operated at 90 > 7d
1.10
0.90
0.75
0.60
O
.O
O
8
0.45 0.
0.30
0.15
10 15 20
Mean Solids Residence (6C), d
25
30
Figure 2-10. Estimates of volatile solids production (see Table 2-11 to Table 2-13 and Figure 2-14).
the estimates contained in Figure 2-8, assuming no ad-
ditional removal once 0C exceeds 10 days.
When using the coefficients defined in Table 2-13, it is
very important to understand that they are coupled pa-
rameters and that they were determined over a particular
range of observed operation with specific wastewaters to
yield some net prediction of response. Practically, the
prediction of the net response is more important than the
actual values used for the coefficients. When the coeffi-
cients are used outside the operating range from which
they were determined, there is an increased opportunity
for error. Additionally, coefficient values that have their
basis in the classical approach can not be interchanged
with coefficient values that have their basis in the
IAWPRC approach.
The coefficients shown in Table 2-13 are often reported
with some suggestion of precision and universal applica-
bility. This should not be assumed. In the case of the
classical BOD5 basis of determination, significantly vary-
ing and sometimes not necessarily sound parameters
have found their way into practice. With the IAWPRC
approach, it should be noted that predicted results are
strongly influenced by not only the coefficients but also
the assumptions used to define the nonbiodegradable
component of the particulates in applied wastewater. Al-
though this approach more correctly duplicates theory, its
75
-------
use still requires utilizing the appropriate model coeffi-
cients for the wastewater in question. Presently, the basis
for the IAWPRC approach's coefficients represents only
a relatively small sampling of wastewaters and treatment
plants in the United States, although the Europeans have
collected a great deal of data on model parameters. En-
thusiasm for any approach must be dampened by the
realization that actual solids production values may be
strongly influenced by plant specific recycles, and the
possible errors that may be imbedded into plant solids
production estimates by failure to account for recycle
impacts.
The foregoing discussion facilitates understanding of the
predicted solids production values found in Figure 2-10.
As shown, the coefficients identified for the classical
BODS and COD approaches are insensitive to whether
the applied wastewater is raw or settled. This, more than
anything else, is the great value of the IAWPRC ap-
proach, for it allows focus and ready understanding that
processing raw wastewaters yields a greater fraction of
Inert volatile biomass.
Interestingly, the plots in Figure 2-10 show similar pre-
dicted results (within 10 percent) through the entire mean
solids residence time range of normally encountered de-
signs and operations (60 of 3 to 30 days) for the classical
BODS and COD basis predictions. Both also fall within
the raw and settled sewage envelope defined by the
IAWPRC approach over a 90 of 3 to 15 days. Such would
be expected in the classical attempt to define one con-
stant coefficient for raw and settled wastewaters over the
range of conventional practice. Undoubtedly, the classical
approach parameters were developed and applied largely
for secondary treatment equivalency facilities. The stimu-
lant for the IAWPRC approach is found with greater use
of older residence time cultures designed to achieve a
variety of processing objectives. The IAWPRC approach
undoubtedly more fairly characterizes the solids produc-
tion derived with high 0C systems; however, informed ad-
justments of the endogenous decay term in the classical
approach would result in the same predictions.
The net conclusion associated with this discussion is that
traditional methods are not necessarily bad when properly
applied. What is important is not the methods but the
predicted result. The mass balances prepared for the
design examples use the predictions of the IAWPRC ap-
proach to illustrate the tradeoffs between raw and settled
wastewater processing.
The oxygen demand that must be satisfied with the bio-
logical reactor represents another significant variable that
must be estimated by the designer. In contrast to the
uncertain (and often arbitrary) application of factors as-
sociated with the BODS removal, the oxygen needs of the
system readily flow from the COD mass balance with
reasonable predictions of the biological reactor's volatile
solids production. Figure 2-9 conceptually, and Figure
2-10 precisely, illustrate the linkage and power of a COD
mass balance in predicting the carbonaceous oxygen de-
mand with the volatile solids production. Use of a COD
mass balance inherently achieves consistency and
checks on the predicted carbonaceous oxygen demand
and volatile solids production. (Where plants don't rou-
tinely perform COD analyses, the designer may be better
served by applying routinely anticipated BOD:COD ratios,
rather than relying totally on the somewhat arbitrary ratios
of oxygen demand to BOD removed or applied.)
When the COD mass balance is applied with a nitrogen
mass balance, the nitrogen available for oxidation is read-
ily identified. This technique is superior to merely assum-
ing that the ammonium found in the applied wastewater
is what must be oxidized (which carries the implicit as-
sumption and conceptual error that all of the organic ni-
trogen in the applied wastewater is removed with the
solids). The total oxygen demand (TOD) of the system is
estimated as the sum of the carbonaceous oxygen de-
mand from the COD mass balance and the ammonia
nitrogen available for oxidation (multiplied by 4.6). This
total oxygen demand approach is recommended for de-
sign, and is illustrated in the mass balances prepared for
the design examples.
Table 2-14 summarizes the remaining assumptions used
to prepare the mass balances. The performance of the
primary treatment system in the more complex Plant B
was developed to illustrate the concept that the perform-
ance of any given process is influenced by the nature and
character of the waste streams applied to it. The assump-
tions shown in Table 2-14 were used in preparing the
mass balances (the pollutant SS characterizations deter-
mine the unit process performance for the other pollutants).
Solids production was estimated by the IAWPRC ap-
proach described in Table 2-13 (45,46). These estimates
reflect raw wastewater solids nonbiodegradability de-
scribed in Table 2-11 in order to demonstrate concepts of
nonbiodegradable volatile matter and buildup (for addi-
tional understanding, see Chapter 5). Classical ap-
proaches can overestimate volatile solids destruction and
air requirements for COD mass balances for high solids
residence time systems and when processing raw waste-
waters. A solids residence time of 15 days was elected
for Plant A to achieve well-stabilized sludge; the same
was used for Plant B to allow direct comparison with the
simpler plant (without primary clarification). No further
decay of recycled VSS from filter backwash and the
sludge processing train was assumed because of the
advanced stabilization of solids generated from the treat-
ment system and for simplicity.
Anaerobic solids stabilization is included in the more com-
plex Plant B in order to illustrate the impact of digester
supernatant. The assumptions shown in Table 2-14 were
made to reflect the understanding that biodegradable
VSS must be solubilized before stabilization can be-
76
-------
Table 2-14. Remaining Assumptions for Example Mass Balances
PRIMARY TREATMENT
Process Stream Parameter
Raw Wastewater TSS
Raw Wastewater Nondegradable VSS
Supernatant, Filter, Backwash Thickener Overflow SS
SECONDARY TREATMENT
Solids Production by IAWPRC approach—See Table 2-13
Solids Residence Time, 60 = 15 d for both Plants A and B
ANAEROBIC DIGESTION
Waste Stream
Biodegradable Raw Wastewater VSS in Primary Sludge
Biodegradable Waste Secondary Solids and Backwash VSS
All Other Recycled VSS
SOLIDS THICKENING AND DEWATERING PROCESSES
Two-Stage Anaerobic Digester
95% Solids Capture' in Thickening and Dewatering
Separate Primary and Secondary Thickening
% Removal in Primaries
65
80
50
% VSS Destroyed
85
10
0
% Applied VSS Remaining as
Solubilized VDS
5
5
0
gin, and that only the biodegradable material can be sta-
bilized. The advanced stabilization of the secondary sol-
ids at the 15-day solids retention time and backwash
solids results in little additional stabilization of secondary
solids in the anaerobic digester.
Ninety-five percent SS capture was assumed for thick-
ening and dewatering processes. If a gravity thickener
was elected (often a good choice for small plants),
high-level SS captures are most assured by avoiding
excessive blankets in the thickener and using down-
stream storage. Wash water, as would be encountered
if a belt filter was used for dewatering, was ignored in
the mass balance.
The example elected not to thicken primary solids sepa-
rately and not to provide cothickening with secondary
solids. This results in direct application of primary sludge
for more conservative design. If a gravity thickener was
provided, consideration should be given to sizing it for
cothickening for additional operating flexibility; although,
in this application, odors may result from mixing the raw
and secondary solids. Feasible operation might be to
base-load the primary sludge into the system, and to have
fractional diversions when sludge processing peaks are
high. Preferential thickening should always be given to
the secondary solids.
A final effluent recycle to the thickener might provide
sufficient odor mitigation but would probably result in a
higher solids loss. A capability to chlorinate this recycle
would provide some measure of protection for sulfide
odors. In a nonrural environment, covering the thickener
with an odor control system should be considered if co-
thickening is a routinely planned mode of operation.
The second stage digester (of the more complex Plant
B) provides solids concentration (gravity thickening) and
storage services. Ninety-percent SS capture efficiency
was used to reflect the combined thickening-storage na-
ture of the process and likely difficulty in capturing an-
aerobically stabilized waste biological sludges.
Federal guidelines for the management of sewage sludge
for beneficial agricultural utilization call for a well-stabi-
lized material, free from pathogens, nuisances, and se-
lected hazardous materials. Sewage sludge is classified
into two categories, Class A and Class B, based upon the
degree of pathogen reduction. Restrictions placed on end
uses of sewage sludge are affected by its pathogen re-
duction classification. Bulk sewage sludge applied to
agricuitrual and nonagricultural land (e.g., forest, public
contact sites, and reclamation sites) must meet at least
Class B requirements. Bulk sewage sludge applied to
lawns and home gardens, and sewage sludge sold or
given away in bags or other containers must meet class
A criteria and.one of 10 vector attraction mandates. A
vector attraction reduction requirement also must be met
when sewage sludge is applied to land. Examples of
77
-------
Class A processes are pasteurization, well-run compost-
ing, and alkaline treatment. Class B processes include
anaerobic digestion, lime stabilization, and aerobic diges-
tion. Proper control of vectors requires steps such as the
following: well-run digestion systems, lime stabilization,
drying, and soil incorporation (5).
2.9.3.5 Prepare Mass Balance
The extent of, and need for, process design mass bal-
ances depends on the decisions to be made. The proc-
essing schematics described earlier in the discussions of
Figures 2-5 and 2-6 differ in their approach to raw settle-
able solids control and stabilization, while remaining in-
sensitive to the specifics of the nitrogen control
technology. Conceptually, such an approach allows for
the identification of the right nitrogen control technology
as a function of the integrated plant works and its proc-
essing objectives. For example, if anaerobic digestion
was provided without equalization and bleed back of its
return, a suspended growth reactor would be preferred
over an attached growth reactor because of its long re-
actor detention time, which would help mitigate the unox-
idfzed nitrogen spikes applied to the reactor.
After determining the conceptual processing sequence of
interest, the next step is to characterize the performance
of the facility under the conditions of interest to the design
issue. Practically, these conditions of interest are best
described in terms of some set condition such as the
average design day to assure a common reference point.
Tables 2-15 and 2-16 describe the resulting average day
mass balance characterization of the simple Plant A and
more complex Plant B treatment facilities, respectively.
The "mg/L equivalent" term is readily converted to pollut-
ant mass by correcting for the actual flow of the facility
and its processing sidestreams. The interrelationship be-
tween the "mg/L equivalent" and the actual pollutant con-
centration is made readily apparent by comparing the
equivalent and actual TSS concentrations in Tables 2-15
or 2-16. The use of "mg/L equivalent" as the first step of
any plant's characterization allows ready understanding
of the interrelationship of processes and pollutants in sim-
ple, easy to follow units.
Pollutant mass balances are prepared through the itera-
tion of repetitive calculations. Typically, three iterations
are performed through the integrated works to achieve
approximate equilibrium values.
Callout 4 in the mass balance for the more complex Plant
B represents the reactor effluent quality for a biological
reactor with a solids residence time of two days. It is
provided because often an existing plant must be up-
graded to achieve a new design objective. As shown, the
reactor effluent soluble BOD5 and COD are slightly infe-
rior to that expected from a system with a solids residence
time of 15 days. However, lower soluble phosphorus and
nitrogen residuals are achieved in the 2-day solids resi-
dence time system because of the higher anticipated de-
gree of sludge synthesis.
Separation of the carbonaceous and nitrogenous oxygen
demands naturally occurs in attached growth systems,
and is the basis of the two-stage activated sludge system
which was applied to several large wastewater treatment
plants in the early seventies. Attached growth systems
are readily amenable to such concepts since they natu-
rally achieve a staged operation. If the phases are un-
coupled through the use of a clarifier, use of a low solids
residence time in the first stage, as suggested in Callout
4, offers the opportunity for reactor savings. This concept
is addressed in more detail in the suspended and at-
tached growth system design chapters of the manual.
As shown in Figures 2-5 and 2-6, the two plants have a
soluble phosphorus residual of about 5 mg/L. Removal
of this phosphorus down to a soluble residual of 0.5 mg/L
is likely to result in some 15 to 30 mg/L of net inert SS
depending on the phosphorus removal strategy (with the
range defined by normal expectations for enhanced bio-
logical phosphorus removal and immobilization by iron
additions). The additional solids mass due to phosphorus
removal can be easily quantified at each processing point
by adjustment of the inert solids (and immobilized phos-
phorus) and by the capture efficiency of each unit proc-
ess. At equilibrium with the capture efficiencies of the
processes, the inert SS with phosphorus removal would
rise by around 20 percent to yield a total incremental
waste solids gain from the liquid processing train of about
18 to 36 mg/L equivalent.
If the phosphorus removal strategy was incorporated into
the biological treatment system, the waste secondary sol-
ids from Plants A and B would increase by 15 to 30
percent and 35 to 70 percent, respectively. This increase
may have significant bearing on the design of suspended
growth systems since it would require an equivalent in-
crease in the reactor tankage to maintain the same MLSS
concentration as encountered without enhanced phos-
phorus removal.
When phosphorus removal is anticipated by metal salts,
the alkalinity demand must be considered to assure that
sufficient alkalinity remains, or is provided, to satisfy the
needs of the nitrification reaction. As discussed earlier in
Section 2.7.2.1, addition of the metal salts to the reactor
effluent, prior to solids separation, is the generally pre-
ferred point of application since it allows the often accom-
panying favorable pH depression to occur without great
concern over protection of the nitrification reaction. Here
the lowest likely target residual alkalinity is approximately
20 to 30 mg/L as CaCO3 in order to keep the product
liquor pH above 6 to 6.5.
2.9.3.6 Use the Mass Balance
The bottoms of Tables 2-15 and 2-16 illustrate the poten-
tial use of the information that can be derived from the
78
-------
Table 2-15. Mass Balance for Plant A (see also Figure 2-5)
mg/L Equivalents3
Normalized Flow
Processing Point
1.
13.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Raw Influent
Recycles0
Total Influent
Reactor Eff.d
Final Effluent
Waste Solids
Solids
Underflow
Filtrate
Cake
Limed Cake9
ISS
52
6
58
58
6
52
3
49
3
46
74
ss
vss
123
6
129
80
9
71
3
68
3
65
65
CBODs
TSS Sol.
175 50
12 —
187 50
138 2
15 2
123 —
6 —
117 —
6 —
111 —
139
Part.
100
2
102
12
1
11
1
10
1
9
Tot.
150
2
152
14
3
11
1
10
1
9
COD
Sol. Part.
100 180
— 10
100 190
20 117
20 13
— 104
— 5
— 99
— 5
— 94
Phosphorus as P
Tot. Sol.
280 5.0
10 —
290 5.0
137 4.9
33 4.9
104 —
5 —
99 —
5 —
94 —
Part.
1.0
0.1
1.1
1.2
0.1
1,1
0.1
1.0
— -
1.0
Tot.
6.0
0.1
6.1
6.1
5.0
1.1
0.1
1.0
—
1.0
Nitrogen
Sol. Part.
24.0 6.0
-r 0.5
24.0 6.5
24.5 6.0
24.5 0.7
— 5.3
— 0.3
— 5.0
— 0.2
— 4.8
asN
Tot.
30.0
0.5
30.5
30.5
25.2
5.3
0.3
5.0
0.2
4.8
Actual
mg/LSS
175
500
183
135
15
5,000
300
25,000
1500
200,000
Flow
Fraction
ofQ
1 .00000
0.02404
1.02404
1 .02404
0.99944
0.02460
0.01992
0.00468
0.00412
0.00056
11. Runoff
Not considered in this example
12. Cake to Land Same as Processing Point 10 in this example
a To find the actual kg mass/d, multiply the mg/L equivalent and the plant flow ML/d; Ib/d mass is calculated by multiplying mg/L by 8.34 and the
plant flow (mgd).
b To find the actual flow rates, multiply the sludge processing point mg/L equivalents by plant flow and divide by expected actual sludge concentration.
Liquid stream recycle flows are determined by difference, with their actual pollutant concentrations determined by dividing the mg/L equivalents
by the flow fraction.
0 Recycles is the sum of processing streams 6, 8, and 11.
d • Nonbiodegradable VSS were assumed to maintain the characteristics found in the influent wastewater; newly synthesized VSS were assumed
to have the following characteristics:
COD/VSS = 1.4; N/VSS = 0.1; P/VSS = 0.02; the BOD:COD ratio was assumed to be 1:10 in the plant effluent and product solids.
• Average day oxygen demand, assuming no natural denitrification, can be calculated from the mass balance as follows:
Carbonaceous Oxygen Demand = Processing Point 2 minus 3 = 290 - 137 = 153 mg/L
Nitrogenous Oxygen Demand = 4.6 mg O2/mg N oxidized (Oxidized N is the soluble nitrogen in Processing Point 3 minus soluble
nonbiodegradable nitrogen and effluent ammonium); assume nonbiodegradable nitrogen is 1.0 mg/L
and effluent ammonium is 0.5 mg/L on the average day
= 4.6 [24.5 - (1.0+ 0.5)] = 4.6(23.0)= 106 mg/L
Total Oxygen Demand, ignoring effluent DO as too small to be significant = 259 mg/L
If this was an activated sludge system, for each million liters/d flow:
- the average day air supply (ignoring mixing requirements), assuming 10% oxygen transfer efficiency and 0.28 kg oxygen/m3 air, is
= [(259 mg/L)(1.0 ML/d)] H- [(0.10 efficiency)(0.28)(1,440 min/d)(60 sec/m)] = 0.107 m3/s (or 226.7 cfm)
- and the MLSS in a 18-hour detention time aerator, for a solids residence time of 15 days is
= (15 days)(138 mg/L/d)(24 hr/18 hr) = 2,800 mg/L
• The alkalinity demand for nitrification is:
(7.1 mg CaCOa/N oxidized) (23.5 mg N oxidized/L) = 167 mg/L, from which, if it is desired to have a residual alkalinity of 50 mg/L as
CaCO3, the facility will need a supplemental alkalinity of (167 + 50) - 120 = 97 mg/L as CaCO3, which corresponds to a lime (CaO)
dose of 97 + 1.8 = 54 mg/L as CaO
8 CaO added at 25% of dry weight mass, which, for this example, was assumed satisfactory for PFRP (Processes for Further Reduction of Pathogens)
requirements.
79
-------
Table 2-16. Mass Balance for Plant B (see also Figure 2-6)
mg/L Equivalents
Normalized Flow3
Processing Point
1. Raw Influent
23. Recycles5
2, Total Influent
3. Prim. Effluent
4. Rx Effluent0
5. Rx Effluent"
6. Sac. Effluent
7. FB. Rx Effl"
8. Final Effluent
9. Backwash
10. Prim. Sludge
SS
ISS VSS TSS Sol.
52 123 175 50
15 22(26) 37(41) 3
67 145(159) 212(216) 53
25 55(59) 80(84) 53
25 58 83 8
25 42 67 2
6 9 15 2
6 14 20 2
14 52
5 10 15 —
42 90 132 —
11,Was.Sec.Sol!ds19 33 52 —
1 2. Thk Overflow
13.ThkSec.SI.
14. SI. to Dig
15. SI. aft. Dig
16. Supernatant
17. SUoDewater
18. Filtrate
19. Cake
20. Limed Cake
21. Runoff
22. Cako to Land
12 3 —
18 31 49 —
60 121 181 —
60 72(77) 132(137) 3
6 7(9) 13(15) 1
54 65(68) 119(122) 2
3 3(5) 6(8) 2
51 62(63) 113(114) —
79 62(63) 141(142)
Not considered in this example
CBODs
Part.
100
3
103
44
20
6
1
2
—
2
59
5
—
5
64
9
1
8
—
• 8
Tot.
150
6
156
97
28
8
3
, 4
2
2
59
5
—
5
64
12
2
10
2
8
COD
Sol. Part.
100 180
6 32
106 212
106 81
30 81
20 59
20 13
20 20
20 5
— 15
— 131
— 46
— 2
— 44
— 175
7 105
3 10
4 95
3 5
1 90
Phosphorus as P Nitrogen
Tot.
280
38
318
187
111
79
33
40
25
15
131
46
2
44
175
112
13
99
8
91
Sol.
5.0
0.4
5.4
5.4
4.9
5.2
5.2
5.1
5.1
—
—
—
—
—
—
0.4
0.2
0.2
0.2
—
Part.
1.0
0.3
1.3
0.6
1.1
0.8
0.2
0.3
0.1
0.2
0.7
0.6
—
0.6
1.3
0.9
0.1
0.8
—
0.8
Tot.
6.0
6.7
6.7
6.0
6.0
6.0
5.4
5.4
5.2
0.2
0.7
0.6
—
0.6
1.3
1.3
0.3
1.0
0.2
0.8
Sol. Part.
24.0 6.0
2.4 1.9
26.4 7.9
26.4 3.1
24.1 5.4
25.7 3.8
25.7 0.8
20.2 1.3
20.2 0.3
— 1.0
— 4.8
— 3.0
— 0.2
— 2.8
— 7.6
2.6 5.0
1.1 0.5
1.5 4.5
1.3 0.2
0.2 4.3
asN
Tot.
30.0
4.3
34.3
29.5
29.5
29.5
26.5
21.5
20.5
1.0
4.8
3.0
0.2
2.8
7.6
7.6
1.6
6.0
1.5
4.5
• Actual
mg/L SS
175
330
190
72
75
60
14
18
5
150
50,000
5,000
380
20,000
35,600
25,900
6,200
40,000
2,400
250,000
Flow
Fraction
of Q
1.00000
.0.11259
1.11259
1.10995
1.10995
1.10995
1.09955
1.09955
0.99955
0.10000
0.00264
Q.01040
0.00795
0.00245
0.00509
0.0050'9
0.00211
0.00295
0.00253
0.00045
Same as Processing Point 16 in this example
* To find the actual flow rates, multiply the sludge processing point mg/L equivalents by plant flow and divide by expected actual sludge concentration.
UquW stream recycle flows are determined by difference, with their actual pollutant concentrations determined by dividing the mg/L equivalents
by the flow fraction.
11 Recycles Is the sum of processing streams 9, 12, 16, and 18; values in parentheses reflect dissolved solids as a result of anaerobic digestion
plus the SS.
" Intermediate point of reactor effluent with solids residence time of 2 days, see text for additional detail.
a Average day oxygen demand, assuming no natural denitrification, can be calculated from the mass balance as follows:
Carbonaceous Oxygen Demand » Processing Point 3 minus 5= 187-79= 108 mg/L
Nitrogenous Oxygen Demand » 4.6 mg Og/mg N oxidized = 4.6 [25.7 - (1.0 + 0.5)] = 4.6(24.2) = 111 mg/L
Total Oxygen Demand, ignoring effluent DO as too small to be significant = 219 mg/L
If this was an activated sludge system, for each million L/d flow:
- the average day air supply (ignoring mixing requirements) and assuming 10% oxygen transfer efficiency and 0.28 kg oxygen/m3
air, Is
» [(219 mg/L)(1.0 ML/d)] -t- [(0.10 efficiency)(0.28)(1,440 min/d)(60 sec/m)] =0.09 nf/s (or 191 cfm)
- and the MLSS In an 18-hr detention time aerator, for a solids residence time of 15 days is
= (15 d)(67 mg/L/d)(24 hr/18 hr) = 1,340 mg/L
• Filter Reactor Effluent reflects net result of methanol addition to achieve 5 mg/L NO3-N removal, to a process stream containing 3 mg/L DO,
where: •
- Methanol COD = 1.5 [DO + 2.9(NO3-N)] = 1.5 [3 + 2.9(5.0)] = 1.5(17.5) = 26 mg/L
- Synthesized Solids ~ 0.25 (COD added) = 0.25 (26) = 5 mg/L VSS \
- Danitrilication In this example will yield about (3.6 mg CaCO^/mg NO3-N)(5 mg/L NO3-N) = 18 mg/L CaCO3 alkalinity
80
-------
mass balances. Inspection of the mass balance results
illustrates several considerations worthy of note, including
the following:
• < Design of the nitrogen control system must address
the total biodegradable nitrogen delivered to the reac-
tor, not just the ammonium nitrogen.
• The net waste solids found in the dewatered sludge
cake and the nitrogen available for oxidation are not
substantially different in the two alternatives (because
of the commonality of the design objective for solids
stabilization).
• The more complex Plant B's use of anaerobic digestion
represents a simple processing tradeoff of generating a
possible useful end product (methane) against the sim-
pler Plant A's use of oxygenation (and mixing) energy,
possibly yielding more optimized reactor volumes and
operating costs against more troublesome recycles.
The principal hydraulic recycle of concern at most waste-
water treatment plants is that associated with intermittent
backwashing of any effluent filtration system (including a
denitrification filter). The solids handling operations rep-
resent the principal source of pollutant mass-related prob-
lems. These problems can be associated with the failure
of the solids processing facility to achieve the desired
solids capture, and/or the discontinuous nature of sludge
wastage and processing practices at any given wastewa-
ter treatment plant. In terms of the latter, it is axiomatic
in wastewater treatment that as plants decline in size,
wastage and sludge processing will become less continu-
ous (as does their monitoring program).
The average day mass balance information defined in
Tables 2-15 and 2-16 can not be used directly for design
beyond serving as a convenient reference point. It must
be adjusted for the controlling peaking factors associated
with the minima and maxima elected for design, the an-
ticipated sludge wastage and dewatering operation, and
the buffering found in the process train itself. These peak-
ing factors should not be confused with a safety factor,
which is applied to reflect uncertainty with the performance
of the chosen technology. The following paragraphs de-
scribe the thought processes used to develop the control-
ling design conditions from the average day mass balance.
• Controlling Design Maxima
The controlling design maxima vary as a function of the
item under consideration. The controlling condition for the
design of the biological reactor can be determined by
comparing the ratios of the maximum week to maximum
month for the design effluent objectives (1.5 from Table
2-10) to the controlling influent pollutant mass peaking
factors used to characterize the influent wastewater (1.3
for the SS and organics from Table 2-12). As is normally
encountered, the maximum month is the controlling de-
sign condition since the controlling pollutant mass peak-
ing factor is less than the allowable deviation of the ef-
fluent standard. Thus, the biological reactor should be
designed for successful operation during the maximum
month. Since the maximum monthly peaks for the nitro-
gen mass are less than for the SS and organics, the
. predicted result of the average day mass balance could
be protectively and simply adjusted upward by the maxi-
mum month peaking factor to adequately characterize
system needs. (Similar thought processes are utilized for
effluent standards written in terms of maximum day limi-
tations. Permit writers should take care in establishing
maximum day limitations and assure that they are needed
since they correspond to an implicit 99.7-percent design
reliability objective. Such objectives may. result in a sub-
stantial increase in the capital cost of any wastewater
treatment facility.)
The controlling design maxima for other processing con-
siderations are not the same. The maximum week flow
might be used with the reactor's operating level of solids
for the maximum month to size the final sedimentation
system. Alternatively, it may be necessary to base the
design on peak wet-weather flow coupled with an as-
sumption about reactor MLSS. Consider also that reactor
MLSS settleability characteristics are variable. The maxi-
mum day's flow might be used to size the return sludge
system if activated sludge was the preferred technology.
The maximum day's peaking factor for the organics and
TKN might be used to size an activated sludge system's
oxygen supply and dissolution system.
• Controlling Design Minima
The controlling design minima can generally be deter-
mined by using the predicted average day mass balance
results, adjusted by the minimum month's peaking factors
for the lowest-flow year in the facility's design life. This
condition is used to develop the minimum needs of the
treatment facility. An example would be the minimum air
supply that may have to be transferred by the aeration
system for an activated sludge plant with a check of
mixing under these conditions.
• Anticipated Sludge Wasting .and Dewatering Operations
The importance of discontinuous recycles is illustrated in
Table 2-17 with the high SS anaerobic digester super-
natant and low SS dewatering filtrate described in the
mass balance prepared for the more complex Plant B
(Table 2-16). As shown for the example, the most stressful
conditions at the plant are encountered on the weekends
and, at least for the anaerobic digestion characterization
of the example, result in a disproportionate recycle of
soluble nitrogen.
It is this disproportionate character of the recycles that
imposes the greatest stress on the nitrogen control tech-
nology—from the standpoint not only of simple reaction
kinetics but also of maintenance of sufficient alkalinity to
avoid transient pH suppressions and ammonium nitrogen
81
-------
losses from the treatment plant. Clearly, as operations
become more discontinuous, any attempt to optimize the
liquid processing train incorporates greater risk because
of the loads associated with the recycles. These loads
may, and often do, have peaks which are far more fre-
quent and severe than anything encountered in the raw
wastewater.
Mitigation measures for these transients include equali-
zation and bleedback, and avoidance of plug flow, short
contact time reactors. With attached growth systems, if
anaerobic digestion is elected for solids stabilization, the
designer would be well advised to use equalization and
bleedback of supernatant and filtrate to avoid sudden
soluble nitrogen loads from the solids processing train to
the short hydraulic contact time reactors.
• Layered Diurnal Peaks
Wastewater flows and loads are not constant during the
operating day. One-half or more of the plant's daily load
may well arrive over an eight-hour period. This corre-
sponds to an eight-hour diurnal peaking factor of 1.5
times the average daily value. Maximum hourly loads
often are more than two times the average daily value.
In general, the smaller the plant (with its smaller collection
system), the more severe the diurnal peaking factor.
There are no hard and fast rules for election of the proper
peaking factor for the raw wastewater loads; they are
largely dependent on the elected process train. The liquid
volume found in the plant, along with the plant's liquid
stream recycles and returns, acts to mitigate the peaks
encountered in the raw wastewater and the recycle
streams. If adequate aeration capacity is available, the
same size tank will produce lower effluent ammonia levels
if it is plug flow rather than complete mix in a dynamic
load situation. Mitigation of processing peaks is also
achieved through the use of complete mix reactors. Fur-
ther, it should be remembered that some excursions over
the effluent standard are allowable.
The elected process peak for a short detention time at-
tached growth reactor should be higher than that selected
for a longer detention time suspended growth system.
Given that this manual is intended to serve the needs of
the designer of the smaller plant and to characterize the
different needs of the suspended growth and attached
growth technologies, a diurnal wastewater processing
peak of 1.3 and 1.6 was elected for the respective tech-
nologies. The corresponding diurnal minima for both proc-
esses, which should be applied to the minimum month
when natural denitrification is anticipated, are about 0.7
and 0.6, respectively. These peaking factors should be
applied as appropriate in the design, but, at minimum,
they must be applied to the nitrogenous component avail-
able for oxidation and denitrification.
The remaining processing peak that should be identified
and checked is consideration of the recycles from the
solids processing train. The analyses in Table 2-17 sug-
gest that the worst-case condition results in a periodic
60-percent transient two-hour increase in the plant's av-
erage day soluble nitrogen load, twice a day. This is the
same peaking factor as elected for the attached growth
technologies and some 20 percent higher than the peak-
ing factor elected for the suspended growth technologies.
Accordingly, no adjustment need be made to the available
nitrogenous component for the attached growth technol-
ogy application. (If the transient solids processing recycle
peak was less than the diurnal wastewater peak, the
correct methodology would end up with a lower overall
combined peak for the mainstream process design.) For
convenience with the design examples, and given that
the suspended growth system inherently offers greater
buffer of applied peaks through its greater liquid volume,
the decision was made to apply the elected process di-
urnal peak also to the total available nitrogen.
• A Word About Dilution, and Lack Thereof
The designer should remember that regulatory effluent
standards apply to the lesser of the permitted concentra-
tion or mass at the rated average annual daily flow. Under
dry weather or draught conditions (or an extremely tight
collection system), the favorable soluble pollutant dilution
in complying with concentration dependent effluent stand-
ards will be lost and noncompliance may be encountered
when compliance would be predicted under the elevated
flow regimes of the average day or maximum month.
Process selections that just barely comply at average day
flows may well fail under seasonal minima. A check
should be made under low flow conditions to ensure com-
pliance. Fortunately, low flows are encountered with
warmer temperatures which, for the suspended growth
technologies, gives some reasonable assurance of com-
pensating increases in reaction rates.
A low flow condition check is particularly relevant with
attached growth technologies as their performance is
largely concentration dependent under oxygen transfer
limiting conditions, as explained in Section 2.6.1. Warmer
temperatures and lower flows yield lower oxygen transfer
to a more concentrated waste stream with a longer resi-
dence time. This may yield conditions where poorer per-
formance on both a mass and concentration basis may
be encountered in the summer, as opposed to the clas-
sical expectation of poorer performance in the winter. This
issue is discussed again in Section 2.7 and more fully in
the later sections of this manual dealing with attached
growth technologies.
Table 2-18, based on the previous paragraphs, summa-
rizes the controlling design conditions for the design
examples.
82
-------
Table 2-17. More Complex Plant B Solids Processing Recycle Impacts on Main Processing Stream (see also Table 2-16
for basis of average day condition) '
PROCESSING STREAM CHARACTERIZATION IN
Processing Point
Raw Influent, total
soluble
LIQUID
SS
175
Case 1 : Continuous Sludge Processing (or Average Day
Supernatant, Factor = 1 .0
Filtrate, Factor = 1.0
Total
% Increase over Raw Influent
Case 2: Sludge Wasting for Two Hours per Day,
A. Weekdays, Worst Condition: Sludge Wastage to
Supernatant, Factor = 24/4 = 6
Filtrate, Factor = '(24/6) (7/5) = 5.6
Total
% Increase over Raw Influent for the Two,
2-Hour Periods
B. Weekends, Worst Condition: Sludge Wastage to
Supernatant, Factor = (24/4)(0.00509/0.00211) =
Total
% Increase over Raw Influent for the Two,
2-Hour Periods
Total Soluble
% Increase over Raw Soluble Influent
for the Two, 2-Hour periods
13
6
19
11
Twice a
Digester
78
34
112
64
Digester
14.5
189
108
0
PROCESSING TRAIN,
CBOD5
150
50
Condition)
2
2
4
3
mg/L unless
COD
280
100
13
8
21
8
Day, and Dewatering for Six Hours
noted otherwise
P
6.0
5.0
0.3
0.2
0.5
8.0
per Day, Five Days
TKN
30.0
24.0
1.6
1.5
3.1
10.0
per Week
for Two Hours and Concurrent Sludge Dewatering
12
11
23
15
for Two Hours per Day,
29
19
14
28
78
45
123
44
Twice a Day,
189
67
44
44
1.8
1.1
2.9
48
with Full Digester
4.4
73
2.9
58
9.6
8.4
18.0
60
23.2
77
16.0
67
83
-------
Table 2-18. Summary of Controlling Design Conditions for Biological Reactor with Design Examples
SUMMARY
• Size Biological Reactor Based on Maximum Month Loads
—Peaking Factor of 1.3 times average day mass balance for SS, BOD6, and COD (per Table 2-12)
—Peaking Factor of 1.2 times average day mass balance for Total Phosphorus and Nitrogen (per Table 2-12)
- Further adjust available Nitrogen by additional Diurnal Peaking Factor in plant as follows:
Suspended Growth
Attached Growth
Diurnal Process Peaking Factor 1.30 1.60
Total Processing Peaking Factor (1.2)(1.3) = 1.56 (1.2)(1.6) = 1.92
Note: With attached growth technologies, it may be appropriate to apply the SS and Organic Peaking Factor because of
oxygen-demanding conditions.
The Total Processing Peak for these pollutants is (1.3)(1.6) = 2.08 ;
—At a minimum, check for concentration compliance at minimum month flow conditions (Peaking Factor = 0.7) with maximum month load.
- Attached growth technologies also should check for oxygen transfer under these conditions.
• Size Oxygen Supply for Suspended Growth Technologies Based on Maximum Day Peaking Factors of 2.1 for Organics and 1.7 for
Available Nitrogen (Table 2-12).
• Size Clarifier for ^Maximum Week Flows (Peaking Factor = 1.9) with Maximum Month's Operating Solids.
PLANT CHARACTERIZATION AND WASTEWATER CHARACTERISTICS, mg/L Equivalent3 and Actual"
SIMPLER PLANT A
(per Figure 2-5 and Table 2-15)
MORE COMPLEX PLANT B
(per Figure 2-6 and Table 2-16)
Maximum Month and Noted
Q
Process Infl., mg/L
SS
BODs
SBODs
COO
SCOD
Reactor Effluent0
SS
COO
Available tf
Carbon Oa
Demand*
Nitrogen Oz
Demand11'8
Total Oa
Demand*
Wastowater Alkalinity
Avg. Day8
1.0
187
152
50
290
100
138
137
24
153
113
266
120
Process Diurnal
Peaka Peak3
Sus.
Growth
1.0 1.0
243
198
65
377
130
179
178
29 38
321
192
513
132 132+
Low Q
Check"
0.7
347
283
93
538
186
256
254
41
188
Maximum Month and Noted
Avg. Day8
1.0
80
97
53
187
106
67
79
26
108
120
228
120
Process
Peak3 Diurnal Peak3
Sus. Att.
Growth Growth
1.0 1.0
104
126
69 110
243
138 221
87
103
31 41 50
227
204
431
132 132+ 132+
Low Q
Check"
0.7
148
180
99
347
197
124
147
44
188
•Average daily flow (ADF) basis, actual Ib/day mass is determined by multiplying the designated mg/L Equivalent by the ADF (mgd) and 8.34.
'Tha mg/L value is value expected if the maximum month pollutant load (mass) was experienced during the minimum month flow.
'For the special Intermediate condition (Callout 4) defined in the average day mass balance for Plant B, the Process Peak values for the Reactor
Effluont are as follows: SS » 108, COD = 144, Available N = 30 mg/L Equivalents.
"Rounded Available N to nearest whole number and determined O2 demand by directly multiplying by 4.6, ignoring soluble effluent refractory organic
N and NH,-N.
•Process Peak demands reflect maximum day, not maximum month.
84
-------
2.10 References
When an NTIS number is cited in a reference, that docu-
ment is available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
703-487-4650
1. Water Environment Federation. 1992. Wastewater
treatment plant design. Manual of Practice No. 8.
Alexandria, VA.
2. U.S. EPA. 1989. Analysis of performance limiting fac-
tors (PLFs) at small wastewater treatment plants.
WH-546/OMPC 10-89. Washington, DC.
3. Water Pollution Control Federation. 1990. Operation
of wastewater treatment plants. Manual of Practice
No. 11. Alexandria, VA.
4. Standards for the Use or Disposal of Sewage Sludge;
Final Rules 40 CFR Part 257 et al. 1993. Federal
Register 58(32) February 19.
5. U.S. EPA. 1992. Environmental regulations and tech-
nology: control of pathogens and vectors in sewage
sludge. EPA625/R-92/013. Cincinnati, OH.
6. Reed, S.C., E.J. Middlebrooks, and R.W. Crites.
1988. Natural systems for waste management and
treatment. New York: McGraw-Hill.
7. Water Pollution Control Federation. 1990. Natural
systems for wastewater treatment. Manual of Prac-
tice No. FD-16. Alexandria, VA.
8. U.S. EPA. 1981. Process design manual: land treat-
ment of municipal wastewater. EPA/625/1-81-013
(NTIS PB88-189436). Cincinnati, OH.
9. U.S. EPA. 1988. Constructed wetlands and aquatic
plant systems for municipal wastewater treatment.
EPA/625/1-88/022. Cincinnati, OH.
10. U.S. EPA. 1980. Design manual: onsite wastewater
treatment and disposal systems. EPA/625/1-80/012
(NTIS PB83-219907). Cincinnati, OH.
11. U.S. EPA. 1983. Design manual on municipal waste-
water stabilization ponds. EPA/625/1-83/015 (NTIS
PB88-184023). Cincinnati, OH.
12. U.S. EPA. 1984. Process design manual for land
treatment of municipal wastewater: supplement on
rapid infiltration and overland flow. EPA/625/
1-81/013a (NTIS PB89-189444). Cincinnati, OH.
13. U.S. EPA. 1992. Wastewater treatment/disposal for
small communities. EPA/625/R-92/005. Cincinnati,
OH.
14. Stone, R.W., D.S. Parker, and D.A. Cotteral. 1975.
Upgrading lagoon effluent for best practicable treat-
ment. JWPCF 7:2019.
15. Water Pollution Control Federation. 1967. Sewage
treatment plant design. Manual of Practice No. 8.
Washington, DC (1967 printing of 1959 edition).
16. Pano, A., and E.J. Middlebrooks. 1982. Ammonia
nitrogen removal in facultative stabilization ponds.
JWPCF 54:344.
17. Reed, S.C. 1985. Nitrogen removal in wastewater
stabilization ponds. JWPCF 57:39.
18. Oswald, W.J., H.B. Gotaas, C.G. Golueke, and W.R.
Kellen. 1957. Algae in wastewater treatment. Sewage
and Industrial Wastes 29(4):437.
19. Golueke, C.G., WJ. Oswald, and H.K. Gee. 1962.
Increasing high-rate pond loading by phase isolation.
Sanitary Engineering Research Laboratory, Univer-
sity of California, Berkeley (April).
20. Tchobanoglous, G., F. Maitiski, K. Thomas, and T.H.
Chadwick. 1989. Evolution and performance of city
of San Diego pilot scale aquatic wastewater treat-
ment system using hyacinths. JWPCF 61:1625.
21. Water Pollution Control Federation. 1977. Wastewa-
ter treatment plant design. Manual of Practice No. 8.
Washington, DC.
22. U.S. EPA. 1975. Process design manual for nitrogen
control. EPA/625/1-77/007 (NTIS PB-259149). Wash-
ington, DC.
23. Sawyer, C.N., and P.L. McCarty. 1978. Chemistry for
environmental engineering. Third edition. New York:
McGraw-Hill.
24. Sanks, R.L. 1980. Water treatment plant design for
the practicing engineer. Ann Arbor, Ml: Ann Arbor
Science.
25. U.S. EPA. 1987. Design manual: phosphorus re-
moval. EPA/625/1-87/001. Cincinnati, OH.
26. U.S. EPA. 1971. Advanced waste treatment as prac-
ticed at Lake Tahoe. Water Quality Office,
17010ELQ08/71. Washington, DC.
27. Mulbarger, M.C. 1982. Technology update trip re-
ports. Prepared for the Washington Suburban Sani-
tary Commission, 201 Western Branch Facilities
Plan.
28. The Soap and Detergent Association. 1989. Princi-
ples and practices of phosphorus and nitrogen re-
moval from municipal wastewater. New York, NY
(September).
85
-------
29. Cohen, J.M. 1972. Nutrient removal from wastewater
by physical-chemical processes. (NTIS PB-213783).
Cincinnati, OH.
30. White, G.C. 1978. Disinfection of wastewater and
water for reuse. New York: Van Nostrand Reinhold
Company.
31. Argo, D.G. 1980. Evaluation of membrane processes
and their role in wastewater reclamation. Final report,
Volume II: Contract 14-34-0001-8520. Prepared by
the Orange County Water District for the Office of
Research and Technology, U.S. Department of the
Interior, Washington, DC (November).
32. Lauer, W.C., et. al. 1991. Process selection for pota-
ble water reuse health effects studies. JAWWA
83{11):52 (November).
33. Parker, D.S., and T. Richards. 1986. Nitrification in
trickling filters. JWPCF 58:896.
34. Parker, D.S., etal. 1989. Enhancing reaction rates in
nitrifying trickling filters. JWPCF 61:632.
35. Lutz, M.P., et al. 1990. Full-scale performance of ni-
trifying trickling filters. Presented at the 63rd Annual
WPCF Conf., Washington, DC (October).
36. Okey, R.W., and O.A. Albertson. 1989. Diffusion's
roles in regulating and masking temperature effects
in fixed film nitrification. JWPCF 54:500.
37. Okey, R.W., and O.A. Albertson. 1989. Evidence for
oxygen limiting conditions in fixed-film nitrification.
JWPCF 54:510.
38. Parker, D.S. 1992. Upgrading trickling filter effluents
with the TF/SC process. Presented at the IWEM 92
Conference (April).
39. Water Pollution Control Federation. 1983. Nutrient
control. Manual of Practice FD-7. Alexandria, VA.
40. Sykes, R.M. 1984. Indeterminancy in mechanistic
biological models. JWPCF 56:209.
41. U.S. EPA. 1989. Design manual: fine pore aeration
systems. EPA/625/1-89/023 (NTIS PB90-125204).
Cincinnati, OH.
42. Schwinn, D.E., and B.H. Dickson, Jr. 1972. Nitrogen
and phosphorus variations in domestic wastewater.
JWPCF 44(11):2059.
43. U.S. EPA. 1987. 1986 needs survey report to Con-
gress. EPA/430/9-87/001. Washington, DC.
44. Jenkins, D., and W.E. Garrison. 1968. Control of ac-
tivated sludge by mean cell residence time. JWPCF
40:1905.
45. Bidstrup, S.M., and C.P.L. Grady, Jr. 1987. A user's
manual for SSSP. Environmental Systems Engineer-
ing. Clemson University, Clemson, SC.
46. Bidstrup, S.M., and C.P.L. Grady, Jr. 1988. SSSC-
simuiation of the single sludge process. JWPCF
60:351.
86
-------
Chapters
Process Chemistry and Kinetics of Biological Nitrification
3.1 Introduction
This chapter presents a review of the process chemistry
and kinetics of biological nitrification for wastewater treat-
ment. An understanding of the process fundamentals is
important for appreciating the factors affecting the design,
operation, and performance of nitrification process sys-
tems. In discussing kinetics, the emphasis in this chapter
is on defining the intrinsic effects of such factors as am-
monia concentration, temperature, and pH. This informa-
tion is intended to serve as the conceptual basis for
Chapter 6, which addresses design aspects of specific
nitrification process systems and the influence of such
kinetic factors on design.
3.2 Fundamentals of Nitrification
The nitrification process is carried out by bacterial popu-
lations that sequentially oxidize ammonium to nitrate with
intermediate formation of nitrite. The two principal genera
of importance for carrying out this process are Nitroso-
monas and Nitrobacter. Both of these groups are classi-
fied as autotrophic organisms because they derive energy
for growth from the oxidation of inorganic nitrogen com-
pounds. In contrast, heterotrophic bacteria derive energy
from the oxidation of organic matter. Another feature of
these organisms is that they use inorganic carbon (carbon
dioxide) for synthesis rather than organic carbon. The two
groups are distinguished from one another by their ability
to oxidize only specific species of nitrogen compounds.
While Nitrosomonas can oxidize ammonium to nitrite but
cannot complete the oxidation to nitrate, Nitrobacter is
limited to the oxidation of nitrite to nitrate. Since complete
nitrification is a sequential reaction, treatment process
systems must be designed to provide an environment
suitable for the growth of both groups of nitrifying bacteria.
3.2.1 Metabolism and Stoichiometry
On a biochemical level the nitrification process involves
more than the sequential oxidation of ammonia to nitrite
by Nitrosomonas and nitrite to nitrate by Nitrobacter. Vari-
ous reaction intermediates and enzymes are involved (1).
Rather than discuss these pathways, however, this man-
ual focuses on the response of the nitrification organisms
to environmental conditions. This information is important
for engineering nitrification process systems that ensure
that the resident nitrifiers are able to carry out their me-
tabolic activities efficiently.
The stoichiometric equation for the oxidation of ammo-
nium to nitrite by Nitrosomonas is:
NHJ + 1.5 O2 -»2H+ + H2O +
(3-1)
The release of free energy by this reaction at conditions
within the cell has been estimated by various researchers
to be between 58 and 84 kcal/mole of ammonium (1,2).
The reaction for the oxidation of nitrite to nitrate by Nitro-
bacter is:
i + 0.5O2->NOi
(3-2)
This reaction has been estimated to release 15.4-20.9
kcal/mole of nitrite at conditions typically found within
microbial cells (2). Thus, Nitrosomonas obtains more en-
ergy/mole of nitrogen oxidized than Nitrobacter. If it is
assumed that the amount of cell mass produced is pro-
portional to the degree of energy release, there should
be a greater mass of Nitrosomonas formed than Nitro-
bacter per mole of nitrogen oxidized. This is indeed the
case, as will be discussed.
The expression for overall oxidation of ammonium by both
groups is obtained by adding Equations 3-1 and 3-2:
NHJ + 2 O2 -» NO5 + 2 H+ + H2O
(3-3)
The equations for the synthesis of Nitrosomonas and
Nitrobacter are shown in Equations 3-4 and 3-5, respec-
tively. These assume that the empirical formulation of
bacterial cells is Q5H7NO2:
1 3 NKJ + 1 5 CO2 -» 10 NO5 + 3 C5H7NO2 + 23 H+ + 4 H2O
Nitrosomonas (3-4)
10
5 CO2 + NHJ + 2 H2O -
Nitrobacter
10
§ + C5H7NO2 + H+
(3-5)
Bacterial cells grow by coupling the reactions that pro-
duce energy (Equations 3-1 and 3-2) with those involving
cell synthesis (Equations 3-4 and 3-5). Thus, cell synthesis
can be described by combining the equations for energy
87
-------
yield and cell synthesis. The efficiency of the organisms
in converting the released energy into biomass dictates
how these equations are combined. Efficiency can be
measured in terms of the observed yield, expressed as
the cell mass produced/mass of substrate utilized. This
yield coefficient is normally specified as the mass of VSS
produced/mass of ammonium or nitrite oxidized.
Yield values calculated from theoretical energy release
relationships are 0.29 g VSS/g of NHJ-N and 0.084 g
VSS/g of NOj>-N (2). Yield values observed in experimen-
tation are lower: for the oxidation of ammonium to nitrite
by Nitrosomonas, they are 0.04-0.13 g VSS/g NHJ-N;
and for the oxidation of nitrite to nitrate by Nitrobacter
0.02-0.07 g VSS/g NOj-N (1). Observed yields may be
lower than theoretical yields because a fraction of the free
energy released by oxidation is diverted to microbial
maintenance functions. The total yield of nitrifiers, when
considering nitrification as a single-step process from am-
monium to nitrate, is 0.06-0.20 g VSS/g NHJ-N oxidized.
The observed yield will vary with changing environmental
conditions and with changes in the rate of growth of the
microbial cells, in part accounting for the range of ob-
served yield values. Observed yield is the net yield of
microbial cells, which takes into account the process of
endogenous decay. The effect of endogenous decay on
net yield, however, is typically not considered significant,
given the uncertainty in estimating actual Nitrosomonas
yields. As such, the net yield coefficient for Nitrosomonas
is often considered an estimate of the true yield coeffi-
cient (3).
Equations for synthesis of Nitrosomonas and Nitrobacter,
using yields of 0.08 g VSS/g NHJ-N and 0.05 g VSS/g
NO2-N, respectively, are as follows:
1.00 NH| + 1.44 O2 + 0.0496CQ2 ->
0.01 C5H7O2N + 0.990 NOi +
0.970 H2O +1.99 H+
(3-6)
1.00 NOi + 0.00619 NH| + 0.031 CO2 +
0.0124 H2O + 0.50 O2->
0.00619 C5H7O2N + 1.00 NOg + 0.00619 H+ (3-7)
Combining these equations, the overall reaction describ-
ing complete nitrification is:
1.00 NHJ + 1.89 O2 + 0.0805 CO2 -»
0.0161 CSH7O2N + 0.952 H2O + 0.984
N05+1.98H+
(3-8)
The implications of Equation 3-8 on the design of nitrifi-
cation systems are significant. The stoichiometric coeffi-
cients imply that per mole of ammonium removed the
nitrification process requires a significant amount of oxy-
gen, produces a small amount of biomass, and results in
substantial destruction of alkalinity through the production
of hydrogen ions. For example, according to Equation
3-8, synthesis and oxidation of 20 mg/L ammonium-nitro-
gen (the equivalent of 25.7 mg/L ammonium) would result
in the consumption of 86.4 mg/L oxygen, the production
of 2.6 mg/L nitrifying organisms, and the destruction of
141.4 mg/L alkalinity (as CaCO3). Values for oxygen utili-
zation, biomass yield and alkalinity destruction coeffi-
cients that are generally accepted in practice for
designing nitrification systems are listed in Table 3-1.
Note that the oxygen utilization coefficient of 4.6 is con-
servative in that it reflects the energy reaction only (Equa-
tion 3-3) and does not consider N used for cell synthesis,
Table 3-1. Oxygen Utilization, Biomass Yield, and Alkalin-
ity Destruction Coefficients Acceptable for Design of Nitri-
fication Systems
Parameter
Oxygen
utilization
Equation
g O2 required
n NIHt-M
Coefficient
4.6
Biomass yield g VSS produced '(as nitrifiers)
g NH|-N
g alkalinity (as CaCO3)
Alkalinity
destroyed
0.1
7.1
g NHJ-N
It is important to understand that in virtually all nitrification
systems treating municipal wastewaters, biodegradable
compounds other than ammonium are present. These
compounds will have an influence on total oxygen utiliza-
tion, biomass production, and alkalinity destruction.
Nitrification reactions take place in an aqueous environ-
ment. Thus, the production of free acid (H+) and the con-
sumption of gaseous carbon dioxide (CO2), as described
by Equations 3-4 and 3-5, will impact the aqueous car-
bonic acid system equilibria (4). This, in turn, can affect
the nitrification reactor pH. As will be discussed in Section
3.3.4, pH affects the growth rate of the nitrifiers. As a
result, engineering decisions on the selection of a nitrifi-
cation reactor system (e.g., pure oxygen versus air-based
systems) will influence the resulting pH conditions.
3.3 Nitrification Kinetics
In the context of the nitrification reaction equations that
have been presented, kinetics can be considered as the
study of the factors influencing the rates of these reac-
tions, and as explanations for these rates (5). Ammonium
removal in the nitrification process occurs through micro-
bial synthesis, or growth, and oxidation, according to
Equations 3-6 and 3-7. In this section, kinetic expressions
will be developed to describe the rate of nitrifier growth
and ammonium oxidation, and the impact that a number
of environmental factors have on these rates will be con-
sidered. Other factors that impact the efficiency and per-
88
-------
formance of nitrification reactors also will be discussed,
including the feed organic carbon (CBOD) to nitrogen
ratio, diffusional limitations, and the influence of reductive
zones. Although reference will be made to data derived
from operating systems, descriptions of these systems as
well as information pertaining to their design and perform-
ance are presented in Chapter 6.
3.3.1 Kinetics of Biomass Growth and Ammonia
Utilization
A description of ammonium and nitrite oxidation can be
derived from an examination of the growth kinetics of
Nitrosomonas and Nitrobacter. Nitrosomonas growth is
limited by the concentration of ammonium, while Nitro-
bacter growth is limited by the concentration of nitrite. The
kinetic equation proposed by Monod (6) is used to de-
scribe the kinetics of biological growth of either Nitroso-
monas or Nitrobacter.
KS-
(3-9)
where:
u. = specific growth rate of microorganisms, d~1
jl = maximum specific growth rate of microorganisms,
d'1
Ks = half-saturation or half-velocity coefficient
(equivalent to the growth-limiting substrate
concentration at half the maximum specific
growth rate), mg/L
S = growth-limiting substrate concentration, mg/L
Nitrite normally does not accumulate in large amounts in
biological treatment systems under steady-state condi-
tions. This is because the maximum growth rate of Nitro-
bacter is considerably higher than the maximum growth
rate of Nitrosomonas and Ks values for both organisms
are less than 1 mg/L N at temperatures below 20°C
(68°F). For this reason, the rate of nitrifier growth can be
modeled with Equation 3-9, using the conversion of am-
monium to nitrite as the rate-limiting step:
N
(3-10)
where:
1
UN = specific growth rate of Nitrosomonas, d
jlN = maximum specific growth rate of Nitrosomonas,
d-1
KN = half-saturation coefficient for Nitrosomonas, mg/L
NHJ-N
N = NHJ-N concentration, mg/L
Although the Monod expression is the most widely ac-
cepted approach for describing microbial growth kinetics
and is acceptable for practical engineering design, it has
certain theoretical deficiencies. These are particularly ap-
parent when the expression is used to describe proc-
esses that may involve multiple substrate-limiting condi-
tions (e.g., microbial growth limited by ammonium or
oxygen under transient versus steady state conditions)
and associated multiple organism groups. It is important
to recognize the shortcomings of the Monod expression
in applications of nitrification kinetics since mass-trans-
port or diffusional resistances, heterotrophic/nitrifier com-
petition, and transient conditions (7) can sometimes
negate the assumption that ammonium conversion to ni-
trite is the rate-limiting step in the nitrification process.
The rate of ammonium oxidation is controlled by the
growth of Nitrosomonas and is related to this growth by
the Nitrosomonas yield coefficient. The relationship be-
tween the oxidation rate and the growth rate of Nitroso-
monas can be expressed as follows:
N
(3-11)
where:
QN = ammonium oxidation rate, g NHJ-N oxidized/g
VSS/d
O.N = maximum ammonium oxidation rate, g NHJ-N
oxidized/g VSS/d
YN = organism yield coefficient, g Nitrosomonas
grown (VSS)/g NHJ-N removed
The growth of microorganisms may be expressed in
terms of their doubling, or generation, time. Generation
times of heterotrophic bacteria, those responsible for car-
bonaceous oxidation or CBOD removal, are normally re-
ported at 10-20 times less than the generation times for
nitrifiers (8). Because of the slow growth rate of nitrifiers,
a sufficient solids retention time (also referred to as the-
mean cell residence time or sludge age) is essential in
nitrification process systems in order to retain an ade-
quate population of these organisms. The solids retention
time in a biological system is normally defined as:
e(
where:
= (total mass of biological solids)
(total mass of biological solids
leaving the system/d)
(3-12)
90 = the solids retention time (or sludge age or mean
cell residence time), d
At steady state, the solids leaving the system will be equal
to the solids produced. Therefore, the growth rate and
solids retention time of the organisms in the system are
related by:
= MN - bN = M-'N
(3-13)
89
-------
where:
U/N as net specific growth rate of nitrifiers, d"1
btf = endogenous decay coefficient for nitrifiers, d"1
With nitrifying organisms, DN is often considered to be
negligible (i.e., bu = 0), in which case the specific growth
rate, u^, is the same as the net specific growth rate,
HV
Values for the maximum specific Nitrosomonas growth
rate and the corresponding half-saturation coefficient
are presented in Table 3-2. These are typical of those
reported in the literature. The values of UN are at least
an order of magnitude smaller than typical u. values for
heterotrophs, implying the need for a much longer solids
retention time for biological systems designed to achieve
nitrification versus only carbon oxidation. The values of
KM presented in Table 3-2, although quite low, exceed
values reported elsewhere (4).
Tabla 3-2. Maximum Specific Growth Rates and Half-
Saturation Coefficient Values for Nitrosomonas at Constant
Temperature (20°C) (Adapted from Reference 9)
mg/L NHJ-N
Reference
1.32
0.84
1.62
3.6
1.0
0.6
10
3
11
The significance of low KN values is clear from examina-
tion of Equations 3-10 and 3-11. When KN is low with
respect to N, the growth rate and the ammonium oxidation
rate are independent of the concentration of ammonium
and the Nitrosomonas organisms are growing at their
maximum rate. In complete mix activated sludge systems,
however, N can be lower than KN, in which case kinetics
approach first order (i.e., growth rate is dependent on
substrate concentration). The independence of growth
rate from substrate concentration, characterized as zero-
order kinetics, has been observed by a number of re-
searchers (12-15).
The maximum specific growth rate coefficient of Nitroso-
monas is highly dependent on the constituents in the
wastewater and should be determined experimentally,
particularly when treating an industrial wastewater or a
municipal wastewater with a significant industrial input. A
Dimple bench scale, laboratory procedure for determining
UN is described in detail elsewhere (16).
A number of environmental factors significantly influence
nitrifier growth rates, thus impacting the minimum cell
residence time required to ensure sufficient buildup and
retention of nitrifiers in a biological system. While factors
affecting process kinetics may not influence the intrinsic
nitrifier growth rates, they will affect the selection of other
process design parameter values. For example, mass
transport or diffusional resistances in attached growth re-
actors will increase the required reactor solids retention
time as the nitrifiers are no longer operating at their in-
trinsic growth rates. The effect of such factors on nitrifi-
cation kinetics, together with the effect of environmental
factors, will be considered in the following sections.
3.3.2 Temperature Effects
The nitrification process occurs over a range of approxi-
mately 4-45°C (39-113°F), with about 35°C (95°F) opti-
mum for Nitrosomonas (17) and 35^-2°C (95-108°F)
optimum for Nitrobacter (18,19). The process has been
shown to be strongly dependent on temperature. Quan-
tifying the temperature effect with confidence is difficult,
as demonstrated by the widely reported observations in
the literature. The collection of conclusive data is compli-
cated, in part, by the fact that both the maximum growth
rate and the half-velocity coefficients of nitrification are
temperature sensitive (4).
Conservative estimates for the maximum growth rate of
Nitrosomonas over the temperature range of 10-30°C
(50-86°F) are presented in Table 3-3 (20). The fact that
the 20°C (68°F) value for (IN in Table 3-3 is less than
those presented in Table 3-2 simply illustrates the vari-
ation in rates reported in the literature.
Table 3-3. Maximum Specific Growth Rate Values for
Nitrosomonas as a Function of Temperature
Temperature, °C UN, d~1
10
20
30
0.3
0.65
1.2
The values in Table 3-3 agree reasonably with the van't
Hoff-Arrhenius equation, which predicts the doubling of
growth rates with each 10°C increment in temperature
(20). Arrhenius-type relationships have been observed by
a number of researchers who have measured the oxida-
tion rate of Nitrosomonas as a function of temperature in
various environments over the 5-30°C (41-86°F) range.
The nitrification rate has been observed to decrease
above 30-35°C (86-95°F) (4). This apparent optimum
temperature range is the result of two interactive proc-
esses: the anticipated increase in reaction rate with in-
crease in temperature and protein denaturation above a
critical temperature. For design purposes, an acceptable
Arrhenius-type expression of the effect of temperature on
the maximum growth rate of Nitrosomonas over a tem-
perature range of 5-30°C (41-86°F) is:
(3-14)
90
-------
where:
3.3.3 Effect of DO Concentration
T = temperature, °C
This is graphically displayed on Figure 3-1. Originally
presented in the 1975 edition of this manual (4), sub-
sequent studies (23-27) have tended to confirm the ex-
pression. Although other expressions have been cited
(4,20,22) and have been found to be as acceptable,
Equation 3-14 has found some consensus among design-
ers and is used as the default value in the International
Association on Water Pollution Research and Control
(IAWPRC) model for suspended growth process design
(see Chapter 5). Although KN has also been reported to
vary according to an Arrhenius-type relationship (11), the
low value of the coefficient and the reported range of
values—even at a constant temperature (Table 3-2)—im-
ply that selecting a constant value of 1.0 mg/L NHJ-N
should be acceptable for design purposes.
The variation in u.N and KN in the literature (even at con-
stant temperature) may be related in part to reactor
biomass concentration differences, according to the work
of Shammas (21). Using biomass derived from identical
fill-and-draw, completely mixed activated sludge reactors,
Shammas concluded that there is an interaction between
biomass concentration, temperature, and pH. Conse-
quently, he developed a relationship that expresses tem-
perature sensitivity as a function of the reactor biomass
concentration.
The concentration of DO has a significant effect on the
rates of nitrifier growth and nitrification in biological waste
treatment systems. By modeling the growth of Nitroso-
monas according to the Monod equation (Equation 3-9),
with DO as the growth-limiting substrate concentration,
values for the half-saturation coefficient have been re-
ported as 0.15-2.0 mg/L O2 (4). Evidence suggests that
the value for the coefficient increases with increasing
temperature (28).
Historically, the influence of DO on nitrification rates has
been controversial. Qualitative observations imply that
under certain conditions complete nitrification can be
achieved in biological systems at DO levels as low as 0.5
mg/L (4). A recent comprehensive study sought to provide
a clearer quantification of the effects of DO on nitrification
and to identify interdependent factors affecting the rela-
tionship (7). The following can be implied from the results
of the study, when defining the relationship between DO
and nitrification kinetics:
• The value of DO at which nitrification is limited can be
0.5-2.5 mg/L in either suspended or attached growth ,
systems under steady state conditions, depending on
the degree of mass-transport or diffusional resistances
and the solids retention time.
• A high solids retention time may be required to en-
sure complete nitrification at low DO concentrations,
and for conditions where diffusional resistances are
significant.
100
80
GC
c
o
1
O
I
60
40
o
8?
20
A Pure Culture
B River Water
C Estuary Water
D Activated Sludge
10 15 20 25
Temperature, °C
30
35
Figure 3-1. Effect of temperature on oxidation of ammonium by Nitrosomonas (adapted from Reference 21).
91
-------
• Under transient conditions of organic shock load-
ing, diffusional resistances and heterotrophic/nitrifier
competition can increase the limiting DO value
significantly.
• Under transient conditions, nitrite conversion to nitrate
can become the rate-limiting step in the nitrification
process; in such conditions, the resulting accumula-
tion of nitrite is not correlated to low DO values.
It can be concluded from this study, as well as from other
recent (29) and past studies (30), that the intrinsic growth
rate of Nitrosomonas is not limited at DO concentrations
above 1.0 mg/L, but that DO concentrations greater than
2.0 mg/L may be required in practice. When designing
the aeration or oxygen addition component of a sus-
pended growth nitrification system, it is recommended
that a minimum DO level of 2.0 mg/L be specified at all
times throughout the biological reactor to prevent peak
load ammonia bleed-through. If significant, occasional
transient conditions are anticipated, consideration should
be given to providing standby DO capacity.
If mass-transport or diffusional resistances are an inher-
ent characteristic of the nitrification reactor, as is the case
with attached growth reactors, the DO level achievable in
designing the oxygen addition component should be rela-
tively high. Recent research work suggests that bulk fluid
DO levels should be near 70 percent saturation. Lower
levels may suggest mass-transfer limitations and limited
ventilation (31). These considerations are discussed in
more detail in Section 6.5.2.
3.3.4 pH and Alkalinity Effects
When the equation describing the complete nitrification
process (Equation 3-8) is written in the context of the
carbonic acid system, a substantial destruction of alkalin-
ity is implied. It can be shown (4) that over a pH range
of approximately 5 to 8 in an aqueous biological reactor,
the equilibrium pH of the reactor will be dictated by the
amount of alkalinity and CO2 present in the system.
Higher pH levels can be maintained at lower alkalinity
levels in systems in which the stripping of CO2 occurs in
the biological reactor. Where the stripping of CO2 does
not occur, as is the case in enclosed systems, the alka-
linity of the wastewater must be 10 times greater than the
amount of ammonium nitrified in order to maintain a pH
greater than 6.0 (32). Recall that the theoretical alkalinity
destruction ratio is 7.1 mg (as CaCO3)/mg of ammonium-
nitrogen oxidized. The observed alkalinity destruction ratio
has generally been equal to or less than the theoretical
value in open systems using air as a source of oxygen (4).
Further information on the effect of particular aeration
systems on the resulting reactor pH is provided in Section
6.4.10.4. The incorporation of a phosphorus removal ca-
pacity into nitrification systems through the addition of
chemicals to the reactor(s) will also affect the reactor
alkalinity. Information on this consideration is presented
in Section 6.4.10.2 and Table 2-3.
Reactor pH conditions have been found to have a signifi-
cant effect on the rate of nitrification, as summarized in
Figure 3-2. The degree of acclimation to the correspond-
ing pH is also annotated on the figure. A wide range of
optimum pH has been reported; an almost universal find-
ing, however, is that as the pH moves to the acid range,
the rate of ammonium oxidation declines. This tendency
has been found to be true for both unacclimated and
acclimated cultures, although acclimation, or selection of
a different population of organisms with time, tends to
moderate pH effects. In one study involving an attached
growth reactor, nitrification declined by 50 percent at pH
6.0 after 1.5 d of acclimation, but no decline in njtrification
performance was evident after acclimation for 10 d (32).
In another study it was found that an abrupt change in
reactor pH from 7.2 to 6.4 had no adverse effect on
nitrification. However, when the pH was abruptly changed
from 7.2 to 5.8, nitrification performance deteriorated
markedly as effluent ammonium levels rose from approxi-
mately zero to 11 mg/L NHJ-N. A return to pH 7.2 caused
rapid improvement, indicating that the lower pH was only
inhibitory and not toxic (45).
For design purposes, it is sufficient to take into consid-
eration that the nitrification rate may drop significantly as
pH is lowered below the neutral range and that for per-
formance stability it is best to maintain pH at 6.5-8.0. The
effect of lower pH conditions, if they are anticipated,
should not be ignored when sizing nitrification reactors,
even though acclimation will attenuate the effect of pH on
the nitrification rate.
3.3.5 Effect of Inhibitors
Nitrifying organisms are susceptible to a wide array of
organic and inorganic inhibitors. As pointed out by Stover
(46), nitrifiers can adapt to many inhibitory compounds
when inhibitors are constantly present in the wastewater
versus when slug discharges occur (e.g., from an acci-
dental industrial discharge). Inhibition can occur through
interference with the general metabolism of the cell or
with the primary oxidative reactions. More important than
distinguishing the mechanism of inhibition, however, is
the need to establish a methodology for assessing the
potential for, or occurrence of, nitrification inhibition in a
biological system. Such procedures have been proposed
by numerous researchers (46-48). More on design con-
siderations that deal with the issue of nitrification inhibition
is provided in Chapter 6 (Section 6.3.1).
Extensive reviews of the influence of selected inorganics
and/or organics on nitrification inhibition have been pre-
pared by Neufeld's group (49), Hockenbury and Grady
(50), Pantea-Kiser's group (47), and Painter (51). While
the data base on nitrification inhibition is extensive, Table
3-4 provides a list of several industrially significant or-
92
-------
ioo r
cc
I
o
CD
a
£
40
20
Acclimated Results
Unacclimated Results
Acclimation Uncertain
Reference Key
A - Reference 36
B - Reference 42
C - Reference 37
D - Reference 38
E - Reference 43
F - Reference 39
G - Reference 40
H - Reference 44
I - Reference 45
J - Reference 33
K - Reference 34
L - Reference 35
M - Reference 41
10.0
Figure 3-2. Effect of reactor pH conditions on rate of nitrification.
ganic chemicals found to cause some degree of nitrifica-
tion inhibition. Care must be taken, however, when inter-
preting reported concentrations of inhibitory compounds,
since acclimation can occur and effectively remove the
inhibitory effect from a system; in a complete mixed sys-
tem, the nitrifiers will normally see significantly lower con-
centrations than present in the influent, and suggested
levels are often considerably higher than could occur in
typical collection systems, particularly where pretreat-
ment programs are in place. The reported data should
be used as references for the relative effect of specific
compounds.
Certain inorganics, including specific metals, are inhibi-
tory to nitrifiers. Sawyer, on reviewing studies carried out
in England, suggested that 10-20 mg/L of heavy metal
can be tolerated due to low ionic concentrations at pH
values of 7.5-8.0. Inorganic compounds identified as po-
tential inhibitors are listed in Table 3-5.
Nitrifying organisms are also sensitive to certain forms of
nitrogen. Un-ionized ammonia (NH3), or free ammonia
(FA), and un-ionized nitrous acid (HNO2), or free nitrous
acid (FNA), are believed to be inhibitory to nitrifiers above
certain concentrations. FA begins to inhibit Nitrosomonas
at a concentration of 10-150 mg/L and Nitrobacter in the
range of 0.1-1.0 mg/L (56). FNA begins to inhibit Nitroso-
monas and Nitrobacter at concentrations of 0.22-2.8
mg/L. The FA and FNA concentrations are directly corre-
lated to pH and temperature, and the concentration, re-
spectively, of ammonia plus ammonium and nitrite plus
93
-------
Table 3-4. Industrially Significant Organic Compounds
Inhibiting Nitrification (Adapted from Reference 50)
Compound
Concentration of Compound
Giving at Least 50 Percent
Inhibition, mg/L
Acetone 2,000
Carbon disulfide 38
Chloroform 18
Ethanol 2,400
Phenol 5.6
Ethylenadlamlne 17
Hexamethylene diamine 85
Aniline <1
Monoethanolamine <200
Table 3-5. Metals and Inorganic Compounds Identified as
Potential Nitrification Inhibitors
Compound
References
Zinc
Frea Cyanide
Perohlorate
Copper
Mercury
Chromium
Nickel
Silver
Cobalt
Thiocyanate
Sodium cyanide
Sodium azlde
Hydrazine
Sodium cyanate
Potassium chromate
Cadmium
Arsenic (trivalent)
Fluoride
Lead
1,51
49
1
1,54
1
1, 53, 55
1, 52, 54, 55
1
51
49
52
52
52
52
52
54
53
53
55
nitrous acid. FA and FNA are present in accordance with
the following equilibrium reactions:
,0 (3-15)
(3-16)
Threshold levels of ammonia plus ammonium-nitrogen,
and nitrite plus nitrous acid-nitrogen at which nitrification
inhibition may begin at a pH of 7.0 and a temperature of
20°C (68°F) are presented in Table 3-6 for illustrative
purposes. (Values for other pH and temperature condi-
Table 3-6. Calculated Threshold Values of Ammonia Plus
Ammonium-Nitrogen and Nitrite Plus Nitrous Acid-Nitrogen
Where Nitrification Inhibition May Begin (from Reference
56)
Equivalent
Nitrite plus
Nitrous Acid-N
atpH
7.0 and 20°C,
mg/L
Inhibitory
FA or FNA
Concentration, mg/L
Equivalent
Ammonia plus
Ammonium-N at
pH 7.0 and 20°C,
mg/L
FA
10 (Nitrosomonas
Inhibition)
Q.~\(Nitrobacter
Inhibition)
1,000
20
FNA
0.22 (Nitrification
Inhibition)
280
tions can be calculated [56].) The calculated values imply
that it is unlikely that nitrification inhibition will occur as a
result of the presence of ammonia plus ammonium and
nitrite plus nitrous acid in the treatment of typical munici-
pal wastewaters. However, sludge discharges into mu-
nicipal systems of highly concentrated industrial wastes
containing these forms of nitrogen can cause inhibition.
If anaerobic digestion is incorporated into a wastewater
treatment plant and if untreated supernatant is returned
to the process, a suitable reduction in the nitrification rate
should be made. The growth rate of Nitrosomonas in a
suspended growth reactor treating municipal wastewater
can be inhibited by introduction of digester supernatant,
according to Gujer (57) and others (58,59). Gujer's results
indicate that the inclusion of digester supernatant recycle,
to the extent that the ammonium-nitrogen concentration
increases by 5 mg/L, can reduce the growth rate of Ni-
trosomonas by approximately 20 percent. The study as-
sumed that the inhibiting compound was produced as a
by-product of anaerobic degradation, versus any change
in process conditions in the suspended growth reactor
resulting from introduction of the digester supernatant.
3.3.6 Effect of Feed Organic Carbon to Nitrogen Ratio
The ratio of the feed biodegradable organic carbon, or
CBOD, to the nitrogen available for nitrification in the
wastewater (i.e., the C:N) is one of the critical factors
affecting the design of nitrification systems. (A discussion
of the C:N may also be found in Chapter 2 (Section
2.4.1.4).) Normally, for all nitrification systems, there is
sufficient organic matter in the reactor feed to enable the
growth of heterotrophic bacteria. Since the yield of het-
erotrophic bacteria is greater than the yield of the autot-
rophic nitrifying bacteria, there is a danger, when
attempting to control the MLSS at a desired level, that
the growth rate of the heterotrophic organisms will be
94
-------
established at a value exceeding the maximum possible
growth rate of the nitrifying organisms. Under such con-
ditions the nitrifiers will be washed out of the system.
Thus, in order to build and sustain a sufficient nitrifying
population, the following condition must be satisfied, rec-
ognizing the relationship between growth rate and solids
retention time:
(3-17)
where:
6£ = design solids retention time, d
0™ = minimum solids retention time for nitrification at
environmental conditions within the reactor, d
The net specific growth rate of the heterotrophic popula-
tion can be expressed as:
(3-18)
where:
U-'H = net specific growth rate of heterotrophs, d"1
YH = true heterotrophic yield coefficient, g VSS
grown/g of substrate CBOD (or COD) removed
qH = rate of organic removal, g CBOD (or COD)
removed/g of active VSS/d
bH = decay coefficient, d"1
(3-19)
The rate of organic removal is defined as:
Sp - S-|
where:
S0 = feed total CBOD (or COD), mg/L
S1 = effluent soluble CBOD (or COD), mg/L
t = reactor hydraulic retention time, d
XT = reactor active total VSS, mg/L
Methods have been proposed for estimating X! in Equa-
tion 3-19 by measuring the reactor total VSS (X) and
correcting for inactive volatile solids as a result of endo-
genous decay (60,61), and accounting for the refractory
VSS entering the reactor with the feed. The issue of
predicting reactor total VSS has been dealt with recently
in a more rigorous fashion in the activated sludge model
development work of the IAWPRC, as referenced in
Chapter 5. Predicting X can be simplified by using an
observed or net biomass yield coefficient (YNEr), as pro-
posed in Chapter 2 (Figure 2-8). It is evident from Equa-
tions 3-17, 3-18, and 3-19 that once 9^ is selected, which
is dictated by the anticipated growth rate of the nitrifiers,
the required reactor hydraulic retention time and total
VSS will depend on the feed biodegradable organic carb-
on concentration, since YH and bH are assumed to be
constant and Si will be at a minimal level in any reactor
in which nitrification is occurring.
It is also evident from the preceding discussion that com-
bining Equations 3-17, 3-18, and 3-19 not only illustrates
how the feed organic carbon effects the nitrification reac-
tor design, but effectively implies a procedure for design.
The first step in the procedure is to define the design
solids retention time. The use of this solids retention time
approach for sizing suspended growth reactors is dis-
cussed further in Section 6.4.1.1.
An alternative to the solids retention time approach calls
for determining a design ammonium oxidation or nitrifica-
tion rate. Equation 3-11 can be expressed as follows if
the nitrification rate is zero order (KN«N) with respect to
ammonium concentration:
NQ-NT
XNt
(3-20)
(3-21)
where:
NO - N! = NHJ-N nitrified, mg/L
XN = reactor active Nitrosomonas as VSS, mg/L
Equation 3-21 represents a simplified relationship for de-
sign, provided XN can be determined. An estimate of XN
can be made with knowledge of the ratio of the feed
biodegradable organic carbon and ammonium-nitrogen,
and the active total VSS:
Y'N (N0-
-X1 (3-22)
where:
Y'N = organism yield coefficient for Nitrosomonas plus
Nitrobacter, g VSS/g NHJ-N removed
Equation 3-22 assumes decay coefficients for nitrifiers and
heterotrophs are negligible. It is clear from Equations 3-21
and 3-22 that wide variation in the maximum ammonium
oxidation rate can be observed under similar environ-
mental conditions, when the rate is calculated based on
the measurement of total VSS. This variation is due to
differences in the feed C:N. Methods are available to
determine X^ by measuring the reactor total VSS and
correcting for the effect of organism decay and inert feed
volatile solids. By combining Equations 3-13 and 3-20,
the minimum solids retention time for nitrification, G™, can
be related to the maximum ammonium oxidation rate:
1
YNqN
(3-23)
This discussion suggests that using the nitrification rate
approach can lead to errors in reactor sizing, unless the
active solids basis to which the rate is normalized is
explicitly understood. For this reason, the solids retention
time approach to design has been favored.
95
-------
3.3.7 Influence of Oxidative-Reductive Environments
Combining the processes of carbon oxidation, nitrifica-
tion, denitrification and biological phosphorus removal in
one or more reactors has become a favored technology
application for control of nutrients from municipal waste-
waters. In these systems, the nitrifiers are exposed to
aerobic, anoxic (i.e., nitrate or nitrite present, but no DO),
and anaerobic conditions. Anoxic or anaerobic zones are
often incorporated into suspended growth systems de-
signed for carbonaceous oxidation and nitrification to pro-
mote biological phosphorus removal and/or to control
filamentous bulking. Recognizing the effect of DO on the
nitrifiers raises concern about the design of nitrification
systems in which the organisms are alternately exposed
to conditions of low or zero DO. It has been reported that
nltrifier activity is unaffected when exposed to anoxic con-
ditions of up to 5 hr (62,63). The results from pilot plant
studies at Burlington, Ontario, indicate that anaerobic
conditions of up to 4 hr duration have no detrimental effect
on nitrification rates (64). The incorporation of anoxic or
anaerobic zones—referred to as selectors—to control fila-
mentous bulking in full-scale suspended growth systems,
had no reported effect on nitrification efficiency at selector
hydraulic retention times of approximately 1.5 hr (65).
Overall, available information leads to the reasonably
safe conclusion that anoxic or anaerobic conditions last-
ing for hours have no tangible impact on nitrifier viability
when acceptable DO concentrations are restored.
3.4 Attached Growth Kinetic
Considerations
The kinetic description of fixed film system performance
involves considerations beyond those presented in Sec-
tion 3.3. Development of a particular biofilm reactor ki-
netic model requires applying the nitrification kinetic
principles relevant for the nitrifying biomass to a reactor
model that describes its hydrodynamics, mass-transfer
characteristics, and any special features of the reactor. In
contrast to suspended growth systems, transport proc-
esses are generally rate controlling in biofilm systems.
Mass-transport or diffusional resistances can or will influ-
ence the nitrification process in the following ways:
• They can negate the assumption that ammonium con-
version to nitrite is the rate-limiting step in the nitrifica-
tion process.
• They will increase the required reactor solids retention
time.
• They will influence the value of DO in the bulk liquid at
which the nitrification rate is limited.
These observations indicate that although the rate rela-
tionships presented in Section 3.3.1 and the qualitative
relationships describing the effect of various environ-
mental factors on nitrification are applicable to systems
influenced by mass-transport or diffusional resistances,
one must account for these influences. Diffusional resis-
tances are of major concern in attached growth reactors
in which growth occurs on or within a solid media. Since
the design of this type of attached growth reactor is com-
plicated by the need to consider the mass-transport proc-
esses and to define the film surface itself, it is primarily
based on empirical results from pilot and full-scale sys-
tems. Nonetheless, reactor design relationships are con-
sistent with biofilm models developed on the basis of
stoleniometry, Pick's Law, and Monod Kinetics.
It has been found that the conditions in the immediate
proximity of a microorganism in a biofilm are not the same
as those measured in the bulk liquid. The concentration
of substrates within the biofilm vary with depth and are
significantly lower than in the bulk liquid, since they must
be transported into and through the biofilm. Since this is
true for DO and ammonium-nitrogen, lower concentra-
tions within the biofilm can result in lower observed rates
of nitrification than would be predicted based on liquid
ammonia concentrations and the amount of attached
biomass.
Figure 3-3 is a schematic representation of the film theory
model normally used to represent a biofilm. The model
considers both external and internal transport limitations.
The incomplete mixing of the bulk liquid with the liquid
phase immediately adjacent to the biofilm surface indi-
cates that external mass-transfer resistance is an impor-
tant consideration. External transport limitations occur
when substrate must diffuse through a stagnant liquid film
at the attached growth surface. For diffusion to occur, a
concentration gradient must exist. Flux of material
through this layer is generally modeled as follows:
= AD(AS/AL)
(3-24)
where:
J = the flux, mass/time
A = the biofilm surface area, length2
D = the diffusion coefficient of the component of inter-
est, Iength2/time
AS = the difference in substrate concentration between
the bulk liquid and the liquid film at the biofilm
surface, mass/length3
AL = the thickness of the stagnant film, length
As indicated in Equation 3-24, the concentration of sub-
strate at the surface will be lower than the concentration
in the bulk liquid. Moreover, after reaching Jhe biofilm
surface, the substrate must diffuse through the biofilm to
reach the microorganisms within the attached growth ma-
trix. This step will further reduce the substrate concentra-
tion within the biofilm since it requires a concentration
gradient.
Some of the observations that can be made based on
the biofilm model are of interest when considering am-
96
-------
Bulk Diffusion
Liquid Layer Biofilm
%
""\^_
^^**»»^
•^ ^
Vs<<^<\^V^V^N<-
'x'V°%'\VV'\'x\'»''\/*'%'\'V
v$%v$v:v$%
t* o N '^/V^/V^V WK/ X* 'X-
/" S /X^N^-VyS^XyX/^^V^ T ^ \
f / / f / tSSff^tf^'^i*.
V>v/ ^**xXyX^X^SxN^xN/^xX^NxN.
/•\^v/Xr^/'N^v/Nx%/p^/v*'V^xNy'v/>'.
/NX^%xV/%xS/iN^X^X/'w^/%/VS<'\
-• '*^S£ ' **" f* ^ f* t* ^ * S*S*f^.
/VVV^r' Sf 'V\V>'\
N
-------
Conf. on Water Poll. Res., pp. 169-199. Tokyo, Ja-
pan.
3. Downing, A.L., Painter, H.A., and Q. Knowles. 1964.
Nitrification in the activated sludge process. J. Inst.
Sew. Purif. 130.
4. U.S. EPA. 1975. Process design manual for nitrogen
control. EPA/625/1-77/007 (NTIS PB-259149). Wash-
ington, DC.
5. Levenspiel, 0.1967. Chemical reaction engineering:
an introduction to the design of chemical reactors,
New York: John Wiley and Sons, Inc. p. 9.
6. Monod, J. 1942. Researches sur la croissance des
cultures. Paris: Hermann and Cie.
7. Stenstrom, M.K., and S.S. Song. 1991. Effects of
oxygen transport limitation on nitrification in the acti-
vated sludge process. JWPCF 63:208.
8. Lawrence, A.W., and P.L. McCarty. 1970. Unified ba-
sis for biological treatment design and operation. Pro-
ceedings of the ASCE. JSED 96(SA3):757.
9. Grady, Jr., C.P.L., and H.C. Lim. 1980. Biological
wastewater treatment: theory and applications, New
York: Marcel Dekker, Inc. pp. 328-329.
10. Stratton, RE., and P.L. McCarty. 1967. Prediction of
nitrification effects on the dissolved oxygen balance
of streams. Environ. Sci. Tech. 1:405.
11. Knowies, G., A.L. Downing, and M.J. Barrett. 1965.
Determination of kinetic constants for nitrifying bac-
teria in mixed culture, with the aid of an electronic
computer. J. Gen. Microbiol. 38:263.
12. Shrinath, E.G., R.C. Loehr, and T.B.S. Prakasam.
1976. Nitrifying organism concentration and activity.
ASCE. J. Environ. Eng. Div. 102:449.
13. Sutton, P.M., K.L. Murphy, and B.E. Jank. 1977. Ni-
trogen control: a basis for design with activated
sludge systems. Prog. Water Tech. 8:467.
14. Huang, C.S., and N.E. Hopson. 1974. Nitrification
rate in biological processes. ASCE. J. Environ. Eng.
Div. 100:409.
15. Hall, E.R., and K.L. Murphy. 1980. Estimation of ni-
trifying biomass and kinetics in wastewater. Water
Res. 14:297-304.
16. Givens, S.W., E.V. Brown, S.R. Gelman, C.P.L.
Grady, Jr., and D.A. Skedsvold. 1991. Biological
process design and pilot testing for a carbon oxida-
tion, nitrification and denitrification system. Environ.
Prog. 10(2):133.
17. Buswell, A.H., T. Shiota, N. Lawrence, and I. Van
Meter. 1954. Laboratory studies on the kinetics of the
growth of Nitrosomonas with relation to the nitrifica-
tion phase of the BOD test. App. Microbiol. 2:21.
18. Deppe, K., and H. Engel. 1960. Untersuchungen
uber Temperatureabhangigkeit der Nitratbildung
durch Nitrobacter winogradski buch. Bel ungehem-
mten und gehemmten wachstum. Zentbl. Bakt.
Parasitkde 11(113):561.
19. Laudelout, H., and L. Van Tichelen. 1960. Kinetics of
the nitrite oxidation by Nitrobacter winogradski. J.
Bact. 79:39.
20. Water Pollution Control Federation. 1983. Nutrient
control. Manual of Practice No. FD-7. p. 28.
21. Shammas, N.K. 1986. Interactions of temperature,
pH, and biomass on the nitrification process. JWPCF
58:52.
22. Downing, A.L. 1968. Factors to be considered in the
design of activated sludge plants. Proceedings of the
Water Resources Symp. No. 1, University of Texas
Press, Austin, TX 190.
23. Gujer, W. no date. A proposed procedure for the
choice of sludge age in the design of a nitrifying
activated sludge system, Swiss Federal Institute.
24. Sawyer, B., et al. 1979. Estimation of the maximum
growth rate of ammonia oxidizing bacteria growing in
municipal wastewater. Paper presented at the 52nd
Annual Conference of the Water Pollution Control
Federation, Houston, TX (October).
25. Rebhun, M., et al. 1977. Combination of physics-
chemical and biological process for wastewater treat-
ment and reuse; International Conference on
Advanced Wastewater Treatment and Reclamation
of Wastewater, IAWPR, South Africa.
26. Becarri, M., et al. 1979. A critical analysis of nitrifica-
tion alternatives. Water Res. 12(2)185-192.
27. Antoniu et al. 1990. Effect of temp, and pH on the
maximum specific growth rate of nitrifying bacteria.
Water Res. 24(1 ):97-101.
28. Stankewich, M.J., Jr. 1974. Biological nitrification with
the high purity oxygenation process. Proceedings of
the 28th Industrial Waste Conference, Purdue Uni-
versity, Lafayette, IN. Ann Arbor, Ml: Ann Arbor Sci-
ence.
29. Hanaki, K., C. Wantawin, and S. Ohgaki. 1990. Nitri-
fication at low levels of dissolved oxygen with and
without organic loading in a suspended-growth reac-
tor. Water Res. 24:297.
30. Loveless, J.E., and J.E. Painter. 1968. The influence
of metal ion concentration and pH value on the
growth of a nitrosomonas strain isolated from acti-
vated sludge. J. Gen. Microbiol. 52:1.
98
-------
31. Logan, B.E., and D.S. Parker. 1990. Discussion of
nitrification performance of a pilot-scale trickling filter
by H.A. Gullicks, J.L. Cleasby. Res. JWPCF 62:33.
32. Haug, R.T., and P.L. McCarty. 1971. Nitrification with
the submerged filter. Report prepared by the Depart-
ment of Civil Engineering, Stanford University for the
Environmental Protection Agency, Research Grant
No. 17010 EPM (August).
33. Forster, J.R.M. 1974. Studies on nitrification in ma-
rine biological filters. Aquaculture 4:387.
34. Stratta, J.M., D.A. Long, and M.C. Doherty. 1982.
Improvement of nitrification in rotating biological con-
tactors by means of alkaline chemical addition. Pro-
ceedings of the First Int. Conf. Fixed-Film' Biol.
Processes, Y.C. Wu, ed., NTIS AD-A126 376.
35. Engel, M.S., and M. Alexander. 1917. Growth and
metabolism of Nitrosomonas europaea. J. Bact.
76:217.
36. Clayfield, G.W. 1974. Respiration and denitrification
studies on laboratory and works activated sludge.
Water Poll. Control 73(1):51. London.
37. Hall, J.B. 1978. Nitrate-reducing bacteria. Microbiol-
ogy-1978, Ed. by D. Schlessinger, American Society
for Microbiology, Washington, DC.
38. Suzucki, I., V. Dular, and S.C. Kwok. 1953. Ammonia
or ammonium ion as substrate for oxidation by M-
trosomonas europaea cells and extracts. J. Bact.
120:556.
39. Barritt, N.W. 1933. Nitrification process in soils and
biological filters. Ann. Appl. Biol. 20:165-184.
40. Wild, H.E., C.N. Sawyer, and T.C. McMahon. 1971.
Factors affecting nitrification kinetics. JWPCF
43:1845.
41. Meyerhof, O. 1917. Untersuchungen uber den at-
mungsvorgany nitrifizierenden bakterien. Pflugers
Archges Physiol. 166:240.
42. Painter, H.A., and J.E. Loveless. 1983. Effect of tem-
perature and pH value on the growth-rate constants
of nitrifying bacteria in the activated sludge process.
Water Res. 17:238.
43. Laudelout, H., R. Lambert, and M.L. Pham. 1976.
Influence du pH et de la pression partielle d'oxygene
sur la nitrification. Ann. Microbiol. Inst. Pasteur
127A:367-382.
44. Kawai, A., Y. Yoshida, and M. Kimala. 1965. Bio-
chemical studies on the bacteria in the aquarium with
a circulation system-ii. Nitrifying activity of the filter
sand. Bull. Jap. Soc. Sci. Fisheries 31(1):65.
45. Pokuska, R.A., and J.F. Andrews. 1975. Dynamics of
nitrification in the activated sludge process. Proceed-
ings of the 29th Industrial Waste Conference, Purdue
University, Lafayette, IN. Ann Arbor, Ml: Ann Arbor
Science, p. 1005.
46. Stover, E.L. 1980. Biological nitrification inhibition
screening procedures for industrial wastewater. Pro-
ceedings of the 34th Industrial Waste Conference,
Purdue University, Lafayette, IN. Ann Arbor, Ml: Ann
Arbor Science, p. 887.
47. Pantea-Kiser, L., R.F. Wukash, and J.E. Alleman.
1990. The effect of inhibitory compounds on biologi-
cal nitrification. Proceedings of the 44th Industrial
Waste Conference, Purdue University, Lafayette, IN.
Chelsea, Ml: Lewis Publishers, p. 465.
48. Blum, D.J.W., and R.E. Speece. 1991. A database of
chemical toxicity to environmental bacteria and its
use in interspecies comparisons and correlations.
JWPCF 63:198.
49. Neufeld, R.D., J.H. Greenfield, A.J. Hill, C.B. Rieder,
and D.O. Adekoya. 1983. Nitrification inhibition biok-
inetics. EPA/600/2-83/111. Cincinnati, OH.
50. Hockenbury, M.R., and C.P.L. Grady, Jr. 1977. Inhi-
bition of nitrification—effects of selected organic com-
pounds. JWPCF 49:768.
51. Painter, H.A., and J.E. Loveless. 1968. The influence
of metal ion concentration and pH values on the
growth of a Nitrosomonas strain isolated from acti-
vated sludge. J. Gen. Microbiol. 57.
52. Tomlinson, T.G., A.G. Boon, and C.N.A. Trotman.
1966. Inhibition of nitrification in the activated sludge
process of sewage disposal. J. Appl. Bact. 29:266.
53. Beg, S.A., R.H. Siddigi, and S. Ilias. 1982. Inhibition
of nitrification by arsenic, chromium, and fluoride.
JWPCF 54:482.
54. Benmoussa, H., G. Martin, Y. Richard, and A.
Leprince. 1986. Inhibition of nitrification by heavy
metal cations. Water Res. 20:1333.
55. Ibrahim, A.M. 1989. Heavy metal inhibition of resting
nitrifying bacteria. Proceedings of the 43rd Industrial
Waste Conference, Purdue University, Lafayette, IN.
Chelsea, Ml: Lewis Publishers, p. 309.
56. Anthonisen, A.C., R.C. Loehr, T.B.S. Prakasam, and
E.G. Srinath. 1976. Inhibition of nitrification by am-
monia and nitrous acid. JWPCF 48:835.
57. Gujer, W. 1977. Discussion on the paper dynamic
nature of nitrifying biological suspended growth sys-
tems. Prog. Water Tech. 9:279.
58. Tomlinson, E.J., and A.M. Bruce. 1979. Problems of
septicity in biological treatment. Proceedings of the
Symposium on Septic Sewage: Problems and Solu-
tions. Institute of Water Pollution Control 84.
99
-------
59. Plaza, E., B. Hultman, and J. Trela. 1990. Effect of
easily degradable carbon source on nitrogen removal
efficiency. Water Sci. Tech. 22:281.
60. Marais, G.V.R., and G.A. Eharna. 1976. The acti-
vated sludge process part 1—steady state behavior.
Water SA 2:163.
61. Eckenfelder, W.W. 1975. Commentary to position pa-
per by R.E. McKinney, The value and use of mathe-
matical models for activated sludge systems. Prog.
Water Tech. 7:35.
62. Christensen, M.H., and P. Harremoes. 1972. Biologi-
cal denitrification in wastewater treatment. Report 2-
72, Department of Sanitary Engineering, Technical
University of Denmark.
63. Wuhrmann, K. 1960. Effect of oxygen tension on
biological treatment processes. Proceedings of the
Third Conference Biological Waste Treatment, Man-
hattan College, NY, p. 27.
64. Sutton, P.M., B.E. Jank, and D. Vachon. 1980. Nutri-
ent removal in suspended growth systems without
chemical addition. JWPCF 52:98.
65. Daigger, G.T., and G.A. Nicholson. 1990. Perform-
ance of four full-scale nitrifying wastewater treatment
plants incorporating selectors. JWPCF 62:676.
66. Williamson, K.L., and P.L. McCarty. 1973. A model of
substrate utilization by bacterial films. Presented at
the 46th Annual Conference of the Water Pollution
Control Federation, Cincinnati, OH (October).
67. Murphy, K.L., P.M. Sutton, R.W. Wilson, and B.E.
Jank. 1977. Nitrogen control-design considerations
for supported growth systems. JWPCF 49:549.
\68. Paulson, C. 1989. Nitrification for the 90's. Water
Engineering and Management, pp. 57-59 (Septem-
ber).
69. Characklis, W.G., et al. 1989. Modeling of biofilm
systems by IAWPRC task group. Institute for Proc-
ess Analysis, Montana State University, Bozeman,
MT (May).
70. Heath, M.S., S.A. Wirtel, and B.E. Rittman. 1990.
Simplified design of biofilm processes using normal-
ized leading curves. JWPCF 62:185.
71. Kincannon, D.F., and Sherrard, J.H. 1973. Trickling
filter versus activated sludge: when to select each
process. Proceedings of the 28th Purdue Industrial
Waste Conf.
72. Mulbarger, M.C. 1991. Fundamental secondary treat-
ment insights. Presented at the 64th Annual Confer-
ence, Water Pollution Control Federation, Toronto,
Ontario.
100
-------
Chapter 4
Process Chemistry and Kinetics of Biological Denitrification
4.1 Introduction
This chapter,reviews the fundamentals of the chemistry,
biochemistry, and kinetics of denitrification, focusing on
the treatment process. The discussion seeks to provide
an understanding of the underlying principles affecting the
performance, design, and operation of denitrification pro-
cesses. Subsequent chapters deal specifically with the
design and operation of these processes.
Biological denitrification involves the microbial reduction
of nitrate to nitrite, and ultimately nitrite to nitrogen gas.
Nitrate and nitrite replace oxygen for microbial respiration
in this reaction; as such, denitrification is commonly
thought to occur only in the absence of molecular oxygen.
The conditions suitable for denitrification—oxygen is ab-
sent but nitrate is present—are commonly referred to as
anoxic.
Since nitrogen gas is relatively biologically inert, denitri-
fication converts nitrogen from a potentially objectionable
form (nitrate) to a form that has no significant effect on
the environment (nitrogen gas). As discussed in Chap-
ter 1, nitrate in water can be objectionable if nutrient en-
richment is a concern and/or if the water is intended to
be potable. Denitrification in wastewater treatment appli-
cations may also provide process benefits in certain situ-
ations, including the development of alkalinity, the
reduction of oxygen demand, and production of an acti-
vated sludge with better settling characteristics.
4.2 Fundamentals of Denitrification
4.2.1 Microbiology
Unlike nitrification, a relatively broad range of bacteria
can accomplish denitrification. Denitrifiers are ubiquitous
in most natural environments, including municipal waste-
waters and sludges (1,2). Many of the microorganisms in
municipal activated sludge systems are denitrifiers, even
in systems that are not specifically designed for denitrify-
ing. The presence of the organisms is due in part to the
fact that they are facultative: they can use either oxygen
or nitrate as their terminal electron acceptor. Denitrifiers
can proliferate in aerobic systems because of their ability
to use oxygen and efficiently oxidize organic matter (2).
The ubiquity of denitrifiers minimizes the need to create
special environmental conditions for their survival, as
must be done for nitrifiers.
4.2.2 Metabolism and Biochemical Pathways
In the process of denitrification, nitrate and nitrite act as
electron acceptors in the respiratory electron transport
chain in the same manner as oxygen. This transport chain
is the fundamental mechanism by which cells generate
energy. The process involves transferring electrons from
a reduced electron donor (e.g., an organic substrate) to
an oxidized electron acceptor (e.g., oxygen, nitrate, ni-
trite, or sulfate). Nitrate or nitrite may serve as a substitute
for oxygen in this chain with only small modifications to
the metabolic system (i.e., the enzymes) of the bacteria.
By using nitrate or nitrite in place of oxygen in the electron
transport chain, however, slightly less energy is gener-
ated. Similarly, more energy is generated using nitrate
than sulfate.
Control systems exist within individual bacteria and natu-
ral microbial populations that ensure the most efficient
form of energy generation is utilized. Thus, if oxygen is
present, it will be used preferentially over nitrate, and if
oxygen is not present, nitrate will be used preferentially
over sulfate. Since the bacteria that reduce sulfate (sul-
fate reducers) cannot compete effectively with nitrate re-
ducers for the available organic matter, sulfate reduction
to sulfide and resulting odor production are not likely to
occur in a treatment system that is anoxic (i.e., where
nitrate is present). Also, significant sulfate reduction is not
likely to occur in a system that may be void of oxygen
and nitrate for short periods of time (e.g., the few hours
in the anaerobic zone of a biological phosphorus removal
activated sludge system), since the sulfate reducers will
not have adequate time to proliferate in the numbers
required to carry out significant sulfate reduction. More-
over, sulfate reducers may be poisoned in the aerobic
zones of such systems.
The control mechanism in denitrifiers that allows them to
switch from oxygen to nitrate occurs at two levels. The
first is the synthesis of the enzymes required for denitri-
fication. In pure cultures, oxygen has been found to re-
press the synthesis of these enzymes. Between 2 and
101
-------
3 hr is typically required for pure cultures to synthesize
(he enzymes when the cells are switched from an aerobic
to an anoxic environment. Research on activated sludge,
however, suggests that the denitrifying enzymes may be
present even in systems that do not have anoxic condi-
tions (3), suggesting that a period for synthesis of the
denitrifying enzymes is not required prior to initiation of
denitrification.
Control of denitrification also occurs at the level of en-
zyme activity. Oxygen will inhibit the activity of the deni-
trifying enzymes. The oxygen concentration at which
denitrification stops has been reported to be 0.2 mg/L in
pure cultures (4). In activated sludge systems the re-
ported values are 0.3 to 1.5 mg/L (5), possibly because
of the diffusion limitations of microbial floe (4).
Research also suggests some species of bacteria can
concurrentiy use nitrite and oxygen as terminal electron
acceptors (6). Studies by Kugelman et al. (7) suggest that
concurrent oxygen and nitrite utilization in activated
sludge systems operated at high DO levels is possible.
This phenomenon was observed in activated sludges
grown in systems with anaerobic/aerobic conditions, such
as those encountered in biological phosphorus removal
systems. Similar results were not observed in sludges
grown in a purely aerobic nitrifying system. At present,
data are insufficient to indicate whether concurrent oxy-
gen utilization and nitrite reduction is a general phenome-
non with anaerobic/aerobic systems and how this
phenomenon might be exploited.
Total nitrogen losses of up to 30 percent are very often
observed in aerobic nitrifying activated sludge systems.
These losses may be due to concurrent oxygen and ni-
trate/nitrite utilization, denitrification occurring inside the
activated sludge floe under anoxic conditions, or denitri-
fication occurring in clarifiers and other unaerated zones.
Due to the unpredictable nature of these losses, they are
not generally relied upon in the design of systems for
nitrate removal.
The list of organic compounds that can serve as organic
substrates (i.e., as carbon and electron donors) for deni-
trification is especially long. Almost any compound that is
degraded with oxygen as the electron acceptor will also
serve as an electron donor with nitrate. Some aromatic
compounds that do not serve as electron donors for ni-
trate exist (e.g., benzene), possibly because of the need
for oxygen in the enzymatic cleavage of the aromatic ring.
The organic compounds that could be used for the deni-
trification of wastewaters include:
• Organics present in municipal and industrial waste-
waters
• Methanol
• Ethanol
• Acetic acid
• Waste organic material
While organics in wastewater and methanol are the two
most commonly used electron donors, electron donor se-
lection will depend primarily on cost and local availability.
4.2.2. 1 Theoretical Stoichiometric Equations
Theoretical Stoichiometric equations can be used to pre-
dict the mass of electron donor (i.e., carbon substrate)
and acceptor (i.e., oxygen, nitrate, or nitrite) consumed,
and the mass of cells produced during any biological
process. Since denitrification involves the transfer of elec-
trons from donor to acceptor, half-reactions can be used
to develop the Stoichiometric equations. Half-reaction
equations for oxygen, nitrate, and nitrite as electron ac-
ceptors are as follows:
(4-1)
(4-2)
(4-3)
NOi + H+ +
H20 + N2
A comparison of equations 4-1 and 4-2 indicates that 8 g
of oxygen (1/4 x 32 g O2/mole) is equivalent to 2.86 (1/5
x 14 g-N/mole) of nitrate nitrogen. Hence, the reduction
of 1 g of nitrate nitrogen is equivalent to the reduction of
2.86 g of oxygen. Using this Stoichiometric equivalence,
the reduction in oxygen demand that is possible when
denitrification is incorporated into a single-sludge acti-
vated sludge system (i.e., raw wastewater is used as the
carbon and energy source) can be estimated. Since 4.6
g of oxygen are required to oxidize ammonia nitrogen to
nitrate nitrogen (Equation 3-3), and 2.86 g of oxygen
equivalents are recovered in the denitrification of the ni-
trate nitrogen, it is theoretically possible to reduce the net
energy expended in providing oxygen for nitrification by
up to 63 percent by using the raw wastewater for denitri-
fication. Since this factor could provide significant energy
savings, it should be taken into account in evaluating flow
scheme alternatives.
Reduction of the electron acceptors (oxygen, nitrate, or
nitrite) requires an electron donor, which can either be
the organic substrate in the raw wastewater or a substrate
added to the source. The most commonly used external
carbon source is methanol when denitrification is accom-
plished as a separate stage. The half-reaction equation
for methanol as the electron donor is:
(4-4)
Half-reactions also can be written for a variety of other
organic compounds serving as electron donors (8,9).
The reactions in Equations 4-2 and 4-4 can be combined
as follows:
102
-------
NOi + 1 CH3OH -> | C02 + £ N2 + 1 H20 + OH~ (4-5)
According to this relationship, 1.9 g of methanol (2.86 g
when expressed as COD) are required per g of nitrate-
nitrogen reduced.
In natural water systems, the carbonic acid buffering sys-
tem is the dominant inorganic component of the water.
Consequently, the hydroxide (OH~) produced during de-
nitrification in natural waters will react with carbonic acid
(carbon dioxide) to produce bicarbonate ions (i.e., bicar-
bonate alkalinity). The following equation takes this con-
version into account (10):
CH3OH + - H2C03 -> N2 + - H2O + HCO3
b b 2. 6
(4-6)
In addition to supplying electrons to produce energy, the
organic substrate also provides the carbon for creating
new cell mass in heterotrophic microorganisms; conse-
quently, more electron donor will be required to reduce a
given amount of nitrate than is presented in Equation 4-5.
The distribution of the organic substrate between incor-
poration into new biomass versus that utilized for the
reduction of nitrate will be a function of the organic com-
pounds making up the substrate, the microbial population
present, and the operating conditions of the system.
Thus, this distribution must be determined experimentally.
The experiments of McCarty et al. (10) provide the basis
to formulate stoichiometric equations for methanol con-
sumption incorporating both the half-reactions above and
the observed biomass yields. These equations are as
follows (11):
(4-7)
(4-8)
(4-9)
- + 1 .08 CH3OH + 0.24 H2CO3 ->
0.056 C5H7NO2 + 0.47 N2 + 1 .68 H2O + HCOj
NO! + 0.67 CHgOH + 0.53 H2CO3 ->
0.04 C5H7NO2 + 0.48 N2 + 1 .23 H2O + HCOj
O2 + 0.93 CH3OH + 0.056 NO3->
0.056 C5H7NO2 + 1 .04 H2O + 0.59 H2CO3 +
0.056 HCO3
In converting to mass, Equation 4-7 suggests 2.47 g of
methanol are required to reduce 1 g of nitrate-nitrogen.
Experimental ratios of methanol to nitrate range from 2.5
to 3.0 g methanol/g nitrate nitrogen. Methanol may also
be required to reduce any nitrite or oxygen present. The
following equation has been used to describe the overall
methanol requirement (10):
where:
M = methanol required, mg/L
NOs~-N = nitrate nitrogen removed, mg/L
NO2-N = nitrite nitrogen removed, mg/L
DO = DO removed, mg/L
Departure of methanol requirements from Equation 4-10
is most likely due to variations in sludge yields among
experimental systems. An expression defining the re-
quired COD to nitrogen ratio has been developed (12) for
any organic substrate:
COD
N
2.86
(4-11)
where:
M = 2.47 (NO3-N) + 1.53 (NO2-N) + 0.87 DO
(4-10)
YS[gET = biomass net yield based on COD, g VSS/g
COD removed
This equation was developed assuming that the COD of
the VSS produced is 1.42 g COD/g VSS, and that the
biomass produced is 10 percent nitrogen.
A second method that can be used to obtain a rough
estimate of the organic substrate requirements is to con-
sider the oxygen demand exerted in a given system by
an organic substrate and convert it to nitrate equivalence
based on a factor of 2.86 mg oxygen/mg nitrate nitrogen.
For example, if it is known from experience for a given
organic substrate and treatment system that the oxygen
demand exerted is 200 mg/L, then 70 mg/L of nitrate
nitrogen could theoretically be reduced.
4.2.2.2 Alkalinity andpH Relationships
As mentioned above, bicarbonate alkalinity is produced
and carbonic acid concentrations are reduced during
denitrification. The theoretical stoichiometry of the bicar-
bonate alkalinity production is 3.57 mg alkalinity as
CaCO3 produced per mg of nitrate nitrogen reduced to
nitrogen gas. The preponderance of investigative evi-
dence strongly suggests that the actual denitrification
alkalinity production is essentially equivalent to stoichi-
ometric predictions.
Since the alkalinity concentration is increased and the
carbonic acid concentration is reduced, the tendency of
denitrification is to partially reverse the effects of nitrifica-
tion and raise the pH of the biological reaction. For waters
low in alkalinity, the recovery of alkalinity through denitri-
fication in a single-sludge system can be a significant
benefit. The recovery of alkalinity and the potential reduc-
tion in energy requirements attributable to denitrification
may in some situations make denitrification attractive
even if total nitrogen limits are not required.
103
-------
4.3 Kinetics of Denitrification
This section presents a partial development of equations
that can be used to describe the rate of denitrification.
The discussion is intended to show the relationship be-
tween denitrification and the most prominent rate-limiting
factors. The material serves as the basis of more detailed,
process-specific kinetic equations development in Chap-
ters 5, 7, and 8.
4.3.1 Biomass Growth and Nitrate Utilization
The kinetics of denitrification can be described using
equations that take the same form as those for other
microfaial reactions, such as nitrification or organic matter
removal. For example, first-order, zero-order, and Monod-
type kinetics have all been used to describe the rate of
denitrification (Section 3.3.1). However, the development
of one set of kinetic expressions to cover the general topic
of denitrification design is not possible given the varied
nature of the configurations used to accomplish denitrifi-
cation and the different organic substrates used. The ex-
pression used will often be a function of the type of
reactor utilized, the organic substrate employed, and the
needs of the user.
Monod-type expressions are developed first in this dis-
cussion in order to illustrate the general concepts of de-
nitrification kinetics. These expressions are often directly
applicable to separate denitrification reactors with supple-
mental substrate addition (e.g., methanol). For single-
sludge denitrification systems, the expressions are more
difficult to apply directly because of factors such as the
heterogeneity of the substrate and the biomass. Expres-
sions that may be used for single-sludge denitrification
systems are briefly discussed in this chapter and again
in Chapter 8. The IAWPRC model described in Chap-
ter 5, however, is based on Monod-type expressions.
A Monod-type expression can be used to relate the
growth rate of denitrifying microorganisms to the concen-
tration of nitrate:
(4-12)
where:
Mo = specific denitrifier growth rate, d"1
Mo = maximum specific denitrifier growth rate, d~1
D = concentration of nitrate nitrogen, mg/L
KD s half-saturation coefficient, mg/L
Such an expression will transform into a zero-order ex-
pression (u,0 = u.0) if KD is considerably less than D, and
A
MD
a first-order expression (u. = (^-) x D) if D is significantly
less than KD.
As discussed in Section 3.3.1, the net specific growth rate
of microorganisms in a system is the inverse of the solids
retention time (SRT):
eR
(4-13)
where:
1-1
60 = solids retention time, d
u.' = net specific growth rate, d
Since the growth rate of denitrifiers is generally very simi-
lar to aerobic heterotrophic organisms (i.e., much greater
than nitrifiers), the minimum solids retention time required
to prevent washout of denitrifiers from a reactor will be
much shorter than for nitrifiers.
Nitrate removal rates can be related to organism growth
rate by using the organism yield as a conversion factor:
Mo
Y~
ID
D
n
+ u
(4-14)
where:
qo = nitrate removal rate, g NO^-N/g VSS/d
YD = denitrifier true yield coefficient, g biomass (VSS)
grown/g NO5-N removed
qo = maximum nitrate removal rate, g NOj-N/g VSS/d
The solids retention time can be related to nitrate removal
rates in a similar fashion:
'(4-15)
7T
"c
where:
bd = denitrifier decay coefficient, d"1
The concentration of organic substrate will also affect the
rate of nitrate removal. Thus, the rate of nitrate removal
can also be expressed as a function of organic substrate
concentration:
„ A O ,. ._,
QD = qo 77 g (4-16)
where:
S = concentration of organic substrate, mg/L
Ks = half-saturation constant, mg/L
Since both nitrate and electron donor concentrations may
limit the rate of denitrification, a more accurate expression
for the rate is a double (or multiple) Monod-type expres-
sion (9,11): ;
O,D =
(4-17)
The values of the half-saturation constant for nitrate, KD,
are reported to be very low: from about 0.1 to 0.2 mg
104
-------
NO^-N/L (1,13,14). It can be seen from examination of
Equation 4-17 with these values for KD that nitrate con-
centrations greater than 1 to 2 mg NOa-N/L have almost
no effect on denitrification rates, provided diffusion limi-
tations are not a factor. Consequently, nitrate concentra-
tions will not affect the rate of nitrate removal unless very
low nitrate Affluent concentrations must be achieved or
diffusion limitations exist such as may be encountered in
attached growth processes.
Values for Ks will depend on the organic substrate, but
are also, generally low. Values of Ks for methanol of
0.1 mg/L (11) to 6.0 mg/L (7), and as high as 72 mg/L
(1.5), have been reported. The low Ks value using metha-
nol implies that to achieve 90 percent of the maximum
denitrification rate, only about 1 mg/L of methanol need
be in the effluent; that is, significant excesses of methanol
above stoichiometric requirements need not be in the
effluent to approach the maximum denitrification rate.
This factor can be of significance in some situations since
excess methanol in the effluent may require removal in a
downstream aerobic process. Although very limited data
are available in the literature, it is likely that the values of
Ks with municipal wastewater as the organic substrate
(expressed as BOD) are likely to be greater than those
with methanol. Consequently, nitrate removal rates are
likely to be a function of the organic substrate concentra-
tion in a system using wastewater as the organic sub-
strate, as will be discussed below.
The impact of oxygen on denitrification can also be added
to denitrification rate expressions as follows (16):
(4-18)
where:
K0 = half-saturation constant for oxygen, mg/L
S0 =. DO concentration, mg/L
The term Ko/(K0+S0) can act as a switching function,
turning denitrification on and off. A value for K0 of
0.1 mg/L has been suggested for denitrification in the
IAWPRC model (see Chapter 5).
The kinetics of denitrification can also be expressed in
terms of the rate of organic substrate removal, since the
nitrate removal is tied directly to the organic removal. In
many instances use of the substrate removal rate may
be preferred, since more information is available on ki-
netic coefficients for substrate removal. The equations will
be identical in form to those above, but the rates of nitrate
removal will be replaced by the rates of substrate re-
moval. For example, the rate of substrate removal can be
expressed as:
where:
qs = substrate removal rate, g COD or BODs/g VSS/d
qs = maximum substrate removal rate, g COD or
BODs/g VSS/d
The value of Ks will remain the same as that used for the
nitrate removal rate (see Equation 4-16).
The rate of nitrate removal can be related to the rate of
substrate removal with the same type of expression com-
monly used to relate oxygen consumption to organic sub-
strate utilization (12):
= (1-1.42YB)
2.86 2.86
(4-20)
where:
Ys = biomass true yield, g biomass (VSS) grown/g
COD removed
In this equation, the first term is used to describe the
fraction of the substrate (expressed as COD) and nitrate
that are used for biomass synthesis, and the second term
describes the fraction of nitrate that is used for endo-
genous respiration. The value 1.42 is an assumed value
for the COD of the VSS produced.
Table 4-1 summarizes some of the kinetic coefficients that
have been determined for suspended growth systems
with methanol as the carbon source. In most cases only
net (or observed) yields are reported or can be calculated
from the available data. The classical relationship (Table
2-13) between the true yield (due to biomass growth) and
the net yield is:
+e0bd)
(4-21)
where:
qs = qs
KS
(4-19)
YDNET = denitrifier net (observed) yield, g VSS/g NO3-N
removed
A value for bd of 0.04 d"1 was used in some cases to
derive calculated YD values for those cases where none
was reported. There is significant variation in the data
represented in Table 4-1, due to differences in test con-
ditions and procedures. Care should be taken when using
these data without consulting the specific reference.
Further, care must be taken when using these data or
any others reported in the literature to determine the basis
for the specific rates. Data may be reported on the basis
of active denitrifying biomass or total volatile solids.
Those listed in Table 4-1, for instance, are for the most
part based on total volatile solids; however, the rate ex-
pressions provided generally in this chapter are based on
active denitrifying biomass. In actual systems the VSS
will be made up of active autotrophic bacteria (nitrifiers),
nondenitrifying active heterotrophic bacteria, denitrifying
105
-------
Table 4-1. Values for Denitrification Yield and Decay Coefficients for Various Investigations Using Methanol
(from Reference 18)
q0. d"1
Variable
Variable
Variable
0.12 to 0.32
0.16 to 0.9
Variable
— *
0.131 to 0.347
0.25
d
0.30
0.12 to 0.24
0.1 to 0.45*
0.3 to 0.56
Y
>D
fr
Qs
Ys
Temp.,°C qs Ys K..
10
20
30
20
20
20
10 to 20
20
16 to 18 :
19 to 20
20
10 3.1 0.17 12.6f
20 10.3 0.18 9.1
16 to 23
16 to 18
= Nitrate removal rate
= Net yield based on nitrate = g VSS/g NOj-N removed
= True yield based on nitrate = g VSS/g NOj-N removed
= Endogenous decay coefficient
= Maximum specific COD removal rate = g COD/g VSS/d
= True yield based on COD = g VSS/g COD
= COD half-saturation constant (based on total COD) = mg/L COD
active heterotrophic bacteria, inert organics in the influent,
and inert products of biomass decay. The IAWPRC model
described in Chapter 5 takes these different fractions into
account. For more simplified design approaches, total
VSS may be used provided the data were obtained from
studies on a system with similar wastewater charac-
teristics, so that the various fractions that make up the
VSS are the same. Alternatively, rates specific to active
biomass may be used if the fraction of the total active
VSS can be estimated. Table 2-13 provides the IAWPRC
equation that can be used to make this estimation.
While a number of methods can be used for sizing a
denitrifying reactor with supplemental substrate addition,
the method chosen will be a function of readily available
kinetic data and the type of reactor used. For a completely
mixed, suspended growth reactor, the rate of nitrate or
substrate utilization (i.e., that which can be estimated
from Monod-type kinetics) can be used in conjunction with
a mass balance equation to calculate a design reactor
volume. An alternative approach (but one that is based
on the same fundamental equations) is to calculate a
design solids retention time from the rate of nitrate or
substrate utilization (Equation 4-15). Similar procedures
can be used from that point as would be used for a
nitrifying activated sludge system (See Chapters 3 and
6). For a suspended growth system, the solids retention
time should be checked to verify that it is greater than
the minimum (1.0 to 2.5 d) required to develop a floccu-
lent biomass.
The kinetic equations developed above may be readily
applied to systems where a single, readily biodegradable
electron donor is used (e.g., methanol). The application
of the equations is somewhat more complicated in sys-
tems that have a complex electron donor source (e.g.,
municipal and many industrial wastewaters). Often only
part of the organic matter in these wastewaters is readily
biodegradable with the remaining fractions being slowly
biodegradable, or nonbiodegradable. The slowly biode-
gradable fraction may be particulate or soluble.
Also, the kinetic expressions above may be applied to
complex electron donor sources by assuming that the
growth of denitrifiers (and heterotrophs in general) only
occurs at the expense of the readily biodegradable frac-
tion and that the slowly biodegradable fraction must be
converted ("hydrolyzed") to a readily biodegradable form
106
-------
prior to its utilization (16). Separate kinetic expressions
can be used to express the rate at which the slowly
biodegradable organics are converted to a readily biode-
gradable form. In addition, growth of denitrifiers will also
occur at the expense of biodegradable organic matter
produced in the reactor from the decay of other cells—a
phenomenon termed endogenous respiration. As pre-
viously indicated, kinetic equations may also be used to
express the rate at which microorganisms decay. All of
these processes are incorporated into the single-sludge
activated sludge model prepared by IAWPRC (16), as
discussed in Chapter 5.
Simplified approaches based on empirical observation
can be used for designing the denitrification process in
single-sludge systems with complex substrates. It has
been observed that the process of nitrate removal exhibits
distinct rates that are a function of the availability of the
substrate (i.e., the location of the anoxic zone in the
process) (7,21). Ekama (21) noted two phases of denitri-
fication in a plug flow, first anoxic zone and a third phase
in a second anoxic zone. The first phase was fairly short
(up to 10 minutes) and may have been due to the con-
sumption of the readily degradable organic fraction.
Burdick et al. (5) observed only one rate in the first anoxic
zone of Bardenpho facilities. The rate was correlated to
the food/biomass (F/M) loading on the anoxic zone:
developed to describe the rate in the second anoxic zone
are provided below (5,12):
= 0.03 - + 0.029
(4-22)
where:
SDNR} = specific denitrification rate in the first anoxic
zone, g NO^-N/g VSS/d
F/M! = F/M loading ratio on the first anoxic zone, g
BOD/g mixed liquor VSS in the first anoxic
zone/d
Specific rates of denitrification in the first anoxic zone
typically range from 0.04 to 0.15 g NOg-N/g VSS/d (22).
This range and Equation 4-22 provide only rough esti-
mates of the denitrification rate that will be observed,
since the rate will be influenced by a number of factors
including the mixed liquor recirculation rate, the influent
wastewater strength, and the denitrifying active, fraction
of the volatile solids. The expressions used in the
IAWPRC model (see Chapter 5) are designed to take
some of these factors into account to give a more accu-
rate estimate of the actual denitrification rate.
Denitrification in second anoxic zones generally is driven
by the endogenous decay of the biomass, so that the
rates will be much lower than in the first anoxic zone.
Specific rates of denitrification in the second anoxic zone
are typically 20-50 percent of the rate in the first zone
(11,22). The rate of endogenous decay will be a function
of the system and the SRT. Two equations that have been
SDNR2 = 0.12x0c0-706
0.175 An
SDNR2 = -
Y
(4-23)
(4-24)
where:
SDNR2 = specific denitrification rate in the second
A anoxic zone, g NOa-N/g VSS/d
Y = net TSS production across the entire
activated sludge system, including inert
solids and biomass, g TSS/g BOD5 removed
An = net amount of oxygen required across the
entire activated sludge system/TBOD
removed, g O2/g BOD
4.3.2 Temperature Effects
As with any microbial activity, nitrate removal rates can
be affected significantly by temperature. Data from the
literature are summarized in Figure 4-1, where they have
been normalized with respect to rate at 20°C (68°F).
These data suggest that temperature exerts a larger ef-
fect below about 20°C (68°F) than above.
The impact of temperature on biological systems is often
described by a Arrenhius-type function:
(4-25)
where:
qD,T = denitrification rate at temperature T(°C), mg
NO5-N/g VSS/d
qo.ao = denitrification rate at 20°C, g NOs-N/g VSS/d
9 = simplified Arrenhius temperature-dependent
constant
Although such a function is useful for modeling denitrifi-
cation, it is limited to a certain temperature range, and
the value of theta (6) is site specific. Values of theta
reported in the literature are summarized in Table 4-2.
Table 4-2. Temperature Correction Coefficients for Model-
ing Denitrification (Endogenous Rate)
6 Reference
1.08
1.09
1.20
1.08
1.03
1.08
23
24
25
25
25
26
107
-------
180
160
140
120
O
<
1
cc
8
O
to
DC
o
I
o>
Q.
110
80
60
40
20
Key
d-1
at20°C
Electron
Donor
r
System
Methanol
Methanol
Methanol
xxxxxx -
1.07 -
0.33 -
-„....... 1.7 1.04 Methanol
• —— • — - - Methanol
— • • — - - Methanol
nnonnn- - Methanol
Sodium Citrate Suspended Growth i
Endogenous Suspended Growth /
Suspended Growth i
Suspended Growth /
Suspended Growth i
Suspended Growth / // /
Suspended Growth i // /
Attached Growth / J/a/a n
Attached Growth - '' /
I
I
10 15
Temperature, °C
20
25
30
Figure 4-1. Effect of temperature on denitrification rates (from Reference 11).
108
-------
4.3.3 pH and Alkalinity Effects
The response of denitrification and aerobic respiration
rates to pH variations should be similar. In general, deni-
trification will be much less sensitive to pH than nitrifica-
tion. Representative observations of the effect of pH on
denitrification rates are shown in Figure 4-2. These data
suggest that denitrification rates are depressed below
pH 6.0 and above pH 8.0. Since denitrification will pro-
duce alkalinity, it,may increase the pH if high concentra-
tions of nitrate are to be removed.
4.3.4 Effects of Inhibitors
Denitrifiers are much less sensitive to inhibitory com-
pounds than are nitrifiers. In general, inhibitors would be
expected to have a similar degree of impact on denitrifi-
cation and heterotrophic aerobic respiration. Conse-
quently, commonly applied concentrations that result in
inhibition (e.g., those published by EPA for activated
sludge and trickling filters [27]) can be used for denitrifi-
cation. The ability of a biomass to acclimate to higher
levels of inhibitory compounds should be taken into ac-
count when reviewing these values. Much higher concen-
trations may be tolerated by acclimated cultures. Specific
literature should be reviewed or pilot tests conducted to
determine actual inhibitory levels.
4.3.5 Effect of Diffusional Limitations
Diffusional limitations will affect the design of fixed film
reactors for denitrification as they will for fixed film reac-
tors for nitrification (see Section 3.4). In general, the
equations presented above are applicable to fixed film
systems only if they are coupled with equations that de-
scribe diffusional limitations. Models have been devel-
oped that take such limitations into account (see
Chapter 5). However, the design of many attached growth
reactors is quite often based on empirical results from
pilot and full-scale systems. Rates of denitrification in
these empirical approaches are often based on media
surface area or media volume.
For reactors that provide very turbulent conditions, such
as fluidized beds, the rate of mass transport may be so
high that diffusion may not limit the rate of reaction. The
design of such reactors may be based on the same equa-
tions as those used for suspended growth reactors, pro-
vided the biomass in the reactor can be estimated.
100
g
I 60
i
D
o
€
ffl
°- 20
i i i i i i i i
6.0
7.0
8,0
9.0
PH
Figure 4-2. Effect of pH on denitrification rates (from Reference 11).
109
-------
4.4 References
When an NTIS number is cited in a reference, that docu-
ment is available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
703-487-4650
1, Christensen, M.H., and P. Harremoes. 1977. Biologi-
cal denitrification of sewage: a literature review.
Water Tech. 8:509-555.
2. Tiedje, J.M., A.L. Sexstone, D.D. Myrold, and J.A.
Robinson. 1982. Denitrification: ecological niches,
competition and survival. Antonie van Leeuwenhoek
48:569-583.
3. Simpkin, T.J., and W.C. Boyle. 1988. The lack of
repression by oxygen of the denitrifying enzymes in
activated sludge. Water Res. 22:201-206.
4. Focht, D.D., and A.C. Chang. 1975. Nitrification and
denitrification processes related wastewater treat-
ment. Adv. Appl. Microbiol. 19:153-186.
5. Burdick, C.R., D.R. Refling, and H.D. Stensel. 1982.
Advanced biological treatment to achieve nutrient re-
moval. JWPCF 54:1078-1086.
6. Robertson, LA., and J.G. Kuenen. 1984. Aerobic de-
nitrification: a controversy revived. Arch. Microbiol.
139:351-354.
7. Kugelman, I.J., M. Spector, A. Harvilla, and D.
Parees. 1991. Aerobic denitrification in activated
sludge. Proceedings of the 1991 Env. Eng. Specialty
Conference, Reno, NV (July).
4
8. McCarty, P.L. 1975. Stoichiometry of biological reac-
tions. Prog. Water Tech. 7:157-172.
9. Grady, C.P.L, and H.C. Lim. 1980. Biological waste-
water treatment. New York, NY: Marcel Dekker.
10. McCarty, P.L, L Beck, and P. St. Amant. 1969. Bio-
logical denitrification of wastewaters by addition of
organic materials. Proceedings of the 24th Purdue
Ind. Waste Conf., Purdue University, Lafayette, IN.
11. U.S. EPA. 1975. Process design manual for nitrogen
control. EPA/625/1-77/007 (NTIS PB-259149).
Washington, DC.
12. Stensel, H.D. 1981. Biological nitrogen removal sys-
tem design. Water. Amer. Inst. of Chem. Engrs., p.
237.
13. Engber, D.J., and E.D. Schroeder. 1974. Kinetics and
Stoichiometry of bacterial denitrification as a function
of cell residence time. University of California at
Davis, unpublished paper.
14. Moore, S.F., and E.D. Schroeder. 1971. The effect of
nitrate feed rate on denitrification. Water Res. 5:445-
452.
15. Stensel, H.D., R.C. Loehr, and A.W. Lawrence. 1973.
Biological kinetics of suspended-growth denitrifica-
tion. JWPCF 45:249-261.
16. Grady, C.P.L., W. Gujer, M. Henze, G.V.R. Marais,
and T. Matsuo. 1986. A model for single-sludge
wastewater treatment system. Wat. Sci. Tech.
18:47-61.
17. Mulbarger, M.C. 1971. The three sludge systems for
nitrogen and phosphorus removal. Presented at the
44th Annual Conference of the WPCF (October).
18. Park, A.K., F.J. Zadick, and K.E. Train. 1973. Sludge
processing for combined physical-chemical-biologi-
cal sludges. EPA/R2-73-250 (NTIS PB-223341).
Washington, DC.
19. Sutton, P.M., K. Murphy, and R.N. Dawson 1975.
Low temperature biological denitrification of waste-
water. JWPCF 47:122-134.
20. Parker, D.S., R.C. Aberley, and D.H. Caldwell. 1977.
Development and implementation of biological deni-
trification for two large plants. Prog. Water Tech.
8:673-686.
21. Ekama, G.A., and G.V.R. Marais. 1984. Biological
nitrogen removal. In: Theory, design, and operation
of nutrient removal activated sludge processes.
Water Research Commission.
22. The Soap and Detergent Association. 1989. Princi-
ples and practices of phosphorus and nitrogen re-
moval from municipal wastewater. New York, NY
(September).
23. Lewandowski, Z. 1982. Temperature dependency of
biological denitrification with organic material addi-
tion. Water Res. 16:19.
24. Barnard, J.L. 1974. Cut P and N Without Chemicals.
Water Was. Engr. 11:33.
25. van Haandel, A.C., and G.V.R. Marais. 1981. Nitrifi-
cation and denitrification kinetics in the activated
sludge process. Research Report no. W39. Univer-
sity of Cape Town, South Africa.
26. Marten, W.L. 1984. A study on the use of denitrifica-
tion to meet oxygen requirements at the MMSD Nine
Springs Wastewater Treatment Plant. Unpublished
report. University of Wisconsin-Madison.
27. U.S. EPA. 1987. Guidance manual on the develop-
ment of local discharge limitations under the pre-
treatment program. EPA Contract 68-01-7043. Wash-
ington, DC.
110
-------
Chapters
Mathematical Modeling of Nitrification and Denitrification
5.11ntroduction
Mathematical modeling is a technique that is increasingly
being used to analyze problems of significance to envi-
ronmental engineering. This analytical approach can be
used as an adjunct to more traditional tools (e.g., those
described in Chapters 6, 7, and 8) to refine and optimize
proposed facility designs. It can also be used as an op-
erational tool to optimize the performance and/or capacity
of an existing facility. Moreover, appropriate mathematical
models can be used to analyze a wider range of options,
for both facility design and operation, than would be pos-
sible by other means, resulting in more informed deci-
sion-making. Analysis of facility design options can lead
to better decisions concerning facility size and operational
flexibility needs. For example, mathematical modeling
can be used to optimize the oxygen transfer system con-
figuration for a single-sludge nitrification/denitrification
system. Similarly, modeling can be used to evaluate a
variety of operational strategies for an existing system,
allowing the selection of the most attractive strategies for
full-scale evaluation.
This chapter discusses available approaches for model-
ing nitrification/denitrification systems. The discussion is
organized so that modeling approaches and the different
types of models are discussed first. Next, requirements
to develop a mathematical model for a particular applica-
tion are discussed, followed by a review of uses of mathe-
matical models. Finally, mathematical models currently
available for use in the design and evaluation of nitrifica-
tion and denitrification systems are presented, and an
example illustrates the use of a mathematical model to
optimize the design of a single-sludge nitrification/denitri-
fication system.
The discussion in this chapter independently addresses
the topic of mathematical modeling. Although a firm un-
derstanding of nitrogen removal technology is necessary
to utilize fully the material presented here, thorough re-
view of the other chapters of this manual is not necessary
for the knowledgeable practitioner to benefit from the dis-
cussion. Users of this manual may elect to proceed di-
rectly to Chapters 6, 7, and 8 if system design is of more
immediate interest.
5.2 Modeling Approaches
A wide variety of models is used to analyze problems in
environmental engineering. For the purposes of this dis-
cussion, however, the models are divided into two cate-
gories: conceptual models and mathematical models.
5.2.1 Conceptual Models
Conceptual models form the basis for most of the design
and operational decisions made in environmental engi-
neering. A conceptual model represents an understanding
of the cause-and-effect relationships at work in a given
system. Figure 5-1, for example, illustrates a conceptual
°2
°2
NH+ N Nitrosomonas NOa'-N Nitrobacter
/ *
HCO,
CO 2
_^ NOg'-N
H2O
HCOs
Figure 5-1. Conceptual nitrification model.
model for the nitrification process. Ammonium nitrogen
(NHJ-N) is first oxidized to nitrite nitrogen (NCvN) through
the action of autotrophic bacteria, most often of the genus
Nitrosomonas. Molecular oxygen (O2) and bicarbonate
alkalinity are consumed as a result of this reaction; reac-
tion by-products include water (H2O) and additional bac-
teria. NO^-N generated in the first step is subsequently
oxidized to nitrate nitrogen (NOi-N) through the action of
a second type of autotrophic bacteria of the genus Nitro-
bacter, co-reactants and by-products for this reaction are
similar to those for the first reaction.
Although not quantitative in nature, this model can be
used to understand many of the factors that affect the
nitrification process in wastewater treatment systems. For
example, it is clear from Figure 5-1 that nitrification will
not occur unless sufficient quantities of the two types of
bacteria—Nitrosomonas and Nitrobacter—are present.
Concentrations will be sufficient only if the environmental
conditions allow these bacteria to proliferate. Factors
111
-------
such as temperature, pH, DO, and specific growth rate
affect the concentration of the nitrifying bacteria and,
based on empirical observations and correlations, condi-
tions that encourage growth of sufficient numbers of these
bacteria in various wastewater treatment systems can be
identified. Using this information, it is possible to design
and operate wastewater treatment systems to accomplish
nitrification (the conversion of NHJ-N to NOj-N).
Based on this example, the characteristics of a concep-
tual model can be identified:
• System components. Conceptual models identify the
components of a system. In Figure 5-1, for example,
the system components include NHj-N, Oa, bicarbon-
ate alkalinity, Nitrosomonas, NOa-N, HaO, Nitrobacter,
and NO5-N. Conceptual models, however, present
one representation of how a system might actually
function.
• Cause-and-effect relationships. Conceptual models
describe the cause-and-effect relationships between
system components. For example, consider the inter-
action of the two bacterial species required to accom-
plish nitrification. Figure 5-1 indicates that NHj-N is
first oxidized to NOa-N by the action of Nitrosomonas.
Subsequently, the generated NOa-N is oxidized to
NOa-N by the action of Nitrobacter. This relationship
indicates that both species must be present in the ap-
propriate amounts to obtain complete nitrification.
Other relationships, such as the effect of aeration on
the loss of COa and the subsequent increase in reac-
tor pH, can also be similarly described. To the extent
possible, the proposed interrelationships typically rep-
resent actual physical, chemical, and/or biological
phenomena.
• Interactions can be described qualitatively and/or
quantitatively. The relationships illustrated in Figure
5-1 are qualitative. However, because the stoichi-
ometry of the reactions presented is known, they
could also be presented in quantitative form. Qualita-
tive information, such as an increase in X leads to an
increase in Y under certain conditions but to a de-
crease in Y under alternate conditions, can also be
presented. Interrelationships may also be presented
as generalized correlations (e.g., effluent NHj-N con-
centrations are correlated with suspended growth
system solids residence time as a function of tempera-
ture). In this latter instance, physical data are pre-
sented graphically.
Conceptual models, as described above, can and have
been used routinely to design nitrogen removal systems.
Designs produced based on the use of these models can
be highly efficient and cost effective. However, as dis-
cussed below, design approaches based on mathemati-
cal modeling procedures can enhance designs developed
using more conventional analytical tools.
5.2.2 Mathematical Models
Mathematical models have many features in common
with conceptual models. The primary difference is that in
a mathematical model the interrelationships between sys-
tem components must be described quantitatively, while
in a conceptual model they may be described either quali-
tatively or quantitatively. Consequently, mathematical
models are in one sense a subset of conceptual models.
However, the restriction that mathematical models can
use only quantitative relationships can lead to significant
practical differences between conceptual and mathemati-
cal models. :
Because of the complexity of the physical, chemical, and
biological systems being modeled, simplifications are fre-
quently introduced into mathematical models to reduce
the mathematical complexity. These simplifications take
several forms. In some instances, the impacts of several
interactions are included in a single factor. For example,
the single half-saturation constant in the Monod-type sub-
strate concentration-reaction rate relationship (the type
often incorporated into many mathematical models) is
often used to represent the impact of more than one
variable. In other instances, approximations are used as
mathematical conveniences to reduce the computations
required to produce a mathematical solution to a problem.
Thus, mathematical models often represent simplifica-
tions of the current understanding of the physical problem
being considered. While some simplifications are intro-
duced to reduce the computational complexity of a pro-
posed problem, others are introduced because of the
difficulty in measuring the stoichiometric and kinetic co-
efficients required by the model.
Mathematical models can be used as a tool to test alter-
native hypotheses concerning the interrelationships be-
tween system components. Alternative models can be
developed based on contrasting hypotheses, and system
responses can then be predicted under defined condi-
tions. Through comparison of detailed model predictions
with the results of actual measurements for bench, pilot,
and/or full-scale systems, the utility of the proposed mod-
els can be analyzed and the most useful one selected.
Although close correspondence between model predic-
tions and 'actual responses does not prove the subject
hypothesis, it does demonstrate that the hypothesis is
useful. This approach can be used as a tool to test alter-
native hypotheses and to select the one that best repre-
sents a particular application. Used in this fashion,
mathematical modeling can lead to enhanced insight con-
cerning both the interactions between process variables
affecting nitrogen removal systems and the relative im-
portance of various process variables in determining
overall process performance. x -,
Although increased computer power, provided by the new
generation of computers, is making the use of complex
simulation models possible, it is important to recognize
112
-------
the value of simplicity. As a general rule,- it is best to use
the simplest model that provides an adequate description
of the system to be analyzed. Simplicity often leads to
clarity in thinking and evaluation, and to the avoidance of
errors caused by a lack of understanding of the funda-
mental features that determine system response. Thus,
while the computer provides the ability to simultaneously
consider many factors, this capability must be balanced
against the value of gaining a basic understanding of a
system. As is typically the case, such issues can be re-
solved best using the direct experience of the practitioner.
Many types of mathematical models can be applied to
engineering applications. For instance, statistical models
are used to analyze existing data sets as a way to identify
and quantify interrelationships between process vari-
ables. Semi-empirical models are used to correlate proc-
ess variables and facilitate engineering calculations. In
contrast, mechanistic models are based on a specific
interpretation of a phenomenon occurring within a system
being modeled. In the following discussion, emphasis is
placed on mechanistic models because they are based
on phenomena that determine system response, while
also offering the best potential for providing a realistic
and unbiased representation of the system being consid-
ered. They also offer the best opportunity to accurately
predict system response over a broad range of operating
conditions.
5.3 Model Development
Since mechanistic mathematical models are developed
to achieve specific objectives, different models will be
used to achieve different purposes. Model development
proceeds most efficiently if a logical approach is used.
One approach consists of three steps: model selection,
model calibration, and. model verification. These are illus-
trated in Figure 5-2 and are discussed in this section.
5.3.1 Model Selection
The first step in model development is model selection.
In this step, the specific form of the model to be used for
a particular application is selected. Factors to be consid-
ered include the existing and possible level of knowledge
concerning the application (i.e., the data available), the
decisions to be made based on the modeling results, and
the available understanding of system mechanics.
The available data affect model selection since they de-
termine the detail possible for the modeling exercise. For
example, if it is not possible to obtain a-detailed charac-
terization of the subject wastewater, it makes little sense
to use a sophisticated model capable of providing detailed
predictions about the impacts of wastewater charac-
teristics on system response. Predictions based on as-
sumed characteristics would be of little value since they
will not, in all likelihood, reflect actual operating condi-
Model Selection
Unacceptable
Fit
Model Calibration
Improper
Model
Acceptable
Fit
Model Verification
Improper
Calibration
Model
Verified
Model Use
Figure 5-2. Stepwise approach to model development.
tions. In such a case, use of a generalized model would
be more appropriate.
The use to which the model is to be put will also deter-
mine model'Structure. A relatively simple model can be
used to size an aeration basin that would allow a com-
plete mix activated sludge system to accomplish single-
stage nitrification. In contrast, a more sophisticated model
is required to size a single-sludge nitrification/denitrifica-
tion system, particularly one using staged reactors for the
anoxic and aerobic zones. Even more sophistication is
required if spatial and temporal variations in process oxy-
gen requirements are to be estimated.
A final consideration in model selection is the level of
current knowledge concerning the underlying mechanics
of the system being simulated. Mathematical models are
one means to represent the current understanding of a
specified system. As such, the predictive capabilities are
no better than the fundamental knowledge that is avail-
able on that system. In other words, mathematical mod-
eling builds on existing knowledge and; although it can
be used to test alternative hypotheses, it cannot be used
by itself to expand existing knowledge concerning a par-
ticular system.
Many different models could be applied to a particular
engineering problem. The knowledge and judgement of
the process modeler, coupled with experience with alter-
113
-------
native process models, form the basis for the initial se-
lection of the model type for a particular application.
5.3.2 Model Calibration
Once a model has been selected, it must be calibrated.
The objective of calibration is to select appropriate values
for the required kinetic and stoichiometric coefficients.
Selection can be based on several factors. In some
cases, the coefficients are fundamental in nature and can
be derived from theoretical considerations. In others, sig-
nificant experience has been acquired with use of the
particular model in similar applications, and that experi-
ence can be used as the basis for the selection of model
coefficients. In still other cases, the model is calibrated
by adjusting model coefficients to match the results of
bench, pilot, and/or full-scale facilities that simulate the
proposed treatment system. The data used for this ap-
proach can vary widely, from a few simple bench scale
measurements to the results from sophisticated pilot
plants, prototypes, and/or full-scale facilities. When ex-
perimental results are used to calibrate a proposed
model, however, the experiments must have been de-
signed to produce a range of responses that is sufficiently
broad to allow for accurate determination of model coef-
ficients. Mathematical models are most accurate when
used to interpolate within the existing data base; extrapo-
lations must be approached with caution. In short, a ra-
tional procedure must be used to select the numerical
values of model coefficients if realistic results are to be
obtained.
Model calibration also involves evaluation of the ade-
quacy of model structure. If the model is unable to prop-
erly predict trends in system response and/or if model
coefficients cannot be adjusted sufficiently to obtain an
acceptable calibration, it may be necessary to adjust the
structure of the model. Again, model responses must be
tested over a sufficiently broad range of conditions to
ensure its suitability for the intended application. Sensi-
tivity analysis is one tool that can be used to assess
model structure. In such an analysis, model imputs are
varied and the resulting impact on model results is deter-
mined. Based on the results, the significance of various
model components can be assessed. If the model is
found to be lacking, model selection must be reconsid-
ered. This so-called feedback loop is considered in Figure
5-2. Once a new model is selected, it must again be
calibrated.
5.3.3 Model Verification
The final step in model development is model verification.
In this step, predictions by the calibrafed model are com-
pared to measured results from a prototype system to
identify similarities and differences between predicted and
measured results. Verification can be considered valid
only if the data set used for verification is different from
the data set used for calibration. (Verification using an
independent set of data is necessary because a properly
calibrated model will always show acceptable results
when compared to the data used for calibration. If accept-
able results are not obtained using calibration data, then
the model is inappropriate or it has not been properly
calibrated.) The model is considered to be validated when
model predictions agree with measured values from an
independent data set within acceptable tolerances. Dis-
crepancies between predicted and measured values in-
dicate improper model selection and/or calibration,
dictating that one or both of these steps must be re-
peated. Once a revised and/or recalibrated model is iden-
tified, it should be verified with yet another set of
independent data. In some instances, additional inde-
pendent data will not be available and the process
modeler must judge whether a model appropriate for its
intended use has been produced.
As with model calibration, the data set used for model
verification must be sufficiently broad to test the calibrated
model over the entire range of the potential application.
5.4 Model Applications
As discussed above, the level of complexity appropriate
for a particular model is dependent on the intended use
for the model. This section discusses potential uses for
nitrification and denitrification process models.
5.4.1 New Facilities
One use of a mathematical model is to design new nitro-
gen control facilities. In such uses, a great deal of judge-
ment must be applied in the selection of an appropriate
model. If there is a lack of information on the charac-
teristics of the wastewater to be treated, then a relatively
simple modeling approach may be most appropriate. On
the other hand, if significant data are available based on
experience with the particular system and/or the type of
wastewater to be treated, then a more sophisticated mod-
eling approach may be utilized. In many cases, model
verification will be difficult to carry out with new facilities
because of a lack of an appropriate data base. A signifi-
cant pilot plant program may need to be conducted if a
sophisticated model is necessary to optimize the design
of a new facility.
Extreme care must be exercised in designing a pilot plant
program that is intended to produce the data necessary
to calibrate a sophisticated process model. Factors of
particular concern include:
• Pilot plant configuration. The configuration of the pilot
plant should fully simulate the type of facility to be
evaluated. At the least, the hydraulic flow pattern,
oxygen transfer characteristics, and other perform-
ance characteristics should be similar to the proposed
facility.
• Pilot plant operating conditions. Operating conditions
for the pilot plant should simulate anticipated full-scale
114
-------
conditions. Factors such as diurnal variations in proc-
ess loadings, seasonal variations in wastewater char-
acteristics (including temperature), and the impacts of
in-process recycle streams (especially those from sol-
ids handling systems) should be considered.
• Experimental design for model calibration. A broad
range of operating conditions should be explored in
the pilot plant program to develop a sufficiently broad
data base to allow accurate estimation of model pa-
rameters. Mathematical models are best used to inter-
polate within a data set; extrapolation beyond the
collected data base should be used cautiously, if at all.
Sensitivity analysis and statistical design are important
concepts in regard jto the experimental design of the pilot
plant program. A sensitivity analysis is accomplished by
varying the numerical value of .the parameters in the proc-
ess model and determining the resulting impact on the
predicted response. A parameter with high sensitivity is
one for which a small variation results in a large variation
in the response predicted by the model. A parameter with
low sensitivity is one that may be varied over a relatively
wide range while producing only a relatively small vari-
ation in the predicted response. Although accurate esti-
mates are required for the sensitive parameters if the
model is to provide accurate predictions, less accuracy
is needed in the estimates for the relatively insensitive
parameters. Consequently, the pilot plant program should
be designed specifically to provide accurate estimates of
the sensitive parameters.
Experimental design is the selection of target operating
conditions for the pilot plant, and variables to be meas-
ured that will allow accurate estimation of the high-sen-
sitivity parameters. The process model itself can be used
as a tool to design the experimental program. The model
can be run over a wide range of potential operating con-
ditions and parameter values, and the results can be used
to select target operating conditions that will provide the
most accurate numerical estimate of the model parame-
ters. It can also be used to identify system responses that
are the most useful in model calibration. For example,
process oxygen requirement values are often more sen-
sitive to suspended growth model parameters than is
process effluent quality. Consequently, calibration of the
model based on measured process oxygen requirements
can lead to more accurate calibration of the model than
measurement of process effluent quality. The process
model can be used to facilitate development of the most
efficient experimental program. It can also be used to
identify variables that are the most useful in accurately
determining the numerical values of model parameters.
Significantly, estimates of model parameters based on
pilot plant data are "coupled," not absolute. While error
in the estimation of one parameter is often compensated
for by a corresponding error in the estimated value of
another parameter, the magnitude of such errors is un-
known. As a result, measured values of model parame-
ters must be considered relative to each other, not as
independent values. Thus, the absolute value of a model
parameter is seldom known and the relative value, in
combination with the relative value of all the other model
parameters, must suffice to produce model predictions.
This consideration is particularly important when esti-
mates of model parameters obtained from one pilot plant
study are compared to or used as inputs for other studies.
5.4.2 Existing Facility Upgrading
Upgrading of existing wastewater treatment facilities often
represents better opportunities to apply sophisticated
mechanistic mathematical models to facility design. In
such cases, information on wastewater characteristics is
available to facilitate model calibration. However, if the
proposed upgrade represents a significant increase in the
level of treatment, the performance characteristics of the
existing system may not closely match those of the pro-
posed system. While the proposed model may be cali-
brated to the existing system, the actual operating
conditions for the existing system may vary considerably
from those associated with the proposed design. Conse-
quently, the opportunities for model verification using data
from the existing system may be limited. (The discussion
of experimental design for model calibration, presented
in Section 5.3.2, is relevant in such cases.) One advan-
tage in modeling an existing facility for an upgrade, in
comparison to modeling a new facility, is that the existing
wastewater and the biomass may be used for some lim-
ited bench scale verification runs. Batch tests can be
designed using the subject model, as discussed above.
If warranted, the availability of the existing wastewater
can allow operation of a pilot plant utilizing the proposed
process.
5.4.3 Existing Facility Optimization or Expansion .
Optimization or expansion of an-existing facility repre-
sents the best opportunity for use of mathematical mod-
eling approaches. The proposed model can be calibrated
and the results verified using data from the existing full-
scale system. The highest degree of calibration and veri-
fication is obtained in these instances since the existing
full-scale operation is the system to be studied. Modeling
can be used as a tool to identify opportunities to optimize
the existing system; both operational and physical modi-
fications can be evaluated. The most desirable modifica-
tions, based on the modeling results, can then be
evaluated in the full-scale system. If results comparable
to those predicted by the model are obtained, the model
can be further used for facility optimization and/or to iden-
tify facility expansion opportunities. If results do not com-
pare favorably, the model can be modified accordingly
and the process repeated. In any event, the capabilities
and limitations of the model can be carefully defined.
A calibrated mathematical model of an existing treatment
system can also be used on an ongoing basis as a tool
115
-------
to optimize facility operations. Alternative operational
strategies can be evaluated using the mathematical
model, and the most promising approaches can then be
implemented on a full-scale basis. The successive inter-
action between use of the model to predict system re-
sponses and monitoring of actual system responses can
lead to development of an invaluable tool for maximizing
the performance and/or capacity of an existing system.
5.5 Available Models
A wide variety of models are available to characterize
full-scale nitrification and denitrification systems. Since
models for suspended growth systems are the most fully
developed and tested, they are discussed in great detail
In this section. Also, however, ongoing research directed
at the development of improved models for fixed film
systems is summarized. It is hoped that efforts in this
area will produce useful models that can be applied to
the design and operation of fixed film nitrogen control
systems.
5.5.1 Suspended Growth Models
While a wide variety of models has been developed to
characterize suspended growth nitrification and denitrifi-
cation systems, the most complete model was developed
by a task group sponsored by the International Associa-
tion on Water Pollution Research and Control (IAWPRC)
(1,2). This model, referred to as Activated Sludge Model
No. 1, was developed spepifically for use in evaluating
long sludge age (i.e., nitrifying) activated sludge systems,
including those incorporating single-sludge denitrification
features. It incorporates features found useful in previous
models; consequently, detailed description of this model
allows coverage of the relevant features of other, less
sophisticated models. A computerized version of the
IAWPRC model is available in the public domain (3). Key
features of the model,* which include wastewater charac-
terization, biomass fractions, model stoichiometry and ki-
netics, and model presentation, are discussed below.
The notation system used to present the model was de-
veloped by another IAWPRC task group. This system is
presented in the literature (4).
5.5.1,1 Wastewater Characterization
Key aspects of Activated Sludge Model No. 1 include the
use of chemical oxygen demand (COD) to characterize
the organic matter in wastewater, rather than BOD5, and
the division of the organic matter into various fractions.
COD is used rather than BODS to allow straightforward
calculation of process oxygen requirements. If both COD
and BODS data are available for a particular wastewater,
development of design COD values is easily accom-
plished. If only BOD5 data are available, it must be con-
verted to biodegradable COD values for input to the
model. Division of the organic matter into various fractions
allows use of a structured modeling approach, which is
necessary to accurately predict spatial and temporal vari-
ations in carbonaceous process oxygen requirements.
Figure 5-3 presents the division of organic matter utilized
by Activated Sludge Model No. 1. The organic matter is
first divided into biodegradable and nonbiodegradable
fractions. Nonbiodegradable organic matter is further di-
vided into soluble (S|) and particulate (X,) fractions ac-
cording to the different processing of the two materials in
the activated sludge system. Soluble, nonbiodegradable
materials simply pass through the system and appear in
the effluent at a concentration equal to the influent con-
centration. In contrast, particulate, nonbiodegradable ma-
terials are enmeshed into the activated sludge mixed
liquor and accumulate as sludge, which is captured in the-
secondary clarifier and returned to the biological reactor.
The concentration factor for a suspended growth system
is the ratio of the solids residence time (SRT) to the
hydraulic residence time (HRT) of the system and often
represents an approximate 20-fold concentration factor.
Biodegradable organic matter is also divided into two
fractions: readily biodegradable (Ss) and slowly bio-
degradable (Xs). The model assumes that slowly biode-
Slowly
(Xs)
Readily
(Ss)
Particulate
(X,)
Soluble
(S,)
Biodegradable
Nonbiodegradable
Figure 5-3. Division of organic matter in Activated Sludge
Model No. 1.
116
-------
gradable organic matter is first hydrolyzed into readily
biodegradable organic matter that is subsequently con-
sumed by heterotrophic bacteria. Note that these frac-
tions do not in general correspond to the division of
biodegradable organic matter into soluble and paniculate
fractions (see Reference 5 for a more detailed discussion
of this topic). Keep in mind that 1) the readily degradable
organic matter will typically be less than the soluble or-
ganic matter, as traditionally defined by standard practice;
and 2) biological methods are generally superior to physi-
cal/chemical methods for characterizing the readily bio-
degradable organic matter in a wastewater stream.
In general, the biodegradable organic matter content
(readily plus slowly biodegradable components) and the
measured BOD5 of a wastewater are related. In fact, they
represent different approaches to characterizing the same
components. In the absence of site-specific data, it may
be assumed that the biodegradable organic matter con-
tent is equal to the ultimate BOD (BODuit).
Nitrogen components are expressed as nitrogen (N) and
are divided into inorganic (ammonium, SNH, and nitrate
plus nitrite, SNO) and organic (soluble biodegradable, SND,
and particulate biodegradable, XND) fractions. Nitrogen
fractions are converted into their oxygen equivalents
when they are involved in oxidation/reduction reactions.
This conversion occurs internal to the model. The
stoichiometric factors are 4.57 mg O2/mg NHJ-N oxidized
and 2.86 mg O2/mg NO§-N reduced. Nondegradable or-
ganic nitrogen (both soluble and particulate materials) is
treated in the same manner as nondegradable organic
matter. DO (S0) and alkalinity (SAu<. molar units) are also
included in the model.
5.5.1.2 Biomass Fractions
Three biomass fractions are included in the model: active
heterotrophic bacteria (XB,H), active autotrophic bacteria
(XB,A), and tne inert products of biomass decay (XP).
Heterotrophic bacteria include those organisms that are
both capable and incapable of denitrification. Correction
factors are used to account for 1) growth of heterotrophic
bacteria (rig) and 2) hydrolysis of slowly biodegradable
organic matter (t|h) under anoxic conditions. These fac-
tors account both for the fraction of heterotrophic bacteria
that are capable of denitrification and for any difference
in activity from aerobic to anoxic conditions. The factors
are empirical in nature and, as such, their numerical val-
ues should be confirmed for a particular application.
Growth of active heterotrophic bacteria occurs when DO
and/or nitrate nitrogen is present; no growth occurs when
both DO and nitrate nitrogen are absent. Autotrophic bac-
teria include both Nitrosomonas and Nitrobacter which
grow only when DO is present.
The model uses an unconventional approach to biomass
decay. Active biomass (both heterotrophic and autotro-
phic) is assumed to decay into slowly biodegradable or-
ganic matter (Xs) and inert product (XP) in a process that
does not require oxygen or nitrate. Oxygen (or nitrate)
utilization occurs only when the slowly biodegradable or-'
ganic matter is subsequently hydrolyzed and then meta-
bolized by heterotrophic bacteria. This has two impacts:
first, the decay rate must be much higher than values
traditionally used in order to produce oxygen consumption
and net biomass production rates consistent with those
observed in practice; second, biomass decay is assumed
to occur under all conditions, but the resulting hydrolysis
of slowly degradable organic matter (Xs) occurs only un-
der aerobic or anoxic conditions. Consequently, slowly
biodegradable organic matter will accumulate under an-
aerobic conditions.
Table 5-1 summarizes the components of the model,
along with their descriptions.
Table 5-1. Activated Sludge Model No. 1: System
Components
Compo- Compo-
nent nent
Number Symbol Definition
1
2
3
4
5
6
7
8
9
10
11
12
13
S,
Ss
X,
Xs
XB.H
XB,A
XP
Soluble inert organic matter, IvKCOD)!."3
Readily biodegradable substrate,
Particulate inert organic matter,
M(COD)L'3
Slowly biodegradable substrate,
M(COD)L3
Active heterotrophic biomass,
Active autotrophic biomass, M(COD)L"3
Products from biomass decay,
M(COD)L3
DO, IvKCOD)!.-3 '
Nitrate and nitrite nitrogen, M(NL)L'3
Ammonia nitrogen,
SND Soluble biodegradable organic nitrogen,
M(N)!--3
XND Particulate biodegradable organic
nitrogen, M(N)L'3
SALK Alkalinity, Molar units
M = mass
L = length
5,5.1.3 Stoichiometry and Kinetics
Table 5-2 summarizes the stoichiometric and kinetic pa-
rameters used in the model, while Table 5-3 summarizes
the kinetic expressions. Aerobic growth of heterotrophic
bacteria includes Monod expressions for readily biode-
117
-------
Table 5-2. Activated Sludge Model No. 1: Kinetic and
Stolchiometrlc Parameters
Kinetic event
Heterotrophic growth and decay
Aulotrophic growth and decay
Correction factor for anoxic growth of
heterotrophs
Ammonificatfon
Hydrolysis
Correction factor for anoxic hydrolysis
Stolchiometrlc coefficient
Hotorotrophic yield
Autotrophic yield
Fraction of biornass yielding decay
products
Mass WMass COD in biomass
Mass N/Mass COD in decay products
Symbols
uHi Ks,
MA
bA
kh, Kx
YH
YA
fp
"XB
IXP
gradable organic matter (Ss) and DO (S0), while anoxic
growth also incorporates a Monod term for nitrate-nitro-
gen (SNO)- Aerobic growth of autotrophic bacteria includes
Monod expressions for both ammonia (SNH) and DO. De-
cay of heterotrophic and autotrophic bacteria are first-or-
der expressions that continue under aerobic, anoxic, and
anaerobic conditions. Ammonification of soluble organic
nitrogen is modeled as a first-order process with regard
to remaining substrate and biomass. Hydrolysis of slowly
biodegradable organic matter (Xs) is modeled as a Monod
function of the ratio of slowly biodegradable organic mat-
ter to heterotrophic bacteria, DO (S0), and nitrate (SNO)-
Hydrolysis of slowly degradable organic nitrogen (thereby
producing ammonia) is proportional to the hydrolysis of
slowly biodegradable organic matter (Xs).
Table 5-4 summarizes example values of the stoichiomet-
ric and kinetic coefficients in Activated Sludge Model
No. 1 as summarized by the IAWPRC task group. These
values are presented for informational purposes only and
should not be viewed as typical of any wastewater or
application. As discussed in Section 5.4, models must be
properly calibrated prior to their use in full-scale applica-
tions. A subsequent section of this chapter discusses ex-
Table 5-3. Activated Sludge Model No. 1: Kinetic Expressions
Component -> i
j Process
-------
Table 5-4. Activated Sludge Model No. 1: Values of Stoichiometric and Kinetic Parameters
Symbol Unit Value at 20°C
Value at 10°C
Stoichiometric parameters
YA g cell COD formed/g N oxidized
YH g cell COD formed/g COD oxidized
fp dimensionless
fa g N/g COD in biomass
to 9 N/g COD in endogenous mass
Kinetic parameters
0.24
0.67
0.08
0.086
0.06
0.24
0.67
0.08
0.086
0.06
/\
HH
Ks ^
KO.H
KNO
^
t|g
•n.h
*h
Kx
A
NH
*»O A
ka
d'1
g COD/m3
g Oa/m3
g NO;-N/m3
d'1
dimensionless
g slowly biodegradable COD/g cell COD/d
g slowly biodegradable COD/g cell COD
d'1
g NHj-N/m3
g O2/m3
m3/(g COD/d)
6.0
20.0
0.20
0.50
0.62
0.8
0.4
3.0
0.03
0.80
1.0
0.4
0.08
3.0
20.0
0.20
0.50
0.20
0.8
0.4
1.0
0.01
0.3
1.0
0.4
0.04
amples in which Activated Sludge Model No. 1 has been
used and identifies actual parameter values selected.
Examination of Table 5-4 indicates values of 0.08 for fp
and 0.62 for bH. In the unconventional approach to
biomass decay described in Section 5.5.1.2 (a death-re-
generation approach rather than an endogenous mass
loss approach), 62 percent of the active mass dies each
day (bH = 0.62), yielding substrate for the regeneration of
38 percent, resulting in an overall net loss of 24 percent.
However, now only 8 percent (/P = 0.08) of the active
mass is not biodegradable. This yields the same mass of
endogenous residue as in the endogenous mass loss
approach where the nonbiodegradable fraction is 0.20
and the decay coefficient is 0.24 (6). Therefore, the co-
efficient values and IAWPRC equation listed in Table 2-13
yield about the same loss of active biomass and formation
of XP as obtained with Activated Sludge Model Number
1 and the coefficients for fp and bH listed in Table 5-4.
5.5.1.4 Model Presentation
As illustrated in Table 5-5, a matrix format is used to
present the model and its structure. Listed across the top
of the matrix are the components of the model; they are
defined and their respective units are listed across the
bottom. Microbial processes incorporated into the model
are listed vertically on the left-hand side. Kinetic expres-
sions are listed on the right-hand side of the matrix for
each corresponding microbial process. Within the matrix,.
Stoichiometric coefficients are listed for each model com-
ponent that is affected by a particular microbial process.
Thus, the reaction rate for transformation of a particular
model component is the product of the Stoichiometric
coefficient and the corresponding kinetic expression. For
example, the reaction rate for readily biodegradable or-
ganic material (Ss) consists of aerobic growth of het-
erotrophic bacteria (with Stoichiometric coefficient -1/YH),
anoxic growth of heterotrophic bacteria (with Stoichiomet-
ric coefficient -1/YH), and hydrolysis of slowly biode-
gradable organic matter (with a Stoichiometric coefficient
of 1). Although this approach may be unusual to many, it
allows the model to be presented in a particularly succinct
fashion.
5.5.1.5 Examples
Experience with the application of Activated Sludge Model
No. 1 to pilot or full-scale wastewater treatment systems
is increasing, and the results indicate there is potential
for its successful application. Several examples of the use
of this model are summarized in Table 5-6. The general
approach used to date involves beginning with default
values for the Stoichiometric and kinetic parameters in the
model (e.g., those presented in Table 5-4) and then ad-
justing the numerical values of these parameters to
achieve an acceptable correlation of model predictions to
pilot and/or full-scale data. Table 5-6 illustrates that this
approach has been generally successful and indicates
the types of adjustments that have been made.
119
-------
£
•a
o
•c
•a,
O.
»J
I
o
to tn
(O X
x.
to X0
^X*
o wx_
1
S CM
It-*
< S 8
«j g §
•S|l
II
•
f
CB §55
*
as
••i? *«s5
I
-S
I
H£
? «,' -|'l
s w-iL'x
cn^S ° o ° •£ E o
sterns
•5 E>
= 11
-'-
il
o
'2
120
!
-
w
'S.
2
|.
o
1
..
eiqepej6opoiq 9)einoi}jBd
E-i(aoo)w 'u
Ojq eiqniog
e-i(aoo)iAi '
+ "HN
'usSojjju eiuiju pue e;BjJ!N
E-i(aoo)w
'(GOO 3AHB68U) ueBAxo
l SSBLUOjq
aiojj BUISUB sjonpojd e)B|no;)jBd
jq AJMOIS
E-i(aoo)w
'ssBiuoiq oindoj}Oje}8Lj eAjpv
'aiBJjsqns
'J9HBOI oiuBBjo yau\ 9)E|noiyBd
E-l(aOO)tM
8|qBpBj6spo|q ,
e|qn|os
2 *" Eo 5 Q a
i, ill 1*1-^1-.
ajSS^oir^iocOtoaJww
E S oS o 2>t35
-------
Table 5-6. Examples: Use of IAWPRC Activated Sludge Model No. 1
Facility
Six Municipal
Facilities
DSM Chemicals,
Augusta, GA
Application
Prediction of process
oxygen requirements
Design pilot plant
study, design
full-scale facility
Adjustments
Required
YH> bH. Tig. 'Hhi MA, bA
Comments
Only minor modifications
required for each facility
Extensive wastewater
characterization, and analysis
of pilot plant data. Industrial
wastewater.
Reference
7
8
Valenton, France;
CMRL Pilot Plant,
France; Henin,
France;
Process evaluations
and designs
Flauil, Switzerland; Measurement of
Zurich, Switzerland; process parameters,
Dietikar, Switzerland; process analysis
Zurich, Switzerland Process Analysis
Renton, WA; Santa
Fe, NM
Process oxygen
requirement
distribution
H, Ks, bH
A .. ., A
M-H, YH. Ks, MA
None
Model modified to incorporate
adsorption of readily
degradable colloids.
Used first-order rate expression 10
for hydrolysis of Xs. Did not
evaluate nitrification.
Used first-order rate expression 11
for hydrolysis of Xs.
None 12
The results presented in Table 5-6 suggest that site-spe-
cific adjustments may be required most often for the maxi-
mum specific growth rates of the heterotrophic (u,H) and
autotrophic (u,A) bacteria. Adjustments in u,H have been
required to produce acceptable oxygen uptake rate/DO
profiles along plug flow biological reactors. Adjustments
in U.A are necessary because of the effect of temperature
and the presence of inhibitory materials on the nitrifying
bacteria. The individual references should be consulted
concerning the magnitude of the required adjustment;
however, adjustments of plus or minus 50 percent are not
uncommon. Also, note the effect of parameter interaction.
For example, an adjustment in HH may a's° necessitate
a corresponding adjustment in Ks. While ongoing appli-
cation of the model and detailed review of the results is
necessary to develop additional insight, the results to date
are certainly encouraging.
Wastewater characteristics must also be assessed appro-
priately to allow the model to predict system responses
accurately. The case histories referenced in Table 5-6
also present insight into this issue.
5.5.2 Fixed Growth Models
In many respects, the modeling of fixed growth systems
is more difficult than the modeling of suspended growth
systems. The increased difficulty results from at least two
factors. One is the effect of diffusional resistance on the
substrate removal rate in such systems, as discussed in
Chapters 3 and 4. To accurately predict the performance
of fixed growth systems, the effects of diffusional resis-
tance must be incorporated into the model. These addi-
tional factors do not need to be considered explicitly in
suspended growth models.
A second factor is the heterogeneous and relatively poorly
characterized conditions that occur within many fixed
growth reactors. For example, in rotating biological con-
tactors (RBCs) the growth media (and, consequently, also
the biological film) is alternately submerged in the com-
pletely mixed reactor basin and exposed to the atmos-
phere. Perhaps the most heterogeneous environment is
that which exists in a trickling filter. In most trickling filters,
influent flow is dosed on a periodic basis using a rotating
distributor. Consequently, the media is alternately wetted
and then exposed directly to the atmosphere. Similarly,
the hydrodynamic conditions within the trickling filter are
poorly characterized (13). Since liquid flow across the
trickling filter media is not sheet flow (i.e., not all the
trickling filter media surface is fully wetted), charac-
terization for modeling is complicated.
Modeling is also hampered by fundamental gaps in
knowledge, such as the lack of information about the
factors that affect the fate of particulate matter in biofilms.
While it is known that certain biofilms will entrap and
subsequently metabolize some of the particulate matter
in an applied wastewater (14), and that entrapped par-
ticulates will reduce the activity of the biomass as they
displace active biomass from the biofilm, the mechanisms
of attachment and metabolism for particulate matter re-
121
-------
main poorly understood. Additional research is needed to
resolve these and other issues so that more mechanisti-
cally correct models can be developed.
In spite of these difficulties, research continues on the
development of models of fixed film processes that accu-
rately predict process performance. An IAWPRC task
force, similar to that which developed Activated Sludge
Model No. 1, is currently working on the development of
a general purpose biofilm model. In addition, researchers
are working on the development of models to predict the
nitrification performance of fixed film processes, either
singly or in combination with carbon oxidation. Two ex-
amples include the RBC model developed by Gujer and
Boiler (15) and the trickling filter model developed by
Logan, as modified and extended for nitrification by
Parker et al. (16).
5.5.3 Available Computer Programs for Modeling
Several microcomputer-based programs have been de-
veloped to facilitate use of the process models described
above. Table 5-7 summarizes several features of these
programs for suspended growth systems and provides
contacts for the software. The information is intended to
Table 5-7. Example Nitrification and Denitrification Mathe-
matical Modeling Microcomputer Packages
Name Features Contact
SSSP
Direct implementation
of Activated Sludge
Model No. 1. Public
domain
ASIM
EFOR
GPS-X
Flexible modeling tool
that can incorporate
Activated Sludge Model
No. 1 or more
complicated model
Activated Sludge Model
No. 1 plus clarifier
model
General purpose
simulator, Activated
Sludge Model No. 1
plus clarifier
C.P.L. Grady, Jr.
Environmental Syst.
Engr.
Rich Environmental
Research Laboratory
Clemson Research
Park
Clemson, SC 29634
Dr. Will! Gujer
Abt. Ingenieurwissen
Schaften
Uberlandstrasse 13
EAWAG
CH-8600 Dubendorf
Switzerland
Mr. Jan Peterson
EFOR
I. Kruger AS
Gladsaxevej 363
DK-2860 Soborg
Denmark
Hydromantis, Inc.
1685 Main St. West
Suite 302
Hamilton
Ontario L8S 1G5
illustrate the availability of such tools, not to serve as an
endorsement of any particular commercial product.
5.6 Example: Analysis of a Single-Sludge
Nitrification/Denitrification System
This section demonstrates the use of the IAWPRC Acti-
vated Sludge Model No. 1 to analyze and optimize a
single-sludge nitrification/denitrification system, using the
influent wastewater characteristics presented in Chapter
2 along with information on the single-sludge nitrifica-
tion/denitrification systems developed in Chapter 8. Thus,
the results of this analysis can be compared to data pre-
sented in Chapter 8.
5.6.1 Development of Wastewater Characteristics
The initial step in the application of the IAWPRC Activated
Sludge Model No. 1 is development of the specific influent
wastewater characteristics required by the model (see
Table 5-1). (See Chapter 2 for a more detailed description
of the selected wastewater treatment plants and the bio-
logical process influent wastewater characteristics.)
First, consider the simple Plant A as described in Figure
2-5 and Table 2-15, and calculate the particulate inert
organic matter (X|). According to Table 2-11, the particu-
late COD is 180 mg/L and it is 30 percent nonbiodegrad-
able; that is 0.3 x 180 mg/L, or 54 mg/L. For simplicity,
use X| equal to 55 mg/L (expressed as COD). Assuming
that the slowly biodegradable substrate (Xs) is equal to
the biodegradable particulate COD (see Section 5.5.1.1),
Xs may be calculated as the difference between the par-
ticulate COD and the particulate inert organic matter,
which is 180 - 55 or 125 mg/L (expressed as COD).
According to Tables 2-11 and 2-15, the soluble COD is
100 mg/L and the soluble nonbiodegradable organic mat-
ter is roughly 20 mg/L, expressed as COD; consequently,
it is assumed that the readily biodegradable substrate
(Ss) is calculated as 100 - 20 or 80 mg/L (expressed as
COD).
According to Table 2-11, the total influent soluble nitrogen
is 24 mg-N/L. Assume that the ammonia-nitrogen (SNn)
concentration is 20 mg-N/L and the influent nitrate plus
nitrite (SNO) concentration is zero. Neglecting nonbiode-
gradable soluble organic nitrogen, the soluble degradable
organic nitrogen (SND) would be the difference between
the influent soluble nitrogen and the ammonia-nitrogen
concentration of 24 - 20, or 4 mg-N/L. The total influent
particulate nitrogen concentration is 6 mg/L. Assuming
that 1 mg/L of this concentration is nonbiodegradable, the
particulate biodegradable organic nitrogen (XND) is 5 mg-
N/L. The influent alkalinity is 120 mg/L as CaCO3. Since
there are 50 mg/L of alkalinity as CaCO3 per mM, this
corresponds to 2.4 mM of bicarbonate alkalinity. In sum-
mary, the biological reactor influent wastewater charac-
122
-------
teristics for the simple wastewater are:
X, = 55 mg/L as COD
Ss = 80 mg/L as COD
Xs = 125 mg/L as COD
SNO = 0 mg-N/L
SNH = 20 'mg-N/L
SND = 4 mg-N/L
XND = 5 mg-N/L
SALK = 2.4 mM
A similar analysis of the mass balance information from
the more complex wastewater treatment Plant B (Figure
2-6 and Table 2-16) yields the following estimates for the
biological reactor influent (primary effluent) wastewater
characteristics:
X, = 29 mg/L as COD
Ss = 86 mg/L as COD
Xs = 52 mg/L as COD
SMO = 0 mg-N/L
SNH = 22.4 mg-N/L
SND = 4 mg-N/L
XND = 2 mg-N/L
SALK = 2.4 mM
For both cases, the influent active heterotrophic biomass
(XB,H) and active autotrophic biomass (XB,A) are assumed
to be zero.
5.6.2 Detailed Analysis
As a further illustration of the use of the model, a design
will be developed for a 18,925 m3/d (5 mgd) average flow
facility with a peak month flow of 28,388 m3/d (7.5 mgd).
The more complex wastewater treatment Plant B is con-
sidered, and the facility is designed to meet a 10 mg-N/L
total nitrogen standard. The design provides an anoxic
volume of 2,914 m3 (0.77 MG) and an' aerobic volume of
5,450 m3 (1.44 MG). The anoxic and aerobic zones are
each divided into three cells in series. The return acti-
vated sludge rate is 72 percent of the peak month flow,
and the internal recycle is 160 percent of the peak month
flow. Oxygen transfer is provided by six slow-speed sur-
face mechanical aerators, each operating at 50 hp. A
uniform aeration pattern is proposed. Results from this
analysis can be compared to an alternative design ap-
proach shown for Example 1 in Chapter 8 (Section 8.5.3).
The public domain computer program SSSP was used to
carry out the analysis in this example. Table 5-8 illustrates
the computer input and output. Critical inputs include a
maximum specific growth rate for autotrophs of 0.45 d"1
(as was used for the example in Chapter 8), a total mean
cell residence time (MCRT), or solids retention time of
12.1 d (giving an aerobic MCRT of 9.4 d), and a KLa in
the aerobic section of 125 d"1 (based on uniform aeration
with the six 50 hp aerators). The design temperature
would be 15°C (59°F). The results indicate the following:
• The calculated MLVSS (mixed liquor volatile sus-
pended solids) in the aeration basin effluent is about
3,000 mg/L as COD, which corresponds to a concen-
tration of approximately 2,110 mg/L when expressed
as VSS (calculated using 1.42 mg COD/mg VSS).
This exceeds the design MLVSS concentration of
1,890 mg/L (expressed as VSS), as presented in
Chapter 8, and indicates a slightly higher solids load-
ing on the secondary clarifier.
• The predicted effluent total inorganic nitrogen concen-
tration (NHj-N plus NOx-N) is 7.7 mg/L, consisting of
0.5 mg/L for NHj-N and 7.2 mg/L for NOx-N. Thus, the
effluent organic nitrogen concentration could be as
much as 2.3 mg/L (10 mg/L minus 7.7 mg/L) and the
effluent total nitrogen concentration would be less
than 10 mg/L. Again, the viability of achieving this
level of performance must be carefully evaluated.
• The NOx-N concentration profile through the anoxic
zone (reactors 1, 2, and 3) is 3.0, 1.8, and 1.2 mg/L.
This indicates that more than enough nitrate is re-
turned to this zone to ensure that the denitrification
potential is fully utilized.
• The ammonia-nitrogen profile through the aerobic
zone (reactors 4, 5, and 6) indicates that the aerobic
zone is fully utilized. The fact that the ammonia-nitro-
gen concentrations is just reduced to 0.5 mg/L in the
last stage of the aerobic zone (reactor 6) indicates that
the entire aerobic zone must be used for nitrification.
• Denitrification is predicted to occur in both the anoxic
and aerobic zones. Based on the unit rates of denitrifi-
cation, it can be calculated that the total amount of ni-
trogen denitrified is 16.8 mg/L. Of this total, 19 percent
is predicted to occur in the aerobic zone (reactors 4, 5,
and 6). Denitrification can occur in the aerobic zone
(at least according to this model) as a result of the use
of Monod kinetics for denitrification. Denitrification is
predicted to occur in an aerated zone when the DO
concentration is relatively low and the NOX-N concen-
tration is relatively high.
• The DO concentration profile in the aerobic zone (re-
actors 4, 5, and 6) indicates that the placement of
aeration capacity is reasonable. A DO concentration
of 1.4 mg/L is predicted for the first stage of the aero-
bic zone (reactor 4), and the DO concentration is pre-
dicted to be 3.6 mg/L in the effluent from the aerobic
zone (reactor 6). A reduced DO concentration in reac-
tor 6 would further optimize process performance.
• Supplemental alkalinity is required, as indicated by a
. reactor effluent alkalinity of 0.3 mM. This corresponds
to an effluent alkalinity of 15 mg/L as CaCOs. Supple-
123
-------
Table 5-8. SSSP Input Parameters and Output for Example Problem
Kinetic and Stoichiometric Parameters
Parameter
Heterotrophic Organisms Autotrophic Organisms
"max-d"1
k, COD, g COD/m3
k, NH4-N, g N/m3
k, O2, g Oa/m3
Yield, g/g
b decay, d'1
Anoxlc growth factor
k, N03, g N/m3
Hydrolysis rate, d"1
Hydrolysis saturation ratio, g COD/g COD
Anoxic hydrolysis factor
Ammonification, m3/g COD/d
Fraction of particulate products, g COD/g COD
N in blomass, g N/g COD
N In part, prod., g N/g COD
O2 saturation concentration, g Oa/m3
4.0
10.0
0.55
0.67
0.62
0.80
0.20
2.20
0.15
0.40
0.16
0.08
0.086
0.06
9.0
Process Configuration
Reactor Specification
Reactor volume, m3
Feed fraction, 0 to 1
Mass trans, coeff. for O2, d"1
Recycle Input, m3/d
Reclrculation input, m3/d
Reclrculation originated from reactor
1
971
1.0
0
20439
45421
6
2
971
0
0
0
0
Reactor
3
971
0
0
0
0
Steady-State
Constituents
Heterotrophic organisms, g COD/m3
Autotrophlc organisms, g COD/m3
Particulate products, g COD/m3
Inert particulates, g COD/m3
Particulate organics, g COD/m3
Soluble organics, g COD/m3
Soluble ammonia N, g N/m3
Soluble nitrate/nitrite N, g N/m3
Soluble organic N, g N/m3
Biodegrad. part, organic N, g N/m3
Oxygen, g Oa/m3
Alkalinity, mo!e/m3
MLVSS, g COD/m3
O2 consumed, g O2/m3/d
Nitrate consumed, g NOg-N/m3/d
0.45
1.0
0.55
0.24
0.05
and Flow Distribution9
Number
4
1817
0
125.0
0
0
Solution19
5
1817
0
125.0
0
0
6
1817
0
125.0
0
0
Location
Feed
0.0
0.0
0.0
29.0
52.0
86.0
22.4
0.0
4.0
2.0
0.0
2.4
1
951.7
140.7
582.6
1191.0
142.8
9.7
6.8
3.0
0.7
10.5
0.1
1.1
3008.8
163.1
221.2
2
952.6
140.7
583.1
1191.0
144.5
3.3
6.8
1.8
0.4
10.7
0.0
1.2
3011.8
4.8
115.4
3
950.1
140.6
583.5
1191.0
146.5
1.6
6.9
1.2
0.3
11.0
0.0
1.2
3011.7
0.2
60.5
4
949.6
141.2
584.5
1191.0
140.5
1.9
4.1
3.7
0.4
10.6
1.4
0.8
3006.7
875.6
22.1
5
949.7
141.6
585.4
1191.0
134.0
2.0
1.7
5.9
0.4
10.3
2.1
0.5
3001.7
819.6
16.9
6
950.0
141.9
586.3
1191.0
127.1
2.0
0.5
7.2
0.5
9.8
3.6
0.3
2996.2
601.0
11.1
• Number of Reactors = 6, Solids Retention Time = 12.1, Average Flow Rate = 28,388 m3/d.
b Warnings: The average was taken of the k, O2 parameters. The alkalinity in reactors 4, 5, and 6 is below the 1 mole/m3 required to sustain
uninhibited biological growth.
124
-------
mental alkalinity of about 35 mg/L as CaCOs would be
required. ......
The model could be used in a number of ways to optimize
the process. First, the variation in process oxygen re^-
quirements through the aerobic zone (reactors 4, 5, and
6) suggests the use of tapered rather than uniform aera-
tion. Several variations were investigated using the
model. A 20-percent increase in the horsepower in the
first stage of the aerobic zone (from 50 to 60 hp) was
simulated by increasing the KLa from 125 d~1 to 150 cT1.
This resulted in a DO concentration in this cell of 2.0
mg/L, which led to less predicted denitrification in the
aerobic zone and an increased process effluent NOX-N
concentration. A reduction in the aerator horsepower in
reactor 6 was simulated by reducing KLa to 60 d"1. This
resulted in a reduction in the DO concentration to 2.9
mg/L (from the previous value of 3.6 mg/L) and a reduc-
tion in the effluent NOX-N to 6.3 mg/L due to increased
denitrification in the aerobic zone. However, based on the
data available, the use. of two-speed aerators with differ-
ent sizes might be appropriate. Sixty hp in the first aerobic
zone (reactor 4), 50 hp in the, second aerobic zone (re-
actor 5), and 40 hp in the third aerobic zone (reactor 6)
would represent a reasonable distribution.
If diurnal flow and load data are available, a more refined
analysis of the aeration system could be conducted. Such
an analysis would facilitate more precise sizing of the
aeration system to meet peak oxygen transfer require-
ments while not providing excess oxygen transfer capac-
ity. Such an analysis would also allow precise sizing of
the biological reactor to treat peak ammonia-nitrogen
loadings and avoid ammonia breakthrough. This, in fact,
is one of the primary benefits of the use of a model such
as Activated Sludge Model No. 1. Because it is a dynamic
model and is capable of quantifying such effects, it allows
more precise sizing of the system to adequately treat the
range of loads applied to it.
The capability to predict denitrification within the aerobic
zone is an interesting aspect of this model. However, such
predictions must be compared with full-scale results as
part of the model calibration and verification exercise de-
scribed in Section 5.3.
5.6.3 Alternative Comparison
The final use of Activated Sludge Model No. 1, which will
be illustrated, involves a comparison of reactor sizes re-
quired for the two plant types identified in Chapter 2 and
for designs intended to meet effluent total nitrogen limits
of 10 and 5 mg/L. The comparison is based on the waste-
water characteristics presented above. Results can also
be compared to design examples 1 and 2 in Section 8.5
and the design detailed in Table 5-8 to achieve a TN
effluent limit of 10 mg/L.
Comparable designs were developed by maintaining
MLVSS concentrations (expressed as VSS) of about
2,300 mg/L. An aerobic solids retention time of 9.4 d was
maintained for the 10 mg-N/L effluent total nitrogen cases
and a more conservative aerobic solids retention time of
11.7 d was used for the 5 mg-N/L effluent total nitrogen
cases. To ensure consistency in the comparison, a uni-
form DO concentration of 2 mg/L was maintained through
the aerobic reactors.
The results of the analysis are presented in Table 5-9.
Process designs consisting of a single anoxic and a sin-
gle aerobic zone are utilized for the 10 mg/L effluent total
nitrogen cases, while dual anoxic zone systems are util-
Table 5-9. Comparison of Process Designs for Complex
and Simple Wastewater Treatment Plants and Effluent Total
Nitrogen of 10 and 5 mg-N/L
TN = 10
TN = 5
Parameter
Complex Simple Complex Simple
HRT, hr3
First anoxic
First aerobic
Second anoxic
Second aerobic
2.5 2.5 2.4
4.6 8.1 5.3 .
— — 2.7
— — 0.5
2.4
9.9
2.7
0.5
TOTAL 7.1 10.6 10.9 15.5
MCRT, d
Total
First aerobic
MLVSS, mg/L
Recirculation input, %b
HAS, %b
NHJ-N, mg-N/L
NOX-N, mg-N/L
Denitrification, mg-N/L
Aerobic zone
denitrification, %
First anoxic zone 1.8 0.0 2.1
effluent
NOX-N, mg-N/L
Effluent alkalinity, 15 .30 30
mg/LasCaCO3
a Based on maximum month flow, without recycle.
b Percentage of maximum month flow.
14.4
9.4
2,312
160
72
0.1
7.1
17.5
21
12.1
9.4
2,360
160
72
0.2
4.9
18.3
38
24.0
11.7
2,384
400
72
0.1
3.6
21.3
17
18.3
11.7
2,327
400
72
0.2
0.9
22.4
31
0.2
40
125
-------
ized for the 5 mg/L effluent total nitrogen cases. The
required hydraulic residence times (HRTs) were 7.1-15.5
hr. The HRT increased by about 4-5 hr as the effluent
total nitrogen limit was reduced from 10 to 5 mg/L. The
difference in total HRT between Plants A and B was
similar.
The results also indicate differences in performance char-
acteristics. For example, effluent NOX-N concentrations
are lower for the simple wastewater treatment plant ex-
amples than for the complex treatment plant examples.
The model predicts a greater increase in denitrification
for the simpler Plant A than for the complex Plant B.
Interestingly, increased denitrification is predicted to occur
in the aerobic portion of the reactor (17 to 21 percent for
the complex wastewater treatment Plant B as compared
to 31 to 38 percent for the simpler wastewater treatment
Plant A). In addition, essentially complete denitrification
is predicted to occur in the first-anoxic-zone for the sim-
pler Plant A (as indicated by low first-anoxic-zone effluent
NOX-N concentrations), while denitrification is not com-
plete for the complex Plant B. Differences in effluent al-
kalinity are predicted for the four cases considered.
The proposed design that would allow the complex Plant
B to meet an effluent total nitrogen of 5 mg/L predicts an
effluent total inorganic nitrogen concentration (NOX-N plus
NHJ-N) of 3.7 mg/L. Alternatives to lower this value, and
thus provide a greater margin of safety relative to the 5
mg/L effluent total nitrogen limit, were investigated. Nei-
ther an increase in the size of the first anoxic zone nor
an increase in the internal recycle rate resulted in signifi-
cant changes in the effluent total inorganic nitrogen con-
centration. However, the effluent total inorganic nitrogen
was significantly reduced by converting the last portion
of the first aerobic zone to anoxic operation. This effec-
tively reduced the size of the first aerobic zone, but sig-
nificantly increased the size of the second anoxic zone.
These modifications could result in a reduction in the
effluent total inorganic nitrogen concentration to about
2 mg/L.
5.7 References
1, Henze, M., C.P.L. Grady, Jr., W. Gujer, G.V.R. Marais,
and T. Matsuo. 1987. Activated Sludge Model No. 1.
IAWPRC Scientific and Technical Reports No. 1. Lon-
don.
2. Henze, M., C.P.L. Grady, Jr., W. Guj'er, G.V.R. Marais,
and T. Matsuo. 1987. Model for single-sludge waste-
water treatment. Water Res. 21:505.
3. Bistrup, S.M., and C.P.L. Grady, Jr. 1988. SSSP-
simulation of single-sludge processes. JWPCF
60:351.
4. Grau, P., P.M. Sutton, M. Henze, S. Elmaleh, C.P.L.
Grady, Jr., W. Gujer, and J. Koller. 1982. Recom-
mended notation for use in the description of biologi-
cal wastewater treatment processes. Water Res.
16:1501.
5. Henze, M. 1991. Methods for wastewater and
biomass characterization. Proceedings of the Work-
shop on Interactions of Wastewater, Biomass and
Reactor Configuration in Biological Treatment Plants.
Copenhagen, Denmark (August). IAWPRC 0273-
1223/92.
6. Dold, P.L., and G.V.R. Marais. 1986. Evaluation of
the general activated sludge model proposed by the
IAWPRC Task Group. Water Sci. Tech. 18:63.
7. Baillod, C.R. 1989. Oxygen utilization in activated
sludge plants: simulation and model calibration.
EPA/600/2-88/065. Cincinnati, OH.
8. Givens, S.W., E.V. Brown, S.R. Gelman, C.P.L.
Grady, Jr., D.A. Skedsvold. 1990. Biological process
design and pilot testing for a carbon oxidation, nitri-
fication, and denitrification system. Presented at the
1990 Summer National Meeting of the American In-
stitute of Chemical Engineers, San Diego, CA (Au-
gust).
9. Lesouef, A., M. Payraudeau, F. Rogalla, and B.
Kleiber. 1991. Optimizing nitrogen removal reactor
configurations by on-site calibration of the IAWPRC
activated sludge model. Proceedings of the Work-
shop on Interactions of Wastewater, Biomass and
Reactor Configuration in Biological Treatment Plants.
Copenhagen, Denmark (August). IAWPRC 0273-
1223/92.
10. Kappeler, J., and W. Gujer. 1991. Estimation of ki-
netic parameters of heterotrophic biomass under
aerobic conditions and characterization of wastewa-
ter for activated sludge modelling. Proceedings of the
Workshop on Interactions of Wastewater, Biomass
and Reactor Configuration in Biological Treatment
Plants. Copenhagen, Denmark (August). IAWPRC
0273-1223/92.
11. Siegrist, H., and M. Tsuchi. 1991. Interpretation of
experimental data with regard to the Activated Sludge
Model No. 1 and calibration of the model for munici-
pal wastewater treatment plants. Proceedings of the
Workshop on Interactions of Wastewater, Biomass
and Reactor Configuration in Biological Treatment
Plants. Copenhagen, Denmark (August). IAWPRC
0273-1223/92.
12. Parker, D.S., M.S. Merrill, and M.J. Tetreault. 1991.
Wastewater treatment process theory and practice:
the emerging convergence. Proceedings of the Work-
shop on Interactions of Wastewater, Biomass and
126
-------
Reactor Configuration in Biological Treatment Plants. 15. Gujer, W., and M. Boiler. 1990. A mathematical model
Copenhagen, Denmark (August). IAWPRC 0273- for rotating biological contactors. Water Sci. Tech.
1223/92. 22:53.
13. Hinton, S.W., and H.D. Stensel. 1991. Experimental 16- Parker, D., M. Lute, R. Dahl, and S. Bernkopf. 1989.
observation of trickling filter hydraulics. Water Res. Enhancing reaction rates in nitrifying trickling filters
25:1390. through biofilm control. JWPCF 61:618.
14. Sarner, E. 1986. Removal of particulate and dis-
solved organics in aerobic fixed film processes.
JWPCF 58:165.
127
-------
-------
Chapter 6
Design Considerations for Biological Nitrification Processes
6.1 Introduction
Biological nitrification in municipal wastewater treatment
is particularly applicable to those cases where an ammo-
nia removal requirement exists, without the need for com-
plete nitrogen removal. Biological nitrification is also a key
step in the biological nitrification-denitrification approach
to nitrogen removal.
This chapter deals with design aspects of specific nitrifi-
cation process systems and the influence of various fac-
tors on nitrification kinetics in these systems. Methods for
the design of nitrification reactors are presented, together
with design examples and operating and performance
information from pilot and full-scale plants.
6.2 Classification of Nitrification Processes
In Section 3.3.6 (Effect of Feed Organic Carbon to Nitro-
gen Ratio), the feed C:N was shown to be a critical factor
affecting the design of nitrification systems. Previous de-
sign documents categorized nitrification systems accord-
ing to the degree of separation of the CBOD removal and
nitrification processes, often using the C:N as a bench-
mark (1,2). A nitrification system was categorized as a
combined or single-sludge carbonaceous oxidation-nitri-
fication system if the CBOD5:TKN was greater than a
certain arbitrary value, typically 4 or 5. Below the selected
value, the system was considered a separate, or two-
sludge, system. In this manual, nitrification systems are
not strictly categorized in this fashion, although systems
at higher and lower feed CBOD5 concentrations are
compared in Section 6.3, and the importance of the feed
C:N is illustrated through completion of design examples
at higher (Section 6.4.2.2) and lower (Section 6.4.3.2)
ratio values.
Biological nitrification reactors can be classified according
to the nature of their biological growth. Activated sludge
systems suspend biological solids in a mixed liquor by
some mixing mechanism; these are termed suspended
growth reactors. Units in which growth occurs on or within
a solid medium are termed attached growth, supported
growth or fixed film reactors. Certain bioreactors contain
microbial films in suspension, resulting from the addition
of very fine inert or active particles to provide microbial
growth sites. An example is the powdered activated carb-
on (PAC) activated sludge reactor. Such reactors are con-
sidered suspended growth reactors since the kinetic
reactions are well described by equations appropriate for
suspended growth (3). In certain nitrification systems, the
growth of both suspended and attached biomass is pro-
moted in the reactor. An example of this combined growth
configuration involves the suspension of highly porous
plastic foam particles in the reactor; attached growth oc-
curs in and around the plastic media while suspended
growth occurs in the liquid phase. Information pertaining
to these systems is presented in Section 6.6.
There are many different configurations of suspended,
attached, and combined growth reactors for nitrification.
New configurations are frequently introduced. A list of
alternatives is presented on Figure 6-1; this does not
include reactors designed to achieve both nitrification and
partial denitrification such as Passveer ditch type sys-
tems, or systems designed to achieve both nitrification
and biological phosphorus removal. Systems designed to
achieve total nitrogen removal (i.e., nitrification and deni-
trification) are discussed in Chapter 8. Descriptive and
design information pertaining to suspended, attached and
combined growth nitrification reactors is provided in Sec-
tions 6.4,6.5 and 6.6, respectively. Design considerations
to incorporate phosphorus removal into suspended
growth nitrification systems are presented in Section
6.4.10.2.
6.3 Comparison of Nitrification Systems at
Higher and Lower Carbonaceous Feed
Concentration
Representative nitrification systems are classified in Table
6-1 according to the nature of the biological growth in the
process reactor(s). Included are systems with both higher
and lower feed CBOD5:TKN values. Table 6-1 also shows
the distribution of total oxygen demand in the process
between carbonaceous sources (CBOD5) and nitroge-
nous sources. It can be seen that in systems with lower
CBOD5:TKN, the proportion of nitrogenous oxygen de-
mand (NOD) !s at least 70 percent of the total. In systems
129
-------
Suspended Growth or
Activated Sludge Reactors
Complete Mix -
Oxidation Ditch -
Step Feed -
Sequencing Batch -
Reactors (SBR)
- Conventional or
Plug Flow
- Extended Aeration
- Contact Stabilization
- High-Purity Oxygen
- Powdered Activated
Carbon (PAC)
Attached Growth,
Supported Growth, or
Fixed Film Reactors
Rotating Biological
Contactor (RBC)
Aerated Biological
Filter
- Trickling Filter and
Biotowers
- Packed-Bed Reactors
- Fluidized Bed
Combined Growth
Reactors
RBC-Suspended -
Growth
- Low Density Biosupport
Media-Activated Sludge
Figure 6-1. A listing of the majority of reactor configurations available for nitrification.
Table 6-1. Classification of Nitrification Facilities (Adapted in Part from Reference 1)
Oxygen Demand
Distribution
(Percent)
Location
Manassas, VA
South Band, IN
Fitchburg, MA
Whittier Narrows, LACSD, CA
Cincinnati, OH
Allentown, PA
Stockton, CA
Nature of
Biological Growth
Suspended
Suspended
Suspended
Suspended
Suspended
Attached
Attached
Reactor Feed
CBOD5:TKN
1.2
1.8
1.0
6.6
7.2
1.9
5.3
CBODg
20
28
18
61
61
30
54
NOD
80
72
82
39
39
70,
46
Reference
4
5
6
7
8
9
10
130
-------
with higher ratios, the proportion of nitrogenous oxygen
demand is lower than 50 percent.
As was shown in Section 3.3.6 (Effect of Feed Organic
Carbon to Nitrogen Ratio), the design solids retention
time, GO, is inversely proportional to the rate of substrate
removal (Equation 3-18). The rate of substrate removal,
qH, is directly proportional to the feed total CBOD, S0,
according to Equation 3-19:
So-s,
(3-19)
Reducing the feed total CBOD only will reduce qH, effec-
tively increasing the design solids retention time. One way
to reduce the feed CBOD is to place an organic carbon
removal step ahead of the nitrification step. This separate
two-stage configuration yields a nitrification process with
a lower carbonaceous feed concentration, reducing the
food available for growth of heterotrophic bacteria in the
second stage. Practical reactor nitrifying biomass concen-
trations can be maintained at long solids retention times
(15 to 25 days), with reasonable hydraulic retention times.
A procedure for reducing the substrate removal rate, but
without separating the carbon oxidation and nitrification
processes, is to increase the biological solids in the sys-
tem. This single-stage configuration uses higher concen-
trations of biological solids (i.e., the MLVSS in a
suspended growth reactor) or increases the volume of the
reactor, while maintaining the concentration of biological
solids at the same level.
The initial development of a suspended growth nitrifica-
tion system with a low feed CBOD5 concentration was
oriented to the isolation of the carbonaceous removal and
nitrification steps so that each could be separately con-
trolled and optimized (11). By placing a carbon removal
step ahead of the nitrification step, the second-stage
sludge would contain a higher percentage of nitrifying
organisms per unit of MLVSS than found in a system
receiving a high feed CBOD5. A higher percentage of
nitrifying organisms makes it easier to increase the reac-
tor solids retention time thus providing a less temperature
sensitive process configuration. The first applications of
two-stage suspended growth nitrification systems were in
the northern portion of the United States where low liquid
temperatures (i.e., less than 10°C [50°C]) were experi-
enced in the wintertime. Such systems were later applied
in moderate climates such as Florida and California (4).
There have been difficulties maintaining a sufficient nitri-
fying sludge inventory in low CBOD5 suspended growth
nitrification systems (12-14). Typically, effluent solids
were observed to fluctuate between 10 and 50 mg/L,
composed primarily of dispersed solids that were not cap-
tured in the secondary clarifier. It has been suggested
that a high fraction of the mixed liquor must be heterotro-
phic in order to maintain good bioflocculation in nitrifica-
tion systems (5). An unstable condition will exist in these
low feed CBOD5 systems if the effluent solids exceed the
net yield of organisms grown in the system and captured
by the clarifier. Several remedies are available. At some
locations, solids from the upstream carbonaceous oxida-
tion suspended growth reactors were periodically trans-
ferred to the nitrification stage to maintain an adequate
solids inventory. In other cases, first-stage pretreatment
configurations were chosen such that lower CBOD5 re-
movals were accomplished when compared to sus-
pended growth systems. This raised the CBOD5 feed to
the nitrification stage, increasing the synthesis of het-
erotrophic bacteria and improving the settleability of the
mixed liquor solids. Examples include primary clarification
with chemical addition (15) and the use of attached
growth reactors (e.g., trickling filter) for partial carbona-
ceous removal. Further information concerning the cou-
pling of attached and suspended growth reactors for
sequential carbonaceous removal and nitrification is pre-
sented in Section 6.6. In another case, 10 percent of the
primary effluent was bypassed around the carbonaceous
removal step to the nitrification step to overcome the low
solids synthesis issue (16). The amount of primary efflu-
ent bypassed can be varied in order to control the solids
retention time according to a procedure presented else-
where (17). Recirculation of filter backwash solids to the
nitrifying system also is effective and is the method of
choice at a number of facilities with effluent filters.
6.3.1 Pretreatment for Removal of Carbonaceous
Material and Inhibitory Compounds
P/etreatment alternatives provide varying degrees of or-
ganic carbon removal ahead of the nitrification step. As
discussed earlier, a high degree of organic carbon re-
moval through pretreatment will lead to the highest am-
monium oxidation rates per unit of MLVSS in the
downstream nitrification stage. This implies that reactor
size will diminish with increasing degrees of carbon re-
moval in the pretreatment stage. But low feed carbon
levels can cause difficulties in separate stage suspended
growth nitrification systems, as discussed in Section 6.3.
Conversely, low levels of organics in the influent to at-
tached growth reactors can be advantageous because of
reduced competition on the media surface between het-
erotrophic and nitrifying organisms and because the low
synthesis of solids, results in very low levels of solids in
the effluent. In some cases this can eliminate the need
for a clarification step, especially if followed by effluent
filtration or some other downstream treatment process
such as denitrification.
The selection of a pretreatment step to reduce the feed
CBOD5 to the nitrification stage can also be beneficial in
protecting the nitrifiers against inhibitory compounds pre-
sent in the wastewater as a result of industrial discharges.
The types of inhibitory compounds removed will vary,
depending on the specific unit operations employed in the
pretreatment step. Biological pretreatment alternatives
131
-------
provide a degree of protection against both organic and
heavy metal inhibitors. An exception would be organics
that are difficult to degrade or are resistant to biological
oxidation, such as, respectively, the solvents perchlo-
roethylene and trichloroethylene which have been identi-
fied as inhibitors to nitrification (5,18). Lime or metal salt
chemical treatment is one of the most effective processes
for removal of a wide range of metals (19).
Federal and state pretreatment regulations are directed
toward eliminating pollutants that are incompatible with the
operation and performance of municipal treatment plants.
An effective industrial pretreatment program should pre-
clude concern for issues regarding inhibition of the biologi-
cal processes at a treatment plant. If there is a problem
with inhibitory levels of organics or metals, the source
should be located. If it cannot be eliminated through some
pollution prevention, waste minimization techniques, then
pretreatment will be necessary before discharge.
Procedures have been developed to screen for nitrifying
inhibitors in the wastewater and to assess the effective-
ness of various pretreatment alternatives at reducing in-
hibition (20,21). These might be considered when there
are major industrial users in the system, and the potential
exists for the discharge of problem compounds. Perhaps
the simplest screening procedure involves batch oxygen
uptake tests using a respirometer to measure oxygen
utilization (21). Composite wastewater samples are sub-
jected to various pretreatments (e.g., alum or powdered
activated carbon via a jar test procedure or to biological
oxidation by batch aeration). Each treated sample is then
split and placed into two respirometers. One respirometer
is used as a non-nitrifying control by treatment with a
nitrification inhibitor such as Allylthiourea or N-Serve (2-
chloro-6 trichloromethyl pyridine). Both respirometers are
inoculated with a small amount of mixed liquor from a
nitrifying activated sludge plant. Alternatively, a seed ac-
climated to the wastewater in question may provide more
meaningful results where the inhibiting compound is bio-
degradable; the compound may be removed either in a
batch pretreatment evaluation or by the heterotrophic
component of an acclimated nitrifying sludge used in the
respirometer study. Differences between the oxygen used
in the control and in the seeded sample can be used to
establish batch nitrification rates. At the end of the test,
the respirometer contents are sampled and analyzed for
the nitrogen species to confirm whether nitrification took
place in the inoculated samples as well as to check the
control. The adequacy of the seed used can also be
checked by running an inoculated, but uninhibited, sam-
ple known to contain ammonia and organics, but no in-
hibitors.
The batch nitrification rates can be examined to deter-
mine the most suitable pretreatment technique among the
options examined. Often, some of the pretreatment tech-
niques will result in little or no nitrification in the inoculated
sample, indicating inadequate removal of the inhibitor(s).
In other cases, the pretreatment techniques will allow
vigorous nitrification in the sample indicating good re-
moval of the inhibitor(s). The particular pretreatment tech-
nique that is effective may also indicate the type of
inhibitor that is interfering with nitrification and may permit
identification and elimination of the source to the system.
For instance, if lime treatment is effective, the problem
may be a heavy metal that can be precipitated by lime.
Alternatively, if biological oxidation is ineffective but acti-
vated carbon treatment allows nitrification to proceed,
then a nonbiodegradable organic is suspect. Subsequent
specific analyses can then be run in the identified cate-
gory of compounds. If the inhibitors cannot be eliminated
by a source control program, often a pilot study of the
process identified by the bench scale procedure can be
justified to confirm the process selection. Pilot studies
also have value in determining the ability of the nitrifiers
to adapt to the toxicants, something the batch test with
an unacclimated seed is not capable of doing. Note that
the use of chemical addition as a pretreatment step to
the nitrification stage may cause significant changes in
alkalinity and pH. Normally this pretreatment approach is
selected in order to achieve a degree of phosphorus re-
moval as well. The effects of chemical addition on alka-
linity and pH are discussed in Chapter 2 (Section 2.7.2).
6.4 Nitrification in Suspended Growth
Reactors
Suspended growth systems were commonly used in Eng-
land to obtain dependable nitrification long before their
use became widespread in the United States. Early U.S:
conventional activated sludge plants often nitrified in the
warmer months of the year or if they were underloaded.
Nitrification became unpopular because of the additional
aeration power cost and the propensity of some sludges
to float in the sedimentation tank as a result of denitrifi-
cation, and it was questioned whether the process was
worth the added expense in many cases (22). As a con-
sequence, ways and means were sought to prevent nitri-
fication, rather than to encourage it, by increasing organic
loading, using tapered aeration, or picking modifications
of the process which were less favorable for nitrification.
This early experience with the process may have led to
some early uncertainty about its reliability. i
At least eight suspended growth reactor configurations
can be designed to incorporate nitrification (Figure 6-1),
all of which are modifications of the activated sludge proc^
ess. It may be appropriate to classify plug flow, complete
mix, contact stabilization, oxidation ditch, and sequencing
batch reactors as aeration basin configurations and the
remaining configurations as process modifications to
these reactors. Simplified schematic representations of
four of the systems are presented in Figure 6-2. Descrip-
tive information and design and performance information
132
-------
Raw
Wastewater
Effluent
Conventional Activated Sludge Plant
Raw Wastewater or
Primary Effluent
MM+M
Aeration Tank
4 4 4 A ''
IF )l
Return Sludge
Complete Mix Plant
Effluent
Raw )
Wastewater
I
Sludge
Sludge
Reae ration
Tank
Contact
Tank
Return Sludge
Reaeration-Aeration Plant
Effluent
, r Excess Sludge
Raw
Wastewater
Sludge
f
Aer
>
ation
^
r
Tank
J-
>
Return Sludge
Step-Feed Plant
Figure 6-2. Suspended growth reactor configurations.
Effluent
Excess Sludge
133
-------
pertaining to each suspended growth configuration are
presented in Sections 6.4.2 through 6.4.9. In Section
6.4.1, the nitrification kinetic theory presented in Chapter
3 is applied to the design of suspended growth reactors.
6.4.1 Application of Kinetic Theory to Design
The nitrification kinetic'theory presented in Chapter 3 may
be directly applied to the design of suspended growth
reactors that are compatible with nitrification. The equa-
tions must be adapted to the specific configuration under
consideration, but in most cases this adaptation is rela-
tively straightforward.
Nitrification kinetic theory can be applied to define the
following parameters:
1. The peaking factor required to handle diurnal tran-
sients in loading to prevent significant ammonium
bleedthrough under peak load conditions.
2. The minimum and design solids retention times under
the anticipated conditions of pH, DO, and tempera-
ture. As discussed in Section 6.4.10.1, the design sol-
ids retention time is calculated through the use of an
overall process design factor, which accounts for in-
fluent loading variations and other factors, such as
process variability (e.g., DO), uncertainty in kinetics,
and the potential presence of inhibitory compounds.
3. The observed organic removal rate and the observed
ammonium oxidation rate specific to the feed C to N
ratio, based on measurement of the total reactor VSS.
4. The required hydraulic retention time in the suspended
growth reactor and the reactor volume.
5, The excess sludge wasting schedule.
Determining the design solids retention time is the first
step in sizing the suspended growth nitrification reactor.
It follows from the relationship between the net specific
growth rate (u/N) and the solids retention time (0C) of the
organisms in the reactor that has been presented in
Chapter 3:
ec=-4- (3-13)
and recognition of the need to ensure that Equation 3-17
is satisfied:
6^>6^ (3-17)
Thus, the solids retention time approach to sizing nitrifi-
cation reactors has a fundamental basis. The procedure
classically has been simplified by making certain assump-
tions when specifying the observed or net yield (YNET) of
total VSS in the suspended growth system. As discussed
in Section 3.3.6, and further detailed in Chapter 5, the
levels of feed inert VSS, organism decay and other fac-
tors will influence the value of the observed yield. Chapter
5 presents information which allows application of the
solids retention time design approach, but in a more
rigorous fashion, accounting for many of these assump-
tions.
The use of experimentally measured ammonium oxida-
tion or nitrification rates for reactor sizing is also valid.
But the limitations of this rate approach must be realized,
as previously discussed in Section 3.3.6.
6.4.1.1 Solids Retention Time Design Approach
The first step in sizing the suspended growth nitrification
reactor is the selection of the solids retention time. To do
this, one must account for the effects of the various kinetic
factors on the growth rate of Nitrosomonas. The com-
bined kinetic expression for Nitrosomonas growth, ac-
counting for the effects of ammonium-nitrogen
concentration and temperature according to Equations
3-10 and 3-14, is:
N
(6-1)
This applies under the constraints that the reactor design
ensures, nonlimiting DO and pH conditions according to
the recommendations of Sections 3.3.3 and 3.3.4. For
example, at T = 20°C, N = 2.5 mg/L and under nonlimiting
DO and pH conditions, Equation 6-1 would yield:
UN = 0.47 (1.63) (0.71) = 0.54 d"1
In calculating u,N, KN was set at 1 .0 mg/L, as suggested
in Section 3.3.2. The calculated u.N represents the maxi-
mum possible nitrifier growth rate under the environ-
mental conditions specified at an NHJ-N concentration of
2.5 mg/L. The solids retention time corresponding to u,N
is the theoretical minimum solids retention time (0™) re-
quired under the specified environmental conditions and
is calculated from Equation 3-13 with bN = 0. That is:
.
c UN
Lawrence and McCarty introduced the concept of a safety
factor (SF) in the application of biological treatment proc-
ess kinetics to design (23). They noted that the safety
factor, which was defined as the ratio of the design solids
retention time to the minimum solids retention time, was
necessary to ensure high treatment performance and
process stability; to provide resistance to toxic upsets;
and to minimize process variations caused by pH ex-
tremes, low DO concentration and inhibitory materials. As
noted earlier in Section 6.4.1, design considerations also
include consideration of a peaking factor (PF) to ensure ;
that ammonium breakthrough does not occur during diurnal
peaks in load. Both the SF and PF concepts are considered
in selecting the overall design factor (DF). In some cases
the PF and DF are equivalent. However, in other circum-
134
-------
stances the product of the SF and PF may be more
appropriate to establish the DF. The design solids reten-
tion time, resulting from the use of an overall DF, will
typically be 1.5 to 3.0 times greater than the minimum
solids retention time. Considerations in selecting the
peaking and design factors are discussed in Section
6.4.10.1.
Once G£ has been determined, Equations 3-13 and 3-10
can be used to calculate, sequentially, the design nitrifier
growth rate and the actual steady state ammonium con-
tent of the effluent. Equation 3-10 must be adapted to
reflect the hydraulic conditions characterizing the sus-
pended growth reactor (e.g., complete mix versus plug
flow) to calculate effluent ammonia levels. This consid-
eration is illustrated in the design examples contained in
Sections 6.4.2.2 and 6.4.3.2.
The concentration of active heterotrophic biomass in the
reactor is represented by the following expression:
V H ,0 q >
Xl-i+bHe0(S°~Sl) t
(6-2)
where the term Yn/(1 + bH90) is the observed yield coef-
ficient, YNET, for active biomass.
This expression in conjunction with Equation 3-19 will
give the following expression:
qH
(6-3)
As shown in Table 2-13, the classical approach to net
VSS production, X, (i.e., the summation of active biomass
plus endogenous decay products plus inert material in
the entering wastewater) is an equation of the same form
as Equation 6-2. The expression is:
X =
Y(A substrate) 60
+b9r
t
(6-4)
and this equation leads to the counterpart of Equation
6-3, namely:
(6-5)
where:
YNET = 9 total VSS produced/g COD removed (feed
total COD minus effluent SCOD)
qoes = 9 COD removed/g total VSS/d
The value for YNET in Equation 6-5 (i.e., based on total
VSS) is dependent on GO and the nature of the wastewater
being treated (i.e., raw, primary, or secondary wastewa-
ter). The values presented in Figure 2-8 may be used for
design in most cases, provided the wastewater does not
contain a large industrial component. If site specific data
are available, they should be used to establish the yield
coefficients. The required hydraulic retention time, t, can
then be calculated from the definition of q0Bs. as derived
from an alteration of Equation 3-19:
QOBS -"
Xt
(6-6)
where:
S0 = feed or influent total COD, mg/L
S1 = effluent soluble COD, mg/L
One must define an allowable level of MLVSS or X in
order to use Equation 6-6. The allowable level of mixed
liquor total suspended solids (MLTSS), and therefore X,
is influenced primarily by the efficiency of the solids-liquid
separation step (Section 6.4.10.5).
The observed ammonium oxidation rate (rN), although not
pertinent to design by this procedure, can be calculated
by modifying Equation 3-21 as follows:
NO-|
Xt
(6-7)
The excess sludge wasting requirements can be calcu-
lated from the definition of the solids retention time ac-
cording to Equation 3-12. The specific design examples
presented in Sections 6.4.2.2 and 6.4.3.2 will illustrate in
detail the use of the solids retention time design approach.
6.4.1.2 Alternative Design Approach
An alternative to the solids retention time design ap-
proach relies on the use of an ammonium oxidation rate,
which is the ammonium oxidized per unit time at a specific
VSS level in the system (g NHJ-N oxidized per day per
g MLVSS). The ammonium oxidation rate is equivalent to
the maximum rate if the rate is zero-order with respect to
ammonium concentration, as previously discussed in
Chapter 3 (Section 3.3.6). That is:
=
(3-20)
Determining the maximum oxidation rate effectively de-
fines the minimum solids retention time, 6™, through the
relationship:
e?=:
1
(3-23)
Often designers that use the ammonium oxidation rate
approach have simply selected a nitrification rate and
used that value to determine t for an allowable level of
total mixed liquor VSS. This, in effect, is equivalent to
using the observed ammonium oxidation rate, rN, as de-
fined by Equation 6-7. This approach is acceptable when
it is based on rate information that is derived specific to
the wastewater in question; the observed maximum nitri-
fication rate relative to the actual, or true, maximum rate
will reflect the level of active Nitrosomonas as VSS (XN)
as a fraction of the total VSS (X) in the reactor. That is,
135
-------
A
the observed maximum rate, rN, is related to true maxi-
mum rate, qVi, by the following equation:
r =
(6-8)
where:
f = fraction of active Nitrosomonas VSS in the
reactor total VSS
It is clear from Equation 6-8 that the fraction of nitrifiers
present in the reactor has a marked effect on the ob-
served nitrification rates. As has been discussed in Chap-
ter 3 (Section 3.3.6), the fraction of nitrifiers present in
the mixed liquor will strongly depend on the wastewater
feed C:N. The influence of this ratio on nitrifier fraction
and nitrification rates was recognized as early as 1940
(24). The wide range in reported nitrification rates which
is demonstrated by Figure 6-3, is likely caused in part by
differences in the wastewater C:N.
The limitations in using the ammonium oxidation or nitri-
fication rate approach to size the nitrification reactor are
obvious from the preceding discussion. There is an ad-
vantage to using the rate approach only when information
on the site-specific rate is available, ideally through pilot
studies. The approach, which is basically the same as
the solids retention time approach, can also be used as
a check against the solids retention based sizing to verify
that the rate information is reasonable.
Once an observed ammonium oxidation rate has been
determined, a design rate value can be established from
the maximum observed rate, accounting for the effects of
the required effluent ammonia-N concentration and other
factors through the use of a design factor. The design
rate can then be used to establish t for an allowable level
of X. Fixing t and X establishes the organic removal rate
that will be observed (Equation 6-6), which, in turn, fixes
the operating 60 for a given observed total VSS yield
(Equation 6-5).
6.4.2 Complete Mix Systems
Many activated sludge systems are designed to operate
on the complete mix principle. When designed in this
fashion the system is often referred to as the complete
mix activated sludge or CMAS process. An example of
the feed and withdrawal arrangement for a complete mix
plant may be found on Figure 6-2. The complete mix
design provides uniformity of load to all points within the
aeration tank, easing the problems of oxygen transfer that
can be encountered at the head end of conventional
plants, and making this configuration attractive for han-
dling organic shock loads. The contents of the complete
mix reactor are homogenous and thus the mixed liquor
feed to the clarifier is identical to the mixed liquor through-
out the reactor. Complete mix systems can be prone to
"bleed through" or short circuiting of the feed to the efflu-
ent, particularly during peak flows.
6.4.2.1 Design Approach
The solids retention time design approach presented in
Section 6.4.1.1 can be directly applied to the design of
complete mix activated sludge systems, using Equation
3-10 (Section 3.3.1) to calculate the actual steady state
ammonium content of the effluent.
^
w 0.4-
1
2 0.3-
1
i
o»
«T 0.2'
1
cc
J
§
1 0.1-
i <
5
Blue Plains, DC BODS:TKN = 1,3© (25)
Blue Plains, DC BOD5:TKN = 5.00 (26)
CCSD, CA BOD5:TKN = 2.4 •(15)
BODs:TKN = 1.8 ±(6)
BODS:TKN =3.0, pH 7.0 (27)
G0 ©
©
S3 0
0
©3
1 © 1
•
® © S _
® 0 © 0
© ® Q
§ .^B- ' "
A ^
J ^ g1 Q S 0 Q @
0 ' 1
1 1 1 1 1 1 1 1 1 1 1 —
Q
©
Q Q
xx \y
Q
Q
_ 1
14 15 16 17 18 19 20 21 22 23 24 25 26 27
Temperature, °C
Rgure 6-3. Variation In observed nitrification rates.
136
-------
6.4.2.2 Design Example No. 1: Nitrification in a Complete Mix Suspended Growth
System at High Carbonaceous Feed Concentration
The following design example presents the solids retention time design approach for the example "complex"
Plant B. The calculations will address sizing requirements to meet the more stringent effluent limits. The
design conditions and effluent limits have been presented in Chapter 2 (Section 2.9 and Table 2-10). Some
of the design information from Chapter 2 is also summarized in Table 6-2. Design calculations are demon-
strated at 15°C. The results at this temperature and at 10°C and 20°C are also presented for discussion
purposes. A simplified process schematic is presented in Figure 6-4.
Table 6-2. Design Conditions for Example 1: Plant B in a Complete
Mix Configuration with Higher Carbonaceous Feed and More Strin-
gent Effluent Requirements
Wastewater Flow Characteristics, ms/d (mgd)
Raw wastewater average flow
Total influent average flow
Primary Effluent Characteristics, mg/L
COD
CBODg
TSS
TKN
Total P
Alkalinity, mg/L (as CaCO3)
Final Effluent Limits, mg/L
18,925 (5.0)
21,055 (5.56)
187a
97a
80a
29.5a
6.0a
168"
87b
72b
26.6b
5,4"
120b
CBODS
TSS
NHJ-N
Total N
Total P
Design Conditions/Assumptions
Reactor temperature, °C
Reactor, MLVSS, mg/L
Reactor minimum DO, mg/L
Reactor pH range
Reactor and effluent VSS/TSS
10
10
2
• 5
'•' 1
15
1,400
2.5
7.0-7.6
0.68°
a Concentration values at average conditions expressed as mg/L equivalents (see
Table 2-16).
b Concentration value at average conditions at total flow of 21,055 m3/d.
c Does not include impact of mineral addition for phosphorus removal.
Complete Mix
Aeration Tank(s)
Primary Effluent
with Recycles
Effluent
Return Sludge
Excess
Sludge
Figure 6-4. Simplified schematic for Design Example No. 1 employing a complete mix suspended growth
reactor for nitrification.
137
-------
6.4.2.2 Design Example No. 1 (continued)
1. Determine the process design factor to be used in sizing the biological reactor. Following from Chapter 2
(Table 2-18), the peaking factor to be used in this example to account for influent loading variations is 1 .56.
Assume that the effluent quality requirement, the anticipated variations in process conditions, and the
uncertainty in the kinetic coefficients warrant a safety factor of 2.0. Compute an overall process design
factor of 3.1 based on the product of PF and SF.
2. Verify that the reactor operating pH can be expected to be in the assumed range of 7.0 to 7.6 (Table
6-2), Alkalinity destroyed can be calculated from the amount of ammonia nitrogen assumed to be oxidized
and from the coefficient presented in Table 3-1. From callout 6 in Table 2-16, the mg/L equivalent soluble
nitrogen was 25.7 mg/L. Allowing for about 1 mg/L equivalent of soluble refractory nitrogen, the actual
concentration of oxidizable nitrogen in the secondary effluent may be estimated as 24.7/1 .09955 or 22.5
mg/L. Even though the effluent limit is 2 mg/L of NHJ-N, the system will normally produce much lower
residual effluent NHJ -N concentrations. Therefore, assume the alkalinity destroyed is given by:
(7.1 mg CaCOg/mg N oxidized) (22.5 mg/L) = 160 mg/L (as CaCO3)
Alkalinity available (Table 6-2) is 120 mg/L (as CaCO3). From Chapter 2 (Table 2-3), and assuming that an
air oxygen transfer system with only 12-percent efficiency is used (see Step 11), it is reasonable to assume
that the pH will remain in the required range (Section 6.4.10.3), provided a minimum residual alkalinity of
50 mg/L (as CaCO3) is maintained. On this basis, the required supplemental alkalinity will be:
(160 mg/L destroyed + 50 mg/L minimum residual - 120 mg/L available)
= 90 mg/L (as CaCO3) to be supplemented
The mass of alkalinity required under average day conditions will be:
= 1,895 kg (4,180 lb)/d (as CaCO3)
The design example peaking factors presented in Table 2-12 can be used to determine alkalinity require-
ments under other loading conditions. For example, the nitrogen mass increases by a factor of 1 .7 on the
maximum day with a corresponding increase in matching alkalinity of only 1.5. Influent flow increases by
a factor of 2.5, and, for this example calculation, it is assumed that return flows to the head of the plant
also increase by this factor. Thus the mass of alkalinity required on the maximum day can be estimated as
follows:
Flow = 2.5 (21,055) = 52,638 m3/d
Influent NH|-N = (1.7/2.5) (22.5) = 15.3 mg/L
Influent alkalinity = (1 .5/2.50)(120) = 72 mg/L as CaCQ3 •
Supplemental alkalinity = [(7.1)(15.3) - 72 + 50] [52,638/103] = 4,560 kg (10,050 lb)/d as CaCO3
Additional alkalinity considerations where metal addition (Al+3 or Fe+3) is used for phosphorus removal are
presented in Step 12.
3. Calculate the maximum nitrifier growth rate under the nonlimiting DO and pH conditions stated (Sec-
A
tions 3.3.3 and 3.3.4). Equation 3-14 should be used to calculate u.N under nonlimiting substrate conditions
(i.e., KN <6 N):
£N = 0.47[ea098] ••. . (6-9)
AtT= 15°C (Table 6-2): ,,
4. Calculate the minimum solids retention time for nitrification. From Equation 3-13 with bN = 0, the correct
expression is:
138
-------
6.4.2.2 Design Example No. 1 (continued)
fim_J_ (6-10)
°C ~ A
For this example:
5. Calculate the design solids retention time according to:
eg = (Process design factor) eg1 (6-11)
For this example:
eg = 3.1 (2.13) = 6.6 d
6. Calculate the design nitrifier growth rate. From Equation 3-13 with bN = 0, the correct expression is:
' =J_ (6-12)
For this example:
7. Calculate the steady state ammonium content of the effluent. Equation 3-10 is directly applicable to
complete mix activated sludge systems, where Nt is the effluent or reactor ammonium-nitrogen content:
A N1 (6-13)
^N = MN K f.
KN + N!
where: N! = effluent NHJ-N, mg/L
Select a value of 1.0 for KN, according to the recommendation in Section 3.3.2. Equation 6-10 becomes:
therefore: N^ = 0.48 mg/L
8. Calculate the organic removal rate. Following from Chapter 2 (Figure 2-10), at eg of 6.6 days, YNeT is
approximately 0.25 g total VSS produced/g COD removed. Therefore, for this example:
ftd- v
wc "NET
qOBs = (R JL OR, = 0.606 g COD/g MLVSS/d
Where plant data are sufficient to develop a relationship between YNET and 60, these data should be used
in place of Figure 2-10. If the municipal wastewater contains an influent composition where the ratios of
degradable VSS to inert VSS are substantially different than used in the example, the net yield should be
calculated for the wastewater in question (see Table 2-13). Finally observed yields are temperature sensitive,
(28) and this additional factor can be considered when calculating net yields. Given the many uncertainties
normally existing with regard to influent wastewater composition and flows, the simple procedure used in
this example will prove adequate for many cases.
139
-------
6.4.2.2 Design Example No. 1 (continued)
9. Determine the reactor hydraulic retention time and the reactor volume at total influent average flow. From
Equation 6-6, the correct expression for t is:
Sp-S, (6-15)
Using the values from Table 6-2 for S0 and X, and an estimate of 18 mg/L for Si (Table 2-16, callout 6)
gives:
The reactor volume can be calculated from:
Volume = Qt (6-16)
where: Q = total influent average flow rate, m3/d
For this example: ;
Volume = (21 ,055)(0.176) = 3,710 m3 (131 ,000 cu ft)
10. Determine the sludge wasting requirements. Sludge is wasted from the system as solids contained in
the effluent from the secondary clarifier and intentionally from the reactor or the return sludge stream/The
total sludge to be wasted under steady state conditions can be calculated from the definition of the solids
retention time (Equation 3-12) expressed for a suspended growth reactor as:
fld "A (6-17)
ec=-g
where:
IA = inventory of VSS under aeration, kg
S = total VSS wasted, kg/d
The inventory of VSS under aeration can be calculated according to:
i XV (6-18)
A~103
where: V = volume of aeration tank, m3
For this example:
, (1,400)(3,
=5,190 kg (11,440 Ib) VSS
Using Equation 6-17 and 0Jj! of 6.6 days, the total VSS to be wasted is:
S = - = 786 kg (1 ,1 730 Ib) VSS/d
D.O
The sludge contained in the effluent will depend upon the efficiency of the secondary clarifier. For purposes
of this example, assume that the effluent TSS are estimated as 14 mg/L and the VSS:TSS ratio in the
reactor is 0.68 (without chemical addition for P removal). The VSS contained in the effluent is:
21,055(14)(0.68)
10
3 = 200 kg VSS/d (442 Ib/d)
140
-------
6.4.2.2 Design Example No. 1 (continued)
Therefore, the sludge to be wasted from the mixed liquor or return sludge is:
786 - 200 = 586 kg (1 ,290 Ib) VSS/d
586/0.68 = 862 kg (1 ,900 Ib) TSS/d
Where there is supplemental chemical addition to the mixed liquor for phosphorus removal (see Step 12),
the sludge wasting calculations and estimated VSS:TSS must also include the impact of the additional inert
solids. Solids-liquid separation considerations (Section 6.4.10.5) will dictate clarifier design including the
return sludge rate and concentration. The mass of sludge to be wasted can then be calculated for average
and maximum month conditions.
11 . Determine aeration requirements. Oxygen requirements for ammonium oxidation can be calculated from
the coefficient presented in Table 3-1. Oxygen requirements for CBOD oxidation are dependent upon the
degree of biomass oxidation as illustrated in Figure 2-10. The COD oxidized can be computed as the
difference between the influent COD and the COD discharged from the system. The discharged COD is
the sum of the soluble effluent COD and the COD of the VSS exiting the system. At the 6.6-day design
6G, the COD in the effluent VSS can be estimated as:
(786 kg VSS/d)(1.42 kg COD/kg VSS) = 1,120 kg (2,470 Ib) COD/d
Carbonaceous oxygen requirements are:
21 ,055(168 - 1 8) _ 1 112(? = 2;04Q kg (4>5QO |b) Og/d
An alternative "rule-of-thumb" methodology to estimate carbonaceous requirements is to select a coefficient
value of from 1 .0 to about 1 .3 (for high^ systems) times the BOD5 removal (influent minus soluble effluent).
For a BOD5 removal of 85 mg/L (Table 6-2 and assuming a soluble effluent BOD5 of 2 mg/L) and a total
carbonaceous requirement of 2,040 kg O2/d, the equivalent coefficient value would be 1.14.
Oxygen required for ammonium oxidation (ignoring the small effluent NHJ-N residual) is:
21.055(22.5)(4.6) =
Total oxygen required under average conditions is:
2,040 + 2,180 = 4,220 kg (9,300 Ib) O2/d
= 47.4 mgO2/hr/L of aeration tank
II
The design example peaking factors presented in Chapter 2 (Table 2-12) can be used to determine oxygen
requirements under other conditions. For example, under maximum day conditions the peak oxygen re-
quirement will be approximately:
2,040(2.1) + 2,180(1 .7) = 7,990 kg (17,610 lb)O2/d
(7,990/3,71 0)(1 03/24) = 89.7 mg O2/hr/L of aeration tank
The efficiency of the oxygen transfer system will dictate the amount of air required to satisfy the total oxygen
demand. The oxygen transfer efficiency depends on a number of factors such as type of aeration device,
reactor depth, alpha value, and operating DO level. An oxygen transfer efficiency of 10 percent requires
141
-------
6.4.2.2 Design Example No. 1 (continued)
approximately 36.2 m3of air/kg O2 transferred (580 cu ft/lb O2). For this example, assume an overall transfer
efficiency of 12 percent. Therefore, the aeration capacity required at average conditions is:
The aertion requirement for the maximum day will be:
More detailed information on aeration requirements and design for suspended growth reactors can be found
in Section 6.4.10.3.
12. Determine impact of phosphorus removal on design. When phosphorus removal is required, the design
must address whether a system with anaerobic and aerobic zones would be appropriate. Phosphorus
removal can be obtained by relying on chemical addition or by using chemical addition as needed to
supplement an enhanced biological removal process. Effluent filtration may also be required.
Chemical addition affects the design approach with regard to alkalinity requirements (Step 2) and the
production of additional SS. The additional SS can affect aeration basin and clarifier sizing and will increase
waste solids production and sludge handling requirements. Information specific to this issue is provided in
Sections 6.4.10.2 and 2.7.2 and also Table 2.3 as well as in other references (2,29,30).
For the example problem, the influent mass of phosphorus is 113.6 kg/d and approximately 15.7 kg/d [786
kg VSS x 0.02] is associated with the waste VSS. Addition of FeCI3 or AI2(SO4)3 • 14 H2O could be selected
for the required P removal.
In calculating additional solids production and alkalinity losses by metal addition, there are two major
pH-dependent competing reactions as illustrated by the following reactions for aluminum.
1) Al+3 + HCO5 => AI(OH)3 + 3 CO2
2) Al+3 + PCT3 =» AIPO4
For the first reaction, there are 5.6 mg CaCO3 alkalinity lost per mg Al reacting and 2.9 mg of AI(OH)3
formed per mg Al reacting. For the second reaction, there is 3.9 mg AIPO4 formed per mg P removed. To
achieve the effluent limit of 1 mg/L total P will probably require from 1 .25 to 1 .75 moles of Al per mole of
P. This chemical addition could be split among the primary clarifier, the biological reactors (see Step 16)
and the effluent filters.
An estimate of solids production at 1 .5 moles Al/mole P follows:
From 1: (1.5 - 1.0)(27/31)(2.9)(113.6 - 15.7) = 124 kg (273 Ib) AI(OH)g/d :
From 2: 3.9 (113.6 - 15.7) = 382 kg (842 Ib) AlPCyd
Total inert solids production equals 506 kg additional TSS/d. Alternatively, one can more easily estimate
the additional inert solids production from the coefficient in Table 2.3 as follows:
(6.1)(113.6 - 15.7) = 597 kg (1,320 Ib) TSS/d
The coefficient 6.1 assumes somewhat different stoichiometry than represented by Reactions 1 and 2 above,
but is an equally acceptable approach for estimation purposes.
142
-------
6.4.2.2 Design Example No. 1 (continued)
Since e£ is 6.6 days, the reactor will carry an additional (6.6)(506) = 3,340 kg (7,365 Ib) inert TSS or an
additional 900 mg/L of additional TSS in the aeration tank. The,total MLSS concentration will be approxi-
mately
1,400
0.68
+ 900 = 2,960 mg TSS/L
This solids estimate should be used to revise the waste sludge production previously presented in Step 10
(which did not consider the impact of alum addition). - ,—
The additional alkalinity depletion from the alum addition is:
(0.5X27/31 )(113.6-15.7)(5.6) = 239 kg (527 Ib) as CaCOg/d
This additional alkalinity loss was not included in Step 2 and would need to be considered when computing
supplemental alkalinity requirements.
13. Evaluate at alternative temperatures (Table 6-3).
Table 6-3. Summary of Example 1 Design Results: Plant B at Higher Carbonaceous Feed
Concentration and More Stringent Effluent Requirements
Value as a Function of Temperature
Design Parameter
ft* d'1
ecm, d
60, d
UN, ^1
N,, mg/L
YNET, g VSS/g COD
t, hr
Reactor volume, m3
VSS inventory (IA), kg
VSS wasted (S), kg/d
Average carbonaceous O2 demand, kg/d
Average total O2 demand, kg/d
10°C
0.288
-' 3'47 "~,
10.8
0.093
0.48
0.22
6.1
5,630
7,500
694
2,170
4,350
15°C
0.47
2.13
6.6
0.152
0.48
0.25"
' ' 4.2
3,710
. 5,190
; 786
2,040
. 4,220
,20°C
0.767
,' 1.30
4.0
0.25
' 0.48
0.30
3.1
2,710
3,790
948
1,810
.3,990
Average O2 demand, mg/L of aeration tank/hr
33.8
47.4
, 61.3
14. Consider evaluating the chosen design with one of the dynamic models discussed in Chapter 5 to
investigate the actual dynamic response in ammonium nitrogen levels .under, the plant diurnal loading
conditions.
15. Consider the design calculations as just one component is sizing Jhe reactor system. Other factors, as
discussed in Chapter 2, are equally important. In some instances, especially for small facilities, prudent
design anticipates that administrative and/or operator support may not be sufficient to ensure that the design
assumptions are fully realized. These constraints can be partially or fully accommodated with larger systems
than dictated solely by kinetic considerations. Thus for the design example, an SRT of from 10 to 15 days
could also be more appropriate when other intangible factors affecting plant operation are also evaluated.
16. To achieve the final effluent limit of 5 mg/L TKN specified in Table 6-2, this design approach requires
the addition of a separate stage denitrification system to reduce the NOg-N. Design of suitable separate
stage systems is discussed in detail in Chapter 7.
143
-------
6.4.3 Conventional or Plug Flow Systems
Conventional systems consist of a series of rectangular
tanks or passes (Figure 6-2) in which the ratio of total
tank length to width is typically between 5 and 10. The
hydraulics of this configuration have been loosely termed
as plug flow because the influent wastewater and return
activated sludge are returned to the head end of the
process and the combined flow must pass along a long
narrow aeration tank prior to exiting the system. The de-
gree to which the process reactors actually approach plug
flow is dependent on the amount of longitudinal mixing.
While complete mix activated sludge systems are typical
for a wide range of industrial application, plug flow hy-
draulic conditions are favored in the design of municipal
wastewater nitrification reactors. As will be shown in Sec-
tion 6.4.3.1, plug flow conditions result in a lower effluent
ammonium concentration than a complete mix condition
at the same design factor, or alternately, the same am-
monium level at a lower design factor.
High-purity oxygen and powdered activated carbon nitri-
fication reactors can be designed as two or more com-
pletely mixed tanks in series or as single tanks with a
number of passes, thus incorporating hydraulic conditions
equivalent to conventional plug flow activated sludge
systems.
6.4.3.1 Design Approach
The approach to designing conventional activated sludge
systems is similar to that for complete mix systems except
for the equations used to predict effluent quality. A plug
flow model may be applied to approximate the hydraulic
regime in these systems, and the Monod expression for
Nitrosomonas growth rate (Equation 3-10) must be inte-
grated over the period of time an element of liquid re-
mains in the nitrification reactor. The following is a
solution for plug flow kinetics that can be adapted to this
problem for conditions in which the sludge recycle ratio
is less than 1.0 (23):
A
(6-19)
Reference 23 should be consulted to determine the form
of Equation 6-19 for recycle ratios greater than 1.0.
Equation 6-19 is evaluated in Figure 6-5 for Design Ex-
ample No. 1, presented in Section 6.4.2.2. This is shown
as a function of the process design factor, DF, recognizing
that from Equation 6-11:
DF = ^ (6-20)
3.5
3.0-
0.5-
0.0
M.N = 0.47 d"'
No = 22.5 mg/L
KN = 1.0mg/L
i i I i i i i i i i i i
1.0 1.4 1.8 2.2 2.6 3.0 3.4
Process Design Factor (DF)
Rgure 6-5. Effect of design factor on steady state effluent ammonia levels in complete mix and plug flow suspended
growth reactors.
144
-------
Lower values of the process design factor can be used
for plug flow nitrification reactors to produce the same
theoretical effluent ammonium concentration found in a
complete mix reactor (Figure 6-5). This means that plug
flow processes are theoretically more efficient at the
same process design factor, or alternatively, require less
aeration tank volume for the same level of nitrification
efficiency.
However, plug flow type reactors also have a potential
disadvantage in that the carbonaceous oxygen demand
is concentrated at the head end of the tank, sometimes
making it difficult to supply enough air in that area for
both carbonaceous oxidation and nitrification. Air diffusion
systems must be specifically designed to handle this con-
centrated load in higher feed CBOD5 systems; otherwise,
the first portion of the tank will not be available for nitrifi-
cation because of inadequate levels of DO. Atypical DO
and nitrification pattern for plug flow tanks in which aera-
tion capability is limited in the front end of the tank is
presented in Figure 6-6. As can be seen from the aeration
tank DO and ammonium nitrogen profile, nitrification is
inhibited in the first portion of the tank because of the DO
suppression that has been incurred for carbonaceous oxi-
dation. Once the DO rises, the ammonium level falls at
a reaction rate that approximates zero order, a reaction
order predicted by kinetic theory (Section 3.3.1). It is
notable that if sufficient aeration capability had been avail-
able in the head end of the tank, virtually complete nitri-
fication probably would have been obtained. If oxygen
supply limitations are present in the head end of the tank,
the plug flow type reactor's advantage over the complete
mix reactor is reduced. The tapered aeration process
modification to reactors designed with plug flow hydrau-
lics represents a solution to this problem (Section 6.4.6).
The degree to which nitrification reactors approach plug-
flow operation can be examined through dispersion the-
ory (32,33). Reactors can be characterized by a
dimensionless axial dispersion number, D/uL, where D is
the axial dispersion coefficient in m2/hr, u is the mean
displacement velocity along the tank length in m/hr and
L is the tank length, m. In the calculation of the axial
dispersion number, u and L are known for any particular
plant design and D must be measured. An acceptable
approximation of D for both fine and coarse bubble dif-
fused air plants can be obtained from the following rela-
tionship (33).
= 3.118W2(A)a35
(6-21)
where:
W = tank width, m
A = air flow per unit tank volume, in standard
m3/min/1,000 m3
The axial dispersion coefficient, D, is zero for true plug
flow plants and infinite (°o) for true complete mix plants.
Plants with a dispersion number, D/uL, less than 0.2 are
usually classified as plug flow reactors, while for complete
Aeration Tank
Baffles
25
20
I
10
o—
I
I
2.5
2.0
1.5
1.0
0.5
50
100 150
Distance Along Tank, ft
200
250
Figure 6-6. DO and ammonium-nitrogen profile in a plug flow system (from Reference 31).
145
-------
mix systems, D/uL is typically greater than 4.0 (34). As
an example calculation, assume a system's four-pass
nitrification tanks have the following characteristics:
Air flow = 51.1 standard m3/min/1,000 m3
Width = 10.7 m (35.1 ft)
Depth = 4.5 m (14.8 ft)
Cross-sectional area of tank = 48 m2 (525 sq ft)
Total length = 329 m (1,080 ft)
Flow each tank (four passes) @ 50% recycle =
85,200 m3/d (22.5 mgd)
From these data, the mean displacement velocity is cal-
culated to be 74.0 m/hr (243 ft/hr). From Equation 6-21,
the dispersion coefficient is:
D » 3.12(10.7)2(51.1 )°-35 = 1,415 m2/hr (15,234 sq ft/hr)
and
D/uL = 0.058
At 0.058, D/uL is significantly less than 0.2; thus, the
system's nitrification tanks closely approach plug flow hy-
draulics. Equation 6-21 can be utilized to evaluate mixing
in actual plant designs. If they closely approach true plug
flow, Equation 6-19 can be used to describe nitrification.
It is probable that most plants operated in the conven-
tional mode do approach plug flow. For those plants with
intermediate values of D/uL, complete-mix kinetics can
be employed, which will yield conservative answers.
The hydraulic configuration of nitrification tanks can also
be designed to discourage back mixing by employing a
series of complete mix tanks. Absolute prevention of back
mixing can be achieved by the use of mixed liquor over-
flow weirs between reactors.
6.4.3.2 Design Example No. 2: Plant B—Separate Stage Nitrification at Low
Carbonaceous Feed Concentration
This design example is based on information provided in Section 2.9 for the more complex Plant B. In this
example it is assumed that the environmental conditions and the solids retention time in the existing
biological reactor are such that only CBOD removal is accomplished. Consequently, a second stage
biological reactor is required for nitrification of the secondary effluent. The design information utilized is
summarized in Table 6-4.
Table 6-4. Design Information for Nitrification of a Low Carbonaceous Feed
Concentration
Wastewater Flow Characteristics, m3/d (mgd)
Raw wastewater average flow
Total secondary effluent average flow
Unsettled Secondary Effluent Concentration8
Total COD, mg/L
Nitrogen available for nitrification, mg/L
Soluble COD mg/L
Alkalinity, mg/L (as CaCO3)
Nitrification Reactor Effluent Characteristics
Soluble, COD, mg/L
VSS, mg/L
Ammonium nitrogen, mg/L
Design Conditions/Assumptions
Reactor temperature, °C
Reactor MLVSS, mg/L
Reactor minimum DO, mg/L
Reactor pH range
18,925 (5.0)
21,055 (5.5)
23. 1b
30b
20b
9b
2.0d
15
1,400
2.5
7.0-7.6
101°
21°
27°
120°
18°
8°
•See Processing Point 4, Table 2-16.
b Concentration values at average conditions expressed as mg/L equivalents.
0 Concentration values at the total flow of 21,055 m3.
" For effluent objective, see Table 2-10.
146
-------
6.4.3.2 Design Example No. 2 (continued)
The design procedure is similar to that used for treating a high carbonaceous feed in a complete mix
suspended growth reactor (Section 6.4.2.2). Plug flow hydraulic conditions in the nitrification reactor and
the level of pretreatment (i.e., primary clarification and a first stage biological reactor) may warrant selection
of an SF lower in value than that used in the complete mix reactor design example. Also the. peak load
may be attenuated in the first stage biological reactor depending upon reactor hydraulic characteristics (i.e.,
degree of mixing and hydraulic retention time, HRT). If an SF of 1.5 is selected, and the process design
factor is computed based on the product of PF and SF, the design solids retention time becomes:
= 1. 56(1 .5)(2. 1.3) = 4.98 d
The corresponding u,'N calculated from Equation 3-13 is 0.20 d"1. The steady state ammonium content for
the plug flow reactor conditions can be calculated from Equation 6-19 to be less that 0.1 mg/L For a
complete mix reactor, effluent NHJ-N would be 0.74 mg/L. Calculating the organic removal rate, hydraulic
retention time and reactor volume require specification of the observed yield, YNFT- The fact that the CBOD
or COD from the first stage reactor is largely due to biological solids makes it more difficult to estimate
YNET-
The VSS concentration in the second stage nitrification system will be largely influenced by the efficiency
of the secondary clarifier (unless some primary effluent is directly bypassed to the nitrification second stage
or unless some other source of VSS such as waste secondary sludge is added to the nitrification reactor).
Assume that the secondary system operates at an SRT of 2 days and the residual effluent degradable
SCOD which will be removed in the nitrification reactor at the 5-day Q$ is approximately 10 mg/L. The
MLVSS in the nitrification reactor include the following components:
• Inert VSS which are in the entering wastewater including inert paniculate biomass decay products
produced in the secondary system.
• Active heterotrophic biomass entering the system plus a small amount grown on the entering biodegrad-
able COD.
• Inert decay products produced in the nitrification reactor from the entering and synthesized active
biomass.
• Autotrophic microorganisms.
If all entering nitrogen available for nitrification (21 mg/L) was converted to NO^-N, the net autotrophic yield
would only be 2.1 mg of nitrifier VSS (Table 3.1) per L of entering wastewater. Ten mg/L of SCOD removal
at a 5-day 6C in the nitrification reactor would produce an additional heterotrophic net yield, YNET, of only
2.5 milligrams per liter of entering wastewater (Table 2-13, IAWPRC model). The sum of the new heterotro-
phic and nitrifier VSS from biomass growth, 4.6 mg/L, is less than the projected 8 mg/L of effluent VSS
from the nitrification system. Hence, the solids balance in the nitrification reactor will be largely controlled
by the amount of VSS entering from the secondary clarifier (or added from other sources as described in
Section 6.3).
Assume that the secondary effluent VSS are 20 mg/L. A simple way to estimate change of this material in
the nitrification reactor is to use the classical approach in Table 2-13. For each liter of entering wastewater
and at the 5-day 9jj! the 20 mg/L of entering VSS will be reduced to:
Therefore, the total solids balance accounting for new growth and biomass decay per liter of entering
wastewater is:
X = 4.6 + 15.4 = 20 mg/L VSS
147
-------
6.4.3.2 Design Example No. 2 (continued)
It is also possible to use the IAWPRC approach in Table 2-13 to estimate the change in the entering VSS
components. Whether or not this extra effort is justified depends on the level of knowledge of the various
VSS components in the secondary system and the degree of confidence in the estimate for the secondary
effluent VSS levels.
The net VSS production is 20 mg/L of entering wastewater, the design MLVSS level is 1,400 mg/L and the
60 is 5 days; the required hydraulic retention time can be computed based on the definition of SRT (i.e.,
90 = mass of solids under aeration divided by the mass of solids wasted per day) as follows:
1,400 =
20(5)
t = 0.071 d = 1.71 hr
Reactor volume = 21,055(0.071) = 1,504 m3 (53,106 cu ft)
IA = 1,400 (1,504)7103 = 2,106 kg (4,640 Ib) VSS
S = lA/eg = 2,106/5 = 421 kg (930 Ib) VSS/d
Effluent VSS = 21,055(8)7103 = 168 kg (370 Ib) VSS/d
Sludge to be wasted = 421 - 168 = 253 kg (560 Ib) VSS/d
Determine the aeration requirements:
Oxygen required for ammonia oxidation:
(21.0)(4.6)(21,055/103) = 2,034 kg (4,480 Ib) O2/d
Assuming 1.42 kg COD/kg VSS, carbonaceous oxygen requirements can be estimated as:
(10 - 2.5(1.42) + 1.42(20 - 15.4))(21,055/103) = 273 kg (602 Ib) O2/d
Total oxygen requirements average 2,307 kg (5,087 Ib) O2/d. This is a mean oxygen requirement of 64
mg/hr/L of aeration tankage. Given that the PF for the maximum monthly load is 1.56, the oxygen demand
for the plug flow reactor will increase to 100 mg/hr/L of aeration tankage under these conditions. As in Case
1, the designer may elect to increase the aeration tank volume to reduce volumetric oxygen demands to a
level compatible with the desired aeration equipment.
6.4.3.3 Performance Information
River Oaks, Hillsborough County, Florida
The River Oaks Advanced Wastewater Treatment Plant
is located in Hillsborough County, Florida, discharging
ultimately to Tampa Bay. In the mid-1980s, the plant's
capacity was increased from 132 Us (3 mgd) to 440 US
(10 mgd) and the plant was upgraded to achieve total
nitrogen removal (35). The configuration of the plant's
aeration tanks results in a conventional plug flow system,
as opposed to complete mix. A high carbonaceous feed
concentration is imposed at the head end of the aeration
tanks, following primary treatment. Carbonaceous oxida-
tion and nitrification are accomplished in three separate,
fine pore, diffused aeration tanks operating in series. A
separate sludge reaeration tank is included as part of the
system and sodium aluminate is added prior to clarifica-
tion of the nitrified reactor contents to accomplish phos-
phorus removal. In addition to handling 440 Us (10 mgd)
of primary treated wastewater, the nitrification system-is
designed to handle filter backwash and recycle water
from the downstream unit operations and return flow from
the equalization basin, installed to accommodate exces-
sive primary effluent during wet weather conditions. Rele-
vant design information for the River Oaks treatment plant
is summarized in Table 6-5. '
The nitrification portion of the system must achieve less
than 0.5 mg/L NHJ-N for the plant to meet its final effluent
total nitrogen limits. Operating and performance informa-
tion pertaining to the nitrification system over a 12-month
period from August 1988 to July 1989 is presented in
Tables 6-6 and 6-7, respectively. The excellent nitrification
performance of this Florida plant is not surprising in7light
of the very favorable year-round climatic conditions, and
148
-------
Table 6-5. River Oaks Advanced Wastewater Treatment
Plant—Design Information for Carbonaceous Oxidation/Ni-
trification System Following Primary Clarification (Adapted
from Reference 35)
Total Design Flow
Average influent flow
Filter backwash and recycle water
Return flow from equalization
Carbonaceous Oxidation/Nitrifica-
tion Reactors
(Secondary Treatment)3
Number of aeration tanks
Number of reaeration tanks
MLSS, mg/L
MLVSS, mg/L
t, hr
9, d
F/M, kg BODg/kg MLVSS/d
Secondary Sedimentation Tanks"
Number
Hydraulic loading rate, m3/m2/d
(gpd/sq ft)
Solids loading rate, kg/m2/d
(Ib/sq ft/d)
50,000 m3/d (13.2 mgd)
37,800 m3/d (10 mgd)
4,500 m3/d (1.2 mgd)
7,600 m3/d (2 mgd)
3
1
3,270
1,750
3.7
5.0
0.3
3
25 (620)
122 (25)
a Phosphorus removal accomplished by sodium aluminate addition to
the nitrification system. F/M represents food-to-mass ratio. Stated val-
ues were calculated at total design flow.
b Hydraulic and solids loading rates to secondary sedimentation tanks
based on total design flow.
Table 6-6. River Oaks Advanced Wastewater Treatment
Plant—Carbonaceous Oxidation/Nitrification System Oper-
ating Conditions August 1988 to July 1989 (Adapted from
Reference 35)
Nitrification System Value*
Parameter
Average
Range
MLVSS, mg/L
MLTSS, mg/L
6, d
F/M, kg BODg/kg MLVSS/d
Temperature, °C
Clarifier loadings
Hydraulic, m3/ma/d
(gpd/sq ft)
Solids, kg/m2/d (Ib/sq ft/d)
2,550
4,350
16
0.12
27
14.8 (364)
125 (25.7)
1,600-3,350
2,750-5,650
8-25
0.09-0.15
•24-30
13.4-18.1
(330-445)
76-198
(15.5-40.6)
* Monthly values.
low loading conditions experienced during this period.
The average monthly temperature did not fall below 24°C
(75°F), and the average solids residence time was 16
days.
Jackson, Michigan
In Jackson, Michigan, winter climatic conditions are less
favorable for nitrification. The Jackson municipal waste-
water treatment plant is a conventional activated sludge
system consisting of seven two-pass, fine pore, diffused
aeration tanks following primary clarification. Phosphorus
removal is accomplished by ferrous chloride addition to
the aeration tanks. The performance of this plant for a
period in 1973 and 1974 was reported previously (1). At
that time, coarse bubble diffusers were used for aeration.
In 1989, the plant installed fine pore aeration and made
other plant modifications (36). The maximum monthly av-
erage capacity of the current system is 1p,160 kg BODs/d
(22,400 Ib/d). The plant typically processes 616 to 748
L/s (14 to 17 mgd) of raw wastewater.
Operating and performance results from 1990 are pre-
sented for the Jackson plant in Table 6-8. The DO in the
carbonaceous oxidation/nitrification reactors typically ex-
ceeded 3.0 mg/L, and the pH was in the neutral range.
During the coldest months the temperature was estimated
at 10°C. The 1990 performance results indicate a dete-
rioration in ammonia oxidation during January and Feb-
ruary. Plant operating personnel said this deterioration
was due to the presence of cyanide, nickel, or zinc in the
wastewater rather than to temperature. The final effluent
zinc concentration during January and February aver-
aged 0.14 mg/L versus the yearly average of 0.09 mg/L.
Operations at an electroplating shop occasionally contrib-
ute up to 15 percent of the influent to the plant and may
be the source of these materials. As discussed in Section
3.3.5, such materials have been identified as potential
nitrification inhibitors.
The results from the coldest months of 1991 appear to
support the claim that lower temperatures are not affect-
ing nitrification. The effluent ammonium-nitrogen values
during January, February, and March of 1991 averaged
0.17 mg/L and the average sludge residence time was
11.7 days. It is worthwhile noting that in order to achieve
an effluent NHJ-N of 1.0 mg/L or less at 10°C (50°F) in
a complete mix system, nitrification theory (Equation 6-1).
would predict that a minimum solids retention time of 7
days would be required (assuming KN is 1.0 mg/L and
operations are under nonlimiting DO and pH conditions).
The Jackson plant results appear to support this, although
there are some inconsistencies in the reported MLVSS, 6,
and t values (Table 6-8).
6.4.4 Extended Aeration and Oxidation Ditch
Systems
Most extended aeration systems are similar to conven-
tional and complete mix plants except that the hydraulic
retention times range from 24 to 48 hours, instead of a
retention time of less than 10 hours that is normally used
in the conventional and complete mix configurations. As
such, the system is often considered simply a process
149
-------
Table 6-7. River Oaks Advanced Wastewater Treatment Plant—Carbonaceous Oxidation/Nitrification System Perform-
ance, August 1988 to July 1989 (Adapted from Reference 35)
Raw Wastewater Influent
Nitrified Effluent
Parameter*
Flow, US (mgd)
BODS, mg/L
TSS, mg/L
TN, mg/L
NHJ-N, mg/L
NO5-N plus NOJ-N, mg/L
Org-N, mg/L
TP.mg/L
* TN represents total nitrogen;
Average
142
118
30.1
21.1
0.17
8.8
6.3
Org-N, organic nitrogen; and
Maximum Month
176
149
37.8
24.9
0.24
12.8
7.0
TP, total
•
phosphorus.
Table 6-8. Jackson, Michigan, Wastewater Treatment Plant Nitrification
Results, 1990 (Reference 36)*
Wastewater
Flow, Us
Month (mgd)
Jan. 443 (13.4)
Feb. 642 (14.6)
Mar. 713 (16.2)
Apr. 713 (16.2)
May 726 (16.5)
June 717 (16.3)
July 660(15.0) •
Aug. 660 (15.0)
Sept. 638 (14.5)
Oct. 678 (15.4)
Nov. 634 (14.4)
Dec. 642 (14.6)
Solids
Recycle
Ratio
0.87
0.73
0.53
0.58
0.58
0.50
0.52
0.50
0.58
0.54
0.53
0.49
MLVSS,
mg/L
2,323
2,671
2,571
2,333
2,248
2,380
2,584
2,723
2,734
2,410
2,003
1,869
e,d
19.1
17.3
11.7
10.1
8.3
10.5
8.3
7.5
15.3
11.2
14.8
14.5
t, hr
9.8
9.1
7.8
8.0
7.9
8.1
8.8
8.8
9.1
8.1
9.1
9.1
Average
339
(7.7)
4
9
17.4
0.15
15.8
1.48
0.53
Maximum Month
418
(9.5)
8
22
21.4
0.31
19.1
1.98
0.86
System Operation Conditions and Performance
BODSJ
Primary
Effluent
92
69
65
84
87
96
89
85
83,
68
72
79
mg/L
Final
Effluent
3
3
2
3
2
2
2
2
2
2
2
3
Ammonia-N, mg/L
Primary
Effluent
17.4
12.7
9.3
12.5
10.8
11.5
11.2
11.7
11.7
9.8
10.6
10.8
Final
Effluent
4.7
4.7
0.4
0.2
0.7
0.1
0.1
0.1
0.3
0.2
0.2
0.2
* Average monthly values are presented.
150
-------
modification of the activated sludge complete mix or plug
flow reactor configuration. Extended aeration plants are
operated to maximize endogenous respiration; conse-
quently, solids retention times of 25 to 35 days are not
uncommon. Because of their long aeration periods, how-
ever, they can suffer from unusual heat losses and low
temperatures in cold climate areas. Extended aeration
plants can be expected to nitrify fully except at tempera-
tures less than about 8°C because of their characteristic
long solids retention time and the resulting accumulation
of the slower growing nitrifying organisms. Nitrification in
activated sludge systems is reported to occur at tempera-
tures as low as 2°C (36°F) (37).
An oxidation ditch system represents a modification of
the activated sludge process in terms of its reactor con-
figuration (see also Section 8.2.5). Wastewater and the
reactor biological solids are pumped around an oval path-
way which typically is configured as a single channel or
concentric multichanneled reactor (Figure 6-7).
Aerators in the form of brush rotors, disc aerators, surface
aerators, draft tube aerators, or fine pore diffusers with
submersible pumps provide mixing and circulation in the
oxidation ditch as well as oxygen transfer. Oxidation
ditches-typically are designed in the extended aeration
mode at hydraulic retention times of greater than 10 hours
and solids retention times of 10 to 50 days.
The increased popularity of oxidation ditch systems over
the past 15 years can be traced to the operating simplicity
and economic attractiveness of this reactor configuration.
Oxidation ditches were first installed in the United States
in the early 1960s (38). Since that time, the number of
oxidation ditches has increased to more than 1,900 in
1991 (39). Although most ditch systems were not de-
signed for nitrification, operation in the extended mode
normally ensures that nitrification will occur, provided suf-
ficient alkalinity is present and there is adequate aeration
capacity to meet nitrification demands.
6.4.4.1 Design and Performance Information for
Extended Aeration Systems
The procedure to follow in designing extended aeration
systems for nitrification is equivalent to that outlined for
complete mix systems (Section 6.4.2.1). If the mixing
conditions within the biological reactor(s) approximate
plug flow, refining the procedure to account for plug flow
hydraulic conditions is normally not required. This also
applies to the design of oxidation ditch nitrification reac-
tors. If the ditch system is designed for carbonaceous
oxidation and nitrification only, versus the inclusion of
partial or complete denitrification, care must be taken that
the aeration component is capable of maintaining a DO
of 2.0 mg/L or above throughout the ditch reactor, as
recommended in Section 3.3.3. If lower DO conditions
are anticipated, selecting the design solids retention time
should account for the effect of DO. Equation 6-1 can be
modified to include a Monod expression for DO effects
as follows:
= 0.47[e(
0.098(1-15).
N
K
DO
K0 + D0
(6-22)
where:
DO = DO concentration, mg/L
K0 = half-saturation coefficient for oxygen, mg/L
A value for K0 of 1.0 mg/L is acceptable for design pur-
poses. As implied by Equation 6-22, ammonium oxidation
will not occur in zones of the oxidation ditch reactor where
the DO concentration is very near zero.
Design and performance information for 17 oxidation
ditch municipal wastewater treatment plants in the United
States has recently been published (39). Although only
12 of the plants were designed for nitrification, the aver-
age effluent ammonia-nitrogen ranged from 0.5 to 2.7
mg/L for all 17 plants. The design hydraulic retention time,
at the average design flow, was 10 to 34 hours and the
design solids retention time, stated for eight of the plants,
was 12 to 48 days.
Oxidation Ditch
Screened and
Degritted
Wastewater
Return Sludge
Final
Clarifier
Excess
Sludge
Figure 6-7. Oxidation ditch system.
151
-------
The Frederick, Maryland, wastewater treatment plant
went on line in February 1988. The system consists of
primary clarification followed by three parallel oxidation
ditch reactors. Aeration is supplied by fine pore- sub-
merged diffusers and each ditch has an intrachannel clari-
fier. The clarified effluent passes through sand filters
before chlorine disinfection and discharge to the Mono-
cary River. The system is not designed for phosphorus
removal beyond that achieved through biomass growth.
The plant Is required to achieve a TKN limit of 2.6 mg/L
from May through October, the "summer" period. Operat-
ing and performance results for the periods of October
through December 1990 (winter) and from May through
July 1991 are presented along with available design in-
formation in Table 6-9. The DO in the oxidation ditch
reactors in the summer period typically varied from 1.5 to
2.0 mg/L; maintaining adequate and uniform DO condi-
tions during the warmer months has been difficult on
occasion (40).
In winter, the DO varies from 2 to 3 mg/L. The tempera-
ture in the oxidation ditch system can be expected to
range from 14°C to 25°C (57-77° F) over the year, ac-
cording to data developed in past years (40). Effluent pH
values are typically between 6.8 and 7.2. Although nitri-
fication was evident during the cooler months, the effi-
ciency was significantly less than in the summer, when
essentially complete nitrification occurred. Nitrification
theory suggests that lower NHJ-N levels should be
achievable in the oxidation ditch reactors during these
lower-temperature months, assuming operation under
nonlimiting DO and pH conditions. Incomplete mixing and
non-uniform DO and/or pH conditions in the reactor,
Table 6-9. Frederick, Maryland, Wastewater Treatment Plant Design Information and Operating Performance Results
(Adapted from Reference 35)
DESIGN
Influent Conditions m3/d (mgd)
Average daily flow
Peak hourly flow
Oxidation Ditch Reactors
Number of oxidation ditches
Hydraulic retention time,8 hr
Intrachannel clarifier retention time, hr
Clarifier hydraulic loading rate, m3/m2/d (gpd/sq ft)
26,500 ma/d (7.0 mgd)
61,300 m3/d (16.2 mgd)
3
12.5
3.5
25 (600)
Effluent Limits," mg/L
BOD6
TSS
TKN
Summer15
8.7
26
2.6
Winter
26
26
(no limit)
PERFORMANCE0
Summer (May, June,
July 1991)
Winter (Oct.,
Nov., Dec. 1990)
Influent
Average daily flow, m3/d (mgd)
Primary effluent BOD5l mg/L
TKN, mg/L
NHJ-N, mg/L
Effluent
BODS, mg/L
TSS, mg/L
TKN, mg/L
NHJ-N, mg/L
NOX-N, mg/L
27,600 (7.3)
116
25.9
16.8
5.6
3.5
2.1
0.53
20.8
32,900
127
22.1
14.4
14.5
10.8
4.7
2.8
14.8
(8.7)
* Calculated at average daily flow.
"Summer is defined as May through October.
e MLSS varied from 2,000 to 2,500 mg/L, and F/M from 0.12 to 0.15.
152
-------
sometimes a characteristic of oxidation ditch systems,
may account for this variation.
6.4.5 Contact Stabilization Systems
The contact stabilization system can be considered a
modification of the activated sludge process in terms of
reactor configuration. Return activated sludge is sepa-
rately aerated in a sludge reaeration or stabilization tank
prior to mixing with the influent wastewater (Figure 6-2).
Backmixing between the contact tank and the sludge
reaeration tank is prevented by providing overflow weirs
or pumps between the tanks. The contact tank has a
relatively short detention time, 0.5 to 1 hour based on
average dry weather flow (ADWF). CBOD5 removal is
accomplished in the contact tank primarily by adsorption.
CBOD5 removals can be high because the bulk of the
organics in municipal wastewater are paniculate or col-
loidal in nature and can be adsorbed to the biological
solids for later oxidation in the sludge reaeration tank.
The system is not well suited for complete nitrification.
Although the overall system solids residence time can be
long, the effective wastewater contact time for nitrification
is normally insufficient to achieve complete nitrification.
Insufficient biological mass is present in the contact tank
to nitrify the ammonium completely and since ammonium
is not adsorbed on the biological floe, ammonium will
bleed through to the effluent. Particularly unfavorable
conditions for nitrification will result during peak hydraulic
and/or organic loads. Although the contact stabilization
reactor configuration is not appropriate for achieving com-
plete nitrification, the system can be considered for partial
nitrification, particularly if there are seasonal require-
ments (41,42).
6.4.6 Step-Feed, Tapered Aeration, and Sludge
Reaeration Systems
The step-feed or step aeration system differs from a con-
ventional plant in that influent wastewater is introduced
at several points along the aeration tank (Figure 6-2). This
distribution of influent flow reduces the initial oxygen de-
mand often experienced in the conventional plant, where
mixing approaches plug flow conditions. A variation to the
step-feed system involves introducing no feed into the
first pass while directing the flow into the remaining down-
stream passes, creating a sludge reaeration zone in the
first pass. This "sludge reaeration plant" is similar to a
contact stabilization process except that the contact times
are normally longer. Generally, no provision is made to
prevent backmixing between the sludge reaeration zone
and the downstream aerated zones. The River Oaks
plant, discussed in Section 6.4.3.3, represents a system
where sludge reaeration occurs in a tank external to the
carbonaceous oxidation-nitrification reactors.
Advantages that are claimed for the step-feed activated
sludge process include production of a better-settling
sludge (43) and the flexibility to vary the solids concen-
tration near the effluent end of the aeration tank while
maintaining a constant reactor solids retention time
(44,45). The ammonia bleedthrough characterizing con-
tact stabilization plants is avoided in step-feed systems
because of the greater contact times employed and be-
cause backmixing of the influent typically occurs.
A process modification to conventional or plug flow acti-
vated sludge systems that regulates the oxygen supply
along the length of the reactor is referred to as tapered
aeration. Operational control is better facilitated when ta-
pered aeration is incorporated into conventional systems.
6.4.6.1 Design and Performance Information
The kinetics of a step-feed system are often better de-
scribed assuming complete mix versus plug flow hydrau-
lics because of the feed pattern and the occurrence of
backmixing. Depending on the tank configuration, how-
ever, some step-feed plants have plug flow characteristics
in each pass. The design approach developed for com-
plete mix systems (Section 6.4.2.1) can usually be
employed for step-feed plants as a reasonable approxi-
mation. The approach essentially allows determination of
the design solids retention time for achieving a required
effluent ammonium level. The step-feed process configu-
ration in combination with sludge reaeratio'n has been
shown to accommodate a reduction in aeration tank size
by allowing an increase in the total mass of solids under
aeration (45).
Tapered aeration is commonly employed in the design of
conventional or plug flow activated sludge systems. The
nitrification design approach developed for these systems
(Section 6.4.3.1) is applicable here. Historically, the ability
to regulate oxygen supply to the aeration tanks made the
tapered aeration configuration attractive when nitrification
was not desired. This feature is no longer relevant in the
context of designing for nitrification, but tapered aeration
does provide a solution to the problem (described in Sec-
tion 6.4.3.1) of supplying sufficient DO to the head end
of plug flow aeration tanks.
As mentioned earlier, there is a danger that if influent is
fed too near the effluent discharge point in the step-feed
process, insufficient time will be available for organic ni-
trogen hydrolysis. Under such conditions elevated levels
of organic nitrogen could be observed in the effluent. This
phenomenon was identified as contributing to the high
concentration of effluent organic nitrogen observed during
a test study completed at the Flint, Michigan, municipal
treatment plant a number of years ago (46).
The Flint test study was conducted in support of a plant
upgrade to comply with state regulatory requirements for
nitrification. At the time of the test, the existing plant had
three aeration tanks, each with four passes, providing a
total capacity of 21,225 m3 (750,000 cu ft). With an av-
erage design BOD5 loading of 11,110 kg (24,500 lb)/d to
153
-------
the aeration tanks at a 75,700 m3/d (20 mgd) flow, the
aeration tank load was 523 g/m3/d (32.7 lb/1,000 cu fl/d).
Flows to the facility were varied, however, to provide a
variation in loading. Three secondary sedimentation tanks
were provided with a design overflow rate of 27.6 m3/m2/d
(678 gpd/sq ft) at ADWF conditions. The plant was usually
operated in a step-feed mode, with one-half the influent
directed to the head ends of the second and third passes.
During the test, which lasted for a period of 10 months,
ferric chloride and polymer were added to the primary
treatment stage for phosphorus removal. This addition
also had the effect of reducing the organic loading to the
aeration tank.
Performance information during eight months of the test
are shown in Table 6-10. While nitrate and nitrite are not
shown, it was reported that a relatively good balance
between ammonium removal and nitrate production was
obtained. Nitrite nitrogen was always less than 0.1 mg/L.
The appearance of high concentrations of organic nitro-
gen was attributed to the low rate of hydrolysis of organic
nitrogen compounds, likely caused in part by the step-
feed process configuration. The 24.1 mg/L of effluent SS
also contributed to the elevated effluent levels of organic
nitrogen.
Table 6-10. Average Nitrification Performance at Flint,
Michigan, for Eight Months (Adapted from Reference 46)
Parameter
BOD5, mg/L
SS, mg/L
TKN, mg/L
Organic nitrogen, mg/L
Ammonia nitrogen, mg/L
Phosphorus, mg/L
Temperature, °C
Raw
Waste-
water
250
300
27.6
13.3
14.3
15.4
7.2-18.3
Settled
Waste- Secondary
water Effluent
131 13.6
140 24.1
23.3 7.8
9.9 6.1
13.4 1.7
2.7 2.3
The effect of temperature and solids residence time (0)
are considered in Table 6-11. Effluent quality deteriorated
somewhat with colder temperatures, with only 75-percent
ammonium removal at 10°C (50°F), at a solids residence
time of six days. This drop is not surprising considering
nitrification theory (Equation 6-1) which would predict that
a minimum solids residence time of approximately seven
days would be required to achieve essentially complete
nitrification.
6.4.7 High-Purity Activated Sludge System
The use of high-purity oxygen versus air for aeration as
a process modification to activated sludge reactors was
first evaluated in the 1950s (47). Both covered and un-
Table 6-11. Effect of Temperature and Solids Residence
Time on Nitrification Efficiency at Flint, Michigan (Adapted
from Reference 46)
NHJ-N Removal,
Temperature, °C
e,d
18 and greater
13
10
7
4
4-5
6
10-12
95
87
75
50*
* Based on bench scale test results.
covered reactors have since been used but only the for-
mer technique has seen wide application at the full-scale
level. The covered reactor approach involves the recircu-
lation of reactor off-gases to achieve efficient oxygen utili-
zation. The system uses a three- or four-staged
oxygenation basin for contact of gases and mixed liquor
(Figure 6-8). High-purity oxygen (>90 percent purity) en-
ters the first stage and flows concurrently with the waste-
water being treated. The gas continues to be used in
successive stages, resulting in the buildup of carbon di-
oxide (released by biological activity) in the gas and in
the liquid.
Advantages that are claimed for the high-purity oxygen
activated sludge systems, when nitrification is required,
include assurance that no DO rate limitations will occur
(48) since reactor DO values are typically greater than 4
mg/L (49), and reduced reactor hydraulic retention time
requirements due to operation at a higher MLSS concen-
tration (48). Reviews of plant information have indicated
that in some cases oxygen activated sludge systems had
reduced energy requirements, produced a better settling
sludge, and provided a greater tolerance for peak organic
loadings than conventional air activated sludge systems
(49).
The buildup of carbon dioxide in pure oxygen systems
typically causes a pH depression, and nitrification can
cause a further pH reduction. While the pH is also de-
pressed by biological carbon dioxide release in conven-
tionally aerated systems (see Section 3.3.4), the pH
depression is less significant than it can be in pure oxy-
gen systems because the evolved carbon dioxide is con-
tinually stripped from the system by the aeration air.
Further information on the effect of the aeration system
on the reactor pH is provided in Section 6.4.10.3.
6.4.7. / Design and Performance Information
The procedure to follow when designing pure oxygen
based nitrification systems is equivalent to that for com-
plete mix (Section 6.4.2.1) or conventional (Section
6.4.3.1) systems, depending on the hydraulic design of
the plant. Historically, the pure oxygen based systems
have been designed as two separate stages, with each
154
-------
Aeration
Tank Cover
Control
-^X g
as ^^ V "•
Agitator
Feed Gas
Wastewater
Feed
Recycle
Sludge
Exhaust
Gas
Mixed Liquor
Effluent to
Clarifier
Figure 6-8. Covered high-purity oxygen reactor with three stages and mechanical aerators.
stage consisting of three or more completely mixed reac-
tors operating in series.
If pH conditions in the nitrification reactor(s) are antici-
pated to be below or near the low end of the optimum
pH range for biological systems (this range was sug-
gested as 6.5 to 8.0 in Section 3.3.4), the effect of pH
should be considered in selecting the design solids re-
tention time for high-purity oxygen systems. Equation 6-1
can be modified as follows:
= 0.47teao98(T-15)]
N
KN
[1 - 0.833 (7.26 - pH)]
(6-23)
where:
pH = anticipated pH in nitrification reactor(s)
The pH effect modification was proposed by Downing and
Knowles (50) and further cited in the 1975 EPA Design
Manual (1). Understand that acclimation or organism
population selection will attenuate the effect of pH on the
nitrification rate, as has been discussed in Section 3.3.4.
The pH correction term in Equation 6-23 can be consid-
ered more applicable to unacclimated cultures within the
pH range of 6.1 to 7.2, and should be considered con-
servative when applied to acclimated cultures.
Widespread utilization of oxygen activated sludge sys-
tems for municipal and industrial wastewater treatment
began in the 1970s. In 1976 the results of a comprehen-
sive study on the effect of pH on pure oxygen nitrification
systems were reported (51). The study, completed at the
EPA-DC Blue Plains treatment plant, consisted of two
carefully controlled pilot investigations involving nitrifica-
tion of a wastewater with a low C:N in:
• four stage, high-purity oxygen activated sludge sys-
tems with and without pH control, and
• four stage, high-purity oxygen and conventional air ac-
tivated sludge systems controlled at the same pH level.
Detailed information concerning operating conditions and
performance results is provided in Reference 51. The
oxygen activated sludge system with pH control produced
a mean effluent ammonium-nitrogen level of 0.2 mg/L.
The effluent value for pure oxygen systems without pH
control was normally equal to or less than 1 mg/L, and
was more variable than observed for the pH controlled
system. In the controlled system, the pH was maintained
at approximately 7.0 by lime addition to the first stage,
while in the uncontrolled system pH conditions of 6.0 or
below were observed in the last stage. The fact that
nitrification was only minimally affected by the low pH
conditions is likely a result of two factors: first, the solids
residence time in the uncontrolled system was well above
the minimum value required (i.e., 9JP); second, it is likely
that acclimation and/or population selection resulted in
the nitrifiers operating at rates approaching those ob-
served under neutral pH conditions. The concentration of
organics and nitrogen species in the effluents were virtu-
ally identical during operation of the oxygen and air acti-
vated sludge systems at the same pH level. The lime
requirements to maintain a pH of 7.0 in the last stage of
each four-stage system were 2 to 3 times greater in the
high-purity oxygen system.
The Town of Amherst Wastewater Treatment Facility in
Amherst, New York, discharges to Tonawanda Creek. The
plant was designed in the late 1970s for an average daily
wastewater flow of 1,050 Us (24 mgd), anticipated in the
year 1990. The plant began operation in 1980. Carbona-
ceous oxidation and nitrification, following primary clarifi-
155
-------
cation and flow equalization, are accomplished in two
separate reactor-clarifier stages using high-purity oxygen
activated sludge. Following the Stage 2 clarifiers, the ef-
fluent undergoes sand filtration and seasonal chlorination
(i.e., from May 15 through October 15) prior to discharge.
Phosphorus removal is accomplished by ferrous sulphate
addition to the Stage 1 reactors. The nitrifying sludge
inventory is maintained at an appropriate level in the
Stage 2 reactors by routing a portion of the equalized,
primary treated wastewater directly to the reactors. The
wastewater flow bypassed to Stage 2 is typically 198 L/s
(4.5 mgd) but ranges from 0 to 308 Us (0 to 7.0 mgd).
Design information pertaining to the carbonaceous oxida-
tion-nitrification system of the Amherst plant is sum-
marized in Table 6-12. Operating and performance
Information over the 12-month period from October 1990
to September 1991 is presented in Table 6-13. The DO
in the Stage 1 and Stage 2 reactors during the period
was quite high, typically exceeding 10 mg/L. The pH de-
crease across the plant, based on raw wastewater and
final effluent determinations, was typically 0.4 units. The
effluent pH ranged from 6.2 to 7.3. The final effluent
alkalinity ranged from 90 to 174 mg/L (as CaCO3). The
temperature in the reactors during the coldest months
Table 6-12. Town of Amherst, New York, Wastewater Treat-
ment Plant Carbonaceous Oxidation-Nitrification System
Design Information Following Primary Clarification (Refer-
ence 52)
Design Flow 90,840 m3/d
(24 mgd)
Carbonaceous Oxidation Reactors (Stage 1)a
Number of parallel trains 2
Number of reactors In each train 3
MUSS, mg/L 6,000
MLVSS, mg/L 3,950
Hydraulic retention time, hr 1.05
F/M, kg BODj/kg MLVSS/d 1.17
Stage 1 Clarifiers"
Number 4
Hydraulic loading rate, m3/m2/d (gpd/sq ft) 18.3 (450)
Solids loading rate, kg/m*/d (Ib/sq ft/d) 112 (23)
Nitrification Reactors (Stage 2)a
Number of parallel trains 4
Number of reactors in each train 3
MLTSS, mg/L 6,000
MLVSS, mg/L 4,450
Hydraulic retention time, hr 2.10
6, d 23.8
F/M, kg BODg/kg MLVSS/d 0.09
Stags 2 Clarifiersb
Number 4
Hydraulic loading rate, m3/ma/d (gpd/sq ft) 15.9 (390)
Solids loading rate, kg/m2/d (Ib/sq ft/d) 98 (20)
* Hydraulic retention time calculated at design flow.
6 Hydraulic and solids loading rates to clarifiers based on design flow.
(i.e., January, February, and March) was estimated be-
tween 10°C (50°F) and 11°C (52°F). The performance
results (Table 6-13) indicate a slight deterioration in am-
monium oxidation during February and March when the
temperature in the reactors was near 10°C and the Stage
2 solids residence time was between 10 and 13 days.
The effluent pH typically was 6.7 to 6.8 during these
months. To achieve an effluent NHJ-N of 1.0 mg/L or less
at 10°C (50°F) in a complete mix reactor, nitrification
theory (Equation 6-23) would predict,that a minimum sol-
ids residence time of 12 days would be required, if KN
was 1.0 mg/L, the pH was 6.7, and operation was under
nonlimiting DO conditions. For the staged oxygen reactor
(i.e., complete mix zones in series) an SRT of less than
12 days would be predicted to be adequate.
6.4.8 Sequencing Batch Reactor Systems
Sequencing batch reactor (SBR) systems represent a modi-
fication of the activated sludge process. SBRs include a
generic system of variable volume activated sludge units in
which aeration, sedimentation, and decant are combined in
a single reactor. Consequently, there are no dedicated sec-
ondary clarifiers or associated return sludge facilities.
SBR manufacturers have adapted the sequence of batch
treatment cycles in various ways. One classification of
SBR systems distinguishes those which operate with con-
tinuous feed and intermittent discharge (CFID) from those
which operate with intermittent feed and intermittent dis-
charge (I FID). I FID reactors are characteristic of the con-
ventional fill-and-draw SBR reactors in that the influent
flow to the reactor is discontinued for some portion of
each cycle. The I FID reactor treats the aqueous waste-
water feed stream through a succession of operating
steps, namely, fill, react, settle, draw, and idle (Figure
6-9). The liquid volume inside the SBR increases from a
set minimum volume to a predetermined maximum vol-
ume during the fill period. Mixing and/or aeration may be
provided during this first step to promote biomass growth,
organics oxidation, and nitrification-denitrification. During
react, flow to the tank is discontinued and aeration and/or
mixing are provided, while sufficient time is allowed for
the microbial reactions to take place. During settle, qui-
escent conditions are initiated and the biomass is allowed
to flocculate and settle prior to removal of the treated
clarified supernatant from the tank to the minimum vol-
ume level. During the idle period, which is a normal com-
ponent in multireactor installations, biomass is retained
in the reactor but no waste is treated. During this period,
excess biomass may be removed from the tank to main-
tain a desired reactor solids residence time.
The CFID reactors receive wastewater during all phases
of the treatment cycle. A key design consideration with
such systems is minimization of short-circuiting between
influent and effluent. This is accomplished by locating the
feed and withdrawal points at opposite ends of the tank,
156
-------
Table 6-13. Town of Amherst, New York, Wastewater Treatment Plant Carbonaceous Oxidation-Nitrification System Oper-
ating Conditions and Performance Results, October 1990 to September 1991 (Reference 52)* . : -
Wastewater
Flow
Month L/S
Oct. 902
Nov. • 871
Dec. 1,183
Jan: 1,016
Feb. 1,007
Mar. 1,197
Apr. 1,042
May 849
June 765
July 770
Aug. 752
Sep. '756
mgd
20.5
19.8
26.9
23.1
22.9
27.2
23.7
.,19.3
17.4
17.5
17.1
17.2
* Average monthly values
Wastewater flow.
MLTSS,
mg/L
Stage
1
7,475
8,298
7,358
8,330
7,492
6,487
7,826
9,357
8,057
8,442
7,361
5,236
Stage
2
5,183
5,063
4,721
5,174
5,707
5,328
4,786
6,279
6,695
5,881
4,454
3,202
HRT, hr
9, d BODS , mg/L
Stage Stage Stage Stage
1
1.2
1.3
0.9
1.1
1.1
0.9
1.1
1.3
1.4
1.4
1.5
1.5
are presented. Flow rate
2
2.4
2.5
1.9
2.2
2.2
1.9
2.1
2.6
2.9
2.9
2.9
2.9
given
1
7.9
6.3
5.2
7.3
7.1
4.8
7.1
9.1
9.6
12.4
11.0
7.7
includes
2 Influent
18.2 110
16.3 119
12.0 94
13.7 101
13.0 102
9.9 82
16.0 94
10.8 124
19.4 122
19.2 122
10.0 125
9.1 147
plant-prooess-water
Stage 1
Effluent
22
11
14
14
13
13
7
25
13
12
36
14
return flow.
Final
Effluent Influent
1 22,1
2 ,22,6
2 16.9
2 18.4
4 18.4
3 16.1
2 17.4
2 24.3
1 24.6
2 24.2 _
1 24.9
1 28.3
Hydraulic retention
TKN, mg/L
Stage 1
Effluent
8.9
10.5
9.1
10.9
11.5
8.9
8.5,
16.6
14.5
.14.6
15.5
9.5
time based
Influent represents the raw Wastewater prior to grit removal. Final effluent represents the effluent after chlorination (when
Final
Effluent
0.6
0.8
0.7
0.9
1.9
2.8
1.1
0.7,
0.7
0.8
0.6
0.7
on stated
required).
, . Feed-
Fill
Operation
Wastewater
Addition
React
Biodegradation
Settle
Clarification
Effluent
Draw
Effluent
Withdrawal
Idle
Waste Excess
Biomass
Figure 6-9. Single-tank SBR system operating steps.
using rectangular reactors with length-to-width ratios of
at least 2 to 1 and providing baffling. ,
Major advantages that are often cited for the SBR tech-
nology include the ability to tolerate peak flows and shock
loads of BOD5, no need for separate clarification and
return sludge pumping systems, and controlled effluent
discharge (53). The first modern, fullrscale plant for SBR
treatment of municipal wastewater in the United States
was the Culver, Indiana, wastewater treatment facility.
Retrofitted for the SBR process, operation was initiated
in May 1980 (54). Since that time, SBR technology has
become widespread in the United States, with more than
150 plants in design or operation (55). SBRs can be
modified to provide carbonaceous oxidation, nitrification,
and biological nutrient removal. Approximately 25 percent
of all SBR systems currently operating were designed to
achieve nutrient removal (53). :
The SBR technology is particularly attractive for treating
smaller wastewater flows. The majority of the plants op-
erating to date were designed at wastewater flow rates
of less than 22 Us (0.5 mgd) (55). The cost-effectiveness
of SBRs may limit their utilization to flows less than 440
Us (10 mgd) (54). Depending on the number of SBR
reactors in a plant and the duration of the discharge cycle,
the downstream units often must be sized for two or more
times the influent flow rate. Plants with four or more sepa-
rate reactors may have the reactor process cycles offset
such that the discharge is nearly continuous.
157
-------
6.4.8.1 Design and Performance Information
Each supplier of SBR system equipment has their own
approach to design (see Section 8.2.6.1). Some SBR
systems are custom designed and the uniqueness of
each of these systems reflects the preferences of the
design engineer. Designs include the use of different tank
configurations, different system hydraulics and a variety
of options for aeration, mixing, effluent discharge, and
sludge wasting. Systems are normally configured to vary
their operation automatically in response to changes in
influent flow rate, or to allow the operator to initiate
changes to the total cycle time or individual step times,
or to make changes during each step (e.g., change length
of time for aeration or mixing during fill step). The steps
and associated conditions and purpose of a complete,
typical cycle for a single tank operated as part of an I FID
SBR system designed to achieve nitrification are de-
scribed in Table 6-14. Nitrification takes place during the
react phase and during the portions of the fill period when
aeration is practiced.
In order to design SBRs for nitrification, an adaptation of
the approach used in the design of complete mix systems
is normally acceptable. The specific calculation procedure
will be dictated by the characteristics of the selected SBR
system. The most important calculation steps are to de-
termine the minimum required aerobic solids residence
time (using Equation 6-1 or the modified versions dis-
cussed earlier, which account for the effects of limiting
DO and pH conditions), and to determine the minimum
volume requirements that will assure adequate time for
settling and decanting. Other critical parameters for the
design of the SBR system can be determined from infor-
mation presented in Section 8.2.6.2 and elsewhere (55).
SBR systems are typically designed and operated at long
solids residence times (>15 days) and low F/M (less than
0.1 kg BODs/kg MLSS/d). Consequently, partial or com-
plete nitrification is nearly always observed (53,55). In a
recent evaluation of 19 SBR treatment plants (53) (all
originally designed for nitrification), influent and effluent
ammonia-nitrogen data were reported for eight of the
plants (Table 6-15). The average effluent ammonium-ni-
trogen concentration for the eight plants was less than
2.0 mg/L, implying that a high degree of nitrification was
achieved in all cases. These efficiencies reflect the long
design solids residence times that are employed and op-
erations that are generally well below the design flow.
6.4.9 Powdered Activated Carbon Activated Sludge
Systems
The powdered activated carbon (PAC) activated sludge
system is a process modification of the activated sludge
process. The addition of PAC to plug flow and complete
mix suspended growth reactors is a more common proc-
ess modification for industrial wastewater treatment, and
has been applied to municipal systems in some in-
stances. PAC is added to the aeration tank, where it is
mixed with the biological solids (Figure 6-10). The mixed
liquor solids are settled and separated from the treated
effluent in a gravity clarifier. Polyelectrolyte will normally
be added prior to the clarification step to enhance solids-
Table 6-14. Typical Cycle for a Single Tank in a Dual Tank
Step Conditions
SBR System Designed for Nitrification (Adapted from Reference 55)
Purpose
FILL Influent flow into SBR
Aeration occurs continually or intermittently
Time = half of cycle time
REACT No influent flow to SBR
Aeration
Time typically = 1 to 2 hours (varies widely
depending on nitrification kinetics, waste
strength, and amount of aeration during fill)
SETTLE No influent flow to SBR
No aeration
Time = approximately 1 hour (depends on
settling characteristics)
DRAW No influent flow to SBR
No aeration
Effluent is decanted
Time = 1 hour (variable)
IDLE No influent flow to SBR
No aeration
Sludge is wasted
Time = variable (determined by flow rate)
Addition of raw wastewater to the SBR; COD removal
and nitrification
Carbonaceous oxidation and nitrification
Allow SS to settle, yielding a clear supernatant
Decant—remove clarified effluent from reactor; 15 to 25
percent of the reactor volume is typically decanted,
depending on hydraulic considerations and SBR
manufacturer's design
Multi-tank system, which allows time for one reactor to
complete the fill step before another starts a new cycle;
waste sludge—remove excess solids from reactors
Note: A typical total cycle time Is 4 to 6 hours.
158
-------
Table 6-15. Nitrification Performance Information for SBR Operating Plants (Adapted from Reference 53)*
Plant Location
Buckingham, PA
Clarkston, Ml
Period of
Evaluation
04/89-04/91
11/89-04/91
Wastewater Flow
m3/d
439
208
mgd
0.116
0.055
Percent of
Design Flow
49
50
BOD5,
Influent
324
192
mg/L
Effluent
8
12
Ammonia-N , mg/L
Influent
25.3
39.1
Effluent
1.1
1.7
(Chateau Estates)
Grundy Center, IA
Marlette, Ml
Mifflinburg, PA
Monticello, IN
(White Oaks Resort)
12/89-11/90
07/90-06/91
10/88-03/91
10/89-05/91
2,176
1,578
2,763
15
0.575
0.417
0.73
0.004
72
60
81
8
195
103
105
131
4
4
12
5
15.8
10.1
7.8
3.1
1.2
0.5
0.4
0.3
Muskegon Heights, Ml
(Clover Estates)
Windgap, PA
01/88-10/90
02/90-10/90
132
2,116
0.035
0.559
78
56
185
160
9
7
21.2
12.9
0.7
,0.6
* Average monthly values based on all data available.
Carbon
Addition
Wastewater
Chemical
Addition
Aeration Tank
Clarifler
Effluent
Regeneration or
Disposal
Figure 6-10. Powdered activated carbon activated sludge system.
liquid separation. If phosphorus removal is necessary,
alum is often added at this point also. Even with polyelec-
trolyte addition, tertiary filtration is normally required to
reduce the level of effluent SS. The clarifier underflow
solids are continuously returned to the aeration tank. A
portion of the carbon-biomass mixture is wasted peri-
odically to maintain the desired solids inventory in the
system.
Demonstrated advantages of PAC addition to suspended
growth reactors include improved solids settling and de-
watering characteristics; the ability of PAC to adsorb
biorefractory materials and inhibitory compounds, improv-
ing effluent quality and reducing the impact of organic
shock loads; reduction in odor, foaming, and sludge bulk-
ing; and improved color and CBOD5 removal (49). Be-
cause PAC is wasted with excess biomass, virgin or re-
generated PAC addition is required to maintain the
desired concentration in the biological reactor. This can
represent a significant cost factor for the system. When
carbon addition requirements exceed 900 to 1,800 kg
(2,400-4,000 lb)/d, wet air oxidation/regeneration (WAR)
is claimed to represent an economical approach to carbon
recovery and waste biomass destruction (56). However,
an ash separation step is needed in this case, affecting
the economics of carbon regeneration and recovery (57).
The economic analysis is further clouded by the inability
to analytically differentiate powdered carbon from back-
ground refractory volatile materials, thus making it difficult
to quantify the value of the volatile suspended material
159
-------
recovered after WAR. Although ash separation processes
have been reported to be effective in at least two munici-
pal PAC activated sludge plants (58,59), the economics
of complete PAC/WAR systems relative to other activated
sludge nitrification systems are unclear (57).
In the United States, PAC activated sludge systems for
nitrification generally have been applied at municipal
treatment plants where industrial sources contribute a
significant fraction of the incoming wastewater. In all in-
stances PAC regeneration was included in the flowsheet
(60). A summary of selected municipal PAC facilities is
presented in Table 6-16.
6.4.9.1 Design and Performance Information
The procedure to follow in designing PAC activated
sludge systems for nitrification involves a modification to
those for complete mix (Section 6.4.1) or conventional
plug flow systems (Section 6.4.3.1) in order to account
for the effects of the addition of PAC. According to the
major supplier of the technology (60), most PAC systems
are designed at MLTSS concentrations of approximately
15 g/L The mixed liquor is composed of volatile activated
carbon, biomass, nonvolatile PAC ash, biomass decay
components, and influent inert material. The relative pro-
portions of these materials are strongly influenced by
whether carbon regeneration via wet air oxidation and a
return of this material to the aerator is practiced. The
intent is to maintain the PAC concentration at approxi-
mately 1.5 times the biomass level in nitrification PAC
reactors (60). The most appropriate PAC concentration
will be dictated by the specific wastewater characteristics
and often cannot be specified without bench or pilot scale
studies. The PAC concentration to be added will depend
on the design solids retention time, the hydraulic retention
time and the required PAC concentration in the reactor.
The PAC concentration to be added can be calculated
from:
(6-24)
where:
RAC0 = influent PAC concentration, mg/L
PACR= mixed-liquor PAC concentration in the reactor,
mg/L
= effluent PAC concentration, mg/L
t = hydraulic retention time, d
The value of PAC! in Equation 6-24 can be estimated by
assuming that the carbon fraction in the effluent TSS is
the same as the fraction of PAC in the MLSS.
As an example, if complete mix hydraulics were employed
for the bioreactor of the PAC activated sludge system,
the design procedure would follow Steps 1 through 8 as
detailed in Section 6.4.2.2. In order to complete Steps 9
and 10, X needs to be selected recognizing that the total
MLSS will now include PAC. Once X and PACR are speci-
fied and t is determined from Equation 6-15, the required
influent PAC concentration can be calculated from Equa-
tion 6-24.
PAC activated sludge nitrification systems are normally
selected when the municipal wastewater contains com-
pounds originating from industrial operations, as
stated previously. Nitrifiers are susceptible to a num-
ber of organic and inorganic inhibitors found in many
industrial wastewaters, as noted in Section 3.3.5 and
Table 6-16. Summary of Municipal PAC/WAR Facilities Reviewed (57)
Permit Limits
Facility
Vomon, CT
Mt. Holly, NJ
E. Burlington, NC
S. Burlington, NC
Kalamazoo, Ml
Bedford Hts., OH
Medina Co., OH
N. O!msted,°OH
Sauget, IL
El Paso, TX
Current/Design
Flow, m3/s
0.18/.28
0.11/.22
0.31/.53
0.30/.42
1.1/2.4
0.15/.15
0.31/.44
0.26/.31
0.70/1.2
0.20/.44
PAC/WARa
Status
MA
MA
MA
AS
MA
MAC
MA
AS
AS
MA
Reason
for PAC"
C
C,S
C.N.T
C.N.T
C.N.T
N,S
N
N.S
T
N,O
BOD5,
mg/L
10
30
12-24
12-24
7-30
10
10
30
20
SDd
TSS, mg/L
20
30
30
30
20-30
12
12
30
25
SD
NHJ-N, mg/L
—
20
4.0-8.0
4.0-8.0
2.0-10.0
5.1
1.5-8.0
2.3-6.9
—
SD
* MA a Modified operation and/or design for ash control. AS = Converted to conventional activated sludge. NAG = Converted to the use of
nonactivated carbon without regeneration.
11C * Color Removal; S = Space; N = Nitrification; T = Toxics; O = Organics.
«Plan to convert to NAG without regeneration.
160
-------
in Section 6.3.1. Researchers have provided evidence
that the addition of PAC to nitrifying activated sludge
systems receiving industrial wastewaters improved nitrifi-
cation rates (61,62). More recently, studies have been
completed with the goal of determining the mechanism
of nitrification enhancement in PAC activated sludge sys-
tems in the presence of adsorbable and nonadsorbable
inhibitors (63). The results indicated that the addition of
the proper amount of PAC can completely nullify the toxic
effects of an adsorbable nitrification inhibitor. A minor
positive effect on' nitrification rates was observed when
PAC was added to a nitrifying activated sludge system
receiving nonadsorbable inhibitors. The activated sludge
used in these studies was not acclimated to the inhibiting
compounds.
6.4.10 Other Design Considerations for Suspended
Growth Nitrification Systems
6.4.10.1 Selection of Peaking Factors, Safety
Factors, and Process Design Factors
The selection of peaking factors and safety factors for
process design should not be confused. Peaking factors
are used to reflect assumed realities under the controlling
conditions of the design. Safely factors are used to reflect
uncertainty in performance under these realities. Whether
or not the two are conceptually multiple to establish an
overall process design factor depends on the application;
judicious application of both peaking and safety factors
can avoid a clearly inappropriate under- or over-designed
condition.
The development of the design example in Chapter 2
introduces the reader to concepts associated with the
development of the peaking factors for process design.
They first reflect the compliance interval of the plant's
effluent objectives (Section 2.9.3.1). These factors then
consider the variability of the influent wastewater charac-
teristics (Section 2.9.3.2) and their impact on the inte-
grated works of the treatment plant (Section 2.9.3.3) for
the processes under investigation (Section 2.9.3.4). The
impact of the wastewater characteristics and their implied
peaking factors are evaluated under the planned modes
of operation through the preparation and use of mass
balances (Sections 2.9.3.5 and 2.9.3.6, respectively). The
developmental discussion and concluding table for the
design example (Table 2-18) show that the elected proc-
essing peaks vary as a function of the unit process and
processing considerations, with the layered impact of the
processing recycles. ,
The designer can influence the process design peaking
factors for the raw wastewater and processing recycles
through equalization and diversion, or split-treatment
strategies, and the elected processes and operating
strategies for the treatment works. Longer hydraulic and
solids residence times, and continuous, as opposed to
discontinuous, processing, serve to mitigate pollutant
mass peaks through the treatment system.
Flow peak mitigation is one of the most important—if not
the most important—concerns in suspended growth sys-
tems because of solids washout concerns. Here, the de-
signer should consider the applicability of automatic high
flow diversions around the reactor and/or the entire bio-
logical treatment system, as well as flow equalization with
the equalization tank dedicated to receiving both the first
and the end-of-storm sewer flushing events. Other peak-
ing factor cbnsiderations, beyond these general consid-
erations, are best considered as a function of the reactor
design.
Safety factors in process design are ultimately expres-
sions of design confidence. They are used when there is
uncertainty. Higher safety factors may be used where the
technology is less demonstrated or more unproven. One
example of a commonly used safety factor in design is
multiplication by two of the reciprocal of the controlling
design Nitrosomonas specific growth rate (|j,N). This solids
residence time is expected to accommodate unknown
variations in temperature, DO, residual ammonium con-
centration, operating variations, and pH (if determined to
be applicable). Conceptually, the sound design approach
would be then to apply this solids residence time under
the controlling design conditions that are established by
correct application of the peaking factors.
6.4.10.2 Incorporation of Phosphorus Removal
Chapter 2 provides the stoichiometry of the phosphorus
removal reactions (Table 2-3), and the considerations as-
sociated with the incorporation of phosphorus removal
with a suspended growth nitrification system are largely
developed in Section 2.7.2 and demonstrated in Design
Example No. 1 (Section 6.4.2.2). A summation of the
salient points as they may influence suspended growth
nitrification systems follows:
• Solids Production:
- All designs must anticipate the production of addi-
tional waste solids due to the,phosphorus removal
step and select a design MLSS concentration with
this in mind.
- Stringent levels of phosphorus removal may not be
obtainable with secondary equivalency levels of ef-
fluent SS (e.g., maximum 30-day average of 30
mg/L) because of the increasing phosphorus con-
centrations in the sludge mass and effluent SS.
• Metal Salt Addition Strategies:
- Must anticipate the attendant alkalinity depletions if
an acid carrier is used with the metal salts.
- May be especially attractive in multipoint addition
strategies for improved solids-liquid separation (with
or without polymers), and with upstream primary
161
-------
clarification as a means of minimizing reactor re-
quirements for cold temperature nitrification, as well
as lower overall dosages.
- Are strongly suggested as backup (and planned
utilization) for strategies that anticipate biologically
enhanced phosphorus removal strategies to ensure
compliance with the plant's effluent objectives.
• Biologically Enhanced Phosphorus Removal Strate-
gies:
— Must pay particular attention to over-aeration (the
best performance is not expected with nitrifying sys-
tems), mainstream solids separation, and the solids
processing train to minimize the likelihood of phos-
phorus resolubilization and return.
- May be best applied with the direct and immediate
application of lime to the waste sludge stream (high
pH phosphorus precipitation) to ensure the perma-
nent removal of the phosphorus contained in the
waste solids beyond normal synthesis or metal pre-
cipitation expectations.
6.4.10,3 Effect of Aeration System on Process Design
The practical realities of the elected aeration system have
a strong influence on process design in that they can
influence the DO and the pH in the system, as well as
the character of the effluent SS. The first two environ-
mental variables can, and do, affect the overall required
solids residence time for nitrification. The character of the
effluent SS may be altered by the mixing energy used in
the oxygen dissolution system.
As explained in Section 2.5.1, the pH of any system
reflects its bicarbonate-carbonate equilibrium with the
carbon dioxide in the liquid and the atmosphere. Besides
the alkalinity demand of the nitrification reaction, variables
include the water surface exchange with the atmosphere
and the carbon dioxide generated by biological stabiliza-
tion of the carbonaceous material. For a biological system
with a nonlimiting alkalinity, lower pH values can be ex-
pected to be promoted by, in order of increasing signifi-
cance, high MLSS concentrations, high loadings of
biodegradable carbon, low physical surface exchange,
and whether the reactor surface is open or closed to the
atmosphere. Open atmosphere operation and isolation of
the nitrification reaction are certain to yield the most fa-
vorable pH conditions because of the relative absence of
any appreciable carbon dioxide release by the biomass.
The closed pure oxygen activated sludge systems, de-
signed for combined carbonaceous and nitrogenous ma-
terial oxidation, can, and do, experience low reactor pH
conditions (7 or less) and in the absence of pH control
provisions, are most safely designed for a longer solids
residence time than air systems since rate increases from
their elevated DO level may not be completely compen-
sating for the rate decreases resulting from the lower pH.
Thereafter, of probable minor importance, higher pH con-
ditions would be expected from a coarse rather than a
fine bubble diffused air supply. With mechanical aeration
systems, the resultant pH will vary as a function of the
system; total reactor coverage as opposed to the single
site application (as classically encountered in ditch con-
cepts) would be expected to result in the most favorable
pH conditions.
DO levels are probably the most important influence on
the overall solids residence time of the system. Systems
that anticipate DO levels below the nonlimiting condition
must adjust the design solids residence time upward to
reflect the very real slowdown in the nitrification replica-
tion rate. This consideration becomes particularly impor-
tant in systems that try to achieve some denitrification in
the nitrification reactor, and leads to the often observed
compromise of elevated ammonium concentrations with
reasonable control of the effluent oxidized nitrogen levels.
The compromise undoubtedly accelerates as design con-
ditions are approached. The significance of this consid-
eration varies with the design intent of the reactor and
.the dissolution system. Manipulation strategies largely
vary with how the air or oxygen is introduced, the reactor
configuration and, of course, the operating concentration
and age of the MLSS.
Coarse or fine-pore diffusion strategies yield the greatest
flexibility and assurance of a nonlimiting nitrification solids
residence time. Brush or single-point aeration devices in
ditches are the opposite extreme and can yield nitrifica-
tion limiting DO levels for a significant portion of the re-
actor volume. Other mechanical aeration systems lie
between these two extremes and may also result in nitri-
fication limiting DO levels in the early stages of a plug
flow design, or find it difficult to achieve elevated DO
levels throughout the reactor's content in a complete mix
design. Investigations at 23 activated sludge plants in
West Germany clearly show that surface aeration sys-
tems did not nitrify as well as diffused aeration plants at
similar loadings and temperatures and comparable oxy-
gen concentrations (as measured at the surface of the
aeration basin) because of the lack of uniformity of ele-
vated DO levels in the aerator (64).
The type of aeration system may also influence the efflu-
ent SS through the resulting localized or systemwide tur-
bulence. Higher levels of effluent SS may be encountered
with increasing mixing intensity as represented by the
root-mean-square velocity gradient, G. In general, G will
increase with a declining oxygen transfer efficiency and
depend on the nature of the oxygen dissolution system.
Fine-pore diffused aeration systems will have lower G
values than coarse bubble systems, and mechanical
aeration will most often have G values similar to those
encountered in coarse systems, with increasingly intense
localized areas of mixing based on the number of aeration
points. Lower level MLSS concentrations provide a
162
-------
means of avoiding high oxygen uptake rates per unit
volume of reactor (and high G designs). Tapered aeration,
including serial staged application of mechanical aeration
systems, and external or internal flocculation clarifiers
provide a means of floe reformation and recovery when
high Gs are unavoidable.
Recent investigations (65), which involved 105 measure-
ments at nine mechanical aeration activated sludge
plants and 150 measurements at 15 diffused aeration
activated sludge plants, reveal average supernatant sam-
ple SS levels of 17 ± 9 and 11+8 mg/L, respectively.
Effluent SS may or may not match these values, depend-
ing on the distance from the mechanical aeration system,
and the shearing/flocculation opportunities in the trans-
port system from the aerator to the point of separation at
the sedimentation tank. Optimum G values are reported
to be on the order of 70 s'1, with little perceptible change
up to a mixing intensity of about 100 s"1; deterioration is
likely once G exceeds 150 s"1. Procedures to calculate G
values for mechanical or pneumatic mixing systems are
readily available (66).
6,4.10.4 Aeration Requirements and Design
Aeration requirements are determined by the procedures
described in Tables 2-15 and 2-16 of Chapter 2. Spatial
and temporal variations in oxygen demand can be deter-
mined through use of mathematical models such as those
described in Chapter 5. Additionally, and especially with
fine pore-diffusion systems, the designer should check for
adequate mixing (G values of 50 to 70 s"1) and, if appli-
cable, blower turndown during the minimum monthly
needs of the design life of the facility.
6.4.10.5 Solids-Liquid Separation
The considerations associated with the design of the sedi-
mentation system for a suspended growth reactor are
described in Section 2.7.1.1. In combined carbonaceous
and nitrogenous oxidation systems, minimal sludge blan-
kets, which are obtained by operating at high return rates
with low return sludge concentrations, will ensure free-
dom from settled sludge denitrification and resultant rising
and floating sludge. Other enhancements may include a
rapid hydrostatic, full-floor sludge collection system as
opposed to a single-point (often a center-well) sludge
collection system; although, the attractiveness of this fea-
ture has yet to be established with a dilute return sludge
concentration (minimal sludge blanket) design and oper-
ating strategy.
6.4.10.6 Seasonal Nitrification Considerations
Seasonal nitrification considerations initiate with the reali-
zation that the plant must fully meet its effluent objectives
for the first full month that they apply. Accordingly, condi-
tions to achieve this compliance must be established in
the preceding month.
The likely controlling environmental parameter is waste-
water temperature and its determining influence on the
required nonlimiting aerobic solids residence time of the
nitrifiers. Long hydraulic residence time reactors are likely
to have somewhat lower reactor temperatures than the
influent wastewater temperature. Surface aerator sys-
tems are likely to have colder reactor temperatures than
diffused aeration systems. To ensure that a sufficient ni-
trifier population will be available at the start of the com-
pliance month, the population must be established in the
previous month. If the average monthly reactor tempera-
ture for the month that precedes the initial seasonal com-
pliance month is lower than the wastewater temperature
during the permit period, the temperature for the month
preceding the permit period can normally be used safely
for design.
After realizing that the proper temperature and DO de-
pendent solids residence time must be in effect to achieve
the required degree of nitrification for the first compliance
month, the designer has many choices on how to achieve
this objective. The following paragraphs describe several
representative approaches.
Total System Responsiveness. Assuming that the coldest
wastewater temperatures occur during the month before
the compliance standard is applicable, simply design the
reactor to achieve the desired 0Jj! for that month. This
approach is the most conservative, with the least amount
of risk; it may also be the most expensive.
Reactor/Clarifier Isolation or Split Treatment. This ap-
proach does the same as the first except that it is limited
to a fraction of the plant. The plant fraction dedicated to
full nitrification is determined by the allowable effluent
ammonium release and the mix of non- and fully nitrified
process streams. Here, the designer might think of con-
stant flow treatment to the nitrifying portion of the plant
and allow the remaining portion of the plant to experience
exaggerated swings associated with the balance of the
flow to the plant. Constant flow strategies to the isolated
nitrification system will also allow higher MLSS concen-
trations in this system, since the operating level does not
have to be established in anticipation of the diurnal and/or
seasonal flow peaks. Waste solids from the nitrifying por-
tion of the plant could also be delivered to the rest of the
secondary treatment system to seed this system and
achieve some favorable ammonium oxidation.
Lower the Applied Suspended Solids (Carbon). If the
plant has primary sedimentation, the applied BOD5 and
SS to the nitrifying activated sludge system may be re-
duced by chemically enhanced primary treatment through
the use of metal salts and/or organic polyelectrolytes.
Such strategies will yield a lower net biomass for the
same solids residence time objective, and may be par-
ticularly attractive as an operating capability as the plant
approaches design conditions.
163
-------
Increase the Aeration Intensity. Whether or not this strat-
egy has any value depends on the certainty of a non-
limiting aerobic nitrifying solids residence time. The
9,900-L/s (225-mgd) Metro Plant in Minnesota reports
that this strategy is successfully used to achieve its sea-
sonal performance objectives (67).
Breakpoint Chlorination Polishing. This strategy was often
conceptually reported in the early years of nitrogen con-
trol technology as the means of achieving seasonal com-
pliance. It is uncertain how many facilities use this
approach. Since it is an all-or-nothing concept, it may be
best applied with a split treatment approach to avoid ex-
cessive chlorine usage. The designer should review the
experience reported for breakpoint chlorination in Section
2.5.2.3 before proposing this concept, and anticipate the
probability of attendant dechlorination needs if the con-
cept is applied.
All seasonal nitrification concepts bring the accompany-
ing certainty of elevated nitrite nitrogen excursions and
the resultant chlorine demands if this is the designed
disinfection process. Section 2.8.5 provides additional in-
formation about this problem. Table 2-3 summarizes the
stoichiometric demand.
6.5 Nitrification in Attached Growth Reactors
Reactors in which growth occurs on or within a solid
medium are termed attached growth, supported growth,
or fixed film reactors. A number of attached growth reactor
configurations—including trickling filters, rotating biologi-
cal contactors, aerated biologial filters, and both packed-
and fluidized-bed systems—can be designed to incorpo-
rate nitrification. Descriptive and design and performance
information pertaining to each attached growth configura-
tion is presented in Sections 6.5.2 through 6.5.4.
6.5.1 Application of Biofilm Models to Design
The design of attached growth processes is primarily
based on empirical results from the operation of pilot and
full-scale systems. Examples of typical design criteria in-
clude volumetric loading rate (mass of substrate per unit
of reactor volume per time), surface loading rate (mass
of substrate per unit of media surface area per time) and
hydraulic retention time (residence time of feed material
in reactor, normally based on empty bed volume, i.e., total
volume occupied by the media). The inability to determine
the VSS concentration easily in most attached growth
reactors precludes the use of VSS specific removal rates
(e.g., mg ammonium oxidized per mg VSS per day) com-
monly employed in the design of suspended growth re-
actors. Although the attached growth reactor design
relationships are generally empirical, in most cases they
are consistent with biofilm models developed based on
stoichiometry, Pick's law, and Monod kinetics. The kinetic
considerations for attached growth systems, and related
diffusional resistance aspects, are discussed in Section
3.4.
The equations that describe the typical biofilm model
(Section 3.4) are presented elsewhere (68). Solving the
equations requires numerical techniques or the use of
pseudoanalytical solutions to approximate the numerical
solutions. A graphical, loading curve approach has been
developed which allows for sizing of an attached growth
reactor without solving the biofilm model (68). The ap-
proach involves the development of families of curves
representing flux (i.e., mass of substrate per unit of biofilm
surface per time) as a function of substrate concentration
in the bulk liquid of a completely mixed biofilm reactor or
reactor segment. The flux and substrate concentration are
expressed as normalized values using appropriate pa-
rameters for normalization. The value of these parame-
ters will be dictated by biofilm kinetic and hydrodynamic
considerations (e.g., diffusion coefficients). The graphical
representations, further simplified by methods proposed
by others (69), provide straightforward relationships for
substrate flux as a function of substrate concentration
under steady state conditions. A steady state mass bal-
ance on a completely mixed attached growth reactor or
reactor segment, yields the following equation:
aV
(6-25)
where:
Ja = substrate flux rate per unit biofilm surface area
Q = volumetric flow rate, volume/time
V = reactor segment volume, volume ;
S0 = feed substrate concentration, mass/volume
S-) = effluent substrate concentration, mass/volume?
a = surface area of biofilm per unit reactor volume,
area/volume :.:
Equation 6-25 can be used directly to calculate required
reactor volume, for given values of the wastewater sub-
strate concentration and the effluent substrate concentra-
tion. The calculation requires definition of the specific
surface area (a) of the support medium and determination
of Ja from the appropriate graphical representation. The
use of this procedure has been illustrated in the design
of an aerated biological filter for CBOD5 removal, assum-
ing in one case completely mixed reactor hydraulics and
in the other, plug flow hydraulics. The procedure is readily
adaptable to the design of nitrification reactors operating
at low CBOD5 feed concentration. Information pertaining
to another practical but fundamentally based approach to
the design of attached growth nitrification reactors is pre-
sented elsewhere (70).
If nitrification is to occur in a biofilm reactor, competition
from the heterotrophs for oxygen and space in the biofilm
must be reduced. Based on the biofilm model and prac-
tice, it has been proposed that the CBOD flux (i.e., J^ in
Equation 6-25, using CBOD as the substrate) must be
164
-------
less than approximately 1 g/m2/d (0.2 lb/1,000 sq ft/d) in
order to obtain reliable nitrification in a biofilm. This pro-
posal is based on measuring CBOD as ultimate CBOD
(71). Other modeling efforts (72), verified through data
from the literature, indicate that for nitrification to be sig-
nificant, oxygen must penetrate more deeply into the
biofilm than biodegradable organic material- In practice
this implies that at a bulk liquid oxygen concentration of
3 mg/L, the concentration of organic material must be
less than 20 mg/L as CBOD (72). Based on field obser-
vations of operating trickling filters, some authors have
suggested that the soluble BOD6 must be less than 20
mg/L before the onset of nitrification, without stating the
corresponding bulk liquid oxygen concentration (73). Oth-
ers claim that significant nitrification can be achieved at
bulk liquid soluble COD levels of up to approximately 60
mg/L without stating the biodegradable fraction of the
COD value (74).
Perhaps the most important implication of the biofilm
model on the design of attached growth reactors for ni-
trification is that below a certain bulk liquid ammonium
and DO level, the model predicts that the specific ammo-
nium oxidation rate will be reduced as a result of limiting
substrate diffusion effects, regardless of the significance
of competition from heterotrophic growth. That is, the spe-
cific ammonium oxidation rate is no longer zero-order; it
is dependent on the bulk liquid ammonium or DO con-
centration. As discussed in Chapter 3 (Section 3.3.1),
nitrifiers under these conditions are no longer operating
at their intrinsic growth rates. Measured ammonium volu-
metric and surface removal rates in attached growth re-
actors treating municipal wastewater with a low soluble
CBOD tend to confirm this theoretical implication (75-85).
In fact the measured or apparent rates in trickling filters,
based on influent and effluent data, appear to be more
influenced by substrate effects than would be expected
theoretically (84). This may be due to patchy or incom-
plete film growth at low bulk liquid ammonium concentra-
tions. Under such conditions the active surface area of
the media is less than calculated.
A recent attempt has been made to relate the solids
residence time in a trickling filter to the surface loading
rate, based on the use of results from field observations
and empirical design relationships developed by various
authors (74). The design information derived from this
effort is empirical, and the' observations should not be
considered design criteria. Generally, they are applicable
to attached growth reactors with a hydraulic regime closer
to plug flow than complete mix:
• The onset of nitrification will likely not occur unless the
soluble BOD5 surface loading rate is less than 9 g/m2/d
(1.8 lb/1,000 sqfl/d).
• .Nearly complete nitrification (effluent ammonium-nitro-
gen of approximately 2 mg/L) will typically be encoun-
tered at a soluble BOD5 loading rate of approximately
2 g/m2/d (0.4 lb/1,000 sq ft/d).
• Low-surface or volumetric ammonium removal rates
are normally encountered in attached growth reactors
that are designed to achieve low effluent ammonium-
nitrogen concentrations (1.5 mg/L or less). This is due
to the difficulty in building up and maintaining a signifi-
cant nitrifier biofilm in the latter stages of a plug flow
fixed film reactor, where the low ammonium concen-
tration limits nitrifier growth.
Therefore, while biofilm modeling can provide significant
insight into the design and performance of attached
growth reactors operating at a low influent soluble
CBOD5, other factors, such as the influence of particulate
influent organics and the nonhomogeneous biofilm thick-
ness and density characteristic to such attached growth
reactors, complicate application of theoretical models. A
reasonable conclusion is that the mechanisms of nitrifi-
cation in attached growth reactors operating at a higher
influent CBOD5 are so complex that design approaches
are necessarily empirical.
6.5.2 Trickling Filters and Biotowers
The trickling filter is an aerobic attached growth reactor
in which a solid surface medium is used to support biofilm
growth. Wastewater is normally introduced at the top of
this attached growth reactor through a distribution system
and flows or trickles down through the media. The dis-
tinction between trickling filters or biotowers and packed-
bed reactors, and aerated biological filters and fluidized
bed reactors is that iii the latter reactors, the hydraulic
design is such that the media are submerged in the re-
actor liquid during normal operation. Although trickling
filters are often designed to allow flooding or submer-
gence of the reactor media for maintenance purposes,
operation in this fashion is not routine.
Information pertaining to the application of trickling filters
and biotowers to municipal wastewater treatment in gen-
eral can be found in other publications (49,86). The fol-
lowing discussions are more specific to use of the
technology for nitrification of municipal wastewaters.
Trickling filter or biotower media traditionally consist of
rocks, slag, or synthetic materials. Rock and slag trickling
filters generally have 1.2 to 3.0 m (4 to 10 sq ft) of media
depth. Plastic media trickling filters are normally con-
structed at depths of 4.6 to 7.6 m (15 to 25 ft) because
of the lighter weight and better ventilation capabilities of
the packing (87). Recent advances in the development
of plastic and other media with different structural con-
figurations have made this technology more efficient and
cost effective. Structural configurations include assem-
blies of plastic corrugated sheets, random plastic ring
structures (e.g., pall rings), horizontal wood-slatted struc-
tures, and polyethylene strip media.
165
-------
Because of their advantages, synthetic media are typi-
cally selected for new plant design instead of rock or slag
trickling filters. Examples of synthetic media types that
are appropriate for nitrification applications are listed in
Table 6-17. Corrugated plastic sheet media fall into two
broad classifications, vertical and cross-flow configura-
tions (88). As shown in Figure 6-11, vertical media have
corrugations that direct flow primarily downward. Cross-
flow media consist of ridged corrugated sheets in which
the ridges on adjacent sheets are normally set at 45- or
60-degree angles to each other, and are touching where
they cross (Figure 6-11). The configuration provides the
opportunity for splitting and combining liquid flow. Studies
have shown that, when compared to vertical flow media,
the cross-flow configuration results in a longer hydraulic
retention or liquid contact, time (89) and appears to pro-
Table 6-17. Comparative Physical Properties of Example Synthetic Trickling Filter Media Suitable for Nitrification of
Municipal Wastewaters
Unit Weight
Specific Surface Area
Media Type
X-FLO 30
X-FLO 42
BlOdek
19050
BlOdek
10250
Accupac
CF-1900
Accupac
CF-1200
Tripac
Tripac
Tripac
BIO-PAC
BIO-PAC
BIO-PAC
BIO-PAC
Sessil 100
Sessil 130
Sessil 200
Characteristics
Plastic sheets, cross-
flow, corrugated
Plastic sheets, cross-
flow, corrugated
Plastic sheets, cross-
flow, corrugated,
60-degree angle
Plastic sheets, cross-
flow, corrugated,
60-degree angle
Plastic sheets, cross-
flow, corrugated,
60-degree angle
Plastic sheets, cross-
flow, corrugated,
60-degree angle
Random pack media,
polypropylene, 1-inch
diameter balls
Random pack media,
polypropylene, 2-inch
diameter balls
Random pack media,
polypropylene, 3.5-
inch diameter balls
Random pack media,
2-inch diameter balls
Random pack media,
1.5-inch diameter
balls
Random pack media,
1-inch diameter balls
Random pack media,
0.62-irtch diameter
balls
Flexible hanging
plastic sheets
Flexible hanging
plastic sheets
Flexible hanging
plastic sheets
kg/m3
24-45
24-45
24-45
24-45
27-63
36-63
99
67
53
53
61
72
112
N/A
N/A
N/A
Ib/cu ft
1.5-2.8
1.5-2.8
1.5-2.8
1.5-2.8
1.6-3.9
2.2-3.9
6.2
4.2
3.3
3.3
3.8
4.5
7.0
N/A
N/A
N/A
nvVm3
98
138
138
223
157
226
279
157
125
102
144
180
331
98
138
223
sqft/
cuft
30
42
42
68
48
69
85
48
38
31
44
55
101
30
42
68
Void
Space, %
>95
>95,
95
95
95
95
90
93
95
94
93
92
88
N/A
N/A
N/A
U.S.
Manufacturer
American Surfpac
Downington, PA
American Surfpac
Downington, PA
Munters
Ft. Myers, FL
Munters
Ft. Myers, FL
Brentwood
Industries
Reading, PA
Brentwood
Industries
Reading, PA
Jaeger Products
Houston, TX
Jaeger Products
Houston, TX
Jaeger Products
Houston, TX
NSW Corp.
Roanoke, VA
NSW Corp.
Roanoke, VA
NSW Corp.
Roanoke, VA
NSW Corp.
Roanoke, VA
NSW Corp.
Roanoke, VA
NSW Corp.
Roanoke, VA
NSW Corp.
Roanoke, VA
166
-------
Vertical Media
* t t
Cross-Flow Media
Figure 6-11. Hydraulic flow patterns in vertical and cross-
flow trickling filter media (adapted from Reference 88).
vide a higher oxygen transfer efficiency (88). These fea-
tures allow a substantially higher BOD5 surface loading
rate than can be imposed on vertical media for equivalent
effluent quality (88).
Rock or slag media trickling filters can be designed to
achieve nitrification by operating them at low organic
loading rates, provided that proper ventilation is achieved.
A trickling filter or biotower based on the use of horizontal
wood slats is currently popular in one commercial em-
bodiment of a combined suspended and attached growth
system. Information pertaining to this combined growth
reactor system is presented in Section 6.6.2.3. The use
of polyethylene strip media in the United States as the
packing for a trickling filter or biotower is a relatively new
development (90). One example is that currently manu-
factured by NSW Corporation of Roanoke, Virginia, and
referred to as Sessil™ media. The Sessil strip media are
2.9 cm (1.14in) wide and are hung from a wood support
structure located at the top of the biotower. The strip
media are typically 6 m (20 ft) in length. Operating results
from the application of this media to nitrification of mu-
nicipal wastewaters have yet to be reported.
The design and performance information presented in
Sections 6.5.2.1 and 6.5.2.2 is specific to conventional
trickling filters employing rock, slag, or plastic media for
the treatment of municipal wastewaters at high and low
CBOD5, respectively.
6.5.2.1 Nitrification Design and Performance
Information at High Carbonaceous Feed
Concentrations
The' number of operating trickling filters designed to
achieve nitrification of municipal wastewater containing a
high CBOD5 concentration, typical of primary treated
wastewater, is limited. In an assessment completed for
EPA in 1991,10 plants were identified that simultaneously
achieved CBOD5 removal and nitrification in a single trick-
ling filter unit, commonly referred to as combined or sin-
gle-stage carbonaceous oxidation-nitrification (87). Six of
the 10 plants used the trickling filter/solids contact
(TF/SC) flowsheet. Note that information pertaining to the
TF/SC flowsheet will be presented later in this section.
A degree of ammonium oxidation has been achieved for
many years in low or standard rate rock media trickling
filters (1). Results from numerous full-scale and pilot plant
studies (Figure 6-12) show that for these filters to attain
nearly complete nitrification (90-percent ammonium re-
moval), the organic volumetric loading rate must be lim-
ited to approximately 80 g BODg/rrvVd (5 lb/1,000 cu ft/d).
The reduction in nitrification at higher loadings was attrib-
uted to domination by heterotrophic bacteria in the micro-
bial film. It was also noted that there were very little data
to judge the effect of temperatures on performance below
20°C, but that lower temperatures would require a reduc-
tion in the volumetric loading rate. The results indicate
some improvement if recirculation around the rock filters
is practiced (Figure 6-12). The results from blast furnace
slag media trickling filters indicate that comparable volu-
metric loadings to those of rock filters are necessary to
achieve nearly complete nitrification based on results
where recirculation is not practiced. Imposition of a recir-
culation ratio of 1:1 to the slag media filters allowed nearly
complete nitrification to be achieved at loading rates over
1.5 times greater than the rock media. The results pre-
sented in Figure 6-12 should be considered approximate.
In addition to operating conditions differing from site to
site, it is not known whether the BOD5 results reflect
CBOD5 or include the effect of ammonium oxidation in
the BOD test.
Historically, the rock media used in trickling filters has
typically had a nominal size of 2.5 to 10 cm (1 to 4 in)
with a corresponding specific surface area of less than
65.6 m2/m3 (20 sq ft/cu ft) (49). The higher specific sur-
face areas of plastic media (Table 6-17) have resulted in
the ability to accomplish nitrification at higher volumetric
loadings, relative to rock media filters. Another factor fa-
voring greater capacity of the plastic media filters is oxy-
gen supply; rock filters often have poor ventilation,
particularly when water and air temperatures are close or
the same.
In studies completed at Stockton, California (91), good
agreement was found between the nitrification perform-
ance of plastic media trickling filters and previous results
with rock media when the performance was expressed
as a function of the BOD5 surface loading rate (Figure
6-13). The Stockton results indicate that 90-percent am-
monium removal was achieved at a BOD5 surface loading
rate (the reference is not specific in differentiating be-
tween BOD5 and CBOD5) of approximately 2.5 g/m2/d
(0.51 lb/1,000 sq ft/d) for both rock and plastic media
trickling filters (Figure 6-12). The corresponding BOD5
volumetric loading rates are 120 g/m3/d (7.5 lb/1,000 cu
ft/d) for the rock media and 220 g/m3/d (13.7 lb/1,000 cu
ft/d) for the plastic media.
167
-------
100
80
60
o>
•g
a
a.
40
20
0 No Recifculation
Q Recirculation
1 kg/rrfVd = 62.4 lb/1,000 sq ft/d
0.1 0.2 0.3 0.4 0.5 0.6
BOD- Volume Loading, kg/rr$d
0.7
0.8
0.9
1.0
Figure 6-12. Effect of BODs volumetric loading on nitrification performance of rock trickling filters (adapted from
Reference 1).
The rock media BOD5 volumetric loading rates reported
at Stockton and for other rock (Figure 6-13) and slag
media trickling filters are comparable; nearly complete
nitrification is accomplished at BOD5 loading rates in the
range from 80 to 120 g/m3/d (5.0-7.5 lb/1,000 cu ft/d).
The soluble BOD5 loading rate of 1.8 g/m2/d (0.4 lb/1,000
sq ft/d) referenced earlier in Section 6.5.1 for nearly com-
plete nitrification (effluent NHJ-N of approximately 2.0
mg/L) agrees with the Stockton results that on a total
BOD5 basis a surface loading of about 2.5 g/m2/d (0.51
lb/1,000 sq ft/d) is necessary for >90-percent ammonia
nitrogen removal with rock and plastic media.
100
80
60
BOD5 Surface Loading, lb/1,000 sq ft/d
1.0 2.0 3.0 4.0
40
20
Legend
49m2/m30 Rock Media (91)
Stockton Pilot Studyr
89 m2/!
Plastic Media (10)
Stockton Plant,
Plastic Media (91) -
Rock Media (84)
5 10 15
BODgSurface Loading, g/m2/d
20
Figure 6-13. Effect of BODs surface loading on nitrifica-
tion efficiency of rock and plastic media trickling filters
(adapted from Reference 73).
The BOD5 volumetric and surface loading rates that are
suggested for nearly complete nitrification are approxi-
mate. The actual loading rates that are required will be
influenced by such factors as the recirculation ratio, the
hydraulic surface loading rate, the filter depth, the specific
hydraulic pattern and retention timejn the trickling filter
(when plastic media are employed), and environmental
conditions. A more detailed discussion of the effects of
such factors on the performance of trickling filters in gen-
eral is presented elsewhere (49). Included is information
concerning appropriate strategies to overcome issues
common to trickling filters, whether they are designed for
carbonaceous oxidation or carbonaceous oxidation and
nitrification, such as biofilm sloughing cycles and the oc-
168
-------
currence of filter flies and snails. A portion of the biofilm
or biological slime will be sloughed either continuously or
periodically during normal trickling filter operation. Accu-
mulation of thick biofilms can result in oxygen limitations,
which will impair performance; continuous and uniform
sloughing, as measured by a relatively constant level of
SS from the trickling filter, provides one indication of a
well-operating unit.
Design Factors. Trickling filters are generally designed
with at least a minimum effluent recycle capability to
maintain a stable hydraulic loading during normal diurnal
variations (87). The studies completed at Stockton on
plastic media trickling filters (91) and those noted pre-
viously from operation of rock and slag media filters (1)
indicate an increase in nitrification efficiency when recir-
culation is practiced. Investigators at Stockton noted an
increase in nitrification performance from 67 to 87 percent
when the recirculation ratio was increased from 2.7 to 3.8
and air circulation was increased by operating more ven-
tilation fans. Both of these actions increased the DO con-
centration in the bulk liquid, resulting in performance
improvement. Effluent recycling can also cause some de-
gree of denitrification to occur in the upper portion of the
trickling filter (73).
The use of effluent recirculation, the magnitude of the
recirculation ratio and the recirculation configuration (e.g.,
before or after secondary clarification) are all factors af-
fecting trickling filter performance: There is general agree-
ment (49) that when consideration is given to both
economics and performance for the typical tricking filter
plant flowsheet (consisting of primary clarification, trick-
ling filter(s), and secondary clarification), the effluent
stream that is recycled should be the one after the trick-
ling filter(s), and selection of a recirculation ratio of 1.0 is
appropriate. Other hydraulic considerations include the
trickling filter hydraulic loading, and the design and op-
eration of the influent flow distributor. These considera-
tions are common to the design of trickling filters for either
carbonaceous oxidation or carbonaceous oxidation and
nitrification and are dealt with in detail elsewhere (49).
The value selected for the minimum hydraulic loading
should ensure complete media wetting under all influent
conditions. The value will depend on the characteristics
of the media employed. Typical values are 1 to 3 m3/m2/hr
(0.41 to 1.22 gpm/sq ft) based on the cross-sectional or
plan area of the filter. Results from one study (92) indicate
that no advantage is gained from effluent recirculation as
long as complete wetting is always accomplished.
The specific hydraulic pattern and retention time in plastic
media trickling filters will influence the BOD5 loading nec-
essary to achieve nitrification. As previously noted, plastic
media with cross-flow characteristics, when compared to
vertical flow media, increase the hydraulic retention time
or contact time between the biofilm and the influent and
provide superior oxygen transfer (88). The results from
pilot studies completed at Garland, Texas (Figure 6-14),
verify the claimed advantage for the cross-flow media,
based on nitrification performance as a function of BOD5
surface loading. The results were also compared to those
from the operation of rock media trickling filters at Stock-
ton, CA (91) and were found to be comparable (Figure
6-14). The results also indicated perhaps a slight effi-
ciency improvement for 60° vs. 45° cross-flow media, and
a significant efficiency advantage for the structured, cor^
rugated plastic media over randomly packed plastic ring
structures (73).
-------
as hydraulic loading. Alternatively, temperature sensitivity
relationships presented in Section 6.5.2.2 for trickling fil-
ters operating at low carbonaceous feed concentrations
are acceptable, although using this approach can lead to
a conservative design.
The importance of DO concentration in the operation of
all trickling filters highlights the need for sufficient venti-
lation. If enough passageways are provided, the differ-
ences in air and wastewater temperatures and humidity
differences between the ambient air and the air in the
trickling filter provide a natural draft. This mechanism may
provide the necessary aeration requirements on occa-
sion, but not consistently (49). After summarizing the lit-
erature regarding ventilation rates in trickling filters it was
recommended that for combined BOD5 removal and ni-
trification systems the air flow requirements are given by
(49,95):
Air flow = (150)(1.25 BOD5 + 4.6 TKN)(PF)/1,440
where air flow is in m3/min, and BOD5 and TKN are in
kg/d
Design Approach. Historically, engineers have selected
an appropriate BOD5 surface loading as a function of
temperature to design trickling filters for nitrification of
municipal wastewater at high CBOD5 (87). As discussed
earlier, a temperature correction is likely not warranted,
at least at temperatures above 15°C (59°F). The use of
a BODS surface loading rate, specific to the media to be
utilized in the biotower, remains an acceptable approach
to biotower sizing for nitrification. More recently, one de-
sign manual has used a manufacturer's empirical corre-
lation for calculating the average TKN surface removal
rate in illustrating an approach to determine the biotower
media requirements (49). TKN removal is due to the com-
bined effects of biomass growth and nitrification; the cor-
relation, as displayed on Figure 6-15, developed in
concert with consulting engineers (49), was derived from
an extensive array of operating results. Alternatively the
design can be supported by pilot plant studies completed
with the wastewater in question. Once a loading or re-
moval rate has been selected and the media surface area
requirements determined, the design procedure is basi-
cally in accord with that presented for the design of trick-
ling filters at low feed CBOD5 (Section 6.5.2.3).
Performance Information. In the assessment completed
in 1991 (87), 2 of the 10 nitrifying trickling filter plants
were identified as operating at a high carbonaceous feed
concentration. These plants are located in Amherst, Ohio,
and Wauconda, Illinois, respectively.
The wastewater treatment sequence at Amherst, Ohio,
involves screening, grit removal, primary clarification,
trickling filters, secondary clarification and chlorine disin-
fection (87). Two trickling filters are operated in series
with no intermediate clarification. Each filter is 27.4 m (90
ft) long, 12.2 m (40 ft) wide, and 5.2 m (17 ft) deep and
is packed with cross-flow plastic media.
Operating and performance information pertaining to the
trickling filters over the 12-month period from February
1989 to January 1990 is presented in Table 6-18. During
this period the wastewater flow was approximately 88 L/s
(2 mgd), equivalent to its design flow. The hydraulic load-
ing rate to the trickling filters averaged 23.0 m3/m2/d (565
gpd/sq ft), and no effluent recirculation was practiced. The
plant is required to meet an effluent ammonium-nitrogen
limit of 3 mg/L during the summer months and 6 mg/L
during the winter months. The raw wastewater tempera-
ture was less than 15°C (59°F) from February 1989
through May 1989, and again from November 1989
through January 1990, based on average monthly values
(87). The average monthly effluent ammonium nitrogen
values during these colder temperature periods ranged
from 1.8 to 4.9 mg/L (Table 6-18). The BOD5 media sur-
face loading ranged from 1.17 g/m2/d (0.24 lb/1,000 sq
ft/d) to 2.05 g/m2/d (0.42 lb/1,000 sq ft/d).
Over the entire period for which effluent NHJ-N data are
available (i.e., April 1989 through January 1990), the av-
erage effluent ammonia-N concentration was 2.5 mg/L,
the average raw wastewater temperature was 15.3°C
(59.5°F), and the average BOD5 surface loading was 1.36
g/m2/d (0.28 lb/1,000 sq ft/d). The Amherst results imply
a temperature dependency for nitrification below 15°C
(59°F). Operation and performance of the trickling filters
would have to be assessed in more detail to determine
whether the temperature effect implies the need for more
trickling filter surface area or the need for modification of
the operating conditions.
Contrary to the Amherst results, nitrification at the Wau-
conda, Illinois, facility appeared to be generally unaf-
fected by temperature, based on operating and
performance information for a two-year period beginning
in January 1987 (87). The effect of temperature may have
been masked by the low effluent ammonia values
throughout the period. Two cross-flow plastic media trick-
ling filters operate in parallel at Wauconda, receiving pri-
mary treated wastewater. Note that the Wauconda
flowsheet includes an aerated solids contact flocculation
step designed to improve solids capture during secondary
clarification. First developed in 1979 (96), this trickling
filter/solids contact (TF/SC) process has seen widespread
use through the 1980s. In its simplest form the TF/SC
process consists of a trickling filter, an aerated suspended
growth contact reactor, a flocculation zone, and a secon-
dary clarifier (Figure 6-16). Normally the contact reactor
is not designed to nitrify (see Section 6.6.2.2 for further
information).
Operating and performance information pertaining to the
Wauconda trickling filters during 1988 is presented in
Table 6-19. During this period the monthly average waste-
water flow was 29 L/s (0.67 mgd), which is less than 50,
170
-------
Standard Deviation ± 1.0
5 10 20
Applied BOD5:TKN
50
Figure 6-15. Correlation between TKN surface removal rate and wastewater BODs:TKN proposed by American Surfpac
Corporation (adapted from Reference 49).
percent of the plant's design flow. The hydraulic loading
rate to the trickling filters was only 0.58 m3/m2/hr (0.24
gpm/sq ft) and no effluent recirculation around the trick-
ling filters was practiced. The plant is required to meet
an effluent ammonia-nitrogen limit of 1.4 mg/L during the
summer months and 4 mg/L during the winter months. In
1988, the raw wastewater temperature was less than
15°C (59°F) from January through May, based on aver-
age monthly values (87). The average monthly secondary
effluent ammonia-nitrogen was 0.24 mg/L and the BOD5
media surface loading was 1.92 g/m2/d (0.39 lb/1,000 sq
ft/d) during this same five-month period (Table 6-19). The
performance of the Wauconda plant may have been in-
fluenced by nitrification occurring in the contact reactor,
as a result of the low flow condition and the introduction
of sloughed trickling filter nitrifiers.
In the assessment completed in 1991 (87), detailed per-
formance information for periods of 12 months or more
was reported for five nitrifying trickling filter plants oper-
ating at a high carbonaceous feed concentration. The
effluent ammonia nitrogen as a function of BOD5 surface
loading for these plants is plotted in Figure 6-17, based
on the mean of the reported average monthly values. The
Wauconda and Ashland plants both incorporate a solids
contact step prior to secondary clarification. These results
imply that BOD5 media surface loading rates below 1.0
g/m2/d (0.2 lb/1,000 sq ft/d) may be best to ensure resid-
ual ammonium-nitrogen concentrations less than 2.0
mg/L.
In summary, the following process factors are some that
affect the design of trickling filters for nitrification at high
carbonaceous feed concentrations:
• Hydraulic loading
• Hydraulic pattern and retention time in the filter media
• DO concentration in the filter liquid
• pH and temperature of the filter liquid under certain
conditions
• Feed TKN concentration
• Feed BOD5 concentration
6.5.2.2 Nitrification Design and Performance
Information at Low Carbonaceous Feed
Concentration
Many operating trickling filters in the United States are
designed to treat municipal wastewaters that receive a
low carbonaceous feed concentration. Many of these
trickling filters are the second stage of a two-stage trick-
ling filter plant with intermediate clarification. When oper-
ated in this fashion, the trickling filter is often referred to
as a separate-stage or tertiary trickling filter. Tertiary trick-
ling filter nitrification has been classified in one municipal
wastewater treatment design manual (49) by conditions
wherein the influent wastewater BOD5:TKN is less than
1.0 and the SBOD5 is less than 12 mg/L.
In general, low levels of organics in the influent to at-
tached growth reactors can be advantageous. Ideally,
such conditions result in low effluent SS, potentially elimi-
171
-------
Table 6-18. Amherst, Ohio, Wastewater Treatment Plant Carbonaceous Oxidation-Nitrification System Operating Condi-
tions and Performance Results, February 1989 to January 1990 (Adapted from Reference 87)*
Month
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Jan.
Waste-
water
Flow
Us
(mgd)
91
(2.06)
106
(2.42)
120
(2.74)
106
(2.42)
102
(2.32)
90
(2.04)
69
(1.57)
67
(1.52)
67
(1.52)
78
(1.78)
77
(1.75)
101
(2.29)
BOD5, mg/L
Primary
Effluent
89
62
62
62
40
61
80
67
56
59
99
48
Secondary
Effluent
12
11
8
10
8
8
7
6
5
4
6
6
TSS, mg/L
Primary
Effluent
74
56
88
118
216
196
192
183
132
99
129
89
Secondary
Effluent
12
12
9
15
10
9
8
9
7
8
9
10
Ammonia-N, mg/L
Primary
Effluent
12.2
14.3
9.3
7.6
8.9
14.3
18.4
17.1
17.0
17.8
—
—
Secondary
Effluent
—
—
3.6
2.2
2.0
1.2
1.6
1.9
1.4
1.8
4.9
4.3
Trickling Filter BOD5 Loading
g/m2/d Ob/
1,000 sq ft/d)
2.05
(0.42)
1.66
(0.34)
1.86
(0.38)
1.66
(0.34)
1.03
(0.21)
1.37
(0.28)
1.37
(0.28)
1.12
(0.23)
0.93
(0.19)
1.17
(0.24)
1.91
(0.39)
1.22
(0.25)
g/m3/d (Ib/
1, 000 cu ft/d)
200
(12.5)
163
(10.2)
184
(11.5)
162
(10.1)
101
(6.3)
136
(8.5)
136
(8.5)
112
(7.0)
93
(5.8)
114
(7.1)
189
(11-8)
120
(7.5)
* Average monthly values are presented. Primary effluent values are calculated assuming 35-percent removal of BOD5 and TSS in primary clarifier.
The primary effluent Is the feed to the trickling filters. Trickling filter surface loading is based on total media surface area of the two trickling filters
operating In series.
Primary
tlllUUHl
Trickling Filter
/
-------
.Table 6-19. Wauconda, Illinois, Wastewater Treatment Plant Carbonaceous Oxidation-Nitrification System Operating Con-
ditions and Performance Results, January 1989 to December 1989 (Adapted from Reference 87)*
Wastewater
Flow
BOD5, mg/L
TSS, mg/L
Ammonia-N, mg/L Trickling Filter BOD5 Loading
Primary Secondary Primary Secondary Primary Secondary g/m2/d (Ib/ g/m3/d (Ib/
Month L/s(mgd) Effluent Effluent Effluent Effluent Effluent Effluent 1,000sqft/d) 1,000 cu ft/d)
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
44
(0.99)
36
(0.83)
35
(0.80)
44
(1.00)
28
(0.64)
24
(0.55)
24
(0.55)
27
(0.61)
24
(0.55)
21
(0.48)
26
(0.60)
21
(0.48)
78
77
101
76
146
128
123
124
109
122
104
115
15
9
11
9
10
8
8
10
9
9
•15
12
70
125
79
104
81
73
59
68
75
81
60
61
10
4
14
20
8
9
4
7
9
6
17
22
12.4
17.6
17.0
14.4
14.1
15.6
14.3
13.3
16.2
17.0
14.6
13.1
0.4
0.3
0.3
0.1
0.1
0.2
0.1
0.3
0.3
3.6
0.7
0.3
1.90
(0.39)
1.56
(0.32)
2.00
(0.41)
1.86
(0.38)
2.30
(0.47)
1.76
(0.36)
1.66
(0.34)
1.86
(0.38)
1.47
(0.30)
1.47
(0.30)
1.56
(0.32)
1.37
(0.28)
188
(11.71)
155
(9.69)
196
(12.26)
185
(11.53)
227
(14.17)
171
(10.68)
164
(10.26)
184
(11.47)
146
(9.09)
142
(8.88)
152
(9.47)
134
(8.37)
'Average monthly values are presented. The primary effluent is the feed to the trickling filters. Secondary effluent is after solids contacVflocculation
clarification steps. Trickling filter surface loading is based on media surface area.
12.5
10-
7.5..
5 .-
w 2.5..
+ Palm Springs, CA (>17°C, Slag Media)
o Wauconda, IL (11 °C-21 °C, Plastic Media)
^ Ashland, OH (13°C-22°C, Plastic Media)
a Amherst, OH (8°C-20°C, Plastic Media)
» Chemung County, NY (11C°-16°C, Rock Media)
1.0 2.0 3.0 4.0
BOD5 Surface Loading, g/rr?/d
5.0
Figure 6-17. Effect of BODs surface loading on nitrification performance (adapted from Reference 87).
173
-------
nating the need for a downstream solids-liquid separation
step (as discussed In Section 6.3.1). Low levels of or-
ganics in the influent result in little biomass production,
which in turn leads to reduced biofilm formation. Pluggage
of voids in the media and resulting bypass or short cir-
cuiting flow patterns are of less concern than when op-
erating at a higher carbonaceous feed concentration.
Nitrification in plastic media trickling filters operating at
low CBODswas first reported in the period from 1973 to
1975 and the results are incorporated into the first EPA
Process Design Manual for Nitrogen Control (1). The re-
sults are expressed in the form of design curves which
specified the surface area required per mass of ammo-
nium oxidized per day as a function of effluent ammo-
nium. Although lacking a fundamental basis, these curves
were used for design for several years after the Manual
was Introduced in 1975. Since that time, research and
development efforts have improved our understanding of
how various factors affect the performance of the trickling
filter when operated at a low CBOD5 concentration.
By operating at a low CBOD5, the influence of bulk liquid
DO and ammonium-nitrogen concentrations, environ-
mental factors (i.e., temperature and pH), and other op-
erating and design conditions (e.g., hydraulic loading rate
and media configuration) on the nitrification process can
be better understood. Nitrifiers are dominant because of
the reduced competition with heterotrophic growth for DO
and other nutrients. Still, previously mentioned (Section
6.5.1) factors such as nonhomogeneous biofilm thickness
and density will generally prevent direct application of
biofilm models for design purposes.
Design Factors. The rate of nitrification in attached growth
reactors, in the absence of significant CBOD5, is influ-
enced by the concentration of both ammonium and DO
In the liquid phase. Depending on their concentration val-
ues, either substrate can dictate the oxidation rate of
ammonium per unit surface area. The nitrification rate no
longer is zero-order with respect to ammonium concen-
tration when the concentration reaches a critical value.
Although somewhat controversial, this value appears to
be in the range from 2 to 5 mg/L NHJ-N (84,97). Ammo-
nium surface removal rates may appear diminished at
lower ammonium concentrations; the reason for the lower
rate, however, may be ineffective surface area due to
poor biofilm coverage at the location in the filter where
the ammonium concentration is the lowest (84). This ap-
parent reduction in rate still implies the need for more
media surface area, or specific operating procedures that
will overcome the issue of incomplete biofilm surface cov-
erage.
At high ammonium concentrations (i.e., greater than 5
mg/L NH|-N), nitrification is limited by oxygen mass trans-
fer across the liquid film and not by the DO concentration
in the bulk liquid phase, according to some researchers
(98). These results imply that oxygen availability in certain
regions of the trickling filter appears to regulate the nitri-
fication rate. Data compiled from five separate studies
(97) indicate that as the ammonium-nitrogen load is in-
creased beyond a critical value, the corresponding re-
moval rate is not predictable (Figure 6-18). The reasons
for the scatter in the data above a critical ammonium-ni-
trogen loading value of approximately 1 g/m2/d (0.2 Ib/sq
ft/d) (Figure 6-18) could be due to differences in oxygen
transfer as a result of different media configurations (i.e.,
vertical versus cross-flow), or due to differences in envi-
ronmental and operating conditions at the different sites
where the studies were completed. The scatter also could
be due to the problem of using average rates when cor-
relating the data (84). The authors attempted to estimate
bulk liquid DO conditions for the different data sets while
separating out the effects of media configuration, operat-
ing conditions, and environmental factors (97). Figure 6-
19 is typical of the results of their analysis, where one
data set is presented from operations using vertical media
at constant hydraulic load and recycle conditions, and at
one temperature condition. The results (Figure 6-19) im-
ply a constant correlation between ammonium-nitrogen
surface loading and removal rate for a given bulk liquid
DO concentration.
Based on the previous discussion of the effects of DO
and NHJ-N concentration on ammonium removal rates,
the NHJ-N profile in a trickling filter operating at low in-
fluent CBOD5 can be predicted. The typical profile will be
a straight line reduction of NHJ-N at a removal rate con-
trolled by oxygen availability in the upper portions of the
filter. Trie removal rate will decrease as the rate-limiting
factor becomes ammonium rather than oxygen diffusion.
The inflection point, although dictated by oxygen mass
transfer, temperature, and degree of biofilm coverage on
the media, will typically occur when the ammonium-nitro-
gen concentration is in the range from 2 to 5 mg/L. Bulk
DO concentrations that are less than 70-percent satura-
tion (i.e., approximately 6 to 7 mg/L) reflect inadequate
trickling filter ventilation (98).
The practice of storage and control of ammonium-laden
anaerobic digester supernatant return streams to the
trickling filter has been proposed as a means of leveling
out the ammonium load to the filter during a 24-hour
period (84). This practice should ensure more consistent
ammonium penetration to the lower filter depths, ideally
eliminating patchy biofilm coverage. A related biofilm con-
trol procedure proposes regular flooding and backwash-
ing to eliminate the growth of certain predators such as
filter fly larvae, which may cause a decrease in nitrifica-
tion rates, and to prevent.repetitive biofilm sloughing (84).
Such procedures may eliminate the need for post clarifi-
cation or filtration, depending on effluent quality require-
ments (84). Others have proposed the use of high-
intensity flushing to control the growth and development
of filter flies and, possibly, snails (49).
174
-------
4.0
f
o
I 2.0
cc
1.0
O, D, A, O Midland
j[_ • Bloom Township
, Anpp
V Sunnyvale
A Garland (Duck Creek)
100% Removal Line
i i i i i i i i i i i i i i i i i i i i i i i i i i i i
0 0.5 1.0 1.5 2,0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
Ammonium-N Loading, g/m2/d
Figure 6-18. Ammonium surface loading versus removal rate (adapted from Reference 97).
4.0+
2.0- -
1.0'
Effluent NH 7-N
j^ 0-1 mg/L
v 1-2 mg/L
p 2-3 mg/L
O 3-4 WQH-
^ 4-5 mg/L
• 5-6 mg/L .
+ 7-10 mg/L
Temperature: 30°C .
Hydraulic Loading Rate: 3.13 nvVm^/hr (1.28 gpm/sq ft)
Estimated Bulk Liquid DO'
7.6 mg/L
5 mg/L
Srrig/L
1.0
2.0
3.0
4.0
5.0
Ammonium-N Loading, g/m2/d
Figure 6-19. Effect of bulk liquid DO conditions on surface loading versus removal rate correlation (adapted from Refer-
ence 97).
175
-------
The effect of temperature on the nitrification rate in trick-
ling filters operating at low CBOD5 will be influenced by
such factors as oxygen availability, influent and effluent
ammonium concentration, and hydraulic loading condi-
tions; thus the reported effect of temperature is variable
(85,99). Zero-order or maximum rates of nitrification ap-
pear to be more influenced by oxygen mass transfer char-
acteristics of the media (84), which in turn are influenced
only moderately by temperature as opposed to intrinsic
nitrifier growth rates. These observations are consistent
with the biofilm model previously noted (Section 6.5.1). A
comparison of the effect of temperature on zero-order
rates versus rates limited by various factors is presented
in Rgure 6-20. The higher nitrification rates at Central
Valley are attributed partially to the use of cross-flow
media, which has oxygen transfer characteristics superior
to those at Midland and Lima, where vertical flow media
are used. The biofilm control procedures adopted in the
Central Valley pilot study were also credited for the higher
nitrification rates. The poor performance at Zurich was
attributed to poor wetting and biofilm coverage (84-).
When nitrifying trickling filters, which are operating at low
influent CBOD5 and near neutral pH conditions, are tem-
porarily exposed to a lower pH, a significant short-term
(less then two days) negative effect on the nitrification
rate will be observed, according to Swiss researchers
(93). Based on other scientists' experience (100), these
researchers concluded that the pH dependence is likely
3,5
3.0
2.5
2.0
1.5
1.0
0.5
Central
Valley
Lima
Midland
Zurich
5 10 15 20 25
Temperature, °C
Rgure 6-50. Effect of temperature on nitrification rates in
trickling filter systems (adapted from Reference 96).
only temporary. The Swiss findings are particularly rele-
vant to nitrifying trickling filters exposed to significant vari-
ations in influent ammonium concentrations when
operating at low effluent alkalinity values (less than 50
mg/L as CaCO3) (93).
Although authors have presented data indicating that a
higher hydraulic loading increases the nitrification rate
(97,99), the hydraulic loading required to maximize the
rate is unknown (49). The data compiled from the five
separate studies discussed earlier indicate that a hydrau-
lic loading above 3.6 m3/m2/hr (1.47 gpm/sq ft) produced
the highest ammonium-nitrogen surface removal rates
(97). Interpretation of the data is complicated by how the
interaction between hydraulic loading and oxygen avail-
ability affects the rates.
The characteristics of the medium employed will dictate
the minimum hydraulic loading needed both to ensure
complete wetting and to affect the ammonium-nitrogen
volumetric removal rate. The superior oxygen transfer
characteristics and higher specific surface area of cross-
flow media favor its use over vertical flow media for nitri-
fication of wastewaters containing a low influent CBOD5.
One author recommended the use of medium density,
cross-flow media (specific surface area of 138 m2/m3, or
42 sq fi/cu ft) over a higher density alternative (specific
surface area of 223 m2/m3, or 68 sq ft/cu ft), citing issues
of poorer wetting characteristics and a greater tendency
for plugging associated with the higher density material
(84). Less efficient wetting characteristics of the higher
density media result in a lower ammonium-nitrogen sur-
face removal rate, since a portion of the surface area is
ineffective. Swiss researchers recommend the selection
of a media with a specific surface area in the range from
150 to 200 m2/m3 (46 to 61 sq ft/cu ft) (101). High density
media trickling filters have proved to be effective at Reno-
Sparks, Nevada (102). The difference may be explained
by the Reno-Sparks investigators' observation that graz-
ing organisms were not a problem, as they were for the
Swiss investigators. Problems with grazing organisms at
pilot scale with fixed distributors may not be realized at
full scale with rotary distributors (103).
The; fact that plugging is often associated with higher
density media implies the need for low CBOD5 and TSS
concentrations in the biotower influent. The role of CBOD5
in reducing ammonium-nitrogen surface removal rates as
a consequence of heterotrophic growth competition, has
been discussed. Extensive biomass growth also can af-
fect the hydraulic integrity of the biotower causing influent
short-circuiting or media plugging, or requiring flow by-
passing. Researchers have found that particulate CBOD5
affects nitrification rates to the same negative degree as
soluble CBOD5 (104).
Design Approach. The preceding discussions imply that
trickling filters operating with low feed CBOD5 should be
176
-------
designed to maximize the fraction of the media volume
operating at the maximum nitrification rate, or the rate
that is not limited by the concentration of NHJ-N. Further-
more, the design should provide the maximum hydraulic
loading possible without the need for recirculation and
should provide adequate air flow to maximize oxygen
transfer to the liquid.
Given these considerations, trickling filters should be con-
structed as deep towers with lower cross-sectional areas,
and should be designed with forced air ventilation. Nitri-
fication biotowers have been constructed with depths as
great as 12.8 m (42 ft); shallower towers can be employed
by operating two units in series. It has been demonstrated
that if biotowers are designed in series, better perform-
ance can be expected, as long as the order of operation
of the two units can be reversed every few days (101).
This operating protocol will lead to development of a more
uniform biofilm in the two towers. Ensuring air flow
throughout the trickling filter requires excess air distribu-
tion within the filter (49). It has been suggested that during
normal and low NHJ-N loading periods, the minimum air
flow rate should supply 50 kg O2/kg of O2 required (49).
During peak loading periods a value of 20 to 30 kg O2
supplied per kg O2 required is claimed to be acceptable
(49). -
The procedure used to design nitrifying trickling filters at
low feed CBOD5 requires specification of the following
parameters:
1. The full range of NHJ-N surface removal rates across
the biotower under the anticipated temperature con-
ditions.
2. The type of media and its oxygen transfer charac-
teristics.
3. The minimum biotower hydraulic loading rate and the
corresponding maximum tower surface area.
4. Any minimum or maximum biotower depth limitations.
5. The oxygen requirement and the minimum air flow
necessary to meet that requirement.
6. The design of the influent distributor.
Approaches to determine the required design parameters
have been proposed by researchers, consultants, and
system suppliers (49). One approach (84,96), supported
by theoretical considerations and pilot plant results, is
based on the efforts of Swiss researchers (70) and others
(92). The approach involves the use of the following equa-
tion to determine the NHJ-N surface removal rate at any
point in the biotower:
JN(Z,T) = JN
(6-26)
JIM (Z,T) = NHJ-N surface removal or nitrification rate
at depth Z and temperature T, g/m2/d
Z = depth in tower, m
T = temperature, °C
JN max = maximum NHJ-N surface removal rate at
temperature T, g/m2/d
N = bulk liquid NHJ-N concentration, mg/L
KN = saturation constant, mg/L
k = empirical parameter describing decrease in
rate with depth
JN max will be dictated by the oxygen transfer efficiency of
the media employed according to:
E(J0
x)
(6-27)
WN max 43
where:
E = media effectiveness factor
Jo max = maximum surface oxygen transfer rate,
g/m2/d
The factor 4.3 in Equation 6-27 represents the mass of
oxygen consumed per unit mass of ammonium-nitrogen
oxidized. Jog max is calculated from the Logan model (92).
Departure of E from a value of 1.0 reflects incomplete
media wetting, impact of predators on the biofilm, or com-
petition from heterotrophs for oxygen (84). The term
N
T;—— in Equation 6-26 is equivalent to a Monod satu-
KM + N
ration term and was used by the Swiss researchers to
describe the influence of ammonium concentration on the
nitrification rate (70). The term e"*2 was used to describe
the nitrification rate decline observed with depth and at-
tributed to poor biofilm growth.
Two solutions to Equation 6-26 have been presented as
follows (70):
H
or for the case where k=o:
z a JN max ..,
) = Nj _ N
{6.28)
(6-29)
where:
where:
a = specific surface area of trickling filter media,
m2/m3
vh = hydraulic loading of trickling filter media
(flow/cross-sectional area), m3/m2/d
NJ = ammonia-nitrogen concentration applied to the
filter, g/m3
Where recirculation is used to obtain the desired hydraulic
loading on the support media, the recirculation ratio, r,
alters the value of ammonia nitrogen applied to the filter.
The relationship is:
Ni,± (6-30)
177
-------
Table 6-20. Calculated Trickling Filter Nitrification Model Parameters from Pilot Plant Studies (Adapted from Reference 84)
JNmax
Period
04/17/87-05/21/87
06/19/87-07/23/87
07/31/87-09/03/87
10/23/87-11/19/87
11/25/87-12/10/87
12/21/87-01/07/88
01/01/88-02/25/88
Test
Number
1
2
3
4
5
6
7
g N/m3/d
2.1
2.9
2.8
3.2
2.3
2.3
2.6
Ib N/1,000 sq
ft/d
0.43
0.59
0.56
0.65
0.46
0.47
0.53
k,
rrf1 (ft'1)
0
0.075 (0.023)
0
0.16 (0.05)
0
0
0
E
0.66
0.89
0.83
0.99
0.71
0.72
0.81
Temp.,
°C
15.5
20.0
21.5
18
15
12
11
Table 6-21. Comparison of Measured and Predicted Nitrification Rates (Adapted from Reference 84)
Plant Reference
Midland, Ml 97,98
Lima, OH 99
Bloom Twp., IL 100
Zurich, Switz. 100
(3.9 m tower)
Zurich, Swilz. 93
(6.8 m tower)
Media
Code
VFc27
VFc27
VFC27
VFd28
XFa68
XFa68
Period of
Designation
Run 9A
Run 12
TP2
TP5
TP6
TP7
9/22-10/10
10/11-10/20
N/A
N/A
N/A
Temp.,
°C
13
7
18
21
22
22
20
17
17-20
17-20
13
Apparent
Zero-Order
Nitrification Rate,
g N/m2/d
(Ib N/1,000 sq ft/d)
1.2
0.93
1.2
1.8
1.5
1.2
1.2
1.1
1.6
1.2
1.1
(0.24)
(0.19)
(0.25)
(0.37)
(0.31)
(0.24)
(0.24)
(0.22)
(0.32)
(0.24)
(0.23)
Predicted
Maximum Rate*
g N/m2/d
(Ib N/1, 000 sq ft/d)
1.4
1.3
1.4
1.4
1.4
1.5
1.3
1.3
1.7
3.1
2.8
(0.28)
(0.26)
(0.28)
(0.28)
(0.28)
(0.32)
(0.27)
(0.27)
(0.35)
(0.63)
(0.58)
E
0.86
0.74
0.88
1.30
1.10
0.76
0.88
0.82
• 0.89
0.38
0.3
* Predicted from Equations 6-26 and 6-27.
If reclrculation is used, an iterative solution is necessary
because recycle effects are included in both the Nj and
VH terms (84).
Values for the terms in Equations 6-26 and 6-27 were de-
veloped through profile sampling of a pilot trickling filter
during studies at the Utah Central Valley Water Reclamation
Facility (84,96). These are summarized on Table 6-20.
The value of KM averaged between 1 and 2 mg/L during
the pilot study (84). The JN max rates .presented in Table
6-20 are the same rates presented in Figure 6-20 for
Central Valley. The value of k in Table 6-15 varied from
0, Implying no rate decrease with media depth, to 0.16.
The range in the value of k is significant.
The nitrification design model was used to analyze certain
data collected from previous studies (84). The data se-
lected (Table 6-21) were from periods when the biotowers
appeared to be operating in the zero-order region
throughout their depths, as evidenced by effluent ammo-
nium levels of 3.0 to 5.0 mg/L. The measured rates thus
represent the apparent maximum or zero-order nitrifica-
tion rate. The predicted maximum rates were calculated
from Equations 6-26 and 6-27, using the Logan model
(92) to calculate J0a max for the specific corrugated sheet
media employed. The measured and predicted rate's
show reasonable agreement for the vertical flow media
data (Table 6-21, VF0 and VFd media), reflected by values
of E near 1.0. The low E values for the cross-flow media
(Table 6-21, XFa media) are believed by the authors to
be caused by poor media wetting. The medium employed
in the Zurich studies was the high density material
(specific surface area of 223 m2/m3, or [68 sq ft/cu ft])
cited previously.
178
-------
The vertical flow media nitrification rates are considerably
lower than those reported from the Central Valley studies
(Table 6-20). The medium employed at Central Valley was
medium density, cross-flow material. The differences in
the media, together with other operating factors, likely
account for the difference in rates. In addition to selection
of an ideal media, several other factors are proposed by
the authors as being important to attaining these high
rates (84):
• High quality secondary effluent (i.e., average TSS and
CBOD5 less than 15 mg/L) as influent to the biotower.
• Provision of flooding capability for prevention of filter
fly larvae growth.
Table 6-22. Annual Operating Information from Five Nitri-
fying Trickling Filters Receiving Low Influent CBOD
(Adapted from Reference 96)
Plant*
A
B
C
D
E
Influent
Wastewater
Flow, L/S
(mgd).
572 (13.01)
70(1.60)
70 (1.60)
582 (13.24)
823 (18.72)
Nitrification
Rate,
gN/m2/d
(lbN/1,000
sqfl/d)
0.29 (0.059)
0.29 (0.059) .
0.20 (0.041)
0.39 (0.080)
0.34 (0.070)
Temp.,
°C
14-24
9-19
14-20
10-23
12-27
a
'§
CD
3
§
10
1.0
0.5
0.4
0.3
0.2
0.1
12 5 10 20304050607080 90 95 98 r
Cumulative Frequency,.% ,
Figure 6-21. Nitrification performance on trickling filters
receiving low enfluent CBODs—Plants A, B, and C (adapted
from Reference 96).
* Plants A, B, and D are located in the Midwest. Plants C and E are
located in the West.
• Use of forced ventilation to prevent stagnation.
• Use of any ammonium-laden supernatant return flows
to even out ammonium loads.
• Use of regular backwashings for removal of SS accu-
mulations; the need for tertiary clarification also may
be eliminated.
Performance Information. Recent surveys identify ap-
proximately 20 operating trickling filters in 1990 that were
designed for nitrification of municipal wastewater at, low
influent CBOD5 (87,96). These trickling filters served as
the tertiary treatment step for trickling filters or activated
sludge systems designed for secondary treatment. De-
tailed operating and performance information suitable for
evaluation was obtained from five of these plants (96).
Annual operating information for the five plants is pre-
sented in Table 6-22. Performance information, in the
form of monthly average effluent ammonia-nitrogen con-
centration plotted against the cumulative frequency, is
presented in Figures 6-21 and 6-22. The results indicate
that all plants were achieving less than 2.0 mg/L ammo-
nianitrogen 50 percent of the time, with four of the plants
at less than 2.0 mg/L 90 percent of the time. The plants
1.0
E 0.5
•| 0.4
o
I 0.3
| 0.2
£
0.1
1 2 5 10 30 50 70 80 90 95 98
Cumulative Frequency, %
Figure 6-22. Nitrification performance of trickling filters
receiving low influent CBODs—Plants D and E (adapted
from Reference 96).
were all operating at somewhat conservative ammonium
surface loading rates. Utilization of conservative practices
in the design of tertiary nitrifying trickling filters to date is
also reflected by the use of a solids-liquid separation
step following the biotower (87,96). Approximately 90
percent of the plants employ either effluent filtration or
clarification.
179
-------
6.5.2.3 Design Example No. 3: Nitrification in a Trickling Filter at Low
Carbonaceous Feed Concentration
This design example is based on information provided in Section 2.9 for the more complex Plant B. The
calculations address sizing requirements to meet the less stringent effluent limits provided in Table 2-10
(with no limit for total N). A trickling filter is to be added to treat the existing secondary effluent. The plant
flows and unsettled secondary effluent quality correspond to the values presented in Table 6-4. As in Design
Example No. 2, assume that 21 mg/L of ammonia nitrogen are available for nitrification. For purposes of
this example, assume also that the existing secondary effluent has the characteristics summarized in Table
6-23. Additional design information used for this example is also summarized in Table 6-23.
Table 6-23. Design Information for Nitrification of a Low Carbonaceous
Feed Concentration and Less Stringent Effluent Requirements
Wastewater Flow Characteristics, m3/d (mgd)
Raw wastewater average flow 18,925(5.0)
Total secondary effluent average flow 21,055(5.5) ":!
Actual Secondary Effluent Concentrations, mg/L
Soluble COD 27
Nitrogen available for nitrification 21
Alkalinity as CaCO3 120
Trickling Filter Reactor Effluent Characteristics, mg/L
Soluble COD 20 ,,-•
Ammonia nitrogen 1.5
Design Conditions/Assumptions
Reactor temperature,°C 15
Reactor pH range 7.0-7.6
Air flow rate (at average secondary loading)
kg O2 supplied/kg O2 required 50
The process design steps are as follows:
1. Determine the process design factor to be used in sizing the trickling filter. Following from Chapter 2
(Table 2-18), the peaking factor developed to account for influent loading variations for attached growth
systems Is 1.92. However, for this situation assume that flow buffering due to upstream unit operations and
effluent recirculation, the effluent quality requirement, anticipated variations in process conditions, and the
uncertainty in the kinetic approach warrant an overall process design factor of 1.5. In sizing a trickling filter,
this determination is equivalent to specifying a design nitrification rate (JN des) equal to the maximum,
nitrification rate divided by the process design factor.
2. Verify that the trickling filter reactor operating pH can be expected to be in the assumed range of 7.0-7.6,
(Table 6-23). Alkalinity destroyed can be calculated from the coefficient presented in Table 3-1 as:
(7.1 mg CaCCVmg N oxidized)(21 mg/L oxidizable N) = 149 mg/L as CaCO3
Alkalinity available (Table 6-23) is 120 mg/L (as CaCO3). It is reasonable to assume that the pH will remain
in the required range since forced air ventilation is employed and a minimum residual alkalinity of 50 mg/L
(as CaCO3) is maintained. On this basis, the required supplemental alkalinity (ignoring any residual NHJ-N)
will be:
(149 mg/L destroyed + 50 mg/L minimum residual) - 120 mg/L available = 79 mg/L (as CaCO3) to be -
supplemented.
The mass of alkalinity required under average and maximum day conditions can be determined following ,
the approach presented in Design Example No. 1 (Section 6.4.2.2).
180
-------
6.5.2.3 Design Example No. 3 (continued)
3. Select the trickling filter media and establish the maximum NHJ-N surface removal rate JN max. Select a
medium density cross-flow media with a specific surface area of 138 m2/m3 (42 sq ft/cu ft). Assume biotower
effluent recirculation is available to maintain a total influent to the biotower equivalent to the total secondary
effluent average flow. Select an average flow hydraulic loading to the media sufficient to provide adequate
wetting. For this example choose a hydraulic loading rate of 80 m/d. Based on pilot or full-scale operating
information and application of Equations 6-26 and 6-27, assume JN max is 2.0 g/m2/d and k equals 0 when
Knis 1.5 mg/L
4. Calculate a maximum design nitrification rate based on the selected process design factor. For this
example:
des =
(6-31)
JN des = 2.0/1 .5 = 1 .33 g NH
Other design parameters already established in this example are:
a=138m2/m3
vh = 80 m/d
NJ = 21.0 mg/L at r = 0
KN = 1.5 mg/L
5. Determine the required media depth. Since k = 0, Equation 6-29 is applicable. To provide an additional
margin of safety, design for an effluent ammonia-nitrogen concentration of 1.5 mg/L. For this example:
des
{ft>-
-^W2'-1^1-5'"1^"-10-2
6. Calculate the total biotower cross-sectional area required based on the desired hydraulic loading rate.
The area is:
7. Determine the number of biotowers and their dimensions. Assume four equally sized biotowers are
selected providing two parallel systems, which will each operate with two trickling filters in series. Assume
that provisions will be made to reverse periodically the biotowers' order of operation in each of the two
parallel systems. For this example, each biotower's area is:
Area = 263/2 = 131.5 m2 (1,415 sq ft)
For circular systems the diameter of each biotower is:
m2
d = 12.9m (42.5 ft)
Since each parallel train consists of two biotowers in series, the depth of each biotower is:
Depth = 10.2 m/2 = 5.1 m (16.7ft) •
8. Determine the media volume required. Media volume can be calculated from:
Media volume = (263)(10.2) = 4(131 .5)(5.1) = 2,683 m3 (94,740 cu ft)
181
-------
6.5.2.3 Design Example No. 3 (continued)
9. Determine oxygen and air flow requirements. As discussed in Section 6.5.2.1 , the recommended air flow
requirements for combined BOD5 removal and nitrification systems are given by (49,95):
m3/min = 150(1 .25 BOD5, kg/d + 4.6 TKN, kg/d)(PF)/1 ,440
For design of tertiary nitrifying trickling filters receiving low levels of influent BOD5, the recommendation
was given that the minimum air flow should correspond to 50 kg O2 supplied/kg O2 consumed (49). Fifty
kg of O2 corresponds to 180 m3 of air, which implies that this recommendation also was intended to include
any additional COD removal that would occur when treating a high quality influent. The recommendation
was also given that during peak loading periods, the oxygen supply should be 20 to 30 kg O2/kg of O2
used (49).
At average conditions, the nitrogenous oxygen demand (ignoring the small residual NHJ-N concentration)
would result in an air flow requirement as follows:
(180)(21 055X21X4.6)
(103)(1,440)
= ^
The peak day nitrogen pollutant mass specified in Chapter 2 (Table 2-12) is 1.7 times the average day
value. Even if there was no attenuation in the peak day loading by virtue of the preceding activated sludge
system, an air supply of 254.2 m3/min (8,980 scfm) would still supply 29.4 kg (64.8 Ib) O2/kg nitrogenous
oxygen demand, which is within the range recommended for the peak loading period. Therefore, base the
design of the aeration system on average conditions.
6.5.3 Rotating Biological Contactors
Rotating biological contactors (RBCs) have been used for
treatment of municipal wastewaters since the 1960s.
Most of the current installations are for carbonaceous
BOD removal only. To a lesser extent, systems have been
constructed for combined organic removal and nitrifica-
tion or for separate stage nitrification.
Two previous EPA publications, Design Information on
Rotating Biological Contactors (105) and Review of Cur-
rent RBC Performance and Design Procedures (106),
discuss the design and operation of RBCs in detail. The
former publication presents a detailed design procedure
for carbonaceous as well as nitrification applications. Lit-
tle has changed in the design and construction of RBC
facilities in the last 10 years. The design approaches
described in these two publications are still commonly
used and valid today. This section presents design pro-
cedures that are based on those discussed in these pub-
lications along with supplemental information from recent
literature and from the major manufacturers.
6.5.3.1 Process Description
An RBC is a fixed film process in which plastic media are
mounted as discs or spiral wound sheets to a rotating
horizontal shaft. The shaft is mounted in a tank such that
approximately 40 percent of the media surface is sub-
merged during its rotation. The contactors are typically
up to 3.6 m (12 ft) in diameter, with shaft lengths up to
8.2 m (27 ft). Size restrictions of the units are due primar-
ily to transportation limitations.
The support medium has a surface configuration, gener-
ally unique to each manufacturer, that maximizes surface
area and enhances turbulence and flow patterns across
the surface. The media are generally available in two
types: a standard medium provides approximately 9,290
to 9,660 m2 (100,000 to 104,000 sq ft) of surface area
per single shaft (3.6 m diameter by 7.6 m [12 ft by 25 ft]
shaft length), and a high density medium provides surface
areas of 13,940 to 14,490 m2 (150,000 to 156,000 sq ft)
for a similarly sized unit. Because spacing between me-
dium surfaces in high density media is decreased, appli-
cations of these media are limited to situations where
biomass is expected to be thin, such as in separate-stage
nitrification systems or the last stages of carbonaceous
or combined carbon oxidation-nitrification systems,
An RBC process train generally consists of several RBCs
in series. The size limitation of an individual RBC unit
requires two or more parallel trains to be employed for
plants greater than 44 L/s (1 mgd) in capacity. Figure 6-23
presents a typical application of RBCs for municipal
wastewater treatment.
The mass transfer of substrate (BOD5 and/or ammonia-
nitrogen) and oxygen is due to the rotation of the partially
submerged discs through the wastewater contained in the
182
-------
Primary
Influent
Primary
Clarification
RBC Stage Number
2
RBC —Underflow Baffle
Secondary
Clarification
Primary
Sludge
Secondary
Sludge
••Secondary Effluent
to Disinfection
Sludge Handling
"*" and Disposal
Figure 6-23. Typical application of rotating biological contactors to municipal wastewater treatment.
tank. The constant rotation of the media through the
wastewater alternately exposes the biofilm, which devel-
ops on the media, to substrate and air. The shafts of
full-scale RBCs (3.6 m [12 ft] diameter) are rotated at one
to two revolutions per minute (rpm); a peripheral velocity
of 18.3 m/min (60 fl/min) is most common (1.6 rpm). In
addition to promoting substrate and oxygen diffusion, the
rotation also regulates biofilm growth by sloughing exces-
sive growth from the discs.
Excessive biofilm growth, generally the most common
symptom of inadequately designed RBC systems, is at-
tributable to excessive first-stage or total-system organic
loading. Excessive growth reduces treatment efficiency
by reducing biofilm surface area and limiting oxygen
transfer. Low DO levels in the biofilm and bulk liquid
promote the development of nuisance organisms, primar-
ily Beggitoia, a sulfur-reducing bacteria. Such organisms
ean:eoat the biofilm surface, severely reducing the effec-
tive surface area available for carbon oxidation and nitri-
fication. Excessive growth can also structurally damage
shafts and media because of the added weight. The prob-
lems of excessive biological growth due to high organic
loading and corresponding low DO in the first stage of
RBG systems may be overcome by using a step-feed
hydraulic flow pattern, an enlarged first stage, and/or sup-
•plemental aeration (107). Nitrification systems require low
^organic loads to promote the development of nitrifying
bacteria. This requirement substantially reduces the po-
tential for excessive growth development and prevents
some of the common problems encountered by carbon
oxidation systems.
All carbonaceous oxidation or combined carbon oxida-
tion-nitrification RBC systems require secondary clarifica-
'tibn for the removal of sloughed biofilm. While a recycle
flow, such as is employed in activated sludge processes,
is flpt necessary, recirculation may enhance nitrification
performance by providing more optimum C:N throughout
the»system (108).
A sufficient biofilm normally develops in approximately
two weeks from startup during normal operation at tem-
peratures above 15°C (59°F). A normal biofilm for a com-
bined carbon oxidation and nitrification system is
generally a brown-gray film in the first and second stages
and a thin, reddish-brown slime in the latter stages (109).
The latter stages may also have a slight algal growth. The
color of biofilm on separate stage nitrification systems has
been characterized as tan to bronze (110). The color
darkens as film thickness increases. Biofilm thickness
reaches equilibrium in 25 to 60 days. In combined carbon
oxidation-nitrification systems, thicker films are .observed
generally in the first and second stages where organic
loadings are at a maximum and carbon oxidation occurs.
Thinner films generally develop on the latter stages where
nitrification occurs.
The active portion of the biofilm is controlled by the dif-
fusion of oxygen and substrate into the film. The depth
of active biomass has been reported to range from 20.to
600 urn (105); greater depths do not contribute to sub-
strate removal. The development of excessively thick
biofilms reduces active RBC surface area and affects
treatment capacity. Biofilm thickness should be controlled
to minimize excessive buildup.
Biofilm thickness can be indirectly monitored using load
ceils to weigh RBC shafts. These devices can be either
permanently installed on the shaft or used periodically.
Excessive biofilm is controlled by natural sloughing. Man-
ual control is needed occasionally and is accomplished
by reverse rotation, air injection, and/or increased rota-
tional speed while the train is still in service. Out-of-serv-
ice methods include caustic cleaning or chlorination.
6.5.3.2 Design Considerations
The two critical design parameters for RBC systems are
the determination of disc area and staging requirements.
While each manufacturer provides his own design curves
and guidance, all are based on similar procedures. The
basic information needed covers design loads ,(flow, or-
ganic, and nitrogen), load variability, and effluent limits.
183
-------
6.5.3.3 Kinetics of Organic Substrate Removal
The organic substrate removal rates of many biological
treatment processes for municipal wastewater are gener-
ally described as following first-order kinetics. The princi-
pal* influence on the reaction rate is the complex nature
of municipal wastewater. During exposure to the mixed
culture that develops in most biological treatment sys-
tems, many compounds degrade readily, some degrade
slowly, and others resist treatment.
Organic removal rates for RBCs have been characterized
In the literature by zero-, first-, and second-order reac-
tions. All manufacturers recommend staging RBCs when
low effluent BOD5 is required, which implies reaction ki-
netics greater than zero-order. Zero-order kinetics apply
only when RBCs are meant to provide a low level or a
roughing level of treatment (40- to 60-percent BOD5 re-
moval). In this case, although the removal of BOD5 may
follow a zero-order reaction rate, it is generally accepted
that the availability of DO controls BOD5 removal.
A first-order kinetic expression was developed based on
an analysis of influent and RBC effluent data from five
pilot and full-scale facilities (111,112). The expression is
as follows:
(6-33)
where:
Sa = secondary effluent total BOD5, mg/L
S| = RBC influent total BODS, mg/L
V » media volume, m3
Q ~ hydraulic loading, m3/min
K areaction constant
The media volume term (it by radius2 by length) does not
include the impact of total available media surface area,
which was found to be a factor also affecting perform-
ance. This expression can be used to develop a family
of design curves similar to those shown in Figure 6-24.
Generally, first-order kinetics best describe design curves
utilized by RBC manufacturers.
A second-order kinetic approach for predicting BOD5 re-
moval has been developed (105,113,114). This approach
Implies that the removal of soluble BOD5 across a series
of RBCs is proportional to the square of the residual
soluble BODs concentration. The equation is as follows:
2kt
(6-34)
Cn s soluble BOD5 in nth stage, mg/L
k s second-order reaction rate constant, L/mg/hr
t3 hydraulic residence time in the nth stage, hr
CM = soluble BOD5 entering the nth stage, mg/L
The second-order reaction rate, k, was found to be 0.083
L/mg/hr based on an analysis of full-scale RBC data for
municipal wastewater. This approach uses soluble BOD5
100
90
^. 80
I
§70
8
m
I
60
50
40
75 mg/L
10 15 20 25 30
Total BOD5 Load, g/m2/d
35
100
Eso
df
O
CD 60
§40
1-20
O
RBC
Influent
BODB
10 15 20 25 30
Total BOD5 Load, g/m2/d
35
Figure 6-24. RBC design curves developed from Equation
6-33 (adapted from Reference 111).
for design and is not sensitive to primary clarifier perform-
ance relative to the level of solids (particulate BOD5)
entering the RBC system. Some consideration should be
given to particulate BOD5, which can exert an oxygen
demand through entrapment of S3 in the biofilm and their
subsequent hydrolysis. The soluble BOD5 within the sys-
tem will thereby increase. A hydrolysis factor could be
used, if known; alternatively, a maximum total BOD5 load-
ing condition can be imposed, which typically would be
30 g BODs/m2/d (6 lb/1,000 sq fl/d). This second-order
expression also can be used to develop a family of curves
similar to those developed with the first-order approach.
Pilot studies, using full-scale equipment (3.6 [12 ft] m
diameter disks), should be performed for wastewaters not
considered "typical" municipal wastewaters. These would
include industrial wastewater and municipal wastewater
184
-------
with a significant industrial component. The results of the
pilot studies can be analyzed, using the first- or second-
order expressions, to develop a series of wastewater-spe-
cific design curves.
Design criteria developed through pilot studies using re-
duced-scale equipment should be used cautiously. Higher
substrate removal rates have been reported for pilot sys-
tems with RBC diameters less than 3.6 m (12 ft). The
higher rates are due to enhanced oxygen transfer effi-
ciencies, which result from the higher rotational speeds
needed to maintain peripheral velocities of 18.3 m/min
(60 ft/min). Appropriate scale-up factors may be needed
when pilot studies are used to develop design data.
6.5.3.4 Kinetics of Nitrification
Nitrification kinetics for RBCs are influenced by tempera-
ture, organic load, and effluent ammonia-nitrogen require-
ments. At NHJ-N concentrations greater than about 5
mg/L, the rate of nitrification for RBCs is oxygen-transfer
limited down to a temperature of 13°C (55°F) (49). Below
this temperature, corrections must be made to the nitrifi-
cation rate. Most manufacturers provide design curves for
temperature correction. Figure 6-25 is an example of
these corrections for both nitrification and BOD5 removal.
The development and balance of heterotrophic and autot-
rophic (nitrifying) organisms differs for combined and
separate stage systems. In nitrifying activated sludge sys-
tems the nitrifier fraction of the mixed bacterial culture is
controlled by the ratio of ammonia nitrogen to BOD. Simi-
larly, the nitrifier fraction of the biofilm on an RBC is a
function of the same ratio. This ratio will change with each
preceding stage in an RBC system, which explains the
3.0
2.5
1 1.5
1.0
gradation in biofilm characteristics across an RBC train,
especially in combined systems.
Organic load influences the rate of nitrification through
the competitive development of heterotrophs and autot-
rophs. Autotrophs, which are slower growing than het-
erotrophs, can dominate the mixed culture population on
a RBC only when the development of heterotrophs is
substrate limited. The following relationship between or-
ganic loading and the rate of nitrification has been devel-
oped (116):
f, = 1.43-0.1M
(6-35)
when:
(4.3 < M < 14.3)
where:
fj = fraction of maximum ammonia removal rate
M = organic substrate load, g COD/m2/d
This relationship has been used to develop Figure 6-26.
As shown, this analysis suggests that the nitrification rate
approaches the maximum rate when the BOD5 loading is
less than approximately 3.5 g BOD5/m2/d (0.7 lb/1,000 sq
ft/d).
A significant degree of nitrification will not occur on RBCs
until the soluble BOD5 is less than 15 mg/L. This phe-
nomenon is due to the autotroph/heterotroph population
dynamics. A review of Figure 6-24, with extrapolation of
the effluent BOD5 vs. loading curves to zero, shows that
15 mg/L BOD5 is reached at loadings of approximately
3.5 g/m2/d (0.7 lb/1,000 sq ft/d). This agrees with the
previous analysis of Equation 6-35 (Figure 6-26).
Nitrification
BOD 5 Removal
40
45 . 50
Wastewater Temperature, °F
55
Figure 6-25. Temperature correction factors for required RBC area (adapted from Reference 115).
185
-------
1.00
0.80
0.60
0.40
0.20
\
\
\
COD
5.0 10.0 15.0
Organic Substrate Loading Rate, g/m2/d
Note; Assume CODSOD = 2:1.
Rgure 6-26. Effect of organic substrate loading on the
rate of nitrification (adapted from Reference 116).
The rate of nitrification has also been shown to approach
a first-order relationship with respect to the effluent con-
centration at effluent ammonium-nitrogen levels less than
5 mg/L; the maximum zero-order removal rate occurs at
levels above 5 mg/L. This function is demonstrated
graphically in Figure 6-27, which uses data from three
municipal nitrifying RBC plants (105). Manufacturers have
incorporated this into their design procedures by specify-
ing nitrogen removal rates as a function of effluent re-
quirements.
4-r
3 *" *
1 --
Legend:
O Gladstone, Mich.
a Guelph, Ontario
A Cleves, Ohio
6.5.3.5 Empirical Procedures
Empirical design approaches are still the most common
design procedures used for determining RBC media area
requirements. Fundamental design procedures and
mathematical models have been developed that incorpo-
rate kinetic parameters, mass transfer of substrate and
DO, and changes in RBC physical configuration due to
biofilm growth and sloughing. These fundamental ap-
proaches are complex and often have not been ade-
quately demonstrated; they are generally not used for
design purposes because of their complexity.
The prevalent empirical design approaches are based on
substrate and hydraulic areal loading rates. Substrate
loading rates control oxygen demand and biofilm thick-
ness. Substrate loading parameters for carbon oxidation
systems include total and soluble BOD or COD, while
loading parameters for nitrification systems include am-
monium nitrogen, TKN, and soluble organic nitrogen.
Generally, nitrification designs should use TKN as the
basis for sizing RBCs. Hydraulic loading rates are gen-
erally based on specific influent substrate concentrations;
as such, they indirectly become a substrate loading rate.
First-stage organic loading conditions are considered the
most critical factor in the design and operation of RBCs
for carbonaceous removal or combined carbon oxidation-
nitrification. Substrate loadings and oxygen demand are
at maximum levels in the first stage. Oxygen demand due
to excessive substrate loadings can easily exceed oxygen
n
O p
o
Note: Temperature = 55°F± 2°F
468
Stage NH ..-N Concentration, mg/L
10
12
Figure 6-27. Full-scale RBC nitrification rates at a wastewater temperature of 13°C (55°F) (adapted from Reference 105).
186
-------
transfer capacity if the first, stage is not sized properly. It
is not unusual for the first stage of a properly designed
carbonaceous oxidation system to contain up to 50 per-
cent of the total system's surface area.
The principal design criterion, with respect to organic re-
moval, is not to exceed soluble BOD5 loadings of about
12.2 g/nf/d (2.5 lb/1,000 sq ft/d) and total BOD5 loads of
31.2 g/m2/d (6.4 lb/1,000 sq ft/d) for any individual stage.
Maximum oxygen transfer rates for full-scale RBCs were
reported to be 6.8 to 7.3 g O2/m2/d (1.4 to 1.5 lb/1,000
sq fl/d) (117). These rates correspond to a maximum BOD
removal capacity of 13.7 to 14.6 g/m2/d (2.8 to 3.0
lb/1,000 sq ft/d) at a presumed COD:BOD of 2:1. Thus,
the maximum soluble BOD5 removal rate generally re-
ferred to in the literature, 12.2 g/m2/d (2.5 lb/1,000 sq
ft/d), is slightly less than the rate that can be supported
by oxygen transfer.
The maximum oxygen transfer rate of 6.8 to 7.3 g O2/m2/d
(1.4 to 1.5 lb/1,000 sq ft/d) also translates to a maximum
nitrogen removal rate of approximately 1.45 to 1.61 g
N/m2/d (0.3 to 0.33 lb/1,000 sq ft/d), which agrees with
the1 maximum zero-order rate shown in Figure 6-27. The
empirical design approaches used by many manufactur-
ers are based on maximum substrate loading rates and
design .curves similar to those in Figure 6-24.
6.5.3.6 Staging Requirements
Substrate (organic and nitrogen) removal by RBCs is
generally described as a zero- or higher-order reaction.
Staging becomes important when low effluent BOD5 or
ammonia-nitrogen concentrations are required; a point at
which the removal rate is often described as a first-order
reaction. Many of the manufacturers recommendations
for staging are a function of substrate removal or effluent
concentration (115,118-120). Table 6-24 presents exam-
ples of manufacturer recommendations.
6.5.3.7 Load Variability Considerations
RBC systems are typically designed for average flow and
loads (BOD5 and ammonia nitrogen) when the ratio of
peak to average design flows is less than 2.5 to 1. In
cases where the ratio is greater than 2.5 to 1, the average
flow/substrate design loading should be reduced by a
factor proportional to the increased variability above the
2.5:1 ratio. Alternatively, flow equalization should be in-
corporated into the process train.
6.5.3.8 Flow Distribution
Flow distribution is an important design consideration for
RBC systems that employ several parallel trains. The lack
of positive flow measurement and control promotes
maldistribution of flow among the trains, leading to over-
loaded and underloaded RBC trains within the same sys-
tem. Poor flow distribution results in variable biofilm
thickness among the first and middle stages of parallel
systems.
6.5.3.9 Design of Combined Systems
The most common application of RBCs for nitrification is
for combined carbon oxidation and nitrification, in which
the media area requirement is the sum of the media
requirements for carbon oxidation and nitrification. The
design of these systems must consider organic loading
as well as nitrogen loading. Insufficient first-stage RBC
area leads to the development of nuisance organisms,
which can effectively reduce the active media area. When
this occurs in the first stage, additional organic removal
must then be accomplished in the latter stages. Nitrifica-
tion activity is displaced further down the train. If there is
insufficient stage/area capacity, nitrification performance
may fail. As such, proper sizing of the RBC system for
carbon oxidation is critical for successful nitrification. Siz-
ing should be checked against maximum expected loads
to ensure that recommended stage and total system load-
ing are not exceeded under these conditions.
Table 6-24. Manufacturer Recommendations for RBC Staging (Adapted from Reference 105)
Carbon Oxidation Nitrification
Envirex (118)
Lyco (115)
Effluent BOD5
>25 mg/L
15-25 mg/L
10-15 mg/L
<10 mg/L
<40% removal
35-65% removal
60-85% removal
80-90% removal
. Number
Stages
1
1-2
2-3
3 or 4
1
2
3
4
Effluent
Ammonia-N
5 mg/L
<5 mg/L
<40% removal
35-65% removal
60-85% removal
80-95% removal
Number
Stages
1
Based on first-
order kinetics
1
3
3
4
187
-------
One design approach described in this section parallels
the approach specified by a major manufacturer. Other
manufacturers' design methods may differ slightly, but
most depend on using typical areal loading rates (hydrau-
lic and substrate mass) and design curves tempered with
maximum loading rates.
When designing a combined system, the sizing of the first
stage for carbon oxidation is critical. The first stage should
be designed for a maximum total BOD5 loading of 31.2
g/ma/d (6.4 lb/1,000 sq ft/d). The soluble BOD5 loading
should not exceed 12.2 to 19.5 g/m2/d (2.5 to 4.0 lb/1,000
sq ft/d). Under these loading conditions, a BOD5 reduction
of 40 to 50 percent can be expected, based on a maxi-
mum BODS removal rate of 13.7 to 14.6 g/m2/d (2.8 to
3.0 lb/1,000 sq ft/d). Conservative design would limit the
soluble BODS loading to 12.2 g/m2/d (2.5 Ib/sq ft/d). ,
The total area for subsequent carbonaceous oxidation
stages is then determined based on the first-stage efflu-
ent BOD5 load and a target soluble BOD5 concentration
of 15 mg/L from the last stages of the carbon oxidation
portion of the total system. In other words, one wants a
soluble BOD5 of less than 15 mg/L before entering the
nitrification stages. This removal efficiency is used with
the first stage effluent BOD5 concentration to determine
the design hydraulic loading (m3/m2/d) from design curves
similar to Figure 6-28. The wastewater flow and the hy-
draulic loading rates are then used to determine the RBC
area required for carbonaceous removal for the sub-
sequent stages. The total area requirement for carbon
oxidation is the sum of the first stage and subsequent
stage areas. Some manufacturers provide design curves
based on effluent BOD5 concentration rather than on per-
cent removed. These curves should be used in the same
manner to determine design loadings.
RBC area requirements for nitrification should be based
on TKN loads, Nutrient uptake and effluent loads should
be subtracted from the influent nitrogen load to determine
the mass of nitrogen to be removed. The overall nitrifica-
tion rate is selected based on the target effluent ammo-
nium-nitrogen concentration. One approach to computing
the additional media area required for nitrification is use
of a manufacturer supplied design curve such as Figure
6-29. Alternatively, the relationship of the nitrification rate
to effluent ammonium nitrogen should be similar to Figure
6-27; the rate data in this,figure and the mass of nitrogen
to be removed provide another.method to compute the
RBC area required for nitrification.
The total area requirement is the sum of the areas re-
quired for carbon oxidation and nitrification. For combined
carbon oxidation/nitrification applications, a minimum of
100
95
Influent BOD5, mg/L
180
150
85
80
75
0.02 0.04 0.06 0.08 0.1 0.12
Hydraulic Load, rrP/mz/d
0.14
0.16
Figure 6-28. Typical RBC design curve (adapted from Reference 118).
188
-------
Wastewater Temperature >55 °C
Influent NH+-N, mg/L
30 25 22 20 19 1817 16 15 14
0 0.04, 0.08 0.12 0.16
Hydraulic Loading Rate, m^m ^d
Figure 6-29. Nitrification design relationships (adapted from Reference 121).
0.20
four stages is generally required. Interstage loads should
be considered when RBC disc area is being allocated to
the stages. Interstage loads beyond the first stage should
not exceed 29 g TBODg/nf/d (6 Ib TBODg/1,000 sq fi/d)
or 12.2 g SBODg/nrfrd (2.5 Ib SBODs/1,000 sq fl/d).
Area and staging requirements should be developed for
summer and winter conditions, and should consider sea-
sonal effluent limits. RBC area requirements for carbon
oxidation and nitrification should be temperature cor-
rected separately, based on manufacturer supplied pro-
cedures such as shown in Figure 6-25. The design should
be based on the condition requiring the most surface
area.
6.5.3.10 Design Approach for Separate-Stage
Nitrification
The design of RBCs for separate-stage nitrification is
similar to the nitrification design for combined carbon oxi-
dation-nitrification systems. Influent soluble BOD5 to a
separate-stage process should be less than 15 mg/L;
otherwise, additional surface area should be provided for
carbon oxidation. Design should be developed for winter
and summer conditions, with the condition requiring the
most area controlling final design. Provision for periodic
flow reversal (e.g., once a week) can also lead to a higher
capacity to handle diurnal NHJ-N fluctuations because of
higher nitrification potential in the last and usually NHJ-N
limited stage(s) (122).
189
-------
6.5.3.11 Design Example No. 4: Nitrification with an RBC System at High
Carbonaceous Feed Concentration
The following design example illustrates RBC design for a combined carbon oxidation-nitrification system.
Effluent limits are for the more stringent criteria specified in Table 2-10. Assume that the plant contains
primary clarification, anaerobic sludge digestion and tertiary filtration such that the primary effluent charac-
teristics and overall plant flow are the same as summarized in Table 2-16. A summary of wastewater
characteristics, effluent limits, and design conditions and assumptions is presented in Table 6-25.
Table 6-25. Design Conditions for Example No. 4
Wastewater Flow Characteristics, m3/d (mgd)
Raw wastewater average flow 18,925 (5.0)
Total influent average flow 21,055 (5.5)
Primary Effluent Characteristics, mg/L
COD 187a 168b
CBOD5 97a 87b
Soluble CBODS 53a 48b
TSS 80a 72b
TKN 29.5a 26.6b
Alkalinity, mg/L (as CaCO3) 120b
Secondary Effluent Permit Limits, mg/L
CBODS 10°
TSS 10°
NHJ-N 2°
Total N 5°
Design Conditions/Assumptions
Reactor temperature, °C 15
Reactor pH range 7.0-7.6
Nitrogen available for nitrification, mg/L equivalents 24.5 (22.0 mg/L)b
•Concentration value at average conditions expressed as mg/L equivalents (Table 2-16).
b Concentration value at average conditions at total flow of 21 ,055 m3/d.
0 From Table 2-10 for year-round monthly maximum permit limits.
1. Evaluate peaking factor requirement. Since the peak to average flow ratio is less than 2.5:1, the design
loadings will be based on average conditions.
2. Evaluate temperature correction. Since the design temperature is greater than 13°C (55°F), the tempera-
ture correction factor from Figure 6-25 for both carbonaceous oxidation and nitrification is 1 .0.
3. Determine media requirements to reduce soluble BOD5 to 15 mg/L. Do not exceed loading limits in the
first stage of 29 g total BODg/m2 (6 lb/1,000 sq ft) or 12.2 g soluble BOD5/m2 (2.5 lb/1,000 sq ft).
(18,925)(97)/29 = 63,300 m2 (681,400 sq ft) required based on total BOD5
(18,925)(53)/12.2 = 82,215 m2 (885,000 sq ft) required based on soluble BOD5
A minimum of nine standard density shafts (9,290 m2 [1 00,000 sq ft]) are required for the first stage providing
a total of 83,610 m2 (900,000 sq ft) of surface area. This minimum requirement corresponds to a first-stage
loading of:
21,055m3/d .„
(9X9.290 mV-
When using figures such as Figure 6-28 for design, a common assumption is that effluent BOD would be
50-percent soluble. Hence to achieve a soluble BOD5 of 15 mg/L would require an overall BOD5 removal
190
-------
6.5.3.11 Design Example No. 4 (continued)
of 65 percent. Based on an extrapolation of the influent BOD5 curves shown in Figure 6-28, a hydraulic
load of about 0.15 m/d should be adequate. Therefore, a total of 140,400 m2 (1,511,000 sq ft) of media
surface is required.
Initally assume a total of five trains with two shafts per train in each of the first stages and one shaft per
train in the second stage. Check the design against the predictions from Equation 6-34 to see if there is
reasonable agreement. At the standard 0.0049 m3/m2 of media (0.12 gal/sq ft) tank sizing, the hydraulic
residence time in the first and second stages is as follows:
(0.0049)(9,290 m2/shaft)(2)(24)
1~ 21,055/5 ~ ' T
t2 = ^- = 0.26hr
From Equation 6-34:
-1 + V1 + 4(0.83)(0.052)(48)
1~ 2(0.52)(0.083) •" '
_ -1 + V1 + 4(0.083)(.26)(23?7y
2~ 2(.26)(0.83)
Consider modifying the design to a total of 18 shafts and 6 trains to ensure adequate carbonaceous
oxidation. Under these conditions, Equation 6-34 would predict soluble BOD5 concentrations of 22.3 and
15.8 mg/L from Stages 1 (12 shafts) and 2, respectively. Additional safety would be provided to deal with
the sustained peak month loading.
4. Determine additional media requirements for nitrification. The soluble, nondegradable TKN from the RBC
equals about 1.0 mg/L and the TKN associated with the waste biological solids equals approximately 3.0
mg/L. The permit limit is 2 mg/L of NHJ-N; to provide a margin of safety this design is for an effluent
concentration of 1 mg/L. Assume that the equivalent NHJ-N concentration available for nitrification is equal
to 22.6 mg/L (26.6 - 4) at the total influent flow including plant recycles of 21,055 m3/d.
For an effluent NHJ-N concentration of 1 mg/L, a hydraulic loading of about 0.057 m3/m2/d (1.4 gpd/sq ft)
would be adequate, according to Figure 6-29. An additional media requirement of 369,400 m2 (3.98 x 106
sq ft) results, which corresponds to an additional 40 shafts of standard density media (9,290 m2 or 100,000
sq ft per shaft) or 27 shafts of high density media (13,940 m2 or 150,000 sq ft per shaft).
As an alternate approach, Figure 6-27 indicates a maximum nitrification rate of 1.46 g NHJ-N/m2/d (0.3
Ib/sq ft/d) would be observed down to about 5 mg/L of NHJ-N. Hence to reduce an influent concentration
of 22.6 mg/L to 5 mg/L requires:
(22'6".5lf'°55) = 253,800 m2 (2.73 x 10" sq ft)
1.4b
At 1 mg/L NHJ-N, the reaction rate from Figure 6-27 is only about 0.5 g/m2/d (0.1 lb/1 ,000 sq ft/d). To reduce
5 mg/L NHJ-N to 1 mg/L at this reaction rate (i.e., the rate that would be observed in one complete mix
stage) would require an additional:
-880 m2 (1 "85 x 1°6 sq ft)
The total requirement for nitrification would be 425,700 m2 (4,58 -x 106 sq ft). This translates to 46 standard
density shafts or 31 high density shafts.
191
-------
6.5.3.11 Design Example No. 4 (continued)
Influent
T
Stage
1
4
5
l l l l I l
I I I I I I
IIIIIIIH
ii i nun
IIIIIIHI
inn mi
IIHiiiii
iiimiii
Ulllllll
iiiiiiiii
iiiiiini
IIMIIIII
IIIIIIHI
IIIIIIHI
111111
111111
111111
111111
IIIIIIHI
IIIIIIIII
l l l I I I
I I I I I I
U)
Q.SS
'
I
ca
55
W--B
Q CD
£2
O)
Figure 6-30. RBC configuration for Design Example 4.
Figure 6-30 provides a convenient layout for this design example showing 18 shafts for carbonaceous BOD5
removal down to a soluble BOD5 of 15 mg/L and 30 high density shafts for NHJ-N removal. Combining two
shafts in the first stage prevents organic overloading and placing two shafts in the third stage will not affect
NHJ-N removal rates since the bulk liquid NHJ-N concentration is >5 mg/L.
It is interesting to reexamine the nitrification design shown in Figure 6-30 on a stage-by-stage basis.
Assuming the maximum nitrification rate of 1.46 g NHJ-N/m2/d (0.3 lb/1,000 sq ft/d) through Stages 3 and
4 (18 shafts total), the NHJ-N concentration would be reduced from 22.6 mg/L to 5.2 mg/L. From Figure
6-27, the concentration in Stage 5 corresponding to the following condition:
(removal rate)(6)(13,940) = (5.2 - Ce)(21,055)
is 2.1 mg/L at a removal rate of 0.77 g NHJ-N/m2/d. In Stage 6, the concentration corresponding to the
following condition:
(removal rate)(6)(13,940) = (2.1 - Ce)(21,055)
is 0.86 mg/L at a removal rate of 0.31 g NHJ-N/nWd. This type of stage-by-stage analysis of a preliminary
design will produce a more accurate estimate of effluent NHJ-N concentrations than just assuming one rate
at the desired effluent concentration as was done earlier in this design example. Whether or not this
increased accuracy is of practical significance is an entirely different matter. This type of analysis demon-
strates that fixed film RBC systems increase in area dramatically as effluent ammonia concentrations
become very stringent.
192
-------
6.5.3.11 Design Example No. 4 (continued)
5. Check alkalinity requirements. Nitrification of 21.6 mg/L of NHJ-N will consume 153 mg/L of alkalinity as
CaCO3. To maintain a residual alkalinity of 50 mg/L as CaCO3 requires:
153 + 50 - 120 = 83 mg/L : .
From Table 2-3, the lime addition required is:
= 970 kg (2,140 Ib) quicklime/d
6. Account for the effect of wastewater temperature. When using the design curves presented in Figures
6-28 and 6-29 for temperatures less than 13°C (55°F), the required media surface areas need to be adjusted
by the factors indicated in Figure 6-25; for temperatures above 13°C (55°F), the design curves do not need
to be adjusted. Some data show that nitrification rates may increase to near 2 g/m2/d (0.4 lb/1,000 sq ft/d)
at temperatures near 18.3°C (65°F ± 5°F), but there was enough variation in the data so that assuming
higher rates would not be justified in the absence of pilot scale data (105).
7. Reduce the nitrate nitrogen to an acceptable level. To achieve the final effluent limit of 5 mg/L total
nitrogen specified in Table 6-25 requires the addition of a separate-stage denitrification system to reduce
the nitrate nitrogen to an acceptable level. Design of such systems is detailed in Chapter 7.
6.5.4 Other Attached Growth Reactors
Packed-bed reactors, biological aerated filters, and
fluidized-bed reactors represent attached growth proc-
esses that have been utilized to some extent for nitrifica-
tion of municipal wastewaters. Unlike trickling filters, the
hydraulic design of these systems is such that the media
, are submerged in the reactor liquid. In packed-bed reac-
tors and biological aerated filters, the media are stationary
during normal operation, held in place by gravity. In the
fluidized-bed reactor, the media are expanded or fluidized
as the incoming flow passes upward through the reactor.
Information pertaining to use of these attached growth
reactors for nitrification of municipal wastewaters is pro-
vided in Sections 6.5.4.1 through 6.5.4.3.
6.5.4.1 Packed-Bed Reactors
A packed-bed reactor, often referred to as a submerged
filter, contains a stationary bed of media which provides
support for biological growth. The influent wastewater or
wastewater plus recycled effluent is normally introduced
at the bottom of the reactor through a flow distribution
system. Methods utilized to supply the necessary oxygen
to support biomass growth have included direct introduc-
tion of air (123) or high-purity oxygen (124) into the bot-
tom of the reactor through a gas distribution system or
injection of air or oxygen into the feed line entering the
reactor. Alternatively high-purity oxygen has been dis-
solved in the feed stream in an oxygenation device prior
to the feed entering the reactor (123).
In 1975, the EPA Process Design Manual for Nitrogen
Control noted that packed-bed reactors for nitrification
were a comparatively recent development, having pro-
gressed from laboratory and pilot status to the point of
commercial availability (1). Since that time packed-bed
reactors have been widely applied for commercial treat-
ment of industrial wastewaters and contaminated ground
waters. Despite continuing interest in packed-bed reac-
tors for nitrification of municipal wastewaters (123-127)
and additional pilot studies, packed-bed reactors have not
been widely applied on a full scale. The lack of informa-
tion clearly demonstrating significant advantages of the
technology relative to alternatives for this application has
limited the acceptance of packed-bed reactors at the full-
scale level for municipal wastewater treatment.
In 1975, it was reported that several types of media in-
cluding stones, gravel, anthracite, and random plastic me-
dia had been successfully utilized in pilot plant studies of
packed-bed reactors. In more recent studies, the media
utilized has normally been either random or corrugated
plastic structures with high void volume (123-127). The
use of such media may eliminate the need for backwash-
ing to control the buildup of reactor SS. If solids buildup
is not prevented or controlled, the hydraulic integrity of
the reactor will be compromised. Design and operating
strategies that minimize the buildup of reactor SS include:
• The use of media with a high void volume (greater than
90 percent).
• The supply of oxygen by the direct introduction of air
into the bottom of the reactor.
• Operation at low influent CBOD5 and SS.
193
-------
As with trickling filters, the efficiency and performance of
nitrifying packed-bed reactors can be expected to corre-
late to the effective surface area for biofilm growth, al-
though growth of active nitrifiers in the voids of the media
may affect this correlation. Thus, both the surface loading
and the volumetric loading are likely to influence nitrifica-
tion efficiency and performance in packed-bed reactors.
Other factors such as the concentration of DO, CBODS,
and ammonium in the reactor, environmental conditions
(i.e., temperature and pH), and media characteristics (i.e.,
surface-to- volume ratio and percent voids) will influence
the correlations between loading and nitrification perform-
ance. Although surface and volumetric loading informa-
tion applicable to the design of packed-bed reactors for
nitrification of municipal wastewaters is available (123-
126), onsite piloting is recommended if the technology is
being considered for use on a full scale.
6,5.4.2 Biological Aerated Filters
In the biological aerated filter (BAF), the media are sub-
merged in the reactor and primary clarified wastewater is
introduced at the top of the reactor. As noted in an EPA-
sponsored study (128), BAF systems are very similar in
both physical appearance and mode of operation to a
downflow water filter or tertiary wastewater filter except
that:
• A coarser, low density media is utilized.
• Air is diffused upward through the media during
operation.
The air is introduced into the media through an air diffu-
sion system located approximately 20 to 25 cm (8 to 10
in) above the filter underdrain system (128). This air is
supplied to promote biomass growth in the voids of the
packed bed and on the media surface above the air dif-
fusion system. The function of the media below the air
diffusion system is to remove SS. As newly grown
biomass and influent SS build up in the reactor, the head
loss across the unit increases. The unit is backwashed
when a predetermined headless is reached. The back-
washing operation involves a series of air scours and
liquid flushes with treated effluent. The intent of this op-
eration is to release SS trapped in the voids of the packed
bed and to control the extent of film growth on the media
surface. The backwash water is normally conveyed to
primary clarification at the head end of the plant. A com-
mon process flow diagram for a complete Biocarbone
BAF system is shown in Figure 6-31. Biocarbone is the
trademark name given to Omnium de Traitement et de
Valorisations (OTVs) commercial embodiment of the
process.
The primary advantage of the BAF is biotreatment and
solids separation in the same reactor eliminating the re-
quirement for separate secondary clarification. Conse-
quently, the technology could reduce the space
requirements for treatment relative to more conventional
technologies such as the activated sludge system.
The first commercial, full-scale BAF system began opera-
tion in 1982 in Soissons, France (129). Since that time a
number of systems have been installed in Europe, Japan,
and North America (128,130). As of 1990, there were
approximately 30 commercial full-scale Biocarbone BAF
systems installed or under construction, designed at
wastewater flows of 1,900 22 Us (0.5 mgd) or greater
(130). The largest Biocarbone BAF system installed to
date is designed to treat approximately 1,056 Us (24
mgd) (129). Most Biocarbone BAF systems in operation
today have been designed for CBOD5 and TSS removal,
but the systems can be designed to nitrify primary or
secondary effluent. :
The original media employed in the Biocarbone BAF was
granular activated carbon. Subsequently, alternative
granular media has been used for economic reasons. The
media in most currently operating BAF systems consists
of a kiln-fired clay or shale particle. Biodamine and Bio-
dagene are the names given to two of the media often
used in the Biocarbone BAF (128). Biodamine is an
angular shaped media whereas Biodagene is more
spherical.
Backwash Water
Figure 6-31. Biocarbone BAF system example flowsheet (adapted from Reference 128).
194
-------
The angularity and size range of the media significantly
affects the BAF treatment performance and operating re-
quirements. The use of smaller media (i.e., in the range
of 2 to 4 mm [0.08 to 0.16 in]), although it offers a superior
effluent quality to that of a system with larger sized media,
normally requires more frequent backwashing (128). The
smaller media has been recommended when nitrification
is required (130).
Process design and performance information on the gen-
eral application of the BAF to municipal wastewater treat-
ment can be found in other publications (128,131,132).
Included is information on the questions of aeration and
backwashing requirements, solids production, and design
of system components. The following discussions present
information on nitrification applications.
Pilot plant studies by the developer of the Biocarbone
BAF system indicate that for a system treating primary
effluent wastewater containing a high CBOD5 concentra-
tion, nitrification is governed in part by the COD volumet-
ric loading. The volumetric loading is based on the volume
occupied by the media (i.e., empty bed volume). The
results (Figure 6-32) indicate that at a COD volumetric
loading above approximately 3.2 kg/m3/d (200 lb/1,000
cu ft/d), nitrification is substantially reduced because of
increased heterotrophic organism growth and associated
oxygen consumption.
Data from a Biocarbone pilot plant study are presented
in Figure 6-33 (130). The results depict the performance
of a 140-L (37-gal) BAF pilot unit receiving primary mu-
nicipal wastewater. A COD volumetric load of less than
2.0 kg/m3/d (125 lb/1,000 cu ft/d) was required to achieve
approximately 90-percent ammonium oxidation in a single
BAF unit. The BAF medium used in the pilot study was
metamorphosed shale with a grain size between 3 and 6
mm (0.12 to 0.24 in). Reactor temperature information
was not provided for the results presented in Figures 6-32
and 6-33.
According to results from the operation in the United
States of a full-scale demonstration Biocarbone BAF plant
treating primary municipal wastewater in the mid-1980s
(132), the BOD5 volumetric loading must be limited to
approximately 1 kg/m3/d (62.4 lb/1,000 cu ft/d) to achieve
near 90-percent ammonium oxidation in a single unit. This
conclusion is based on operation at temperatures as low
as 11°C (52°F) using a vitrified clay medium with an
effective size of 3.4 mm (0.13 in) and a uniformity coef-
ficient between 1.5 and 1.6. Other more recent full-scale
Biocarbone BAF plant assessments indicate that to
achieve an average effluent ammonia-N concentration of
2.5 mg/L in the treatment of primary effluent, the COD
volumetric loading must be limited to approximately 5
kg/m3/d (312 lb/1,000 cu ft/d). The volumetric loading rate
results indicate that carbonaceous oxidation and nearly
complete nitrification of primary treated wastewater can
§
o
cu
DC
E
"c
o
90
85
80
75
70
65
60
-V
Media Type: Biodamine
Media Depth: 1.6m (5.25 ft)
Media Size: 2-5 mm (0.08-0.20 in.)
1.6 3.2 4.8 6.4
COD Volumetric Loading, kg/m3/d
Figure 6-32. Effect of COD volumetric loading on Biocar-
bone BAF nitrification performance in the treatment of pri-
mary effluent (adapted from Reference 128).
be achieved in single BAF units at an empty-bed hydrau-
lic retention time of approximately 1.5 to 3.5 hours.
BAFs are typically designed to treat municipal waste-
waters with low carbonaceous feed concentration, such
as that characteristic of secondary effluent. In an EPA-
sponsored, detailed assessment of BAFs (128), informa-
tion derived from operation of a full-scale BAF unit
treating secondary effluent was used to develop a design
approach to predict the empty-bed hydraulic retention
time required to achieve nitrification. At an influent BOD5
and TSS concentration of approximately 20 mg/L, a hy-
draulic retention time of 0.83 hr was predicted to be re-
quired to reduce the ammonium nitrogen from
approximately 21 to 7 mg/L. These results translate to an
ammonium-nitrogen loading of 0.58 kg/m3/d (36 lb/1,000
cu ft/d). Other reports indicate that over 90-percent re-
moval of ammonium nitrogen is achievable at comparable
volumetric loading rates at temperatures as low as 13.5°C
(56.3°F) (129,133).
195
-------
T>
§
0)
CC
c
_ • Removal Efficiency (% N Removed)
o Nitrogen Removed (kg N/rrP/dx 100)
1 kg/m3/d = 62.4 lb/1,000 ft3/d
1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7
COD Volumetric Loading, kg/m3/d
Figure 6-33. Effect of COD volumetric loading on nitrification of primary treated wastewater in a Biocarbone BAF pilot
unit (adapted from Reference 130).
Although full-scale application of BAFs for municipal
wastewater treatment has become widespread in recent
years, particularly in Europe (130), the amount of oper-
ating and performance information on U.S. installations
Is limited. The lack of an extensive data base on nitrifi-
cation applications suggests that onsite piloting may be
warranted before selecting a technology.
6.5.4.3 Fluidfzed-Bed Reactors
In the conventional biological fluidized-bed reactor, often
referred to as an expanded-bed reactor, wastewater or
wastewater plus recycled effluent is introduced at the
bottom of the reactor at a hydraulic loading rate or upflow
velocity sufficient to expand the bed media, resulting in a
flutdized state. The fluidized media particles provide a
vast surface area for biological growth, in part leading to
the development of a biomass concentration approxi-
mately 5 to 10 times greater than that normally main-
tained in a conventional suspended growth reactor. To
date, the media employed in most full-scale fluidized-bed
reactors have either been silica sand or granular activated
carbon.
The mechanical components and subsystems critical to
the development of fluidized-bed commercial systems
are:
• The device or method to distribute the influent flow to
the reactor.
• The device or method to transfer oxygen in a controlled
fashion to the fluidized-bed reactor in aerobic applica-
tions of the technology. The oxygenation system is
particularly critical in the treatment of wastewaters con-
taining medium to high concentrations of oxygen de-
manding material (i.e., O2 requirements greater than
25 mg/L).
• The device or method to control the expansion of the
fluidized bed due to biofilm growth. The bed height
control system is particularly critical in treatment appli-
cations where the net yield of biomass is significant.
Further details concerning the design of the critical com-
ponents have been presented elsewhere (134).
Although the development of water and wastewater sys-
tems using a fluidized bed of biomass can be traced back
to the 1940s in England (135), media-based fluidized-bed
reactors were not developed until the early 1970s. Re-
searchers at Manhattan College in New York, at the EPA
Municipal Environmental Research Laboratory in Cincin-
nati, OH, and at the Water Research Centre in Medmen-
ham, England, can be credited for the initial application
of media-based fluidized-bed reactors to water and
wastewater treatment. The Manhattan College re-
searchers were granted a U.S. patent in 1974 (assigned
to Ecolotrol, Inc.) for the application of the fluidized-bed
process configuration to "denitrifying wastewater" '(136).
In a paper published in 1970 by researchers from the
University of Michigan, biological activity was observed
in expanded-bed activated carbon reactors and was be-
lieved to be the reason for the observed nitrate reduction
(137).
The ability of the biological fluidized-bed process configu-
ration to intensify biological reaction rates through accu-
mulation of high concentrations of active biomass has
attracted attention for many years. The results from; labo-
ratory and field pilot scale studies have consistently illus-
trated the technical advantages of the fluidized bed over
most other suspended and attached growth reactor con-
figurations in many wastewater treatment applications. In
1981, a comprehensive account of ongoing fluidized-bed
process development activities 'was published based' on
a 1980 seminar held in Manchester, England (138).-AI-
though hailed at that time as the most significant devel-
196
-------
opment in the wastewater treatment field in the last 50
years, it also was claimed that no full-scale plants were
yet in operation (139). Since that time, even though more
than 70 commercial, fluidized-bed reactors have been
installed in North America and Europe, wider use of the
technology has been hampered by such factors as me-
chanical scale-up issues, slow development of economi-
cally attractive system configurations, and proprietary
constraints (140). •
According to a 1991 state-of-the-art review of fluidized
beds for water and wastewater treatment, the technology
was being applied largely for industrial versus municipal
wastewater treatment at current operating full-scale in-
stallations in North America and Europe (140). Although
full-scale fluidized-bed industrial systems are operating
under conditions that result in nitrification, few, if any,
systems have been installed for nitrification of municipal
wastewaters on a full scale. A limited number of reactors
have been installed for denitrification of municipal waste-
water (Section 7.4).
Information useful for the process design of full-scale
systems for nitrification of municipal wastewater derived
from the results of fluidized-bed pilot plant studies (141
to 147) is summarized as follows:
• A half-order model appears appropriate to describe the
kinetics of ammonium oxidation in fluidized-bed reac-
tors under nonlimiting DO conditions.
• The volumetric removal rate and the specific ammo-
nium oxidation rate decrease significantly at low reac-
tor ammonium concentrations.
• The fluidized-bed hydraulic retention time required to
achieve nitrification down to ammonium levels of 2
mg/L or less ranges from 10 to 40 minutes. This HRT
is for treatment of municipal wastewaters containing
less than 50 mg/L of CBOD5 and approximately 20
mg/L of oxidizable nitrogen compounds, and providing
that the reactor is designed to promote the buildup of
at least 8.5 g/L of volatile attached solids and that
nonlimiting DO conditions are achieved. The actual
HRT required will depend on such factors as the con-
centration of carbonaceous BOD in the wastewater,
the system hydraulics (i.e., plug flow versus complete
mixing conditions), and the reactor temperature and
pH conditions.
If the use of the fluidized bed for nitrification is being
considered, onsite piloting is recommended given the lim-
ited amount of full-scale operating and performance in-
formation on this application.
6.6 Combined or Coupled Suspended and
Attached Growth Systems
6.6.1 General
A variety of approaches combine suspended and at-
tached growth components into a single treatment proc-
ess. For purposes of this section, a single treatment
process means any system in which the wastewater con-
tacts both a suspended growth component and an at-
tached growth component prior to final settling. For
example, a roughing filter may precede an activated
sludge system or a trickling filter may precede a solids
contact reactor. In each of these examples, an attached
growth and suspended growth system have been com-
bined. In other systems, called coupled systems, sus-
pended and attached growth biomass operate within the
same reactor(s). In coupled systems, settled solids from
the final clarifier may or may not be recycled back through
the fixed biomass component in each application, al-
though this option is available in all the coupled systems
to be discussed in this section.
Combining or coupling suspended and attached growth
systems can offer several process advantages including
protection against biomass washout, improved handling
of industrial discharges or toxic shock loads, improved
SVIs and SS settling velocities, and overall ease of op-
eration. This section describes approaches used for nitri-
fication, including:
• Trickling filter/activated sludge (TF/AS)
• Trickling filter/solids contact (TF/SC) .".'.'-
• Aerated biological filter/activated sludge (ABF/AS)
• Coupled Systems
- Captor
- RingLace
- Linpor
- FAST ;
- Bio-2-Sludge
- Monitor
Fundamental aspects of these systems are presented
along with selected data about several of them.
6.6.2 Trickling Filter/Suspended Growth
Trickling filter/suspended growth processes encompass-
ing a variety of applications have been summarized (148).
According to some reports, the trickling filter may serve
as a roughing filter with typical loadings of 1.6 to 3.2 kg
BODs/rrrVd (100-200 Ib BOD/1,000 cu ft/d) placed imme-
diately upstream of an activated sludge system. Or the
trickling filter may be sized to ensure the necessary level
of nitrification ahead of a solids contact reactor. In either
197
-------
configuration, recycle solids from the final settler are re-
turned to the suspended growth reactor only.
To distinguish the TF/SC process from TF/AS, EPA de-
fined TF/SC as having the following distinguishing fea-
tures:
• The primary function of the solids contact tank is to
increase solids capture and particulate BOD removal.
• The majority of the soluble BOD removal occurs in the
trickling filter.
• Return sludge solids are mixed with the trickling filter
effluent.
• The solids contact tank is not designed to nitrify (the
solids residence time is less than approximately two
days) although nitrification may occur in solids contact
if the trickling filter is the source of the nitrifying organ-
isms.
* The aerated solids contact time is one hour or less
based on total flow, including recycle.
In yet another process variation (ABF/AS), the settled
solids are recycled through the trickling filter (Figure 6-34)
to achieve both a coupled system within the filter as well
as a combined trickling filter/activated sludge system.
6.6.2.1 Trickling Filter/Activated Sludge
The most common application of the TF/AS configuration,
where nitrification must be achieved, is to use the trickling
filter as a "roughing filter" for partial oxidation of the in-
coming carbonaceous material and to design the acti-
vated sludge system with a sufficient solids residence
time to ensure that nitrification can be sustained. With
this approach the roughing filter can be designed to
achieve any level of treatment desired. Alternatively, for
upgrading of existing trickling filter facilities, the benefits
of existing infrastructure may be retained. This upgrading
approach was taken at existing trickling filter plants in
Livermore and Lompac, CA, Corvallis, OR, and El Lago,
TX (149). Factors such as odor production, ventilation, or
media clogging will control maximum feasible trickling
filter loadings. Irrespective of the initial trickling filter load-
ing in the range of 0.74 to 3.2 kg BOD5/m3/d (46 to 200
Ib BODs/I.OOO cu ft/d), the solids residence times in the
activated sludge system controlled the final effluent qual-
ity with regard to soluble BOD and control of nonsettle-
able influent solids (150).
A survey of seven roughing filter/activated sludge facilities
found a mean trickling filter design loading of 0.78 kg
BODg/nf/d (48 Ib BODg/1,000 cu ft/d) (148). Organic
loadings can be much higher. Data from plastic media
trickling filter studies at Sacramento, California, indicate
oxidation in the tower was about 40 percent of the applied
BOD5 over the loading range of 1.6 to 6.4 kg BODs/m3/d
(100 to 400 Ib BODg/1,000 cu fl/d), with oxidation deter-
mined by the difference between influent and unsettled
effluent BOD5 values (151). Removals in the 1.3 m (4.25
ft) deep rock filter with 5 to 10 cm (2 to 4 in) media at
Livermore, California, were not as good, even at the lower
organic loadings shown in Table 6-26.
The unsettled wastewater exiting the trickling filter con-
sists of a mixture of incoming soluble BOD not metabo-
lized in the filter, incoming particulate degradable SS not
precipitated or bioflocculated within the filter, active
sloughed biomass and associated captured particulates
not yet metabolized, and biomass decay products and
other inert components in the entering wastewater. Fur-
thermore, trickling filters normally do not slough at a uni-
form rate, adding additional day-to/day variability to
unsettled effluent quality. All these factors are variables
that need to be included in a rigorous conceptual ap-
proach to mathematically describing combined TF/AS
performance. Although developing mathematical expres-
sions to describe the performance of the suspended
growth system is reasonably straightforward, the equa-
tions require a series of input parameters for which reli-
able data are normally unavailable. Suffice it to say that
*• WAS
Figure 6-34. ABF/AS process schematic (adapted from Reference 148).
198
-------
Table 6-26. Trickling Filter Removals at Livermore, Califor-
nia (Adapted from Reference 149)
Filter Loading
Year
1968-69b
1970
1971
1972
1973
1974
kg BODs/mVd
1.92
0.88
0.83
1.00
0.82
0.90
Ib BOD,/
1,000 cu
ft/d
120
55
52
62
51
56
BODS
Removal8
(%)
32
29
28
15
18
32
a Based on unsettled trickling filter effluent.
b During operation of one filter.
design of coupled TF/AS systems involves considerable
empiricism.
A conservative design approach for a TF/AS system
would be to ignore the BOD removal (influent minus un-
settled effluent) occurring in the trickling filter, and base
the activated sludge system design on the influent waste-
water characteristics prior to the trickling filter. For rough-
ing filter applications, where no nitrification occurs in the
trickling filter, this approach has no impact on the mini-
mum solids residence time, 9™, required to sustain nitri-
fication in the activated sludge system. It will result in
overestimation of the net solids production in the AS re-
actor (for any 00) due to some active biomass decay in
the TF, and an overestimate of oxygen requirements due
to the combined impact of organic carbon oxidation and
active biomass decay in the TF. One method of estimating
these impacts has been presented (149). An alternative
is to reduce biomass production and oxygen demand in
the activated sludge unit by the biomass decay and esti-
mated removals (based on unsettled effluent), respec-
tively, which will occur in the trickling filter.
6.6.2.2 Nitrifying Trickling Filter/Activated Sludge
and Nitrifying Trickling Filter/Solids Contact
When the trickling filter is designed to provide the level
of nitrification required (Section 6.5.2.1), design of a cou-
pled trickling filter/solids contact system is reasonably
straightforward. Hydraulic retention times in solids contact
vary from 3 to 60 minutes; a typical range of solids resi-
dence times is 0.2 to 1.0 day (152). Others indicate a
general approach is to maintain 0C between 1 and 2 days
(153). Total organic loadings in TF/SC studies at Chino
Basin and Garland were 0.2 to 1.1 kg BOD/m3/d (12 to
69 Ib BODs/1,000 cu ft/d); for both studies, effluent TSS
decreased with increasing 00 up to between 0.5 and 1.0
days and were essentially constant at 90 equal to 1.0 day
(150). When short solids contact times are used, a sludge
reaeration tank is almost always needed to aerate solids
adequately for proper bioflocculation. Where soluble BOD
levels are low, as they will be for a nitrifying trickling filter,
solids reaeration, either in a separate tank or by using
step feed, is usually the preferred approach. Studies at
Tolleson and Medford indicate that most flocculation oc-
curs in less than 12 minutes (154). Secondary clarifiers
are typically flocculator clarifiers. The combination of floc-
culator clarifiers in deep secondary tanks is reported to
allow a monthly average effluent SS level of approxi-
mately of 10 mg/L at average overflow rates of 48 m3/m2/d
(1,200gpd/sqft) (152).
It is also possible to have partial nitrification in the trickling
filter and design for an additional increment of nitrification
in the solids contact or activated sludge reactor. One
conceptual approach is to design the SC or AS system
with a GC that is < 0™ (or to use a process design factor
near 1.0) and rely on the continual sloughing of nitrifying
organisms from the trickling filter to sustain nitrification in
the suspended growth reactor. For example, the theoreti-
cal relationship shown on Figure 6-35 illustrates the ad-
ditional nitrification that can be achieved even when the
design solids residence time, 9^ is less than 0™ (the
safety factor is less than 1.0) (155). This relationship
assumes that nitrifiers continuously slough from the trick-
ling filter at a uniform rate, dictated by the level of nitrifi-
Trickling Filter
Nitrification
Efficiency
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Nitrification Safety Factor
Figure 6-35. Effect of upstream nitrification on the effluent
NHj-N from downstream suspended growth unit. Influent
NHJ-N is 20 mg/L and temperature is 20°C (adapted from
Reference 155).
199
-------
cation efficiency in the trickling filter. If the discharge of
nitrifying organisms from sloughing fluctuates widely, the
relationship in Figure 6-35 will not hold. Nitrification in the
contact tank at Garland reportedly was inconsistent on a
month-to-month basis indicating erratic seeding from the
nitrifying trickling filter (156).
When the modeling approach used to develop Figure
6-35 was applied to data from the Duck Creek TF/SC
plant, the results shown in Figure 6-36 were obtained.
The Duck Creek plant consists of both solids reaeration,
which is one-third of the total aerated volume, and solids
contact; actual contact is 24 to 44 minutes. Even where
the total solids residence time in the aeration basins was
less than Q$, additional nitrification occurred in the solids
contact tank, indicating a reasonably constant degree of
nitrifier sloughing from the TF. From 1988 to 1990, the TF
loading averaged a moderate 0.4 kg/m3/d (25 Ib
BODs/1,OOOcuft/d).
6.6.2.3 Aerated Biological Filter/Activated Sludge
The concept of returning activated sludge to an ABF was
developed in 1966 by a wastewater treatment plant op-
erator (157). In 1973, the original concept was modified
by combining an aeration basin with the aerated biofilter
(ABF/AS). General design and performance charac-
teristics of the ABF/AS process have been reported (158).
The process is currently marketed by Infilco Degremont.
As shown in Figure 6-34, recycle solids are mixed with
the primary effluent and pumped through the tower. This
approach provides a high initial F/M to the recycled solids
which should contribute to improved sludge settleability.
With this arrangement, it is not possible to take any direct
measurements of BOD5 or COD reduction across the
tower. Consequently, the contribution of the biotower to
organic removals can be determined only indirectly. One
approach to determining biotower contribution to satisfy-
ing the total oxygen demand is to measure the oxygen
uptake rate in the aeration basin and compare this rate
to the total COD reduction (influent COD minus effluent
and waste COD) across the combined system (including
additions for nitrogenous demand). Using a similar ap-
proach, it was estimated that the biotower might satisfy
from 17 to 33 percent of the total oxygen demand (159).
Neptune Microfloc, which formerly offered the ABF/AS
technology, summarized data from three pilot plant stud-
ies where nitrification was occurring (160). These data
are presented in Table 6-27.
6.6.3 Other Inert Support Media
A variety of systems include addition of an inert support
media to an activated sludge aeration basin (161). These
media include freely moving porous pads (e.g., Captor
and Linpor), trickling filter media (e.g., FAST and Bio-2-
Sludge), racks of open weave media fixed in place (e.g.,
Ringlace), and modular media systems (Monitor). Un-
doubtedly, new approaches will continue to appear and
some existing systems may not survive the competitive
pressures of the marketplace. None of these systems is
in widespread use now. The systems have certain proc-
ess advantages but the cost benefits associated with their
implementation are still being defined. The intent of this
section is to provide a brief overview of some available
systems.
1987
1988
1989
1990
Year
Figure 6-36. Comparison of predicted and actual monthly average effluent NHj-N for Duck Creek (adapted from
Reference 155).
200
-------
Table 6-27. Nitrification Performance from ABS/AS Pilot Studies (Adapted from Reference 160)
Influent (mg/L) Effluent (mg/L) Bio Cell Aeration Basin
Location
Corvallis,
Oregon
Rochester,
Minnesota
Bend, Oregon
BOD
56
92
113
140
198
286
221
180
167
167
186
148
141
162
184
TSS
64
122
153
296
296
181
174
122
101
122
198
173
141
204
207
NHJ-N
15.0
16.3
16.0
36.0
46.0
25.0
24.8
21.1
12.7
13.3
17.4
13.7
11.2
11.9
12.3
BOD
20
8
6
13
15
12
8
13
14
8
11
Q
5
11
12
TSS
29
29
23
37
51
27
18
16
5
16
30
19
18
25
27
NHJ-N
2.2
1.5
0.9
1.8
3.4
1.1
0.9
2.2
0.7
9.5
5.1
0.5
0.1
1.5
0.5
kg BOD./
m3/d
0.69
1.64
1.35
1.70
5.62
3.43
2.66
2.76
2.65
2.49
4.33
2.97
3.53
4.17
3.85
MLVSS
(mg/L)
1,690
2,870
2,350
2,735
2,820
3,210
3,580
3,070
3,625
3,010
2,810
2,680
2,420
2,780
2,870
Temp.
(°C)
11
17
20
22
23
17
17
19
23
10
13
13
15
15
18
HRT
(hr)
3.8
2.6
4.1
3.9
3.4
8.3
7.1
4.4
3.7
8.7
5.4
6.2
5.1
3.4
4.1
6.6,3.1 Captor and Linpor
Both Captor (Ashbrook-Simon-Hartley) and Linpor (Linde
AG) systems use porous pads freely suspended in the
aeration basin. Linpor sponges are approximately cubical
with sides of about 10 to 12 mm (0.4 to 0.5 in) and Captor
sponges are about 12 by 25 by 25 mm (0.5 by 1 by 1
in). While not required, either system may include recycle
of settled solids from the final settler back through the
pads. The recycle option leads to a combined at-
tached/suspended growth system. The pads are retained
in the aeration basin via fine screens. The Captor system
may include an airlift pump and pad conveyor system for
squeezing excess solids from the porous pads. At one
time, it was believed that this cleaning procedure would
control solids loss from the pads in secondary treatment
applications, such that existing aeration basins could be
upgraded to nitrification by confining the pads to the up-
stream reactor segment where the bulk of the BOD and
SS removal would occur. This upgrade would allow for a
conventional activated sludge nitrification system in the
downstream reactor segment. However, the ability to con-
trol solids loss never met expectations (162). The Linpor
approach for combined carbonaceous removal and nitri-
fication recycles settled secondary solids back through
the pads. When treating secondary effluent in a separate
stage nitrification application, there is normally no solids
recycle.
Captor has been installed in Moundsville, West Virginia.
Primary effluent is fed to a Captor zone, with 40 to 70
minutes hydraulic residence time which is directly fol-
lowed by a conventional activated sludge reactor. Partial
nitrification is achieved in the Captor zone. The only other
full-scale Captor installation in the United States is in
Opelousos, Louisiana, where Captor nitrifies secondary
effluent, and is not followed by a final clarifier. One inter-
esting application of Captor was a pilot plant investigation
of lagoon effluent at Vesper, Wisconsin (163). The lagoon
effluent was fed to a Captor reactor with no final clarifier.
The ability to nitrify 0.1 to 0.16 kg NHJ-N/m3/d (6 to 10
lb/1,000 cu ft/d) at wastewater temperatures near 0°C
(32°F) was demonstrated.
For combined carbonaceous removal and nitrification, the
volume of pads in a Linpor system equals roughly 10 to
30 percent of the aeration tank volume. The fixed film
biomass typically has an equivalent MLSS concentration
of 1,200 to 3,800 mg/L and accounts for 25 to 60 percent
of the total biomass in a typical aeration system (161).
The Freising, Germany, plant was converted to a Linpor
system in 1984 by the addition of a pad volume equal to
20 percent of the aeration tank volume. The primary rea-
sons for the conversion were frequently occurring poor
sludge settleability, limited space, and cost effectiveness.
Before conversion, the plant could only maintain an MLSS
concentration of about 2,600 mg/L. Following conversion,
a much higher MLSS concentration could be maintained
and an average of 65-percent nitrification was achieved
over a wastewater temperature range of 10°C (50°F) to
17°C (63°F), even though the DO concentration averaged
201
-------
only 1.7 mg/L. Like Captor, the Linpor system has been
employed as a fixed film system for nitrification of secon-
dary effluent, where no final clarification or solids recycle
occurred (164).
6.6.3.2 FAST and Bio-2-Sludge
Submerged trickling filter media systems include the
FAST system (Smith and Loveless), Bio-2-Sludge (Weber
Engineering), as well as other systems (165,166).
In a FAST system, typically 75 percent of the aeration
tank volume is occupied by media. The media is 100
percent submerged and diffused aeration forces waste-
water flow up through the media. A combined fixed film,
suspended growth system was installed at Stow, Massa-
chusetts, as illustrated in Figure 6-37. This plant was
upgraded from a complete mix activated sludge system
when new discharge standards limited total effluent nitro-
gen to 10 mg/L. The existing aeration tank was converted
to a FAST system followed by an anoxic reactor equipped
for methanol addition. The process design allows for
either1 raw wastewater or methanol to be used as the
carbon source. Methanol was used during startup; proc-
ess performance during the first seven months is sum-
marized in Table 6-28.
In contrast to FAST systems, the Bio-2-SIudge approach
includes trickling filter media in only about 25 percent of
the aeration tankage, but these systems always combine
both attached and freely suspended (i.e., from sludge
recycle) biomass.
Bio-2-Sludge systems use just about any synthetic media,
as long as the openings are at least 2 by 2 cm (0.8 by
0.8 in). Typical media surface area ranges from 90 to 120
m2/m3 (27.5 to 36.6 sq ft/cu ft). Racks to hold the media
are constructed to provide approximately 0.5 m (1.64 ft)
of clear space between the top of the racks and the liquid
surface. This arrangement allows for normal side roll mix-
ing in the tank.
A combined carbonaceous oxidation-nitrification system
using Bio-2-Sludge was installed in Schoemberg, Ger-
many. Prior to plant conversion, the poor settling charac-
teristics of the MLSS would not permit maintenance of
more than 1,000 to 1,500 mg/L MLSS in the aeration
tanks. After conversion and plant expansion, the MLSS
settleability improved greatly (mean SVI of 82 mL/g) and
secondary effluent quality averaged 0.4 mg/L NHJ-N
(161).
Some investigators contend that in submerged biological
filter systems with sludge recycle, the filter media is popu-
lated mainly by protozoa (166). The media then has only
an indirect influence by making it easier to maintain a
lower F/M ratio, due to the resulting improved SVIs.
6.6.3.3 Ringlace
In a Ringlace system (Ringlace Systems, Inc.), polyvinyl
chloridene (PVCE) string on racks are installed in the
To Dosing
Chamber
Legend:
RAS Return Activated Sludge
WAS Waste Activated Sludge
Figure 6-37. Nitrogen control schematic for Stow, Massachusetts (adapted from Reference 166).
202
-------
Table 6-28. Average Influent and Effluent Data for Stow, Massachusetts (Adapted from Reference 167)
BODS ,mg/L TSS, mg/L NHJ-N, mg/L
Month
Influent
Effluent
Influent
Effluent
Influent
Effluent
Mar.*
Apr.*
May
June
July
Aug.
Sept.
650
490
300
620
445
230
240
. 13
65
1
21
11
3
24
424
309
280
310
456
311
211
18
6
9
12
2
1
1
56 :
35
41
22
32
38
39
4
8.1
0.4
0.1
0.3
0.1
0.2
* Startup period.
aeration tank. Each string has numerous loops of the
same material, thus greatly increasing the surface area
available for fixed film growth. According to a European
distributor, typically 25 to 50 percent of an aeration tank
would contain Ringlace at a density of 120 to 300 lineal
m/m3 (11 to 28 ft/cu ft) of aeration tank (161). One U.S.
distributor indicated the racks typically occupy about 50
to 80 percent of the basin volume with a mean density
based on the entire basin volume of about 120 lineal
m/m3 (11 ft/cu ft).
These systems have been installed in Japan, with and
without sludge recycle. The systems typically have been
at small scale and lightly loaded. The Olching, Germany,
plant added Ringlace in 1987. Before conversion, the
maximum operational MLSS concentration achievable
was about 1,500 mg/L, which resulted in a process F/M
ratio of 0.6 to 0.7 kg BODs/kg MLSS/d. After conversion,
the freely suspended biomass varied from 3,500 to 4,500
mg/L. Fixed biomass on the Ringlace material was esti-
mated at 6.5 g/m which is equivalent to an additional 790
mg/L of MLSS. The overall operating F/M of the system
decreased to approximately 0.2 and the desired levels of
nitrification could be achieved (161). Ringlace Systems Inc.
is scheduled to supply the media for a technology evaluation
in one of the aeration basins at the 440 L/s (10 mgd) Annapo-
lis Wastewater Reclamation Facility.
6.6.3.4 Monitor
The final system to be mentioned is Monitor (KLV Tech-
nologies)—a new system that consists of a media-filled
biochamber constructed on pontoons and floated into po-
sition. This system is suitable for lagoon installations that
require upgrading for nitrification. Wastewater is pumped
into and through the aerated media. A Monitor system in
Ontario County, New York, was reported by the manufac-
turer to maintain ammonia-nitrogen concentrations below
0.5 mg/L in cold wastewater temperatures (168).
6.6.3.5 Design Considerations
When fixed film and freely suspended biomass systems
are combined, the overall contribution from each system
component typically is uncertain. Because of mass trans-
fer limitations for both substrate and DO, a unit of at-
tached biomass usually is not equivalent to a unit of freely
suspended biomass in terms of ability to oxidize carbon
or ammonia. Since some media systems are more porous
or open than others, the benefits of alternative media
systems should not be compared solely on the mass of
biomass attached. Furthermore, the advantages fre-
quently associated with such systems (e.g., improved set-
tleability) cannot be related to a specific design
parameter, such as specific surface area available per
unit of loading. Given current levels of understanding,
pilot plant evaluations remain the best design approach
whenever these systems are under consideration.
6.7 References
When an NTIS number is cited in a reference, that docu-
ment is available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
703-487-4650
1. U.S. EPA. 1975. Process design manual for nitro-
gen control. EPA/625/1-77/007 (NTIS PB-259149).
Washington, DC.
2. Water Pollution Control Federation. 1983. Nutrient
control. Manual of Practice No. FD-7.
3. Lee, S.E, H.S. Shin, and B.C. Paik. 1988. Treatment
of CR(VI)-containing wastewater by addition of
powdered activated carbon to the activated sludge
process. Proceedings of the 42nd Industrial Waste
Conference, Purdue University, Lafayette, IN. Chel-
sea, Ml: Lewis Publishers, p. 121.
203
-------
4. Mulbarger, M.C. 1971. The three sludge system for
nitrogen and phosphorus removal. Presented at the
44th Annual Conference of the Water Pollution Con-
trol Federation, San Francisco, CA (October).
5. Tenney, M.W., and W.F. Echelberger. 1972. Re-
moval of organic and eutrophying pollutants by
chemical-biological treatment. EPA Report no. R2-
72-076 (NTIS PB-214628). Washington, DC.
6. Rimer, A.E., and R.L. Woodward. 1972. Two-stage"
activated sludge pilot plant operations. Fitchburg,
MA. JWPCF44:101.
7. County Sanitation Districts of Los Angeles County.
Monthly operating reports. Whittier Narrows Water
Reclamation Piant. April 1973 to March 1974.
8. Schwer, A.D. 1971. Metropolitan sewer district of
Greater Cincinnati. Letter communication to D.S.
Parker (March).
9. U.S. EPA. 1974. Design manual: upgrading existing
wastewater treatment plants. EPA/625/1 -71/004a
(NTIS PB249148). Washington, DC.
10. Stenquist, R.J. 1974. Carbon oxidation-nitrification
in synthetic media trickling filters. JWPCF 46:2327.
11. Barth, E.F., R.C. Brenner, and R.F. Lewis. 1968.
Chemical-biological control of nitrogen in wastewa-
ter effluent. JWPCF 40:2040.
12. Weddle, C.L, D.G. Niles, E. Goldman, and J.W.
Porter. 1971. Studies of municipal wastewater reno-
vation for industrial water. Presented at the 44th
Annual Conference of the Water Pollution Control
Federation, San Francisco, CA (October).
13. Greene, R.A. 1973. Complete nitrification by single-
stage activated sludge. Presented at the 46th An-
nual Conference of the Water Pollution Control
Federation, Cleveland, OH (October).
14. Mulbarger, M.C. 1971. Private communication to
D.H. Caldwell.
15. Horstkotte, G.A., D.G. Niles, D.S. Parker, and D.H.
Caldwell. 1974. Full-scale testing of a water recla-
mation system. JWPCF 46:181.
16. Newton, D.f and T.E. Wilson. 1973. Oxygen nitrifi-
cation process at Tampa. In: Speece, R.E., and J.F.
Malina, Jr., eds. Applications of commercial oxygen
to water and wastewater systems. Austin, TX: The
Center for Research in Water Resources.
17. Wilson, T.E., and M.D.R. Riddel. 1974. Nitrogen re-
moval—where do we stand? Water Was. Eng.
18. Environmental Quality Analysts, Inc. 1974. Letter
report to valley community services district (March).
19. Parker, D.S., et. al. 1975. Lime use in wastewater
treatment design and cost data. EPA/600/2-75/038.
Cincinnati, OH.
20. Stover, E.L. 1980. Biological nitrification inhibition
screening procedures for industrial wastewater.
Proceedings of the 34th Industrial Waste Confer-
ence, Purdue University, Lafayette, IN. Ann Arbor,
Ml: Ann Arbor Science, p. 887.
21. Stensel, H.D. 1975. Oasis wastewater charac-
terization study, Chattanooga Moccasin Bend
wastewater treatment plant. Report prepared by Air
Products and Chemicals, Inc.
22. Sawyer, C.N. 1939. Activated sludge oxidatiohs, III.
Factors involved in prolonging the high initial rate
of oxygen utilization by activated sludge-mixtures.
Sewage Works J. 11:595.
23. Lawrence, A.W., and P.L. McCarty. 1970. Unified
basis for biological treatment design and operation.
Proceedings of the ASCE. JSED 96(SA3):757.
24. Sawyer, C.N. 1940. Activated sludge oxidations, V.
The influence of nutrition in determining activated
sludge characteristics. Sewage Works J. 12:3.
25. Heidman, J.A., D.F. Bishop, and J.B. Stamberg.
1975. Carbon, nitrogen and phosphorus removal in
staged nitrification-denitrification treatment. EPA/
600/2-75/052. Cincinnati, OH.
26. Heidman, J.A. 1979. Sequential nitrification-denitri-
fication in a plug flow activated sludge system.
EPA/600/2-79/157 (NTIS PB80-134075). Cincin-
nati, OH.
27. Sawyer, C.N., H.E. Wild, Jr., and T.C. McMahon.
1973. Nitrification and denitrification facilities,
wastewater treatment. Prepared for the EPA Tech-
nology Transfer Program.
2.8. Design information report: sludge management sys-
tems. 1990. Water Environ, and Tech. 2(7):62.
29. U.S. EPA. 1987. Phosphorus removal design man-
ual. EPA/625/1-87/001. Cincinnati, OH.
30. Sutton, P.M., K.L. Murphy, B.E Jank, and B.A. Mon-
aghan. Reliability of nitrification systems with inte-
grated phosphorus precipitation. Research Report
No. 64, Canadian Environ. Protection Service Re-
port 75-3-21.
31. Nagel, C.A., and J.G. Haworth. 1969. Operational
factors affecting nitrification in the activated sludge
process. Presented at the 42nd Annual Conference
of the Water Pollution Control Federation, Dallas,
TX (October). (Available as a reprint from the
County Sanitation Districts of Los Angeles County.)
204
-------
32. Murphy, K.L., and P.L. Timpany. 1967. Design and
analysis of mixing for an aeration tank. Proceedings
of the ASCE. JSED 93(SA5):1.
33. Murphy, K.L., and B.I. Boyko. 1970. Longitudinal
mixing in spiral flow aeration tanks. Proceedings of
the ASCE. JSED 96(SA2):211.
34. Metcalf and Eddy, Inc. 1972. Wastewater engineer-
ing. New York: McGraw-Hill.
35. Tetreault, M.J., D.S. Parker, R.L. Pope, J.B. Sai-
gaonkar, and A.E. Fox, Jr. 1989. Separate-stage
denitrification—key to achieving a 3-3-2-0.6 AWT
effluent. Presented at the 62nd Annual Conference
of the Water Pollution Control Federation, San
Francisco, CA (October).
36. City of Jackson, Ml. Wastewater treatment plant
operating records. January 1990 to August 1991.
! i ' . . . . •
37. Oleszkiewicz, J.A., and S.A. Berquist. 1988. Low-
temperature nitrogen removal in sequencing batch
reactors. Water Res. 22;1163.
38. Ettlich, W.F. 1978. A comparison of oxidation ditch
plants to competing processes for secondary and
advanced treatment of municipal wastes.
: EPA/600/2-78/0511. Cincinnati, OH.
39. U.S. EPA. 1991. Evaluation of oxidation ditches for
nutrient removal. Report prepared by HydroQual,
Inc. Washington, DC.
40. Emon, D. 1991. Personal communication from plant
operator, City of Frederick, MD, to D. Sexton, Hy-
droQual, Inc. (November).
41. Gujer, W., and D. Jenkins. 1975. A nitrification
model for the contact stabilization activated sludge
p process. Water Res. 9:561.
42. Gujer, W., and D. Jenkins. 1974. The contact stabi-
''•--• lization process—oxygen and nitrogen mass bal-
ances. University of California, Sanitary
Engineering .Research Laboratory, SERL Report
74-2 (February).
43.itTorpey, W.N. 1948. Practical results of step aera-
'."'!' tiqn. Sewage Works J. 20:781.
4^:%uhV, H.O., etai. 1984. Making full use of step-feed
'"r" 'capability. JWPCF 56:325.
45v,Bajley, W., W. White, C. Wilbur, T.E. Wilson, and R.
!''Jl'§izzarh. 1989. Sludge reaeration in two-step nitrifi-
•': Ration. Water Sci, Tech. 21:1347.
46;t;Beckman, W.J., R.J. Avendt, T.J. Mulligan, and G.J.
".' Kehrberger. 1972. Combined carbon oxidation-nitri-
^^fidation. JWPCF 44:1-916.
47. Budd, W.E., and G.F. Lambeth. 1957. High-purity
oxygen in biological sewage treatment. Sewage
Works J. 29:237.
48. Stankewich, M.J., Jr. 1972. Biological nitrification
with the high-purity oxygenation process. Proceed-
ings of the 27th Industrial Waste Conference, Pur-
due University, Lafayette, IN. Ann Arbor, Ml: Ann
Arbor Science, pp. 1-23.
49. WEF and ASCE. 1992. Design of municipal waste-
water treatment plants. WEF Manual of Practice
No. 8 and ASCE Manual and Report on Engineering
Practice No. 76. WEF, Alexandria, VA.
50. Downing, A.L., and G. Knowles. 1963. Population
dynamics in biological treatment plants. Presented
at the 3rd IAWPRC Conference, Munich, Germany.
51. Heidman, J.A. 1976. Experimental evaluation of
oxygen and air activated sludge nitrification sys-
tems with and without pH control. EPA/600/2-
76/180. Cincinnati, OH.
52. Town of Amherst, New York, wastewater treatment
plant design and operating information. October
1990 to September 1991. Provided by Amherst in
written communication with HydroQual, Inc.
53. U.S. EPA. 1991. Sequencing batch reactors for ni-
trification and nutrient.removal. Report prepared by
HydroQual, Inc. Washington, DC.
54. Irvine, R.L., L.H. Ketchum, R. Breyfogfe, and E.F.
Barth. 1983. Municipal application of sequencing
batch treatment. JWPCF 55:484.
55. Deeney, K.J., J.A. Heidman, W.W. Schuk, K.S.
Young, and A.J. Condren. 1991. Implementation of
sequencing batch reactor technologies in the
United States. Presented at the 64th Annual Con-
ference of the Water Pollution Control Federation,
Toronto, Ontario (October).
56. Randall, T.L., W.M. Copa, and M.J. Dietrich. 1986.
Leachate treatment by a powdered activated carb-
on process. Presented at the 59th Annual Confer-
ence of the Water Pollution Control Federation, Los
Angeles, CA (October).
57. Deeny K.J., J.A. Heidman' and A.J. Condren: 1990.
Performance of activated sludge/ powdered acti-
vated carbon/wet air regeneration systems.
EPA/600/S2-90/012. Cincinnati, OH.
58. Depuydt, K., and R. Amundson. 1991. Solving an
ash buildup challenge. Pollution Engineering, p. 73
(December)." -:
59. Gumbita, R. 1992. Liverpool wastewater'treatment
plant, Medina County, OH. Personal communication
to P.M. Sutton (January).
205
-------
60. MeidI, J.A. 1991. Personal communication from
Zimpro Passavant Environmental Systems, Inc., to
P.M. Sutton (October).
61. Beftens, L. 1979. Powdered activated carbon in an
activated sludge unit. J. Effluent and Water Treat-
ment 9:129.
62. Leipzig, N.A. 1980. Effectiveness of the powdered
activated carbon activated sludge system in remov-
ing ammonia from an organic chemical production
wastewater. Proceedings of the 35th Industrial
Waste Conference, Purdue University, Lafayette,
IN. Ann Arbor, Ml, pp. 889-897.
63. Ng, A.S., and M.K. Stenstrom. 1987. Nitrification in
powdered-activated carbon- activated sludge proc-
ess. J. Environ. Eng. 113:1285.
64. Maier, W., and K. Krauth. 1988. Optimizing nitrifica-
tion in aeration basins with surface aerators. Water
Sci. Tech. 20(4/5):23.
65. Das, D., T.M. Keinath, D.S. Parker, and E.J.
Wahlberg. 1993. Floe breakup in activated sludge
plants. Water Environ. Res. 65(2):138-145.
66. Metcalf and Eddy. 1991. Wastewater engineering,
third edition. New York: McGraw-Hill.
67. Hart J., R. Defore, and S.C. Chiesa. 1986. Activated
sludge control for seasonal nitrification. Water Pol-
lution Control Federation 58(5):358 (May).
68. Heath, M.S., S.A. Wirtel, and B.E. Riffman. 1990.
Simplified design of biofilm processes using nor-
malized loading curves. JWPCF 62:185.
69. Cannon, F.S. 1991. Discussion of the paper simpli-
fied design of biofilm processes using normalized
loading curves. Res. JWPCF 63:90.
70. Gujer, W., and M. Boiler. 1986. Design of a nitrifying
tertiary trickling filter based on theoretical concepts.
Water Res. 20:1353.
71. Rittman, B.E. 1987. Aerobic biological treatment.
Environ. Sci. Tech. 21(2):128.
72. Harremoes, P. 1982. Criteria for, nitrification in fixed
• film reactors. Water Sci. Tech. 14:167.
73. Parker, D.S., and T. Richards. 1986. Nitrification in
trickling filters. JWPCF 58:896.
74. Mulbarger, M.C. 1991. Fundamental secondary
treatment insights. Presented at the 64th Annual
Water Pollution Control Federation Conference.
Toronto, Ontario.
75. Gullicks, H.A., and J.L. Cleasby. 1990. Nitrification
performance of a pilot-scale trickling filter. JWPCF
62:40.
76. Williamson, K.L., and P.L. McCarty. 1973. A mode!
of substrate utilization by bacterial films. Presented
at the 46th Annual Conference of the Water Pollu-
tion Control Federation, Cleveland, OH (October).
77. Duddles, G.A., and S.E. Richardson. 1973. Applica-
tion of plastic media trickling filters for biological
nitrification. Report prepared for EPA, EPA-R2-73-
199. Washington, DC.
78. Antonie, R.L. 1974. Nitrification of activated sludge
effluent with the Bio-Surf process. Presented at the
Annual Conference of the Ohio Water Pollution
Control Association, Toledo, OH (June 7-13).
79. Duddles, G.A., S.E. Richardson, and E.F. Barth.
1974. Plastic medium trickling filters for biological
nitrogen control. JWPCF 46(5):937-946. . .
80. Richardson, S.E. 1973. Pilot plants define parame-
ters for plastic media trickling filter nitrification. Pre-
sented at the 46th Annual Conference of the Water
Pollution Control Federation, Cleveland, OH (Octo-
ber).
81. Sampayo, F.F. 1973. The use of nitrification towers
at Lima, OH. Presented at the Second Annual Con-
ference Water Management Association of Ohio,
Columbus, OH (October).
82. Veuxw, M., and J.C. Merkens. 1970. Recent studies
of high-rate biological filtration. Water Poll. Control
69:113-148.
83. Brown and Caldwell. 1975. Report on tertiary treat-
ment pilot plant studies. Prepared for the City of
Sunnyvale, CA (February).
84. Parker, D.S., M.P. Lutz, R. Dahl, and S. Bernkopf.
1989. Enhancing reaction rates in nitrifying trickling
filters through biofilm control. JWPCF 61:618.
85. Okey, R.W., and O.E. Albertson. 1989. Diffusion's
role in regulating rate and masking temperature ef-
fects in fixed film nitrification. JWPCF 61:500.
86. Water Pollution Control Federation. 1988. O&M of
trickling filters, RBCs, and related processes. Man-
ual of Practice OM-10. Alexandria^ VA.
87. U.S. EPA. 1991. Assessment of single-stage trick-
ling filter nitrification. Report prepared by Hydro-
Qual, Inc. Washington, DC. ,,
88. Parker, D.S., and D.T. Merrill. 1984. Effect of plastic
media configuration on trickling filter performance.
JWPCF 56:955. '.:
89- Sarner, E. 1980. Plastic packed trickling filters. Arm
Arbor, Ml: Ann Arbor Science Publishers.
90. Randall, A.A., C.W. Randall, A. Pramanik, and A.
Shah. 1990. Optimization and performance'of a
Sessil media tower for the pretreatment of a^syn-
206
-------
thetic fiber manufacturing wastewater. Presented at
the 63rd Annual Conference of the Water Pollution
Control Federation, Washington, DC (October).
91. Brown and Caldwell. 1980. Converting rock trickling
filters to plastic media. EPA/600/2-8/120. Cincinnati,
OH.
92. Logan, B.E., S.W. Hermanowicz, and D.S. Parker.
1987. A fundamental model for trickling filter proc-
ess design, JWPCF 59:1017.
93. Siegrist, H., and W. Guj'er. 1987. Demonstration of
mass transfer and pH effects in a nitrifying biofilm.
Water Res. 21:1481.
94. Okey, R.W., and O.E. Albertson. 1989. Diffusion's
role in regulating rate and masking temperature ef-
fects in fixed film nitrification. JWPCF 61:500.
95. Albertson, O.E., and R.W. Okeyk. 1989. Trickling
filters need to breathe, too. Presented at Iowa
WPCF meeting, Des Moines, IA (June).
96. Parker, D.S., M.P. Lutz, and A.M. Pratt. 1990. New
trickling filter applications in the U.S.A. Water Sci.
Tech. 22(1/2):215.
97. Okey, R.W., and O.E. Albertson. 1989. Evidence of
oxygen-limiting conditions during tertiary fixed film
nitrification. JWPCF 61:510.
98. Logan, B.E., D.S. Parker, and R.G. Arnold. 1990.
Oa limitations in CHU and NH utilizing biofilms. Pre-
sented at the American Society of Consulting Engi-
neers Conference on Environmental Engineering,
Washington DC (July).
99. Guilicks, H.A., and J.L. Cleasby. 1990. Cold-ciimate
nitrifying biofilters: design and operation considera-
tions. Res. JWPCF 62:50.
100. Haug, R.T., and P.L. McCarty. 1972. Nitrification with
the submerged filter. JWPCF 44:2086.
101. Boiler, M., and W. Gujer. 1986. Nitrification in tertiary
trickling filters followed by deep-bed filters. Water
Res. 20:1363.
102. Parlin, L., and J. Peel. 1990. Biological ammonia
removal at Reno Sparks, Nevada: nitrification util-
izing high density plastic media towers. Presented
at the California Water Pollution Control Association
Annual Conference (April).
103. Parker, D.S. 1992. Personal communication from
Brown and Caldwell Consulting Engineers to P.M.
Sutton (June).
104. Figueroa, L., and J. Silverstein. 1988. The effect of
paniculate BODs on nitrification in a biofilm proc-
ess. Presented at the Rocky Mountain Water Pol-
lution Control Association Annual Conference,
Snowmass, CO.
105. Brenner, R.C., et al. 1984. Design information on
rotating biological contactors. EPA/600/2-84/106.
Cincinnati, OH.
106. Roy F. Weston, Inc. 1985. Review of current RBC
performance and design procedures. EPA/600/2-
85/033. Cincinnati, OH.
107. Surampalli, R.Y., and E.R. Baumann." 1993. Effec-
tiveness of supplemental aeration and an enlarged
first-stage in improving RBC performance. Environ.
Prog. 12(1 ):24-29 (February).
108. Klees, R., and J. Silverstein; 1990. The use of re-
circulation to improve nitrification in rotating biologi-
cal contactors. Presented at the Annual WPCF
Conference (October).
109. Antonie, R.L. 1976. Fixed biological surfaces—
wastewater treatment: the rotating biological con-
tactor. CRC Press, Cleveland, OH.
110. Stratta, J.M., and D.A. Long. 1981. Nitrification en-
hancement through pH control with rotating biologi-
cal contactors. Final report prepared , for USA
Medical Research and Development Command,
under Contract No. DAMD17-79-C-9110 by Penn-
sylvania State University'(September).
111. Benjes, H.H., Sr. 1977. Small community wastewa-
ter treatment facilities—biological treatment sys-
tems. U:S. EPA Technology Transfer National
Seminar, Chicago, IL.
112. Benjes, H.Hi, Jr. 1990. Capital and O&M cost esti-
mates for attached growth biological wastewater
treatment processes. EPA/600/52-89-003. Cincin-
nati, OH. ,
113. Opatken, E.J. 1982. Rotating biological contac-
tors—second order kinetics. In: Proceedings of the
First International Conference on Fixed-Film Bio-
logical Processies, Vol. I, EPA/600/9-82/023a, Kings
Island, OH, pp. 210-232. :
114. Opatken, E.J. 1986. An alternative RBC design:
second order kinetics. Environ. Prog. 5(1):51-56
(February). ' .
115. Lypo, Inc. 1992- 'Rotating biological surface design
manual. Wastewater Division, a subsidiary of U.S.
Filter Corporation, P.O. Box 181, 29 Vanderburg
Road, Marlboro, NJ 07746 (March).
116. Pano, A., and J.E. Middlebrooks. 1983. Kinetics of
carbon and ammonia nitrogen removal in RBCs.
JWPGF 55(7):956-965 (July).
117. Scheible, O.K., and J.J. Novak.' 1980. Upgrading
primary tanks with rotating biological contactors. In:
Proceedings of the First National Symposium on
Rotating Biological Contactor Technologies, Vol II.
207
-------
EPA/600/9-80/046b (NITS PB81-124547). Cincin-
nati, OH.
118. Envirex, Inc. 1984. Specific RBC process criteria.
Envirex, Inc., 1901 South Prairie Ave., Wauksha,
Wl 53186-7360.
119. Walker Process Equipment. 1992. RBC perform-
ance evaluation. Written correspondence with Hy-
droQual, Inc. Walker Process Equipment, Division
of McNish Corporation, 840 North Russel Ave.,
Aurora, IL 60506 (June).
120. Klargester, Inc. 1984. Standard projects engineering
catalogue. Klargester, Inc., 500 Andalusia Blvd.,
Cape Coral, FL 33909 (September).
121. Autotrol Corporation. 1978. Autotrol wastewater
treatment systems design manual. Bio-Systems Di-
vision, Milwaukee, Wl.
122. Boiler, M., W. Gujer, and G. Nyhuis. 1990. Tertiary
rotating biological contactors for nitrification. Water
Sci. Tech. 22(1/2):89-100.
123. Ruston, B. 1984. Wastewater treatment with aer-
ated submerged biological filters. JWPCF 56:424.
124. Haug, R.T., and P.L. McCarty. 1972. Nitrification with
the submerged filter. JWPCF 44:2086.
125. Bonhomme, M., F. Rogalla, G. Boisseau, and J.
Sibony. 1990. Enhancing nitrogen removal in acti-
vated sludge with fixed biomass. Water Sci. Tech.
22(1/2):127-136.
126. Schlegel, S. 1988. The use of submerged biological
filters for nitrification. Water Sci. Tech. 20:177.
127. Saintpierre, O. 1988. Tertiary nitrifying immerged
biofllter with plastic media. Environ. Tech. Letters
9:1059.
128. Condren, A.J. 1990. Technology assessment of the
biological aerated filter. EPA/600/2-90/015. Cincin-
nati, OH.
129. Rogalla, F., and J. Sibony. 1992. Biocarbone aer-
ated filters—ten years after: past, present, and
plenty of potential. Presented at the IAWPRC 16th
Biennial Conference, Washington, DC (May).
130. Rogalla, F., M. Payraudeau, G. Bacquet, M. Bour-
bigot, J. Sibony, and P. Filles. 1990. Nitrification and
phosphorous precipitation with biological aerated
filters. Res. JWPCF 62:169.
131. Water Pollution Control Federation. 1988. O&M of
trickling filters, RBCs, and related processes. Man-
ual of Practice OM-10, Alexandria, VA, pp. 98-100.
132. Stensel, H.D., R.C. Brenner, K.M. Lee, H. Melcer,
and K. Rakness. 1988. Biological aerated filter
evaluation. ASCE J. of Environ. Eng. 3:655.
133. Dillon, G.R., and V.K. Thomas. 1990. A pilot-scale
evaluation of the Biocarbone process for the treat-
ment of settled sewage and for tertiary nitrification
of secondary effluent. Water Sci. Tech. 22(1/2):305-
316.
134. Sutton, P.M., and P.N. Mishra. 1990. Biological
fluidized beds for water and wastewater treatment:
a state of the art review. Presented at the 63rd
Annual Conference of the Water Pollution Control
Federation, Washington, DC (October).
135. Pugh, N.J. 1945. Treatment of doubtful waters for
public supplies. Transactions of Institution of Water
Engineers 50:80.
136. U.S. Patent No. 3,846,289. 1974.
137. Weber, W.J., Jr., et al. 1970. Physiochemical treat-
ment,of wastewater. JWPCF 42:83.
138. Cooper, P.F., and B. Atkinson, eds. 1981. Biological
fluidized bed treatment of water and wastewater.
Chichester, England: Ellis Horwood Publishers.
139. Cooper, P.F., and B. Atkinson, eds. 1981. Biological
fluidized bed treatment of water and wastewater,
editors' preface. Chichester, England: Ellis Hor-
wood Publishers, p. 12.
140. Sutton, P.M., and P.N. Mishra. 1991. Biological
fluidized beds for water and wastewater treatment.
Water Environ, and Tech. 3(8):52.
141. Jeris, J.S., R.W. Owens, R. Hickey, and F. Flood.
1977. Biological fluidized bed treatment for BOD
and nitrogen removal. JWPCF 49:816.
142. Dearborn Environmental Consulting Services. 1980.
Pilot-scale assessment of the biological fluidized
bed for municipal wastewater treatment (Novem-
ber).
143. Sutton, P.M., W.K. Shieh, P. Kos, and PR. Dunning.
1981. Dorr-Oliver Oxitron System fluidized bed
water and wastewater treatment process. In: Coo-
per, P.F., and B. Atkinson, eds. Biological fluidized
bed treatment of water and wastewater. Chichester,
England: Ellis Horwood Publishers, p. 285.
144. Stevens, O.K., P.M. Berthouex, and T.W. Chapman.
1982. Dynamics and simulation of a biological
fluidized bed reactor. Proceedings of the First Inter-
national Conference on Fixed-Film Biological Proc-
esses, Kings Island, OH, 3:1247.
145. Green, M.K., and P.J. Hardy. 1985. The develop-
ment of a high-rate nitrification fluidized bed proc-
ess. Water Poll. Control 84:44.
146. Cooper, P.F., and S.C. Williams. 1990. High-rate
nitrification in a biological fluidized bed. Water Sci.
Tech. 22(1/2):431-442.
208
-------
147. Tanaka, H., and I.J. Dunn. 1982. Kinetics of biofilm
nitrification. Biotech, and Bioeng. 34:669.
148. Harrison, J.R., and J.W. Filbert. 1984. A survey of
combined trickling filter and activated sludge proc-
esses. JWPCF 56:1073.
149. Stenquist, R.J., D.S. Parker, and W.E. Loftin. The
coupled trickling filter-activated sludge process: de-
sign and performance. Final Report for EPA Con-
tract No. 68-03-2175. Cincinnati, OH.
150. Newbry, B.W., G.T. Daigger, and D. Taniguchi-Den-
nis. 1988. Unit process tradeoffs for combined trick-
ling filter and activated sludge processes. JWPCF
60:1813.
151. Albertson, O., and W.W. Eckenfelder. 1984. Analy-
sis of process factors affecting plastic media trick-
ling filter performance. Proceedings of the Second
International Conference on Fixed-Film Biological
Processes, p. 345.
152. Parker, D.S., and R.N. Matasci. 1989. The TF/SC
process at ten years old: past, present and future.
Presented at the 62nd Annual Water Poll. Control
Fed. Conference, San Francisco, CA (October).
153. Harrison, J.R., and P.L Timpany. 1988. Design con-
siderations with the trickling filter solids contact
process. Proceedings of the National Conference
of Environmental Engineering, CSCE/ASCE, Van-
couver, BC, p. 753.
154. Matasci, R.N., et al. 1985. Full-scale studies on the
trickling filter/solids contact process. Final Report
for EPA Contract No. 68-03-1818 (NTIS PB86
183100/AS) Cincinnati, OH.
155. Daigger, G.T, et al. 1991. Process and kinetic
analysis of nitrification in coupled trickling filter/ac-
tivated sludge processes. Presented at the Water
Poll. Control Fed. Conference, Toronto, Canada.
156. Parker, D.S. 1992. Personal communication from
Brown and Caldwell Consulting Engineers to O.K.
Scheible (May).
157. Hemphill, B.W., and K.P. Lange. 1984. Full-scale
experience with wood media filters using return
sludge application. Proceedings of the Second In-
ternational Conference on Fixed-Film Biological
Processes, p. 440.
158. Arora, M.L, and M.B. Umphres. 1987. Evaluation
of activated biofiltration and activated biofiltra-
tion/activated sludge technologies. JWPCF 59:183.
159. Harrison, J.R. 1980. Survey of plants operating ac-
tivated biofilter/activated sludge. Presented at the
1980 C.W.P.C.A. Northern Regional Conference
and Training School.
160. Slechta, A.F., and G.P. Mattli. 1976. Activated bio-
filter process for biological wastewater treatment.
Presented at the Wastewater Treatment and Reuse
Seminar, S. Lake Tahoe, CA (October).
161. Condren, A.J., B. Rusten, and J.A. Heidman. 1989.
An analysis of high biomass systems. Presented at
the 62nd Annual Water Poll. Control Federation
Conference, San Francisco, CA (October).
162. Heidman, J.A., R.C. Brenner, and H. Shah. 1988.
Pilot-plant evaluation of porous biomass supports.
ASCE. J. Environ. Eng. Div. 114:1077.
163. Golla, P.S., and A.J. Lin. 1991. Cold temperature
nitrification using the Captor process. Presented at
the 46th Purdue Industrial Waste Conference, La-
fayette, IN (May).
164. Morper, M., and A. Wildmoser. 1990. Improvement
of existing wastewater treatment plants; efficiencies
without enlargement of tankage by application of
the Linpor process—case studies. Water Sci. Tech.
22(7/8):207.
165. Huang, C.S. 1982. The Air Force experience in
fixed-film biological processes. Proceedings of the
First International Conference on Fixed-Film Bio-
logical Processes 3:1777.
166. Schlegel, S. 1988. The use of submerged biological
filters for nitrification. Water Sci. Tech. 20(4/5):177.
167. Cote, P.M. 1991. Total nitrogen removal from do-
mestic sewage in fixed film and suspended growth
activated sludge treatment systems. J. New Eng-
land Water Poll. Cont. Assoc. 25(1 ):33.
168. Monitor—biological treatment systems for bioecol-
ogy restoration. KLV Technologies, Victor, NY.
209
-------
-------
Chapter?
Design Considerations for Denitrification Processes with Supplemental
Substrate Addition
7.11ntroduction
As discussed in Chapter 2, the process of biological de-
nitrification can be used to remove nitrogen from waste-
water when the nitrogen is predominantly in the form of
nitrate. In municipal applications, nitrogen is present in
raw wastewater primarily in organic and ammonium-nitro-
gen form and must be converted (nitrified) to an oxidized
form (nitrite or nitrate) before biological denitrification can
take place. Nitrification and denitrification can occur sepa-
rately in distinct systems (Chapter 6) or in one system
where the process would be termed single-sludge (Chap-
ter 8). Basic chemistry, microbiology, and kinetics are all
relevant to nitrification and denitrification (Chapters 3 and
4, respectively).
This chapter presents design criteria for alternative deni-
trification systems, including suspended growth and at-
tached growth processes, that use a supplemental carbon
source. The attached growth processes that are dis-
cussed include the downflow packed-bed and upflow
fluidized-bed systems. Other systems have been investi-
gated but are not currently in full-scale use (1).
7.1.1 Classification of the Supplemental Substrate
Addition Denitrification Process
The supplemental substrate addition process generally
involves a separate-stage denitrification process, using a
distinct biological process to remove nitrate. Nitrification
is accomplished prior to the denitrification step in either
a combined carbon-oxidation and nitrification process or
in a separate system.
A combined carbon-oxidation and nitrification system with
a separate denitrification system is known as a two-
sludge system. When oxidation, nitrification, and denitri-
fication occur separately, the process is called a
three-sludge system. Regardless of whether oxidation
and nitrification are combined, a separate denitrification
process is used for both the two-sludge and three-sludge
processes. The addition of supplemental substrates to
single-sludge systems is addressed in Chapter 8.
7.1.2 Substrate Selection
A supplemental carbon source is needed in separate-
stage denitrification systems because the preceding
carbon-oxidation and nitrification step has removed nearly
all of the degradable carbonaceous material from the
wastewater. Several alternative sources of carbon have
been considered at various plants; however, only a few of
these, such as methanol, acetate, and raw sewage, have
actually been tested. Alternative sources include (2,3,4):
• Methanol
• Raw sewage or primary effluent
• Ethanol
• Methane
• Acetate
• Ketones
• Molasses and sugars
• Brewing and distilling wastes
The ideal supplemental substrate should be inexpensive,
readily available all year, essentially free of nitrogen (as
well as phosphorus, in many cases), and should have
suitable kinetics. (Substrate selection is discussed in
Chapter 4.) Neither raw sewage nor primary effluent is
generally suitable as a carbon source because of high
ammonia, organic nitrogen, and SS levels; however, if the
introduced nitrogen levels can be tolerated, these sub-
strates may be acceptable. In most cases, methanol is
the most appropriate choice and is being used in the vast
majority of installations because of its availability, low
cost, favorable sludge production, low volatile organic
compound (VOC) emissions potential, and lack of nitro-
gen and phosphorus. Because of its near universal use,
methanol is the only supplemental substrate considered
in this chapter.
The methanol dose should be monitored and controlled
carefully to prevent overdosing, which can adversely af-
fect the effluent BOD; however, some systems are re-
ported to work well with little attention to monitoring the
211
-------
methanol dose. The methanol dose is based on the ni-
trogen removal required; the methanol to nitrogen ratio
(M:N) is typically 2.5-3.0. (See Chapter 4 for discussion
of the stoichiometry.) Details regarding methanol proper-
ties and handling are presented at the end of this chapter.
7.2 Suspended Growth Systems
7.2.1 Description
Suspended growth denitrification is an activated sludge
process. The biomass is kept in suspension in a reactor
by mixing before being allowed to settle out in a clarifier,
and the majority of the settled biomass is recycled to the
reactor. In the standard activated sludge process, oxygen
Is supplied to act as an electron acceptor for the oxidation
of aerobic carbonaceous matter in wastewater. In the
denitrification process, nitrate acts as an electron ac-
ceptor and methanol is supplied to the reactor to serve
as the carbonaceous matter. The continuous flow-through
reactors are typically operated with an average detention
time of 2-3 hr (5). An aeration basin or zone must be
provided after the denitrification reactor to strip the nitro-
gen gas bubbles produced during the denitrification proc-
ess and oxidize any methanol that remains. Nitrogen gas
bubbles must be removed prior to the final clarifier to
prevent problems with floating sludge caused by adher-
ence of the nitrogen bubbles to the sludge. A portion of
the settled sludge is returned to the basin and the remain-
der is wasted to maintain the desired solids residence
time (9C). A typical suspended growth denitrification sys-
tem with supplemental substrate addition is shown sche-
matically in Figure 7-1.
A suspended growth system with methanol addition has
the following reported or potential advantages:
• Because the rate of denitrification should be greater
than if wastewater were used as the carbon source,,
reactors can be smaller.
• Excess methanol is oxidized more easily than in at-
tached growth systems.
Methanol
Nitrified,
Influent"
• Because nitrification and denitrification occur in differ-
ent stages, each process can be optimized separately.
• A high degree of nitrogen removal is possible.
• A plant retrofit can likely be accomodated.
• Suspended growth processes are well-understood by
most operators although experience with separate-
stage suspended growth systems for denitirification is
limited.
A separate-stage suspended growth system with metha-
nol addition has the following disadvantages:
• Methanol is an added expense.
• A separate clarification step is required.
• A greater number of unit processes.is required for total
nitrogen removal when compared to a single-sludge
system.
7.2.2 Design Considerations
The design considerations for suspended growth denitri-
fication systems are similar to the requirements for stand-
ard activated sludge plants. Key factors are hydraulic
residence time (HRT), solids residence time (0C), and
methanol requirements. Systems are designed on the
basis of reaction kinetics and full- and pilot-scale experi-
ence. Chapter 4 discusses the basic kinetics of sus-
pended growth denitrification. Table 4-1 summarizes
some of the kinetic coefficients that have been deter-
mined for suspended growth systems where methanol
is used as the carbon source. There is significant vari-
ation in the data presented in this table, due to differences
in test conditions and procedures. Care must be taken
if these data are used without consulting the "specific
reference.
The aerobic reactor that follows the anoxic reactor (to
strip out nitrogen gas bubbles and oxidize excess metha-
nol) should be mildly aerated so the floe does not, break
up and sedimentation is not hindered (6). The detention
time of the aerobic reactor is typically 20-60 min for ni-'
4
1
k.
>.
a, a. a.
U U \J
" .
• • . »
e « o «
t
f
Clarifier
\. — /^
• i v • •• •
t
••-••: •
Anoxic Aerobic
(Reaeration/Excess
Methanol Removal)
Rgure 7-1. Schematic of suspended growth system.
212
-------
trogen gas stripping and excess methanol oxidation. Re-
quirements for oxidizing excess methanol are further dis-
cussed in Section 7.5.7; kinetic equations for methanol
oxidation are generally used to estimate the required re-
actor residence time.
The design considerations for solids-liquid separation
(clarification) for suspended growth denitrification sys-
tems are the same as those for nitrification and standard
activated sludge systems (see Chapter 2). This section
will not attempt to discuss aspects of clarifier design. The
reader is directed to WPCF FD-8 (7) and WEF MOP 8
(8) for a more detailed discussion.
7.2.3 Design Example
The design of suspended growth denitrifying reactors with
methanol addition can be based on the kinetic equations
presented in Chapter 4. The design procedure presented
here is based on those equations, although other proto-
cols can be followed. The equations and the kinetic co-
efficients that are used are intended for illustration
purposes only.
A design example for a suspended growth activated
sludge system receiving a nitrified effluent follows. In an
actual design situation, it is necessary to develop a flow
and materials balance such as generated for the more
complex Plant B shown in Figure 2-6 and illustrated in
Table 2-16. Assume such an analysis yields the nitrified
effluent criteria summarized in Table 7-1. The effluent
criteria are for year-round total nitrogen limits and the
more stringent BOD and TSS limits shown in Table 2-10.
Table 7-1 also lists the effluent limits and the effluent
values used for this design example.
Table 7-1. Design Example: More Complex Plant B with
Year-Round Effluent Limits—Suspended Growth System
Design
Nitrified Effluent Final
Characteristic Effluent Limits Effluent
Minimum Monthly Temp.
15°C
Average Flow, ms/d
Average Flow, mgd
TSS, mg/L
CBOD5, mg/L
TKN, mg/L
(NOg + NCQ-N, mg/L
TN, mg/L
18,930
5
15
3
0.7
24.5
25.2
—~
10
10
— . *
- —
5
— •
8
8
—
•— •
3
Note: Design final effluent values are lower than the effluent limits for
reliability.
7.2.3.1 Design Example: Suspended Growth Denitrification System
1. Select the kinetic coefficients and other design constants to be used. The selection of the kinetic
coefficients can be critical to the design. The values in Table 4-1 may be used as a starting point. Since
there is significant variation in the values reported in the literature, care must be taken when selecting
values, and consideration should be given to conducting pilot studies. When considering,the need to conduct
a pilot study, one must weigh its cost against the potential savings that may result from the use of less
conservative kinetic-coefficients that might be developed from the pilot evaluation. The studies would also
provide site-specific information that might allow the application of lower design (or safety) factors.
The following design constants have been selected for purposes of illustration:
• True (gross) yield: Ys (g VSS/g COD removed) = 0.18
• Decay coefficient: bd (g VSS/g MLVSS/d) = 0.04
• Maximum substrate removal rate: qs (g COD/g VSS/d) = 10.3
• Half-saturation constant: Ks (mg/L COD) = 9.1
• Temperature correction coefficient: 6 = 1.08
• Maximum month MLSS, mg/L = 3,000
• Design MLSS, mg/L = 2,000
• Assumed MLVSS/MLSS = 0.75
213
-------
7.2.3.1 Design Example (continued)
2. Estimate the amount of methanol allowed in the effluent. Start by estimating the effluent BOD5 that will
come from the solids in the effluent.
• Assume BOD5 of the effluent VSS = 0.5 mg BODs/mg VSS (see Chapter 2)
• Assume VSS of the effluent TSS = 0.75 mg VSS/mg TSS (see Chapter 2)
. • Thus, effluent BOD5 from TSS = 8 x 0.75 x 0.5 = 3 mg/L
Estimate the amount of BOD5 allowed in the effluent that may be from excess methanol. Assume that the
BOD5 of the secondary effluent will pass through the denitrification system unchanged. Thus, the maximum
amount of BOD5 in the effluent because of excess methanol may be estimated as:
• Allowable BODS from methanol = design effluent BOD5 (Table 7-1) - secondary effluent BOD5 (Table
7-1) - BOD5 from TSS = 8 mg/L - 3 mg/L - 3 mg/L = 2 mg/L
• Convert this BOD5 to a COD value based on a BOD5:COD ratio of 0.57 for methanol (9)
• Allowable COD from methanol = 2 mg/L BOD5 -*• 0.57 = 3.5 mg/L COD
This is the maximum amount of COD allowed in the effluent because of excess methanol. The volume of
the anoxic zone can be reduced by allowing the methanol concentration in the anoxic reactor to be greater
than this amount, thus increasing the rate of denitrification. The excess methanol could be removed in the
subsequent aerobic reactor. There will be a tradeoff between the costs for supplying this excess methanol
and the capital costs of the larger anoxic reactor that would be required if the methanol concentration were
to be kept lower. For this example, a methanol concentration of about half the saturation constant will be
used, yielding a COD of 4.5 mg/L in the anoxic reactor. This is equivalent to 3 mg/L of methanol, based
on the stoichiometric CODrmethanol ratio of 1 .5.
3. Estimate what the concentration of nitrate in the effluent must be to achieve the design effluent total
nitrogen concentration. First, assume that the soluble, nonbiodegradable organic nitrogen in the effluent is
1 mg N/L. Second, assume that the TKN:VSS ratio of the solids in the effluent is 0.08 mg N/mg VSS.
• Effluent particulate N = effluent TSS x VSS:TSS x 0.08 = 8 x 0.75 x 0.08 = 0.48 mg N/L
• Thus, total organic N = 1 mg N/L + 0.48 mg N/L = 1 .48 mg N/L
• Based on the calculations in Chapter 6, it can be assumed that the ammonia concentration in the nitrified
effluent is 0.5 mg N/L
• Thus, the total effluent TKN (ammonium plus organic nitrogen) anticipated from the denitrification reactor
= 0.5 + 1.48 = 1.98 mg N/L, say 2.0 mg N/L
• Thus, the allowable nitrate in the effluent = design effluent TN - TKN = 3 mg N/L - 2.0 mg N/L = 1 .0 mg
N/L
• Base the design on an effluent nitrate of 1 .0 mg N/L
• Thus, the average nitrate removal = nitrified effluent nitrate - design effluent nitrate = 24.5 mg N/L - 1 .0
mg N/L = 23.5 mg N/L
4. Calculate substrate removal rate. Equations based on organic substrate removal (on a COD basis) are
used for this example. Assuming that the methanol will be the only rate-limiting factor,
• This rate can be related to an equivalent nitrate removal rate:
q^d-I^Y + ^ bd
214
-------
7.2.3.1 Design Example (continued)
41 1 4?
i + 0.04
= 0.91 g NOg-N/g VSS/d
Based on the nitrate removal rates presented in Table 4-1 , this value appears reasonable.
• Convert to a rate at the minimum monthly temperature:
= 2.32gCOD/gVSS/d
5. Calculate minimum required solids residence time (9[T) based on the above rate:
•r = Ysqs -bd
ftm s^sis a
o
= (0.1 8)(2.32) - 0.04 = 0.38 d"1
6. Select the peaking factor for design. From Table 2-18, a total process peaking factor of 1 .56 was selected.
7. Calculate the design SRT (9^):
• g£ = 9? x design peaking factor = 2.6 d x 1 .56 = 4.1 d
8. Calculate methanol dose required for denitrification from Equation 4-10. Assume the nitrified effluent DO
is 2 mg/L.
• Methanol dose = 2.47 x (nitrate removed) + 0.87 x (DO) + design methanol in anoxic reactor = 2.47 x
23.5 mg N/L + 0.87 x 2 mg/L + 3 mg/L = 63 mg methanoI/L
• Convert to COD: 63 mg/L x 1.5 = 94.5 mg COD/L, or 1,790 kg (3,940 Ib) COD/d
9. Calculate required anoxic HRT and reactor volume. For simplicity, assume that the incoming VSS in the
nitrified effluent do not decay further in the denitrification reactor. Also the small residual effluent methanol
COD concentration can be ignored for a slightly more conservative design.
,,nT Ys(S0-S)9g (VSS)(SRT)
-'' '. ' Fill I - --I T . mm . fnj-v
(1 + b9$ MLVSS MLVSS
0.1 8 (94.5-0) (4.1) (15) (4.1) (0.75)
~ (1 + (0.04) (4.1 ) ) (2,000) (0.75) + 2,000 (0.75)
= 0.04 + 0.031 =0.071 d
'"'•• = 1.704 hr
Use a design HRT of 1 .75 hr.
• Calculate the volume of the reaction/postaeration tank, assuming a 40-min HRT to provide adequate
stripping of nitrogen gas and removal of the excess methanol
• The total reactor HRT is 2.4 hr, 73% anoxic and 27% aerobic:
Volume = 18,930 m3/d x 2.4 hr x (1 d/24 hr) = 1,893 m3 (500,000 gal)
215
-------
7.2.3.1 Design Example (continued)
• Use three reactors, 4.6 m (15 ft) deep with a Iength:width ratio of 3:1: •
Reactor Volume = 1,893 m3/(3 reactors)
= 631 m3 (22,280 cu ft) each
Reactor Volume = (3W) (W) (Depth)
631 m3 = (3W2) (4.6)
Width = 6.8 m (22.3 ft)
Length = 3 x 6.8 m = 20.4 m (67 ft)
Thus, each of 3 reactors is 20.4 m x 6.8 m x 4.6 m (67 ft x 22.3 ft x 15 ft) deep. Allow approximately 0.6
m (2 ft) freeboard in final tank sizing (total depth of 5.2 m [17 ft]).
Compartmentalizing the reactor to promote plug flow characteristics will ensure the best performance.
10. Calculate the size of the denitrification clarifiers. Check sizing based on a surface overflow rate (SOR)
and a solids loading rate (SLR) and select the larger of the two. Use maximum MLSS of 3,000 mg/L.
• Assume a design SOR of 20.4 m3/m2/d (500 gpd/sq ft) at average conditions
• Area required for SOR = 18,930 m3/d H- 20.4 m3/m2 = 928 m2 (10,000 sq ft)
• Assume a design SLR of 122.3 kg/m2/d (25 Ib/d/sq ft) at the maximum month MLSS concentration of
3,000 mg/L (3 kg/m3). Assume a return sludge rate of 0.6 the incoming flow
• Size required = (1 + 0.6) x 18,930 m3/d x 3 kg/m3 -s- 122.3 kg/d/m2 = 743 m2 (8,000 sq ft)
• Thus, SOR controls the design sizing. Use three clarifiers to provide adequate surface area:
(928 m2)/3 clarifiers = 309 m2/clarifier
K D2/4 = 309 m2
D m 19.9 m (65 ft)
Use three 19.9-m (65-ft) diameter clarifiers.
11. Calculate the horsepower required for mixing the denitrification reactor:
• Assume 13.2 hp/1,000 m3 (50 hp/Mgal) (see Chapter 2)
Anoxic Basin Volume = 1,893 m3 x 73% anoxic = 1,382 m3 (370,000 gal)
hp » (1,382) (13.2/1,000) = 18.5 hp/3 basins, or 6.2 hp/basin
12. Calculate rate of waste sludge production.
Net Yield = YSN6T=Ys/(1+bd9g)
Ys^ = (0.18) / [(1+0.04 (4.1)]
= 0.155 kg VSS/kg COD removed
= (0.155 kg VSS/kg COD) / (0.75 kg VSS/kg TSS) •
= 0.21 kg TSS/kg COD removed
Waste Sludge = Q (S0 - S) (YSNET) + Q (Influent TSS removed)
= [(18,930) (94.5 - 0) (0.21) + (18,930) (15 - 8)]/(1,000)
= 376 + 133 = 509 kg (1,120 lb)/d
A process flow diagram is shown in Figure 7-2 for the separate-stage suspended growth denitrification
system.
216
-------
Mixer
Methanol
Feed
Nitrified
Influent
from
Nitrification
System
Denitrification Reactors
Reaeration Zone
Denitrification Clarifiers
Figure 7-2. Design example schematic of suspended growth system.
7.2.3.2 Alternate Design Approach
As previously indicated, there are alternate design ap-
proaches to the one illustrated in Section 7.2.3.1. An
alternative procedure*would be to rely on rate data such
as presented in Figure 7-3. At 15°C (59°F), denitrification
rates of 0.05-0.18 kg NOg-N/kg MLVSS/d are shown. This
can be compared to the rate for the design in Section
7.2.3.1, which equates to a removal rate of 0.21 kg
NOg-N/kg MLVSS/d for the denitrification of 23.5 mg/L of
NOs-N in the 1.75-hr detention time anoxic reactor, oper-
ated at an MLVSS of 1,500 mg/L. Using the data in Figure
7-3 as an alternate design basis would produce a larger
anoxic reactor whose size would depend on the rate
selected.
7.2.4 Case Study: River Oaks Advanced Wastewater
Treatment Plant (Hillsborough County, Florida)
The River Oaks Advanced Wastewater Treatment Plant
(AWTP) was upgraded in three phases beginning in 1986.
The upgrades increased capacity from 131 L/s (3 mgd)
to 438 L/s (10 mgd) and included the addition of sus-
pended growth denitrification with methanol addition. The
completed plant includes flow equalization, headworks,
primary clarification, aeration, secondary clarification with
final flocculation, suspended growth denitrification, and
disinfection. A flow schematic is shown on Figure 7-4.
The denitrification system includes two 3,940-m3 (1.04-
Mgal) tanks operated in parallel with 16 cells each. The
cells are further divided into anoxic and aerobic zones.
The first 10 cells operate in the anoxic mode; the next 4
can operate in either the anoxic or aerobic mode; and the
last 2 cells operate only in the aerobic mode.
The plant discharges to Tampa Bay and was designed to
meet the following stringent effluent limits: BOD5 =
2 mg/L, TSS .= 2 mg/L, TN = 1.2 mg/L, and TP = 0.4
mg/L, all based on a maximum monthly flow of 526 L/s
(12 mgd). The limits have since been relaxed to 5, 5, 3
and 1 mg/L, for BOD5, TSS, TN, and TP, respectively, on
an annual average. Monthly plant operating data are
shown in Table 7-2.
The River Oaks AWTP has demonstrated the ability to
keep TN levels below 1.0 mg/L when operated at approxi-
mately 70 percent of design flow. The treatment proc-
esses employed have provided for flexible and reliable
treatment and have demonstrated the ability to remove
nitrogen to low levels. The methanol used for denitrifica-
tion represents about 10 percent of the plant O&M costs.
7.3 Downflow Packed-Bed Systems
7.3.1 Description
The downflow packed-bed system is an attached growth
process. Physically, it is identical to a deep-bed downflow
sand filter. Denitrifying microorganisms attach to the filter
media which provides the support system for their growth.
Methanol is added upstream of the packed-bed filter and
the nitrified effluent is filtered through the media. The
packed-bed filter system is well suited for denitrification
because it provides the necessary hydraulic detention
time for the biological reaction to take place, and the open
water surface in the filter is sufficiently limited to minimize
transfer of oxygen by surface aeration. A schematic of a
typical system is shown in Figure 7-5, and a cross-section
schematic is shown in Figure 7-6. The packed-bed sys-
tem is a proprietary process of TETRA Technologies,
217
-------
i
-> Below)
Preliminary
Treatment
Clarification
Aeration
(BOD Removed
and Nitrification)
Clarification
Denitrification
Tanks .
Flocculation/
Clarification
Filtration
Post-Aeration
Disinfection
Figure 7-4. Schematic of River Oaks Advanced Wastewater Treatment Plant (from Reference 11).
which holds a patent based on a previous Dravo patent
(12). A representative installation list is shown in Table 7-3.
The filter is composed of coarse, hard, round, high-den-
sity media. These media filter out solids and serve as a
support system for the the denitrifying microorganisms.
The packed-bed system also eliminates the requirement
for downstream filtration or clarification.
7.3.2 Design Considerations
As denitrification occurs, nitrogen gas accumulates in the
filter media, which increases the headloss. The nitrogen
gas bubbles are periodically released from the media by
taking the filter off line and applying backwash water for
a few seconds—a process commonly called the nitrogen
release cycle (NRC) or filter bumping (13). The frequency
of the NRC is a function of both nitrate removal and a
minimum acceptable time between cycles, usually no less
than 1 hr. Usually, a filter needs to be bumped once every
4-8 hr, depending on the nitrogen loading.
As with a conventional gravity filter, SS gradually accu-
mulate in the filter, and part of the nitrogen removal
will occur because nitrogen in the SS is removed. Physi-
cal-removal mechanisms may include straining, sedi-
mentation, and particle interaction, Chemical-removal
mechanisms may include adsorption, coagulation-floccu-
lation, and biological activity. As with all filters, no one
218
-------
Table 7-2. River Oaks Wastewater Treatment Plant: Monthly Performance Data
Plant Influent
Denitrification Process
Influent
Denitrification Process
Effluent
Final Plant Effluent
Flow Hor)c TS
Month Us mgd mg/L m<
S TKN NHJ-N BOD5 TSS TKN NHj-N NOx-N BOD5 TSS TKN NHj-N NOx-N
I/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L
SEP 90 305 6.96 167 183 31.6 23.7 2.0
OCT90 316 7.21 174 156 34.0 23.4 2.0
NOV90 292 6.66 212 173 34.9 24.5 3.0
DEC 90 308 7.03 237 200 36.5 28.0 3.0
JAN 91 329 7.54 215 185 33.4 27.3 3.0
FEB91 334 7.62 194 133 35.6 19.6 5.6
MAR 91 317 7.23 213 167 38.7 26.2 3.6
APR 91 290 6.61 200 142 38.0 26.0 3.7
MAY 91 317 7.23 173 146 37.7 24.7 2.4
JUN91 309 7.04 177 130 34.5 23.4 3.2
JUL91 340 7.77 167 127 31.5 21.5 2.9
AVG. 314 7.17 194 158 35.1 24.4 3.1
Methanol
Storage Tank
Methano
Pump
To
0.5 1.27 0.11 13.20 2.0
0.7 1.59 0.12 11.37 2.0
0.8 1.47 0.13 16,30 3.0
0.8 1.56 0.14 15.90 4.0
0.1 6.00 2.39 13.32 2.0
10.7 2.65 0.55 13.36 4.2
7.4 2.09 0.26 14.79 4.3
4.9 1.64 0.17 14.54 3.6
3.7 1.79 0.18 12.74 6.0
3.6 1.53 0.13 12.39 4.8
7.1 1.31 0.14 10.21 3.9
3.7 2.08 0.39 13.47 3.6
Nitrified Influent A
. Methanol
Feed
4
7.0 0.93 0.10 0.56
6.0 0.42 0.10 0.07
9.0 1.04 0.10 0.33
11.0 2.40 0.10 0.56
6.0 4.62 2.35 0.71
8.0 1.23 0.15 0.76
8.1 1.36 0.11 0.48
5.2 1.41 0.10 0.14
5.7 1.28 0.10 0.19
8.2 1.40 0.10 0.23
5.8 1.04 0.10 0.24
7.3 1.56 0.31 0.39
k
k. f
Clear Well Denitrified Effluent
Backwasl
Water
/" 1
BOD5 TSS TKN NHj-N NOx-N TN
mg/L mg/L mg/L mg/L mg/L mg/L
1.00 1.00 0.64 0.10 0.47 1.12
1.00 0.00 0.73 0.10 0.10 0.80
1.00 1.00 0.71 0.10 0.30 1.01
1.00 1.00 0.74 0.10 0.58 1.33
1.00 1.00 3.08 2.28 0.72 3.80
1.00 0.60 0.88 0.17 0.83 1.71
0.83 0.55 0.72 0.10 0.55 1.21
0.87 0.13 0.75 0.10 0.13 0.89
0.80 0.32 0.89 0.10 0.25 1.14
1.09 0.50 0.87 0.10 0.34 1.21
1.65 0.40 0.79 0.10 0.20 0.98
1.02 0.59 0.98 0.30 0.41 1.38
I
Media "^
Support Gravel
Underdrain
i
k.
Backwash
Air
f J Backwash Supply
\ / Pump
Air Scour Blower
Note:
Backwash waste
not shown
Figure 7-5. Linear schematic of downflow packed-bed system.
219
-------
Media
Support
Gravel
Underdrain
Backwash
Air
Filtering
Mode
Backwash Water
Backwash
Mode
Figure 7-6. Cross-section schematic of downflow packed-bed system.
mechanism can account for the total effect. The SS
trapped in the filter are removed by backwashing, usually
with an air/water wash in a step that also sloughs off
excess biomass. Part Of the backwash waste may be
returned to the filter to reseed the filter with denitrififying
microorganisms. The requirement for reseeding is de-
pendent on the influent TSS, the efficiency of backwash-
ing, and the presence of methanol-acclimated biomass.
Therefore, it may not be desirable to completely clean the
filter during backwash. Return flows equivalent to 880-
1,170 m3/m2/d (15-20 gpm/sq ft) of filter for 1-2 min may
be used if reseeding is needed. After backwashing, a loss
of denitrification capability may be observed because of
the loss of accumulated biomass. A cleaner filter placed
In operation requires some time to reestablish the
biomass and may experience some decreased perform-
ance when first placed back in service. Therefore, several
small filters will produce a better blended effluent than a
few large filters. If the accumulation of SS were allowed
to continue without backwashing, the filter would become
clogged arid the frequency of gas-bumping would have
to be increased. Under normal conditions, headlosses will
require that the filter be automatically backwashed .every
1-5 d, which is comparable to the backwashing frequency
for a filter used only for SS removal. '"'•-!' -
,.-. .-' , •:-.• : IM-. •::;.::• ••'
The design of denitrification filters must consider filtration,
denitrification kinetics, and the limitations on .{he. fre-
quency of the NRC. The process hydraulic loading rate
is generally 58-117 m3/m2/d (1-2 gpm/sq ft).with an
empty bed contact time of 30 min or greater. Figure 7-7
illustrates typical design curves based on empty bed con-
tact time.
Knowledge of reaction rates is necessary to size an
attached-growth denitrification reactor. In estimating
biomass reaction rates, the level of biomass effective in
220
-------
Table 7-3. Packed-Bed Denitrification (Partial Installation List)
Capacity
Facility and Location
Tampa, Florida
Hookers Point AWTP
Seminole County, Florida
NW Area Regional
WW Facility Expansion
Port Orange, Florida
Hillsborough County, Florida
Valrico Wastewater Facility
U.S. Home
Brandon, Florida
Purity Farms
Clearwater, Florida
Hillsborough County, Florida
Dale Mabry AWTP
Piney Orchards, Maryland
Hillsborough County, Florida
Falkenburg RD AWTP
Altamonte Springs, Florida
Florida Cities Water Co.
Fiesta Village
Fort Myers, Florida
Kanapaha Wastewater
Treatment Plant
Description of Packed-Bed Denitrification System
Twelve 97 m2 (1 ,050 sq ft) filters
Nineteen 93 m2 (1 ,000 sq ft) filters
Two 46 m2 (500 sq ft) filters
Six 52 m2 (560 sq ft) filters
Three 46 m2 (500 sq ft) filters
One 19m2 (200 sq ft) filter , , :
One 9.3 m2 (100 sq ft) filter
Five 60 m2 (650 sq ft) filters
Four 9.3 m2 (100 sq ft) filters
Five 46 m2 (500 sq ft) filters
Seven 56 m2 (600 sq ft) deep-bed filters for tertiary *
filtration, denitrification, and virus control of municipal
sewage treatment plant effluent
Four 37 m2 (400 sq ft) filters for nitrate reduction and
SS removal ' . .
Six 46 m2 (500 sq ft) filters
Us
4,208
110
351 ,
. 132
33
10
264
53
264
110
548
„ . 96
220
308
• ;'770
mgd
96.0
2.5
8.0
3.0
0.75
0.23
6.0
1.2
6.0
2.5 (Avg.)
12.5 (Peak)
2.2 (Avg.)
5.0 (Peak)
7.0 (Avg.)
17.5 '(Peak)
Gainesville, Florida
Parkland III Expansion
..(slip, New York ,
Fairfield Village
New York
Southampton Hospital
Southampton, New York
Blue Ridge Condo. Medford
Broo'khaven, New York
'•• »-;•"-!0"V~ '• - ' ..•
Parkland III
'York ' •
Parr Village
Yaphank, New York
i ••!.-.; f-" . :,i
Deep-bed gravity denitrification-effluent polishing system.
including four 5.6 m2 (60 sq ft) filters .....'"
Two 5.6 m2 (60 sq ft) deep-bed sand filters for effluent
polishing and denitrification
Two 4.7 m2 (50 sq ft) deep-bed sand filters for effluent
polishing and denitrification
One deep-bed sand filter system. System includes-threes
deep-bed gravity filter cells 5.6 m2 (60 sq ft) each
One deep-bed gravity filtration system for effluent polishing
and denitrification. System includes four deep-bed filter
cells 5.6 m2 (60 sq ft) each • •''• •
Three 4.8 m2 (52 sq ft) deep-bed sand filters
11,;
4
0.085'
. 0.2" ;
0.45
221
-------
100
80
60
40
20
J
/
"
/
/
/
V
/ ^
_ ;
^-— _WE
"- W£
n Hook«
• Kana
A Fiesta
^ Edga
0 Richn
• Blue F
— . — • —
ter Tempera
ter Tempera
1 — - —
ure16-19°C
ure11-15°C
3rs Point AWT, Tampa, FIc
>aha WWTP, Gainesville,
Village, Ft. Myers, Florida
rtown, Massachusetts
nond Pilot Study
'lains Denitrification Study
—A
rida
Florida
20 40
Empty Bed Detention Time, min
60
80
Figure 7-7. Typical design curves for empty-bed contact time (from Reference 14).
denitrification must also be known. One approach is to
estimate the amount of biomass on the media surface
and then use measured reaction rates per unit of
biomass. This approach is of limited usefulness in design
applications because there are insufficient data available
at the present time to predict the level of biomass that
will develop on the media. Biomass development is de-
pendent on hydraulic regime, type of media, loading,
means for promoting sloughing, and possibly the tem-
perature of operation.
Denitrification rates can also be expressed in terms of
nitrate removal rates per unit of filter surface. One ap-
proach uses mass of nitrate removed/unit area of filter
surface/d (1). Table 7-4 shows rates for selected applica-
tions. (These rates are also given for each of the design
examples presented in this section.) On this basis, high
surface-removal rates would reflect extensive biological
film development, whereas low surface-removal rates
would reflect minimal surface film development! The sur-
face denitrification rate varies considerably among
the various denitrification column configurations and is
affected by the loadings under which the process is
operated. " ...
As the applied nitrogen concentrations increase, the rate
of nitrogen gas accumulation increases, forcing more fre-
quent NRCs. As a practical matter, the maximum NRG
frequency that can be tolerated is typically one every hour
(15). The relationship among nitrate concentration, load-
ing, media underdrain configuration, and NRG is not well
defined. The effect of influent nitrate concentrations and
loading rate on effluent quality is shown in Figure 7-8
using, data' from actual, operating experience at 20°C
(68°F). As.influent nitrate levels increase, the loading or
application rate decreases for a given effluent criterion.
Loading rates will also decrease predictably with de-
creased temperatures. Figure 7-9 illustrates the relation-
ship between NO^-N removal and loadings based on
limited data from two installations. This curve was created
by calculating the pounds of nitrate denitrified/sq ft/NRG
(which ranged between 0.03 and 0.09) and then stand-
ardizing the concentration based on the concentration
that could be denitrified between two NRCs, 2 hr apart
(19). While the data presented in this curve are based on
actual experience, recent media and underdrain modifica-
tions have decreased NRG frequency at some plants (8).
As discussed above, downflow packed-bed systems are
effective in removing both SS and nitrogen. However,
• when'considering the use of this process as both an
effluent filter and a denitrification system, an important
design factor—cost—should be kept in mind. It has been
claimed that combining the functions of filtration and de-
nitrification reduces tankage and equipment requirements
.and therefore yields .cost savings irvp.la.nts requiring filtra-
tion ('19)". However, column loading criteria are different
222
-------
Table 7-4. Selected Downflow Packed-Bed Application Rates
Facility
Hookers Point"
Kanapahab
Dale Mabryb
Tetra Design
Value0
Average
Hookers Point"
Kanapaha"
Dale Mabry"
Tetra Design
Value0
Average
Temp.,
°C
20
NA
24
15
Temp.,
°C
20
NA
24
15
Media
Depth, m
• 1-7
1.8
1.8
1.8
Media
Depth, ft
5.5
6
6
6
kg NOX-N
removed/m2
reactor/d
2.7
2.3
0.54
2.4
2.0
Ib NOX-N
removed/sq ft
reactor/d
0.55
0.48
0.11
0.5
0.41
kg NOX-N
removed/1,000 m2
media surface area/d
1.6
1.3
0.29
1.4
1.15
Ib NOX-N
removed/1,000 sq ft
media surface area/d
0.33
0.27
0.06
0.28
0.23
kg NOX-N
removed/1 ,000
m3 reactor/da
1,600
11280,
290
1,330
1,125
Ib NOX-N
removed/1 ,000
cu ft reactor/da
100
80
18
83
70
a Based on 985 m2/m3 (300 sq ft of surface area/ou ft) of reactor.
b Based on operating data, design criteria may be higher.
0 Based on design example in Chapter 7.
9 _
8 -
7 _
6 _
I 4_|
£
3
2 _
1 _
Treatment of Nitrified Effluent Sewage
Summer Temperature Conditions, 20°C
Performance Characterization at Stoichiometric Methanol Feed = 3 Ib Methanol/lb NOj -N
I I
4 5
Surface Loading, gpm/sq ft
Figure 7-8. Effect of nitrate concentrations on loading rates in downflow packed-bed systems (from Reference 16).
223
-------
0.5
Hydraulic Loading, gpm/ft2
Figure 7-9. Downflow packed-bed denitrification performance (from References 17 and 18).
for the functions of filtration and nitrogen removal. For
effluent filtration, fairly high hydraulic loadings can be
applied (234-351 m3/m2/d [4-6 gpm/sq ft]). However, for
filters 0.9-1.8 m (3-6 ft) deep acting as denitrification
columns, available data indicate that hydraulic loading
should be between 29-88 m3/ma/d (0.5-1.5 gpm/sq ft) at
a wastewater temperature of 10°C (50°F) (1). Thus, to
accomplish denitrification at 10°C (50°F), it could be nec-
essary to have filter surface areas five times as large as
those required for filtration alone. Consequently, an eco-
nomic analysis must be done to determine the most eco-
nomic process configuration for the particular system.
The physical design characteristics of downflow packed-
bed systems are essentially the same as for any deep-
bed filter system. Much like a typical filter, the downflow
packed-bed filter system consists of three major compo-
nents: the media, the underdrain, and the backwash fa-
cilities. Typical design criteria are shown in Table 7-5. No
attempt has been made to discuss all factors in filter
design. The reader is directed to the Process Manual for
Suspended Solids Removal (20), to Water Treatment
Plant Design (21), and to MOP 8 (8).
The media must be carefully selected to ensure proper
size, shape, hardness, and density. Also, it must be large
224
-------
Table 7-5. Typical Design Criteria for Downflow
Packed-Bed System
Media Type
Media Depth
Backwashing
Application Rates
Reseeding
Nitrogen Release Cycle
Flow Control
Sand: Effective size 1.8-2.3 mm
Sphericity: 0.8-0.9
Specific gravity: 2.4-2.6
Dual media: coal also used
1.2-1.8m (4-6 ft) . .
Duration: 5-10 min
Water: 350-470 m/d
(6-8 gpm/sq ft)
Air: 1.5-1.8 m/min
(5-6 cfm/sq ft)
60-120 m/d
(1-5 gpm/sq ft)
Duration: 1-2 min
Rate: 880-1,170 m/d
(15-20 gpm/sq ft)
Duration: Up to 5 min
Interval: 1-6 hr
Rate: 293 m/d (5 gpm/sq ft)
Variable declining rate or
Influent flow control or
Effluent flow control
enough to facilitate bed penetration but small enough to
provide a high surface area for microorganisms and to
prevent solids breakthrough (19). Media typically range
in size from 1.8 to 2.3 mm. Coarser, deeper media filters
are often used for denitrification rather than the final me-
dia filters that are typically found in municipal wastewater
applications.
Rounded media are typically used because of the im-
proved backwash characteristics of round grains, which
tend to rotate during backwash and create a vigorous
mutual scrubbing action. The scrubbing results in particu-
larly clean media after backwashing. However, this agita-
tion necessitates that the filter media be hard.
The density of the media is also a key factor. Bed expan-
sion is a function of particle size, shape, density, and
water flow rate. Higher density media require a higher
flow rate to provide greater bed expansion. TETRA Tech-
nologies uses high-density sand with a specific gravity of
2.5 to minimize media loss during backwash because
the denser media are more readily retained in the filter
vessel.
While TETRA Technologies' underdrain system consists
of precast concrete blocks, many other underdrain sys-
tems such as nozzles and high-density plastic distribution
blocks (see Figure 7-10) have been used in filtering ap-
plications other than denitrification. These alternative un-
derdrain systems are discussed more thoroughly
elsewhere (22,23). The underdrains are designed to dis-
tribute the backwash water and air and to collect the
filtered water. The backwash system must operate in con-
junction with the underdrain, which distributes the back-
wash water evenly to the media. The backwash system
consists of backwash water pumps and, if desired, back-
wash blowers. The manufacturer generally uses a com-
bined air/water wash: for 15 min, air is pumped at 1.8
m3/m2/min (6 cfm/sq ft); water is pumped at 351 m3/m2/d
(6 gpm/sq ft). The air/water wash is followed by a water-
only wash at 468 m3/m2/d (8 gpm/sq ft), for 5 min. A
separate air/water wash is also beneficial in scouring the
media. Rates as high as 1,170 m3/m2/d (20 gpm/sq ft)
would be required for full-bed expansion. Air entrainment
in the influent and the media should be minimized so that
nitrogen removal is not hindered. The water-only wash
can be used to purge the media of entrained air. In a
denitrification filter, this is particularly important because
anoxic conditions must be maintained.
Backwash Water
Trough
Air Scour
Supply
Gravel Support
Underdrain
Backwash Water
Trough
Nozzles
Nozzle
Bushes
Filter Media
Underdrain
System
Filter Effluent and
Backwash Water
Supply
High-Density Plastic Underdrain
Formwork
Panels
Nozzle Underdrain
Support
Wails
Figure 7-10. Alternative filter underdrain systems.
225
-------
Presenting useful design criteria based on kinetic reac-
tions is difficult, partly because of the variability of the
systems but also because of the proprietary nature of the
process. However as explained below, there are some
useful approaches for characterizing denitrification on at-
tached growth systems.
7.3.2.1 Design Examples
The design examples that are presented for downflow
packed-bed systems are based on the discussion in
Chapter 4 about the kinetics of denitrification and the
fundamental equations used in system design. These ex-
amples also draw from the influent criteria discussion in
Chapter 2. The application rates used in this example are
for illustration purposes only and are based on typical
data; actual rates must be selected on a site-specific
basis. The cases presented below involve different efflu-
ent requirements (see Table 7-6). A process flow sche-
matic for a downflow packed-bed system is provided in
Figure 7-11.
Table 7-6. Influent Characteristics and Design Effluent
Limits for Denitrification Design Examples: Downflow
Packed-Bed System
Case 1 Case 2
Nitrified Effluent Effluent
Characteristic Effluent Limits Limits
Minimum Monthly Temp.
Average Flow, m3/d
Average Flow, mgd
Peak Day Week, m3/d
Peak Day Week, mgd
TSS, mg/L
CBOD, mg/L
COD, mg/L
TKN, mg/L
(NOi+NOS-N, mg/L
NHJ-N, mg/L
TN, mg/L
—
—
—
—
15
3
33
1.8
23.4
0.05
26.5
15°C
18,930
5
28,400
7.5
30
30
—
—
7
2
10
15°C
18,930
5
28,400
7.5
10
10
31 .
—
3
2
6
Nitrified
Influent from-
Nitrification
System
Filter
Influent
Methanol Feed
Blowers
Air Supply
Backwash Pumps
(at the Ballast Pond)
Filter Effluent
Filter
(Typical of 6)
Gullet
(Typical)
Figure 7-11. Design example schematic of downflow packed-bed system.
226
-------
Design Example: Case 1—Downflow Packed-Bed Denitrification System
1. Calculate nitrate removed:
(23.4 - 7 mg/L) (18,930 m3/d)/1,000 = 311 kg (684 Ib) NOx-N/d
• Secondary effluent meets TSS discharge criteria, so TSS removal is not a parameter of concern; the
filter will reduce the TSS to 5. mg/L -•--....?•
• Calculate TSS removed: ---..-...-•
(15-5 mg/L) (18,930 m3/d)/1,000 = 189 kg (417 Ib) TSS/d
2. Determine filter area based on nitrogen removal: .
• 2.45 kg NOx-N/m2/d (0.5 Ib NOx-N/sq ft/d) is assumed based on TETRA Technologies operating
experience at 15°C
(311 kg NOx-N/d) / (2.45 kg NOx-N/m2/d) =127 m2 (1,368 sq ft) of filter
3. Calculation of filter area based on solids filtration is not needed for effluent limits.
• Nitrogen removal conditions control
• Assume: 127m2 (1,368 sq ft) of filter area
• Determine number and size of filters
, « Assume: six filters, with five in-service
127 m2/5 filters = 25.4 m2 (274 sq ft)/filter
• Assume: six 3 m x 8.5 m (10 ft x 28 ft) filters with one out of service for backwash
Total filter area = (5 filters) (25.5 m2/filter) = 127.5 m2 (1,400 sq ft)
• Calculate actual loading rate:
(18,930 m3/d) •*• (127.5 m2) = 148 m3/m2/d (2.52 gpm/sq ft)
4. Calculate backwash requirements:
• Assume: 1) maximum backwash rate of 468 m3/m2/d (8 gpm/sq ft) on one filter; 2) average backwash
rate of 351 m3/m2/min (6 gpm/sq ft) on one filter
(25.5 m2) (468 m3/m2/d) / (1,440) = 8.3 m3/min (2,200 gpm) maximum backwash rate
(25.5 m2) (351 m3/m2/d) / (1,440) = 6.2 m3/min (1,640 gpm) average backwash rate
• Assume: backwash air rate of 1.8 m3/m2/min (6 cfm/sq ft) on one filter
(25.5 m2) (1.8 m3/m2/d) = 45.9 m3/min (1,680 cfm) backwash air.
• Calculate volume of backwash flow:
Duration = 15 min at average backwash flow and 5 min at maximum flow
Volume = (6.2 m3/min)(15 min) + (8.3 m3/min)(5 min)
'=134.5 m3 (35,400 gal)
5. Calculate backwash frequency: ....:. .
• Assume: 7.34 kg/m2 (1.5 Ib/sq ft) TSS removed/backwash (based on TETRA Technologies operating
experience) - •-;\' . •;:•;•'.,•.;:'• .,„;'• '"' '•'•-• --•'<•
Backwash = (189 kg TSS/d) + [(7.34 kg TSS/rji2) (127.5 m2/filter)] = 0.20 backwashes/d/filter
Or— '. ; .; .'.' '''"'• " :
• One backwash every 120 hr/filter (with six filters, this is approximately one backwash per day)
• The backwash volume generated per day is:
(1 backwash/d) (134.5 m%ackwash) = 134.5 m3 (35,400 gal)/d
227
-------
Design Example: Case 1 (continued)
This equals 0.7 percent of the average flow; therefore, the actual hydraulic application rate will be
0.7 percent higher (i.e., 148 m3/m2/d x 1.007 = 149 m3/m2/d).
6. Calculate frequency of nitrogen release cycle (NRC):
• Assume: 0.245 kg NOx-N/m2/NRC (0.05 Ib NOx-N/sq fl/NRC) (Reference 18)
(311 kg NOx-N/d) / [(0.245 kg NOx-N/m2/NRC) (127.5 m2)] = 9.9 NRC/d
(24 hr/d) / (9.9 NRC/d) = 2.4 hr/NRC
NRC frequency = 2.4 hr
7. Calculate methanol requirements.
• Assume: 3 kg methanol/kg of NOX-N removed.
Methanol = (311 kg NOx-N/d) (3 kg methanol/kg NOX-N)
» 933 kg (2,055 Ib) methanol/d
(933 kg/d) (1 L/0.79 kg) = 1,181 L (312 gal)/d
Methanol dose = 49.2 L (13 gal)/hr
• Alternatively, methanol requirements can be computed from Equation 4-10. If the influent DO to deni-
trification Is assumed to be 3 mg/L, and minimal residual methanol, the methanol requirement is:
(311 kg)(2.47) + (0.87)(3)(18,930 m3/d/1,000) « 817 kg (1,800 Ib)
8. Calculate biomass production
• From suspended growth design example (Section 7.2.3.1), use 0.18 kg VSS/kg of COD removed
• Assume: All COD removed is methanol
COD removed = (933 kg methanol/d) (1.5 kg COD/kg methanol) = 1,400 kg/d
VSS = (0.18 kg VSS/kg COD) (1,400 kg COD/d)
= 252 kg (555 Ib) VSS produced/d
At 75 percent volatile:
• TSS produced = 252/0.75 = 336 kg (740 Ib) TSS/d
• Compare to yield based on nitrate removed:
Assume: Yield = 0.5 kg/kg of NOX-N removed (TETRA Technologies recommendation).
(311 kg NOx-N/d) x (0.5 kg biomass/kg NOX-N removed)
* 156 kg (344 Ib) biomass/d
9. Calculate approximate motor horsepower
• Assume: Two backwash pumps are in service, with one on standby ,
• Capacity of each = (8.3 m3/min)/2 = 4.15 m3/min = 1,100gpm
• Assume: Capacity is 4.5 rrrVmin @ 9.14 m TDH (1,200 gpm @ 30 ft TDH)
hp = [(Q)(TDH)(Sp. Gravity)] / [(3,960)(Pump Eff.)(Motor Eff.)]
Where:
Q = 1,200 gpm
TDH = 30 ft (assumed) «
Sp. Gravity = 1 '
Pump EFF = 75% (assumed)
Motor EFF = 90% (assumed) *
hp = [(1,200) (30) (1)] / [(3,960) (0.75) (0.90)]
= 13.5hp ji)
Kw = (hp) (0.746) = 10.1 Kw
228
-------
Design Example: Case 1 (continued)
• Calculate blower horsepower; assume that one blower is in service and one is on standby.
Capacity = 45.9 m3/min(1,680 cfm) ,.
hp = t(w)(R)(T1)-(550)(n)(e)n(P2/Pi)n-1] , ,,
Where:
w = weight flow of air, Ib/sec
R =' gas constant (53.5)
T! = absolute inlet temperature, °R
P! = absolute inlet pressure, psia
P2 = absolute outlet pressure, psia
n = (k- 1)/k = 0.283 for air (k = 1.395 for air)
e = efficiency (usual range for compressors is 70 to 80 percent; assume 70 percent)
The density of air at 20°C (68°F) is 1.20 kg/m3 (0.075 Ib/cu ft); therefore, the weight flow of air is:
w = (45.9 m3/min) (1 min/60 sec) (1.20 kg/m3)
= 0.92 kg (2.0 lb)/sec
• Assume an inlet temperature of 68°F, an inlet pressure of 14.7 psia, and an average outlet pressure of
22.7 psia (pressure may be 3 to 4 psi greater at startup).
hp = [(2.0)(53.5)(460 + 86)/(550)(0.283)(0.7)] [(22.7/14.7)a283 - 1] ,
= 70.2 hp
kW = (73.7) (0.746) = 52.3 kW
Design Example: Case 2—Downflow Packed-Bed Denitrification System
1. Calculate Nitrate Removed:
(23.4 - 3 mg/L) (18,930 m3/d)-^ (1,000) = 386 kg (850 Ib) NCyN/d .
« Calculate mass of TSS removed:
Effluent requirement is 10 mg/L TSS; however, the filter will produce an effluent of approximately
5 mg/L TSS. See Case 1 for solids removal of 189 kg TSS/d (417 !b/d).
2. Determine filter area based on nitrogen removal:
(386 kg NOx-N/d) / (2.45 kg NOx-N/m2/d) =158 m2 (1,700 sq ft) of filter
As a conservative measure, an effluent nitrate concentration of 0 mg/L could also be assumed; removal
would then be equal to 23.4 mg/L:
(23.4 - 0 mg/L) (18,930 m3/d) -5- (1,000) = 443 kg (976 Ib) NOx-N/d
and the resulting filter area would be 181 m2 (1,950 sq ft) ' '
3. Calculation of filter area based on solids filtration is not necessary because of effluent limits. However,
a typical average filtration rate would be 176 m3/m2/d (3 gpm/sq ft): '
(18,930 m3/d) / (176 m3/m2/d) = 108 m2 (1,162 sq ft) of filter
Therefore, nitrogen requirements control the filter sizing.
Determine number and size of filters; assume six filters, with five in service and one in backwash:
229
-------
Design Example: Case 2 (continued)
158 ma/5 filters = 31.6 m2 (340 sq ft)/filter
Assume six filters each 3 m x 10.7 m (10 ft x 35 ft) with one out of service for backwash.
Total filter area = (5 filters) (3 m) (10.7 m) = 160.5 m2 (1,750 sq ft)
160.5 m2/5 filters = 32.1 m2 (345 sq ft) / filter
Calculate actual loading rate:
(18,930 m3/d) / (160.5 m2) = 118 m3/ma/d (2.0 gpm/sq ft)
4. Calculate backwash requirements
• Assume: Maximum backwash rate of 468 m3/m2/d (8 gpm/sq ft) on one filter
• Assume: Backwash rate of 351 m3/m2/d (6 gpm/sq ft) on one filter
(32.1 m2) (468 m3/m2/d) / (1,440) =10.4 m3/min (2,750 gpm) maximum backwash rate
(32.1) (351) / (1,440) =7.8 m3/min (2,070 gpm) average backwash rate
• Assume: Backwash air rate of 1.8 m3/m2/min (6 cfm/sq ft) on one filter
(32.1 m2) (1.8 m3/m2/d) = 57.8 m3/min (2,100 cfm) backwash air
• Calculate volume of backwash water (see Case 1):
Volume = (10.4 nvVmin) (5 min) + (7.8 m3/min)(15 min)
= 169 m3 (44,600 gal)
5. Calculate backwash frequency (see Case 1):
Backwash = (189 kg TSS/d) / [(7.34 kg TSS/m2) (160.5 m2)]
= 0.16 backwashes/d/filter
—or—
One backwash every 150 hr/filter or approximately one backwash/d.
6. Calculate frequency of nitrogen release cycle (NRC):
• Assume: 0.245 kg NOx-N/m2/NRC (0.05 Ib NOx-N/sq fl/NRC) (Reference 18)
(386 kg NOx-N/d) / [(0.245 kg NOx-N/m2/NRC) (160.5 m2)] = 9.8 NRC/d
(24 hr/d) / (9.8 NRC/d) = 2.5 hr/NRC
NRC frequency = 2.5 hr
7. Calculate methanol requirements (see Case 1):
• Assume: 3 kg methanol/kg of NOX-N removed
Methanol = (386 kg NOx-N/d) (3 kg methanol/kg NOX-N)
= 1,158 kg (2,550 Ib) methanol/d "
• Alternatively, use Equation 4-10 to compute methanol requirements.
8. Calculate biomass production (see Case 1):
COD removed = (1,158 kg methanol/d) (1.5 kg COD/kg methanol) = 1,737 kg/d
VSS = (0.18 kg VSS/kg COD) (1,737 kg COD/d)
= 313 kg (688 Ib) VSS produced/d
At 75 percent volatile TSS:
TSS produced = 313 / 0.75 = 417 kg (918 Ib) TSS/d
9. Calculate motor horsepower (see Case 1 for methodology).
230
-------
7.3.3 Case Studies
Two case studies of downflow packed-bed systems fol-
low. Operating data for other downflow packed-bed sys-
tems are provided in Table 7-7.
7.3.3.1 Hookers Point Wastewater Treatment Plant
(Tampa Florida)
The 4,208-L/s (96 mgd) Hookers Point Wastewater Treat-
ment Plant (WWTP) includes preliminary treatment, pri-
mary treatment, biological treatment, post aeration, and
effluent disinfection. The biological treatment system in-
cludes two-stage carbonaceous oxidation/nitrification .us-
ing high-purity oxygen and a separate-stage downflow
packed-bed denitrification system with methanol feed. A
flow schematic is shown in Figure 7-12.
The downflow packed-bed denitrification system consists
of 20 filters measuring 3 m x 32 m (10 ft x 105 ft). Each
filter is filled with 142 cm (56 in) of coarse sand (2.3 mm),
loaded at an average rate of 59-117 m3/m2/d (1-2 gpm/sq
ft) and having an empty-bed contact time of 45 min at
average flow.
The Hookers Point WWTP receives domestic wastewater,
with a 30-percent contribution from breweries (13). The
influent wastewater has a BOD5 of 224 mg/L, TSS of 221
mg/L, and TKN of 32 mg/L. The current effluent limits of
the plant are 5 mg/L for BOD5 and TSS, 3 rhg/L for TN
on an annual average basis, and 7.5 mg/L for total phos-
phorous (TP). A summary of monthly plant operating data
is provided in Table 7-8. The data show that an average
month's effluent is below 3 mg/L TN 83 percent of the
time, with an average over this three-year period of
2.33 mg/L. (Note that the effluent limit was changed to 3
mg/L TN in October 1990. Prior to that time, the limit was
4 mg/L TN in summer and 5 mg/L in winter.) The maxi-
mum month values appear to be attributable to decreased
performance of the upstream nitrification system. The av-
erage effluent TSS is 2 mg/L and is relatively stable.
Hookers Point has a process loading rate of 1.32 kg
NOx-N/m2/d (0.27 Ib/sq ft/d). The brewery waste may con-
tribute significantly to the background nitrogen removal
by synthesis. The plant's overall efficiency in removing
nitrogen and SS has been 93 percent and 99 percent,
respectively.
7.3.3.2 Dale Mabry Advanced Wastewater Treatment
Plant, Tampa, Florida
The flowstream of the 263-L/s (6-mgd) Dale Mabry Ad-
vanced WWTP consists of screening, grit removal, oxi-
dation ditch with anoxic zone, secondary clarification,
downflow packed-bed denitrification filters, post-aeration,
and disinfection. A schematic is provided in Figure 7-13.
There are two oxidation ditches (HRT = 19.2 hr) and five
secondary clarifiers. The solids are thickened on site and
hauled off site by a contract hauler. The denitrification
system consists of five filters, each 3 m (10 ft) wide by
19.8 m (65 ft) long, with 1.8 m (6 ft) of 2.3-mm sand. The
design average daily filter loading is 123 m3/m2/d (2.1
gpm/sq ft). The plant discharge standards are 5, 5, 3, 1
mg/L for BOD5, TSS, TN, and TP, respectively, on an
annual basis. The monthly limit for TN is 3.75 mg/L. The
wastewater is primarily domestic, and the NOX-N concen-
tration in the inlet to the denitrifying filter is 6-15 mg/L.
A summary of monthly plant operating data is provided
in Table 7-9. The data show that the Dale Mabry WWTP
can consistently meet its effluent limit, with an average
TN of 2.08 mg/L. Based on the available data, the plants
nitrate removal rate is 0.54 kg NOx-N/m2/d (0.11 Ib/sq ft/d).
Table 7-7. Operating Data for Selected Downflow Packed-Bed Systems
Average Rate
Facility
Hookers Point
Fiesta Village
Altamonte Springs
Faulkensand Road,
Hillsborough Co.
Dale Mabry
Port Orange
Florida
Location
Tampa
Ft. Myers
Altamonte
Tampa
Tampa
Port Orange
Capacity
Us
4,208
220
548
264
264
526
mgd
96.0
5.0
12.5
6.0
6.0
12.0.
m3/m2/d
123
117
123
29
123
123
gpm/
sqft
2.1
2.0
2.1
0.5
2.1
2.1
Number
of
Denit.
Filters
20
4
7
5
5
7
Media
Depth
Filter Size
3 m x 32 m
(10 ft x 105 ft)
3 m x 13.4 m
(10 ft x 44 ft)
3 m x 18.3 m
(10 ft x 60 ft)
3 m x 15.2 m
(10 ft x 50 ft)
3 m x 19.8 m
(10 ft x 65 ft)
3 m x 17.1 m
(10 ft x 56 ft)
m
1.47
1.83
1.83
1.22
1.83
1.07
in
54
72
72
48
72
42
Media
Size,
mm
2.3
3.0
Dual
Media
3.0
2.3
1.8
Note: All plants have a 3 mg/L TN permit limit.
231
-------
Methanol
• " • *
HPC
^=
)
=^
'
'4/%>
, ... .(
Second-Stage
Aerobic
(Nitrification)
RAS
Packed-Bed System
(Denitrification)
Backwash
Supply
\
WAS
WAS
Rgure 7-12. Schematic of Hookers Point Advanced Wastewater Treatment Plant.
Methanol
Preliminary
Treatment
Oxidation Ditch
BOD Removal &
Nitrification
Clarification
Backwash
Supply
Downflow Packed-
Bed Denitrification
Filters '
Disinfection
Figure 7-13. Schematic of Dale Mabry Wastewater Treatment Plant.
The nitrogen and SS removal efficiencies have been
94 percent and 99 percent, respectively, with effluent SS
from 0.67 to 2.2 mg/L. These data are an indication of
the favorable performance of the downflow packed-bed
systems in regard to both SS and nitrogen removal. Vari-
ations in performance were generally attributable to diffi-
culties with the methanol feed equipment; methanol feed
pumps and a flow meter which paced the methanol feed
were both replaced in 1991.
7.4 Upflow FIuidized-Bed Systems
7.4.1 Description
In a fluidized-bed reactor, nitrified secondary effluent
passes upward through a column at a flow rate sufficient
to produce a fluidized bed of media (typically sand). De-
nitrifying microorganisms attach to the sand, and the ni-
trified effluent passes through the media. Typically, small
factory-assembled reactors are constructed as a column
that is 2.4-4.3 m (8-14 ft) in diameter with varying height.
The more common larger columns are fabricated in the
field from steel or concrete. Column heights are variable.
Because the media are fluidized and the particles are not
in contact with other particles, an extremely large surface-
area-to-volume ratio is provided for supporting growth of
denitrifying microorganisms. The main advantage of the
fluidized-bed system is the small reactor size made pos-
sible by the increased efficiency associated with the de-
velopment of a high biomass concentration, and the
vertical configuration of the reactor. The specific surface
area available for biological growth is from 244-305
m2/m3 (800-1,000 sq ft/cu ft) of reactor volume, as com-
pared to approximately 91 m2/m3 (300 sq ft/cu ft) in a
downflow packed-bed system. A typical flow schematic
for a small, proprietary reactor is shown in Figure 7-14.
Fluidized-bed reactors are discussed further in Chapter 6.
To date, fluidized-bed reactors have been installed at a
limited number of facilities, where they have been imple-
mented in aerobic, anoxic, and anaerobic modes. The
systems have been used for a variety of applications such
as treating chemical wastes, contaminated groundwater,
fish hatchery wastes, municipal wastes, and high-strength
industrial wastewater. Currently, the technology has been
232
-------
Table 7-8. City of Tampa, Hookers Point Wastewater Treatment Plant: Monthly Performance Data
Plant Influent
Flow
Month
JAN 89
FEB 89
MAR 89
APR 89
MAY 89
JUN 89
JUL 89
AUG 89
SEP 89
OCT 89
NOV 89
DEC 89
JAN 90
FEB 90
MAR 90
APR 90
MAY 90
JUN 90
JUL 90
AUG 90
SEP 90
OCT 90
NOV 90
DEC 90
JAN 91
FEB 91
MAR 91
APR 91
MAY 91
JUN 91
JUL 91
AUG 91
SEP 91
OCT 91
AVERAGE
Us
2,362
2,266
2,209
2,349
2,090
2,226
2,831
3,936
3,024
2,687
2,336
2,507
2,371
2,481
2,323
2,612
2,108
2,261
2,709
2,577
2,319
2,297
2,980
1,937
2,314
2,104
2,814
2,038
2,038
2,134
868
2,319
2,481
2,862
2,253
mgd
53.9
51.7
50.4
53.6
47.7
50.8
64.6
89.8
69.0
61.3
53.3
57.2
54.1
56.6
53.0
60.0
48.1
51.6
61.8
58.9
52.9
52.5
68.0
44.2
52.8
48.0
64.2
46.5
46.5
48.7
19.8
52.9
56.6
65.3
51.4
BOD
mg/L
383
299
275
275
242
255
214
216
182
215
200
172
176
276
313
280
241
268
231
222
256
267
222
281
247
227
281
258
241
281
276
258
254
198
253
TSS
mg/L
220
238
214
225
185
205
206
168
162
176
210
196
151
191
275
236
273
278
167
171
191
183
201
197
184
281
293
195
261
237
210
217
225
172
218
TN
mg/L
32.3
32.3
34.5
34.7
33.8
32.9
28.9
27
24.8
30
33.7
30.4
25.1
28.9
35.5
36.2
37.2
34.1
27.8
29.7
38.6
32.3
36.1
32.1
34.9
35.5
36.5
31.6
31.9
34.7
32.1
32.3
27.1
34
Denitrifi-
cation
Process
Influent
NOX-N
mg/L
16.8
17.5
18.8
19.7
16.4
17.8
15.0
15.1
12.4
18.2
17.4
15.8
16.0
16.0
17.0
18.0
17.0
16.8
16.8
8.5
11.9
13.9
14.4
13.8
14.4
13.4
15.5
15.8
15.2
18.6
13.2
13.3
14.1
14.42
Final Plant Effluent
BOD
mg/L
2.0
3.0
2.0
2.0
2.0
2.0
2.0
3.0
5.0
3.0
2.0
3.0
2.0
2.9
2.8
2.8
3.1
2.0
1.3
2.2
2.0
2.9
3.8
3.8
2.0
3.0
3.0
2.0
2.8
3.0
2.0
2.4
2.0
2.3
2.64
TSS
mg/L
0.9
1.9
1.6
0.9
1.0
1.2
1.5
1.7
2.5
1.4
2.0
1.7
2.3
2.6
1.6
1.2
1.5
1.5
2.8
2.1
1.8
1.8
2.9
0.9
1.9
2.0
1.8
1.8
3.6
2.3
2.0
1.9
1.5
1.7
2.0
TKN
mg/L
1.47
1.81
1.75
1.48
1.43
1.37
1.82
1.43
2.02
1.61
1.62
1.66
1.62
2.75
1.74
1.62
1.59
1.41
0.08
2.45
2.76
1.40
1.58
1.65
1.49
1.58
1.89
1.55
1.57
1.65
1.66
1.58
1.57
1.58
1.68
NHJ-N
mg/L
0.00
0.03
0.22
0.21
0.11
0.13
0.40
0.20
0.45
0.22
0.19
0.06
0.16
1.06
0.13
0.12
0.14
0.16
0.30
0.94
1.38
0.02
0.03
0.48
0.02
0.03
0.04
0.03
2.08
0.11
0.28
0.16
0.16
0.36
0.34
NOX-N
mg/L
0.13
0.91
1.72
2.27
0.94
1.21
1.26
1.33
1.13-
1.47
0.94
0.69
0.19
2.05
0.55
0.60
1.37
1.33
0.31
1.61
1.36
1.46
0.41
0.48
0.44
1.41
0.50
1.58
0.64
1.16
0.72
1.92
1.02
1.00
1.01
TN
mg/L
2.80
2.72
3.47
3.75
2.37
2.58
3.07
2.76
3.15
3.08
2.56
2.34
2.41
2.70
2.36
1.22
2.85
2.83
2.22
3.06
3.32
1.86
1.99
2.12
.1.86
1.39
2.62
2.13
2.89
2.09
2.37
2.50
2.88
2.54
2.33
233
-------
Table 7-9. Dale Mabry Wastewater Treatment Plant: Monthly Performance Data
Plant Influent
Month
JAN 90
FEB 90
MAR 90
APR 90
MAY 90
JUN 90
JUL 90
AUQ 90
SEP 90
OCT 90
NOV 90
DEC 90
JAN 91
FEB 91
MAR 91
APR 91
MAY 91
JUN 91
JUL 91
AUQ 91
AVERAGE
Flow BOD
L/s mgd mg/L
240 5.48 214
238 5.42 199
234 5.33 193
231 5.27 204
230 5.24 195
238 5.42 182
241 5.50 182
239 5.46 183
236 5.39 179
234 5.33 185
232 5.29 210
235 5.37 218
222 5.06 188
213 4.87 188
155 3.54 220
148 3.39 203
148 3.39 209
147 3.36 213
156 3.55 202
165 3.77 191
191 4.36 201
Effluent
Recycle
Flow
1 ' f — 1
Nitrified \ I
Effluent _,,_,. ,
Fluidlzation
Pump
TSS TN
mg/L mg/L
192 34.7
150 33.9
185 34.2
189 35.2
169 34.5
142 32.2
152
32.2
156 32.3
166 32.1
154 33.1
182
35.7
180 34.6
165
146
188
133
182
177
178
181
169
34.4
33.7
34.7
33.2
33.2
32.9
32.9
31.7
33.5
^
V-
»k *•
r
Process
Influent Final Plant Effluent
TSS NOX-N BOD TSS NHJ-N NOX-N TN
mg/L mg/L mg/L mg/L mg/L mg/L mg/L
9 5.75 4.13 2.00 0.99 1.18 3.30
13 11.73 3.70 1.25 0.12 1.95 2.82
9 10.32 4.52 1.14 0.18 1.75 2.67
10 9.11 2.02 2.23 0.15 1.79 2.34
8 6.70 2.22 2.61 0.23 0.97 1.99
5 6.82 2.85 2.20 0.47 1.30 2.58
9 11.30 2.20 1.74 0.11 1.57 2.46
3 11.20 2.11 1.68 0.10 1.30 2.03
6 10.10 2.26 1.43 0.11 0.61 1.12
2 11.03 1.78 1.00 0.12 0.97 1.61
3' 10.95 1.75 0.81 0.12 1.34 2.12
5 9.07 1.91 1.77 0.23 1.04 1^94
3 8.25 2.02 0.67 0.16 1.38 2.73
4 8.52 2.05 1.04 1.16 1.02 2.88
4 12.25 1.36 1.02 0.22 1.43 2.27
4 13.77 1.34 1.11 0.11 0.96 1.70
6 12.43 1.67 1.68 0.11 2.31 3.14
5 12.06 2.21 0.81 0.10 0.67 1.36
4 12.51 1.93 0.87 0.11 1.53 2.17
4 15.18. 1.89 1.02 0.11 1.19 1.87
4.2 11.34 1.85 1.10 0.22 1.20 2.08
.s s^^ Growth Control Device
s^ /\. Waste Biomass to
i,..-| >",• Solids Handling
^ ' ' fc/^ f'
Growth Control
Pump
Fluid Bed
Reactor
t t t V«. .
AAAAAAAAA
Sheared Media &
Return Biomass
Methanol Feed
-HI
\ 1
Methanol
Feed Pump
Methanol
Storage Tank
Figure 7-14. Schematic of upflow fluidized-bed system.
234
-------
more commonly used for industrial wastewater rather
than municipal wastewater. Concerns over municipal ap-
plications have included mechanical scale-up factors,
proprietary constraints, and economically unattractive
system appurtenances (24). However, as discussed be-
low, there are successful municipal applications; Table
7-10 lists several industrial and municipal installations
with fluidized-bed reactors. Some of these plants are not
operating in the denitrification mode and others were de-
signed solely for nitrification. Table 7-11 lists the loading
at four of these plants, operating in the denitrification
mode.
The principal commercial suppliers of fluidized-bed sys-
tems are Dorr-Oliver, Envirex, and Ecolotrol. Both Dorr-
Oliver and Envirex systems were developed on the basis
of Ecolotrol process patents. Currently, Envirex is the only
manufacturer actively marketing the fluidized-bed reactor
for denitrification applications in the United States. Ta-
ble 7-12 summarizes the types of reactors in use.
The principle of the fluidized-bed reactor is the same,
regardless of the application. The uniformly round media
become fluidized when the nitrified effluent passes up-
ward through the bed. A thin active biomass forms on all
sides of the media and causes the media to become less
Table 7-10. Full-Scale Applications of Upflow Fluidized-Bed Technology for Nitrogen Control3
Facility and Location Application
Reactor Design Information
(Approximate Dimensions)
Pensacola Wastewater Treatment
Plant"
Pensacola, FL
Reno-Sparks Wastewater Treatment
Plant"
Reno, NV
Rancho California Wastewater
Reclamation Plantb
Riverside County, CA
IBM Plant"
Somers, NY
Anglian Water Authority
Ipswich, England
Aquafuture Fish Hatchery Plant
Turner Falls, MA
Department of Energy Facility
Fernald, Ohio
Dworshak National Fish Hatchery
Ahsahka, ID
General Motors Plant
Sandusky, OH
General Motors Plant
Ypsilanti, Ml
Hazardous Waste Treatment
Storage and Disposal Facility
Hatfield, PA
Sherwood Medical Plant
Deland, FL
Denitrification of nitrified sanitary
wastewater (not operating)
Denitrification of nitrified sanitary
wastewater
Denitrification of nitrified sanitary
wastewater (partially fluidized bed)
Denitrification of nitrified sanitary
wastewater
Denitrification of river water to
potable quality
Ammonia removal from fish hatchery
water
Denitrification of nuclear fuel
processing wastewater
Ammonia removal from fish hatchery
water
Aerobic carbonaceous oxidation of
sanitary wastewater and aerobic
carbonaceous oxidation and
nitrification of wastewater from
automotive manufacturing operations
Aerobic carbonaceous oxidation and
nitrification of wastewater from
automotive manufacturing operations
Denitrification of industrial wastewater
Denitrification of wastewater from
cleaning operations during production
of medical products
4 reactors, each 33.4 m2 by 5.8 m
high (360 sq ft by 19ft)
4 reactors, each 74.3 m2 by 7.9 m
high (800 sq ft by 26 ft)
3 reactors, each 22.7 m2 by 4.6 m
high (244 sq ft by 15ft)
1 reactor, 2.8 m2 by 4.3 m high
(30 sq ft by 14ft)
1 reactor, 5.2 m2 by 7.0 m high
(56 sq ft by 23 ft)
2 reactors, each 4.7 m2 by 2.4 m
high (51 sq ft by 8 ft)
4 reactors, each 1.2 m2 by 11.0 m
high (13 sq ft by 36 ft)
7 reactors, each 14.3 m2 (154 sq ft)
by 7.3 m (24 ft) high (2 reactors)
and 8.5 m (28 ft) high (5 reactors)
3 reactors: two 8.8 m2 by 7.0 m high
(95 sq ft by 23 ft), and one 5.9 m2
by 7.0 m high (64 sq ft by 23 ft)
4 reactors, each 65.5 m2 by 8.2 m
high (705 sq ft by 27 ft)
2 reactors, each 10.5 m2 by 7.6 m
high (113sqftby25ft)
1 reactor, 3.2 m2 by 5.5 m high
(34 sq ft by 18ft)
a Partial listing; includes only media-based fluidized
b Additional process design information is shown in
•bed systems. Some plants are no longer operating and others are used for nitrification (25).
Table 7-11.
235
-------
Table 7-11. Process Design Information for Upflow Fluidized-Bed Systems (from Reference 24)
Facility
Parameter*
PensacolabpC
Reno-Sparks
Rancho, CA°
* Sea Table 7-10 for additional Information.
6 Modified design as developed by Dorr-Oliver, Inc.
9 No longer operated for denitriflcation.
11 Equalization provided to achieve a constant wastewater flow rate.
• Based on mean wastewater flow and fluldized-bed/empty-bed volume.
1 Based on total flow to the reactor (plant flow plus recycle).
IBM0
Mean wastewater flow, Us
Mean wastewater flow, mgd
Maximum wastewater flow, Us
Maximum wastewater flow, mgd
Influent NOJ-N, mg/L
Effluent NOJ-N, mg/L
Design wastewater temperature, °C
Estimated reactor blomass, mg/L VSS
Hydraulic retention time,8 min
Hydraulic loading rate,' m3/m2/d
Hydraulic loading rate,' gpd/sq ft
Estimated settled sand depth, m
Estimated settled sand depth, ft
Fluidized-bed height, m
Fluidized-bed height, ft
1,052
24
1,490
34
20
<6
18
NA
8.5
672
11.4
1.8
6
4
13
1,883
43
2,400
55
18
2
13
18,000
13.8
550
9.3
2.4
8
4.9
16
263
6d
21
2.5
22
28,000
10
336
0.8
1.2
4
2.4
8
113
1d
54
8
10
NA
26
578
1.3
1.5
5
2.7
9
Table 7-12. Types of Fluidized-Bed Denitrification
Systems (from Reference 26)
Oxitron System
— Devebped by Dorr-Oliver
— System based on Ecolotrol process patents
— Uncertain regarding system marketing in North America
— Dorr-Oliver Europe marketing systems in Europe
Rex aerobic fluidized-bed process, anaerobic and biological
denitrification configuration
— Developed by Envirex/Ecolotrol based on Ecolotrol
process patents
- Sold in North America by Envirex
Custom engineered systems
- Developed by consulting engineering firms
— Normally designed and operated under conditions falling
outside the limits of Ecolotrol patents
dense than clean media. The lower density particles (i.e.,
the particles with the thickest biomass film) move toward
the top. The open nature of the fluidized bed minimizes
the chance for bed plugging.
7.4.2 Design Considerations
The upflow fluidized-bed system usually consists of a
reactor vessel in the form of an above-ground steel and
fiberglass tower or in-ground concrete reactors. The flow
rate and strength of waste determines the size of the
reactor vessel. The reactor size is dependent on tempera-
ture; at 15°C (59°F), the design loading rate is 6,420 kg
NO5-N/1.000 m3/d (400 lb/1,000 cu fl/d) (27). Other load-
ing rates are shown in Table 7-13.
When the fluidized bed system is operated for denitrifica-
tion, methanol is fed to the nitrified influent by injection
into the recycle line (see Figure 7-14). The reactor oper-
ates as a plug flow process; however, the high recycle
ratio of reactor effluent to plant flow (10:1 to 20:1 for high
strength waste treatment and 2:1 to 5:1 for municipal
denitrification) emulates a complete mix system. The high
recycle ratio also helps protect the reactor from shock
loads and is required to achieve bed fluidizatiori. The
amount of recycle is dictated by a maximum allowable
fluid-bed height; structural considerations often control
bed height. The amount of settled media depth is a func-
236
-------
Table 7-13. Selected Upflow Fluidized-Bed Loading Rates
Facility and Location
Mineola, NYb
Reno-Sparks, NVb
Rancho, CAb
Envirex0
Temp., °C
20
13
25
15
Reactor
Height, m
4.7
6.2
1.2
2.2
kg NOX-N
removed/1,000 kg NOX-N
m2 media surface removed/1,000 ,
areaa/d m3 reactor/d Reference
2.0
0.78
1.17
3.4
5,380
2,010
3,000
6,420
28.
29
Average
1.8.4
4,200
Facility and Location
Mineola, NYb
Reno-Sparks, NVb
Rancho, CAb
Envirex0
Temp., °C
20
13
25
15
Reactor
Height, ft
15.5
20,5
4
7.5
lb NOX-N
removed/1 ,000
sq ft media
surface areaa/d
0.41
0.16
0.24
0.7
lbNQx-N
removed/1,000
cu ft reactor/d
335
125 -
187
400
S '
Reference
28
29
Average
0.38
262
a Based on 2,625 m2/m3 (800 sq ft of surface area/cu ft) of reactor.
b Based on actual plant data; design criteria may be different.
0 Based on publishecLEnvirex literature; see design example in this chapter.
tion of the type and strength of waste being treated and
is variable between 30 and 60 percent of bed volume,
with a typical value of 50 percent.
Design and operating parameters that are important for
maintaining high biomass concentrations in the reactor
include'(29): 1) type, size, and volume of media, 2) media
density, 3) reactor cross sectional area, 4) biofilm thick-
ness, and 5) expanded bed height, including expansion
due to hydraulic flow and biofilm growth. The first three
items can be established during design; however, items 4
and 5 become important operating parameters that affect
fluidization and bipmass concentrations.
The typical media sand is hard and round, with a size
range of 0.3-0.6 mm. Its uniformity coefficient is 1.25-
1.50. To ensure proper fluidization at the given flow rates,
the'particle density is important. Sands with a specific
gravity.of 2.6 have,been used.
Two interrelated factors are the control of biofilm thick-
ness, and expanded bed height. An important design con-
sideration is proper bed height selection and the
media-biomass growth control system. The fluidized-bed
system;results in biofilm. growth; the effective diameter of
the particles increases and the density is reduced. As a
result, fhe bed expands beyond the expansion of clean
media); The lower density, larger particles (i.e., particles
with, the thickest biomass film) move toward the top. Con-
trolljng^the thickness of the film is necessary to prevent
media- from being transported out of the reactor. A me-
chanjeal film thickness or growth-control device is used
at.the (top of the reactor to, in effect, control bed expan-
sion. When the bed height reaches the bottom of the
growth-control device, the growth-control pump is de-
signed to start automatically. The growth-control system
must minimize media carryover from the reactor to the
effluent stream, minimize media loss and abrasion or
contribution to SS in the effluent or recycling streams, and
ensure that the returned media are clean.
The growth control system removes the upper layers of
the media from the top of the reactor. The slush, or
sheared biomass, and media slurry are then passed
through a separation device, and the media are returned
to the reactor. The sheared media have a thinner layer
of biomass and are returned to the reactor to retain the
proper amount of microorganisms. The sheared biomass
is wasted to the solids-handling system. The clear zone
above the expanded media is usually the height of the
growth-control mechanism plus a freeboard of 2.4-3 m
(8-10 ft). The clean media returned to the reactor have
an effective biofilm thickness of less than 20 rnm (30).
The biomass wasted from the Envirex growth-control sys-
tem is typically 0.5-1.5 percent solids. All known full-
sized systems use a pump to shear the biomass from the
media. The downstream mechanisms that separate the
media-biomass slurry include (26): ,
• External wedge bars '
• Screens with a spray wash
• Hydraulic separators internal to the reactor
• External cyclones
• Patented separation system (Ecolotrol/Envirex)
237
-------
• Combination of the above
The most significant problems concerning the above sys-
tems are bed media loss and the release of biomass into
the effluent (24). In other words, the systems may not
adequately separate the sheared biomass and the media.
As a result, either the media remain in the waste biomass
and are lost, or the sheared biomass being returned with
the media is discharged with the effluent. The ideal media-
separation system must ensure there is no media carry-
over from the reactor into the effluent and recycle stream.
The system must not cause loss of media or contribute
biomass SS to the effluent and recycle stream.
Excess biomass is also removed from the media by con-
tinuous sloughing off through death and abrasion.
Sloughed off biomass can be a problem, but in a properly
designed system, the effluent SS will generally be 15-20
mg/L for municipal wastewater applications.
The influent distribution manifold is another critical design
feature of the fluidized-bed system. The manifold must:
• Achieve uniform distribution of flow across the entire
reactor area to maximize reaction kinetics.
• Prevent plugging and media escape.
• Minimize abrasive wear.
• Maintain uniform flow throughout the media depth and
minimize the movement of biomass through the reactor.
• Minimize shearing of biomass above the influent dis-
tribution manifold, thereby promoting uniform biofilm
buildup throughout the media.
Most problems with distribution manifolds can be attrib-
uted to plugging. However, plugging can be prevented by
removing solids from the influent stream and with a hy-
draulic design that prevents media backflow into the
distribution manifold (24). The most common influent dis-
tribution systems consist of a header manifold pipe with
lateral pipes branching out. Nozzles attached to the lat-
erals point downward at or near the reactor bottom.
A key advantage of the fluidized-bed system is its mobility
and small footprint, which is conducive to factory assem-
bly of small reactors. Therefore, some reactors are as-
sembled in the factory and have no moving parts; the
methanol feed system, growth-control pump, and influent
pump are the only required mechanical components.
However, the majority of installed systems are in-ground
concrete units or above-ground steel units that are fabri-
cated on site. The main disadvantages of the fluidized-
bed system are the limitations on reactor size, energy
requirements (dominated by high recycle rates), difficul-
ties in biomass control and media selection (media loss
and biomass in the effluent), and the imprecision in proc-
ess control because of difficulty in monitoring biomass
concentration.
The design of fluidized-bed reactors is based on a com-
bination of hydraulic and biological kinetic models. A num-
ber of models have been suggested (31-33). A half-order
kinetic model proposed by Stephenson and Murphy (33),
combined with an Arrhenius temperature relationship (1),
was reported to adequately describe denitrification of mu-
nicipal wastewater over a range of temperatures. This,
results in a curve similar to that shown in Figure 7-15.
7.4.3 Design Examples
The design approaches presented here for upflow
fluidized-bed systems are based on the discussions in
Chapter 4 about the kinetics of denitrification and the
fundamental equations used in system design. As with
preceding system design examples in this chapter, these
draw from the theory discussed in Chapter 4 and the
influent criteria discussed in Chapter 2. The application
rates used are for illustration purposes only and are
based on recommendations from Envirex. The use of
more conservative rates may be appropriate, depending
on the system application. Onsite pilot testing is recom-
mended to generate design criteria (see Chapter 6). The
examples use two different effluent requirements. The
calculations for Case 1 are presented in more detail than
those for Case 2; thus Case 1 should be referred to for
a complete methodology. Table 7-14 lists the nitrified ef-
fluent characteristics (influent to the denitrification reac-
tor), the final effluent limits, and the values used for
design. A process flow schematic for an upflow fluidized-',
bed system is provided in Figure 7-16.
Table 7-14. Upflow Fluidized-Bed System Design
Examples—Influent Characteristics and Effluent Limits
Characteristic
Minimum Monthly Temp.
Average Flow, m3/d
Average Flow, mgd
Peak Week Flow, m3/d
Peak Week Flow, mgd
TSS, mg/L
CBOD, mg/L
COD, mg/L
TKN, mg/L
(NO;+NOi)-N, mg/L
NHJ-N, mg/L
TN, mg/L
Nitrified
Effluent
—
—
15
3
33
1.8
23.4
.05
26.5
Effluent
Limits
15°C
18,930
5
28,396
7.5
30
30
—
' —
7
2
10
Effluent
Limits
15°C
18,930
." 5
28,396
7.5
, 10
10
31
—
3
1
5
238
-------
Design Example: Case 1-Upflow FIuidized-Bed Denitrification System
1. Calculate nitrate removed: .••,.--
(23.4 - 7 mg/L) (18,930 m3/d) / 1,000 = 311 kg (684 Ib) NOx-N/d
2. Calculate reactor volume: ._..'..."
• Assume:
Loading rate = 6.42 kg NOx-N/m3/d (400 Ib NCyN/1,000 cu ft/d) [See Figure 7-15.]
• Calculate volume of reactor:
Volume = (311 kg NOx-N/d) / (6.42 kg NOx-N/m3/d) -'.,-;
= 48.4 m3 (1,710 cu ft)
• Assume: 3.65-m (12-ft) diameter reactor
Area = 7t (3.65)2/4 = 10.5 m2 (113 sq ft)/reactor
• Assume: Two reactors in service and one on standby
• Calculate bed height:
(48.4 m3) / [(10.5 m2/reactor) (2 reactors)] = 2.3 m (7.6 ft)
• Use 3 m (10 ft) high bed with 1.8 m (6-ft) of freeboard for solids separation for a 4.9 m (16 ft) high
reactor, based on manufacturer's standard.
Total volume of in-service reactors = 2(10.5 m2) (3 m)
= 63 m3
= 2,225 cu ft •>-.-.•
= 16,640 gal :
3. Calculate HRT at average flow:
(63 m3)(1,440 min/d)/18,930 m3/d = 4.8 min
Calculate HRT at peak flow:
(63 m3)(1,440 min/d)/28,396 m3/d = 3.2 min
4. Check flux rate:
Flux rate = flow/area
(18,930 m3/d) / (10.5 m2) / (2) = 901 m3/m2/d (15.3 gpm/sq ft) at average flow ;
(28,396 m3/d)/(10.5 m2) / (2) = 1,352 m3/m2/d (23.0 gpm/sq ft) at peak week >f low' . ,
5. Calculate actual nitrogen loading based on selected reactor:
(311 kg NOx-N/d) / (63 m3) = 4.94 kg NOx-N/m3/d (308 Ib NCyN/1,000 cu ft/d),
6. Calculate recycle rate:
• Maintain reactor flow rate equal to peak flow rate of 19.72 m3/min (5,208 gpm) since the peak flow rate
provides adequate fluidization for the media in this example. Flux rate should be between 880-1,470
m3/m2/d (15-25 gpm/sq ft). . ..y: -
• Recycle required at average flow is equal to 19.72-13.15 = 6.57 m3/min (1,736 gpm).
7. Calculate methanol required: .-,„.' •
• Calculated nitrate removed = 311 kg/d (684 Ib/d) ' '"".".'•':.
• Assume: 3 kg methanol/kg of NOX-N removed • -. ,
Methanol = (311 kg NOx-N/d) (3 kg methanol/kg NOX-N) , =. ; , • ... . , :
= 933 kg (2,055 Ib) methanol/d , •- : , -,,-...•-... , ;
239
-------
Design Example: Case 1 (continued)
(933 kg/d) (1 L/0.79 kg) = 1,181 L (312 gal)/d
Methano! dose = 49.2 L (13 gal)/hr
• Alternatively, methanol requirements can be computed from Equation 4-10. If the influent DO to deni-
trification is assumed to be 3 mg/L, the methanol requirement (neglecting residual methanol) is:
(311 kg)(2.47) + (0.87)(3)(18,930 m3/d) / 1,000 = 818 kg (1,800 Ib)
8. Calculate biomass produced:
• From the suspended growth design example (Section 7.2.3.1), use 0.18 kg VSS/kg COD removed.
Assume that all COD removed is methanol:
COD removed = (933 kg methanol/d) (1.5 kg COD/kg methanol) = 1,400 kg/d
VSS = (0.18 kg VSS/kg COD) (1,400 kg COD/d)
= 252 kg (485 Ib) VSS produced/d
At 75 percent volatile:
TSS produced = 252 / 0.75 = 336 kg (741 Ib) TSS/d = 1.08 kg TSS/kg NOg-N
« Typical biomass production is 0.4-0.8 kg TSS/kg NOX-N removed
• Calculate excess biomass flow rate (assume 1 percent solids):
Flow = (336 kg TSS/d) (1/0.01) (1 L7 kg) (1/1,440 min/d)
= 23.3 L/min (6.2 gpm)
9. Calculate horsepower of pumps:
• Fluidization pump:
Total pump capacity is 19,720 L/min (5,208 gpm) (peak week flow)
With 2 pumps: 19,720/2 = 9,860 L7min (2,605 gpm)
• A typical fluidized-bed configuration requires a fluidization pump with approximately 12.2 m (40 ft)
TDH. This should be verified for the actual reactor configuration.
hp = [(Q)(TDH)(Sp. Gravity)] / [(3,960)(Pump Eff.)(Motor Eff.)]
Where:
TDH = 40 ft (assumed)
Pump EFF s 75% (assumed)
hp m [(2,605) (40) (1)] / [(3,960) (0.75) (0.90)]
= 39.0 hp
Kw = (hp) (0.746) = 29.0 Kw
• Calculate horsepower of growth control pump:
Base flow rate on biomass flow rate of 3 gpm for each pump.
hp = [(3) (40) (1)] / [(3,960) (0.75) (0.90)]
= 0.04 hp
Kw B (hp) (0.746) = 0.03 Kw
240
-------
Design Example: Case 2—Upflow Fluidized-Bed Denitrification System
1. Calculate nitrate removed:
(23.4 - 3 mg/L) (18,930 m3/d) + 1,000 = 386 kg (850 Ib) NOx-N/d
2. Calculate reactor volume:
• Assume:
Loading rate = 6.42 kg NOx-N/m3/d (400 Ib NCyN/1,000 cu ft/d) [See Figure 7-15.]
• Calculate volume of reactor:
Volume = (386 kg NOx-N/d) / (6.42 kg NOx-N/m3/d)
= 60.1 m3 (2,120 cu ft)
• Assume: 3.65-m (12-ft) diameter reactor
Area = n (3.65)2/4 = 10.5 m2 (113 sq ft)/reactor
• Assume: Three reactors in service and one on standby
• Calculate bed height:
(60.1 m3) / [(10.5 m2/reactor) (3 reactors)] = 1.9 m (6.2 ft)
• Use 3 m (10 ft) high bed with 1.8 m (6 ft) of freeboard for solids separation for a 4.9 m (16 ft) high
reactor, based on manufacturer's standard.
Total volume of in-service reactors = 3(10.5 m2)(3 m)
= 94.5 m3
= 3,340 cu ft
= 24,970 gal
3. Calculate HRT at average flow:
(94.5 m3)(1,440 min/d)/18,930 m3/d = 7.2 min
Calculate HRT at peak flow:
(94.5 m3)(1,440 min/d)/28,396 m3/d = 4.8 min
4. Check flux rate at average flow:
Flux rate = flow/area
18,930 m3/d / [(10.5 m2) (3)] = 601 m3/m2/d (10.2 gpm/sq ft) at average flow
28,396 m3/d / [(10.5 m2) (3)] = 901 m3/m2/d (15.3 gpm/sq ft) at peak flow
5. Calculate actual nitrogen loading based on selected reactor:
(386 kg NOx-N/d) / (94.5 m3) = 4.1 kg NCyN/m3/d (250 Ib NOX-N/1,000 cu ft/d)
6. Check recycle rate (see Case 1).
7. Calculate methanol required (see Case 1).
8. Calculate biomass production (see Case 1).
9. Calculate horsepower of pumps (see Case 1).
The process flow schematic is similar to Figure 7-16, except that for this example four reactors are used.
241
-------
2,400
2,000
1,600
1,200
800
400
10 15 20 25
Temperature, °C
30
Figure 7-15. Temperature vs. loading rate for upflow
fluidized-bed system (from Reference 1).
ment, phosphorus and BOD removal in a sidestream
phosphorous-removal system, nitrification biotowers, de-
nitrification upflow fluidized-bed reactors, post aeration,
effluent filtration, and disinfection. The solids-handling
system consists of thickening, anaerobic digestion, and
dewatering.
The denitrification system consists of four upflow fluidized-
bed towers measuring approximately 8.2 m (27 ft) in di-
ameter by 6.2 m (20.5 ft) high. The hydraulic residence
time at average daily flow is 13.8 min, and the solids
residence time (60) is 8.5 d. The denitrification system,
manufactured by Envirex, was designed to produce efflu-
ent with a nitrate level of 2 mg/L. A summary of monthly
plant operating data is provided in Table 7-15. The data
indicate that the Reno-Sparks plant has consistently met
its effluent requirements, with an average effluent ammo-
nia level of 0.16 mg/L and a NOX-N level of 0.29 mg/L.
The plant's efficiency in removing total nitrogen has been
94 percent. The removal rate of the fluidized-bed reactors
has been 6.4 kg NOx-N/m2/d (1.3 Ib/sq ft/d), and the plant
has regularly produced an effluent TN of less than 3 mg/L
and an average effluent TN of 1.78 mg/L. The one event
over 3 mg/L TN was 3.55 mg/L. Note that the design
loadings are lower than those used in the design
example.
7.4.4 Case Studies
7.4.4.1 Reno-Sparks Wastewater Treatment Facility,
Cities of Reno and Sparks, Nevada
A flow schematic for the 1,753-L/s (40-mgd) Reno-Sparks
Wastewater Treatment Facility is shown in Figure 7-17.
The plant consists of preliminary treatment, primary treat-
7.5 Methanol Handling, Storage, Feed
Control, and Excess Methanol Removal
Methanol is a chemical not normally dealt with in the
operation of wastewater treatment plants. Care must be
exercised in the design and operation of methanol han-
dling, storage, and feeding facilities to ensure the safe
and proper use of the methanol.
Nitrified
Influent from
Nitrification
System
the
Recycle
[O*
i
S ^
r
\
'
.. — ^
Fluidization
Pump
r
ffcr
^ — -O
Fluidized-Bed
Reactor Tower
Growth Control
Pump
Denitrification
System Effluent
Waste Sludge to
Disposal
Figure 7-16. Design example schematic of upflow fluidized-bed system.
242
-------
Equalization
Methanol
Plant
Influent
Preliminary
Treatment
Clarification
(Continued
Below)
hoi
Aeration Clarification
(BOD Removal)
Nitrification
Upflow Fluidized-
Bed Denitrification
Filtration
Disinfection
Post-Aeration
Figure 7-17. Schematic of Reno-Sparks Wastewater Treatment Plant.
Table 7-15. Reno-Sparks Wastewater Treatment Plant: Monthly Performance Data
Plant Influent
Flow
Month
JAN 90
FEB 90
MAR 90
APR 90
MAY 90
JUN 90
JUL 90
AUG 90
SEP 90
OCT 90
NOV 90
DEC 90
JAN 91
FEB 91
MAR 91
APR 91
MAY 91
JUN 91
JUL 91
AVERAGE
Us
1,133
1,153
1,174
1,156
1,169
1,221
1,202
1,233
1,211
1,175
1,144
1,192
1,168
1,122
1,149
1,103
1,137
1,171
1,210
1,169
mgd
25.86
26.30
26.79
26.37
26.66
27.87
27.42
28.14
27.64
26.81
26.10
27.20
26.64
25.60
26.21
25.17
25.95
26.72
27.36
26.67
BODS NHJ-N
mg/L mg/L
167
171
178
177
183
175
152
164
168
149 .
162
150
132 21.1
130 22.2
135 22.5
142 22.7
186 22.0
167 21.0
160 20.1
160 21.7
Denitrification Denitrification
Process Influent Process Effluent
TN NHJ-N
mg/L mg/L
0.37
0.53
0.31
1.37
0.67
0.35
0.30
0.91
0.56
0.15
0.37
1.13
32.3 0.45
32.3 0.46
32.2 0.35
32.2 0.16
31.9 0.59
31.9 0.08
31.6 0.06
32.1 0.48
NOX-N
mg/L
14.44
14.58
14.80
14.21
12.84
14.02
14.07
13.70
14.18
11.03
13.17
13.92
14.04
14.26
14.88
15.56
15.07
13.72
13.85
14.02
TSS NOX-N
mg/L mg/L
2.12
3.91
0.80
0.82
0.19
0.06
0.04
0.29
0.12
0.03
0.10
0.22
22 0.24
20 0.16
18 0.39
20 0.31
21 0.17
17 0.03
13 0.08
19 0.53
BOD
mg/L
5
8
5
6
5
3
3
6
7
3
4
4
3
3
3
4
4
2
3
4
Final
TSS
mg/L
5
5
4
2
3
2
2
10
18
2
3
6
4
3
4
4
4
4
4
5
Plant Effluent
NHJ-N
mg/L
0.13
0.24
0.27
0.29
0.08
0.11
0.10
0.32
0.41
0.05
0.19
0.33
0.17
0.18
0.10
0.02
0.03
0.02
0.03
0.16 .
NOX-N
mg/L
0.79
0.43
0.88
0.90
0.15
0.16
0.06
0.14
0.11
0.01
0.07
0.37
0.35
0.28
0.42
0.28
0.14
0.03
0.03
0.29
TN
mg/L
2.53
2.25
2.79
3.55
1.38
1.43
1.11
1.82
2.50
1.01
1.26
2.00
1.64
1.58
1.70
1.53
1.47
1.18
1.01
1.78
243
-------
7.5.1 Characteristics and Properties of Methanol
Methanol (CHaOH), which is known by a variety of
names—such as methyl alcohol, methyl hydrate carbinol,
and wood alcohol—is a colorless liquid and is noncor-
rosive except to aluminum, lead, and some natural rub-
bers at normal atmospheric temperature. It is normally
supplied pure (99.90 percent). Key properties of methanol.
are listed in Table 7-16. Additional data are available in
the literature (34,35). Material safety data sheets
(MSDSs) from chemical companies can be extremely
useful sources of information about the properties, use,
Table 7-16. Key Properties of Methanol
Density
Vapor density (air = 1.00)
Vapor pressure
@ 0°C
@ 10°C
@ 20°C
© 30°C
@40°C
© 50°C
Solubility
Viscosity © 20°C
Evaporation rate (butyl acetate = I)
Combustible limits, percent by
volume in air at standard
temperature and pressure
Flash point (tag open cup)
Flash point (tag closed cup)
0.7913 g/mL @ 20°C
(6.59 Ib/gal)
1.105 @ 15°C
29 mm Hg
52 mm Hg
96 mm Hg
159 mm Hg
258 mm Hg
410 mm Hg
Miscible in all
proportions with water
0.614 cps
4.6
7.3 to 36
16°C (61 °F)
12°C (54°F)
handling requirements, and hazards associated with
methanol; this should be made available to operations
personnel.
Taken internally, methanol is highly toxic. It also is harmful
if the vapors are inhaled or if it comes in contact with skin
for prolonged or repeated exposure. Fire and explosion,
however, are the primary dangers associated with metha-
nol. Again, individuals involved in handling methanol
should be made aware of these hazards. Federal, state,
and local regulations for safety should be posted, along
with supplier information.
7.5.2 Regulations and Standards
The shipping, unloading, storage, and handling of any
flammable chemical, including methanol, are governed by
a number of stringent requirements. These include federal
regulations imposed by the Department of Transportation
(DOT) and by the Occupational Safety and Health Act
(OSHA); state safety orders and codes; municipal ordi-
nances; guidelines issued by independent associations
such as the National Fire Protection Association (NFPA)
and the Manufacturing Chemists Association (MCA); and
precautions imposed by insurance companies. These re-
quirements must be studied before methanol facilities are
designed, and all must subsequently be followed.
7.5.3 Delivery and Unloading
Methanol is shipped in 208 L (55 gal) metal drums, tank
wagons, tank trucks, and tank railcars. Other methods of
shipping, not discussed at length here, are barge, metal
drums (smaller than 208 L [55 gal]), and glass and metal
cans. Tank wagons normally hold 3,785-15,140 L (1,000-
4,000 gal), tank trucks 15,140-34,060 L (4,000-9,000
gal), and tank railcars 22,700-37,850 L (6,000-10,000
gal). Tank railcars and tank trucks represent the most
economical shipping mode for most facilities. However,
for pilot work and small plants, 208 L (55 gal) metal drums
may be appropriate. Because methanol is classified by
the DOT as a flammable liquid, all shipping containers
must be approved and labeled in accordance with appli-
cable DOT regulations. Additionally, methanol fumes must
be properly vented during transport.
The recommended method of unloading methanol from
any container is pumping. Some barges and tank wagons
have their own pumps for unloading,,Tank railcars and
trucks can be unloaded from the top or bottom, and the
methanol can be pumped or conveyed* by gravity or
syphoning. The preferred method of unloading is pumping
from the top via an eductor tube. Because of, thei in-
creased likelihood of spillage when unloading from the
bottom, railcars or trucks must be approved for .bottom
unloading. They must incorporate valving that is approved
by the Association of American Railroads (AAR) andJhat
meets DOT requirements. This valving helps contaiathe
product by safely controlling flow. Additional precautions
such as fusible link valves and excess flow valves, may
be used. ;, , .-,'.'.^'..',*'•
Air pressurization of the tank (air padding) must heverrbe
used for methanol unloading. However, top unloading us-
ing a displacement method with an inert gas such "as
carbon dioxide or nitrogen for padding may be permissi-
ble if the unloading procedures follow the chemical sup-
plier's specifications exactly. Unloading procedures
should be validated against current supplier information
and pertinent regulations. •• •'•'• •''-3r~ ••'
General requirements for the design' of unloading faeWfies
for methanol apply to both tank raitcars and truc|s.'The
unloading area should be located away from the;pfahVs
principal traffic areas. Also, all storage facilities shoulcf be
located outside because of the fire hazard, and alPeguip-
ment in the vapor area must be explosion-proof-^Cllfs I,
Group D, Division 1 or 2—-per the National Jifeclfical
Code. Tools used in unloading should be of the honsp'ark-
244
-------
ing type. Unloading should occur during daylight hours
because the safety and lighting requirements for night
operation are particularly extensive. Ample fire extin-
guishers, safely blankets, deluge showers, eye washes,
no-smoking signs, and unloading-area signage are also
required.
If top unloading is utilized, approach platforms are re-
quired for access to the top of the tank. In all unloading
setups, all equipment must be grounded, including the
shipping vessel, interconnecting piping, pumps, and ap-
proach platforms. Also, bonding jumpers must be used to
provide a favorable continuous system, and the ground-
ing system must be checked routinely.
Static electricity buildup must be minimized because a
spark discharge can easily start a fire or cause an explo-
sion. Refer to National Fire Protection Association codes
for ways to prevent excessive static electricity.
Truck unloading areas must faciltate truck maneuverabil-
ity, upon both entering and leaving the area. When the
number of individual unloading spots is determined, fac-
tors such as frequency of use, amount of space, and
ability to unload several truck simultaneously should be
considered. To facilitate truck traffic flow, it is best to have
parallel unloading setup areas such that trucks can drive
straight through.
Rail unloading requires additional considerations. Un-
loading areas must have derailers or a closed switch at
a minimum of one car length away from the car being
unloaded. A primary issue in rail unloading is who uncou-
ples and parks the cars—the plant or the railroad. It is
preferable to have private sidings so the railroad can drop
off or pick up cars at any time without disrupting plant
operations. Also, the cost of having the railroad uncouple
and park cars at night, over weekends, or on holidays
can be high. In addition, because the railroad usually
cannot guarantee arrival time of shipments, safety provi-
sions and lighting must be provided for night operations.
Two sidings should be provided—one for empty cars and
one for full cars. It is recommended that sufficient storage
be provided to accommodate the contents of a full rail
car. The cars can be maneuvered by plant personnel
using railcar movers that operate both on rails or on a
roadway surface. In the case of short distances, winches
can be used. By paving the railroad yard area-, both truck
and rail unloading can be practiced. This is advantageous
in the event of strikes affecting either kind of transport.
Although unloading equipment is typically made of steel,
many other materials (except aluminum) are acceptable,
, providing they can withstand pressure loads and are com-
pletely grounded. Pumps must be of the nonsparking
typeVisuch as bronze-fitted steel pumps with bronze im-
pellers. Many materials are compatible with methanol, so
seal's and gaskets can be constructed from common ma-
terials^ Pumps may be either centrifugal or the positive-
displacement types; however, positive displacement
pumps must have relief valves. Because of the widely
varying heads encountered during unloading, a pump
should be carefully selected. Piping should have as few
joints as possible and should have a Schedule 40 rating
as a minimum. Splash guards at joints may be desired in
traffic areas. Valves may be of the gate-, plug-, or dia-
phragm-type, and may be made of iron or steel with
bronze trim. Neoprene can be used for the plugs in the
plug valve or for the diaphragm in the diaphragm valves.
Some refineries on the West Coast have adopted a stand-
ard of using cast steel valves in all lines carrying flam-
mable materials, to prevent damage during a fire.
Couplings must be leak-proof, It is preferable to have a
valve next to the coupling to limit material leakage and
waste during disconnection. If flexible hose connections
are used, a coupling with an integral valve can be used.
A strainer should be used ahead of any pumping or stor-
age equipment.
Care must be exercised to avoid overfilling the storage
vessel. A high-level alarm and pump shutoff should be
used. Because of the cost of methanol, it may be desir-
able to have a flow meter in the unloading piping to
monitor flow level. All vessels must be vented during
unloading or loading.
7.5.4 Storage
In order to accommodate delays in methanol delivery,
facilities should have the capacity to store a two- to four-
week supply. The volume of storage will be determined
by various site and cost requirements; however, storage
of less than a two-week supply would provide too small
a buffer for expected delivery delays and strikes. Tank
truck deliveries require in-plant storage; with rail deliver-
ies, the railcars can be used for storage, though charges
(demurrage) are levied by the carriers for time on site in
excess of a fixed time. For small plants, demurrage may
be cost effective; however, carriers may have a time limit
on the loaning of cars or have excessive'demurrage
charges.
Although methanol may be stored in vertical or horizontal
aboveground tanks, which can be located inside or out-
side, it is strongly recommended that all methanol equip-
ment and tanks be located outside to minimize dangers
associated with the possibility of explosions caused by
methanol fumes. If interior storage is required, refer to
NFPA codes for detailed requirements. An exception to
this rule involves drums that, if riot stored indoors, must
be shaded from direct sunlight of constantly sprinkled with
water.
Methanol .tanks should be laid out in accordance with the
requirements of the NFPA. There should also be a dike
around each aboveground tank or group of tanks to con-
tain 125. percent of the largest tank's volume in .case of
a rupture or fire. If the tanks are not made of steel, care
245
-------
must be taken to ensure that a fire will not rupture the
entire group of tanks, thereby causing methanol to over-
flow the dike. Fire protection is critical, especially when
the tanks are near other structures. For large volumes of
methanol storage, low-expansion alcohol-type foam is
used for fire extinguishing. For small fires, dry chemical
or carbon dioxide extinguishers can be used. Tempera-
ture detectors may be used for sensing fire and initiating
automatic release of foam. Water should not be used.
Storage tanks are normally made of steel, but most com-
parable materials are satisfactory (except for aluminum
and lead). Tank size depends on the required capacities
and any size limitations imposed by the tank material.
Piping, valves, and other components should be as de-
scribed in Section 7.5.3. Tank fittings should include:
• An inlet with dip tube to prevent splash and static
electricity.
• An anti-siphon valve or hole on the inlet to prevent
back siphonage.
* A vent pipe with a pressure-vacuum relief valve (36)
with fiame arrester.
• An outlet connection.
• A drain connection.
• Various openings for depth gauges, sample points, and
level switches.
Tanks should also have manholes for access. Extreme
corrosion will take place if the tank is drained dry and not
cleaned. The tank must also be grounded. Because of
increasing air pollution requirements, venting must be
controlled by conservation vents or by maintaining a slight
negative pressure in the tank using a small ejector.
To maintain the proper quantity of materials, a diaphragm
level sensor or float should be used. Low- and high-level
alarms are needed to protect against overfilling and
against pumping out settled material from the bottom of
the tank. The high-level alarm should be separate from
the tank sensors for a fail-safe design.
7.5.5 Transfer and Feed
Methanol must be controlled during transfer from the stor-
age vessel to the point of feed, which is accomplished by
gravity-feed or with pumps. Generally, pumps are easier
to control. Transfer pumps should always have positive
suction pressure and should be protected by a strainer.
As with all methanol use, it is desirable to mount all
equipment outside. There are two basic pumping arrange-
ments that can be used: 1) positive displacement chemi-
cal feed pumps with adjustable stroke, speed, or both,
where flow rate can be measured continuously or cali-
brated periodically, and the flow rate is based on speed
and stroke settings; or 2) centrifugal or regenerative tur-
bine pumps with variable-speed drives controlled by a
flow meter. Each arrangement has its own particular limi-
tations and must be considered on an individual basis for
each installation. However, positive displacement pumps
are more common in chemical feed applications because
they are more accurate and less dependent on discharge
pressure.
All pumps and piping should be uniform, as noted in
Section 7.5.3. All piping should be tested for 1.5 times
the maximum system pressure for 30 min, with zero
leakage.
7.5.6 Feed Control
Because methanol is expensive and an overdose can
result in a high effluent BOD, it is essential to pace the
methanol feed accurately with the oxidized nitrogen load.
Simply pacing methanol dose against plant flow will lead
to inaccuracies since this approach does not account for
daily and diurnal variations in the nitrate concentration.
Nitrification
Effluent
Legend
FT Flow Transmitter
FFIK Ratio Station
Wet Chemical Analyzer
Auto/Manual Control Station
Analog Multiplier
Methanol from Storage
Methanol Feed Pump
Flow Meter
Figure 7-18. Schematic of automatic feed forward control system for methanol pacing.
To Denitrification
System
246
-------
There are two methods that can be used to control metha-
nol dose: automatic feed forward control and manual con-
trol. Manual control is more commonly used.
Feed forward control using plant flow and nitrification sys-
tem effluent nitrate is shown in Figure 7-18. Feed ratio is
approximately three parts methanol per one part nitrate
nitrogen by weight (see Chapter 4). This control method,
which rarely proves successful and is not commonly
used, requires continuous online measurement of nitrate
using an online, automated, wet chemistry analyzer or
frequent grab samples for methanol measurement.
.In the online measurement system, the AIT output is pro-
portional to nitrate concentration in the nitrification efflu-
ent. The manual control station (HIK) provides means to
select either the analyzer output or to enter a manual
concentration value in case of analyzer failure. The output
of HIK is multiplied by a signal proportional to flow from
the ratio stations (FFIK) to obtain a signal proportional to
the required methanol flow ratio. This signal may then be
fed to a chemical proportioning pump, as shown in the
figure, or may be the .setpoint of a flow control loop. FFIK
provides a means to adjust the methanol feed ratio. The
dependability of this control procedure is predicated on
the reliability of the automated wet chemical analyzer.
These analyzers require very careful routine maintenance
and calibration and are not commonly used.
The most common technique for feed control is the man-
ual method in which operators dose methanol until a
hydrogen sulfide smell is recognized (13). However, this
method can make it difficult to control the chlorination
system because of intermediate formation of nitrite-
nitrogen.
7.5.7 Removal of Excess Methanol
Unless specific measures are taken to provide for metha-
nol removal, addition of methanol above stoichiometric
requirements (see Chapter 4) will cause methanol to ap-
pear in the denitrification process effluent. In one reported
instance, a methanol overdose caused an effluent BOD5
of 106 mg/L (1). Placing total reliance on the methanol
feed control system to prevent methanol overdoses may
be unrealistic in small plants, where a trained technician's
attention can be expected to be infrequent. The provision
of a reliable methanol control system and a methanol
removal system as a backup should allow nearly fail-safe
operation in terms of preventing effluents containing high
levels of organics.
A modification of the suspended growth denitrification
process, geared to prevent methanol bleedthrough, is
shown in Figure 7-1. After denitrification, mixed liquor
passes to an aerated stabilization tank where facultative
organisms "switch over" from using nitrate to using DO
and oxidize any remaining methanol. While there is some
uncertainty about the length of time required for faculta-
tive bacteria to make the switch and complete methanol
oxidation, it has been reported that 30 min of aeration in
an aerated stabilization tank at a Burlington, Ontario, pilot
plant was insufficient, as high effluent methanol values
were periodically observed in the system (1). However,
other plants, including River Oaks in Florida, have re-
ported 30-60 min to be acceptable (4). A period of about
48 min has been found to be sufficient for methanol oxi-
dation (1). Therefore, a 1 -hr aeration is adequate on the
basis of experience to date. There must be a sufficient
population of microorganisms acclimated to the presence
of methanol to oxidize the excess methanol. A sudden
shock load of methanol will not be oxidized readily be-
cause the microorganisms will not be acclimated to the
high dose. Anaerobic systems for methanol oxidation are
not recommended as a rule because they require even
more time for acclimation. Further details concerning sol-
ids-liquid separation are presented in Section 7.2.2.
In attached growth denitrification systems, the provision
of an aeration basin after the denitrification column does
not ensure oxidation of excess methanol. This is because
the mass of facultative organisms in the column's effluent
is not sufficient to oxidize the carbon biologically. The
mass is insufficient because denitrifying organisms are
retained in the column's media; only a few pass into the
column's effluent. Therefore, systems that remove excess
methanol are ineffective in attached growth systems.
Careful monitoring and alert operators are necessary to
ensure that excess methanol is not discharged.
7.6 References
When an NTIS number is cited in a reference, that docu-
ment is available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
703-487-4650
1. U.S. EPA. 1975. Process design manual for nitrogen
control. EPA/625/1-77/007 (NTIS PB-259149). Wash-
ington, DC.
2. Henderson, D., and G. Parlin. 1990. Pilot testing and
full-scale operation of fluidized-bed biological denitri-
fication. San Francisco, CA: Kennedy/Jenks/Chilton.
3. Monteith, H.D., T.R. Briddle, and P.M. Sutton. 1980.
Industrial waste carbon sources for biological denitri-
fication. Prog. Water Tech. 12(127).
4. Shrinde, J.R., and S.K. Bhagat. 1982. Industrial
waste carbon sources for biological denitrification.
JWPCF 54(370).
5. Soap and Detergent Association. 1991. Principals of
practice for nitrogen,and phosphorus removal from
municipal wastewater. Ann Arbor, Ml: Lewis Publish-
ing.
247
-------
6. Parker, D.S., W.J. Kaufman, and D. Jenkins. 1971.
Physical conditioning of activated sludge • floe.
JWPCF 43(9):1817-1833.
7. Water Pollution Control Federation and American So-
ciety of Civil Engineers. 1985. Clarifier design. Man-
ual of Practice FD-8 (Facilities Development).
Washington, DC.
8. Water Environment Federation. 1992. Wastewater
treatment plant design. Manual of Practice No. 8.
Alexandria, VA.
9. Pitter, P., and J. Chuboda. 1990. Biodegradabiiity of
organic substances in the aquatic environment. Boca
Raton, FL: CRC Press.
10. Parker, D.S., B.C. Aberley, and D.H. Caldwell. 1977.
Development and implementation of biological deni-
trification in two large plants. Prog. Water Tech.
8(4/5):673.
11. Tetreault, M., and D. Parker et al. Separate stage
denitrification—key to achieving a 3-3-2-0.6 AWT ef-
fluent. Presented at the 62nd Annual Water Pollution
Control Federation Conference, San Francisco, CA
(October 15-19).
12. Chen, J. 1980. Plant scale operation of a biological
denitrification filter system. Presented at the Ameri-
can Society of Civil Engineers Expo, FL (October
27-31). Preprint No. 80-586.
13. Pickard, D.W., R.E. Bizzurri, and T.E. Wilson. 1985.
Six years of successful nitrogen removal at Tampa,
Florida. Presented at the 58th Annual Water Pollution
Control Federation Conference, Kansas City, MO
(October 6-10).
14. Savage, E.S. 1983. Biological denitrification deep
bed filters. Paper presented at the Filtech Confer-
ence, Filtration Soc., London, England.
15. Water Pollution Control Federation. 1990. Nitrogen
control. Manual of Practice No. 9. Alexandria, VA.
16. TETRA Technologies. Denite System corporate bro-
chure. Coraopolis, PA.
17. Wilson, T.E., et al. 1980. Design and operation of
nitrogen control facilities at Tampa and the NSSD.
Paper presented at the U.S. EPA Int. Sem. on Control
of Nutrients in Municipal Wastewater Effluent, San
Diego, CA.
18. Savage, E.S., and J.J. Chen. 1972. Development
report for nutrient removal project at Indian Lake
Treatment Plant, North Huntington, PA. Dravo Corp.,
Pittsburgh, PA.
19. Savage, E.S., and U.J. Chen. 1973. Operating expe-
riences with columnar denitrification. Dravo, Corp.,
Pittsburgh, PA.
20. U.S. EPA. 1975. Process design manual for sus-
pended solids removal. EPA/625/1-75/003 (NTIS PB-
259147). Washington, DC.
21. American Water Works Association. 1989. Water
treatment plant design. 2d ed. Denver, CO.
22. Leopold A. Mueller Co. Corporate brochure no. ASU-
100. Zelienople, PA.
23. Patterson Candy International. Corporate brochure.
Type K L&A Water Treatment Corp. Pub. no. 4400.2.
London, England.
24. Button, P.M. 1990. Biological fluidized beds for water
and wastewater treatment: a user's forum. Ann Arbor,
Ml (May).
25. Sutton, P.M. 1991. Biological fluidized beds for con-
trol of nitrogen in municipal wastewaters: review of
commercial installations in the United States. Bethel,
CT: P.M. Sutton and Associates (March).
26. Sutton, P.M., and P.N. Mishra. 1990. Biological
fluidized beds for water and wastewater treatment: a
state of the art review. Presented at the 1990 WPCF
Conference, Washington, DC (October).
27. Envirex, Inc. 1987. Waukasha, Wl. Company litera-
ture.
28. Jerris, J., et al. 1975. Pilot scale, high rate biological
denitrification. JWPCF 47(8):2043 (August).
29. MacDonald, D.V. 1990. Denitrification by fluidized
biofilm reactor. Water Sci. Tech. 22(1/2):451.
30. Sutton, P.M., and Mishra, P.N. 1991. Biological
fluidized beds for water and wastewater treatment.
Water Environ, and Tech. 3(8):52 (August).
31. Mulcaly, L.T., and E.J. LaMotta. 1978. Mathematical
model of the fluidized bed reactor. Department of
Civil Engineering, University of Massachusetts.
32. Shieh, W.K., P.M. Sutton, and P. Kos. 1981. Oxidation
system fluidized bed wastewater treatment proc-
esses. JWPCF 53:1574.
33. Stephenson, J.P., and K.C. Murphy. 1980. Kinetics of
biological fluidized bed wastewater denitrification.
Prog. Water Tech. 12:159.
34. Alberta Gas Chemicals, Inc. 1989. Methanol material
safety data sheet. Parsippany, NJ.
35. Austin, G.T. 1974. Industrially significant organic
chemicals—Part 7. Chemical Engineering 81(13):
152-153.
36. National Fire Protection Association. 1990. Flamma-
ble and combustible liquids code. NFPA 30. Battery
Park, Quincy, MA.
248
-------
Chapters
Design Considerations for Single-Sludge
Nitrification-Denitrification Processes
8.1 Introduction
Single-sludge nitrification-denitrification processes were
first developed and applied in the 1960s. Since then,
these processes have gained popularity, particularly in
small- to medium-sized plants. Driving factors include
less-critical energy and onsite tankage considerations, an
increase in the general understanding of basic process
principles, marketing efforts by companies promoting pro-
prietary single-sludge systems, and a perception that
such systems offer potential cost advantages over multi-
ple-sludge nitrogen removal processes and systems with
separate-stage denitrification. Single-sludge systems for
nitrogen removal basically combine carbonaceous re-
moval, ammonia oxidation, and nitrate reduction within
the same process, using modified versions of the acti-
vated sludge process with a single sedimentation step for
separation of the biological sludge. As the enforcement
of effluent nitrogen limits became more prevalent in vari-
ous parts of the country, increased efforts were made to
develop new or modified versions of the single-sludge
process. As a result, there is now a wide variety of system
configurations from which to choose. Single-sludge sys-
tems have been developed with various combinations of
single or multiple anoxic zones, oxidation ditches, se-
quencing batch reactors, and cyclical aeration systems.
Some manufacturers have obtained patent rights for vari-
ous types of systems or portions of systems and impose
a licensing fee for their use. Nonproprietary versions of
some of these systems have been developed by other
manufacturers. Manufacturers of proprietary systems
generally offer performance guarantees and design as-
sistance.
Single-sludge systems are available with a variety of
design layouts, reactor configurations, inlet feed arrange-
ments, compartmentalization or baffling, mixing proc-
esses, return sludge requirements, internal recycle
patterns, aeration processes, integrated phosphorus re-
moval techniques, performance capabilities, process con-
trol requirements, and miscellaneous support approaches
and controls. This chapter is intended to assist the reader
in screening, evaluating, and/or selecting, if appropriate,
a single-sludge system. The chapter also provides infor-
mation on the types of systems, design considerations,
and features of various configurations, support systems,
performance capabilities, operational requirements, and
other factors to consider in designing new plants, plant
expansions, and retrofits of existing plants.
Single-sludge systems offer several advantages over
multiple-sludge systems or separate-stage systems.
Without intermediate clarifiers or separate denitrification
units, there is a potential cost advantage, if the costs of
larger reactor tankage and energy requirements do not
exceed these benefits. Factors to consider compared to
separate sludge/stage systems are space availability, re-
duction in alkalinity consumption, use of wastewater carb-
on as a carbon source for denitrification in lieu of
methanol and lower oxygen requirements.
Potential limitations or disadvantages to consider com-
pared to separate sludge/stage systems include greater
sensitivity to toxicity or inhibition without a separate up-
stream biological treatment step, lower nitrogen removal
efficiency, higher energy usage (compared to separate
stage), larger volumes of reactor tankage, and greater
site requirements. These potential disadvantages are site
specific and all or none may apply to a particular situation.
The major facto'r—in addition to the effluent nitrogen
limit—in evaluating and comparing a single-sludge sys-
tem to other systems is cost comparison in terms of capi-
tal outlay and operation and maintenance. Single-sludge
systems can more readily be used in retrofitting existing
activated sludge plants for nitrogen removal, particularly
if the plant has excess capacity. Single-sludge systems
can be followed by a separate stage for denitrification
where more stringent nitrogen limits are imposed. The
separate stage may need to be operated during winter
only while operating the single-sludge system exclusively
for nitrification. During warmer months, the single-sludge
system would be used for nitrogen removal without the
separate stage, thus eliminating methanol costs.
This chapter provides an overview of generic types and
classification of single-sludge processes and describes
249
-------
available single-sludge systems, including operational pa-
rameters, typical performance and process design fea-
tures, considerations for selection and design of
single-sludge systems, and process design examples.
8.2 Classification of Single-Sludge
Processes
8.2.1 Classification System
Single-sludge systems are generally classified according
to their flow regime, staging of anoxic and aerobic se-
quences, or method of aeration. All the classifications and
their component processes require nitrification to occur
in an aerobic zone or reactor, followed by denitrification.
For denitrification to occur, nitrates must be present to-
gether with an organic carbon source. Organic carbon
can be provided by the endogenous activity of the micro-
organisms (i.e., by depleting the cell's mass) or by an
exogenous source such as the BOD of the influent waste-
water or primary effluent. To use endogenous activity as
the carbon source, plant flow would be conveyed sequen-
tially through a combined BOD removal/nitrification step
in an aerobic zone or reactor, and then to the endogenous
anoxic zone or reactor to denitrify the nitrates. Alterna-
tively, the influent BOD can be exploited for denitrification
by either: 1) recycling nitrates to an anoxic zone or reactor
that precedes the aerobic zone, 2) operating alternate
anoxic/aerobic conditions within a single zone or reactor,
or 3) conveying the flow sequentially through alternating
anoxic/aerobic zones. Since denitrification cannot occur
without nitrification occurring first, systems are designed
and sized to completely nitrify the oxidizable influent TKN.
Thus, conventional parameters such as F/M ratio, reten-
tion time, oxygen transfer rate, and solids retention time
(Gc) are used in sizing the aeration equipment and tank
volume. Denitrification can then be achieved by convey-
ing the oxidized nitrogen in the form of nitrates to an
anoxic zone. A summary of the categories and charac-
teristics of the general single-sludge classifications is pro-
vided below. The classifications are presented in
approximate order of familiarity and complexity.
1. Multistage. Processes are most commonly configured
as suspended growth treatment. Variations in aeration
conditions are achieved spatially in different reactors
as flow is conveyed through the process train. This
process class can be further subdivided into:
a. Single Anoxic Zone (Section 8.2.2). Uses one an-,
oxic stage for denitrification and represents one of
the simplest configurations for nitrogen removal in
a single-sludge system. The most common con-
figuration to achieve denitrification involves recy-
cling nitrified mixed liquor to an antecedent anoxic
zone, where exogenous carbon provided by the
influent wastewater can be used by the facultative
denitrifiers. Nitrates that are not recycled will be
discharged to the final clarifier. Configurations util-
izing endogenous carbon denitrification are not
generally used at full scale. Examples of single
anoxic zone processes include anaerobic/an-
oxic/oxic (A2/O), Modified Ludzack-Ettinger (MLE),
Virginia Initiative Plant (VIP), and University of
Capetown (UCT) processes.
b. Multiple Anoxic Zones (Section 8.2.3). Uses more
than one anoxic zone. Two anoxic zones are most
commonly used. The carbon source for denitrifica-
tion may be either endogenous or exogenous,
however, endogenous denitrification should be
preceded by an exogenous denitrification reactor
for maximum nitrogen removal. Endogenous deni-
trification is commonly used to denitrify the nitrates
that were not recycled to the antecedent exoge-
nous denitrification reactor.
Exogenous denitrification can be achieved by the
following design strategies: 1) recycling nitrified
mixed liquor to an antecedent anoxic zone, 2) step-
feeding raw wastewater or primary effluent to an
anoxic zone containing nitrates, or 3) supplement-
ing the depleted carbon in the nitrified mixed liquor
with methanol. For systems that denitrify by em-
ploying two exogenous zones with internal recycle
and no endogenous zone, the final effluent nitrate
concentration is controlled by the recycle rate
since the aerobic zone is not followed by another
anoxic zone. This process configuration does not
achieve effluent TN (total nitrogen) concentrations
as low as configurations that have an endogenous
anoxic zone following BOD removal/nitrification.
Step-feeding raw wastewater or primary effluent to
provide substrate for exogenous denitrification re-
quires a final aeration step to-nitrify the ammonia
that bypasses the initial BOD removal/nitrification
process.
The Bardenpho and Modified UCT processes are
examples of dual anoxic zone processes. :
2. Multiphase/Cyclical Aeration (Section 8.2.4). Cyclical
technologies are generally a modification of the acti-
vated sludge process. Alternating anoxic/aerobic se-
quences are achieved in continuous flow reactors or
compartments by pulsing the aeration source. The
aeration frequency or intensity should be adjusted
such that the DO in the reactor does not exceed 2
mg/L during the aerobic phase. If several alternating
reactors or zones are used in series, raw wastewater
or primary effluent may be step-fed to those reactors
in which wastewater organic carbon has been de-
pleted or is present in rate-limiting concentrations.
3. Oxidation Ditches (Section 8.2.5). Oxidation ditches
are perhaps the simplest treatment scheme, but are
250
-------
less common in the United States than conventional
activated sludge configurations. Wastewater flows in
a continuous circuitous path and aeration is provided
at fixed points along the flow path. Anoxic conditions
are achieved between the aerators as oxygen is de-
pleted. The hydraulic retention time of an oxidation
ditch is generally longer than in multistage systems.
4. Sequencing Batch Reactors (SBFts) (Section 8.2.6).
SBR technologies are among the oldest technologies.
By pulsing the aeration mechanism on a timed cycle,
alternating aerobic and anoxic conditions are
achieved on a temporal basis within a single reactor,
as opposed to a spatial basis, and all reactions and
settling occur in the same reactor.
Several manufacturers of proprietary processes can pro-
vide various modifications to achieve nitrogen or phos-
phorus removal, or both. In cases where more than one
process is available, the process discussion will be limited
to the process designed exclusively for nitrogen control.
Examples of processes with multiple variations include
the Bafdenpho (four- or five-stage), Kruger, and Schreiber
processes. Although the five-stage Bardenpho process is
more common than the four-stage, only the latter will be
discussed in detail since it is designed exclusively for
nitrogen removal.
8.2.2 Single Anoxic Zone
8.2.2.1 Background and Process Description
The simplest continuous-flow single-sludge configura-
tions rely on a dedicated compartment or tank for deni-
trification. The earliest investigation of single-sludge
nitrification-denitrification processes for domestic waste-
water was documented by Wuhrmann (1), but a concur-
rent system was developed by Ludzack and Ettinger (2).
These two systems are presented schematically in Fig-
ures 8-1 and 8-2, respectively. The difference between
these two systems is related to the carbon source for the
denitrifying population. The Wuhrmann process places
the denitrification reactor after the combined carbon oxi-
dation/nitrification step, thus, this configuration has also
been termed postdenitrification. The electron donor (carb-
on source) in a postdenitrification process train must be
provided from endogenous decay, which is an intracellular
depletion of organic carbon. The Wuhrmann process was
not tested at full scale, but Christensen (3) was able to
demonstrate 88-percent TN removal. Subsequent studies
of the Wuhrmann process determined it to be unsuitable
for full-scale application because of high effluent turbidi-
ties (presumably caused by lack of a post-aeration com-
partment, and/or long solids residence times (9C), the
potential for increased effluent ammonia levels from lysed
Effluent
> WAS
£2 Anoxic Zone
[f^j Aerobic Zone
Figure 8-1. Wuhrmann process.
Effluent
Influent
(Q)
*WAS
[23 Anoxic Zone
EH Aerobic Zone
;,{.**; . •' ' ' ; '
Figure 8-2. Ludzack-Ettinger process.
251
-------
organisms, and low denitrification rates. The Wuhrmann
design pioneered single-sludge nitrification-denitrification
processes, but this process has not been used at
full scale without modifications such as step-feed
arrangements or supplemental carbon addition. Wuhr-
mann's effort provided the basic comprehension of the
nitrification-denitrification process and microbiology for fu-
ture refinements and modifications.
The system developed by Ludzack and Ettinger differed
from the Wuhrmann system by placing the anoxic deni-
trification zone ahead of the aerobic zone, using external
(exogenous) carbon provided by the raw wastewater. This
type of process is termed predenitrification. The nitrate
source was provided by directing the return activated
sludge to the anoxic reactor. Conventional underflow ra-
tios of 0.2:1 to 0.5:1 would not be expected to provide
sufficient nitrates to optimize the amount of denitrification,
and thus would be rate limiting.
Barnard (4) improved the Ludzack-Ettinger process by
providing an additional internal MLSS recycle from the
aerobic stage to the anoxic stage to return nitrified MLSS
at a regulated rate. This modification ensures adequate
nitrates for the heterotrophic denitrification population.
Process control and specific denitrification rates were en-
hanced with these modifications; consequently, process
performance was improved. TN removals of 88 percent
were achieved. The MLE, by Barnard, schematically illus-
trated in Figure 8-3, was not extensively implemented at
full scale, but was the progenitor of proprietary configu-
rations, such as the A2/O, Bardenpho, UCT, and VIP.
Variations of the MLE process design have been investi-
gated by German and Japanese researchers (5-7).
Schreiber and Menzel (6) proposed looped reactors,
which place an anoxic reactor concentrically within the
external aerobic reactor. Influent is received in the interior
anoxic reactor, which may then be directed to the outer
ring at the desired rate, either by a baffle system (Menzel
process) or by a dedicated internal recycle line
(Schreiber).
A proprietary single anoxic zone configuration is the A2/O
(anaerobic/anoxic/oxic) process, patented by Air Prod-
ucts, Inc. Originally developed for phosphorus removal as
the A/O process (anaerobic/oxic), nitrification-denitrifica-
tion was accommodated with the addition of an anoxic
zone between the anaerobic and aerobic zone. Although
the anaerobic zone is not required for nitrification-
denitrification removal, it may be used at the start of the
treatment train as an anaerobic "selector" for nitrification-,
denitrification in scenarios that do not require phosphorus
removal. The anaerobic selector is used to control and
maintain tank conditions to promote the profligation of
zoogleal organisms, while suppressing the growth of fila-
mentous organisms in the anoxic and aerobic reactors.
Anoxic compartments located at the head end of the
biological treatment train have demonstrated similar
benefits (8). A discussion of selectors is provided in Sec-
tion 8.4.7. A schematic diagram of the A2/O process is
presented in Figure 8-4. The A2/O process is now mar-
keted and licensed by Kruger Company.
The UCT process was developed at the University of
Capetown in South Africa to surmount one of the inherent
limitations of the MLE and A2/O processes—the interfer-
ence of nitrates on phosphorus removal processes. This
was accomplished by: 1) returning activated sludge to the
anoxic zone instead of to the anaerobic zone, and 2)
providing an additional recycle from the anoxic zone to
the anaerobic zone. The UCT process schematic is
shown on Figure 8-5. The purpose of these modifications
is to denitrify nitrates returned by the RAS (reutrn acti-
vated sludge) line before they are recycled to the anaero-
bic zone. A further refinement of the UCT process to
accommodate lower strength wastewaters in the United
States was investigated in Norfolk, Virginia. This process
became known as the VIP process. Although the VIP and
UCT processes are schematically similar, there are two
fundamental differences: 1) the VIP process uses multiple
complete mix cells instead of a single anaerobic reactor;
this modification is intended to enhance phosphorus up-
take by allowing a higher concentration of residual or-
ganics in the first anaerobic cell; and 2) because of the
lower-strength wastewaters in the United States, a higher
system rate (i.e., shorter 6C) is afforded in the VIP process
to increase the proportion of active biomass in the mixed
liquor; this allows a smaller reactor volume and a shorter
0C. The VIP process is patented, but its developers have
waived the process fee.
8.2.2.2 Typical Design Criteria
Owing to their process limitations, the Wuhrmann and
original Ludzack-Ettinger processes are not commonly
used. The more recent predenitrification single anoxic
zone processes are favored. Discussion of design criteria
will be limited to the A2/O, VIP/UCT, and MLE processes.
Endogenous postdenitrification zones (e.g., those used in
the Wuhrmann process) are used in some processes that
employ multiple anoxic zones.
The design procedure for a single-sludge, single-anoxic
zone nitrification-denitrification system consists of sizing
the aerobic zone to nitrify the influent oxidizable TKN
completely^ and then sizing the anoxic zone and deter-
mining the required recycle rate.
The procedure for sizing the aerobic zone can be deter-
mined by conventional 90 or nitrification rate considera-^
tions used in activated sludge nitrification applications, as
discussed in Chapter 6. In summary, the sizing of the
aerobic zone should consist of the following steps: -
1. Select the design aerobic GO-
2. Calculate secondary sludge production.
252
-------
Nitrified Recycle (100-400% Q)
Influent
. (Q)
WAS
\/A Anoxic Zone
E3 Aerobic Zone
Figure 8-3. Modified Ludzack-Ettinger process.
Nitrified Recycle (100-500% Q)
J2| Anaerobic Zone
7% Anoxic Zone
IB Aerobic Zone
Figure 8-4. A /O process with nitrification-denitrification.
Effluent
Anoxic Recycle
(100-200 %Q)
Nitrified Recycle (100-200% Q)
Anaerobic Zone
£2 Anoxic Zone
ETTI Aerobic Zone
Effluent
>WAS
Figure 8-5. University of Capetown (UCT) process.
253
-------
3. Calculate the required aerobic zone solids inventory
based on e£
4. Determine tank volume based on the solids inventory,
settling properties, peaking factors, and the design
MLSS.
The size of the anoxic zone should be based on the
amount of nitrates to be denitrified. The required nitrate
recycle rate is determined by the design effluent nitrate
concentration. From mass balance considerations, the
required combination of internal mixed liquor recycle rate
and return activated sludge recycle rate can be calculated
using Equation 8-1:
rate (Q|, m3/d) to be calculated. Equation 8-1 may be
algebraically manipulated to express Q| and QRAS in
terms of a desired percent removal, as-provided in Equa-
tion 8-2:
I + RAS = (% Removal) + (1 - % Removal)
where % removal is expressed as a decimal
(8-2)
[CTKNJ/(Q CJ] - 1 = RAS + I
(8-1)
where:
= total mass of oxidizable TKN (nitrates)
produced in the aerobic reactor, g/d
* influent TKN - effluent TKN - TKN
„ assimilated in waste sludge, g/d
Q = plant influent flow rate, m3/d
Ca =s desired effluent nitrate concentration, mg/L
I s nitrified internal recycle ratio of recycle rate
to plant influent flow rate
RAS = return activated sludge recycle ratio of return
sludge rate to plant influent flow rate
Typically, the return sludge flow rate (QRAS. m3/d) will be
calculated during the aerobic zone BOD/nitrification de-
sign step as the recycle rate necessary to maintain the
design MLSS leaving only the required internal recycle
The anoxic zone should be sized to denitrify the nitrates
recycled to it. The mass of nitrates recycled can be simply
expressed as:
(Q) (Ce) (I + RAS) = nitrates recycled to anoxic
zone, g/d (8-3)
The required anoxic solids inventory can be calculated
based on the selected SDNR (specific denitrification rate),
from which the anoxic volume can be derived. Typical
design criteria for single anoxic zone single-sludge ni-
trification-denitrification processes are provided in Table
8-1.
The design criteria for the A2/O process presented in
Table 8-1 reflect data compilation from three full-scale
plants. The design criteria presented for the VIP were
obtained from the pilot-scale study performed at the Lam-
berts Point Wastewater Treatment Plant in Norfolk, Vir-
ginia, which represents the only full-scale application of
the VIP/UCT process. The VIP criteria differ from the A2/O
criteria because of the different objectives of each proc-
ess and the conditions and influent characteristics at each
site. The VIP process is designed to optimize nitrogen
Table 8-1. Typical Design Criteria for Single Anoxic Zone Predenitrification Systems
Parameter
MLSS, mg/L"
HRT, hr
Anaerobic13
Anoxic
Aerobic
Ocd
F/M, g BOD5 applied/g MLVSS/d
RAS recycle, %Q
Internal recycle
Nitrified recycle, %Q
Anoxic recycle, %Q°
Mix Power, hp/Mgal
Anaerobic
Anoxic
A2/O
3,000-5,000
0.5-1
0.5-1
3.5-6
5-10
0.15-0.25
20-50
100-200
50
50
VIP/UCT
1,500-3,000
1-2
1-2
2.5-4
5-10
0.1-0.2
50-100
200-400
50-200
70
70
Generic Single
Anoxic Zone
1,500-4,000
0.5-2
0.5-2
2.5-6
5-10 '
0.1-0.3
50-100
100-400
40-70
40-70
* RflRftrf nn fntat mn«
-------
removal by providing two internal recycles. This modifi-
cation affords a greater total recycle of nitrates for deni-
trification without affecting phosphorus removal
processes. The A2/O process is generally operated at a
higher MLSS than the VIP and at a lower RAS rate. The
lower RAS rates in the A2/O process are required to
ensure that the anaerobic selector is not overloaded with
nitrates, which would adversely affect phosphorus
removal.
8.2.2.3 Process Performance
Single anoxic zone systems will typically achieve total N
effluent concentrations of <10 mg/L, and long-term aver-
age effluent total N concentrations of 8 mg/L can reliably
be achieved (9). Lower total N concentrations would re-
quire an additional anoxic zone or a separate denitrifica-
tion step.
As demonstrated in Equation 8-2, higher recycles are
required to achieve lower effluent nitrate concentrations.
However, practical limitations on the recycle ratios, due
to the energy required to pump large volumes, detract
from the viability of single anoxic zone technologies
where effluent nitrogen limitations are <5 mg/L, or at fa-
cilities where >80 percent TN removal is required. Return
sludge rates are generally limited to 100 percent of the
plant flow because of design solids considerations. Con-
sequently, higher internal recycle rates are necessary to
achieve lower effluent nitrogen levels. The increased
capital and O&M costs and the effect of higher pumping
rates on reactor retention time must be evaluated and
compared with the benefit of enhanced nitrogen removal
performance.
The theoretical efficiency of a single anoxic reactor sys-
tem may be derived from a mass balance analysis on
nitrates for a given reactor train, as illustrated in Figure
8-6 The quantity of total nitrates produced in the aerobic
zone, assuming complete nitrification of the available
TKN, is the oxidizable TKN or TKNOX. A fraction of these
nitrates will be recycled to the anoxic zone via the internal
MLSS and RAS recycles and the remainder will be
passed to the effluent. The recycled fraction (/NO3) may
be quantified mathematically as:
/N03 = (Q| + QRAS) * (Q + QI + QRAS) (8-4)
where Q, QI, and QRAs are the plant influent, internal
recycle, and return sludge flow rates, respectively, ex-
pressed in consistent units of volume per time unit.
For convenience, the recycle flows may be expressed as
a multiple of the influent plant flow, yielding:
TKNOX
C9 =-
(TKNOX)
QRAS
Q
I + RAS
1+RAS+l
Mass of nitrates produced in aerobic = Nitrates recycled + Ne
/ I + RAS x
TKNOX =
-------
/NO3 = (I + RAS ) -s- (1 + I + RAS) (8-5)
where:
l = Qi/Q
RAS =
The fraction of nitrates removed shown in Equation 8-5
for a single anoxic system is equivalent to the percent
removal. Figure 8-7 illustrates the relationship between
percent removal and total recycle ratio. The minimum
effluent nitrate concentration (Ce) will depend on the ni-
trate that is not recycled to the anoxic tank, expressed
as:
Ce-(1-/N03){TKNox-Q) (8-6)
or, expressed in other terms as:
Ca - (TKNOX) * [(Q) (1 + I + RAS)] (8-7)
Thus, the mass fraction of the total nitrate formed in the
aerobic zone that is removed in a predenitrification sys-
tem Is a function of the internal recycle rate from the
aerobic zone and of the return sludge rate, as presented
in Equation 8-7.
A plot of theoretical oxidizable nitrogen removal rate ver-
sus internal recycle for typical return sludge rates
(i.e., 50-100 percent) is illustrated in Figure 8-8. The fig-
ure demonstrates that the maximum removal efficiency
for a single anoxic reactor is 85 percent of oxidizable TN,
at realistic recycle rates (i.e., <400 percent). These rela-
tionships do not consider denitrification that may occur in
the final clarifier and within the floe particle in the aerobic
zone, as hypothesized (10,1). Consequently, observed
nitrate removals may be greater than results predicted by
theoretical considerations.
The A2/O process has been implemented at the Largo
Wastewater Treatment Plant in Florida. Performance data
from that plant are presented in Table 8-2.
The MLE process was also used at Maitland, Ontario, to
treat a high-strength industrial wastewater (nitrate = 175
mg/L, NHJ-N = 190 mg/L, BOD5 = 1,230 mg/L). During
optimum conditions, TN removals of 93 percent were
obtained (10).
There are no known facilities configured on the UCT proc-
ess in the United States. The VIP process was designed
for 70-percent removal of both TN and TP; however, lower
TN removal is anticipated in winter because of lower
nitrification rates (9). Theoretically, the UCT and VIP proc-
esses should achieve higher degrees of nitrogen removal
than the A2/O process as a result of both the extensive
use of baffles and compartments, and the denitrification
conditioning of the RAS in the anoxic stage. By denitrify-
ing the RAS in the anoxic zone, the nitrate interference
on phosphorus removal is alleviated. This affords a higher
RAS recycle rate and enhanced nitrogen removal effi-
ciency. Actual performance data for the VIP process in-
dicated that effluent TN was 6.8-11.3 mg/L, and the over-
all average was <8 mg/L; results comparable to those
obtained by the similar A2/O process. Experimental re-
sults from the VIP plant are presented in Table 8-2 (11).
Table 8-2. Performance Summary of Single Anoxic Zone
Processes
Parameter
Q, m3/d
BOD inf., mg/L
TKN inf., mg/L
BOD/TKN
TKN eff., mg/L
NHJ-N inf., mg/L
NHJ-N eff., mg/L
NO^-N eff., mg/L
Total N eff., mg/L
N removal, %
A2/O (9)
Largo, FL
39,360
204
23.5
8.7:1
2.2
—
—
5.7
7.9
66
VIP (11)
Pilot
Norfolk, VA
151,400
115
24.4
4.7:1
2.4
—
1.0
5.3
7.7
68
MLE
Landis, NJ
19,300
414
34.7
11.9:1
1.4
—
—
4.4
4.4
83
The MLE and A2/O processes can be optimized by vary-
ing the internal recycle and return sludge rates. However,
if phosphorus removal is required, the return sludge rate
must be minimized to control the nitrate level to ensure
an anaerobic condition. This nitrate interference of the
RAS recycle was cited as a limitation to phosphorus re-
moval and was the impetus for development of the UCT
process. If adequate volume exists, the MLE or A2/O can
be implemented at an existing facility requiring minor
modifications. These modifications may include installa-
tion of baffles, mixers, and internal recycle pump, as long
as the existing reactor volumes can provide adequate
retention times and sufficient aeration capacity is avail-
able.
8.2.2.4 Process Design Features
Anoxic reactors most commonly use a continuously
stirred tank reactor (CSTR) configuration, however, bench
and pilot scale studies have investigated plug flow (12)
and concentric circular reactors (5). These optional flow
regimes did not appear to offer significant process im-
provement over the CSTR (6). Plug flow regimes offer
better reaction kinetics; however, the increased oxygen
demand for nitrification can result in organic overloading
at the influent end of the reactor. This factor should be
considered when designing the aeration system.
The division of a single aeration tank into anoxic and
aerobic zones (and anaerobic zones for A2/O) can be
sufficiently achieved by a nonrigid baffle system; however,
256
-------
90
80
•< 70
"5 60
I
cc
50
40
£ .
s 30
0
1 20
1-
10
/
7
23 4 5 6
Total Recycle Ratio of Plant Influent Flow, I + R
Figure 8-7. Theoretical oxidized TKN removal performance for a single anoxic zone process as a function of total
recycle rate.
100
60
a.
1
I 50 :
40
30
Return Sludge Rate, % of
Plant Influent Flow
50
100
0 100 200 300 400 500 600 700 800 900 1000
Internal Recycle Rate, % of Plant Influent Flow
Figure 8-8. Theoretical oxidized TKN removal performance for a single anoxic zone process as a function of internal
recycle rate.
257
-------
extensive baffling within each zone as employed in the
VIP process did not appear to improve nitrogen removal
significantly. An exogenous carbon source for a single
anoxic zone system with a predenitrification configuration
is normally preferred, since postdenitrification arrange-
ments typically require methanol addition. Monitoring of
the denitrification rate should be performed to assess the
heterotrophic denitrification capacity for the specific influ-
ent COD or BOD. Bypassing the primary effluent to the
anoxic reactor was suggested as a means to ensure a
higher COD.TKN ratio for improved denitrification per-
formance (6).
Since theoretical nitrogen removal performance consid-
erations are derived based on complete nitrification, TN
removal performance is limited by the nitrification effi-
ciency in the aerobic reactor. The chemoautotrophic nitri-
fiers are susceptible to upsets from shock loadings of
BOD and ammonia, and the nitrification rate is highly
sensitive to changes in pH, alkalinity, and temperature.
The sensitivity of the nitrifiers provides further support for
a predenitrification configuration, which can serve as a
buffer zone against shock loads to the aerobic reactor.
Thus, process optimization will require monitoring of
these parameters in the aerobic basin, and adjustments
such as alkalinity control and increasing the solids reten-
tion time may be necessary.
Design features and considerations for the aerobic zone
follow the same procedure as the aeration basin of a
nitrification system. A detailed discussion of design for
nitrification systems is provided in Chapter 6. Additional
design considerations for nitrification-denitrification proc-
esses are discussed in Section 8.4.
Minimal DO should be introduced to the anoxic zone by
influent and recycle flows or by surface transfer. Reduced
denitrification rates at DO levels above 0.2 mg/L have
been observed (13). Thus, nitrified internal recycle flow
rates from the aerobic zone may require adjustment if
excess DO is introduced in the anoxic zone. This problem
can be mitigated in design by locating the internal recycle
line inlet from the aerobic tank in a relatively unaerated
corner of the tank where anoxic conditions may prevail.
Also, submerged mixers should be designed not to en-
train excessive air as a result of surface turbulence, but
to provide sufficient mixing to ensure maximum disper-
sion and exposure of recycled nitrate and substrate to the
denitrifying organisms.
Recycling of mixed liquor from the aerobic to the anoxic
zone may typically involve high-volume, low-head pump-
ing conditions. These applications may be achieved more
economically by installing low-head submersible propeller
pumps, sewage pumps, or vertical turbine pumps directly
in the aerobic basin, rather than by constructing a sepa-
rate dry pit pump gallery.
Both the MLE and A2/O process require only one MLSS
internal recycle, thereby limiting process flexibility to only
the RAS and internal recycle (IR). Additional flexibility and
ability to bypass primary settling or step feed as needed
may be achieved by providing interconnecting gates and
channels. If phosphorus removal at the facility is required,
RAS flow to the anaerobic zone must be minimized to
limit nitrate interference. The UCT and VIP processes
circumvent this limitation by conditioning RAS in the an-
oxic zone. This modification will enhance phosphorus re-
moval and will also entail a higher degree of process
monitoring, control, and operator sophistication. However,
the nitrified recycle and RAS rates must be carefully con-
trolled so that the nitrate load does not exceed the deni-
trification potential of the anoxic reactor, and result in a
nitrate load to the anaerobic reactor that would cause a
subsequent reduction of phosphorus removal.
A brief list of monitoring and control requirements for
single anoxic zone systems is outlined in Table 8-3.
Table 8-3. Monitoring Requirements and Rationale for Single Anoxic Zone Reactors
Ludzack Ettinger, MLE
A2/O, VIP, UCT
Reactor
Parameter Rationale
Parameter
Rationale
Anaerobic N/A , N/A
Anoxic DO Will reduce denitrification rate
QI Controls NO3 load
Aerobic DO High DO may inhibit denitrification;
low DO may inhibit nitrification
Alkalinity, pH Nitrification consumes alkalinity;
may require pH control
DO, Nitrates
Orthophosphates
DO
N03
Q,
DO
Alkalinity, pH
Presence of nitrates and DO will
mitigate fermentive organisms
Control to verify release
Will reduce denitrification rate
Inadequate load can cause excess
phosphate release
Controls NO3 load
High DO may inhibit denitrification;
low DO may inhibit nitrification
Nitrification consumes alkalinity;
may require pH control
258
-------
8.2.3 Dual Anoxic Zones
8.2.3.1 Background and Process Description
TN effluent concentrations <8 mg/L cannot be consis-
tently obtained using single anoxic zone processes with-
out an additional attached growth filter or methanol
supplement. TN effluent concentrations <6 mg/L can be
practically attained in a suspended growth system without
methanol addition by placing an endogenous anoxic zone
in series after the aerobic zone. While the A2/O process
does use two unaerated zones, the first (anaerobic) zone
is not used for enhanced nitrogen removal but is provided
for phosphorus removal or as an anaerobic selector. The
first documented case of a second anoxic zone for deni-
trification was credited to Barnard, depicted schematically
in Figure 8-9. This process served as a precursor to the
process he later patented as the Bardenpho process.
Phosphorus removal was later accommodated in the Bar-
denpho process by placing an anaerobic reactor at the
head of the treatment train, resulting in a five-stage proc-
ess also illustrated in Figure 8-9. The anaerobic fifth stage
can be included in facilities that are not required to re-
move phosphorus as an anaerobic selector to suppress
the growth of filamentous organisms.
The UCT process, described in Section 8.2.2, was also
further modified by providing two anoxic zones (instead
of one as in the original UCT) and two separate internal
recycle lines. The purpose of this modification was to
control the return sludge (RAS shown in Figure 8-10) and
the nitrate recycle separately and also to reduce the NO3
load to the anaerobic reactor. Although the Modified UCT
process uses dual anoxic zones, the second anoxic zone
is not an endogenous denitrification reactor as was de-
scribed for the Bardenpho process. Instead, the second
anoxic zone in the Modified UCT is used only to denitrify
' recycled nitrates from the aerobic zone, and the first an-
oxic zone is exclusively used as an exogenous denitrifi-
cation reactor to denitrify the RAS before recycle to the
anaerobic zone. This allows increased recycle rates to
the second anoxic zone for denitrification, and reduces
nitrate interference of phosphorus removal in the anaero-
bic reactor.
A nonproprietary multi-anoxic zone process (7) is illus-
trated in Figure 8-11. This design incorporated a three-
stage sequence of aerobic-anoxic basins and a step feed
to the second and third stages to supply the exogenous
carbon source. The staging of the aerobic-anoxic zones
served the purpose of an internal recycle, thereby offset-
ting O&M requirements with a larger capital cost associ-
ated with increased tank volume requirements. This
configuration would presumably not offer the degree of
process control compared to a design that included both
IR and RAS.
The Bardenpho process is marketed in the United States
by EIMCO. The patent describes a four-stage process,
with one nitrified internal recycle and an activated sludge
4-Stage (Nitrogen Removal) Process
Nitrified Recycle (400% Q)
Influent
(Q)
5-Stage (Phosphorus and Nitrogen Removal) Process
Nitrified Recycle (400% Q)
Final * Effluent
Clarifier
>WAS
K>d Anaerobic Zone
Y7\ Anoxic Zone
im Aerobic Zone
Figure 8-9. Bardenpho process.
259
-------
return. EIMCO administers a one time royalty fee for the
process, which can include startup, training, and guaran-
tee of performance (9).
8.2.3.2 Typical Design Criteria
Design criteria for single-sludge dual anoxic zone systems
(i.e., MLSS, recycle rates, retention time, and mixing en-
ergy) are similar to criteria presented in Section 8.2.2.2 for
single anoxic zone systems. The most significant difference
in dual anoxic zone design criteria from single anoxic zone
design criteria relates to whether provisions for phosphorus
removal are required. The long system 60's which improve
nitrogen removal, have been shown to adversely affect
phosphorus removal. The four-stage Bardenpho, for in-
stance, will typically be designed with a longer 9C than con-
figurations such as the A2/O or VIP that are designed for
phosphorus removal. The provision of a longer 80 typically
results in a lower sludge production rate.
If phosphorus removal is desired, a five-stage Bardenpho
can be selected by providing an anaerobic stage at the
front of the four-stage Bardenpho treatment train. As a
result, the first three stages of the five-stage Bardenpho
process are similar to the A2/O or VIP configuration. How-
ever, the final anoxic endogenous stage of the Bardenpho
process affords two important process enhancements
over processes that use single anoxic zones. The first is
Anoxic Recycle
(100-200% Q)
the additional degree of denitrification and consequent
lower effluent TN concentrations. Second, the resulting
reduced nitrate load to the final clarifier, which is recycled
in the RAS to the anaerobic stage, reduces the potential
for nitrate interference of phosphorus removal in the five-
stage Bardenpho process. These features permit the use
of higher internal recycle rates for a Bardenpho system
than can be used with single anoxic zone systems that
remove phosphorus and consequently improve nitrogen
and phosphorus removal performance.
The procedure for sizing the first aerobic and anoxic
zones of a dual anoxic zone process is identical to the
procedures and concepts used for single anoxic zone
systems. The first aerobic zone should be sized to nitrify
the oxidizable influent TKN. The first anoxic zone of the
Bardenpho system should be sized to completely denitrify
the internal and RAS recycled nitrates. The first anoxic
zone of a Modified UCT process should be .sized to de-
nitrify nitrates in the RAS.
The second anoxic zone of a Bardenpho is sized to de-
nitrify the nitrates not recycled to the first anoxic zone.
The nitrate load to this zone is the difference between the
oxidizable TKN and the nitrate reduced in the first anoxic
zone. The tank volume will also be a function of the nitrate
mass loading, temperature, MLSS, and SDNR. Since en-
dogenous denitrification rates are much slower than ex-
Nitrified Recycle
(100-200% Q)
f Final\ Efflu
Effluent
Influent
(Q)
Anaerobic Zone
Anoxic Zone
Aerobic Zone
Figure 8-10. Modified UCT process.
Influent
Step Feed
Effluent
E3 Anoxic Zone
CD Aerobic Zone
Figure 8*11. Multi-anoxic zone with step feed.
WAS
260,
-------
ogenous rates, the second basin will typically have a
higher volume per mass of nitrates applied.
The second anoxic zone of a Modified UCT process
should be sized to denitrify the oxidizable TKN recycled
from the aerobic reactor and the nitrates not recycled in
the first anoxic zone.
Typical values used in the design of the four-stage Bar-
denpho and Modified UCT system are presented in
Table 8-4,
The modified UCT has two fundamental process differ-
ences compared to a five-stage Bardenpho:
1. The Modified UCT is designed to optimize phosphorus
' removal.
2. No endogenous denitrification is provided. Thus, the
Modified UCT would be unable to attain effluent TN
concentrations consistently lower than 5 mg/L. Design
r for the Modified UCT involves similar design concepts
to a single anoxic zone process. Typically, the second
anoxic zone of a Modified UCT system is larger than
the first anoxic zone due to the relationship of SDNR
to influent COD. Since the COD to the second reactor
is lower than to the first and less easily degradable,
a lower. SDNR will be experienced, necessitating a
longer anoxic retention time. However, the rate in the
second anoxic reactor will be greater than the endo-
•genous rate in the second anoxic stage of the Bar-
' denpho system.
The distinguishing characteristic of the Modified UCT
process is the complexity of internal recycling require-
ments, which exceed those of the Bardenpho design with-
out offering a comparable degree of TN removal. As has
been discussed, this lower degree of efficiency is caused
by the phosphorus removal provision of the Modified UCT.
Table 8-4. Typical Design Criteria for Dual Anoxic Zone
Systems
4-Stage
Modified
Parameter
F/M, g BODg/g MLVSS/d
ec,d
MLSS, mg/L
HRT, hr
Anaerobic
1st Anoxic
Aerobic
2nd Anoxic
Reaeration
RAS, %
Internal recycle, %
Bardenpho
0.1-0.2
10-40-
2,000-5,000
— . • '
2-5
4-12
2-5
0.5-1
100
400-600
UCT
0.1-0.2
10-30
2,000-4,000
1-2
,2-4
•4-12
2-4
" —
100
100-600
To accomplish nitrification-denitrification without an inter-
nal recycle, primary effluent or raw wastewater can be
step-fed to the anoxic zones. For a system such as that
illustrated in Figure 8-11, the optimum step-feed ratio can
be derived or estimated from the influent wastewater
characteristics. Each aerobic zone should be sized to
completely nitrify all the influent TKN discharged to that
zone. Likewise, the anoxic zone should be sized to com-
pletely denitrify the nitrates produced in the preceding
aeration basin. The influent step feeds to the anoxic
zones should be balanced such that the influent COD to
each anoxic zone is sufficient to optimize exogenous ni-
trate respiration.
8.2.3.3 Process Performance
The Bardenpho design has achieved TN effluent concen-
trations of 3 mg/L and 90-percent removal afforded by the
endogenous postdenitrifjcation stage. The Bardenpho
process has been used at several plants in the United
States. A list of typical performance data is included in
Table 8-5. : -.;-••• .•:. ' ,.-•'"
In contrast, the Modified UCT process has not been employed
in the United States. Consequently, data for this configuration
are unavailable for an assessment of the process.
A mass balance analysis provides insight to the higher*
nitrogen treatment efficiency obtainable with a Bardenpho
process compared to the Modified UCT system. The theo-
retical nitrate removal efficiency for each system can
, be described by considering their respective recycle
patterns.
The fraction of total nitrates produced in the nitrification
tanks that are removed is a function of the internal recycle
rate from the aerobic to anoxic zone. This rate is typically
400 percent and 200 percent of the influent flow for the
Bardenpho and Modified UCT, respectively. The nitrate
fraction recycled from the first 'aerobic zone to the first
anoxic reactor of the Bardenpho is (assuming a typical
RAS rate of 100 percent):
Nitrate to first anoxic zone
= [(4Q) -4- (4Q + 2Q)] x TKNOX = (2/3) TKNOX ; (8-8),
Thus, one-third of the TKN oxidized in the aerobic zone
passes through to the second anoxTc zone. If no denitri-
fication is assumed to occur in the second anoxic zone
(an unrealistic assumption) then the total amount of ni-
trates not recycled to the first anoxic zone, one-third
TKNpx would De passed to the clarifier. The nitrates then
recycled to the anaerobic reactor of a five-stage Barden-
pho process at an RAS of 100 percent of the influent, Q,
would be:
Nitrates recycled to ariaerobjc reactor in RAS
= (1/3)(TKNox)(Q)/(2Q)k(1/6);TKNQX (8-9)
The total fraction of nitrates recycled (calculated in Equa-
tions 8-8 and 8-9) would be:'>! vv c ~ .,.:.? .:"-.;;•:.-«
261
-------
Table 8-5. Summary of Bardenpho Plant Operating Data
Plant
Tarpon Springs, FL
PalnieUo, FL
Ft. Myers-Central, FL
Ft. Myers-South, FL
Payson, AZ
Environmental Disposal Corp., NJ
Eastern Service Area, Orange County, FL
Kelowna, BC, Canada
Hills Development, Pluckemin, NJ
Flow, m3/d
(mgd)
10,068 (2.66)
4,656(1.23)
23,429 (6.19)
18,622 (4.92)
2,574 (0.68)
818 (0.216) '
12,112 (3.2)
12,491 (3.3)
908 (0.24)
Influent
BOD5,
mg/L
NA
160
135
144
196
190
175
188a
169
Influent
TKN,
mg/L
NA
36.60
23.30
25.40
32.80
17.20
30.60
24.20
18.3b
Effluent
Total N,
mg/L
4.4
2.9
2.7
5.1
3.2
2.8
1.9
1.8
2.7
%N
Removal
NA
92
88
80
90
84
94
91
85
•COD
"NHfrNonly
Total fraction nitrates recycled
» [(2/3) + (1/6)] TKNOX = (5/6) TKNOX
(8-10)
and the fraction of nitrates discharged to the effluent
would be:
Nitrates In effluent
= [(1) - (5/6)] TKNOX = (1/6) TKNOX (8-11)
The preceding discussion can be written in general for
any combination of internal recycle or return activated
sludge recycle rate, or overall percent removal in the
second anoxic zone. From these theoretical considera-
tions, a model can be developed to explain the enhanced
performance of the dual anoxic zones of the Bardenpho
process compared to the Modified UCT or to a single
anoxic zone process. The fraction of nitrates recycled in
the Internal recycle flow to the first anoxic zone (fNO3)
In a Bardenpho is:
/N03 = (Q,) * (Q + Q, + QRAS)
(8-12)
or, if the recycle rates are expressed in terms of Q, as
shown in Figure 8-12.
/N03 = (I) * (1 + RAS + I)
where:
= Q/Q
(8-13)
RAS
The fraction of nitrates not recycled but passed to the
second anoxic zone is then:
Fraction of nitrates passed to 2nd anoxic zone
= (1-/N03) (8-14)
The amount of nitrates removed in the endogenous de-
nitrification reactor can be expressed as a percent re-
moval of the total oxidized nitrogen load passed to the
second anoxic zone. For discussion purposes, the re-
moval from endogenous nitrate respiration expressed as
a decimal will be denoted as ER. The ER nitrate removed
is itself a function of the nitrate load that can be evaluated
from the mechanistic considerations introduced in Chap-
ter 4. The mass of nitrates removed in the second anoxic
reactor using these parameters may be expressed as:
Mass of nitrates removed in endogenous anoxic reactor
= [ER (1 - /N03)J (TKNOX) (8-15)
The fraction of nitrates passed to the aerobic reactor and
final clarifier is simply the difference from unity of the
nitrates passed through the endogenous denitrification
reactor, expressed as:
Fraction of nitrates passed to clarifier
= (1 -ER) (1 -/N03) (8-16)
The expression for nitrates passed to the effluent is then
the portion of remaining nitrates that are not recycled to
the anoxic reactor in the RAS. From expressions similar
to before, the nitrates recycled in the RAS may be ex-
pressed as a recycled fraction of the total flow to the
clarifier, or: .
Fraction of nitrates recycled in RAS
= [(1 - ER) (1 - /NOg)] [(RAS) / (1 + RAS)](8-17)
The fraction of nitrates passed to the effluent, /Ne, may
be calculated by a mass balance relationship as follows:
Nitrates passed to clarifier = nitrates recycled in RAS +
nitrates in final effluent
(1-ER)(1-/N03)
= [(1 - ER) (1 -/N03)] [(RAS) /(1 + RAS)] + /Ne (8-18)
262
-------
1
J
(fN03) (TKNOX)
1+l+RAS
TKNOX
d-
=R)
(ER) (1-/N03) (TKNOX)
t ^
(1-fN03)(TKNox) '//tf//, (1-ER)(1-fN03)(TKNox) \\\\\\\\\\^\\ (
^'/////J> I:::::::.:::::: [SeCOI
1+RAS W/W^ illiilliilliiii \ciar
"" \ 1 + RAS /
RAS
= /N03
+ QRAS +Q,
1+RAS+l
Nitrates passed to clarifier = Nitrates recycled + Ne
RAS and I expressed as fraction of Q,
i.e., I = 1 for 100% recycle
(
1
R) (1-fN03) (TKNOX)
Fraction removed = 1 -
(23 Anoxic Zone
OH;] Aerobic Zone
RAS
\
1+RAS )
Q
RAS
Q
= RAS
RAS
1+RAS)J
Figure 8-12. Bardenpho mass balance schematic.
Solving for /Ne:
/Ne = (1 - ER) (1 - /N03) {1 - [RAS / (1 + RAS)]}
(8-19)
Percent removals as a function of the internal recycle rate
(manifested in the /NO3 term) are displayed in Figure
8-13 for various endogenous removal rates and two RAS
rates.
Figure 8-13 demonstrates that under normal conditions,
the Bardenpho process can remove 83 percent of the
oxidized TKN if no endogenous denitrification is consid-
ered. If 50 percent of the nitrates to the second anoxic
zone are removed through endogenous nitrate respiration
(a conservative estimate), the nitrate removal perform-
ance increases to approximately 93 percent.
A similar analysis can be made for the Modified UCT
system, illustrated in Figure 8-14. The nitrified internal
recycle (I) and RAS recycle directly affect the nitrate re-
moval performance. Assuming all nitrates in the RAS
stream are removed in the first anoxic zone, the anoxic
recycle (IA) does not affect denitrification. The expression
for the fraction of nitrates recycled in the second anoxic
zone of the Modified UCT is the same as the expression
for the fraction removed in the first anoxic zone of the
Bardenpho process:
/ NO3 = I -s- (1 + RAS + I)
(8-20)
However, the amount of nitrates passed to the effluent
from the final clarifier in the Modified UCT is only affected
by the RAS recycle, since there is no additional denitrifi-
cation afforded by a second anoxic zone sequent to the
aerobic zone. This is evident mathematically if ER in the
Bardenpho process is set equal to zero, in which case
the Bardenpho would achieve the same level of perform-
ance as a single anoxic zone system.
Since denitrification in the Modified UCT occurs only in
the exogenous denitrification zones, the concentration of
nitrates in the discharge from the aerobic reactor will
remain constant. Thus, the mass of oxidized TKN in the
effluent (or that portion not recycled) can be calculated
by simply expressing the combined recycle (I + RAS) as
a fraction of the total nitrified flow (1 + I + RAS), as
provided in Equation 8-21:
Ne = {1 - [(I + RAS) / (1 + I + RAS)]} TKNOX (8-21)
The above expression may be algebraically manipulated
to verify its equivalence with the expression for Ne derived
in Figure 8-14.
Figure 8-15 demonstrates that the maximum feasible per-
cent nitrogen removal of the Modified UCT process is
<90 percent .
263
-------
100
Q.
40
30
Normal Bardenpho operating condition
I = 400%,RAS = 100%
0.1 0.2 0.3 0.4 0.5 0.6
Fraction of Nitrates Recycled to First Anoxic Zone,
0.7
Endogenous Removal Rate
(R) in Second Anoxic Zone
R=0 %
" R=50%
•• R=100%
0.8
Figure 8-13. Bardenpho process nitrate removal as a function of recycle rates and denitrification performance.
(TKNOX)
I = Q i /Q
1+RAS 1+l+RAS
(1-f-NOs) (TKNOX)
TKNOX
(1-fN03) (TKNOX)
[ Secondary)
1+RAS ~~\ Clarifier / N0
_RAS_\
1+RAS/
1+RAS+l
Nitrates passed to Clarifler » Nitrates recycled In RAS + N8
(1 - /N03 ) (TKNo*) - (1 - f NO 3) (TKNOX)
/Q
N,-(1-/N03)CTKNOX)
Fraction Removed n 1 - N< = 1 - {(1 - f NO3 )(1 -rrrr^ )}
r?n A , , (TKNOX) 1+RAS
[23 Anoxic Zone °a
ITT} AerobteZone
^2 Anaerobic Zone
Rgure 8*14. Modified UCT mass balance schematic.
264
-------
100
90
40
Maximum Feasible Nitrate Removal Percentage = 88.75 %
HAS Rate
%Q
50
100
150
200
300
100
150 200 250 300
Internal (Nitrified) Recycle Rate, % of Influent Q
350
400
Figure 8-15. Modified UCT process nitrate removal as a function of internal and return sludge recycle rates.
The theoretical mass balance necessary to optimize a
staged sequence of aerobic-anoxic zones with step feed
can be derived using similar concepts to those used for
recycle systems. A definition schematic is presented in
Figure 8-16. The influent wastewater mass load of TKN
and biodegradable COD is denoted with the subscript
zero. The mass of TKN and biodegradable COD to the
first aerobic basin is proportional to the step-feed ratio S,
such that:
SA TKNrj
SA COD0
(8-22)
(8-23)
where SA is the ratio of the flow diverted to the first anoxic
basin (volume/time) to the total process influent flow (Q).
In Figure 8-16, SA + SB +SC + SD = 1.0.
If the aeration system is sized to provide for oxidation of
the incoming oxidizable fraction of TKN,, the nitrate load
to the first anoxic reactor, NO3(,), can be expressed' sim-
ply as:
NO3(1) = TKN, - 0.015 COD, - Nref, (8-24)
where NO3(,j is the nitrate mass. The term 0.015 COD,
accounts for the nitrogen incorporated into cell mass, and
Nref, denotes the refractory nitrogen mass that was in-
cluded with TKN,.
The COD necessary for denitrification in the first anoxic
zone must be provided by step feeding. The mass of
biodegradable COD step-fed into the first anoxic zone is
given by:
COD2 = SB COD0
(8-25)
A fraction of TKN0 will also be present in the step feed.
This fraction is given by:
TKN2 = SB TKN0
(8-26)
Both TKN2 (including its refractory organic nitrogen com-
ponent, Nref2) and Nref1 will be passed to the second
aerobic zone.
The nitrates that are denitrified in the first anoxic zone
will consume 2.9 g of COD per g of nitrate denitrified
(Chapter 2). Thus, the influent wastewater biodegradable
COD passed to the second aerobic zone is:
COD3 = COD2 - 2.9 NO;
3(1)
(8-27)
The above expressions can be expanded to describe a
series of subsequent reactors as a function of the influent
wastewater TKN, biodegradable COD and step-feed ratios.
N03(2) = TKN3 - 0.015 COD3 - Nref3
= TKN2 - 0.015 COD3 - Nre{2 (8-28)
N03(3) = TKN5 - 0.015 COD5 - Nref5
= TKN4 - 0.015 COD5 - Nref4 (8-29)
COD4 = Sc COD0 (8-30)
265
-------
TKN0
COD,
Nref.
Nitrates formed - Influent TKN- TKN synthesized- TKN in effluent
TKN synthesized = 0.015 x biodegradable COD
Anoxfc Zone
Aerobic Zone
Figure 8-16. Multi-anoxic zone step-feed process.
CODS = COD4 - 2.9 NO3(2)
COD6 = SD COD0
TKN3 = TKN2 + Nre{1
TKN4 = Sc TKN0
TKN6 = TKN4 + Nref3
TKN6 = SD TKN0
Nref3 = Nrefi + Nref2
NrefS = Nre(3 + Nref4
(8-31)
(8-32)
(8-33)
(8-34)
(8-35)
(8-36)
(8-37)
(8-38)
If adequate COD and retention time are available to de-
nitrify all nitrate, the effluent will only contain TKN intro-
duced with the flow SD and Nraf5. These equations can
be modified to account for nitrates or ammonia in the
return sludge flow or to make further refinements in the
nitrogen balance due to biomass synthesis in each of the
reactor zones. A computation sheet containing the pre-
ceding equations can be used to optimize the step-feed
ratios and percent removals. The theoretical removals as
a function of the influent COD:TKN are presented in Fig-
ure 8-17.
Figure 8-17 demonstrates step-feed processes can theo-
retically achieve >90-percent removal if the COD:TKN
>6:1, The process can be further optimized by supple-
menting the last anoxic stage with methanol, or by pro-
viding a final endogenous reactor with post aeration.
8.2.3.4 Process Design Features
The Bardenpho process has been designed with plug
flow, CSTR, and oxidation ditch flow regimes. However,
the combined oxygen requirements of nitrification and
carbonaceous oxidation can cause oxygen depletion in a
plug flow aeration zone. The Bardenpho process incor-
porates many of the same process design features as the
100
95
•= 90
I
I
las
I
80
75
70'
4 4.5 5 5.5 6
6.5 7 7.5 8 8.5 9 9.5 10
COD:TKN
Figure 8-17. Theoretical percent nitrogen removal as a
function of COD:TKN for a triple anoxic zone process with
step feed.
266
-------
single anoxic zone processes. Design considerations
should include the use of baffles for compartments, pump
capacity for internal recycle requirements, and mixers in
the anoxic zone to ensure maximum contact of nitrates
and wastewater carbon with the microorganisms. The
Bardenpho process as a retrofit option was also deter-
mined to represent a viable option for existing plants that
have their permits revised requiring nutrient removal (9).
If -sufficient tank volume exists, modifications may only
require the installation of baffles and internal MLSS recy-
cles. Plants that are not currently nitrifying may also re-
quire increased aeration capacity.
A five-stage Bardenpho plant should be designed to by-
pass the anaerobic zone in the event of a shock hydraulic
or high DO load. For additional process operability and
control, a prefermentation tank can be provided; alterna-
tively, the anaerobic zone can be divided into compart-
ments with baffles.
As was described for single anoxic zone systems, moni-
toring of the reactors in the Bardenpho process is re-
quired to ensure optimization of process performance.
Suggested monitoring parameters and rationale are pro-
vided in Table 8-6.
8.2.4 Multi-anoxic Zones
8.2.4.1 Background and Process Description
Alternating aerobic and anoxic zones can be achieved in
a continuous-flow, activated sludge system by cycling the
aerators on and off. This type of intermittent or pulsed
aeration in an- activated sludge facility is termed cyclical
nitrogen removal (CNR). CNR processes can be most
effectively applied at existing plants that have revised
permits that impose nitrogen removal. Research, and de-
velopment of the CNR process has primarily been per-
formed at a few existing plants, requiring only minor
process modifications to convert to CNR. These modifi-
cations may be as minimal as installing baffles or timers
to cycle aeration equipment, but may include providing
internal recycle pumps and piping, or providing step-feed-
ing capability. Thus, potential cost savings can be ex-
pected by implementing a CNR process when compared
Table 8-6. Monitoring Requirements and Rationale for Bardenpho Reactors (9)
Four-Stage Process Five-Stage (Phosphorus Removal) Process
Parameter
Rationale
Parameter
Rationale
Anaerobic N/A
N/A
1st Anoxic DO
N03
Aerobic
IR rate
DO
Alkalinity, pH
2nd Anoxic NO3
DO
Will reduce denitrification
rate
Inadequate load reduces
amount of denitrification
Controls NO3 load
High DO may inhibit
denitrification rate; low
DO may inhibit nitrification
Nitrification consumes
alkalinity; may require pH
control
High nitrification in
aerobic zone may
overwhelm endogenous
denitrification capacity
resulting in NO3 in effluent
High DO will inhibit DO
endogenous denitrification
DO, Nitrates
Orthophosphates
DO
N03
IR rate
DO
Alkalinity, pH
N03
Presence of electron
acceptors will inhibit
fermentive organisms
Control to verify
phosphate release
Will reduce
denitrification rate
Inadequate load can
cause excess phosphate
release
Controls NO3 load
High DO' may inhibit
denitrification rate; low
DO may inhibit
nitrification
Nitrification consumes
alkalinity; may require
pH control
High nitrification in
aerobic zone may
overwhelm endogenous
denitrification capacity
resulting in NO3 in
effluent
High DO will inhibit
endogenous
denitrification
267
-------
with conversion to a proprietary nitrogen removal proc-
ess, if it is applicable.
One study of aeration cycling was conducted at the Blue
Plains Pilot Plant in Washington, DC (14). Although the
performance during the study was subject to upsets, 84
percent nitrogen removal was achieved. The F/M ratio
(BODg/MLVSS/d) was kept at 0.1, which was low enough
to permit a mixed culture for nitrification and denitrifica-
tion. Researchers achieved TN concentrations <7 mg/L
at the onset of the study in cold weather conditions.
Process upsets at Blue Plains were attributed to experi-
mentation with unsuccessful modifications that were later
discontinued and to the lack of experience associated
with the first attempt at the CNR technology. The process
was susceptible to sludge bulking, necessitating opera-
tional changes that reduced denitrification. The Blue
Plains facility had experienced bulking under normal op-
eration; therefore, the cycling of the aerators could not be
stated with certainty to be the causative factor. The deni-
trification process was interrupted by ceasing the aeration
cycling to eliminate the bulking organisms. The addition
of FeCI3 to primary influent was initiated in warm weather
to promote phosphorus removal, but FeCI3 also reduced
the COD:TKN from 10:1 to 8:1 causing a decrease in
denitrification. Despite these operational modifications
and upsets, the process did achieve TN effluent concen-
trations <8 mg/L and greater than 80-percent nitrogen
removal.
Denitrification was observed to be the limiting rate reac-
tion in summer months at Blue Plains. The nitrification
rate was not observed to be sensitive to pulsing aeration
and was observed to be relatively constant during the
aeration cycles. Denitrification rates were observed to
decrease, ostensibly as a result of depletion of the carbon
source. This effect could be surmounted by providing step
feeding during the anoxic cycle. The implication from
these studies to the design of cyclic operation confirms
the prudence of performing nitrification and denitrification
rate tests to determine cycling frequency and duration. It
was further suggested that long 9c's were necessary to
ensure a high ratio of nitrifiers to denitrifiers to optimize
CNR performance (12).
TN removals of 80 percent in summer and just under 80
percent in winter were achieved at the Owego, New York,
wastewater treatment plant (15). High 90, solids inventory
control, and high COD.TKN were determined to be the
key operational parameters. The process schematic for
the Owego facility is presented in Figure 8-18. Sub-
sequent investigations (16) at the Barnstable, Massachu-
setts, wastewater treatment plant corroborated the
Owego results.
An innovative alternating cyclical aeration process for ni-
trification-denitrification using countercurrent aeration is
known as the Schreiber process (Figure 8-19). The
RAS
15-HP Timer Controlled
Mechanical Aerators
E3 Anoxic Zone
Cj] Aerobic Zone
Figure 8-18. Town of Owego, NY, water pollution control
Rotating Bridge
Diffused Aerator
Influent
Effluent
RAS
-» WAS
Anoxic Zone
Aerobic Zone
Figure 8-19. Schreiber process.
Schreiber process achieves alternating anoxic-aerobic
zones within a single reactor by transferring air through
submerged diffusers attached to a rotating arm. The
mixed liquor typically rotates at a velocity less than the
moving bridge. The moving diffuser concept is intended
to prevent bubble rise in a common vertical path and to
prevent inducement of vertical currents. The manufac-
turer claims that this will maximize oxygen transfer by
completely dispersing bubbles within the mixed liquor and
increase the bubble detention time. Anoxic conditions can
be achieved in the zone in front of the moving diffusers,
while aerobic conditions exist in the zone immediately
after the diffusers pass by that zone. Alternatively by
using turbidity for process control, the single basin is
cycled through oxic, anoxic, and anaerobic conditions.
Mixing is maintained by the bridge rotation without aera-
tion in the anoxic and anaerobic phases.
8.2.4.2 Typical Design Criteria
CNR design incorporates similar considerations as single
anoxic zone processes. Aeration capacity, solids retention
time (SRT), solids inventory, and BOD.TKN are the most
important design parameters. Bypassing primary settling
to ensure a high COD.TKN for retrofit applications has
been'suggested (15); calculations should determine the
268
-------
adequacy of existing reactor basin volume, aeration ca-
pacity, and settling capacity. Design criteria are presented
in Table 8-7.
CNR can be used to nitrify and denitrify without the use
of an internal recycle. However, the capability to provide
internal recycle should be considered as a process option
or an on-demand basis.
8.2.4.3 Process Performance
CNR systems can consistently produce effluent TN con-
centrations <8 mg/L and >80 percent TN removal. Al-
*
Table 8-7. Cyclical Aeration Design Criteria
Parameter
F/M, g/BODj/g MLVSS/d
Aerator on, min
Cycle off, min
ec, d
COD-.TKN
Aerobic DO, mg/L
Anoxic DO, mg/L
MLSS, mg/L
CNR at Owego
0.06-0.13
15-45
15-30
13-32
10:1
1-1.5
<0.3
2,600-4,000
Schreiber
0.05
*
*
25
0.5-1.5
2,000-7,000
1 Load oriented with turbidity control.
though the CNR process has not been used or investi-
gated extensively, pilot and full-scale operating results are
presented in Table 8-8, along with full-scale operating
results for the Schreiber process.
The CNR system offers flexibility, but requires more op-
erator attention and expertise compared to other acti-
vated sludge modifications. Factors that introduce
complexity to the process are the monitoring of nitrate,
DO, and solids inventory, and adjustments in aeration
cycles and step feeding that may be required to optimize
nitrification and denitrification.
8.2.4.4 Process Design Features
Practical experience at full scale has suggested that the
best performance for a continuous-flow, nonproprietary
CNR system can be obtained using at least three basins
in series and is recommended for design applications
(15). The recommendation for a minimum of three basins
in series is predicated on the provision of step feeding to
the downstream basins. The CNR process with step feed
is analogous to the Miyaji process (7) discussed in Sec-
tion 8.2.3. The process performance considerations intro-
duced in Section 8.2.3.3 indicated that process
performance is enhanced by increasing the number of
reactors.
Existing plants that use a plug flow configuration can be
modified to achieve CNR by dividing the reactor into com-
Table 8-8. Cyclical Aeration Operating Results
Process
Location
Q, rii3/d
HRT.hr.
60, d.
TKN in, mg/L
TKN out, mg/L
NHJ-N in, mg/L
NHJ-N out, mg/L
Noxout, mg/L
Total N removed, %
COD:TKN
F/M, g BOD/g MLVSS/d
CNR (16)
Barnstable, MA
5,450
9
15
N/A
N/A
22.3a
'.3.2
3.0
77b
7.8°
0.08?
0.24e
CNR (15)
Owego, NY
1,820
13-16
20-24
39.9
3.6
26.2
1.4
4.8
80
10.5
0.089
CNR (12)
Blue Plains Wash., DC
N/A
10.1
22.2
21.3
2.2
N/A
1.0
3.0
76
9.3
0.089
*
Schreiber (17)
Clayton County, GA
8,970
N/A
N/A
24.5
1.4
16 ' •
0.5
2.4
,84.5
N/A
Schreiber (17)
Jackson, TN
31,260
N/A
47.7
16.9
3.0
13.3
1.2
3.3
63
*
a Primary effluent
b Based on influent NHJ-N.only. Actual percent removed is higher, based on TKN.
0 Ratio based on influent BOD to primary effluent TKN.
dWinter' . . .- '- -•-.....
8 Summer,
N/A = Data not available
269
-------
partments with baffles. The provision of step-feed capa-
bility will distribute the BOD load to each compartment
and ensure an exogenous carbon source for denitrifica-
tion during each aerator off-cycle. Step-feed inputs should
be located one-half to three-quarters of the distance from
the influent end of the tank. The organic carbon demand
In the downstream portions of the tank will vary primarily
with temperature. Rapid depletion of available carbon has
been observed in the upstream portion of the aeration
tank in summer months when warmer temperatures
cause increased growth rates. Though not always neces-
sary to meet discharge requirements in a CNR process,
higher return sludge pumping rates can improve perform-
ance. Internal recycle pumping of mixed liquor from the
effluent to the influent end of the aeration tank, or from
the point of the last anoxic cycle zone to the head end
of the aeration tank may also improve performance. In
many retrofit situations where the plant is operating below
design capacity, the existing return sludge pumps may
provide sufficient recycle flow (i.e., 100-200 percent of
plant influent) to meet effluent TN concentrations <8 mg/L
Step feeding may not be required in cold weather as BOD
depletion may not occur as rapidly in the head end of the
tank.
Aeration cycling can be accomplished at timed intervals
or by using set point signals from DO probes. Multiple
aerators in a single basin can have staggered operation
to avoid excessive power current draws. Timers should
have programmable control to permit aeration cycling cor-
responding to diurnal and workday/weekend variations.
Systems that use mechanical surface aerators may not
require submerged mixers during the anoxic phase. The
residual turbulence imparted by mechanical surface aera-
tors has been observed to suspend the floe to provide
sufficient liquid-solids contact for a 30-minute anoxic cy-
cle (15). Although solids separation may not be of concern
with 15-45 minute anoxic cycles and higher MLSS, low-
energy mixing may enhance substrate availability for the
biomass. Cycling of air in diffuser systems to different
compartments of the basin can be accomplished by using
electrically operated butterfly valves. Oxygen transfer in
excess of that required for complete nitrification during
the aerobic cycle should be avoided; at existing installa-
tions this will require careful monitoring to determine the
actual oxygen transfer capacity of the aeration system.
Excess aeration (DO >2 mg/L) will delay denitrification by
prolonging the lag period required for DO depletion. Sub-
surface mixers may be desirable in diffuser systems since
the residual turbulence with diffused air is less than that
provided by mechanical surface aerators.
Several unique concepts are incorporated in the design
of the Schreiber process (Figure 8-19). These include the
mixed liquor velocity, rotating bridge velocity, DO profile
(a transitional curve dependent on bridge location), and
the variation of DO concentration for subsequent bridge
revolutions. Since the rate of oxygen transfer demand
varies, automated systems are provided with the
Schreiber process. Types of programmable control that
can be provided include DO probes to cycle blowers on
and off or a patented O2 minimizer. The O2 minimizer is
designed to vary aeration in response to the turbidity of
a settled MLSS sample, which varies with the quantity of
oxygen supplied to satisfy biochemical and nitrogenous
oxygen demand.
8.2.5 Oxidation Ditches
8.2.5.1 Background and Process Description
Oxidation ditch processes include technologies that use
looped trenches that provide a continuous circulation path
for the wastewater. Aerators within the flow path simulta-
neously provide the necessary aeration of the waste
stream as well as fluid propulsion. This type of aeration
results in an aerobic zone for nitrification in that portion
of the loop immediately downstream of the aerators, while
immediately upstream of the aerators, oxygen deficient
conditions usually prevail, thereby providing anoxic con-
ditions for denitrification.
Oxidation ditch technology was pioneered in the late
1950s by the Dutch engineer Pasveer, whose purpose
was to develop an inexpensive wastewater treatment
process for small communities that required a minimum
of operator attention. The Pasveer ditch consisted of an
oval loop, in which flow was induced and aerated by a
rotating aerator. Over 3,000 ditches were installed based
on the Pasveer configuration. The foremost limitation of
Pasveer's process was the areal requirement to accom-
modate the reactor loop, particularly for large populations.
Subsequent studies in the late 1960s and early 1970s
surmounted these limitations by modifying the reactor
configurations.
The next significant oxidation ditch investigation after the
Pasveer effort was initiated in Austria (18). Treatment
improvement was attempted at the Vienna-Blumenthal,
Austria, plant by using multiple aerators within each loop
and then connecting the tandem loops in series instead
of in parallel. The process schematic is illustrated in Fig-
ure 8-20. Denitrification in the second loop was attempted
by reducing the aeration within the loop, thereby providing
a process configuration similar to the postdenitrification
Wuhrmann concept. Ironically, nitrification-denitrification
was observed to occur between the rotors of the first tank,
which receives the raw influent, because of higher respi-
ration rates afforded from the higher COD load. Sub-
sequent studies using oxidation ditches hypothesized that
denitrification could also be occurring at the interior por-
tions of the floe particle where an oxygen deficient con-
dition existed (1,19).
270
-------
Final Settling Tanks
Effluent
Influent
Figure 8-20. Vienna-Blumenthal Wastewater Treatment Plant.
Aeration Discs
Channel
4
x-*-"
3 "~
x* - . —
2 —
' 1 —
X N.
-^ , ^^.
•>- Flow
= _ ^ ^"N
/""NX
Q
CH Motors
f
t
Sedimentation Tank
Effluent
Plant Influent
Figure 8-21. Orbal oxidation ditch.
The areal requirement of the early oxidation ditch designs
was reduced with the development of larger, more efficient
aeration devices, and the development of deeper basins. In
addition, the flow path per unit area was increased by either
arranging the flow loops concentrically or by folding the flow
oval in half, as shown in Figures 8-21 and 8-22,
respectively. The former configuration became
trademarked as the Orbal Process, and the latter the
Carrousel System by their developers. The Orbal patent by
Envirex applies only to the aeration disk equipment itself
and not to the reactor configuration process flow train. The
license for the Carrousel System is held by Eimco Process
Equipment Company of Salt Lake City, Utah. The
capitalized form of .the words will be used-in this manual to
distinguish the trademarked systems ,.
The earliest reference to,the Carrousel process was
made in 1968 (2Q), when It was held as an innovative
solution to implementing ditch technology to serve large
populations. The Carrousel, process developers cited the
requirement for excessive surface area (and thus space),
an infeasible capital expenditure for numerous surface
271
-------
Influent
Aerator
Effluent
Figure 8-22. Carrousel oxidation ditch.
aerators, and the concomitant energy costs. The Car-
rousel design was developed to provide adequate aera-
tion and to maintain adequate flow velocities, while
simultaneously maximizing utilization of surface area. Li-
censed by El MOO since 1976, the Carrousel process is
a proprietary process in the United States which has been
Installed at numerous U.S. facilities and at over 500 sites
worldwide. The Carrousel process is widely used in Hol-
land (its place of origin) and to a lesser extent in Germany
and England.
The orbal process was developed primarily in South Af-
rica as another optional arrangement of oxidation ditch
channels that would maximize utility of land area and
minimize costs of aeration equipment and power, while
simultaneously providing alternating aerobic and anoxic
conditions. The orbal system consists of concentrically
arranged channels, as shown in Figure 8-21. Raw waste-
water or primary effluent is introduced into the innermost
or outermost channel and is conveyed to the other chan-
nels via ports at the bottom of the channel dividing walls.
The Installation date of the first orbal plant was not avail-
able from a review of the literature, but 50 orbal plants
were In operation by 1972 in South Africa (21).
The nitrification-denitrification option of the patented Or-
bal process by Envirex is termed the Sim-Pre process,
schematically illustrated in Figure 8-23. Sim-Pre is an
acronym for simultaneous nitrification-denitrification/pre-
denitrification. The Sim-Pre process incorporates an in-
ternal recycle from the innermost to the outermost
channel. This aspect of the process represents predeni-
trificatlon. The simultaneous nitrification-denitrification
phase occurs in the first aeration channel. Since the aera-
tion demand exceeds the supply, anoxic conditions are
attained along the flow path in the area upstream of the
aerators. Thus, the outer channel of the Sim-Pre process
is operated analogous to a conventional oxidation ditch.
I. Kruger of Denmark, represented in the United States
by I. Kruger, Inc., in conjunction with the University of
Denmark, has developed two ditch-type processes for
nitrogen removal. These employ multiple ditches of large
total volume and flexible, simple timing controls of rotors
(mixing and aeration) and weir levels. The patented
BioDenitro process is designed to remove nitrogen, while
the BioDenipho is a modification to remove phosphorus
and nitrogen. The basic BioDenitro configuration consists
of two identical aeration ditch tanks and a clarifier. The
process can be applied to any type of complete mix aera-
tion tank system, as long as both aeration (e.g., brush
aerators or diffused air) and mixing equipment (typical
slow-speed propeller mixer) are installed. The separate
provision of mixing and aeration thus allows tanks or
ditches with high-sidewall depths, e.g., 4.6 m (15 ft).
However, instead of creating anoxic and aerobic zones
within each tank as in a conventional oxidation ditch, the
BioDenitro process achieves aerobic or anoxic conditions
alternately within each looped reactor. The wastewater
feed is alternated between the two tanks to provide a
carbon donor source for the desired microbial reactions.
Two designs of the BioDenitro plant are available. The
type DE plant consists of two oxidation ditches and a final
clarifier. The operating sequence is shown in Figure 8-24.
This four-phase alternating influent contact operation is
an application of a process first introduced by Christensen
(3). The BioDenipho process for phosphorus removal can
272
-------
Influent
(Q) J
Nitrified Recycle (400 % Q) Predenitrification
Anoxic (Large Volume) and
Aerobic (Small Volume) Zones
Simultaneous Denitrification
Final
Effluent
I Clarifier /
1 st Aeration
Channel
(65% of Total
Reactor Volume)
2nd Aeration
Channel
(23% of Total
Reactor Volume)
RAS(100%Q)
3rd Aeration
Channel
(12% of Total
Reactor Volume)
WAS
I22 Anoxic Zone
Q Aerobic Zone
Figure 8-23. Orbal Sim-Pre process.
Influent
Return Sludge
Phase A
Effluent
Influent
Return Sludge
Phase B
Effluent
Influent
Return Sludge
Phase C
Effluent
Q
Return Sludge
Phase D
Phases A & C (Main Operational Phases): 60-90 min.
Phases B & D (Intermediate Operational Phases): 15-30 min.
Effluent
Rotor Mixing
Aeration
N = Nitrification
DN = Denitrification (Anoxic)
S = Sedimentation
Figure 8-24. Kruger BioDenitro process (Type DE).
273
-------
be provided by including an anaerobic phase to the
BioDenitro process.
The second BioDenitro process design uses the T-ditch
configuration in which first and third oxidation ditches
primarily provide denitrification and settling; the middle
oxidation ditch always serves as an aeration and flow
distribution unit. The BioDenitro T process is a six-phase
cycle and is schematically illustrated in Figure 8-25. This
system is very much a sequencing batch reactor of
unique design and operating features. As shown in Fig-
ures 8-24 and 8-25, neither the T nor the DE process
uses an internal recycle. The T process does not use an
RAS recycle since separate clarifiers are not provided.
8.2.5,2 Typical Design Criteria
Typical design criteria are unavailable for oxidation
ditches since their designs vary. However, criteria used
from existing plants are presented in Table 8-9.
Design for denitrification in oxidation ditches is similar to
other anoxic reactor designs. Methods of computing the
nitrogen available for nitrification and denitrification were
introduced in Section 8.2.2; these are applicable for oxi-
dation ditch designs as well.
Oxidation ditch technologies are generally operated in the
extended aeration mode, with long hydraulic and solids
retention times and higher MLSS than are commonly
used in activated sludge plants. Conceptually, the oxida-
tion ditch configuration is an endless channel. Only a
portion of the mixed liquor is withdrawn in each cycle,
providing a high internal recycle ratio. The orbal configu-
ration incorporates properties similar to plug flow, de-
pending on the transfer and distribution method of mixed
liquor between successive channels.
The size of the ditch should be determined based on
maintaining channel velocities of 0.3-0.6 m/s (1-2 fps),
or a loop flow time of 10-45 min. The design of the
channels can vary, but will incorporate considerations
used in the design of single anoxic zone systems, such
as sludge concentrations, F/M ratios, RAS rate, tempera-
ture extremes, and desired effluent quality. The F/M ratio
has been suggested as the critical design parameter (21),
Effluent
/"
r"
"\
hi t
f\
Mill
K
^\
HIM
i i
/T
e
r\
i
j
Phase
Phase
Phases A & D (Main Operational Phases): 60-120 minutes
Phases B, C, E & F (Intermediate Operational Phases): 30-60 minutes
Cycle Duration: 4-8 hours
Rotor Stopped
DN Denitrificatton-Anoxic
N • Nitrification-Aerobic
S Sedimentation
Mixing
4H4f Aeration
Rgure 8-25. Kruger BioDenitro process (Type,T).
274
-------
Table 8-9. Operating Parameters at Various Oxidation Ditches
Type
Location
Q, m3/d
MISS, mg/L
F/M, g BODg/g
MLSS/d
Oxidation
Ditch (22)
Vienna-
Blumenthal,
Austria
41,290
5,800
0.17
Orbal
(23)
189
4,300
0.22a
Orbal
(21)
, Modder-
fontein,
S. Africa
2,385
3,030
0.093
Orbal
(21)
S. Wit-
bank,
S.Africa
167
8,830
0.03
Orbal (1)
Extended
Aeration
Huntsville,
Texas
98
8,000
0.027b
Orbal (1)
Split Feed
Huntsville,
Texas
114
8,800
0.01 5b
Oxidation
Ditch (19)
Carrol-
wood,
Florida
13,250
3,060
NA
BioDenitro
DE
Frederiks-
sund,
Denmark
6,000
3,000-
5,500e
0.08e
Bio-
Denitro
T
Odense,
Denmark
15,000
3,000e
RAS, %Q
9C, d
HRT.hr
Volumetric loading,
g BOD/m3/d
Ib BOD/1 ,000 cuft/d
190
7 •
7
985
61
200
144-240
9-15
115
15
11
416
26
31
194-226
15-18
110
>50
33
120
8
95
>50
28
90
6
—
44
17
150°
9d
30-80"
15-309
14
707
44
NA
15-30"
22
282
18
JCOD/MLSS b g BODs/g MLVSS/d °g COD/m3/d d Ib COD/1,000 cu ft/d e Design value
and can be used as a design basis if the minimum sludge
concentration is known.
The required aeration capacity is difficult to ascertain ac-
curately during design. Factors that affect aeration re-
quirements include wastewater characteristics, flow
variations, temperature, and RAS rate. Further design
considerations, such as sizing aerobic and anaerobic or
anoxic zones for TN removal, are complex and difficult to
predict. However, a suggested approach would be to de-
sign for complete nitrification between the aerators. By
varying the oxygen transfer and monitoring the DO profile
in the channel, the operating nitrification rate can be as-
certained for the actual COD and temperature of the
wastewater. The oxygen transfer rate can then be further
adjusted to achieve oxygen deficient conditions down-
stream of the aerators. Several researchers have sug-
gested that nitrification and denitrification can best be
obtained by specifying the required anoxic volume and
then monitoring and controlling the DO profiles (24,25).
Also, flow and load equalization has been suggested as
a means to reduce shock loads, or in the case of Orbal
configurations, enhanced denitrification has been re-
ported by step-feeding influent to the inner channels (1).
The use of temporal sequencing in the design of the
Kruger BioDenitro process is unique. As shown in Figures
8-24 and 8-25, this method of operation allows operator
flexibility to optimize the anoxic-aerobic volumes by vary-
ing the time duration of the specific phases. During anoxic
phase, the aeration is stopped and the sludge is main-
tained in suspension by submerged propeller mixers.
Thus no oxygen is introduced while the anoxic zone is
created. Similarly, the aeration equipment is operated to
establish a defined aerobic volume. Typically, the stand-
ard BioDenitro type DE process volume is proportioned
to 57-67 percent for nitrification and 33-43 percent for
denitrification. The denitrification volume can be changed
to accommodate fluctuations in flow, level, or temperature
simply by changing the duration of the cycle. Further
flexibility may be obtained by adding the third aerobic
ditch of the T process. The third ditch ensures that all
mixed liquor is aerated before entering the final clarifiers
and permits the main ditches to increase the portion of
time allotted for denitrification.
Design of an Orbal Sim-Pre process follows the standard
procedure used by Envirex for the design of its Orbal
process. Denitrification is afforded by the internal recy-
cling of mixed liquor from the innermost to the outermost
channel. The Orbal process design by Envirex uses a
single factor of 0.93 to account for a, (3, temperature, and
elevation. Each of the three channels are designed to
achieve different DO levels. The suppliers of the Sim-Pre
process recommend that the channels should be de-
signed for 0, 1, and 2 mg/L DO for the outer, middle, and
inner channels, respectively, although the actual DO will
vary widely along the channel. The channel'volume split
in the Sim-Pre mode is generally 65/23/12 for outer, mid-
dle, arjd inner channels, corresponding to a base oxygen
demand split of 70/18/12, respectively. Thus, for a total
O2 requirement of 908 kg (2,000 Ib) O2/d, the correspond-
ing oxygen delivery to each channel would be 590, 209,
275
-------
and 109 kg (1,300, 460, and 240 Ib) O2/d. With these
requirements, the number of discs, immersion, and rota-
tional speed can be selected.
Typical design criteria for the Orbal process are presented
in Table 8-10.
8.2.5.3 Process Performance
Oxidation ditch designs vary and, consequently, oxidation
ditch performance also varies. Performance results from
a number of facilities are presented in Table 8-11. Nitro-
gen removals range from 65 to 97 percent. The perform-
ance results reported by one manufacturer (26) indicate
that removals as high as 90 percent have been attained.
In certain cases, Envirex has guaranteed process effluent
total nitrogen of 3 mg/L.
The Kruger process incorporates a unique feature of flexi-
bility due to its timed alternating aeration and mixing se-
quences, incorporating aspects of an SBR. However, it
differs from many SBRs in that the Kruger process pro-
vides continuous influent and effluent flows. The phase
lengths may be varied to achieve specific treatment ob-
jectives, or to respond to variations in load, flow, or tem-
perature. Kruger claims that the operator can vary the
operation of the plant within two minutes in response to
an impetus from the control room. The process can be
provided with a sophisticated monitoring and operation
Table 8-10. Design Parameters for Orbal and Orbal Sim-Pre Process
Row, rr»3/d BOD Load, g/m3 MLSS, mg/L 9C, d
HRT, hr
Depth, m
<760
760-1,889
1,890-3,784
3,785-7,569
>7,570
Table 8-11.
•type
Location
200
240
240-288
288
320
Nitrogen Removal
Oxidation
Ditch (22)
Vlenna-
Blumenthal,
Austria
Performance
Orbal
(21)
4,000-5,000
4,000-5,000
4,000-5,000
4,000-5,000
5,000-6,000
31-38
26-32
20-32
21-27
24-29
for Various Oxidation Ditch-Type Plants
Orbal Orbal Step Oxidation
(21) Feed(1) Ditch (19)
South Modder-
Wltbank, fontein, Huntsville,
South Africa South Africa TX
Carrolwood,
FL
24
20
16.6-20
16.6
15
Oxidation
Ditch (19)
Frankfort,
KY
Kruger T
(27)
Faaborg,
Denmark
1 '2-2.4
1,5-2.4
1.8-3.0
2.4-3.7
2.4-3.7
' • V
Kruger DE
(27)
Frederiks-
sund,
Denmark
FLOW
m3/d(mgd) 41,256(10.9)227(0.06) 2,271(0.6) 2,271(0.6) 13,248(3.5) 14,383(3.8)15,000(3.96)8,327(2.2)
BOD (COD), mg/L *'"[,
Influent 245 319 200 168 (250) (205) ; 300 "}
Effluent 12 35 6 (20) (23) 6 " 9
TKN, mg/L
Influent 30 52.3 34 19.4 25 16.8 ' 36" "if
Effluent 3.1 8.4 10 0.7 0.6 1.1 ' -;
BOD(COD):TKN
Influent 6.1 ' 5.9 8.7 (10.0) (12.2) 8.3 ,
NHJ-N, mg/L :
Influent 17.9 39.2 21 17.8
Effluent 3.6 6.7 7.3 0.6 0.3 4.8 0.5
NOX-N, mg/L
Effluent 0.9 0.7 2.2 1.1 0.04 2.5 '4.3 '.,,'lJ?
Total N, mg/L
Effluent 4 9.1, 12.2* 1.8 : 0.64 3.6' ,9.1 - 3.5 S
%N removal 87 86 '65 91 97 79 80 90 '
276
-------
PC-based control system. For example, the plant can
have, in the PLC or timer control, preprogrammed/preset
operational modes for the weekend, weekdays, summer,
high-load, etc. The operator would have the discretion to
activate these operating sequences.
Oxidation ditch technologies were developed to accentu-
ate simplicity, minimize operator control, and provide cost
and energy savings. The ditch configuration is the most
energy efficient design available because it conserves
fluid momentum. This flow regime augments stability, con-
sistency, and energy efficiency (27). The key operational
parameters in an oxidation ditch are the oxygen transfer
rate and MLSjS. The MLSS may be controlled by the waste
activated sludge (WAS) rates from the clarifier (or from the
ditch for Kruger T). In a single oxidation ditch loop, aeration
can be optimized by placing the aerators at selected points
along the flow path. Orbal configurations can incorporate
features of a compartmented plug flow reactor by main-
taining individual loops at a desired average DO level. For
example, the outermost channel can be maintained at
aerobic conditions, while the second channel is main-
tained at anoxic conditions. Nitrates formed in the aerobic
channel will pass to the second (anoxic) channel, where
denitrification can occur. Step-feeding raw wastewater or
influent to the anoxic channel will ensure a carbon source
to promote substrate nitrogen respiration. If active recy-
cling of mixed liquor between channels is not provided,
the oxygen transfer capacity is the most direct means to
control nitrate loads. Orbal plants can reduce aeration
equipment costs by extending a single aerator shaft
across all the channels. Oxygen transfer efficiency and
capacity may be adjusted in Kruger, Orbal, and conven-
tional single-loop oxidation ditches by adjusting the sub-
mergence of the aeration discs or rotors, varying the
rotational speed, or changing the number of aerators on
each shaft in the channel. Cycling of the aerators can also
be used, as long as sufficient alternate impetus is provided
to propel the mixed liquor in the channel.
The Orbal configuration also is conducive to split-feed
options and has brought about enhanced nitrogen remov-
als. The following hypotheses were offered to explain the
efficacy of split-feed approaches (1):
1. Higher DO levels and enhanced nitrification were
achieved by lowering the organic load and oxygen
demand in the first channel. :
2. Raw wastewater addition into the second channel low-
ered the oxygen concentration in the channel and
extended the denitrification zone.
3. Raw wastewater addition to the second channel pro-
vided an exogenous carbon source to promote deni-
. trification. -
8.2.5.4 Process Design Features
Oxidation ditches are a hybrid flow regime and incorpo-
rate aspects of plug flow and a CSTR. The concentric
channels of Orbal configurations have been mathemati-
cally proven to approach an ideal plug flow reactor
(21,28). As the wastewater recycle (i.e., the average num-
ber of passes around the loop) is increased, the system
was also determined to approach an ideal CSTR. At long
retention times (extended aeration), the limiting nutrient
concentration is lower in a plug flow reactor. Although
superior effluent quality may be realized as a result of the
advantages of a plug flow regime, the provision of equali-
zation facilities for plug flow regimes has been suggested
because of their lesser ability to accommodate shock
loads. .
As discussed, alternating oxygen levels can be provided
by varying the aeration capacity or by installing concentric
channels in the Orbal configuration and maintaining the
desired condition within each channel. The Orbal ar-
rangement can be further modified to promote denitrifica-
tion by providing step-feed capability. Additional influent
feed points may also be included in a single oxidation
ditch loop in an attempt to ensure an exogenous carbon
source.
Several investigators have commented on the suscepti-
bility of oxidation ditch technologies to poor settling char-
acteristics (23,29) which were postulated to be
exacerbated by excessive rotational energy input by the
aeration mechanism, causing floe shear. Poor settling
characteristics were also,attributed to placement of an
anoxic zone near the effluent withdrawal for the clarifiers.
Presumably, the settling was hindered by entrapment of
nitrogen gas. in the floe, causing a floating sludge. This
condition was corrected by ensuring that the ditch effluent
withdrawal point was preceded by an aerated zone.
Aeration of oxidation ditches is also a critical operational
parameter. Continuous monitoring of the DO is required
to develop an oxygen profile that provides anoxic and
aerobic zones. The oxidation ditch can be automated by
using DO probes to coordinate aerator cycles, weir mo-
tors, or a combination of the two. To a certain extent,
ditches are somewhat self- regulating in the case of hy-
draulic load increases. As the water surface of the.ditch
rises, the rotors will be more submerged, thereby impart-
ing a higher oxygen transfer to the waste. This feature
can be modulated mechanically if desired.
8.2.6 Sequencing Batch Reactors
8.2.6.1 Background and Process Description :
An SBR, as it is commonly referred to today, is a fill-and-
draw, variable reactor volume technology. The prototype
for the activated sludge concept (30) was developed on
. a fill-and-draw( basis. Shortly, after that initial study, the
277
-------
emphasis switched to continuous flow "conventional" ac-
tivated sludge. Further developments with SBR technol-
ogy were not pursued because of limitations of equipment
and engineering experience—namely the ability to dissi-
pate the hydraulic energy of the effluent discharge, op-
erator attention, and expertise, and the susceptibility of
clogging of the air diffusers during the settling period.
Recent innovations in aeration devices, control logic, level
sensors, solenoids, and hydraulic energy dissipators
have surmounted these limitations and revitalized interest
in SBR technology. The resurgence of interest in SBRs
was initially limited to small treatment applications; how-
ever, the need for greater treatment efficiencies due to
increasingly stringent effluent limits has resulted in the
adoption of SBR technology in installations as large as
660 Us (15 mgd) (31).
The SBR consists of a self-contained treatment system
incorporating equalization, aeration, anoxic reaction, and
clarification within one basin. Intermittently fed SBRs con-
sist of the following basic steps:
1. Fill—The fill operation consists of adding the waste
and substrate for microbial activity. The fill cycle can
be controlled by float switches to a designated volume
or by timers for multireactor systems. A simple and
commonly applied mode to control the fill cycle is
based on reactor volume, resulting in fill times in-
versely related to influent flow rates. The fill phase can
include many phases of operation and is subject to
various modes of control, termed static fill, mixed fill,
and react fill. Static fill involves the introduction of
waste influent with no mixing or aeration. This type of
fill method is most common in plants requiring nutrient
control. In such applications, the static fill will be ac-
companied by a mixed fill stage such that the micro-
organisms are exposed to sufficient substrate, while
maintaining anoxic or anaerobic conditions. Both mix-
ing and aeration are provided in the react fill stage.
The system may alternate among static fill, mixed fill,
and react fill throughout the fill cycle.
2. React—The purpose of the react stage is to complete
reactions initiated during fill. The react stage may be
comprised of mixing or,aeration, or both. As was the
case in the fill cycle, desired processes may require
alternating cycles of aeration. The length of the react
phase may be controlled by timers, by liquid level
controls in a multitank system, or when the desired
degree of treatment has been attained, verified by
monitoring of reactor contents. Depending upon the
amount and timing of aeration during fill, there may or
may not be a dedicated react phase.
3. Settle—Liquid-solid separation occurs during the settle
phase, analogous to the operation of a conventional final
clarifler. Settling in an SBR can demonstrate higher ef-
ficiencies than a continuous-flow settler, since total
quiescence is achieved in an SBR.
4. Draw—Clarified effluent is decanted in the draw
phase. Decanting can be achieved by various appa-
ratus, the most common being floating or adjustable
weirs. The decanting capability is one of the opera-
tional and equipment limitations of SBR technology.
Adaptation or development of equipment compatible
with a fluctuating liquid level is required.
5. Idle—The final phase is termed the idle phase and is
only used in multibasin applications. The time spent in
the idle phase will depend on the time required for the
preceding basin to complete its fill cycle. Sludge wast-
age will typically be performed during the idle phase.
A typical SBR process sequence schematic is shown in
Figure 8-26.
Denitrification can occur during the fill or react stages by
cycling the aerators, and during the settle and draw pe-
riod. An obvious advantage of an SBR systems with low
flows is that the reactor contents can be retained until the
desired level of treatment is achieved, providing that suf-
Typical
Percent of
Cycle Time |nf|uen,
25
Fill
Operation
Anoxic/Aenobic
Cycles
35
React
Anoxic/Aerobic
Cycles
20
Settle
Aerators/Mixers Off
20
Draw
Effluent
Aerators/Mixers Off
WAS
Figure 8-26. Sequencing batch reactor.
278
-------
ficient tankage exists to equalize or accommodate the
additional influent.
Several proprietary process and equipment innovations
have been developed to enhance treatment, simplify op-
eration, or control sludge characteristics. All proprietary
SBR manufacturers will guarantee TN effluent concentra-
tions <5 mg/L To illustrate the variety of options available,
the proprietary aspects of five SBR manufacturers are
discussed below.
• Aqua SBR—Jhe Aqua SBR system provided by Aqua-
Aerobic Systems, Inc., is not a patented process, but
the. process does include a proprietary floating direct
drive mixer, an effluent decanter, and a microprocessor
control system. The floating decanter is designed to
prohibit MLSS from entering the decanter during mixed
or react phases, and it also withdraws supernate 30
cm (0.5 ft) below the water surface to mitigate scum
losses to the effluent. If long settling times are pro-
vided, clear effluent can be obtained at high SVIs.
• Omniflo—Jet Tech, Inc., has developed SBR equipment
and also has a patented logic control for their aeration
system. The proprietary equipment includes dry pit
pumps! headers, manifolds, influent distribution hard-
ware, jet aerators, and decanter apparatus. A proprietary
aspect of the SBR process provided by Jet Tech is the
Batch Proportional Aeration System. The function of
this aeration system is to relate the volumetric change
rate during the fill phase to the aeration capacity re-
quirements by sensing the DO level in the reactor,
optimizing nitrification and denitrification cycles.
• Fluidyne—Jhe Fluidyne Corp. offers a system with ef-
fluent decanters fixed in position to the reactor wall.
The device excludes MLSS entry during aeration.
These systems also commonly employ jet aeration
with a combination of aeration and static conditions
during fill.
• CASS—The Cyclic Activated Sludge System (CASS)
was developed and is marketed by Transenviro, Inc.
CASS uses a similar sequence of operation as other
batch technologies, but is configured with a proprietary
captive selector reactor. The selector can also receive
continuous flow. The selector is a baffled compartment
that receives raw wastewater or primary effluent where
it is mixed with RAS or internally recycled MLSS. The
selector then conveys flow to the reactor basin. By
limiting or eliminating aeration to the selector, oxygen
deficient conditions can be attained, while concurrent
high substrate levels are maintained. This mode of
operation is claimed to favor the propagation of floe
.formers and to inhibit growth of filamentous strains
(32). A process schematic is presented in Figure 8-27.
Influent
Fill
Influent Aerators/Mixers On
t
RAS (20 % Q)
Influent Aerators/Mixers Off
RAS (20 % Q)
Aerators/Mixers Off
#
Decant
RAS(20%Q)
Figure 8-27. Cyclical Activated Sludge System.
1
RAS (20 % Q)
Influent
^
/
/
/
Aer<
ators/Mixers O
Settle
T
RAS (20 % Q)
E2 Anoxic Zone
O Aerobic Zone
S Settled Sludge
279
-------
ICEAS—A modified batch system is available from
Austgen-Biojet (ABJ). The ABJ system is termed Inter-
mittent Cycle Extended Aeration System (ICEAS) and
is depicted schematically in Figure 8-28. The distin-
guishing features of ICEAS is that continuous inflow is
incorporated in all phases, compared to other variable
volume processes that do not receive continuous in-
flow. Noncontinuous inflow operation can be provided,
if requested. Austgen-Biojet maintains that the continu-
ous inflow mode is preferable to noncontinuous flow
operation, as the distribution box used by ABJ will
ensure that variations in load and flow are distributed
evenly between the reactors and prevent diurnal vari-
ations or shock loads from continually overloading one
reactor. The manufacturer asserts an additional advan-
tage of the ICEAS flow regime is that continuous flow
via the distribution box reduces the valving and head-
works engineering compared to requirements for a
noncontinuous flow SBR. A complete ICEAS treatment
cycle consists of three phases: aeration, settle, and
draw. Since influent is received during all phases,
ICEAS does not offer total quiescence during the settle
phase, a characteristic of an intermittently fed SBR.
Although ICEAS is proprietary, no royalty or license
fees are imposed. ICEAS uses a patented anoxic se-
lector to provide denitrification and to promote growth
Influent
Operation
Anoxlc/Aerobic
Cycles
Anoxic/Aerobic
Cycles
Influent
Settle
Aerators/Mixers Off
Influent
Draw
Effluent
Aerators/Mixers Off
Figure 8-28. Intermittent Cycle Extended Aeration System.
of zoogleal microorganism, and to inhibit filamentous
strains. The ABJ selector has characteristics similar to
the patented CASS selector, but ABJ claims to be the
developer of the original selector concept.
As noted in the previous section, some Kruger ditch con-
figurations (Figure 8-25) are variations of the SBR con-
cept in a unique looped reactor design.
8.2.6.2 Typical Design Criteria
A unified approach to SBR technology has yet to be
developed (33); however, the principles used to design
nitrification-denitrification facilities in single anoxic or dual
anoxic zone systems, such as flow and loadings, may be
applied with some modifications. One factor to consider
.specifically for the design of an SBR is the flow volume
which will determine whether one reactor will suffice (gen-
erally for flows <2 Us [0.05 mgd]) or whether a two-vessel
system is required. Additional vessels should be consid-
ered for sites that experience a wide transient variation
in either organic or hydraulic loading. Conditions, includ-
ing wet weather with ingress of surface or ground waters,
may be accommodated by effecting more frequent decant
cycles, without causing washout of the reactor biomass.
The SBR process can accommodate peak hourly flows
3-10 times as large as the design flow without adverse
effects, if excess capacity is available. The F/M ratio must
be determined by the desired effluent quality which in turn
dictates reactor sizing.
The critical operational feature is the cycle time for fill,
react, settle, and draw, and the amount of oxygen that is
supplied. A typical cycle for an intermittent-feed, intermit-
tent-discharge SBR based on average flow conditions is
four-hour duration; two hours allocated to fill/aeration/an-
oxic react, one hour to settling, and one hour to decant
and idle. The total time for a batch cycle consists of the
time allowed for each component phase. Design cycle
times in full-scale plants have varied from 2 to 24 hours
(34). A suggested strategy is presented in Figure 8-29.
Some typical design criteria are presented in Table 8-12.
SBR reactors have been constructed with a variety of
shapes including rectangular, oval, circular, and with
sloped sidewalls. Design bottom water levels after decant
are typically 3-4 m (10-13 ft) and design top water levels
are typically 4.3-5.5 m (14-18 ft). A freeboard of 1 m (3
ft) is common. The design mixed liquor volume can be
calculated from the selected MLSS concentration, which
decreases throughout the fill cycle. The MLSS concen-
tration at the end of the draw phase is that of a settled
mixed liquor and is similar to that in a conventional clari-
fier underflow (35). Once the tank volumes have been
calculated, the cycle times can be determined. If the cycle
times are unsatisfactory, the tank volumes can be ad-
justed accordingly.
280
-------
Hours
FM I FMR,
BOD and SS Removal
FM i FMR
A/AX
BOD, SS, and N Removal
AX
AX
ICEAS Process
Austgen-Biojet
F-Fill
FM - Mixed Fill
FMR - Aerated Mixed Fill
R - React
S - Settle
D - Decant
I - Idle
A - Aerobic
AX - Anoxic
Figure 8-29. Suggested operating strategies for SBR systems.
Table 8-12. Typical Design Criteria for Sequencing
Batch Reactors
Parameter
BOD load, g/d/m3
Cycle time, hr
Fill (aeration)
Settle
Draw
MLSS, mg/L
MLVSS, mg/L
HRT, hr
60, d
F/M,'gBOD/gMLVSS/d
SBR ICEAS
80-240
1-3
0.7-1
0.5-1.5
2,300-5,000
1,500-3,500
15-40 36-50
20-40 —
0.05-0.20 0.04-0.06
The sizing of aeration equipment is done according to
criteria for complete nitrification and BOD removal, except
that the required oxygen transfer must be accomplished
in a shorter period. The actual amount of aeration time
per cycle must be considered when sizing the aeration
equipment.
8.2.6.3 Process Performance
Nitrogen removal in SBR systems can be considerably
higher than in conventional activated sludge systems.
Performance results from full-scale facilities are provided
in Table 8-13. Available data on full-scale SBR facilities
detailing nitrogen removal are limited, because of the rela-
tively recent application of this technology.
One of the primary features of SBR technology is the
flexibility to exercise control as a function of time rather
than space (as in conventional flow-through systems).
Several key aspects include:
• The SBR system can tolerate shock loads and peak
flows because of the equalizing basin characteristics
of the fill phase.
• Periodic effluent discharge may permit retention of re:
actor contents until desired clarity or treatment quality
is achieved.
• A fraction of the total volume may be used during low
flow periods, resulting in lower aeration requirements.
If aerators or blowers have turn-down capability, O&M
costs may be reduced.
• No RAS or internal recycles are required; however,
some systems (e.g., CASS) include recycle to an an-
tecedent basin or selector chamber.
• With intermittently fed SBRs, clarification occurs under
total quiescence, thereby eliminating short-circuiting.
Consequently, small floes will settle in an SBR that
would be washed out in a continuous-flow regime.
• Filamentous growth can be controlled by operational
strategies along with adjustments during the fill phase.
SBR operation is somewhat more sophisticated than
other systems of comparable size. However, the advent
of reliable automation of liquid level sensors with pro-
grammable logic controllers and decanting devices sig-
nificantly simplifies operation. Particular attention to
process performance and monitoring is required to opti-
mize the performance and to determine the optimum
aeration cycle frequency.
281
-------
Table 8-13. Summary of SBR Plant Operating Data (36)
Influent
Influent TKN
Effluent
TKN Influent Effluent Effluent Effluent %N
Plant
Nonproprietary
Culver, IN
Cass
Deep River, CT
Cass
Dundee, Ml
Nonproprietary
Grundy Center, IA
Aqua SBR
Grundy Center, IA
Aqua SBR
Rock Falls, IN
Aqua SBR
Oak Hill, Ml
Jet Tech
Oak Pt., Ml
Jet Tech
Cow Creek, OK
Jet Tech
Del City, OK
ICEAS
Buckingham, PA
ICEAS
Burkeville, VA
ICEAS
Shiga Kogen
Flow
m3/d, mgd
N/A
189 (0.05)
N/A
1,249(0.33)
3,028 (0.8)
530 (0.14)
416 (0.11)
227 (0.06)
9,841 (2.6)
13,248 (3.5)
492 (0.13)
530 (0.14)
757 (0.2)
BOD5,
mg/L
170
100
123
210
140
109
220'
142
119
115
349
296
484
(Total N),
mg/L
N/A
54.5
28.9
N/A
28.0
39.8
N/A
N/A
24.0
(28.3)
N/A '
35.7
(36.9)
(Total N),
mg/L
N/A
3.6
2.2
N/A
4.4
1.8
N/A
N/A
2.7
(5.4)
N/A
3.6
(5.4)
NHJ-N,
mg/L
20.0
40.4
16.9
17.3
19.0
35.9
25.0
19.0
17.0
17.6
29.2
19.3
N/A
NHJ-N,
mg/L
1.0
1.3
0.5
0.8
1.6
0.6
0.6
0.6
1.8
0.9
0.6
0.3
N/A
NOX-N,
mg/L
N/A
1.0
4.9
2.8
0.5
1.0
3.5
2.8
1.9
3.5
0.9
1.0
N/A
Total N,
mg/L
1.0*
4.6
2.7
3.6*
4.9
2.8
4.1*
3.4*
4.6
5.4
1.5*
4.6
5.4
Re-
moval
88
92
75
90
83
93
84
82
81
81
95
87
85
N/A - Data not available
* Based on effluent NHJ-N + NOX-N
8.2.6.4 Process Design Features
SBR technology requires unique and innovative strate-
gies to accomplish each phase of the process cycle.
Large facilities that require dual vessels can accommo-
date continuous flow by alternating fill cycles between
reactors; single-vessel facilities except for ICEAS sys-
tems will require flow equalization or a selector. Compart-
ments or baffles may be included within a selector to
control the hydraulic regime and sludge characteristics.
Several criteria have been proposed that can be used to
design an appropriate selector (8,37,38). The CASS proc-
ess by Transenviro is a proprietary SBR that includes an
integral selector as part of the process.
The process control of an SBR requires relatively sophis-
ticated coordination of probes, valves, timers, and level
sensors. The recent advances and cost reductions of
microprocessors have been some of the causes of the
revival of interest in SBR technology, permitting auto-
mated control of the timing and sequence of process
phases and operation. The use of timers and DO monitors
can be used to reduce costs attributable to over aeration,
thereby reducing the lag period of DO depletion and al-
lowing the maximum time for denitrification to occur.
Maintenance of the desired solids inventory is facilitated
in an SBR since aeration and settling occur in the same
reactor and the entire sludge mass may be retained if
required. There is no set interval dedicated to sludge
wasting in an SBR, but wasting is only conducted as
performance requirements dictate.
The mixing requirements of an SBR reactor are similar
to flow-through systems. Since all reactions occur in the
same basin, some aeration systems cannot generally be
used to provide mixing during anoxic cycles. However,
the jet aeration systems offer independent control of mix-
282
-------
ing and oxygen transfer. The varying liquid volume re-
stricts the feasibility of fixed mechanical surface aerators.
The most common aeration system in SBRs are diffused
bubblers; but both the floating aerator as manufactured
by Aqua SBR and diffused bubble aeration systems will
benefit from submerged mixers used to ensure proper
agitation of the reactor contents under anoxic conditions.
Mixing considerations in SBR processes during anoxic
cycles are similar to considerations discussed for single
and oval anoxic zone processes.
8.3 Process Selection Considerations
Selection of the most cost-effective and efficient type of
single-sludge system for application to a particular waste-
water treatment facility will be influenced by a number of
factors, including:
• Effluent limits
• Wastewater characteristics
• Site constraints
• Existing facilities
• Cost
The following discussion of selection factors will focus on
single-sludge systems but will include some consideration
of separate-sludge systems. In many cases, combining
at least BOD removal and nitrification into a single-sludge
system will be cost effective considering the additional
cost associated with separating these two processes.
However, there will be situations where effluent limits,
wastewater characteristics, or physical constraints re-
quire a separate-stage denitrification system.
8.3.1 Effluent Limits
The first criteria for selecting process alternatives will be the
permitted discharge limits. In general, process selection will
be driven by the form and degree of nitrogen removal re-
quired With requirements for phosphorus removal being sec-
ondary. Treatment alternatives may be broadly divided into
systems that remove ammonia (nitrification), those that re-
move nitrogen (nitrification and denitrification), and those
that remove both nitrogen and phosphorus.
8.3.1,1 Nitrification
Permits limiting ammonia, and not TN, will require a treat-
ment system consisting of an aerobic section only, using
a single-sludge or a separate-sludge system. However,
even if not required by the discharge permit, it may be
desirable to include denitrification in the treatment sys-
tem. Denitrification", which reduces the nitrates produced
during the nitrification process, offers features in a single-
sludge, system that can reduce operating costs. These
include the elimination or the reduction of chemical re-
quired -for pH adjustment in low-alkalinity wastewaters
subject to pH depression as a result of the nitrification
process and the reduction of aeration requirements be-
cause of the consumption of wastewater carbon (BOD)
by nitrates. There is a potential advantage to including
denitrification in a system that includes biological phos-
phorus removal. Nitrates interfere with, biological phos-
phorus removal. Denitrification will reduce or eliminate the
recycle of nitrates back to the anaerobic zone and thus
improve the system's phosphorus removal capabilities.
The presence of nitrates can also create problems in the
final settling tanks. If the DO level in the influent to the
settling tank is low, denitrification will occur when a source
of wastewater carbon is available or more likely because
of endogenous metabolism. These conditions are likely
to occur below the sludge blanket. If denitrification is
significant, nitrogen gas bubbles will become attached to
sludge particles causing them to rise. An operational so-
lution to this problem is to maintain the minimum possible
sludge blanket in the final clarifiers. Rapid sludge removal
collectors may also be utilized.
8.3.1.2 Denitrification
Permits that limit the TN level will require the treatment
process to nitrify and denitrify. As discussed above, sin-
gle-sludge systems are capable of complete nitrification.
Process selection becomes more a matter of how to im-
plement denitrification. The options available, in order of
increasing complexity, are single-sludge systems with al-
ternating aerobic/anoxic phases, single-sludge systems
with one aerobic/anoxic zone, single-sludge systems with
two or more aerobic/anoxic zones, and finally, separate-
sludge systems.
These basic options encompass a variety of treatment
processes—some that require relatively simple opera-
tional modifications to implement and others that require
more extensive additions to the existing facility.
8.3.1.3 Treatment Performance
The following discussion of nitrogen removal performance
• of the various treatment alternatives is intended as a
general guideline for the selection of alternatives for fur-
ther evaluation. The performance attributed to each proc-
ess is generally conservative. Actual treatment plant
performance data for various processes have been pro-
vided elsewhere in the manual. In some cases, however,
there is little full-scale operating data from which to make
engineering judgements. The discussion of performance
is divided into ranges of effluent TN. In reality, process
performance can not be so neatly delineated because
treatment efficiency depends on many factors that make
generalizations difficult. These factors include.wastewater
characteristics, variability of flows and loads, tempera-
ture, presence of inhibiting compounds, and the design
features of the plant (e.g., flexibility of operation, clarifier
loadings, and recycle flows).
283
-------
In many cases it may be possible to improve the perform-
ance of a process by modifying its operation or by adding
relatively minor equipment or process improvements. For
example, the nitrogen removal performance of a system
with a single anoxic zone may be improved—within lim-
its—by increasing the mixed liquor recycle rate. A CNR
system's performance may be improved by adding step
feed or an internal recycle. Methanol trimming may be
added to a second anoxic zone to improve denitrification
kinetics. If there is adequate excess capacity, a second
anoxic zone may be added.
When evaluating wastewater treatment alternatives for
biological nutrient removal (BNR), the tendency to lock
the design criteria for a particular process into a narrow
set of numbers should be avoided. For example, the Bar-
denpho process is generally a low-load process with rela-
tively long hydraulic retention times. If justified by the
specific circumstances, however, it can be designed to
operate at a higher rate, similar to the A2/O or VIP proc-
ess. The effluent limits and wastewater characteristics
should be the driving force in the design and not precon-
ceived limits on design criteria.
Actual plant operating data for the various processes may
be found in Section 8.2, which also presents detailed
descriptions and design criteria of the single-sludge proc-
esses. The following discussion approaches the task of
selecting a treatment process for nitrogen removal from
the standpoint of the maximum allowed permit level for
the discharge of TN. Where appropriate, the phosphorus
removal capability of the system is also included.
a. Total Nitrogen: 8-12 mg/L
Basically, any of the single anoxic zone systems will be
capable of achieving TN residuals of 8-12 mg/L in a
typical domestic wastewater on an average annual basis.
Multl-anoxic zone systems such as some types of oxida-
tion ditches or the CNR process would also be candidates
for providing cost-effective treatment.
An MLE type of approach (Figure 8-3) can be relatively
easily retrofitted into an existing wastewater treatment
plant through the installation of baffles, mixers, and an
internal recycle capacity. In a diffused air system, the
diffusers would have to be relocated to create an anoxic
zone and to ensure that there is adequate air for nitrifi-
cation in the oxic zone. Some diffusers could be left in
the anoxic zone to provide mixing if the DO concentration
were kept at a low level. An MLE process, which includes
an internal recycle, is capable of meeting an 8-mg/L TN
permit. Without the internal recycle, the process becomes
the basic Ludzack-Ettinger schematic (Figure 8-2) which
should be able to meet a 12-mg/L TN permit.
The UCT, A2/O, and VIP processes may be implemented
for a TN limit in the 8-12 mg/L range. However, these
processes also include an anaerobic zone or selector for
phosphorus removal and bulking control. Selection of one
of these processes will depend more on the wastewater
characteristics. The UCT or VIP process would be fa-
vored when a high degree of phosphorus removal is
required and when the TBOD5:TP is low (<20:1). Both
the A2/O and VIP processes are high-rate systems with
relatively low 0c's and short hydraulic retention times.
As such, they may offer more opportunities for imple-
mentation in an existing treatment plant with limited
excess capacity.
Oxidation ditches are capable of performing in this range
of TN removal. The performance of a ditch system cannot
be predicted with as much certainty as other activated
sludge systems. Acceptable performance will require field
monitoring for optimizing the operation, such as determin-
ing the DO profile. The process can then be optimized by
adjusting the DO level, which is done by turning aerators
on or off or by controlling the oxygen transfer rate by
varying the aerator submergence and/or horsepower.
SBRs can produce an effluent with <8 mg/L TN but close
attention to operating conditions is required to optimize
performance. SBRs are well suited for relatively small but
highly variable flows, since equalization of flows is inher-
ent in the process.
The CNR process has been demonstrated to produce a
TN effluent of <10 mg/L and, with attention to operating
conditions, can consistently achieve TN levels of <8 mg/L.
This may require the ability to step feed with multiple
compartments, especially in warmer weather, to provide
a carbon source for denitrification. In existing plants with
little excess capacity, CNR offers .the potential to meet a
seasonal TN permit with relatively small modifications.
During the warmer periods, when reaction rates for car-
bonaceous BOD and nitrification are higher, the aerobic
volume of the basin may be smaller, thus allowing capac-
ity for the anoxic operating condition. During colder peri-
ods, if all the existing tank volume is required for the
aerobic processes, the system would not be able to de-
nitrify. If the permit was not based on a seasonal TN limit,
then either additional tank volume would be required or
a separate-stage denitrification system would be needed.
If the permit places only a seasonal restriction on TN,
then the separate-stage denitrification system would not
be required.
b. Total Nitrogen: 6-8 mg/L
Typically, an effluent TN limit of 8 mg/L is presented as
the limit of performance for single anoxic zone systems,
such as for the A2/O, UCT, and VIP processes. However,
these systems should be capable of producing effluent
TN concentrations of 6-8 mg/L with further enhance-
ments. As the TN limit decreases, the internal mixed liq-
uor recycle rate from the aerobic to the anoxic zone must
increase. The ability to recycle 100-400 percent p'f the
flow should be provided. Also, the design criteria must be
284
-------
more conservative in order to provide the treatment effi-
ciency and reliability required. A second anoxic zone may
be added with methanol addition to reduce remaining
nitrates.
Dual anoxic zone processes should be considered in this
range of performance although they are capable of .even
greater TN removal. The Bardenpho system, for example,
is capable of achieving TN removals to <3 mg/L. How-
ever, in the 6-8 mg/L TN range, it is possible to implement
the Bardenpho process without effluent filtration if ade-
quate settling capacity is provided.
Oxidation ditches are capable of achieving effluent TN
levels down to 6 mg/L. The literature reports a wide va-
riety of treatment efficiencies for ditch systems principally
because the operating conditions and wastewater char-
acteristics vary so greatly. The inherent internal recycle
rate of a ditch system often exceeds 100 times the influ-
ent flow. This tends to dilute the influent BOD, which
improves the competitive advantage of the nitrifying or-
ganisms. The nitrifiers are able to compete for the avail-
able DO over a larger fraction of the aerobic volume.
Thus, nitrification is maintained over a larger fraction of
the mixed liquor volume. However, the dilution of BOD,
which is the carbon source for denitrification, also tends
to decrease the denitrification rate. Another key to reliable
nitrogen removal performance in an oxidation ditch is to
provide the ability to closely monitor and control the DO,
as was discussed in Section 8.2.5.4.
The BioDenitro process by Kruger can consistently pro-
duce an effluent with a TN level of 6 mg/L. Effluent con-
centrations of 3 mg/L have been obtained, but consistent
nitrogen removal to this effluent concentration has not yet
been demonstrated.
c. Total Nitrogen: 3-6 mg/L
Dual anoxic zone processes must be considered to
achieve TN residuals of 3-6 mg/L because of the addi-
tional denitrification that is required and provided by the
second anoxic zone. The Bardenpho process is typically
the single-sludge system applied at TN limits <3 mg/L.
This process was developed as a nitrogen removal proc-
ess :in contrast to the A/O or A2/O processes, which typi-
cally operate at higher loading rates and shorter hydraulic
retention times (HRTs).
The Modified UCT process has two anoxic zones, but the
second anoxic zone follows in series after the first. As
was discussed in detail in Section 8.2.3, the Modified UCT
performance is similar to single anoxic zone processes.
Single anoxic zone processes are not able to achieve
effluent nitrogen concentrations of 3-6 mg/L for typical
waste'waters. Since the Modified UCT is intended as a
compromise between nitrogen and phosphorus removal,
this process would be expected to be similar or slightly
better to that of the UCT process.
Separate-sludge systems can also be designed for TN
permit levels of <3 mg/L. Separate-stage denitrification
using a downflow filter may be a desirable approach if a
strict TSS limit (<10 mg/L) is also imposed by the dis-
charge permit. If effluent filtration is not desired as a
component of the denitrification system, then an upflow
packed-bed reactor may be used for denitrification. Dur-
ing the warmer months, it may be possible to implement
cyclical aeration in the nitrification stage to achieve deni-
trification and reduce the methanol requirements in the
denitrification reactor, thus saving operating costs.
8.3.1.4 Phosphorus Removal
Many discharge permits limiting* the discharge of TN will
also include a TP limit. Very low levels of effluent TP can
be achieved through biological processes (for a detailed
discussion of these systems, see Reference 39).
Where moderate levels of nitrogen removal are required,
single-sludge, single anoxic zone systems such as the
A2/O, VIP, and UCT processes can be applied. These
systems can produce effluents with <1 mg/L TP although
results are variable. Selection among the A2/O, UCT, and
VIP processes depends primarily on the wastewater char-
acteristics. Reliable performance to 51 mg/L will require
chemical precipitation in addition to biological removal.
As discussed previously, dual anoxic zone systems are
typically applied where low levels of TN are also required.
In this case, the modified or five-stage Bardenpho proc-
ess would be considered since it includes an anaerobic
zone at the beginning of the process train. Effluent TP
levels of <3 mg/L have been reported with these systems.
To achieve a <1-mg/L TP limit, and to consistently pro-
duce an effluent with <2 mg/L TP, provisions for chemical
addition should be included. TP levels below 0.5 mg/L
may require effluent filtration in addition to chemical treat-
ment. A phosphate detergent ban or use of a selector
may allow plants to achieve effluent phosphorus concen-
trations consistently below 1 mg/L; however, testing
would be required to verify this.
Biological nutrient removal systems designed for both
nitrogen and phosphorus removal present some unique
operating considerations since the two processes have
conflicting requirements. The mechanism for biological
phosphorus removal in a mainstream process such as
UCT, A2/O, VIP, or modified Bardenpho is through the
assimilation of enhanced levels of phosphorus in the bac-
teria selected by the anaerobic zone and subsequent
elimination of this phosphorus through waste sludge.
Therefore, biological phosphorus removal is optimized at
a shorter 00 than is typically employed for nitrification.
Sludge production and subsequent phosphorus removal
is promoted at the shorter 90's. Generally, the compromise
is to operate at the shortest 9G required to achieve effluent
nitrogen limits.
285
-------
Nitrates also present a problem for biological phosphorus
removal systems. For "luxury" uptake of phosphorus to
proceed, the bacteria must first assimilate organic sub-
strates in the anaerobic zone and store them within their
cells. Nitrates that may be present in the RAS allow fac-
ultative heterotrophic bacteria to compete for soluble sub-
strates required by the phosphorus-removing bacteria.
This can degrade the phosphorus-removal performance
of systems such as the A2/O or modified Bardenpho,
which recycle RAS directly to the anaerobic zone. The
UCT and VIP processes avoid this problem by returning
the RAS to the anoxic zone where the nitrates are re-
duced by denitrification. Mixed liquor is then recycled from
downstream of the anoxic zone to the anaerobic zone.
A key selection criteria, therefore, for BNR systems which
must remove both nitrogen and phosphorus is the
TBODS:TP ratio. If this ratio is greater than 20:1, then the
recycle of nitrates to the anaerobic zone may not be a
problem because there is an adequate supply of excess
organic substrate. In this case, either the A2/O or the
modified Bardenpho process should be considered for
stricter limits on TN. These systems do not have the costs
associated with an additional recycle from the anoxic to
the anaerobic zone. If the TBOD5:TP is less than 20:1,
then the VIP or UCT process should be considered. For
the same reasons, the TKN:BOD is important. A high
TKN:BOD may indicate there is insufficient carbon to de-
nitrify completely, resulting in a carryover of nitrates to
the anaerobic zone.
For wastewater treatment plants that must remove phos-
phorus and are required to nitrify but not denitrify, it may
still be advantageous to include denitrification as a com-
ponent of a single-sludge system. This will reduce the
interference of nitrates with the biological phosphorus re-
moval mechanism and will also provide the cost savings
associated with the return of alkalinity and reduction in
aeration requirements.
Mainstream biological phosphorus-removal systems will
require effluent filtration in order to achieve low levels of
TP (<0.5 mg/L). This is because the solids that are carried
by the clarified effluent will contain an enhanced level of
phosphorus, which can have a significant impact on the
effluent TP levels. Typical activated sludge solids will con-
tain approximately 2 percent phosphorus, whereas the
solids from a biological phosphorus removal system will
contain typically 4-8 percent phosphorus.
8.3.1.5 Time Basis of Permit Limits
A permit with a TN limit that is based on a yearly average
obviously provides greater flexibility than a permit that
requires the same discharge limit on the basis of a maxi-
mum monthly average. Similarly, a monthly average per-
mit provides more freedom than one based on a
maximum weekly average. With a shorter time period,
fewer excursions above the permit limit can be tolerated.
The shorter time periods will therefore require a more
conservative process design approach to improve reliabil-
ity. Operating experience at biological nutrient removal
facilities has indicated that typically the maximum month
effluent TN will be 1.4 times the average annual value
(40). Phosphorus removal is more variable. Typically, the
maximum month effluent TP will be 2 times the average
annual value.
8.3.1.6 Seasonal Limits
A permit requiring nitrification or nitrogen removal only in
the summer offers more opportunities for process selec-
tion than does a permit that requires year-round nitrogen
control. For example, a plant that has limited excess ca-
pacity may not be able to implement denitrification using
CNR year-round, but may be able to do so in the summer
when reaction rates are greater. Because the aerobic
volume required for nitrification is less in the summer,
there is generally more flexibility in existing plants to
achieve the seasonal permit limits through operational
modifications of the existing process. Seasonal permits
are common for ammonia removal when the concern is
related to fish toxicity. Having to nitrify only in the summer
reduces the aeration tank volume required and thus re-
duces capital costs.
Most instances of noncompliance with effluent nitrogen
limits occur in the first month when the seasonal restric-
tions apply. To ensure that the system can achieve the
level of removal when required, the wastewater tempera-
ture during the month prior to the month the seasonal
limits are to take effect should be included when estab-
lishing the design temperature.
8.3.1.7 Solids Limits
A well-operated treatment plant with conservatively de-
signed final clarifiers should be able to consistently pro-
duce an effluent with <15 mg/L TSS. Effluent <10 mg/L
will generally require effluent filtration. *
Once filtration is required, more consideration may be
given to separate-stage nitrification-denitrification be-
cause the sand filter with methanol addition can also
provide the required denitrification. However, the analysis
must also consider the present worth of the added op-
erations and maintenance costs. Since separate-stage
denitrification requires methanol addition, the savings in
aeration costs and chemical for alkalinity adjustment as-
sociated with single-sludge systems will not be realized.
A low level of TN required by a discharge permit does
not necessarily imply that filtration must be added. Al-
though effluent solids do contain organic nitrogen, the
amount of organic nitrogen that would appear in the TN
of the final effluent may be small. A TSS level of 30 mg/L
in the final effluent would contribute only approximately 2
mg/L to the TN level. A requirement for very low levels of
286
-------
TN (<1 mg/L) will require filtration because of the particu-
late organic nitrogen and the nondegradable soluble or-
ganic nitrogen fraction. Often discharge permits requiring
very low levels of effluent TN also limit effluent TSS to a
level where filtration is required.
8.3.2 Wastewater Characteristics
The characteristics of the wastewater to be treated will
affect treatment performance and thus the selection of an
effective process. Of primary concern to single-sludge
nitrification-denitrification systems is the ratio of BOD5 to
TKN. Organic carbon is required by the denitrifying or-
ganisms. The BOD5:TKN is an indication of the supply of
necessary carbon, with a high ratio favoring denitrifica-
tion. Also significant is the presence of readily available
organic carbon, which is indicated by SBOD5:BOD5. A
high proportion of soluble and readily degradable BOD
would favor dehitrification and improve reaction rates.
Wastewater temperature is a critical parameter since it
affects the growth rate of nitrifiers and thus the design
90 and also the rate of denitrification. The availability
of adequate capacity in an existing wastewater treat-
ment plant is therefore significantly affected by the design
temperature.
The pH of a wastewater also affects nitrification and de-
nitrification rates. The optimum pH range for nitrification
is generally accepted to be 6.5-8.5. For denitrification the
optimum pH is 7.0-8.0. Since nitrification consumes al-
kalinity, the natural bicarbonate alkalinity in a wastewater
is of concern. If the alkalinity remaining after ammonia
oxidation is <50 mg/L (as CaCO3) then provision must be
made to supplement the alkalinity. Single-sludge nitrifica-
tion-denitrification systems will have alkalinity returned to
the process as a result of the denitrification reaction.
Approximately 50 percent of the alkalinity lost through
nitrification can be regained during denitrification in a sin-
gle-sludge system, if all nitrates are denitrified.
Variability in flows and loads will negatively affect process
performance. Wastewater treatment facilities with highly
variable flows should consider the addition of flow equali-
zation. An equalization basin can also dampen peak loads
caused by internal recycle flows such as digester super-
natant returns and dewatering operations. Because of the
concern for the BOD5:TKN as described previously, the
possibility of BOD and TKN peaking at different times
should also be investigated.
Wastewaters that include a significant contribution from
industry should be investigated for substances that may
be inhibiting to the nitrification process. A separate-stage
nitrification system affords some protection to the nitrifiers
by providing buffering capacity in the first-stage carbona-
ceous BQD removal process. Inhibitory compounds can
significantly reduce the rate of nitrification. The addition
of powdered activated carbon may enhance nitrification
rates in these cases.
Very high ammonia concentration (in the NH3 form) can
be toxic to nitrifiers. The amount of ammonia present as
NH3 is dependent on both pH and wastewater tempera-
ture, with its relative concentration increasing as the pH
and wastewater temperature increase (Figure 1-5).
Collection systems that suffer from a high degree of infil-
tration/inflow or that contain combined sewers will pro-
duce a dilute wastewater. Such wastewater will exhibit
lower rates of denitrification because of the lower con-
centration of organic carbon.
Wastewaters that include septage may contain a rela-
tively higher fraction of refractory TKN. This form of nitro-
gen is resistant to treatment and may require long Gc's to
achieve even partial oxidation.
8.3.3 Site Constraints
If the space available for plant upgrade or expansion is
limited, then single-sludge systems, which do not require
intermediate clarifiers, should be considered. Although
the reactor volume required for single-sludge nitrification-
denitrification will be greater than that required for a sepa-
rate-sludge system, the total combined volume including
reactor and settling capacity may be less for a single-
sludge system.
In situations where sufficient space is not available to
expand immediately adjacent to the existing activated
sludge tankage, it may be necessary to implement a
separate-stage system and use the existing tankage for
carbonaceous BOD removal only. The second stage can
be built elsewhere, though pumping may be required.
Alternatively, it may be possible to divert some of the plant
influent to new tankage, thus allowing nitrification-denitri-
fication to be incorporated into existing tankage. Plans for
upgrade and expansion should also consider future re-
quirements so that process selection and site planning
do not preclude treatment alternatives in the future.
Where space is a concern, consideration should be given
to higher rate processes such as the VIP process. This
process, which was developed to optimize biological
phosphorus removal, operates at a total 80 of 5-10 d
under warm weather conditions compared to the modified
Bardenpho or UCT processes, which typically operate at
e0's of 10-25 d. Also, the VIP process promotes higher
rates by creating multiple compartments instead of single
complete mix reactors for the anaerobic, anoxic, and
aerobic zones. This approximates a plug flow reactor
which increases the substrate concentration in the initial
compartments of each zone, thereby increasing the reac-
tion rate. The overall result is that TN removal can be
achieved in a smaller volume and in less area. The higher
rates associated with separate-stage denitrification as
287
-------
compared to denitrification in a single-sludge system, can
also reduce land requirements.
SBR systems offer the possibility of a high degree of
treatment in a relatively small space. Since settling occurs
in the aeration basin, there is no need for a final settling
tank. Also, this allows operation at relatively high MLSS
concentrations, which can decrease volume requirements.
There are also design considerations that can reduce
land requirements. These include the use of common wall
construction with rectangular settling tanks, the use of
deeper aeration tanks with fine bubble aeration, and the
use of methanol to increase denitrification rates.
8.3.4 Existing Facilities
The nature of the existing facilities will have an effect on
the process selection when upgrading for nitrogen re-
moval, especially when attempting to make maximum use
of the existing facilities to reduce costs. Usually, a single-
sludge system can be more easily retrofitted into an ex-
isting activated sludge plant than can a separate-stage
system. This is particularly true if there is sufficient excess
capacity available to allow the anoxic zone or zones to
be incorporated without building additional tankage. A
separate-stage system will require intermediate clarifiers
and process tanks and may require an intermediate
pumping station.
Effluent limits will govern process selection, but where a
single-sludge, single anoxic zone system is applicable, a
high-rate design can be more easily retrofitted where ex-
cess capacity is limited. Additional baffling within zones
to approximate plug flow kinetics can improve nitrification
and denitrification efficiency. The VIP process exploits this
type of design.
There may be occasions where the configuration of the
existing aeration basin does not allow the basin to be
divided into the proper size compartments or the baffles
to be located where desired to create separate aero-
blc/anoxic zones. For example, a mechanically aerated
basin will typically be divided into compartments with the
aerator at the center, and it may be difficult to install a
new baffle at any location other than where the aeration
compartments are already divided. One possible solution
is to dedicate the first compartment to an anoxic condition
but then employ cyclical aeration in the second compart-
ment. The total effective anoxic volume may then be var-
ied by changing the on/off time of the aerator in the
second basin. These schemes would require internal re-
cycle of nitrates to the anoxic zone.
Existing facilities may limit the level of MLSS that can be
carried in a system either because of the loadings on the
final clarifier or the capacity of the return sludge pumps
and piping systems. This could limit the performance of
a system that must nitrify year-round, if in winter the
system cannot carry enough MLSS to allow nitrification
to continue. If the capacity of the clarifiers or the return
sludge system cannot be improved, then a possible so-
lution is to install fixed-film media in the aerobic section
of the tank (Section 6.6). This would increase the nitrifier
population and the effective 90 of the nitrifiers without
burdening the clarifiers. This technology has been applied
in Japan and Europe, but is relatively new to the United
States.
Nitrification and denitrification reaction rates are tempera-
ture dependent. If the existing facilities do not provide
enough tank volume to nitrify and denitrify year-round in
a single-sludge system, then it may still be cost effective
to provide a single-sludge system and add a separate-
stage denitrification step. During the summer, when reac-
tion rates are greater, the single-sludge system may be
adequate to denitrify and thus may provide the cost sav-
ings associated with the return of alkalinity and the use
of nitrates to oxidize organics. During the winter, when
more of the existing tankage is required to nitrify, the sepa-
rate-stage system could be operated for denitrification.
8.3.5 Costs
8.3.5.1 Capital Cost
The cost for upgrading existing wastewater treatment fa-
cilities or adding new facilities for biological nutrient re-
moval is site specific and varies considerably. Such
factors as the actual BOD and TKN loads, the nature of
the existing facilities, site conditions, degree of new ver-
sus retrofit facilities will have major impacts on the design
and cost of the facilities and it is difficult to provide mean-
ingful generalizations relating cost to design flow.
In many cases, single-sludge systems will have a lower
capital cost than separate-sludge systems, primarily be-
cause a single-sludge system does not require interme-
diate clarifiers. Estimates based on studies in the
literature indicate that a separate-sludge system can typi-
cally cost 15-20 percent more than a single-sludge sys-
tem. General guidelines for costs of various components
of a biological nutrient removal system have been pub-
lished (40). These guidelines provide approximate costs
based on a range of assumptions and are not intended
to replace a detailed cost estimate for a specific installation.
8,3.5.2 Operational Cost
Single-sludge systems offer several potential advantages
over separate-stage systems that can reduce their oper-
ating costs. Aeration requirements are reduced in a sin-
gle-sludge system when wastewater is used as the
carbon source for denitrification in the anoxic zone. Ni-
trates replace oxygen as the electron acceptor in oxidiz-
ing carbonaceous BOD in the denitrification reaction. The
net affect is to reduce the aeration required for BOD
removal by as much as 25 percent. This is partially offset
288
-------
by extra mixing energy required by anoxic reactors and
larger aeration tanks.
In addition, the use of wastewater as the carbon source
can eliminate the need for methanol addition as in sepa-
rate-stage denitrification systems. Methanol addition adds
a significant operating cost and is a hazardous material
to handle. This is potentially or partially offset by the high
capital and operation cost of mixed liquor recycle.
The need for supplemental alkalinity is also reduced or
eliminated in a single-sludge system. Approximately one-
half of the alkalinity lost during the nitrification reaction
can be recovered during the denitrification reaction; how-
ever, this is not of significance in alkaline or adequately
buffered waters.
Single-sludge systems have been shown to produce less
sludge than a separate-sludge system. The relatively long
GG in single-sludge systems results in increased endo-
genous respiration and thus less excess biomass to be
wasted. Also, BOD oxidation by nitrates under anoxic
conditions minimizes heterotrophic biomass production.
Therefore, when evaluating single-sludge systems for ni-
trogen removal, the impact on sludge production should
be considered.
Treatment systems with permits requiring both phospho-
rus and nitrogen removal may utilize biological phospho-
rus removal rather than just chemical precipitation to
reduce operating costs. Chemical precipitation for phos-
phorus removal may reduce the amount of alkalinity pre-
sent in the wastewater to the point where supplemental
alkalinity is required to avoid pH depression. By incorpo-
rating biological phosphorus removal it may be possible
to limit the loss of alkalinity to the point where chemical
supplementation is not required.
8.4 Design Considerations
8.4.1 Primary Settling
The use of primary settling tanks will provide the usual
benefits associated with such systems including the re-
duction of rag accumulations on aeration equipment; the
reduction of nonbiological floatables in the aeration tanks
and final settling tanks; and process improvements re-
lated to the capture of solids from return flows such as
digester supernatant and thickener overflow, and from
septage discharges. The removal of BOD in the primaries
will reduce the volume required by the biological reactor
for carbonaceous BOD removal and nitrification. How-
ever, there are additional factors to consider with a sin-
gle-sludge system for nitrification and denitrification.
Primary settling will also reduce the BOD:TKN, which
may reduce the rate of denitrification that can be
achieved. This may not be a significant problem if a large
fraction of the total BOD is soluble, in which case, the
removal of particulate BOD may not adversely affect de-
nitrification. A BOD:TKN >5 favors denitrification. If pri-
mary settling is contemplated, a short settling period or
the use of fine screens should be considered; alterna-
tively, provision should be made to bypass a portion of
the raw wastewater around the primary settling tanks to
increase available carbon for denitrification.
8.4.2 Aeration Systems
The aeration system must be sized to handle the in-
creased oxygen demand imposed by nitrification and
must be capable of delivering the total amount of oxygen
required for complete carbonaceous BOD removal and
nitrification under peak loading conditions and changing
seasonal conditions. Plug flow designs must consider the
greater oxygen demand at the head end of the tank.
Additional cost savings can be obtained by installing a
DO monitoring/aeration control system to vary the blower
output in response to the oxygen demand.
Seasonal and diurnal variations in total oxygen require-
ments can cover a large range. Diffused air systems, with
the turn-down capability inherent in blower equipment and
the ability to taper the aeration capacity, can take advan-
tage of these variations and provide savings in operating
costs. Fine pore aeration systems are recommended over
coarse bubble because of their increased oxygen transfer
efficiency, but with that comes an increased potential for
fouling. Where cyclical aeration is used, ceramic-fine bub-
ble diffusers should not be employed; flexible membrane-
type diffusers have been used with cyclical start/stop
operation in small systems and should be considered. For
cyclical systems, electrically operated butterfly valves
should be provided on air headers to allow cycling of the
air supply to various tank compartments.
Mechanical surface aerators with conservative service
factors require less maintenance but do not have the
same degree of turn-down capability as diffused air sys-
tems. Also important in northern climates is the tendency
of mechanical aerators to increase heat loss. Mechanical
aerators are frequently used for cyclical nitrogen removal
systems because they can be easily cycled on and off at
set intervals using programmable timers. Aerator cycles
may be staggered to avoid high-ampere draws upon
aerator startup, or connected to a variable frequency
drive. Timers should be adjustable to allow each on-, or
off-cycle to vary over a range of 30 minutes to a few hours
as well as allow various cycle patterns at different times
of the day and different days of the week. In existing
plants that are being operationally modified for nutrient
removal, mechanical aerators can be converted to mixers
for use in an anoxic zone. One oxidation ditch technology
uses variable-speed rotors combined with weir level con-
trol to yield a highly flexible range of aeration and mixing
conditions, which provides conditions that transcend
some of the above issues.
289
-------
Submerged turbine aerators provide some of the benefits
of both the diffused air and mechanical surface aeration
system. They do offer some turn-down capability, at least
in regard to the air supplied to the diffuser. An anoxic
zone can be easily created by shutting the air completely
off, in which case the turbine would serve as a mixer.
DO monitoring should be considered for any system that
incorporates aerobic and anoxic zones. DO information
is critical to optimizing system performance. This is es-
pecially true when operating a plant in the CNR mode or
when attempting to operationally modify an oxidation
ditch for BNR. Automated DO control should be consid-
ered in most systems to save energy and to control the
process. For cyclical aeration, the DO level during the
aerobic phase should be maintained at 1-2 mg/L.
8.4.3 Mixers
Submerged propeller mixers or turbine mixers are typi-
cally used to maintain the MLSS in suspension in the
anoxic zone. The location of the mixer(s) is critical to
proper operation and the manufacturer must be consulted
regarding this matter. The objective is to provide mixing
energy without turbulence, which would entrain air, and
to avoid dead spots, which could become anaerobic. Mix-
ers are desirable during the anoxic phase with cyclical
aeration but are not mandatory if the off-cycle is short.
Consideration should be given to aerator designs that can
provide mixing during the anoxic cycle, such as sub-
merged jets, or variable-speed, variable-depth controls on
the Kruger ditches.
8.4.4 Recycle Pumping
The pumping of nitrified mixed liquor from an aeration
zone to an anoxic zone to recycle nitrates for denitrifica-
tion Is typically required. This will often require pumping
from one end of the aeration tank to the other, over a
tank wall or flow channel, or through the aerator basin
wall. In these cases, the water level in the aerobic and
anoxic zones is approximately the same, and the system
head will normally be low. However, pumping volumes as
large as four times the plant influent flow may be required.
Larger pumping volumes are impractical since the mar-
ginal increase in nitrogen removal via internal recycle
decreases significantly for recycle rates >400 percent of
the influent flow. This concept is discussed in detail in
Section 8.2.3.3. The rate of nitrate recycle controls the
denitrification process in the first anoxic zone and estab-
lishes the maximum efficiency achievable assuming the
wastewater organic content is sufficient. Multiple smaller
pumps should be provided in lieu of a few large pumps
to control the recycle rate as changing conditions dictate
to optimize the process. DO concentrations in recycle
streams should be kept to a minimum.
8.4.5 Reactor Design
Aerobic and anoxic zones should be designed to allow
for flexibility in operation to optimize the various proc-
esses by the use of channels, piping, gates, and valves
such that alternate feed points or tanks and compart-
ments can be used for influent, internal recycles, and
return sludge. Control of DO levels, solids inventory, re-
cycle rates, sludge blanket levels, and tankage in service
is necessary to optimize virtually all of the processes
given the impact of changes in diurnal loadings, seasonal
loadings, and temperature changes.
Submerged baffles are desirable to divide the anoxic
zones into compartments operated in series to simulate
a plug flow type configuration. Multiple compartments in
the nitrification zone may be desirable to avoid short-cir-
cuiting of ammonia and to ensure that the internal recycle
flow to the anoxic zone has been fully nitrified. For cyclical
or multiple anoxic zone nitrogen removal processes, the
ability to step-feed influent flow to downstream compart-
ments may be desirable to provide wastewater as a carb-
on source during denitrification in lieu of an internal
recycle.
8.4.6 Secondary Settling
Biological nutrient removal systems are susceptible to the
same operational problems experienced with typical ac-
tivated sludge systems and may be plagued by some
additional problems due to the presence of anaerobic and
anoxic zones. Bulking sludge may occur with the growth
of filamentous organisms. One possible cause of bulking
sludge is a condition of low DO. This situation may occur
if close control of the aeration system is not maintained
during periods when oxygen demand is increasing such
as in plants that transition seasonally from carbonaceous
BOD removal only to operation with nitrification. The use
of chlorine to control the growth of filamentous organisms
can be effective. However, this practice may be harmful
to the performance of plants that also incorporate biologi-
cal phosphorus removal, since the chlorine can also oxi-
dize the soluble organic substrates required for efficient
biological uptake of phosphorus. Excessive anoxic reten-
tion periods may also promote bulking sludge. The total
anoxic period should not exceed the time required for
denitrification of the nitrate mass returned via the recycles
(38).
The nuisance organism Microthrix parvicella, which pro-
duces scum and is difficult to eliminate with anoxic selec-
tors only, has been reported at biological nutrient removal
plants. Design of BNR facilities must assume that foam
and scum will occur and provide adequate facilities for
the collection and disposal of scum and floating solids
from clarifiers.
The addition of an aerobic stabilization zone prior to the
final settling tank has been reported to improve settling
290
-------
performance. Improved performance is likely the result of
increasing the DO level in the influent to the final settling
tank, thus preventing denitrification. Nitrogen gas, pro-
duced by denitrification, attaches to sludge particles caus-
ing them to rise. Also, the additional aerobic detention
can prevent denitrification by oxidizing remaining waste-
water organic matter or any remaining methanol if it is
used in a postdenitrification stage. This would eliminate
a carbon source for denitrification. However, this ap-
proach has a potential negative impact on systems that
recycle RAS to the anoxic zone. The RAS is more likely
to have a level of DO that will tend to decrease the
denitrification rate.
*
Another possible solution for systems that may be
plagued by rising sludge is to provide rapid sludge re-
moval equipment, such as vacuum collector final settling
tanks.
8.4.7 Selectors
Several researchers have observed poor sludge settling
characteristics in nitrogen removal processes (14,38,41).
Nuisance filamentous organisms Microthrix parvicella,
Sphaerotilus natans, Nocardia, and Types 021N and
1701 have been identified in bulking sludge samples; and
their presence has been determined to induce bulking
conditions. The organism most often identified in bulking
sludges is M. parvicella, which has been characterized
as a low F/M microorganism. Low F/M organisms exhibit
a higher growth rate at low substrate levels (38). Conse-
quently, they will proliferate at low F/M, suppressing the
growth of floe forming bacteria. Conversely, at high sub-
strate concentrations, floe-forming zoogleal organisms
maintain higher growth rates and are able to outcompete
filamentous organisms. Thus, bulking sludges due to M.
parvicella can be suppressed by providing a zone with
high substrate loading conditions.
Other causative factors of bulking sludges include anoxic
mixing sequences (41), low BOD:N and BOD:P (42), and
low DO levels. As a result of the variety of relationships
recorded that were determined to cause bulking, no single
process variable has been acknowledged as a process
control parameter (43). Mixing return sludge with influent
wastewater in one or more in-series contact chambers
for a short duration prior to directing the stream into a
complete mix basin has been suggested (44-46); the
pre-react chamber described was termed a "selector"
since it affects the selection of nonfilamentous organisms.
Biomass grown under a substrate gradient loading con-
dition has been observed to control sludge bulking in both
aerobic and anaerobic selectors (41). This observation
has been confirmed at several full-scale plants (8).
The following have been recommended for effective se-
lector design (43):
1. Selector should be designed with a sharp soluble or-
ganic substrate gradient.
2. Substrate leakage from the selector should be mini-
mized. The selector should be designed to remove
more than 90 percent of soluble substrate.
3. Microbial activity (determined from the substrate up-
take rate) should be maintained as high as possible.
While aerobic, anaerobic, and anoxic selectors have all
been found to control bulking effectively, the type of aera-
tion was determined to influence selector performance
(8). Mechanical aerators were hypothesized to produce a
DO gradient in the aeration basin, which, in conjunction
with an aerobic selector, resulted in a poorly settling
sludge. An anaerobic plug flow selector, however, was
effective in controlling sludge bulking when placed ahead
of aeration basins with surface mechanical aerators.
Whether effective selector performance with mechanical
surface aerators was due to the plug flow regime or the
anaerobic conditions, or both, the flow regime and an-
aerobic condition in concert, could not be determined.
While both aerobic and anaerobic selectors can control
bulking, anaerobic selectors can also provide the benefit
of phosphorus removal without requiring additional aera-
tion capacity.
8.5 Process Design Examples
8.5.1 Introduction
The process design examples in this section illustrate
three of the different process types. The design examples
illustrate sizing calculations for two different plant scenar-
ios (Figures 2-5 and 2-6) with two different sets of effluent
limitations (Table 2-10). Both plants in the scenarios from
previous examples are activated sludge plants and have
an average daily design flow of 220 L/s (5 mgd). Plant A
does not have primary settling or separate digestion.
Plant B has primary settling tanks and anaerobic di-
gesters followed by mechanical sludge dewatering. For
each plant, two sets of effluent limitations are imposed.
One set consists of secondary treatment standards with
nitrogen removal on a seasonal basis«to meet a TN limit
of 10 mg/L. The other set requires advanced waste treat-
ment for BOD5 and SS with a TN limit of 5 mg/L on a
year-round basis; effluent filtration is provided. For the
more stringent limitations, the impact on nitrogen removal
of imposing an additional limit of 1.0 mg/L TP is also
considered for that design example. Table 8-14 summa-
rizes the two sets of effluent limits.
The process design examples in this chapter expand
upon the examples in Chapter 6, which illustrated the
design of single-sludge nitrification systems and sizing of
the aeration tanks. The following examples illustrate proc-
ess design and sizing for single-sludge nitrification-deni-
trification systems, which primarily involves sizing of the
291
-------
Table 8-14. Design Example Effluent Limits
Effluent Limits,
mg/L
30-day 7-day
Effluent 1 (seasonal)
TN
NHJ-N
CBODS
TSS
10
2
30
30
15
3
45
45
Effluent 2 (year round)
TN
NHJ-N
CBODg
TSS
5
2
10
10
7.5
3
15
15
nitrification and anoxic zone(s) or phases in the aeration
tanks to achieve nitrogen removal. The design examples
also illustrate other design features, requirements, and/or
impacts on support systems. Specifically, the examples
Identify reactor size, typical reactor configuration, aera-
tor/mixing requirements, waste and return sludge require-
ments, and internal recycle rates. All calculations shown
are based on designing for 15°C (59°F) water tempera-
ture. The sizing results for designs at 10°C (50°F) and
20°C (68°F) are also summarized for comparison. The
plant scenarios illustrated are not intended to suggest
optimum approaches but to demonstrate calculations un-
der different conditions.
The following design features are some of the factors to
be considered during the facility sizing and design
phases:
• Internal Recycle Rate. The amount of denitrification in
systems with internal recycle is controlled by the rate
of recycle to the anoxic zone. There is a practical limit,
or point of diminishing returns, even when the influent
BOD:TKN is adequate (at least 3:1) such that waste-
water carbon is the carbon source. Internal recycle
pumps should generally be sized to provide an upper
limit of 3 to 4 times the influent flow rate, except in
unusual cases.
• DO Control. Automatic DO control for the aerobic
zones is desirable to reduce energy consumption and
to prevent high DO levels in the internal recycle to the
anoxic zone, which could adversely affect the denitri-
fication process. DO levels in the anoxic zone should
be less than or equal to 0.3 mg/L at all times. A tapered
aeration system is appropriate for a plug flow configu-
ration.
• RAS and WAS Pumping Rates. Variable-speed pumps
or flow-control valve arrangements should be provided
to control and vary pumping rates to adjust to changes
in influent loadings, reactor temperatures, and taking
tanks in and out of service.
• Internal Recycle Pumping. Pumps should be located
at the end of the nitrification zone where they will mini-
mize DO levels in the recycle flow. Recycle flow should
be returned to the anoxic zone via piping and should
be submerged at the point of discharge. Multiple
pumps are desirable to vary the internal recycle rate
depending on changing conditions.
• Multiple Basins. Multiple basins should be provided to
allow taking basins out of service during warm weather
or low loading periods; therefore, flexibility in piping,
valves, gates, and channels is desirable to operate the
system as needed.
For the four plant and effluent scenarios selected as de-
sign examples, three of the major classifications for nitro-
gen removal are used to illustrate process design for that
type of system. Each type of system used is intended to
be generic and does not reflect sizing techniques for any
particular system offered by manufacturers either proprie-
tary or nonproprietary. The following systems are used
for the design examples:
• Design Example No. 1. Single Anoxic Zone—Plant B
(complex plant) and less stringent limits.
• Design Example No. 2. Dual Anoxic Zones—Plant B
(complex plant) with more stringent limits.
• Design Example No. 3. Multiple Anoxic Phases (cycli-
cal)—Plant A (simple plant) with less stringent limits.
• Design Example No. 4. Dual Anoxic Zones—Plant A
(simple plant) with more stringent limits.
8.5.2 Summary of Design Procedures
The following is an outline of procedures used in design-
ing single-sludge nitrification-denitrification systems:
1. Determine influent characteristics, effluent limita-
tions, time basis of limits (e.g., monthly, weekly),
peaking factors, and design temperature based on
weekly or monthly minimum average temperature for
the time period that the nitrogen limits are in effect.
2. Prepare mass balances for the entire plant as shown
in Tables 2-15 and 2-16 for the annual average,
maximum monthly, and maximum weekly or peak
day conditions that could affect the design calcula-
tions. The mass balances should reflect the impact
of all recycle streams and any intermittent dis-
charges, such as septage or landfill leachate.
3. Calculate the level of treatment required for denitrifi-
cation and TN removal. All systems generally will be
designed to achieve complete nitrification. Select
type(s) of single-sludge process configurations re-
292
-------
quired to achieve the desired level of treatment to
meet the effluent limits with a margin of safety.
4. Calculate the volume and MLSS required for the ni-
trification zone based on aerobic design 9c and con-
trolling conditions at the final clarifier.
5. Determine the size of the first anoxic zone based on
the degree of denitrification required and/or achiev-
able with various internal recycle rates, where ap-
plicable. RAS rates should be included with internal
recycle rates for single anoxic zone systems with
predenitrification. Select the denitrification rate
based on the carbon source to be used and adjust
for temperature and peaking factors or maximum
design loading. Where feasible, denitrification rate
studies should be conducted prior to selecting the
denitrification rate used in design. Wastewater typi-
cally would serve as the carbon source where the
influent (feed to secondary system) BOD5:TKN is at
least 3:1. Atrial-and-error solution might be required
to size the anoxic zone, since the denitrification rate
is dependent on the anoxic F/M ratio (availability of
COD as the carbon source).
6. Size the second anoxic zone based on the nitrate
loading that was not denitrified in the first anoxic
zone, the additional denitrification required, and the
selected denitrification rate using an endogenous
carbon source. The rate should be adjusted for tem-
perature and maximum loading used for design. Al-
ternatively, methanol can be used in the second
anoxic zone.
7. Size the post-aeration zone to achieve a residual DO
level of 1-2 mg/L prior to the secondary clarifier.
8. Determine WAS and RAS pumping requirements to
cover the full range of possible conditions.
9. Calculate aeration requirements for nitrification and
mixing requirements for the anoxic zone. Generally,
the aeration system should be sized for nitrification
without the oxygen demand savings from denitrifica-
tion if sized on a maximum monthly basis. Peak day
and short-term peak demands should also be con-
sidered in determining total aeration capacity. The
minimum oxygen demand condition should be deter-
mined as well to ensure that the aeration system has
adequate turn-down capability to control DO levels
as desired.
10. Determine alkalinity requirements to ensure a residual
of at least 50 mg/L as CaCO3. Alkalinity produced by
denitrification should be included in the calculations.
11. Prepare final mass balance to check sizing of unit
processes and redo calculations as necessary.
293
-------
8.5.3 Design Example No. 1: Plant B with Less Stringent Limits
From the mass balances in Table 2-16, the inlet wastewater characteristics (primary effluent with recycles)
and secondary effluent characteristics for the more complex plant without nitrification-denitrification are as
follows:
mg/L Equivalents
Characteristic
VSS
TSS
CBODg
TCOD
SCOD
TN
Alkalinity, as CaCO3
Primary
Effluent
55
80
97
187
106
29.5
120
Secondary
Effluent
9
15
3
33
20
26.5
—
Design Conditions:
Reactor temperature = 15°C
Reactor MLSS = 3,000 mg/L
MLVSS = 63%
Reactor pH range = 7.0-7,6
Reactor minimum DO = 2.0 mg/L
Secondary effluent NHJ-N = 1.0 mg/L
Secondary effluent Org-N = 2.0 mg/L
Secondary effluent NOi-N = 5.0 mg/L
To meet seasonal limits of 2 mg/L NHjj-N and 10 mg/L TN, design for 1.0 mg/L equivalent NHJ-N and 8
mg/L equivalent TN at maximum monthly loadings. Secondary effluent Org-N of approximately 2.0 mg/L
equivalents represents the nonbiodegradable fraction of soluble TKN and nitrogen associated with effluent
VSS.
With an influent raw TKN concentration of 30 mg/L and effluent limits of 2 mg/L NHJ-N and 10 mg/L TN,
essentially complete nitrification and a minimum of 67-percent TN removal efficiency are required. A single
anoxic zone-type process, such as the MLE process, is adequate to meet these limits. The plant configu-
ration is as shown in Figure 2-6, for the more complex Plant B using a single aerobic zone preceded by
an anoxic zone for nitrogen removal with internal recycle, mixers in the anoxic zone, and mechanical surface
aerators. Plant B has effluent filtration, but this feature is not required to meet the effluent limits in this
design example.
In this example, it is assumed that neither the assimilation of TKN (3.0 mg/L) nor the percent volatile MLSS
(63 percent) will be affected by the solids retention time (60) and that the recycle stream characteristics will
remain as shown in the mass balance in Table 2-16. Therefore, the mass balance is unchanged.
Removal Requirements Across Secondary Process:
At average annual loadings (concentrations in mg/L equivalents)
BOD5 Removed = (18,925 m3/d) (97 - 3 mg/L)/1,000 = 1,779 kg (3,922 lb)/d
TN Removed = (18,925 m3/d) (29.5 - 8.0 mg/L)/1,000 = 407 kg (897 lb)/d
TN Removed in Waste Solids by Assimilation = (18,925 m3/d) (3.0 mg/L)/1,000
= 57 kg (125 lb)/d
TN Removed by Nitrification-Denitrification = 407 - 57 = 350 kg (772 lb)/d
294
-------
8.5.3 Design Example No. 1 (continued)
1. Sizing of nitrification zone. Similar to the design procedure outlined in Section 6.4.2.2, determine the
size of the nitrification reactor.
Calculate maximum nitrifier growth rate, £N:
(IN = 0.47 ea°98 (3-14)
£N= 0.47 d'1 at 15°C
Calculate specific growth rate, |XN:
UN = (IN [N/(KN + N)] (3-10)
i
where:
N = 1.0mg/L
KN = 1.0mg/L
UN = (0.47)[1.0/(1.0 + 1.0)] = 0.23 d'1
Calculate the minimum solids retention time, 6|P:
6? = I/UN (3-13)
eg1 = 1/0.23
9? = 4.35 d
Calculate design solids retention time, GO, using a design factor (Chapter 6):
6^ = (PF)(SF)(9S1) •
where:
PF = ,1.56 (Table 2.18)
SF = 1.25
eg = 1.56 X 1.25x4.35 = 8.5 d
Calculate the organic removal rate, QOBS:
qOBs = 1/(6g YNET) (6-5)
From Figure 2-10 at eg of 8.5 d, YNET is 0.24 g total VSS/g COD removed.
qoBS = 1/[(8.5)(0.24)] = 0.49 g COD/g MLVSS/d
Determine reactor hydraulic retention time, t:
t = (S0 - S^^qoBS X) (6-6)
where:
S0 =187 mg/LCOD
S, = 20 mg/L COD
X = 1,890 mg/L MLVSS
t = (187 - 20)/(0.49 x 1890) = 0.18 d = 4.3 hr
Calculate reactor volume, VN:
VN = Qt
Since t in Equation 6-6 was computed based on the mg/L equivalent concentrations, the flow Q to compute
the reactor volume is 18,925 m3/d (5 mgd) and not the actual flow of 21,056 m3/d (5.56 mgd) given in Table
2-16.
VN = (18,925)(0.18) = 3,410 m3 (0.90 Mgal)
295
-------
8.5.3 Design Example No. 1 (continued)
2. Determine various design parameters to check validity.
Actual Retention time, t:
t = (3,410/21,056) x 24 = 3.9 hr at Q of 21,056 m3/d (5.56 mgd)
t = 3.9/1.5 = 2.6 hr at Qmax mo of 31 ,584 m3/d (8.34 mgd)
Food/Mass:
F/M = (Qavg x BOD)/(VN x MLVSS)
F/M = (18,925 x 97)/(3,410 x 1,890) = 0.28 g BOD/g MLVSS/d
Specific nitrification rate required, SNRmin:
Nitrification required = 29.5 - 3.0 - 2.0 - 1 .0 = 23.5 mg/L NHJ-N
At design average:
SNRmln = (18,925 x 23.5)/(3,410 x 1,890) = 0.069 g NHJ-N/g MLVSS/d
At max. monthly:
SNRmln = 0.069 x PF, where PF = 1 .2 for influent TKN (Table 2-12)
n = 0.069 x 1.2 = 0.083 g NHJ-N/g MLVSS/d
Check minimum rates required against actual rates measured by testing. If rate tests are not performed,
check rates reported in the literature for similar 00 and COD:TKN or CBOD:TKN ratios. In this case, 0C =
8.5 d, COD:TKN = 6.3, and CBOD:TKN = 3.3. If minimum rates calculated above are too low, increase the
process design factor (DF) for sizing reactor (see Section 6.4.1.1).
3. Sizing of anoxic zone for denitrification. Size anoxic zone based on SDNR and adjust based on design
temperature and selected PF for design condition. This example is based on the maximum monthly.
As previously determined, nitrates produced in nitrification zone = 23.5 mg/L equivalents
Denitrification required = 23.5 - 5.0 = 18.5 mg/L or 350 kg NO§-N/d
Total NO^-N available in recycle (internal and RAS) streams = 23.5 mg/L
Select SDNR from rate tests, rates reported in the literature under similar conditions, or Eq. 4-22. In this
example, determine the denitrification rate from Figure 8-30 (40), which shows the specific rate of nitrate
removal as a function of the F/M ratio in the anoxic zone, and use it as the average rate for staged
compartments. By trial and error:
SDNR = 0.09 g NOg-N/g MLVSS/d at 20°C with wastewater as carbon source at anoxic F/M of
0.34 g BOD/g MLVSS/d at maximum monthly conditions.
Adjust nitrate removal rate using Eq. 4-25 at 6 = 1.08 (Table 4-2) for T = 15°C:
SDNR15.C = (0.09)(6T-20) = (0.09)(1.08)15-20 = 0.061 g NO§-N/g MLVSS/d
For maximum month where PF = 1.2: MLVSS required = 350 kg NOjj-N/d x 1.2 PF 4-
0.061 g NOi-N/g MLVSS/d
from Table 2-1 2
= 6,885 kg (15,179 Ib) MLVSS
Anoxic volume, VAN:
296
-------
8.5.3 Design Example No. 1 (continued)
0.12 :•
•a 0.1 •
I
\ o-08 '
OJ
E 0.06 '
IE
g 0.04
1
* 0.02
0.15
0.2 0.25 0.3
Anoxio F/M, g BOD/g MLVSS/d
0.35
0.4
Figure 8-30. Denitrication rate as a function of anoxic F/M (adapted from 40).
Calculate hydraulic retention time in anoxic zone at actual flow:
At Qavg, t = (3,643/21,056) x 24 = 4.2 hr
AtCWmo, t = 4.2/1.5 = 2.8 hr
t = 4.2/1.5 = 2.8 hr
System eg = aerobic eg + anoxic eg = 8.5 + [(3,643/3,410) x 8.5] = 17.6 d
4. Determine PAS rate.
To maintain MLSS = 3,000 mg/L at QRAS = 7,000 mg/L
3,000 x Q
QRAS —"
• = 0.75Q
RAS' (7,000-3,000)
At Q = 21,056 m3/d (5.563 mgd), QRAS = 15,790 m3/d (4.17 mgd).
At Qmaxmo = 31,584 m3/d (8.34 mgd), QRAs = 23,690 m3/d (6.26 mgd).
5. Size anoxic internal recycle rate. From previous calculations of denitrification required and NO^-N
returned to anoxic zone, the denitrification efficiency required equals 18.5/23.5 x 100 percent, which is 78.7
percent.
For single anoxic zone system:
Q|
Denit. eff . =
x 1 00%
297
-------
0.787 =
8.5.3 Design Example No. 1 (continued)
Q, + 0.75Q
Q,=
Q + Q|+0.75Q
[(0.787X1.75)-0.75]Q
(1-0.787)
Q, = 2.95Q
At Qmax mo = 31 ,584 m3/d (8.34 mgd), Q, = 93,160 m3/d (24.6 mgd).
6. Determine alkalinity requirements to maintain residual alkalinity of 50 mg/L equivalents as CaCO3 with
Influent alkalinity = 120 mg/L as CaCO3.
mg CaCO3 ,
Alkalinity demand = 7.14 — H - — — x 23.5 mg/L NHj-N oxidized = 168 mg/L as CaCO,
mg NHJ-N
Supplemental alkalinity addition required = (168 + 50) - 120 = 98 mg/L as CaCO3
3.6 mg CaCOo
Savings with denitrification = - - - - — x 18.5 = 67 mg/L as CaCOS
mg NHj-N reduced
Average supplemental alkalinity required with denitrification = 98-67
= 31 mg/L as CaCO3
Size the maximum capacity of the feed system on peak conditions in a similar fashion to prevent violation
of pH limits, usually daily. Size the system for peak day demand.
7. Determine mixing requirements in anoxic zone. At 50 hp/Mgal of anoxic volume, the minimum total hp
required equals 48 where V^N = 3,643 m3 (0.96 Mgal).
The number and size of each mixer is based on the number of anoxic compartments and the compartment's
configuration. Verify mixing requirements based on reactor depth and configuration. With six compartments,
8 hp is required per compartment. Therefore, each mixer is 10 hp for next standard size unit.
8. Determine waste sludge requirements as outlined in Section 6.4.2.2.
6* = IA/S (6-17)
where
IA = VSS under aeration, and
S = VSS wasted daily
IA = (1,890 mg/L MLVSS x 3,410 m3)/1,000
= 6,445 kg MLVSS (14,210 Ib MLVSS)
Using (Eq. 6-17):
S = 6,445/8.5 = 758 kg (1 ,670 Ib) VSS/d to be wasted
Solids contained in the effluent = 9 mg/L VSS from Table 2-15
Sludge wasting from RAS = 758 - (9) (18,925)/103 = 607 kg (1 ,340 Ib) VSS/d
At MLVSS/MLSS = 63%:
Average WAS = 963 kg (2,120 Ib) TSS/d
298
-------
8.5.3 Design Example No. 1 (continued)
Determine WAS pumping rate at XRAS =. 7,000 mg/L:
Average QWAS = = 138 m3/d (36,400 gpd)
Similarly, both WAS mass and WAS pumping requirements should be determined from mass balances at
peak conditions.
9. Determine aeration requirements under various design conditions.
Average Conditions
1 ft Q9^ m /H v fCT7 — r-
BOD5 Removed = ' b ' ^ """^ = 1,779 kg (3,922 lb)/d
K,,,+ K, •_,- ., 18,925 m3/dx23.5 mg/L „„,_, ,„„....,,
NHJ-N oxidized = —! o— — = 445 kg (981 lb)/d
103
For this example, 1.1 kg O2/kg BOD5 removal was assumed for the carbonaceous demand. The range for
nitrification systems is about 1.0 to 1.3, depending on 60 and the temperature (see Design Example 6-1 for
more information).
Total oxygen demand = (1.1 x 1,779) + (4.6 x 445) = 4,004 kg (8,827 lb)/d
Peak Day Conditions (Peaking Factors From Table 2-12):
Total oxygen demand = (2.1 x 1.1 x 1,779) + (1.7 x 4.6 x 445)
= 7,589 kg (16,730 lb)/d
Savings in O2 demand with denitrification from wastewater BOD5:
O2 saved = 2.9 mg O2/mg NOJ-N x 18.5 mg/L NO^-N
= 53.7 mg/L or 1,015 kg (2,237 lb)/d avg.
Oxygen demand with denitrification:
Avg O2 demand = 4,004 - 1,015 = 2,989 kg (6,590 lb)/d
Peak day O2 demand = 7,589 - (1.7)(1,015) = 5,864 kg (12,926 lb)/d
Mechanical aeration sizing
At 50 Ib O2/hp/d for mechanical aerators, the minimum hp required is 259. For series arrangement of
aeration basins, aerators should be sized to meet the higher oxygen demand in the first basin. Two-speed
motors and adjustable submergence are recommended to meet the varying conditions.
10. Atypical flow configuration with baffles in the anoxic zone is shown in Figure 8^-31. Nitrified flow is
recycled from the end of the nitrification zone to the head of the first anoxic compartment. Six mixers
required at 10 hp for anoxic compartments. The following table summarizes reactor volumes at 10°C and
20°C for comparison. The volume changes reflect the impact of temperature in Equations 3-14 and 4-25,
as well as the impact of 9C on the observed yield, YNET, from Figure 2-10.
299
-------
8.5.3 Design Example No. 1 (continued)
Volume, m3 Design Temperature
10°C 15°C 20°C
Van
VNK
VTot
5,370
4,980
10,350
3,640
3,410
7,050
2,270
2.380
4,650
To Secondaiy
Clarifiers
RAS
j
Primary Effluent
with Recycles
To Secondary
Clarifiers
Anoxic Zone
Aerobic Zone
O Mechanical Aerator
9 Mixer
•^ Internal Recycle Pump
Figure 8-31. Design Example No. 1—Single anoxic zone system.
300
-------
8.5.4 Design Example No. 2: Plant B with More Stringent Limits
The inlet wastewater characteristics (feed to secondary system) and secondary effluent characteristics from
the mass balance in Table 2-16 are the same as in Example No. 1 for the purposes of this example. To
meet the more stringent effluent limitations for SS, effluent filtration is required.
Design Conditions: ,
Reactor temperature = 15°C Reactor minimum DO = 2.0 mg/L
Reactor MLSS = 3,000 mg/L Secondary effluent NHJ-N = 1.0 mg/L
Volatile MLSS = 63% Secondary effluent Org-N = 2.0 mg/L
Reactor pH range = 7.0-7.6 Secondary effluent NOg-N = 1.5 mg/L
To meet the year-round limit of 5 mg/L TN, design for 4.5 mg/L equivalents TN or less at maximum monthly
loading and prior to filtration. Secondary effluent Org-N of approximately 2.0 mg/L equivalents represents
the nonbiodegradable fraction of soluble TKN and nitrogen associated with effluent VSS. Approximately 1
mg/L of nitrogen associated with the VSS will be removed by effluent filtration.
With an influent raw TKN concentration of 30 mg/L and an effluent limit of 5 mg/L TN, essentially complete
nitrification is required and a very high level of denitrification is required (>90 percent). To meet this limit,
a dual anoxic zone type process configuration is required to provide the additional denitrification required
in a second anoxic zone. The plant configuration is as shown in Figure 2-6 with effluent filtration. The
process configuration would consist of two anoxic zones and two aerobic zones for nitrogen removal with
internal recycle, mixers in the anoxic zones, and diffused aeration.
For this example, it is assumed that the assimilation of NHJ-N (3.0 mg/L equivalents) and percent volatile
MLSS (63 percent) will not be affected by the 6C to remain consistent with the mass balance in Table 2-16.
Removal Requirements Across Secondary Process: ,
At average annual loadings (concentrations in mg/L equivalents):
BOD5 Removed = 1,779 kg (3,922 lb)/d
TN Removed = 1 8-925 x (29-5 ~ 4-5> = 473 kg (1 )043 |b)/d
TN Removed in Waste Solids by Assimilation = 57 kg (125 lb)/d
TN Removed by Nitrification-Denitrification = 473 - 57 = 416 kg (917 lb)/d
1. Sizing of nitrification zone (first aerobic zone.) Following the same procedure used in Example No. 1,
the reactor volume VN = 3,410 m3 (0.90 Mgal) using the same design factor. Since the limits are more
stringent and year round, a more conservative design factor may be considered if daily or seasonal variations
are significant.
2. Sizing of first anoxic zone for denitrification. Size first anoxic zone based on specific denitrification rate
with wastewater as the carbon source and adjust based on design temperature and selected PF.
Allowing 1 .5 mg/L equivalents of NOa-N in the final effluent, the denitrification required with both anoxic
zones = 23.5 - 1 .5 = 22.0 mg/L or 416 kg (917 Ib) NOg-N/d.
Nitrates produced in first aerobic zone = 23.5' mg/L equivalents
NOg-N in internal recycle stream to first anoxic zone = 23.5 mg/L equivalents
Determine the maximum percent denitrification removal in the first anoxic zone based on the practical limit
for internal recycle from the end of the first aerobic zone to the first anoxic zone.
301
-------
8.5.4 Design Example No. 2 (continued)
Select Q| = 450 percent or 4.5Q as practical limit. While Q| can be higher, denitrification efficiency increases
at a decreasing rate as shown in Figure 8-8.
For dual anoxic zone systems, any return of nitrates to the anoxic zone from QRAs will be small because
of the removals in the second anoxic zone and can be ignored.
Q,
Denitrification efficiency = — — - _ — x 100%
Q + Q + QRAS
QRAS s 0.75 Q from Example No. 1
45Q
Denitrification efficiency = — — ' x 100 % = 72% maximum
NOJ-N Removed in first anoxic zone = 0.72 x 23.5 mg/L = 16.9 mg/L or 320 kg (706 lb)/d
NOJ-N Removed required in second anoxic zone = 5.1 mg/L or 96 kg (213 lb)/d
Select SDNR as in Example No. 1 . Since the amount of denitrification required in the first anoxic zone is
similar to that in Example No. 1 , the anoxic F/M and SDNR will be approximately the same.
At 15°C, SDNR! = (0.09)(1.08)'5 = 0.061 g NOi-N/g MLVSS/d
320kgNO;-N/dx1.2PF
For maximum month, MLVSS required = - - - - - = 6,295 kg (13,880 Ib)
0.061 g NOg-N/g MLVSS/d
.. 6,295 kg MLVSS x 103 0 00 . 3,_00,, „
V*N1 s 1.890 rnoyL MLVSS = 3'331 m (°'88 Mgal)
Calculate actual retention time
At Q of 21,056 m3/d (5.56 mgd), t = 3.8 hr
At Gnu* mo of 31 ,584 m3/d (8.34 mgd), t = 2.5 hr
3. Determine RAS rate.
QRAS - 0-75 Q similar to Example No. 1 .
4. Internal recycle rate.
*
As determined previously, Q| = 4.5 Q
At maximum monthly flow, Q| = 142,130 m3/d (37.5 mgd)
5. Size second anoxic zone for denitrification with endogenous carbon.
Determine SDNR2 from Eq. 4-23 as a function of ec:
SDNR2 = 0.12 x 6C -°-706 at 20°C
By trial and error, estimate that the system 60 = 25 d.
SDNR2 = 0.12 x 0.103 = 0.0124 g NO§-N/g MLVSS/d •
Adjust for T = 15°C using Eq. 4-25 and = 1.03 from Table 4.2:
SDNR2 = 0.0124 x (1.03)'5 = 0.011 g NO§-N/g MLVSS/d
removal required in second zone = 96 kg (213 lb)/d, as previously determined.
302
-------
8.5.4 Design Example No. 2 (continued)
96kg/dNO;-N
MLVSS required = 96 kg/d = 8,727 kg (19,240 Ib)
0.11 g NO^-N/g MLVSS/d
8,727 kg MLVSS x 1.2 x103 _ _..-• 3 .. '_ .. .. •
V*"2 = 1.890 mg/L MLVSS = 5,541 m3 (1.47 Mgal)
At Q of 21,056 m3/d (5.563 mgd), t = 6.3 hr
AtQmaxmo, t = 4.2hr
6. Determine alkalinity requirements as in Example No. 1.
7. Determine mixing requirements as in Example No. 1.
50 hp/Mgal required for mixing in the anoxic zones
VANI + VAN2 = 3,331 + 5,541 = 8,872 m3 (2.35 Mgal) '
Total hp required = 118 hp
8. Determine waste sludge as in Example No. 1.
9. Size post-aeration zone.
Size for 0.5 hr detention time at Qmax mo = 31,584 m3/d (8.34 mgd). ' ,
VPA= 660m3 (0.17 Mgal)
10. Calculate aeration requirements with fine pore diffused aeration system.
Similar to Example No. 1 for carbonaceous oxidation and nitrification, but with post-aeration zone to raise
DO from 0 to 2.0 mg/L.
Total oxygen demand = 4,004 + 42 = 4,046 kg (8,920 lb)/d avg.
= 7,684 kg (16,940)/d peak day * • • '•
Savings in O2 demand with denitrification:
O2 savings = 2.9 mg O2/mg NO§-N denit. x 22 mg/L NO^N denit.
= 63.8 mg/L or 1,207 kg (2,660 lb)/d avg.
Oxygen demand with denitrification:
Total oxygen demand = 4,046 - 1,207 = 2,839 kg (6,260 lb)/d avg.
= 7,684 - (1.7)(1,207) = 5,632 kg (12,416 lb)/d peak day
At 12.5 percent O2 Transfer Efficiency (assumed at 4.6 m [15 ft] diffuser submergence): :
Peak air required = (5,632 kg O2/d)/[(0.125)(0.28 kg O2/m3 air)(1,440 ^miri/d)]
= 112 nrvVmin (3,955 cfm) -
Provide three blowers plus one standby -
Blower Capacity = 38 m3/min (1,340 cfm)/blower
Check mixing requirements at end of nitrification zone with tapered aeration system to ensure that the air
provided for oxygen demand is adequate to meet the mixing requirement.
11. Reactor configuration is similar to reactor configuration in Example No. 1, but add second anoxic zone
and post aeration as shown in Figure 8-32. Nitrified flow is recycled from the end of the nitrification zone
303
-------
8.5.4 Design Example No. 2 (continued)
RAS
I
Primary Effluent
with Recycles
To Secondary
Clarifiers
First Anoxic Zone
Second Anoxic Zone
Aerobic Zone
Mixer
Internal Recycle Pump
Rgure 8-32. Design Example No. 2—Dual anoxic zone system.
to the head of the first anoxic compartment. The following table summarizes reactor volume requirements
at 10°C, 15°C, and 20°C for comparison:
Volume, m3 Design Temperature
10°C 15°C 20°C
VANI
VNi,
VANS
VPA
VTot
4,950
4,980
6,450
660
17,040
3,330
3,410
5,540
660
12,940
2,110
2,380
4,670
660
9,820
With diffused aeration, use .a tapered aeration pattern to match higher O2 demand at the head end of the
nitrification zone. Provide one mixer for each anoxic compartment. For six compartments, each mixer would
be 20 hp.
12. To meet a TP limitation of 1.0 mg/L for the process configuration in this example, the logical process
selection would be to incorporate biological phosphorus removal with nitrogen removal followed by chemical
addition for phosphorus polishing to meet the limit consistently. This process can be accomplished by adding
an anaerobic selector ahead of the first anoxic zone, typically 1-2 hr nominal retention time. The internal
recycle of nitrified flow would continue to be returned to the first anoxic zone. RAS, however, would be
recycled to the head of the anaerobic selector. To operate at maximum efficiency for biological phosphorus
removal, it would be necessary to operate at the minimum 90 necessary to achieve nitrification and at
maximum denitrification efficiency to minimize the return of nitrates to the anaerobic selector. In addition,
the internal recycle rate and anoxic zone volume in use must be carefully monitored to prevent an excessive
anoxic period (i.e., inadequate NOg-N to denitrify) as this can cause an excess release of phosphates—a
secondary release phenomenon, which occurs without storage of BOD.
304
-------
8.5.4 Design Example No. 2 (continued)
Consequently, extensive process monitoring and process control are required to maintain proper recycle
rates, solids inventory, DO control, and sludge blanket levels. Polishing would be accomplished by adding
chemicals such as ferric chloride and alum to the post-aeration zone ahead of the secondary clarifiers
followed by effluent filtration for SS removal to meet the limit of 1 mg/L TP.
To meet the phosphorus removal requirement by chemical precipitation only, the impact on the nitrogen
removal system would primarily result from an increase in sludge production (chemical sludge) and reduced
fraction of volatile solids in the mixed liquor. As the system would be limited in its ability to carry MLSS,
the volume of aerobic and anoxic tankage could increase significantly, as well as pumping requirements
for WAS and RAS.
The use of metal salts to precipitate phosphorus will cause a loss of alkalinity. Therefore, an increase in
supplemental alkalinity addition could be required to maintain a residual alkalinity of 50 mg/L as CaCO3.
8.5.5 Design Example No. 3: Plant A with Less Stringent Limits
From the mass balances in Table 2-15, the inlet wastewater characteristics for the less complex plant (raw
influent plus recycles) and secondary effluent characteristics without nitrification-denitrification are as follows: •
mg/L Equivalents
Characteristic
VSS
TSS
CBODg
TCOD
SCOD
TN
Alkalinity, as CaCO3
Design Conditions:
Reactor temperature = 1 5°C
Reactor MLSS = 3,000 mg/L
MLVSS = 58%
Reactor pH range = 7.0-7.6 .
Aeration
Tank Feed
129
187
152
290
100
30.5
120
Aerobic phase DO
Secondary effluent
Secondary effluent
Secondary effluent
Secondary
Effluent
9
15
3
33
20
25.2
—
= 2.0 mg/L
NHJ-N = 1.0 mg/L
Org-N = 2.0 mg/L
NOg-N = 5.0 mg/L
To meet seasonal limits of 2 mg/L NHJ-N and 10 mg/L TN at maximum monthly loading, design for 1.0
mg/L equivalents NHJ-N and 8.0 mg/L equivalents TN. Secondary effluent Org-N of approximately 2.0 mg/L
equivalents represents the nonbiodegradable fraction of soluble TKN and nitrogen associated with effluent
VSS.
305
-------
8.5.5 Design Example No. 3 (continued)
With raw influent TKN = 30 mg/L and effluent limits of 2 mg/L NHJ-N and 10 mg/L TN, essentially complete
nitrification and a minimum of 67-percent TN removal efficiency are required. To meet these limits, a single
anoxic zone or phased system, such as the cyclical aeration process, can be used. The plant configuration
is as shown in Figure 2-5 for the less complex plant using cyclical nitrogen removal with mechanical aerators
to provide alternating aerobic anoxic cycles within the same basin with mixing during the off-cycle.
Removal Requirements Across Secondary Process:
At average annual loadings (concentrations in mg/L equivalents)
BODS Rem. = (18,925 x 149)71,000 = 2,820 kg (6,217 lb)/d
TN Rem. = [18,925 x (30.5 - 8.0)]/1 ,000 = 426 kg (939)/d
1 R OP'S x *5 "3
TN Rem. in Waste Solids by Assimilation = ' 3 = 100 kg (221 lb)/d
TN Rem. Required by Nitrification-Denitrification = 426 - 100 = 326 kg (719 lb)/d or 17.2 mg/L
NOg-N produced = 30.5 - 5.3 - 2.0 - 1 .0 = 22.2 mg/L
Denitrification required during anoxic phase = 17.2 mg/L
1. Sizing of nitrification or aerobic phase. Similar to Example No. 1, 6{j = 8.5 d for nitrification. Without
separate digestion facilities, however, assume that sludge stabilization is required within the aeration tanks.
For this example, to achieve stabilization, choose GO = 20 d for 10-20°C. Following the same procedure
as in Example No. 1, the following calculation can be made:
qoss = 1/(20 x 0.27) = 0.185 g COD/g MLVSS/d
t = (S0-S|)/(q0BS ' X) = (290 - 20)/(0.185 x 3,000 x 0.58) = 0.84 d
Vaer = 0.84 x 18,925 = 15,900 m3 (4.20 Mgal); [(8.5/20) (15,900)] or 6,760 m3 (1.78 Mgal) required for
nitrification
where Q = 18,925 m3/d was used to determine required aeration volume because t was computed based
upon mg/L equivalent concentrations.
2. Check design parameters.
At Q = 19,379 m3/d (5.12 mgd), the actual retention time t = 19.7 hr
1 fl Q9<^ v 1 W?
F/M = 15700X1.740 " °-1° 9 BOD/S
3. Sizing of anoxic phase for denitrification. Size based on specific denitrification rate, and adjust based
on design temperature and selected PF. For a cyclical system, the denitrification rate can vary between the
rate in a dedicated anoxic basin with internal recycle and with wastewater as the carbon source, and the
rate with endogenous decay. At the beginning of the off-cycle, the unmetabolized wastewater COD level
will be very low. At the midpoint of the off-cycle, where the offcycle duration equals one-half the retention
time in that basin, the COD level will be similar to a dedicated anoxic zone with recycle. With the high
CODfTKN for this example, it is estimated that the overall SDNR will equal the endogenous rate for 25
percent of the anoxic phase and the wastewater carbon rate for 75 percent of the cycle.
Since the quantity of denitrification required is approximately the same as in the first anoxic zone of Example
No. 2 but the rate is less, assume the anoxic F/M will be similar. Therefore, determine SDNR from weighted
average. From the previous example, SDNR2 = 0.011 g NOg-N/g MLVSS/d and SDNR! = 0.061 g NO^-N/g
MLVSS/d. The weighed average is 0.049 g NOi-N/g MLVSS/d.
306
-------
8.5.5 Design Example No. 3 (continued)
Denitrification required = 326 kg (719 lb)/d
For maximum month, MLVSS required = (326 x 1.2 PF)/0.049 = 7,984 kg (17,601 Ib) MLVSS.
VAN = (7,984 x 103)/1 ,740 = 4,589 m3 (1 .21 Mgal)
Check anoxic F/M:
F/M = (18,925)(152)/[(4,589)(1,740)] = 0.36, which is approximately equal to assumed 0.34 from
Example No. 1 for trial-and-error solution.
Calculate actual hydraulic retention time for anoxic phase:
At Q = 19,379 m3/d (5.12 mgd), t = 5.7 hr
At Qmax mo, t = 3.8 hr
4. Determine ratio of anoxic and nitrification periods (cycling of on/off periods).
VNit = 6,760 m3 (1 .78 Mgal)
VAN = 4,589 m3 (1 .21 Mgal)
On/off ratio = 3:2
With two trains and three cycled compartments and one final continuously aerated compartment per train,
the retention time in each cycled compartment is approximately 4.69 hr. Off-cycle duration should equal at
least one-half the retention time in each compartment to maintain high average COD:TKN during the anoxic
cycle. The basins following the cycled basins would be aerated continuously for sludge stabilization.
Off-cycle = 2.4 hr
On-cycle = 3.6 hr
5. Determine RAS rate similar to previous example. No internal recycle is required if step-feeding to
downstream compartments is provided to allow use of wastewater carbon as the primary carbon source for
denitrification.
6. Determine alkalinity requirements as in Example No. 1.
7. Determine mixing requirements during anoxic phase. Mixers are desirable to obtain full liquids-solids
contact while the air is off and to distribute raw wastewater carbon source for denitrification particularly for
downstream compartments with step feed.
8. Calculate waste sludge and aeration requirements similar to previous examples. The aeration rate
required for each cycled basin, however, should be increased by 67 percent to account for the off-cycle.
The aeration required for the final basins would be based on sludge stabilization requirements.
9. A typical flow configuration is shown in Figure 8-33. The first three basins in each train would have
cycled aeration and the final basin would be continuously aerated. Step feeding of influent is provided to
each of the three cycled basins. The following summarizes reactor volumes at 10°C, 15°C, and 20°C for
comparisons.
307
-------
8.5.5 Design Example No. 3 (continued)
Volume, m3 Design Temperature
10°C 15°C 20°C
VNB
VAN
VPA
VTO,
9,870
6,400
6,020
22,290
6,760
4,590
9,140
20,490
4,250
3,100
11,640
18,990
The total volume of each cycled compartment is the sum of VNit and VAn and the total volume required for
sludge stabilization is the sum of VNrt and VPA. The total volume for sludge stabilization should be based
on state and federal guidelines or requirements for stabilization at various temperatures.
Step Feed
RAS
I
Primary Effluent
with Recycles
To
Secondary
Clarifiers
Step Feed
AnoxicZone
Aerobic Zone
Sludge Stabilization
Mechanical Aerator
Mechanical Aerator and Mixer
Rgure 8-33. Design Example No. 3—CNR process.
308
-------
8.5.6 Design Example No.4: Plant A with More .Stringent Limits
To meet the more stringent limitations with the less complex plant configuration, again a dual anoxic zone
type of process configuration would be required similar to that in Example No. 2 as shown in Figure 8-34.
The procedures for sizing would be identical to those used in Example No. 2 except the total aerobic volume
would be governed by the volume required to achieve sludge stabilization. The volume required is summa-
rized for various temperatures as follows:
Volume, rri3 Design Temperature
10°C 15°C 20°C
VANI
VAer
VANS
VPA
5,040
9,870
8,370
6,020
29,300
3,390
6,760
. 7,190
9,140
26,480
2,150
4,250
6,060
11,640
24,100
RAS
i
Primary
Effluent
with Recycles
First Anoxic Zone
Second Anoxic Zone
h • \ Nitrification Zone
ESjgjj Sludge Stabilization
;<<««<«5>>>>S «
4V«
-------
8.6 References
When an NTIS number is cited in a reference, that docu-
ment Is available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
703-487-4650
1. Applegate, C.S., B. Wilder, and J.R. DeShaw. 1980.
Total nitrogen removal in a multichemical oxidation
system. JWPCF 52(3):568-577.
2. Ludzack, F.J., and M.B. Ettinger. 1962. Controlling
operation to minimize activated sludge effluent nitro-
gen. JWPCF 34(9).
3. Christensen, M.H. 1975. Denitrification of sewage by
alternating process operation. Prog. Water Tech*.
7(2):339-347.
4. Barnard, J.L. 1973. Biological denitrification. Water
Poll. Control (G.B.) 72(6):705-720.
5. Krauth, K. 1982. Practical experiences with nitrifica-
tion and denitrification processes. Proc. Japanese-
German Workshop on Wastewater and Sludge
Treatment, Tsukuba (October).
6. Kuribayashi, M. 1982. Study of nitrified liquor recy-
cled biological nitrification-denitrification processes.
Proc. Japanese-German Workshop on Wastewater
and Sludge Treatment, Tsukuba (October).
7. Mtyaji, Y, et al. 1980. Biological nitrogen removal by
step feed process. Prog. Water Tech. 12(6):193-202.
8. Daigger, G.T., and G.A. Nicholson. 1990. Perform-
•ance of four full-scale nitrifying wastewater treatment
plants incorporating selectors. Res. JWPCF
62(5):676-683.
9. Bowker, R.P.G., J.M. Smith, and H. Shah. 1988. As-
sessment of cost and effectiveness of biological dual
nutrient removal technologies in the Chesapeake Bay
drainage basin: Volume 1. EPA CBP/TRS 17. An-
napolis, MD.
10. Sutton, P.M., and Bridle, T.R. 1980. Biological nitro-
gen control of industrial wastewater. Water—1980
AICHE Symposium Series, 177.
11. Daigger, G.T., G.D. Waltrip, E.D. Romm, and L.M.
Morales. 1988. Enhanced secondary treatment incor-
porating biological nutrient removal. JWPCF
60(10): 1833-1842.
12. Heidman, J.A., I.J. Kugelman, and E.F. Barth. 1976.
Plug-flow single-stage nitrification-denitrification acti-
vated sludge. Presented at the 49th Annual WPCF
Conference, Minneapolis, MN.
13. Randall, C.W., J.L. Barnard, and H.D. Stensel. 1992.
Design and retrofit of wastewater treatment plants for
biological nutrient removal. Volume 5. Water Quality
Management Library. Lancaster, PA: Technomic Pub-
lishing Company, Inc.
14. Bishop, D.F., J.A. Heidman, and J.B. Stamberg.
1976. Single-stage nitrification-denitrification. JWPCF
48(3):520-532.
15. Schwinn, D.E., D.F. Storrier, and D.G. Thome. 1977.
Full-scale operation of a single-stage nitrification-
denitrification plant. EPA/600/12-77/088. Cincinnati, OH.
16. Schwinn, D.E. 1988. Final report on nitrification-de-
nitrification research at the water pollution control
plant, Town of Barnstable, MA. Stearns & Wheler
Engineers and Scientists.
17. Schreiber, Inc. 1991. The Simultech nutrient removal
system. Schreiber Corporation, Inc., Trussville, AL.
18. Matsche, N.F. 1972. The elimination of nitrogen in
the treatment plant of Vienna-Blumenthal. Water Res.
6:485-486.
19. Rittman, B.E., and W.E. Langeland. 1985. Simulta-
neous denitrification with nitrification in single-chan-
nel oxidation ditches. JWPCF 57(4):300-308.
20. Koot, A.C.J., and J. Zeper. 1972. Carrousel, a new
type of aeration-system with low organic load. Water
Res. 6:401-406.
21. Drews, R.J.L.C., W.M. Malan, P.G.J. Meiring, and B.
Moffaltt. 1972. The Orbal extended aeration activi-
ated sludge plant. JWPCF 44(2):221-231.
22. Matsche, N.F, and G. Spatzierer. 1975. Austrian
plant knocks out nitrogen. Water and Was. Eng.
12(1):19-24 (January).
23. Drews, R.J.L.C., and A.M. Greef. 1973. Nitrogen
elimination by rapid alternation of aerobic/"anoxic"
conditions in "Orbal" activated sludge plants. Water
Res. 7:1183-1194.
24. VanderGeest, A.T., and W.C. Witvoet. 1977. Nitrifica-
tion and denitrification in Carrousel systems, Prog.
Water Tech. 8:653.
25. U.S. EPA. 1978. A comparison of oxidation ditch p|ant
to competing processes for secondary and advanced
treatment of municipal wastes. EPA/600/2-78/051.
Cincinnati, OH. '•
26. Bundgaard, E., and F.M. Nielsen. 1989. Stability of
effluents from biological nutrient removal plants—
Danish long-term operating and optimization experi-
ence. Presented at the 62nd Annual WPCF
Conference, San Francisco, CA. •
310
-------
27. Norcross, K.L. 1986. Sequencing batch reactors: an
overview. Proceedings of the 6th Annual Kansas En-
gineering Conference, Lawrence, KS.
28. Powell, E.O., and J.R. Lowe. 1962. Theory of multi-
stage continuous cultures. Proceedings of 2nd Sym-
posium on Continuous Cultivation of Organisms,
Prague, Czechoslovakia.
29. Inomae, K., H. Araki, K. Koya, Y. Awoya, T. Kusuda,
and I. Matsuo. 1987. Nitrogen removal in an oxidation
ditch with intermittent aeration. Water Sci. Tech., Vol.
19.
30. Ardern, E., and T.W. Lockett. 1914. Experiments on
the oxidation of sewage without the aid of filters. J.
Soc. Chem. Ind. 33(10):523.
31. Austgen Biojet Waste Water Systems, Inc. 1991. De-
sign Manual, Version 1.1, No. 44 (May).
32. Goronszy, M.C-1990. CASS, cyclic activated sludge
system: superior batch reactor technology, Transen-
viro, Inc. Aliso Viejo, CA.
33. Shubert, W.M. 1986. SCR-sequencing batch reactor.
Design manual for Aqua-Aerobic Systems, Inc.
34. Deeny, K.J., J.A. Heidman, W.W. Schuk, K.S. Young,
and A.J. Condren. 1991. Implementation of sequenc-
ing batch reactor technologies in the United States.
Presented at the 64th Annual WPCF Conference,
Toronto, Ontario (October).
35. U.S. EPA. 1986. Sequencing batch reactors. Summary
report for the Center for Environmental Research Infor-
mation. EPA/625/8-86/011. Cincinnati, OH.
36. Henderson and Bodwell Consulting Engineers. 1990.
Sequencing batch reactor wastewater treatment sys-
tems, Plainview, NY.
37. Goronszy, M.C. 1987. Nitrogen removal and sludge
.bulking control in cyclically operated activated sludge
systems. Proceedings of the 61st Meeting of Ohio
^Water^Pollution Control Association, Akron, OH.
38. Daigger, G.T., M.H. Robbins, Jr., and B.R. Marshall.
1985. The design of a selector to control low F:M
filamentous bulking. JWPCF 57(3).
39. Bowker, R.P.G. and H.D. Stensel. 1987. Phosphorus
removal: design manual. EPA/625/1-87/001. Cincin-
; nati, OH.
40. Soap and Detergent Association. 1991. Principles
and practice of phosphorus and nitrogen removal
from municipal wastewater. New York, NY.
"!'••
41. Goronszy, M.C., and J. White. 1988. Biological nutri-
ent removal with sludge bulking control in a batch
activated system. Proceeding of the 14th Biennial
International Association on Water Pollution Control
and Research Conference, Brighton, UK.
42. Rensik, J.H. 1974. New approach to preventing
sludge bulking. JWPCF 46:1888.
43. Linne, S.R., and S.C. Chiesa. 1987. Operational vari-
ables affecting performance of the selector—com-
plete mix activated sludge process. JWPCF 59(7).
44. Chudoba, J., et al. 1973. Control of activated sludge
filamentous bulking: I. Effect of hydraulic regime or
degree of mixing in an aeration tank. Water Res.
7:1163.
45. Chudoba, J., et al. 1973. Control of activated sludge
filamentous bulking: II. Selection of microorganisms
by means of a selector. Water Res. 7:1389.
46. Chudoba, J., et al. 1974. Control of activated sludge
filamentous bulking: III. Effect of sludge loading.
Water Res. 8:231.
8.7 Bibliography
Albertson, O.E. 1987. The control of building sludges:
from early innovators to current practice. JWPCF
59(4):172.
Irvine, R.L., D.V.S. Murphy, M.L. Arora, J.L. Copeman,
and J.A. Heidman. 1987. Analysis of full-scale SBR op-
eration at Grundy Center, Iowa. JWPCF 59(3):132-138.
Sen, D., C.W. Randall, and T.J. Grizzard. 1992. Critical
design and process control features to optimize biological
nutrient removal in temperature climates. Proceedings of
7th IAWPRC Conference, Washington, DC.
Sen, D., et al. 1991. Impact of solids handling units on
the performance of biological phosphorus and nitrogen
'removal systems. Presented at the 54th Annual WPCF
Conference, Toronto, Canada.
Sen, D., et al. 1991. Process design and operational
modifications of oxidation ditches for biological nutrient
removal. Proceedings of 6th IAWPRC Conference,
Praque, Czechoslovakia.
Sutton, P.M., K.L. Murphy, and B.E. Jank. 1977. Nitrogen
pontrol: a basis for design with activated sludge systems.
Prog. Water Tech. 8(4/5).
U.S. EPA. 1991. BNR plant assessment interim data sum-
mary report. Office of Wastewater Enforcement and Com-
pliance. Prepared by HydroQual, Inc., New Jersey
(October).
U.S. EPA. 1975. Process design manual for nitrogen control.
EPA/625/1-77/007 (NTIS PB-259149). Washington, DC.
Water Research Commission. 1984. Theory, design, and
operation of nutrient removal activated sludge processes.
South Africa.
U.S. GOVERNMENT PRINTING OFFICE: 1994-550-001/00161
311
-------
------- |