PROCESS DESIGN MANUAL
FOR
NITROGEN CONTROL
U.S. ENVIRONMENTAL PROTECTION AGENCY
Technology Transfer
October 1975
-------
ACKNOWLEDGEMENTS
This design manual was prepared for the Office of Technology Transfer of the U.S.
Environmental Protection Agency. Coordination and preparation of the manual was carried
out by the firm of Brown and Caldwell, Walnut Creek, California, under the direction of
Denny S. Parker with assistance from Richard W. Stone and Richard J. Stenquist. Chapters
7, 8, and portions of Chapter 9 were prepared by Gordon Gulp of Culp/Wesner/Culp.
Clair N. Sawyer and Perry L. McCarty served as consultants to the U.S. EPA for the purpose
of reviewing portions of the text. U.S. EPA reviewers were Edwin F. Earth and Irwin J.
Kugelman of the U.S. EPA National Environmental Research Center, Cincinnati, Ohio, and
Robert S. Madancy of the Office of Technology Transfer, Washington, D.C.
NOTICE
The mention of trade names of commercial products in this publication is for illustration
purposes and does not constitute endorsement or recommendation for use by the U.S.
Environmental Protection Agency.
11
-------
ABSTRACT
This manual presents theoretical and process design criteria for the implementation of
nitrogen control technology in municipal wastewater treatment facilities. Design concepts
are emphasized as much as possible through examination of data from full-scale and pilot
installations.
Design data are included on biological nitrification and denitrification, breakpoint
chlorination, ion exchange and air stripping. One chapter presents the concepts involved in
assembling various unit processes into rational treatment trains and presents actual case
examples of specific treatment systems that incorporate nitrogen control processes.
111
-------
CONTENTS
Chapter Page
ACKNOWLEDGEMENTS ii
ABSTRACT iii
CONTENTS v
LIST OF FIGURES xviii
LIST OF TABLES xxiii
FOREWORD xxvii
1 INTRODUCTION 1-1
1.1 Background and Purpose 1-1
1.2 Scope of the Manual 1-2
1.3 Guide to the User 1-2
2 NITROGENOUS MATERIALS IN THE ENVIRONMENT AND
THE NEED FOR CONTROL IN WASTEWATER EFFLUENTS 2-1
2.1 Introduction 2-1
2.2 The Nitrogen Cycle 2-1
2.2.1 The Nitrogen Cycle in Surface Waters and Sediments 2-5
2.2.2 The Nitrogen Cycle in Soil and Groundwater 2-5
2.3 Sources of Nitrogen 2-8
2.3.1 Natural Sources 2-8
2.3.2 Man-caused Sources 2-9
2.4 Effects of Nitrogen Discharge 2-12
2.4.1 Biostimulation of Surface Waters 2-12
2.4.2 Toxicity 2-13
2.4.3 Effect on Disinfection Efficiency 2-13
2.4.4 Dissolved Oxygen Depletion in Receiving Waters 2-14
2.4.5 Public Health 2-14
2.4.6 Water Reuse 2-16
-------
CONTENTS - Continued
Chapter Page
2.5 Treatment Processes for Nitrogen Removal 2-16
2.5.1 Conventional Treatment Processes 2-16
2.5.2 Advanced Wastewater Treatment Processes 2-17
2.5.3 Major Nitrogen Removal Processes 2-17
2.5.3.1 Biological Nitrification-Denitrification 2-18
2.5.3.2 Breakpoint Chlorination 2-18
2.5.3.3 Selective Ion Exchange for Ammonium Removal 2-19
2.5.3.4 Air Stripping for Ammonia Removal 2-19
2.5.4 Other Nitrogen Removal Processes 2-20
2.5.5 Summary 2-20
2.6 References 2-21
3 PROCESS CHEMISTRY AND BIOCHEMISTRY OF
NITRIFICATION AND DENITRIFICATION 3-1
3.1 Introduction 3-1
3.2 Nitrification 3-1
3.2.1 Biochemical Pathways 3-1
3.2.2 Energy and Synthesis Relationships 3-2
3.2.3 Alkalinity and pH Relationships 3-4
3.2.4 Oxygen Requirements 3-6
3.2.5 Kinetics of Nitrification 3-6
3.2.5.1 Effect of Ammonia Concentration on Kinetics 3-6
3.2.5.2 Relationship of Growth Rate to Oxidation Rate 3-7
3.2.5.3 Relationship of Growth Rate to Solids
Retention Time 3-8
3.2.5.4 Kinetic Rate Constants for Temperature and
Nitrogen Concentration 3-8
3.2.5.5 Effect of Dissolved Oxygen on Kinetics 3-12
3.2.5.6 Effect of pH on Kinetics 3-13
3.2.5.7 Combined Kinetic Expressions 3-14
VI
-------
CONTENTS - Continued
Chapter » Page
3.2.6 Population Dynamics 3-17
3.2.7 Nitrification Rates in Activated Sludge 3-21
3.2.8 Nitrification Rates in Trickling Filters and Other
Attached Growth Systems 3-27
3.2.9 Effect of Inhibitors on Nitrification 3-27
3.3 Denitrification 3-29
3.3.1 Biochemical Pathways 3-29
3.3.2 Energy and Synthesis Relationships 3-30
3.3.3 Alkalinity and pH Relationships 3-34
3.3.4 Alternative Electron Donors 3-34
3.3.5 Kinetics of Denitrification 3-36
3.3.5.1 Effect of Nitrate on Kinetics 3-36
3.3.5.2 Relationship of Growth Rate to Removal Rate 3-36
3.3.5.3 Solids Retention Time 3-37
3.3.5.4 Kinetic Constants for Denitrification 3-37
3.3.5.5 Effect of Carbon Concentration on Kinetics 3-40
3.3.5.6 Effect of pH on Kinetics 3-41
3.3.5.7 Combined Kinetic Expression 3-42
3.3.6 Effect of DO on Denitrification Inhibition 3-43
3.4 References 3-43
4 BIOLOGICAL NITRIFICATION 4-1
4.1 Introduction 4-1
4.2 Classification of Nitrification Processes 4-1
4.3 Combined Carbon Oxidation-Nitrification in Suspended
Growth Reactors 4-2
4.3.1 Activated Sludge Modifications 4-2
4.3.1.1 Complete Mix Plants 4-2
4.3.1.2 Extended Aeration Plants 4-4
4.3.1.3 Conventional or Plug Flow Plants 4-4
Vll
-------
CONTENTS - Continued
Chapter Page
4.3.1.4 Contact Stabilization Plants 4-4
4.3.1.5 Step Aeration and Sludge Reaeration Plants 4-6
4.3.1.6 High Rate and Modified Activated Sludge 4-6
4.3.1.7 High Purity Oxygen Activated Sludge Plants 4-6
4.3.2 Utility of Nitrification Kinetic Theory in Design 4-7
4.3.3 Complete Mix Activated Sludge Kinetics 4-7
4.3.3.1 Effect of Temperature and Safety Factor on
Design 4-13
4.3.3.2 Consideration in the Selection of SF 4-13
4.3.4 Extended Aeration Activated Sludge Kinetics 4-20
4.3.5 Conventional Activated Sludge (Plug Flow) Kinetics 4-20
4.3.5.1 Considerations in the Selection of the Safety
Factor 4-23
4.3.5.2 Kinetic Design Approach 4-23
4.3.6 Contact Stabilization Activated Sludge Kinetics 4-23
4.3.6.1 Design Example 4-24
4.3.7 Step Aeration Activated Sludge Kinetics 4-30
4.3.8 Operating Experience with Combined Carbon Oxidation-
Nitrification in Suspended Growth Reactors 4-30
4.3.8.1 Step Aeration Activated Sludge In a Moderate
Climate 4-30
4.3.8.2 Step Aeration Activated Sludge in a Rigorous
Climate 4-32
4.3.8.3 Conventional Activated Sludge In a Rigorous
Climate 4-32
4.4 Combined Carbon Oxidation-Nitrification In Attached Growth
Reactors 4-35
4.4.1 Nitrification with Trickling Filters in Combined Carbon
Oxidation-Nitrification Applications 4-35
viii
-------
CONTENTS - Continued
Chapter Page
4.4.1.1 Media Selection 4-35
4.4.1.2 Organic Loading Criteria 4-37
4.4.1.3 Effect of Media Type on Allowable Organic
Loading 4-39
4.4.1.4 Effect of Recirculation on Nitrification 4-40
4.4.1.5 Effect of Temperature on Nitrification 4-40
4.4.1.6 Effect of Diurnal Loading on Performance 4-41
4.4.2 Nitrification with the Rotating Biological Disc Process in
Combined Carbon Oxidation-Nitrification Applications 4-41
4.4.2.1 Loading Criteria for Nitrification 4-43
4.4.2.2 Effect of Temperature 4-43
4.4.2.3 Effect of Diurnal Load Variations 4-43
4.5 Pretreatment for Separate Stage Nitrification 4-45
4.5.1 Effects of Pretreatment by Chemical Addition 4-46
4.5.2 Effects of Degree of Organic Carbon Removal 4-50
4.5.3 Protection Against Toxicants 4-51
4.6 Separate Stage Nitrification with Suspended Growth Processes 4-52
4.6.1 Application of Nitrification Kinetic Theory to Design 4-52
4.6.2 Solids Retention Time Approach 4-53
4.6.2.1 Choice of Process Configuration 4-53
4.6.2.2 Choice of the Safety Factor 4-54
4.6.3 Nitrification Rate Approach 4-55
4.6.4 Effect of the BODs/TKN Ratio on Sludge Inventory
Control 4-57
4.6.5 Comparison of the Use of Conventional Aeration to
the Use of High Purity Oxygen 4-58
4.6.5.1 High Purity Oxygen Nitrification With and
Without pH Control 4-59
IX
-------
CONTENTS - Continued
Chapter Page
4.6.5.2 Comparison of Conventional Aeration and
High Purity Oxygen at the Same pH 4-59
4.7 Separate State Nitrification with Attached Growth Processes 4-61
4.7.1 Nitrification with Trickling Filters 4-62
4.7.1.1 Media Type and Specific Surface 4-62
4.7.1.2 Loading Criteria 4-62
4.7.1.3 Effect of Recirculation 4-66
4.7.1.4 Effluent Clarification 4-66
4.7.1.5 Effect of Diurnal Load Variations 4-66
4.7.1.6 Design Example 4-68
4.7.2 Nitrification with the Rotating Biological Disc Process 4-69
4.7.2.1 Kinetics 4-70
4.7.3 Nitrification with Packed-Bed Reactors 4-72
4.7.3.1 Oxygenation Techniques 4-73
4.7.3.2 Media Type, Backwashing and Loading Criteria 4-73
4.8 Aeration Requirements 4-76
4.8.1 Adaptability of Alternative Aeration Systems to Diurnal
Variations in Load 4-79
4.8.2 Oxygen Transfer Requirements 4-80
4.8.3 Example Sizing of Aeration Capacity 4-84
4.9 pH Control 4-85
4.9.1 Chemical Addition and Dose Control 4-86
4.9.2 Effect of Aeration Method on Chemical Requirements 4-86
4.10 Solids-Liquid Separation 4-89
4.11 Considerations for Process Selection 4-94
4.11.1 Comparison to Physical-Chemical Alternatives 4-94
-------
CONTENTS - Continued
Chapter Page
4.11.2 Choice Among Alternative Nitrification Systems 4-94
4.12 References 4-99
5 BIOLOGICAL DENITRIFICATION 5-1
5.1 Introduction 5-1
5.2 Denitrification in Suspended Growth Reactors Using
Methanol as the Carbon Source 5-1
5.2.1 Denitrification Rates 5-3
5.2.2 Complete Mix Denitrification Kinetics 5-4
5.2.2.1 Effect of Safety Factor on Steady-State
Effluent Quality 5-8
5.2.2.2 Effect of Diurnal Load Variations on Effluent
Quality 5-8
5.2.3 Plug Flow Denitrification Kinetics 5-10
5.2.4 Effluent Quality from Suspended Growth
Denitrification Processes 5-12
5.2.4.1 Experience at Manassas, Va. 5-12
5.2.4.2 Experience at the CCCSD's Advanced
Treatment Test Facility. 5-13
5.3 Denitrification in Attached Growth Reactors Using Methanol
as the Carbon Source 5-15
5.3.1 Kinetic Design of Attached Growth Denitrification
Systems 5-15
5.3.2 Classification of Column Configurations 5-17
5.3.2.1 Nitrogen Gas Filled Denitrification Columns -
Packed Bed 5-17
5.3.2.2 Submerged High Porosity Media Columns -
Packed Bed 5-22
5.3.2.3 Submerged Low Porosity Fine Media
Columns - Packed Bed Configuration 5-23
XI
-------
CONTENTS - Continued
Chapter Page
5.3.2.4 Submerged High Porosity Fine Media Columns -
Fluidized Bed 5-29
5.3.2.5 Comparison of Attached Growth
Denitrification Systems 5-32
5.4 Methanol Handling, Storage, Feed Control, and Excess
Methanol Removal 5-32
5.4.1 Properties of Methanol 5-33
5.4.2 Standards for Shipping, Unloading, Storage and
Handling 5-33
5.4.3 Methanol Delivery and Unloading 5-34
5.4.4 Methanol Storage 5-36
5.4.5 Transfer and Feed 5-37
5.4.6 Methanol Feed Control 5-37
5.4.7 Excess Methanol Removal 5-38
5.5 Combined Carbon Oxidation-Nitrification-Denitrification
Systems with Wastewater and Endogenous Carbon Sources 5-39
5.5.1 Systems Using Endogenous Respiration in a Sequential
Carbon Oxidation-Nitrification-Denitrification System 5-39
5.5.2 Systems Using Wastewater Carbon in Alternating
Aerobic/Anoxic Modes 5-42
5.5.2.1 Aerobic/Anoxic Sequences in Oxidation Ditches 5-42
5.5.2.2 Denitrification in an Alternating Contact Process 5-48
5.5.2.3 The Bardenpho Process 5-49
5.5.2.4 Alternating Aerobic/Anoxic System Without
Internal Recycle 5-52
5.5.2.5 Kinetic Design of Alternating Aerobic/Anoxic
Systems 5-55
5.6 Solids-Liquid Separation 5-58
5.7 Considerations for Process Selection 5-60
5.7.1 Comparison to Physical-Chemical Alternatives 5-61
5.7.2 Choice Among Alternative Denitrification Systems 5-61
5.8 References 5-64
xii
-------
CONTENTS - Continued
Chapter Page
6 BREAKPOINT CHLORINATION 6-1
6.1 Process Chemistry 6-1
6.1.1 Chemical Stoichiometry 6-1
6.1.2 The Breakpoint Curve 6-5
6.2 Process Application Considerations 6-5
6.2.1 Chlorine Dosage Requirement 6-6
6.2.1.1 Effect of Pretreatment 6-6
6.2.1.2 Effect of pH and Temperature 6-7
6.2.1.3 Initial Mixing of Chlorine 6-7
6.2.2 Residual Nitrogenous Materials 6-9
6.2.3 Alkalinity Supplementation 6-12
6.2.4 Reaction Rates 6-12
6.2.5 Effect on Total Dissolved Solids 6-13
6.2.6 Reactions with Organic Nitrogen 6-14
6.2.7 Disinfection 6-15
6.3 Process Control Instrumentation 6-15
6.3.1 Process Control System 6-15
6.3.1.1 Chlorine Dosage Control 6-15
6.3.1.2 pH Control 6-18
6.3.2 Process Control Components 6-18
6.4 Dechlorination Techniques 6-18
6.4.1 Sulphur Dioxide Dechlorination 6-19
6.4.1.1 Stoichiometry 6-19
6A.I.2 Reaction Rates 6-20
6.4.1.3 Significance of Sulphur Dioxide Overdose 6-20
6.4.1.4 Process Application and Control 6-20
xiii
-------
CONTENTS - Continued
Chapter Page
6.4.2 Activated Carbon Dechlorination 6-21
6.4.2.1 Stoichiometry 6-21
6.4.2.2 Process Application 6-22
6.5 Design Example 6-22
6.6 Considerations for Process Selection 6-24
6.7 References 6-25
7 SELECTIVE ION EXCHANGE FOR AMMONIUM REMOVAL 7-1
7.1 Chemistry and Engineering Principles 7-1
7.1.1 Basic Concept 7-1
7.1.2 Ion Exchange Principles 7-2
7.1.3 Properties of Clinoptilolite 7-3
7.1.3.1 Selectivity 7-3
7.1.3.2 Mineralogical Classification 7-5
7.1.3.3 Total Exchange Capacity 7-7
7.1.3.4 Chemical Stability 7-7
7.1.3.5 Physical Stability 7-8
7.1.3.6 Density 7-9
7.2 Major Service Cycle Variables 7-9
7.2.1 pH 7-9
7.2.2 Hydraulic Loading Rate 7-9
7.2.3 Clinoptilolite Size 7-9
7.2.4 Pretreatment 7-10
7.2.5 Wastewater Composition 7-10
7.2.6 Length of Service Cycle 7-10
7.2.7 Bed Depth 7-11
7.2.8 One Column vs. Series Column Operation 7-12
7.2.9 Determination of Ion Exchanger Size 7-14
7.3 Regeneration Alternatives 7-16
7.3.1 Basic Concepts 7-16
xiv
-------
CONTENTS - Continued
Chapter Page
7.3.2 Regeneration Process 7-17
7.3.2.1 High pH Regeneration 7-17
7.3.2.2 Neutral pH Regeneration 7-17
7.3.2.3 Effects on Effluent TDS 7-19
7.3.3 Regenerant Recovery Systems 7-19
7.3.3.1 Air Stripping of High pH Regenerant 7-19
7.3.3.2 Air Stripping of Neutral pH Regenerant 7-22
7.3.3.3 Steam Stripping 7-24
7.3.3.4 Electrolytic Treatment of Neutral pH
Regenerant 7-26
7.4 Considerations in Process Selection 7-27
7.5 References 7-28
8 AIR STRIPPING FOR NITROGEN REMOVAL 8-1
8.1 Chemistry and Engineering Principles 8-1
8.1.1 Basic Concept 8-1
8.2 Environmental Considerations 8-1
8.2.1 Air Pollution 8-2
8.2.2 Washout of Ammonia from the Atmosphere 8-4
8.2.3 Noise 8-5
8.3 Stripping Tower System Design Considerations 8-5
8.3.1 Type of Stripping Tower 8-5
8.3.2 pH 8-5
, 8.3.3 Temperature 8-6
8.3.4 Hydraulic Loading 8-7
8.3.5 Tower Packing 8-9
8.3.5.1 Packing Depth 8-9
8.3.5.2 Packing Material and Shape 8-9
xv
-------
CONTENTS - Continued
Chapter Page
8.3.5.3 Packing Spacing and Configuration 8-10
8.3.6 Air Flow 8-11
8.3.7 Scale Control 8-12
8.4 Ammonia Recovery or Removal From Off-Gases 8-13
8.4.1 Acid Systems 8-13
8.4.2 Nitrification-Denitrification 8-15
8.5 Stripping Ponds 8-17
8.6 Considerations in Process Selection 8-19
8.7 References 8-20
9 TOTAL SYSTEM DESIGN 9-1
9.1 Introduction 9-1
9.2 Influence of Effluent Quality Objectives on Total System
Design 9-1
9.3 Other Considerations in Process Selection 9-4
9.4 Interrelationships with Phosphorus Removal 9-4
9.4.1 Alternative Systems 9-5
9.4.2 Considerations in System Selection 9-8
9.4.2.1 Phosphorus Removals Obtainable 9-8
9.4.2.2 Impacts on Sludge Handling 9-9
9.4.2.3 Reliability 9-12
9.4.2.4 Flexibility of Operation in Multipurpose
Treatment Units 9-12
9.4.2.5 Cost 9-12
9.5 Case Examples 9-12
9.5.1 Case Examples of Nitrification for Ammonia Reduction 9-13
9.5.1.1 Jackson, Michigan 9-13
9.5.1.2 Valley Community Services District, California 9-16
xvi
-------
CONTENTS - Continued
Chapter Page
9.5.1.3 Livermore, California 9-22
9.5.1.4 San Pablo Sanitary District, California 9-23
9.5.2 Case Examples of Nitrification-Denitrification for
Nitrogen Removal 9-29
9.5.2.1 Central Contra Costa Sanitary District, California 9-29
9.5.2.2 Canberra, Australia 9-39
9.5.2.3 Washington, D.C. 9-40
9.5.2.4 El Lago, Texas 9-49
9.5.3 Case Examples of Breakpoint Chlorination for Nitrogen
Removal 9-55
9.5.3.1 Sacramento, California 9-55
9.5.3.2 Montgomery County, Maryland 9-62
9.5.4 Case Examples of Selective Ion Exchange for Nitrogen
Removal 9-70
9.5.4.1 Upper Occoquan Sewage Authority, Va. 9-70
9.5.4.2 Rosemount, Minnesota 9-77
9.5.5 Case Examples of Air Stripping for Nitrogen Removal 9-81
9.5.5.1 South Lake Tahoe, California 9-81
9.5.5.2 Orange County Water District, California 9-87
9.6 References 9-91
APPENDIX A - GLOSSARY OF SYMBOLS A-1
APPENDIX B - METRIC EQUIVALENTS B-l
xvu
-------
LIST OF FIGURES
Figure No. Page
2-1 The Nitrogen Cycle 2-2
2-2 The Nitrogen Cycle in Surface Water 2-6
2-3 The Nitrogen Cycle in Soil and Groundwater 2-7
2-4 Allowable Effluent Discharge into the Thames Estuary 2-15
3-1 Temperature Dependence of the Maximum Growth Rates of Nitrifiers 3-9
3-2 Temperature Dependence of the Half Saturation Constants of Nitrifiers 3-9
3-3 Comparison of Effect of Temperature on Nitrification in Suspended
Growth and Attached Growth Systems 3-11
3-4 Effect of Dissolved Oxygen on Nitrification Rate 3-13
3-5 Effect of pH on Nitrification Rate 3-15
3-6 Effect of Solids Retention Time on Effluent Ammonia Concentration
and Nitrification Efficiency 3-20
3-7 Effect of Ammonia Concentration on Nitrification Rate 3-24
3-8 Effect of Temperature and Fraction of Nitrifiers on Nitrification Rate 3-25
3-9 Effect of BODs/TKN Ratio on Nitrification Rate - Experimental
Attached Growth System • 3-26
3-10 Effect of Temperature on Denitrification Rate 3-39
3-11 Effect of pH on Denitrification Rate 3-41
4-1 Modifications of the Activated Sludge Process 4-5
4-2 Effect of the Safety Factor on Steady State Effluent Ammonia Levels
in Suspended Growth Systems 4-16
4-3 Diurnal Variations at the Chapel Hill, N.C. Treatment Plant 4-17
4-4 Effect of SF on Diurnal Variation in Effluent Ammonia 4-19
4-5 DO and Ammonia Nitrogen Profile in a Plug-Flow System 4-21
4-6 Nitrification Efficiency as a Function of Process Parameters 4-25
4-7 Effect of Organic Load on Nitrification Efficiency of Rock Trickling
Filters 4-38
4-8 A Typical Rotating Biological Disc Process 4-42
4-9 Effect of BOD5 Concentration and Hydraulic Load on Nitrification
in the RBD Process 4-44
4-10 Temperature Correction Factor for Nitrification in the RBD Process 4-45
4-11 Pretreatment Alternatives for Separate Stage Nitrification 4-46
4-12 Rancho Cordova Wastewater Treatment Facility Effluent Ammonia
Characteristics, March 19-20, 1974 4-54
4-13 Observed Nitrification Rates at Various Locations 4-56
4-14 Covered High Purity Oxygen Reactor with Three Stages and
Mechanical Aerators 4-58
4-15 Surface Area Requirements for Nitrification - Midland Michigan 4-64
XVlll
-------
LIST OF FIGURES - Continued
Figure No. Page
4-16 Surface Area Requirements for Nitrification - Lima, Ohio 4-65
4-17 Surface Area Requirements for Nitrification - Sunnyvale, California 4-67
4-18 Nitrification Rates as a Function of Stage Effluent Concentration 4-70
4-19 Design Relationships for a 4-Stage RED Process Treating Secondary
Effluent 4-71
4-20 Schematic Diagram of a Packed-Bed Reactor 4-72
4-21 Temperature Dependence of Detention Time for Complete
Nitrification, (<2 mg/1 NH|-N) at Steady State in the PER 4-74
4-22 Relation Between Ammonia Peaking and Hydraulic Peaking Loads for
Treatment Plants with No In-Process Equalization 4-77
4-23 Relationship of Maximum/Minimum Nitrogen Load Ratio to
Maximum/Average Flows 4-78
4-24 Relationship of Aeration Air Requirements for Oxidation of
Carbonaceous BOD and Nitrogen 4-83
4-25 Effect of Temperature on Thickening Properties of Oxygen Activated
Sludge at MLSS = 4000 mg/1 4-91
4-26 Effect of Temperature on Thickening Properties of Oxygen Activated
Sludge at MLSS = 7000 mg/1 4-92
5-1 Suspended Growth Denitrification Systems Using Methanol 5-2
5-2 Observed Denitrification Rates for Suspended Growth Systems
Using Methanol 5-3
5-3 Effect of Safety Factor on Effluent Nitrate Level in Suspended Growth
System 5-9
5-4 Effect of Diurnal Variation in Load on Effluent Nitrate Level in
Complete Mix Suspended Growth System . 5-11
5-5 The Three Sludge System as Tested at Manassas, Va 5-12
5-6 ATTF System for Nitrogen and Phosphorus Removal 5-14
5-7 Design Details of Nitrogen Gas Filled Denitrification Column 5-19
5-8 Typical Process Schematic for Submerged High Porosity Media Columns 5-22
5-9 Surface Denitrification Rate for Submerged High Porosity Media Columns 5-24
5-10 Nitrification-Denitrification Flow Sheet Utilizing Low Porosity Fine
Media in Columns 5-25
5-11 Column Depth vs. Specific Surface Area 5-27
5-12 Fluidized Bed Denitrification System 5-29
5-13 Volume Denitrification Rate for Submerged High Porosity Fine Media
Columns 5-31
5-14 Feedforward Control of Methanol Based on Flow and Nitrate Nitrogen 5-38
5-15 Sequential Carbon Oxidation-Nitrification-Denitrification 5-40
5-16 Denitrification Rates Using Endogenous Carbon Sources 5-41
xix
-------
LIST OF FIGURES - Continued
Figure No. Page
5-17 Pasveer Ditch or Endless Channel System for Nitrogen Removal 5-43
5-18 Vienna-Blumenthal Wastewater Treatment Plant 5-44
5-19 Alternating Contact Process 5-49
5-20 Operational Sequencing of One of Two.Aeration Tanks in Alternating
Contact Process 5-50
5-21 The Bardenpho System - Sequential Utilization of Wastewater Carbon
and Endogenous Carbon 5-51
5-22 Blue Plains Alternating Anoxic Aerobic System 5-52
5-23 Effect of Temperature on Peak Denitrification Rates with Wastewater
as Carbon Source 5-57
5-24 Comparison of Denitrification Systems 5-63
6-1 Relative Amounts of HOC1 and OC1~ at Various pH Levels 6-2
6-2 Effects of pH and Temperature on Distribution of Ammonia and
Ammonium Ion in Water 6-3
6-3 Theoretical Breakpoint Chlorination Curve 6-6
6-4 Effect of C^NH^-N on Nitrogen Residuals in Lime Clarified
Filtered Secondary Effluent 6-11
6-5 Comparison of Germicidal Efficiency of Hypochlorous Acid,
Hypochlorite Ion, and Monochloramine for 99 Percent Destruction
of E. Coli at 2-6 C „ 6-16
6-6 Breakpoint Chlorination Control - Functional Schematic 6-17
7-1 Selective Ion Exchange Process 7-2
7-2 Generalized Ion Exchange Isotherms 7-4
7-3 The 23 C Isotherms for the Reaction, (Ca)z + 2(NHf )N = 2(NH|)Z. +
(Ca)N with Hector Clinoptilolite and IR 120 7-5
7-4 Selectivity Coefficients vs. Concentration Ratios of Sodium or
Potassium and Ammonium in the Equilibrium Solution with Hector
Clinoptilolite at 23 C for the Reaction (Y)z + (NH^ = (NH^ + (Y)N 7.6
7-5 Selectivity Coefficients vs. Concentration Ratios of Calcium or Magnesium
and Ammonium in the Equilibrium Solution with Hector Clinoptilolite
at 23 C for the Reaction (X)z + 2(NHj)N = 2(NHj)z + (X)N 7-7
7-6 Isotherms for Exchange of NH^ for K+, Na+, Ca"*"1", and Nig*"1" on
Clinoptilolite 7-8
7-7 Minimum Bed Volumes as a Function of Influent NH^-N Concentration
to Reach 50 Percent Breakthrough of Ammonium 7-12
7-8 Ammonium Breakthrough Curves for a 6 ft Clinoptilolite Bed at
Various Flow Rates , 7-13
7-9 Effect of Bed Depth on Ammonium Breakthrough at 9.7 BV/hr 7-14
xx
-------
LIST OF FIGURES - Continued
Figure No. Page
7-10 Variation of Ammonium Exchange Capacity with Competing Cation
Concentration for a 3 ft Deep Clinoptilolite Bed 7-15
7-11 Ammonium Elution with 2 Percent Sodium Chloride Regenerant 7-18
7-12 Example Ion Exchange - Air Stripping System for High pH Regenerant 7-20
7-13 Flow Diagram of Neutral pH Regeneration System Using Air Stripping 7-23
7-14 Typical Elution Curve 7-25
7-15 Simplified Flow Diagram of Electrolytic Regenerant Treatment
System 7-26
8-1 Ammonia Stripping Process 8-2
8-2 Types of Stripping Towers 8-6
8-3 Effect of Temperature on Ammonia Removal Efficiency Observed
at Blue Plains Pilot Plant 8-7,
8-4 Percent Ammonia Removal vs. Surface Loading Rate for Various
Depths of Packing 8-8
8-5 Illustrative Packing Configuration $-10
8-6 Effect of Packing Space on Air Requirements and Efficiency of
Ammonia Stripping , 8-11
8-7 Process for Ammonia Removal and Recovery 8-14
8-8 Ammonia Elimination System 8-16
8-9 Ammonia Stripping Pond System 8-|8
9-1 Alternate Process Sequencing for Systems Yielding Combined Nitrogen
and Phosphorus Removal - Systems with Coagulant Addition to
Primary Sedimentation 9-6
9-2 Alternative Process Sequencing for Systems Yielding Combined
Nitrogen and Phosphorus Removal - Systems with Coagulant Addition
after Primary Treatment 9-7
9-3 Schematic Flow Diagram - South Lake Tahoe, California Plant 9-10
9-4 Jackson, Michigan Waste water Treatment Plant Flow Diagram 9-15
9-5 Valley Community Services District (Calif.) Wastewater Treatment
Plant Flow Diagram 9-19
9-6 Holding Basin at the Valley Community Services District (California)
Wastewater Treatment Plant 9-21
9-7 City of Livermore Water Reclamation Plant (Calif.) Flow Diagram 9-22
9-8 Aeration Tank at the Livermore Water Reclamation Plant (California)
with Roughing Trickling Filters in Background 9-25
9-9 San Pablo Sanitary District Treatment Plant (California) Flow Diagram 9-27
9-10 Liquid Process Flow Sheet - CCCSD Water Reclamation Plant ft-33
xxi
-------
LIST OF FIGURES - Continued
Figure No. Page
9-11 Nitrification-Denitrification System at the CCCSD Water Reclamation
Plant 9-36
9-12 Solids Flow Diagram at the CCCSD Water Reclamation Plant 9-38
9-13 Process Flow Diagram for the Lower Molonglo Water Quality
Control Centre (Canberra, Australia) 9-41
9-14 Section Through Nitrification Tanks at the LMWQCC, Canberra,
Australia 9-44
9-15 Washington, D.C. Blue Plains Treatment Plant Flow Diagram of
Primary and Secondary Systems 9-46
9-16 Washington, D.C. Blue Plains Treatment Plant Flow Diagram of
Nitrification and Denitrification Systems 9-47
9-17 Washington, D.C. Blue Plains Treatment Plant Flow Diagram of
Filtration and Disinfection Systems 9-48
9-18 El Lago, Texas Wastewater Treatment Plant, Flow Diagram 9-51
9-19 El Lago, Texas Denitrification Columns, Coarse Media Type on Right
and Fine Media Type on Left 9-53
9-20 Hypochlorite Generation Schematic - Sacramento Regional
Wastewater Treatment Plant 9-59
9-21 Plan and Section of the Breakpoint Facility at the Sacramento
Regional Wastewater Treatment Plant 9-60
9-22 Flow Diagram of the Montgomery County, Maryland Plant 9-63
9-23 Membrane Cell Used for Hypochlorite Production 9-64
9-24 Overall System Using Membrane Cells for Hypochlorite Production 9-65
9-25 Schematic of Montgomery County, Maryland Breakpoint
Chlorination Process 9-67
9-26 Flow Diagram - Upper Occoquan Sewage Authority Plant (Virginia) 9-71
9-27 Plan and Section of Ion Exchange Beds at Upper Occoquan Plant 9-73
9-28 Added Details - Ion Exchange Beds at Upper Occoquan Plant 9-74
9-29 Plan View of ARRP Module - Upper Occoquan Plant 9-75
9-30 Section of ARRP Module - Upper Occoquan Plant 9-76
9-31 Schematic of Rosemount, Minnesota Plant 9-79
9-32 Orange Co. Ammonia Stripping/Cooling Tower Section 9-88
9-33 Overall View of the Orange County Water District 9-89
9-34 Stripping Tower Packing Module at the Orange County Water
District Plant 9-90
xxii
-------
LIST OF TABLES
Table No. Page
2-1 Estimated Nitrogen Loadings for the San Francisco Bay Basin 2-11
2-2 Effect of Ammonium Removal on Total Oxygen Demand of
Wastewater Treatment Plant Effluent 2-14
2-3 Effect of Various Treatment Processes on Nitrogen Compounds 2-21
3-1 Relationships for Oxidation and Growth in Nitrification Reactions
in Relationship to the Carbonic Acid System 3-4
3-2 Alkalinity Destruction Ratios in Experimental Studies 3-5
3-3 Maximum Growth Rates for Nitrifiers in Various Environments 3-10
3-4 Half-Saturation Constants for Nitrifiers in Various Environments 3-10
3-5 Relationship between Nitrifier Fraction and the BOD5/TKN Ratio 3-23
3-6 Compounds Toxic to Nitrifiers 3-28
3-7 Comparison of Energy Yields of Nitrate Dissimilation vs. Oxygen
Respiration for Glucose 3-30
3-8 Relationships for Nitrate Dissimilation and Growth in Denitrification
Reactions 3-32
3-9 Combined Dissimilation-Synthesis Equations for Denitrification 3-33
3-10 Values of Denitrification Yield and Decay Coefficients for Various
Investigations Using Methanol 3-38
4-1 Classification of Nitrification Facilities 4-3
4-2 Calculated Design Parameters for a 1 mgd Complete Mix Activated
Sludge Plant 4-14
4-3 Design Data Whittier Narrows Water Reclamation Plant 4-31
4-4 Nitrification Performance at the Whittier Narrows Water Reclamation
Plant 4-33
4-5 Average Nitrification Performance at Flint, Michigan for 8 Months 4-34
4-6 Effect of Temperature and Solids Retention Time on Nitrification
Efficiency at Flint, Michigan 4-34
4-7 Nitrification Performance at the Jackson, Michigan Wastewater
Treatment Plant 4-36
4-8 Comparative Physical Properties of Trickling Filter Media 4-37
4-9 Organic Nitrogen Reductions in Nitrifying Trickling Filters 4-39
4-10 Loading Criteria for Nitrification with Plastic Media at Stockton 4-40
4-11 Effect of Recirculation on Nitrification in Rock Trickling Filters at
Salford, England 4-41
4-12 Effect of Alum Addition to Wastewater on Alkalinity 4-47
4-13 Comparison of Process Characteristics for Oxygen Nitrification
Systems with and without pH Control at Blue Plains, Washington, D.C. 4-60
4-14 Comparison of Process Characteristics of Conventionally Aerated and
High Purity Oxygen Systems with pH Control at Blue Plains,
Washington, D.C. 4-61
xxiii
-------
LIST OF TABLES - Continued
Table No. Page
4-15 Commercial Types of Plastic Media for Separate Stage Nitrification
Applications 4-63
4-16 Nitrification in Separate Stage Rock Trickling Filters 4-66
4-17 Packed Bed Reactor Performance When Treating Secondary
; Effluents 4-75
4-18 Peaking Factors Versus Frequency of Occurrence for Primary
Treatment Plant Effluent 4-79
4-19 Relation of Oxygen Transfer Efficiency to Aerator Power Efficiency 4-83
4-20 Air Requirements for Nitrification Activated Sludge Plants 4-84
4-21 Effect of Oxygen Transfer Efficiency and Residual Alkalinity on
Operating pH 4-88
4-22 Comparison of Nitrification Alternatives 4-95
5-1 Denitrification Performance: Final Four Months of Operation at
Manassas, Virginia 5-13
5-2 ATTF Performance Summary, April 16 to July 15, 1972 5-16
5-3 Types of Denitrification Columns and Measured Denitrification Rates 5-18
5-4 Summary of Operation - Nitrogen Gas Filled Denitrification Column 5-21
5-5 Neptune-Microfloc Media Designs for Denitrification 5-26
5-6 Comparison of Suspended Solids Removal Efficiency for Submerged
Fine Media Denitrification Columns 5-28
5-7 Properties of Methanol 5-33
5-8 Pilot Tests of Wuhrman's Sequential Carbon Oxidation-Nitrification
Denitrification System 5-40
5-9 Design Data for the Vienna-Blumenthal Treatment Plant 5-45
5-10 Operation and Performance of the Vienna-Blumenthal Plant
24-Hour Investigations 5-46
5-11 Operation and Performance of Oxidation Ditch Operated for
Nitrogen Removal in South Africa 5-47
5-12 Performance of the "Bardenpho" Process at Pretoria, South Africa 5-51
5-13 Summary of Operation and Performance of the Blue Plains Alternating
Aerobic/Anoxic System 5-54
5-14 Observed Nitrification and Denitrification Rates for Blue Plains
Alternating Anoxic/Aerobic System 5-55
5-15 Effect of Stabilization Tank on Denitrified Effluent at the Central
Contra Costa Sanitary District's Advanced Treatment Test Facility 5-59
5-16 Denitrification Process Parameters at the Central Contra Costa
Sanitary District's Advanced Treatment Test Facility 5-59
5-17 Comparison of Denitrification Alternatives 5-62
XXIV
-------
LIST OF TABLES - Continued
Table No. Page
6-1 Effect of Pretreatment on C^NH^-N Breakpoint Ratio 6-8
6-2 Effect of Pretreatment on Formation of Nitrogenous Residuals at
Breakpoint 6-10
6-3 Effects of Chemical Addition on Total Dissolved Solids in Breakpoint
Chlorination 6-13
6-4 Effect of Breakpoint Chlorination on Soluble Organic Nitrogen 6-14
7-1 Influent Composition for Selective Ion Exchange Pilot Tests at
Different Locales 7-11
9-1 Effluent Nitrogen Concentrations in Treatment Systems Incorporating
Nitrification-Denitrification 9-2
9-2 Effluent Nitrogen Concentrations in Treatment Systems Incorporating
Ion Exchange 9-2
9-3 Effluent Nitrogen Concentrations in Treatment Systems
Incorporating Breakpoint Chlorination 9-3
9-4 Effluent Phosphorus Concentration from Alternative Systems 9-11
9-5 Design Data, Jackson, Michigan Wastewater Treatment Plant 9-14
9-6 VCSD Wastewater Treatment Plant Design Data 9-17
9-7 Nitrification Performance at Valley Community Services District
Wastewater Treatment Plant, California 9-20
9-8 Nitrogen Analyses on 24 Hour Composite Effluent Samples at the
Valley Community Services District Treatment Plant 9-21
9-9 Design Data - Livermore Water Reclamation Plant 9-24
9-10 Nitrification Performance at the Livermore Water Reclamation Plant 9-26
9-11 Design Data, San Pablo Sanitary District Treatment Plant 9-28
9-12 Nitrification Performance at the San Pablo Sanitary District
Treatment Plant 9-30
9-13 Average Process Loading Conditions at the San Pablo Sanitary
District Treatment Plant During Special Test, May 19th to July 8th,
1974 9-31
9-14 Performance Summary for the San Pablo Sanitary District Treatment
Plant During Special Testing, May 19th to July 8th, 1974 9-31
9-15 Central Contra Costa Sanitary District Water Reclamation Plant -
Design Data 9-34
9-16 Lower Molonglo Water Quality Control Centre, Design Data 9-42
9-17 Anticipated Performance Data and Effluent Standards - Blue
Plains Plant 9-45
9-18 Design Data, El Lago, Texas Wastewater Treatment Plant 9-52
9-19 Initial Performance of Fine Media Denitrification Columns at
El Lago, Texas - June 4 to July 6, 1973 9-54
xxv
-------
LIST OF TABLES - Continued
Table No. Page
9-20 Initial Performance of Coarse Media Denitrification Columns - at
El Lago, Texas - July 8 to August 31,1973 9-54
9-21 Subsequent Performance of Fine Media Denitrification Columns at
El Lago, Texas - October 1 through October 31, 1974 9-55
9-22 Design Criteria for Hypochlorite Production Facility Sacramento
Regional Wastewater Treatment Plant 9-57
9-23 Capital Cost Breakdown for Breakpoint Chlorination at the
Sacramento Regional Wastewater Treatment Plant 9-61
9-24 Total Annual Cost Breakdown for Breakpoint Chlorination at the
Sacramento Regional Wastewater Treatment Plant 9-61
9-25 Design Criteria for Hypochlorite Production Facility at the
Montgomery County Facility 9-66
9-26 Breakpoint Chlorination Design Criteria for the Montgomery County
Facility 9-68
9-27 Estimated Costs of Breakpoint Chlorination at the Montgomery
County Plant 9-69
9-28 Design Criteria Selective Ion Exchange Process for Ammonium
Removal at the Upper Occoquan Plant 9-72
9-29 Regeneration and Regenerant Recovery System Design Criteria at
the Upper Occoquan Plant 9-77
9-30 Estimated Costs of Selective Ion Exchange at the Upper Occoquan
Plant 9-78
9-31 Rosemount Ion Exchange Design Criteria 9-80
9-32 Design Data, Ammonia Stripping Tower at South Lake Tahoe,
California 9-82
9-33 Operating Costs for Ammonia Stripping for Continuous Operation of
Tahoe Air Stripping Tower at 7.5 mgd 9-84
9-34 Design Data and Estimated Nitrogen Removals for All-Weather
Ammonia Stripping at South Tahoe, California 9-85
xxvi
-------
FOREWORD
The formation of the United States Environmental Protection Agency marked a new era of
environmental awareness in America. This agency's goals are national in scope and
encompass broad responsibility in the area of air and water pollution, solid wastes,
pesticides, and radiation. A vital part of EPA's national water pollution control effort is the
constant development and dissemination of new technology for waste water treatment.
It is now clear that only the most effective design and operation of wastewater treatment
facilities, using the latest available techniques, will be adequate to meet the future water
quality objectives and to ensure protection of the nation's waters. It is essential that this
new technology be incorporated into the contemporary design of waste treatment facilities
to achieve maximum benefit of our pollution control expenditures.
The purpose of this manual is to provide the engineering community and related industry
with a new source of information to be used in the planning and design of present and
future wastewater treatment facilities. It is recognized that there are a number of design
manuals and manuals of standard practice, such as those published by the Water Pollution
Control Federation, available in the field that adequately describe and interpret current
engineering practices as related to traditional plant design. It is the intent of this manual to
supplement this existing body of knowledge by describing new treatment methods, and by
discussing the application of new techniques for more effectively removing a broad
spectrum of contaminants from wastewater.
Much of the information presented is based on the evaluation and operation of pilot,
demonstration, and full-scale plants. The design criteria thus generated represent typical
values. These values should be used as a guide and should be tempered with sound
engineering judgment based on a complete analysis of the specific application.
This manual is one of several available through the EPA Office of Technology Transfer to
describe recent technological advances and new information. Future editions will be issued
as warranted by advancing state-of-the-art to include new information as it becomes
available, and to revise design criteria as additional full-scale operational information is
generated.
xxvii
-------
CHAPTER 1
INTRODUCTION
1.1 Background and Purpose
Man's influence on the environment is receiving increasing public and scientific attention.
The quality of some of the nation's water bodies has been subjected to continuing
degradation as a result of man's activities. While there has been considerable success in
reversing this trend, one roadblock to greater progress often has been the lack of the
necessary technology to reliably and economically remove the pollutants which are the
cause of degradation of receiving waters. While conventional technology is well developed
for removing organics from wastewater, the processes for the control of nitrogen in
wastewater effluents have been developed only recently.
The beginnings of the implementation of nitrification on a significant scale occurred in the
U.S. as late as the 1960's. The practice of nitrification was widespread in England much
earlier. The first implementation of full nitrogen removal was as late as 1969 at South Lake
Tahoe in California and even this installation encountered many problems. A flurry of
research and development activity on the various nitrogen control methods occurred very
recently beginning in the late 1960's and continues to date. Recent legislation and state
regulatory activities have spurred many localities into nitrogen control projects.
Nitrogen control techniques are divided into two broad categories. The first group of
nitrogen control processes is involved with the conversion of organic and ammonia nitrogen
to nitrate nitrogen, a less objectionable form. These processes are termed nitrification
processes. The second category involves processes which result in the removal of nitrogen
from the wastewater, not just merely the conversion of nitrogen from one form to another
form in the wastewater. This latter group includes biological nitrification-denitrification, ion
exchange, ammonia stripping and breakpoint chlorination.
The purpose of this manual is the dissemination of the available data on the nitrogen control
techniques developed to date. Further, this manual is not simply an assembly of data,
rather, data from a variety of sources has been scrutinized and reasonable design criteria
drawn on the basis of all available sources. Where design procedures come directly from a
single investigator, appropriate reference is made to the work.
This manual could not have been prepared five years ago because of the state of nitrogen
control technology at that time. It may well be that continuing research will require an
update of this manual in the future. Nonetheless, the body of knowledge on nitrogen control
techniques is now well developed and municipalities and local agencies have a firm basis upon
which to plan those wastewater treatment facilities which require nitrogen control
techniques.
1-1
-------
1.2 Scope of the Manual
This manual presents theoretical and process design information on a number of nitrogen
control processes. While all of the possible nitrogen removal processes are discussed, details
are presented only on those general methods which are most technically and economically
feasible, as evidenced by their actual or planned full-scale application. One exception to this
is nitrogen control in oxidation ponds; material on nitrogen control in oxidation pond
systems was not included because of the paucity of generally applicable design information.
Another exception is land treatment; nitrogen removal by land treatment systems is beyond
the scope of this manual.
The information in this manual was developed from the following sources: (1) the
experience of the individuals involved in the preparation of the manual, (2) the EPA
research, development and demonstration program, (3) the literature, (4) from progress
reports on on-going projects, (5) from private communication with investigators active in
the field, and (6) from operating personnel at existing wastewater treatment plants.
1.3 Guide to the User
A perusal of the table of contents will give the reader a fairly complete picture of the
subject matter contained in this manual. The following chapter-by-chapter description is
oriented toward providing a general description of the contents of each chapter.
Chapter 2, Nitrogenous Materials in the Environment and the Need for Control in
Wastewater Effluents, describes the sources of nitrogen compounds entering water bodies,
the nitrogen transformations which take place in the environment, and the effects of
nitrogen compounds as pollutants. Also given in Chapter 2 is a general introduction into the
various types of nitrogen control methods and their applicability to the individual chemical
forms of nitrogen. Chapter 2 is useful for establishing the rationale for nitrogen removal.
Chapter 3, Process Chemistry and Biochemistry of Biological Nitrification and Denitrifica-
tion, is a presentation of the basic factors affecting the growth of nitrifying and denitrifying
organisms. With an understanding of these factors on a fundamental level, the design
concepts evolved in Chapters 4 and 5 can be better appreciated. However, should the reader
decide not to involve himself in basic theory, Chapters 4 and 5 are designed to stand by
themselves without requiring reference to Chapter 3 except when detailed explanations of
individual points are required.
Chapter 4, Biological Nitrification, presents design criteria for a wide variety of nitrification
processes. Since it has been anticipated that the greatest number of manual users will be
concerned with ammonia oxidation, as opposed to nitrogen removal, Chapter 4 presents
more material than any other chapter. Both combined carbon oxidation-nitrification and
separate stage nitrification systems are described with details, whether given on attached
growth or suspended growth processes. The alternative methods for pretreatment for
1-2
-------
organic carbon removal prior to separate stage nitrification systems are presented. Sections
are included on aeration, pH control, and solids-liquid separation.
Chapter 5, Biological Denitrification, completes the sequence of the three chapters on the
biological approach to nitrogen removal. Design information is provided for both attached
growth and suspended growth denitrification systems. For those systems using methanol as
the carbon source for denitrification, a section is included describing the methods for
chemical handling. The increasingly popular systems using wastewater carbon sources are
described in detail. Chapter 5 concludes with a section on solids-liquid separation and a
qualitative comparison of the alternative denitrification techniques.
Chapter 6, Breakpoint Chlorination, is the first of a set of three chapters on physical-
chemical techniques for nitrogen removal. Basic process chemistry is presented along with a
host of process design considerations for breakpoint chlorination applications. Because of
the importance of process control, details of methods are given. Information is presented on
the removal of toxic chlorine residuals.
Chapter 7, Selective Ion Exchange for Ammonium Removal, is a presentation of the design
concepts involved in the use of clinoptilolite, a natural zeolite exchange material, for
ammonium removal from wastewater. Ion exchange fundamentals are discussed along with
clinoptilolite properties. Process loading and regeneration relationships are presented.
Alternative methods of regenerant recovery are described.
Chapter 8, Air Stripping for Ammonia Nitrogen Removal, describes the application of
ammonia stripping to wastewater treatment. The air pollution aspects of the method are
discussed and general conclusions drawn. The major factors affecting design and process
performance are described. The problem of equipment scaling and its control is given
detailed consideration. Methods of removing ammonia and controlling the carbon dioxide
levels in the stripping air are described.
Chapter 9, Total System Design, describes the concepts involved in assembling various unit
processes into rational treatment trains that can accomplish not only nitrogen removal, but
organics removal and phosphorus removal (where it is required). The main thrust of Chapter
9 is to present actual examples of treatment systems that incorporate the nitrogen control
processes described in the previous chapters of this manual. Design concepts that evolved to
suit local circumstances are given emphasis.
1-3
-------
CHAPTER 2
NITROGENOUS MATERIALS IN THE ENVIRONMENT AND THE
NEED FOR CONTROL IN WASTEWATER EFFLUENTS
2.1 Introduction
Various compounds containing the element nitrogen are becoming increasingly important in
wastewater management programs because of the many effects that nitrogenous materials in
wastewater effluent can have on the environment. Nitrogen, in its various forms, can deplete
dissolved oxygen levels in receiving waters, stimulate aquatic growth, exhibit toxicity
toward aquatic life, affect chlorine disinfection efficiency, present a public health hazard,
and affect the suitability of wastewater for reuse. Biological and chemical processes which
occur in wastewater treatment plants and in the natural environment can change the
chemical form in which nitrogen exists. Such change may eliminate one deleterious effect of
nitrogen while producing, or leaving unchanged, another effect. For example, by converting
ammonia in raw wastewater to nitrate, the oxygen-depleting and toxic effects of ammonia
are eliminated, but the biostimulatory effects may not be changed significantly.
It is important, therefore, prior to the detailed discussions of nitrogen removal processes
which form the principal content of this manual, to review the chemistry of nitrogen and
the effects that the various compounds can have. Several specific aspects are discussed in
this chapter. First, the nitrogen cycle for both surface water and soil/groundwater
environments is described, with emphasis on the important compounds and reactions
associated with each. Second, sources of nitrogen, both natural and man-caused, are
discussed. Important elements of the latter category include domestic and industrial
wastewater, urban and suburban runoff, surface and subsurface agricultural drainage, and
emissions to the atmosphere which may eventually enter the aquatic environment through
precipitation or dustfall. Then, the effects of nitrogen discharge to surface water,
groundwater, and land are summarized. And finally, introductory to the following chapters,
a brief discussion is presented on the relationship between the various nitrogen compounds
and process removal efficiency.
2.2. The Nitrogen Cycle
Nitrogen exists in many compounds because of the high number of oxidation states it can
assume. In ammonia or organic compounds, the form most closely associated with plants
and animals, its oxidation state is minus 3. At the other extreme its oxidation state is plus 5
when in the nitrate form. In the environment, changes from one oxidation state to another
can be brought about biologically by living organisms. The relationship between the various
compounds and the transformations which can occur are often presented schematically in a
diagram known as the nitrogen cycle. Figure 2-1 shows a common manner of presentation J
The atmosphere serves as a reservoir of N2 gas from which nitrogen is removed naturally by
2-1
-------
electrical discharge and nitrogen-fixing organisms and artificially by chemical manufactur-
ing. Nitrogen gas is returned to the atmosphere by the action of denitrifying organisms. In
the fixed state, nitrogen can undergo the various reactions shown. A general description of
the nitrogen cycle is presented here, and aspects of particular importance to surface water
and soil/groundwater environments are discussed in the following sections.
FIGURE 2-1
THE NITROGEN CYCLE (AFTER REFERENCE 1)
FECAL
MATTER
ORGANIC
N
ANIMAL
PROTEIN
ORGANIC
N
PLANT
PROTEIN
ORGANIC
N
2-2
-------
Transformation reactions of importance include fixation, ammonification, assimilation,
nitrification and denitrification.2 These reactions can be carried out by particular
microorganisms with either a net gain or loss of energy ; energy considerations often play an
important role in determining the reaction which occurs. The principal compounds of
concern in the nitrogen cycle are nitrogen gas, ammonium, organic nitrogen, and nitrate.
These compounds and their oxidation states are shown below:
-3 0 +3 +5
NH, / NH+ - N~ - NO! - NO!
j . * T" 2^ 2t j
Organic
Derivatives
It is important to note that at neutral pH values there is very little molecular ammonia
in wastewater as most is in the form of the ammonium ion (NH4). The distribution
of ammonia and ammonium as a function of pH is discussed in Section 6.1.1.
Fixation of nitrogen from N2 gas to organic nitrogen is accomplished biologically by
specialized microorganisms. This reaction requires an investment of energy. Biological
fixation accounts for most of the natural transformation of nitrogen to compounds which
can be used by plant and animal life. Lightning fixation has been estimated to account for
approximately 15 percent of the total which occurs naturally. 3 Industrial fixation was
initially developed in the early 20th Century for manufacture of both fertilizer and
explosives. Presently, nitrogen fixed by industry is about half the amount that is removed
from the atmosphere by natural means.
Ammonification is the change from organic nitrogen to the ammonium (NH3/NH4) form.
This occurs to dead animal and plant tissue and to animal fecal matter.
Protein (organic N) + microorganisms - »- NH-/NH.
Nitrogen in urine exists principally as urea. Urea is hydrolyzed by the enzyme urease to
ammonium carbonate.
H0NCONH0 + 2H00 Enzyme* (NH,),,CO,
^ i 2. Urease °> z •*
Assimilation is the use of ammonium or nitrate compounds to form plant protein and other
nitrogen-containing compounds.
NO^ + CO 2 + green plants + sunlight - *• protein
NH3/NH4 + CO2 + green plants + sunlight - »• protein
2-3
-------
Animals require protein from plants or from other animals. With certain specific exceptions,
they are incapable of converting inorganic nitrogen forms into organic forms.
The term "nitrification" is applied to the biological oxidation of ammonium, first to the
nitrite, then to the nitrate, form. The bacteria responsible for these reactions are termed
chemoautotrophic because they use inorganic chemicals as their source of energy. Generally,
the Nitrosomonas genera are involved in conversion of ammonium to nitrite under aerobic
conditions as follows:
2NH* f 302 bactena» 2NO~ + 4H+ + 2H2O
The nitrites are in turn oxidized to nitrate generally by Nitrobacter according to the
following reaction:
-1 o_ bacteria. 2No:
^ J
The overall nitrification reaction is as follows:
To oxidize 1 mg/1 of ammonia-nitrogen requires about 4.6 mg/1 of oxygen when synthesis of
nitrifiers is neglected. The nitrate thus formed may be used in assimilation as described
above to promote plant growth, or it may be used in denitrification, wherein through
biological reduction, first nitrite and then nitrogen gas are formed. A fairly broad range of
bacteria can accomplish denitrification, including Psuedomonas, Micrococcus, Achromo-
bacter, and Bacillus. In simplified form, the reaction steps are as follows:
NO~ + 0.33 CH3OH - •> NO" + 0.33 CO2 + 0.67 H<0
(organic carbon
source)
NO~ + 0.5 CH3OH - *• 0.5N2 + 0.5 H2O + OH" + 0.5 CO2
(organic carbon
source)
Here methanol is used as the example organic carbon source, although many natural and
synthetic organic compounds can serve as the carbon source for denitrification.
2-4
-------
Oxidation of organic matter to carbon dioxide and water furnishes energy for bacteria.
Either oxygen or nitrate may be used for the oxidation, but the use of oxygen results in the
release of more energy. When both oxygen and nitrate are present, bacteria preferentially
use oxygen. Therefore, use of nitrate for denitrification can only occur under anoxic
conditions, an important consideration when attempting to remove nitrate from wastewater.
Nitrite, since it is an intermediate in the nitrification and denitrification processes, can link
the nitrification and denitrification steps directly without passing through nitrate. First,
nitrite is formed from oxidation of ammonium by Nitrosomonas, then nitrite can be
denitrified to nitrogen gas. By this route less oxygen is required for nitrification and less
organic matter (energy) is required for denitrification. This is a special case, however, and
not broadly applicable to municipal wastewater treatment.
In discussing the nitrogen cycle, it is useful to differentiate between the surface water and
sediment environment and the soil/groundwater environment. This aids in understanding the
roles that nitrogenous compounds play in each and the problems which can be encountered.
2.2.1 The Nitrogen Cycle in Surface Waters and Sediments
A modified representation of the nitrogen cycle applicable to the surface water environment
is presented in Figure 2-2.4 Nitrogen can be added by precipitation and dustfall, surface
runoff, subsurface groundwater entry, and direct discharge of wastewater effluent. In
addition, nitrogen from the atmosphere can be fixed by certain photosynthetic blue-green
algae and some bacterial species.
Within the aquatic environment ammonification, nitrification, assimilation, and denitrifica-
tion can occur as shown in Figure 2-2. Ammonification of organic matter is carried out by
microorganisms. The ammonium thus formed, along with nitrate, can be assimilated by
algae and aquatic plants; such growths may create water quality problems.
Nitrification of ammonium can occur with a resulting depletion of the dissolved oxygen
content of the water. To oxidize 1.0 mg/1 of ammonia-nitrogen, 4.6 mg/1 of oxygen is
required.
Denitrification produces nitrogen gas which may escape to the atmosphere. Because anoxic
conditions are required, the oxygen-deficient hypolimnion (or lower layer) of lakes and the
sediment zone of streams and lakes are important zones of denitrification action.4
2.2.2 The Nitrogen Cycle in Soil and Groundwater
Figure 2-3 shows the major aspects of the nitrogen cycle associated with the soil/ground-
water environment.^ Nitrogen can enter the soil from wastewater or wastewater effluent,
artificial fertilizers, plant and animal matter, precipitation, and dustfall. In addition,
2-5
-------
fHECIflTATION
AND
DUSTFALL
FIGURE 2-2
THE NITROGEN CYCLE IN SURFACE WATER (AFTER REFERENCE 4)
RUNOFF
WASTEWATER
EFFLUENT
ATMOSPHERE
-------
FIGURE 2-3
THE NITROGEN CYCLE IN SOIL AND GROUNDWATER (AFTER REFERENCE 5)
PRECIPITATION
AND
DUSTFALL
WASTEWATER
AND
WAfTEWATER
EFFLUENT
-------
nitrogen-fixing bacteria convert. nitrogen gas into forms available to plant life. Man has
increased the amount of nitrogen fixed biologically by cultivation of leguminous crops (e.g.,
peas and beans). It is estimated that nitrogen fixed by legumes now" accounts for
approximately 25 percent of the total fixed.3
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. Most
of the remainder is ammonium (NIfy), which is tightly bound to soil particles.
The nitrate content is generally low due to assimilation by plant roots and leaching by water
percolating through the soil. Nitrate pollution is the principal groundwater quality problem
in many areas. Denitrification, which is the dominating reaction below the aerobic top layer
of soil, rarely removes all nitrates added to the soil from fertilizers or wastewater effluents.
Thus, most of the nitrogen which is not assimilated by plant growth eventually enters the
groundwater table in the nitrate form.
2.3 Sources of Nitrogen
Nitrogenous materials may enter the aquatic environment from either natural or man-caused
sources. Further, the quantities from natural sources are often increased by man's activity.
For example, while some nitrogen may be expected in rainfall, the combustion of fossil
fuels or the application of liquid ammonia agricultural fertilizers with subsequent release to
the air through volatilization can increase rainfall concentrations of nitrogen substantially. It
is useful to have an understanding of the various sources of nitrogenous materials and to
have an appreciation of the quantities of nitrogen which may be expected from each.
Although the source of nitrogen causing a specific pollution problem is often obvious,
difficulty may be encountered in determining which of several possible sources is most
important. As an example, if a stream with excessive aquatic growths due to nitrogen
receives effluent from a sewage treatment plant, drainage from fertilized cropland, and
runoff from pastures or feedlots, the contribution of nitrogen from the treatment plant may
be a small fraction of that from the other two sources. Thus, in analyzing a nitrogen
pollution problem, care must be taken to ensure that all possible sources are investigated
and that the amount to be expected from each is accurately estimated. Once an estimate is
made, nitrogen control measures can be oriented toward the more significant sources.
2.3.1 Natural Sources
Natural sources of nitrogenous substances include precipitation, dustfall, nonurban runoff,
and biological fixation. Amounts from all may be increased in some way by man. It may be
quite difficult to determine quantities which might be expected under completely natural
conditions.
In order to find levels of nitrogenous substances in precipitation which are as close to
"natural" as possible, it is necessary to take samples far from urban or agricultural areas.
2-8
-------
Even these values may be suspect, however. In one review of nutrient levels in precipitation,
total nitrogen in rainfall in Sweden was cited as 0.2 mg/1.6 The average concentration of
nitrogen in western snow samples, mainly in the Sierra Nevada Mountains, was 0.15 ppm of
ammonia-nitrogen, 0.01 ppm of nitrite-nitrogen and 0.02 ppm of nitrate-nitrogen. How
representative such values are of "natural" conditions cannot be determined 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. One study found that
runoff from forested land in Washington contained 0.13 mg/1 of nitrate-nitrogen and 0.20
mg/1 of total nitrogen.7
Biological fixation may add nitrogen to both soil and surface water environments. Of
particular interest is the role of fixation in eutrophication of lakes. Certain photosynthetic
blue-green algae, such as the species of Nostoc, Anabaena, Gleotrichia and Calothrix, are
common nitrogen fixers.^
As much as 14 percent of the total nitrogen entering eutrophic Lake Mendota, Wisconsin,
was added by fixation.^ The role of nitrogen fixation in oligotrophic lakes has not been
established.
2.3.2 Man-caused Sources
The activities of man may increase quantities of nitrogen added to the aquatic
environment from three of the sources discussed above: precipitation, dustfall, and
nonurban runoff. These sources are increased principally by fertilization of agricultural land
and the combustion of fossil fuels.
Other man-related sources include runoff from urban areas and livestock feedlots, municipal
wastewater effluents, subsurface drainage from agricultural lands and from septic tank leach
fields, and industrial wastewaters.
Nitrogen concentrations in raw municipal wastewaters are well documented.^'"'" Values
generally range from 15 to 50 mg/1, of which approximately 60 percent is ammonia-
nitrogen, 40 percent is organic nitrogen, and a negligible amount (one percent) is nitrite- and
nitrate-nitrogen. Unless wastewater treatment facilities are designed to remove nitrogen
specifically, most will pass through the treatment works to the receiving waters or land
disposal site. An estimate for the total amount of nitrogen discharged into sewerage systems
in domestic wastewater is 0.84 million metric tons per year in the United States."
Nitrogen discharged into individual septic tank systems can also create pollution problems.
It has been estimated that up to 25 percent of the national population utilizes individual
systems, 9 contributing up to 0.23 million metric tons of nitrogen annually. In a
well-operating septic tank system, most of the nitrogen leaving the tank will be converted to
nitrate in the leaching field. This may then percolate downward to a groundwater table.
2-9
-------
Problems from high nitrate concentrations occasionally occur when septic tank waste
disposal is located near shallow wells used for water supply, particularly on the fringes of
urban areas where the population density may be fairly high.
The nitrogen content of industrial wastes varies dramatically from one industry to the next.
Among those industries whose wastewater nitrogen contents may be quite high are meat
processing plants, milk processing plants, petroleum refineries, ice plants, fertilizer
manufacturers, certain synthetic fiber plants, and industries using ammonia for scouring and
cleaning operations.^
Feedlot runoff constitutes a source of nitrogen which has become significant as a result of
the increased number of concentrated, centralized feedlots. Ammonium is a major
constituent of feedlot waste as a result of urea hydrolysis. Ammonia-nitrogen concentra-
tions may reach 300 mg/l,4>8,10 an(j organjc nitrogen concentrations of up to 600 mg/1
have been reported.^>10 The total annual nitrogen load from livestock in the U.S. is
estimated to be 6.0 million metric tons.4 While the majority of the animals are apparently
still raised on small farms, the trend toward feedlot operations is continuing, and unless
steps are taken to prevent drainage and runoff, serious localized problems can occur.
Urban runoff can contribute significant quantities of nitrogen to receiving waters during and
after periods of precipitation. Average concentrations which have been reported are 2.7 mg/1
total nitrogen in Cincinnati, 11 2.1 mg/1 total nitrogen in Washington, D.C.,12 2.5 mg/1 total
nitrogen in Ann Arbor, Michigan,^ and 0.85 mg/1 organic nitrogen in Tulsa, Oklahoma. 14
Sanitary or combined sewer overflows can also add to the nitrogen load.
The use of artificial fertilizers has increased the nitrogen concentrations which can be
expected in nonurban runoff. In rural Ohio, runoff from a 1.45 acre field planted in winter
wheat contained an average of 9 mg/1 total nitrogen. 15 For agricultural land in Washington,
the nitrate-nitrogen concentration was 1.25 mg/lJ On a 75-acre site in North Carolina
which consisted of grassed pasture, wooded pasture, corn field, and orchard, the mean
nitrogen concentration in the runoff was 1.2 mg/1.16
Subsurface irrigation drainage from fertilized cropland can contain high concentrations of
nitrates. In agricultural areas of California's San Joaquin Valley, monitoring of subsurface
tile drainage systems between 1966 and 1968 showed average nitrate-nitrogen concentra-
tions of 19.3 mg/1.17
In the same way that increased nitrogen concentrations in nonurban runoff and subsurface
drainage have been caused by man's activities, increased nitrogen levels in precipitation and
dustfall have also resulted. For example, high ammonium concentrations in spring rains in
California are due to the use of liquid ammonium fertilizers there." Most atmospheric
nitrogen (other than nitrogen gas), however, is associated with soil picked up by the wind
and can be returned to earth by gravitational settling (dry fallout) or in precipitation, and
several studies have been conducted to determine the quantities to be expected from such
2-10
-------
sources. The 10-year average of ammonia- plus nitrate-nitrogen concentrations in rainfall at
Geneva, New York, was 1.1 mg/1.6 Snow samples from Ottawa, Canada, over 17 years
contained an average of 0.85 ppm inorganic nitrogen." Rainwater from the same area for
the same period had concentrations of 1.8 mg/1 ammonia-nitrogen and 0.35 mg/1
nitrate-nitrogen. In rainfall measurements at Cincinnati, Ohio, total and inorganic nitrogen
concentrations were 1.27 and 0.69 mg/1, respectively.^ For a rural area near Coshocton,
Ohio, the respective concentrations were 1.17 and 0.80 rrig/1.^
A study made near Hamilton, Ontario, was cited^ which related dustfall to rainfall. It was
found that the nitrogen fall totaled 5.8 Ib per acre per year. Approximately 61 percent of
the nitrogen came down on rainy days, which constituted 25 percent of the days monitored
during the test.
In a study on dustfall in Seattle^ the fall rate for soluble nitrate-nitrogen was 0.63 Ib per
acre per year. The concentration of nitrate-nitrogen in the total dustfall was 700 ppm.
As a summary to this discussion of sources of nitrogen, Table 2-1 shows estimates of
nitrogen quantities discharged from various sources in the San Francisco Bay Basin,
California. 1 ^ The bay basin has a population of about 4,500,000 people, a land area of
4,300 square miles, and a water surface area of about 450 square miles. Because of the high
population density, the greatest amount of nitrogen discharged is from municipal and
industrial sources. This table is presented only as an example. Care must be taken for each
case to accurately evaluate the significance of each source.
TABLE 2-1
ESTIMATED NITROGEN LOADINGS FOR
THE SAN FRANCISCO BAY BASIN
Identified Nitrogen Source
Municipal wastewater, before treatment
Industrial wastewater, before treatment
Vessel wastes , before treatment
Dustfall directly on Bay
Rainfall directly on Bay
Urban runoff
Non-urban runoff
Nitrogen applied to Irrigated agricultural land
Nitrogen from dairies and feedlots
Total
Nitrogen mass emission.
thousand Ib per. year
(thousand kg per year)
55,000 (26,000)
35,000 (16,000)
130 ( 60)
1,300 ( 590)
870 ( 390)
3,100 ( 1,400)
4,100 ( 1,900)
2,000 ( 900)
13,000 ( 6,000)
118,000 (53,000)
Percent
of
total
49
30
0.1
1 .1
0.8
2.7
3.6
1.7
11
100
From Reference 19
a
'A major source not Included is biological fixation
An estimated 50 percent percolates to giroundwater
2-11
-------
2.4 Effects of Nitrogen Discharge
It was previously noted that nitrogenous compounds discharged from wastewater treatment
facilities can have several deleterious effects. Although biostimulation of receiving waters
has generated the most concern in recent years, other less well publicized impacts can be of
major importance in particular situations. These impacts include toxicity to fish life,
reduction of chlorine disinfection efficiency, an increase in the dissolved oxygen depletion
in receiving waters, adverse public health effects — principally in groundwater, and a
reduction in the suitability for reuse.
2.4.1 Biostimulation of Surface Waters
A major problem in the field of water pollution is eutrophication, excessive plant growth
and/or algae "blooms" resulting from over-fertilization of rivers, lakes, and estuaries. Results
of eutrophication include deterioration in the appearance of previously clear waters, odor
problems from decomposing algae, and a lower dissolved oxygen level which can adversely
affect fish life.
Four basic factors are required for algal growth: nitrogen, phosphorus, carbon dioxide, and
light energy. The absence of any one will limit growth. In special cases, trace micronutrients
such as cobalt, iron, molybdenum and manganese may be limiting factors under natural
conditions.
Good generalizations concerning which factor is growth limiting and at what concentration
are difficult to make. Light and carbon dioxide are essentially impossible to control. Both
nitrogen and phosphorus are present in waste discharges and hence subject to control. The
questions which must usually be answered when faced with a eutrophication problem are: is
nitrogen or phosphorus (or neither) the limiting nutrient, and if either one is, can the
amount entering the receiving water be significantly reduced by removing that nutrient from
the waste stream? In some cases algal assay procedures may allow a conclusion as to which
nutrient is limiting. Under some circumstances, however, removal of both nitrogen and
phosphorus may be undertaken to limit algal growth.
Eutrophication is of most concern in lakes because nutrients which enter tend to be
recycled within the lake and build up over a period of time. 9 A river, by contrast, is a
flowing system. Nutrients are always entering or leaving any given section. Accumulations
tend to occur only in sediment or in slack water, and the effects of these accumulations are
normally moderated by periodic flushing by 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 it contains. 9 Thus, upwelling,
which brings nutrient-rich waters to the surface, may result in periodic blooms of algae or
other aquatic life. While in some estuaries discharges from wastewater treatment plants may
increase nitrogen concentrations to the level where blooms occur, the high dilution provided
2-12
-------
by a direct ocean discharge probably eliminates the danger of algae blooms caused by such
discharges. In summary, while nitrogen in wastewater treatment plant effluents can in
particular cases cause undesirable aquatic growths, determination of the limiting constituent
and other sources of that constituent (such as feedlot runoff or fixation) should be made
before the decision is made to require nitrogen removal from municipal wastewaters.
2.4.2 Toxicity
The principal toxicity problem is from ammonia in the molecular form (NH3> which can
adversely affect fish life in receiving waters. A slight increase in pH may cause a great
increase in toxicity as the ammonium ion (NH4) is transformed to ammonia in accordance
with the following equation.
NH* + OH~ —- NH3 + H20
Factors which may increase ammonia toxicity at a given pH are: greater concentrations of
dissolved oxygen and carbon dioxide; elevated temperatures; and bicarbonate alkalinity.
Reported levels at which acute toxicity is detectable have ranged from 0.01 mg/K to over
2.0 mg/l^O of molecular ammonia-nitrogen.
2.4.3 Effect on Disinfection Efficiency
When chlorine, in the form of chlorine gas or hypochlorite salt, is added to wastewater
containing ammonium, chloramines, which are less effective disinfectants, are formed. The
major reactions are as follows:
NH* + HOC1 i=^ NH2C1 (monochloramine) + H2O + H+
NH2C1 + HOC1 :^=^ NHC12 (dichloramine) + H2O
NHC12 + HOC1 y*" NC13 (nitrogen trichloride) + H2O
Only after the addition of large quantities of chlorine does free available chlorine exist. If
the effluent ammonia-nitrogen concentration were 20 mg/1, about 200 mg/1 of chlorine
would be required to complete the reactions with ammonium and organic compounds. Only
rarely in wastewater treatment is this level of chlorine addition ("breakpoint" chlorination)
used. Therefore, as a practical matter, the less effective combined chlorine residuals
(monochloramine and dichloramine) must be relied upon for disinfection. This results in
increased chlorine dose requirements for the same level of disinfection. Further information
on the relative effectiveness of free chlorine and combined residuals is presented in Section
6.2.7.
2-13
-------
2.4.4 Dissolved Oxygen Depletion in Receiving Waters
Ammonium can be biologically oxidized to nitrite and then to nitrate in receiving waters
and thereby add to the oxygen demand imparted by carbonaceous materials. Table 2-2
shows a typical example of the removal of total oxygen demand obtainable with varying
degrees of treatment. If either conventional biological treatment or physical-chemical
treatment is utilized to provide 90 percent BOD5 removal, an effluent will be discharged
which still contains over 100 mg/1 of oxygen demand. This high level of oxygen demand
may cause significant oxygen depletion in the receiving water if insufficient dilution is
available. Nitrification (or ammonia nitrogen removal) will reduce the total oxygen demand
of the effluent to less than 40 mg/1.
The Potomac Estuary in the United States22 and the Thames Estuary in Great Britain^^ are
examples of estuaries which are greatly affected by nitrification. Figure 2-4 shows, as a
function of the degree of nitrification provided by wastewater treatment facilities, the
estimated discharge into the Thames Estuary which will cause the maximum oxygen
depletion to be 10 percent of saturation. The calculation assumes an effluent BOD5 of 20
mg/1, an effluent organic plus ammonia-nitrogen concentration of 19 mg/1, and discharge at
a point 10 miles above London Bridge. From the figure, the allowable discharge for
non-nitrified effluent is about 12 mgd, while for completely nitrified effluent, over 40 mgd
can be discharged.
2.4.5 Public Health
The public health hazard from nitrogen is associated with the nitrate form and is limited
principally to groundwater where high concentrations can occur. Nitrate in drinking water
TABLE 2-2
EFFECT OF AMMONIUM REMOVAL ON TOTAL OXYGEN
DEMAND OF WASTEWATER TREATMENT PLANT EFFLUENT
Parameter
Organic matter, mg/1
Organic oxygen demand, mg/1
Organic and ammonia nitrogen, mg/1
Nitrogenous oxygen demand (NOD) mg/1
Total oxygen demand (TOD) mg/1
Percent of TOD due to nitrogen
Percent organic oxygen demand removed
Percent of TOD removed
Raw
wastewater
250
375b
25
115C
490
23.5
-
-
Final effluent
Organic carbon
removal
only
25
37
20
92°
129
71.3
90
73.7
With ammonium
and organic carbon
removal
20
30
1.5
7c
37
18.9
92
92.5
"After Reference 21
Taken as 1.5 times organic matter
cTaken as 4.6 times the nitrogen level
2-14
-------
was first associated in 1945 with methemoglobinemia, a sometimes fatal blood disorder
which affects infants less than three months old. When water high in nitrate is used for
preparing infant formulas, nitrate is reduced to nitrite in the stomach after ingestion. The
nitrites react with hemoglobin in the blood to form methemoglobin, which is incapable of
carrying oxygen. The result is suffocation accompanied by a bluish tinge to the skin, which
accounts for the use of the term "blue babies" in conjunction with methemoglobinemia. In
suspect areas water should be analyzed for both nitrite and nitrate since either form will
cause methemoglobinemia.
FIGURE 2-4
ALLOWABLE EFFLUENT DISCHARGE INTO THE THAMES ESTUARY
(AFTER REFERENCE 23)
50
m
TJ
O»
E
LJ
O
CL
<
X
o
UJ
ID
U_
UJ
UJ
_l
CD
I
o
40
BOD5 OF WASTEWATER EFFLUENT: 20 mg/l
POINT OF DISCHARGE : IOMILES
ABOVE LONDON BRIDGE
MAXIMUM OXYGEN DEPLETION:
10 PERCENT OF SATURATION
100
PERCENT NITRIFICATION
OF WASTEWATER EFFLUENT
2-15
-------
Since 1945 about 2,000 cases of methemoglobinemia have been reported in the U.S. and
Europe, with a mortality rate of seven to eight percent. Because of difficulty in diagnosing
the disease and because no reporting is required, the actual incidence may be many times
higher. !0
The EPA's interim primary drinking water standard (40 CFR Part 141) for nitrate is 10 mg/1
as nitrogen. This standard is exceeded most often in shallow wells in rural areas.
2.4.6 Water Reuse
While direct wastewater reuse for domestic water supply is not yet a reality because of
public health considerations, plans for industrial reuse are being carried out in several areas.
When reclaiming wastewater for industrial purposes, ammonia may need to be removed in
order to prevent corrosion. Further, nitrogen compounds can cause biostimulation in
cooling towers and distribution structures.
2.5 Treatment Processes for Nitrogen Removal
In the past several years the number of processes utilized in wastewater treatment has
increased rapidly. Many of these processes have been developed with the specific purpose of
transforming nitrogen compounds or removing nitrogen from the wastewater stream. Others
can remove several compounds, including significant amounts of nitrogen. Still others may
remove only a small amount of nitrogen or a particular form of nitrogen which is a small
fraction of the total.
In determining which method is most suitable for a particular application, consideration
must be given to six principal aspects: (1) form and concentration of nitrogen compounds in
the process influent, (2) required effluent quality, (3) other treatment processes to be
employed, (4) cost, (5) reliability, and (6) flexibility. Great care must be taken in
developing and evaluating alternatives.
Presented below are brief descriptions of the various processes employed in wastewater
treatment facilities which, to varying degrees, remove nitrogen from the waste stream.
Process characteristics, compound selectivity, and normal range of efficiency are presented.
It is stressed that this discussion is descriptive and is intended only to provide an
introduction to the following chapters of this manual.
2.5.1 Conventional Treatment Processes
Nitrogen in raw domestic wastewaters is principally in the form of organic nitrogen, both
soluble and particulate, and ammonia. The soluble organic nitrogen is mainly in the form of
urea and amino acids. Primary sedimentation acts to remove a portion of the particulate
organic matter. This generally will amount to less than 20 percent of the total nitrogen
entering the plant.
2-16
-------
Biological treatment will remove more particulate organic nitrogen and transform some to
ammonium and other inorganic forms. A fraction of the ammonium present in the waste
will be assimilated into organic materials of cells formed by the biological process. Soluble
organic nitrogen is partially transformed to ammonium by microorganisms, but concentra-
tions of 1 to 3 mg/1 are usually found in biological treatment effluents.24 Through these
processes, an additional 10 to 20 percent of the total nitrogen is removed when biological
treatment and secondary sedimentation follows primary sedimentation. Thus, total nitrogen
removal for a conventional primary-secondary facility will generally be less than about 30
percent.
2.5.2 Advanced Wastewater Treatment Processes
Advanced treatment processes designed to remove wastewater constituents other than
nitrogen often remove some nitrogen compounds as well. Removal is often restricted to
particulate forms, and overall removal efficiency is rarely high.
Tertiary filtration can remove a significant fraction of the organic nitrogen present. Overall
removal depends on the amount of nitrogen in the suspended organic form. As noted above,
most of the organic nitrogen in secondary effluent is insoluble, but ammonium usually
accounts for the majority of the total nitrogen. Carbon adsorption, used to remove residual
organics, will also remove organic nitrogen. The amount of organic nitrogen remaining at
that point in the treatment scheme will generally be quite small.
Electrodialysis and reverse osmosis are tertiary processes used primarily for reduction of
total dissolved solids. Nitrogen entering such systems is mainly in the ammonium or nitrate
form. Electrodialysis can be expected to remove about 40 percent of these forms; reverse
osmosis, 80 percent. However, these processes are not currently in use for treatment of
municipal wastewater.
Chemical coagulation, often utilized for phosphate removal, also aids in removal of
particulate matter, including particulate organic nitrogen. While chemical coagulation does
not remove ammonium directly, lime addition is used prior to ammonia stripping (discussed
in Section 2.5.3.4) in order to raise the pH and allow the process to proceed.
Land disposal may be used to remove nitrogen. Removal occurs when the effluent is used
for irrigation purposes with the nitrogen assimilated by growing crops which are
subsequently harvested. However, nitrogen removal by land treatment systems is not within
the scope of this manual.
2.5.3 Major Nitrogen Removal Processes
The major processes considered in this manual are nitrification-denitrification, breakpoint
chlorination (or superchlorination), selective ion exchange for ammonium removal, and air
2-17
-------
stripping for ammonia removal (ammonia stripping). These are the processes which are
technically and economically most viable at the present time.
2.5.3.1 Biological Nitrification-Denitrification
Biological nitrification does not increase the removal of nitrogen from the waste stream over
that achieved by conventional biological treatment. The principal effect of the nitrification
treatment process is to transform ammonia-nitrogen to nitrate. The nitrified effluent can
then be denitrified biologically. Nitrification is also used without subsequent biological
denitrification when treatment requirements call for oxidation of ammonia-nitrogen.
Oxidation of ammonium can be as high as 98 percent. Overall transformation to nitrate
depends on the extent to which organic nitrogen is transformed to ammonia-nitrogen in the
secondary stage or is removed by another process. Nitrification can be carried out in
conjunction with secondary treatment or in a tertiary stage; in both cases, either suspended
growth reactors (activated sludge) or attached growth reactors (such as trickling filters) can
be used.
Biological denitrification can also be carried out in either suspended growth or attached
growth reactors. As previously noted, an anoxic environment is required for the reactions to
proceed. Overall removal efficiency in a nitrification-denitrification plant can range from 70
to 95 percent.
2.5.3.2 Breakpoint Chlorination
Breakpoint chlorination (or superchlorination) is accomplished by the addition of chlorine
to the waste stream in an amount sufficient to oxidize ammonia-nitrogen to nitrogen gas.
After sufficient chlorine is added to oxidize the organic matter and other readily oxidizable
substances present, a stepwise reaction of chlorine with ammonium takes place. The overall
theoretical reaction is as follows:
3C12 + 2NH* *- N2 + 6HC1 + 2H+
In practice, approximately 10 mg/1 of chlorine is required for every 1 mg/1 of
ammonia-nitrogen. In addition, acidity produced by the reaction must be neutralized by the
addition of caustic soda or lime. These chemicals add greatly to the total dissolved solids and
result in a substantial operating expense. Often dechlorination is utilized following
breakpoint chlorination in order to reduce the toxicity of the chlorine residual in the
effluent.
An important advantage of this method is that ammonia-nitrogen concentrations can be
reduced to near zero in the effluent. The effect of breakpoint chlorination on organic
nitrogen is uncertain, with contradictory results presented in the literature. Nitrite and
nitrate are not removed by this method.
2-18
-------
2.5.3.3 Selective Ion Exchange for Ammonium Removal
Selective ion exchange for removal of ammonium from wastewater can be accomplished by
passing the wastewater through a column of clinoptilolite, a naturally occurring zeolite
which has a high selectivity for ammonium ion. The first extensive study was undertaken in
1969 by Battelle Northwest in a federally sponsored demonstration project. Regeneration of
the clinoptilolite is undertaken when all the exchange sites are utilized and breakthrough
occurs.
Filtration prior to ion exchange is usually required to prevent fouling of the zeolite.
Ammonium removals of 90-97 percent can be expected. Nitrite, nitrate, and organic
nitrogen are not affected by this process.
2.5.3.4 Air Stripping for Ammonia Removal
Ammonia in the molecular form is a gas which dissolves in water to an extent controlled by
the partial pressure of the ammonia in the air adjacent to the water. Reducing the partial
pressure causes ammonia to leave the water phase and enter the air. Ammonia removal from
wastewater can be effected by bringing small drops of water in contact with a large amount
of ammonia-free air. This physical process is termed desorption, but the common name is
"ammonia stripping."
In order to strip ammonia from wastewater, it must be in the molecular form (NH3) rather
than the ammonium ion (NHjj) form. This is accomplished by raising the pH of the
wastewater to 10 or 11, usually by the addition of lime. Because lime addition is often used
for phosphate removal, it can serve a dual role. Again, nitrite, nitrate, and organic nitrogen
are not affected.
The principal problems associated with ammonia stripping are its inefficiency in cold
weather, required shutdown during freezing conditions, and formation of calcium carbonate
scale in the air stripping tower.
The effect of cold weather has been well documented at the South Lake Tahoe Public
Utility District where ammonia stripping is used for a 3.75 mgd tertiary facility. The
stripping tower is designed to remove 90 percent of the incoming ammonium during warm
weather. During freezing conditions, the tower is shut down. One mechanism of scale
formation is attributed to the carbon dioxide in the air reacting with the alkaline wastewater
and precipitating as calcium carbonate.22 in some instances, removal with a water jet has
been possible; in other applications the scale has been extremely difficult to remove. Some
factors which may affect the nature of the scale are: orientation of air flow, recirculation of
sludge, pH of the wastewater, and chemical makeup of the waste water. 2 2
2-19
-------
2.5.4 Other Nitrogen Removal Processes
In addition to the processes listed above, there are other methods for nitrogen removal
which might usefully be discussed. Most are in the experimental stage of development or
occur coincidentally with another process.
Use of anionic exchange resins for removal of nitrate was developed principally for
treatment of irrigation return waters.22 TWO major unsolved problems are the lack of resins
which have a high selectivity for nitrate over chloride and disposal of nitrogen-laden
regenerants.
Oxidation ponds can remove nitrogen through microbial denitrification in the anaerobic
bottom layer or by ammonia emission to the atmosphere. The latter effect is essentially
ammonia stripping but is relatively inefficent due to a low surface-volume ratio and low pH.
In a study of raw wastewater lagoons in California, removals of 35-85 percent were reported
for well-operated lagoons. ^
Nitrogen in oxidation ponds is assimilated by algal cultures. If the algal cells are removed
from the pond effluent stream, nitrogen removal is thereby effected. Methods for removal
of algae are summarized in the EPA Technology Transfer Publication, Upgrading
Lagoons. 25
It was noted previously that in secondary biological treatment and in nitrification, some
nitrogen is incorporated in bacterial cells and is removed from the waste stream with the
sludge. If an organic carbon source such as ethanol or glucose is added to the wastewater,
the solids production will be increased and a greater nitrogen removal will be effected.
Disadvantages are that large quantities of sludge are produced and that difficulties occur in
regulating the addition of the carbon source, with high effluent BOD5 values or high
nitrogen levels resulting. 2
2.5.5 Summary
Table 2-3 summarizes the effect of various treatment processes on nitrogen removal. Shown
is the effect that the process has on each of the three major forms: organic nitrogen,
ammonium, and nitrate. In the last column is shown normal removal percentages which can
be expected from that process. Overall removal for a particular treatment plant will depend
on the types of unit processes and their relation to each other. For example, while many
processes developed for nitrogen removal are ineffective in removing organic nitrogen,
incorporation of chemical coagulation or multimedia filtration into the overall flowsheet
can result in a low concentration of organic nitrogen in the plant effluent. Thus, the
interrelationship between processes must be carefully analyzed in designing for nitrogen
removal. Further discussion of process interrelationships is presented in Chapter 9.
2-20
-------
TABLE 2-3
EFFECT OF VARIOUS TREATMENT PROCESSES ON NITROGEN COMPOUNDS
Treatment process
Conventional treatment processes
Primary
Secondary
Advanced wastewater treatment processes
Filtration0
Carbon sorption
Electrodialysis
Reverse osmosis
Chemical coagulation0
Land application
Irrigation
Infiltration/percolation
Major nitrogen removal processes
Nitrification
Denitrification
Breakpoint chlorination
Selective ion exchange for ammonium
Ammonia stripping
Other nitrogen removal processes
Selective ion exchange for nitrate
Oxidation ponds
Algae stripping
Bacterial assimilation
Effect on constituent
.Organic N
10-20% removed
15-25% removedb
urea -*. NH3/NHj
30-95% removed
30-50% removed
100% of suspend
organic N removed
100% of suspend
organic N removed
50-70% removed
_». NH3/NH4
-*• NH3/NH4
limited effect
no effect
uncertain
some removal, un-
certain
no effect
nil
partial transformation
to NH3/NHJ
partial transformation
to NH3/NH4
no effect ,
NH3/NH4
no effect
;10% removed
nil
nil
4 J% removed
85% removed
ni..
-»• N03
-»• plant N
-»• NOJ
--NO;
no effect
90-100% removed
90-97% removed
60-95% removed
nil '
partial removal
by stripping
-». cells
40-70% removed
NO3
no effect
nil
nil
nil
40% removed
85% removed
nil
-». plant N
-» N2
no effect .
80-98% removed
no -effect
no effect
no -effect
75-90% removed
partial removal by
nitrification-
df. nitrification
-*. cells
limited effect
Removal of
total nitrogen
entering process,
percent9
5-10
10-20
20-40
10-20
35-45
80-90
20-30
40-90
0-50
5-10
70-95
80-95
80-95
50-90
70-90
20-90
50-80
30-70
Will depend on the fraction of influent nitrogen for which the process is effective, which n.ay depend on other processes
in the treatment plant.
Soluble organic nitrogen, in the form of urea and amino acids, is substantially reduced by secondary treatment.
GMay be used to remove particulate organic carbon in plants where ammonia or nitrate are removed by other processes.
2.6 References
1. Sawyer, C.N., and P.L. McCarty, Chemistry for Sanitary Engineers. New York,
McGraw-Hill Book Co., 1967.
I
2. Christensen, M.H., and P. Harremoes, Biological Denitrification in Wastewater
Treatment. Report 2-72, Department of Sanitary Engineering, Technical University of
Denmark, 1972.
3. Delwiche, C.C., The Nitrogen Cycle. Scientific American, .223, No. 3, pp 137-146
(1970).
2-21
-------
4. Martin, D.M., and D.R. Goff, The Role of Nitrogen in the Aquatic Environment.
Report No. 2, Department of Limnology, Academy of Natural Sciences of Phila-
delphia, 1972.
5. Sepp, E., Nitrogen Cycle in Groundwater. Bureau of Sanitary Engineering, State of
California Department of Public Health, 1970.
6. McCarty, P.L., et al, Sources of Nitrogen and Phosphorus in Water Supplies. JAWWA,
59, pp 344 (1967).
7. Sylvester, R.O., Nutrient Content of Drainage Water for Forested, Urban, and
Agricultural Areas. Alfjae and Metropolitan Wastes, Robert A. Taft Sanitary Engineer-
ing Center, Tech. Rep. W61-3, 1963.
8. Reeves, T.G., Nitrogen Removal: A Literature Review. JWPCF, 44, No. 10, pp
1896-1908(1972).
9. Nitrogenous Compounds in the Environment. Hazardous Materials Advisory Com-
mittee (to the EPA), EPA-ASB-73-001, December, 1973.
10. Kaufman, W.J., Chemical Pollution of Ground Waters. JAWWA, 66, No. 3, pp 152-159
(1974).
11. Weibel, S.R., Anderson, R.J., and R.L. Woodward, Urban-Land Runoff as a Factor in
Stream Pollution. JWPCF, 43, p 2033 (1971).
12. American Public Works Association, Water Pollution Aspects of Urban Runoff.
FWPCA Report No. WP-20-15, January, 1969.
13. Burn, R.J., Krawezyk, D.F., and G.T. Harlow, Chemical and Physical Comparison of
Combined and Separated Sewer Discharges. JWPCF, 40, pp 112 (1968).
14. Avco Economic Systems Corporation, Storm Water Pollution From Urban Land
Activity. EPA Report No. 11034 FKL 07/70, July, 1970.
15. Weibel, S.R., et al, Pesticides and Other Contaminants in Rainfall and Runoff.
JAWWA, 58, pp 1075 (1966).
16. Dept. of Biological and Agricultural Engineering, North Carolina State University at
Raleigh, Role of Animal Wastes in Agricultural Land Runoff. EPA Report No. 13020
DGX 08/71, August 1971.
17. California Departmen't of Water Resources, Nutrients From Tile Drainage Systems.
EPA Report No. 13030 ELY 5/71-3, May 1971.
2-22
-------
18. Johnson, R.E., Rossano, A.T., Jr., and R.O. Sylvester, Dustfall as a Source of Water
Quality Impairment, ASCE, JSED, 92, No. SA1, pp 145 (1966).
19. California State Water Resources Control Board, Tentative Water Quality Control Plan,
San Francisco Bay Basin. November, 1974.
20. Brown and Caldwell/Dewante and Stowell, Feasibility Study for the Northeast-Central
Sewerage Service Area. Prepared for County of Sacramento, Department of Public
Works, November 1972.
21. Ehreth, D.J., and E. Barth, Control of Nitrogen in Wastewater Effluents. Prepared for
the EPA Technology Transfer Program, March, 1974.
22. Nitrogen Removal from Wastewaters. EPA Advanced Waste Treatment Research
Laboratory, ORD-17010, October, 1970.
23. Effects of Pollution Discharges on the Thames Estuary. Department of Scientific and
Industrial Research, Water Pollution Research Technical Paper No. 11, Her Majesty's
Stationery Office, 1964.
24. Parkin, G.F., and P.L. McCarty, The Nature, Ecological Significance, and Removal of
Soluble Organic Nitrogen in Treated Agricultural Wastewaters. Stanford University,
prepared for the Bureau of Reclamation, Contract USDI- 14-06-200-6090-A,
September, 1973.
25. Caldwell, D.H., Parker, D.S., and W.R. Uhte, Upgrading Lagoons. Prepared for the EPA
Technology Transfer Program, August, 1973.
2-23
-------
CHAPTER 3
PROCESS CHEMISTRY AND BIOCHEMISTRY OF
NITRIFICATION AND DENITRIFICATION
3.1 Introduction
The purpose of this chapter is to present a treatment process-oriented review of the
chemistry and biochemistry of nitrification and denitrification. An understanding of this
subject is useful for developing an appreciation of the factors affecting the performance,
design, and operation of nitrification and denitrification processes. Subsequent chapters deal
with design aspects of nitrification (Chapter 4) and denitrification (Chapter 5). Since these
latter chapters are laid out to be used without reference to this chapter, review of the
theoretical material in this chapter is not mandatory.
Biological processes for control of nitrogenous residuals in effluents can be classified in two
broad areas. First, a process designed to produce an effluent where influent nitrogen
(ammonia and organic nitrogen) is substantially converted to nitrate nitrogen can be
considered. This process, nitrification, is carried out by bacterial populations that
sequentially oxidize ammonia to nitrate with intermediate formation of nitrite. Nitrification
will satisfy effluent or receiving water standards where reduction of residual nitrogenous
oxygen demand due to ammonia is mandated or where ammonia reduction for other reasons
is required for the treatment system. The second type of process, denitrification, reduces
nitrate to nitrogen gas and can be used following nitrification when the total nitrogenous
content of the effluent must be reduced.
3.2 Nitrification
The two principal genera of importance in biological nitrification processes are Nitro-
somonas and Nitrobacter. Both of these groups are classed as autotrophic organisms. These
organisms are distinguished from heterotrophic bacteria in that they derive energy for
growth from the oxidation of inorganic nitrogen compounds, rather than from the
oxidation of organic matter. Another feature of these organisms is that inorganic carbon
(carbon dioxide) is used for synthesis rather than organic carbon. Each group is limited to
the oxidation of specific species of nitrogen compounds. Nitrosomonas can oxidize
ammonia to nitrite, but cannot complete the oxidation to nitrate. On the other hand,
Nitrobacter is limited to the oxidation of nitrite to nitrate. Since complete nitrification is a
sequential reaction, treatment processes must be designed to provide an environment
suitable to the growth of both groups of nitrifying bacteria.
3.2.1 Biochemical Pathways
On a biochemical level, the nitrification process is more complex than simply the sequential
oxidation of ammonia to nitrite by Nitrosomonas and the subsequent oxidation of nitrite to
3-1
-------
nitrate by Nitrobacter. Various reaction intermediates and enzymes are involvedJ More
important than an understanding of these pathways is knowledge of the response of
nitrification organisms to environmental conditions.
3.2.2. Energy and Synthesis Relationships
The stoichiometric reaction for oxidation of ammonium to nitrite by Nitrosomonas is:
NH* + 1.5O2 +• 2H+ + H2O + NO~ (3-1)
&M*
The Jess of free energy by this reaction at physiological concentrations of the reactants has
been estimated by various investigators to be between 58 and 84 kcal per mole of
ammonia. *>2
The reaction for oxidation of nitrite to nitrate by Nitrobacter is:
N0~ + 0.5 02 »» N0~ . (3-2)
This reaction has been estimated to release between 15.4 to 20.9 kcal per mole of nitrite
under in vivo conditions. 2 Thus, Nitrosomonas obtains more energy per mole of nitrogen
oxidized than Nitrobacter. If it assumed that the cell synthesis per unit of energy produced
is equal, there should be greater mass of Nitrosomonas formed than Nitrobacter per mole of
nitrogen oxidized. As will be seen, this is in fact the case.
The overall oxidation of ammonium by both groups is obtained by adding Equations 3-1
and 3-2:
NH* + 202 —*• N0~ + 2H+ + H23 + C5H?NO2 + H+ (3-5)
Nitrobacter
3-2
-------
Equations 3-1, 3-4 and 3-5 have terms showing the production of free acid (H+) and the
consumption of gaseous carbon dioxide (CC>2)- In actual fact, these reactions take place in
aqueous systems in the context of the carbonic acid system. These reactions usually take
place at pH levels less than 8.3. Under this circumstance, the production of acid results in
immediate reaction with bicarbonate ion (HCOs) with the production of carbonic acid
(H2CO3). The consumption of carbon dioxide by the organisms results in some depletion of
the dissolved form of carbon dioxide, carbonic acid (H2CO3). Table 3-1 presents the
modified forms of Equations 3-1 to 3-5 to reflect the changes in the carbonic acid system.
As will be later described in Sections 3.2.3 and 3.2.5.6, the variations occurring in pH
resulting from changes in the carbonic acid system can significantly affect nitrification
process performance.
The equations for energy yielding reactions (Equations 3-1 and 3-2) can be combined with
the equations for organism synthesis (Equations 3-4 and 3-5) to form overall synthesis-
oxidation relations by knowledge of the yield coefficients for the nitrifying organisms.
Experimental yield values for Nitrosomonas range from 0.04 to 0.13 Ib VSS grown per Ib
ammonia nitrogen oxidized. 1 Experimental yields for Nitrobacter are in the range from 0.02
to 0.07 Ib VSS grown per Ib of nitrite nitrogen oxidized.l Values based on thermodynamic
theory are 0.29 and 0.084 for Nitrosomonas and Nitrobacter, respectively.2 The
experimentally based yield may be lower than theoretical values due to the diversion of a
portion of the free energy released by oxidation to microorganism maintenance functions.2
Equations for synthesis-oxidation using representative measurements of yields and oxygen
consumption for Nitrosomonas and Nitrobacter are as follows: 3,4
+ 76 (X, + 109HCO" »- C^HJSKX, + 54NCX, + 57tt,O (3-6)
Z* 3 J I £ £ £
Nitrosomonas
+ 104H2C03
400NO" + NH* + 4H,,CO, + HCO~ + 195 Oo —». C^H-NO0 ,, 7,
2* H" 2* j J Zi j I £ \.^~' )
Nitrobacter
3H2O + 400NO~
Using Equations 3-6 and 3-7, the overall synthesis and oxidation reaction is:
NH^ + 1.8302 + 1.98HC03 —*~ 0.021 C5H?NO2 + 1.041H2O (3-8)
+ 0.98NO~ + 1.88H2CO3
3-3
-------
In these equations, yields for Nitrosomonas and Nitrobacter are 0.15 mg cells/mg NH^-N
and 0.02 mg cells/mg NO2-N, respectively. On this basis, the removal of twenty- mg/1 of
ammonia nitrogen would yield only 1.8 mg/1 of nitrifying organisms. This relatively low
yield has some far reaching implications, as will be seen in Section 3.2.7. Oxygen
consumption ratios in the equations are 3.22 mg O2/mg Nffy-N oxidized and 1.11 mg
O2/mg NOJ-N oxidized, which is in agreement with measured values.^
3.2.3. Alkalinity and pH Relationships
Equation 3-3A (Table 3-1) shows that alkalinity is destroyed by the oxidation of ammonia
and carbon dioxide (H2CO3 in the aqueous phase) is produced. When synthesis is neglected,
it can be calculated that 7.14 mg of alkalinity as CaCOs is destroyed per mg of ammonia
nitrogen oxidized. The effect of synthesis is relatively small; in Equation 3-8, the ratio is
7.07 mg of alkalinity per mg of ammonia nitrogen oxidized. Experimentally determined
ratios are presented in Table 3-2; differences between the experimental and theoretical
ratios are due either to errors in alkalinity or nitrogen analyses or the inadequacy of theory
TABLE 3-1
RELATIONSHIPS FOR OXIDATION AND GROWTH IN NITRIFICATION
REACTIONS IN RELATIONSHIP TO THE CARBONIC ACID SYSTEM
Reaction
Equation
Equation
No.
Oxidation -
Nitrosomonas
Oxidation -
Nitrobacter
Oxidation -
overall
Synthesis -
Nitrosomonas
Synthesis -
Nitrobacter
NH. +1.50_+ 2HCO
4 £
NO. +0.5O.-*-NO
NO, + 2H.CO, + H.O
L Z O £t
2HCO -^
13NH4 + 23HCO3
3-1A
3-2
3-3A
3-4A
10NO
3-5A
3-4
-------
TABLE 3-2
ALKALINITY DESTRUCTION RATIOS IN EXPERIMENTAL STUDIES
System
Suspended growth
Suspended growth
Suspended growth
Attached growth
Attached growth
Attached growth
mg alkalinity destroyed
' mg NH~4-N oxidized
6.4
6.0
7.1
6.5
6 . 3 to 7 . 4
x .7.3
Reference
5
6
7
8
' ' 9 '
2
As CaCO ; the theoretical value is 7.1
O
to completely explain the phenomenon. A ratio of 7.14 mg alkalinity destroyed per mg of
ammonia nitrogen oxidized may be used for engineering calculations.
These changes may have a depressing effect on pH in the nitrification system, as the
relationship for pH in the system is:
pH = pKj - log
(3_9)
Since nitrification reduces the HCOs level and increases the H2CO3 level, it is obvious that
the pH would tend to be reduced. The effect is mediated by stripping of carbon dioxide
from the liquid by the process of aeration and the pH is elevated upwards. If the carbon
dioxide is not stripped from the liquid, such as in enclosed high purity oxygen systems, the
pH can be depressed as low as 6.0. It has been calculated that to maintain the pH greater
than 6.0 in an enclosed system, the alkalinity of the wastewater must be 10 times greater
than the amount of ammonia nitrified.2
Even in open systems where the carbon dioxide is continually stripped from the liquid,
severe pH depression can occur when the alkalinity in the wastewater approaches de-
pletion by the acid produced in the nitrification process. For example, if in a wastewater
20 mg/1 of ammonia nitrogen is nitrified, 143 mg/1 of alkalinity as CaCOs will be destroyed.
In many wastewaters there is insufficient alkalinity initially present to leave a sufficient
residual for buffering the wastewater during the nitrification process. The significance of pH
depression in the process is that nitrification rates are rapidly depressed as the pH is reduced
3-5
-------
below 7.0 (see Section 3.2.5.6). Procedures for calculating the operating pH in aeration
systems are presented in Section 4.9.
3.2.4 Oxygen Requirements
The theoretical oxygen requirement for nitrification, neglecting synthesis, is 4.57 mg O2/mg
NH^-N (Equation 3-3). Synthesis has an effect on oxygen requirements; the oxygen
requirement is calculated to be, from Equation 3-8, 4.19 mg C>2/mg NH^-N. An oxygen
requirement sufficiently accurate to be used in engineering calculations for aeration
requirements is 4.6 mg O2/mg NH^-N.
The oxygen demand for nitrification is significant; for instance if 30 mg/1 of ammonia
nitrogen is oxidized by the nitrification system, about 138 mg/1 of oxygen will be required.
Caution: in virtually all practical nitrification systems, oxygen demanding materials other
than ammonia are present in the wastewater, raising the total oxygen requirements of
nitrification systems even higher (see Section 4.8).
3.2.5 Kinetics of Nitrification
A complete review of the kinetics of biological systems is beyond the scope of the manual;
however, several excellent reviews are available. 10,11,12 Rather, the basics of biological
kinetics are drawn upon to usefully portray a mathematical description of the oxidation of
ammonia and nitrite. In the succeeding portions of this section, the impact of a variety of
environmental factors on the rates of growth and nitrification are considered. A combined
kinetic expression is then formulated which accounts for the effects of ammonia
concentration, temperature, pH and dissolved oxygen concentration.
At several points, reference is made to data developed from various types of nitrification
processes. Comprehensive descriptions of the various nitrification processes are presented in
Chapter 4 and will not be reproduced herein. One distinction that needs to be clearly
understood in discussions in this chapter is the difference between combined carbon
oxidation-nitrification processes and separate stage nitrification processes. The combined
carbon oxidation-nitrification processes oxidize a high proportion of influent organics
(BOD) relative to the ammonia nitrogen content. This causes relatively low populations of
nitrifiers to be present in the biomass. Separate stage nitrification systems, on the other
hand, have a relatively low BODs load relative to the influent ammonia load. As a result,
higher proportions of nitrifiers are obtained. Separate stage nitrification can be provided in
municipal treatment applications when a high level of organic carbon removal is provided
prior to the nitrification stage. This level of treatment is generally greater than provided by
primary treatment. Other differences between these classes of processes can be drawn, but
these are left for detailed discussion in Section 4.2.
3.2.5.1 Effect of Ammonia Concentration on Kinetics
A description of ammonia and nitrite oxidation can be derived from an examination of the
3-6
-------
growth kinetics of Nitrosomonas and Nitrobacter. Nitrosomonas' growth is limited by the
concentration of ammonia nitrogen, while Nitrobacter 's growth is limited by the
concentration of nitrite.
The kinetic equation proposed by Monodl3 is used to describe the kinetics of biological
growth of either Nitrosomonas or Nitrobacter:
M
K + S
s
where: M = growth rate of microorganisms, day ,
A _1
M = maximum growth rate of microorganisms, day ,
K = half velocity constant = substrate concentration, mg/1,
at half the maximum growth rate and
S = growth limiting substrate concentration, mg/1.
Since the maximum growth rate of Nitrobacter is considerably larger than the maximum
growth rate of Nitrosomonas, and since Ks values for both organisms are less than 1 mg/l-N
at temperatures below 20 C, nitrite does not accumulate in large amounts in biological
treatment systems under steady state conditions. For this reason, the rate of nitrifier growth
can be modeled with Equation 3-10 using the rate limiting step, ammonia conversion to
nitrite. For cases where nitrite accumulation does occur, other approaches are
available. 14, 15, 16
3.2.5.2 Relationship of Growth Rate to Oxidation Rate
The ammonia oxidation rate can be related to the Nitrosomonas growth rate, as follows:
where: MXJ = Nitrosomonas growth rate, day ,
H~, = peak Nitrosomonas growth rate, day ,
MN +
q^T = -y— = peak ammonia oxidation rate, Ib NH. - N
N oxidized/lb VSS/day,
q-^ = ammonia oxidation rate, Ib NH4 - N oxidized/lb VSS/day
Y^ = organism yield coefficient, Ib Nitrosomonas grown (VSS) per
Ib NH4 - N removed,
N = NH^ - N concentration, mg/1, and
KN = half-saturation constant, mg/1 NH4 - N, mg/1.
3-7
-------
In Equations 3-10 and 3-11, only the effect of ammonia concentration is considered; in later
sections, the effects of temperature, pH, and dissolved oxygen are considered.
3.2.5.3. Relationship of Growth Rate to Solids Retention Time
The growth rate of organisms can be related to the design of activated sludge systems by
noting the inverse relationship between solids retention time and growth rate of nitrifiers:
0 = 4_ (3-12)
c M
where: 6 = solids retention time, days.
The solids retention time can be calculated from system operating data by dividing the
inventory of microbial mass in the treatment system by the quantity of biological mass
wasted daily. Equations applicable for this calculation are presented in Section 4.3.3.
3.2.5.4 Kinetic Rate Constants for Temperature and Nitrogen Concentration
The most widely accepted kinetic constants for the nitrifers are those presented by Downing,
and coworkers. 1 ' > * ° Their results are presented in Figures 3-1 and 3-2. As can be seen, both
the maximum growth rate, /i, and the half saturation constants, Ks for Nitrosomonas and
Nitrobacter are markedly affected by temperature. Further, the maximum growth rate for
Nitrosomonas in activated sludge was found to be considerably less than for Nitrosomonas
in pure culture.
Kinetic constants found by other investigators are summarized in Tables 3-3 and 3-4. The
observations of maximum growth rates of Nitrosomonas of Gujer and Jenkins %
Wuhrmann^ Loehr, et al.H, Poduska and Andrews^, and Lawrence and Brown24; are
closer to the pure culture values of Nitrosomonas rather than the activated sludge values in
Figure 3-1. This suggests that some additional parameter such as dissolved oxygen (DO) may
have been limiting Downing's activated sludge measurements. ^ The influence of DO is
discussed in the next section. For illustrative use in this manual the pure culture values of
Downing, et al. for Nitrosomonas are used for considering the effect of temperature and
ammonia on growth and nitrification rates. The expressions for the half saturation constant
for oxidation of ammonia-nitrogen, KN, is:
sN (3-13)
T = temperature, C.
The expression for the effect of temperature on the maximum growth rate of Nitrosomonas
2N = 0.47e°-098 day'1 (3-14)
3-8
-------
FIGURE 3-1
4.0
r 3.0
2 2.0
TEMPERATURE DEPENDENCE OF THE MAXIMUM
GROWTH RATES OF NITRIFIERS
lu
O
(t
1.0
°-8
O.6
0.4
$
S
X O.2
S
< a
O.I
I
I
I
0.098 (T-15) —
= O.I8e
O.II6(T-I5)
I
12
16 20
T, TEMPERATURE, C
24
28
32
FIGURE 3-2
TEMPERATURE DEPENDENCE OF THE HALF SATURATION
CONSTANTS OF NITRIFIERS
16 2O
T, TEMPERATURE, C
28
32
3-9
-------
TABLE 3-3
MAXIMUM GROWTH RATES FOR NITRIFIERS
IN VARIOUS ENVIRONMENTS
Organism
Nitrosomonas
Nitrobacter
a.. , day- 1 at stated temperature , C
' JN
8
0.25
12
0.40
0.34
IS
0.21
0.28
16
0.57
20
0.71
0.48
0.5
. 21
0.85
0.65
0.34
23
0.37
1.08
1.44
25
0.17
0.55
0.53
prtf
Kei .
4
4
19
20
21
22
11
23
15
24
11,23
15
r i
tiii vironmsnt
Activated sludge, wash out
Activated sludge, math model
Activated sludge
Activated sludge
Activated sludge
Activated sludge
Synthetic river water
Activated sludge
Activated sludge
Synthetic river water
Activated sludge
TABLE 3-4
HALF-SATURATION CONSTANTS FOR
NITRIFIERS IN VARIOUS ENVIRONMENTS
Organism
Nitrosomonas
Nitrobacter
Ks, mg/l-N at stated temperature, C
15
2.8
0.5 to 1.0
0.7
20
3.6
0.5 to 1.0
1.0
0.5
1.1
0.07
25
0.37
3.4
3.5
0.25
0.7
5
28
5
30
10
6
32
8.4
Ref.
20
11,23
1,25
2
1,26
1,27
28
20
11,23
1,29
1,30
1,31
1,26
28
Environment
Activated sludge
Synthetic river water
Lab culture
Warburg analyses
Lab culture
Lab culture and activated
sludge
Lab culture
Activated sludge
Synthetic river water
Lab culture
Lab culture
Lab culture
Lab culture
Lab culture
Somewhat differing temperature effects have been found for attached growth systems.
Huang and Hopson's summary, with some modifications, is shown in Figure 3-3 for attached
growth systems.34 Downing and coworkers' relationship for Nitrosomonas (Equation 3-14)
is also shown for comparison purposes. Comparing the suspended growth and attached
nitrification data, one can conclude that attached growth systems have an advantage in
withstanding low temperatures (<15 C) without as severe losses in nitrification rates.
However, measurement of nitrification rates for suspended growth systems are not normally
made on the same basis as attached growth systems. In suspended growth systems, rates are
3-10
-------
expressed on a per unit of biomass basis (MLVSS is used). Precise measurement of biomass
is normally not possible in attached growth systems so other parameters are used such as
reaction rate per unit surface or volume. Therefore, attached growth systems can
compensate for colder temperature conditions by the effective slime growth growing
thicker. Thus, if rates could be expressed on a unit biomass basis for both system types,
reaction rate variation with temperature might be more similar.
It could be argued that compensation for low temperature in suspended growth systems
could be provided by an increase in mixed liquor level, much as an increase in slime growth
occurs in attached growth systems. However, suspended growth systems are limited by
reactor-sedimentation tank interactions which at cold temperatures might prohibit this due
to reduction of thickening rates of the sludge (cf. Section 4.10).
Other differences in reaction rates shown in Figure 3-3 may arise from the fact that some
determinations were on separate stage nitrification systems while others were made on
combined carbon oxidation-nitrification processes.
FIGURE 3-3
o
o
fc
o
-------
3.2.5.5 Effect of Dissolved Oxygen on Kinetics
The concentration of dissolved oxygen (DO) has a significant effect on the rates of nitrifier
growth and nitrification in biological waste treatment systems. The Monod relationship has
been used to model the effect of dissolved oxygen, considering oxygen to be a. growth
limiting substrate, as follows:
D0
= u
M
N K0 + DO
(3-15)
where: DO = dissolved oxygen, mg/1, and
Kn = half-saturation constant for oxygen, mg/1.
U2
British investigators found that the KQ2 value was about 1.3 mg/1 at an unspecified
temperature. 3 5 One U.S. investigator has suggested a relationship that would indicate
half-saturation constants of 0.15 mg/1 at 15 C and 0.42 mg/1 at 25 C, but did not provide
supporting data. 14 Studies conducted by the Los Angeles County Sanitation Districts at its
Pomona Water Renovation Plant represent one of the most careful attempts to evaluate the
effect of DO on nitrification rate.3° This facility is a combined carbon oxidation-
nitrification plant. Sludge samples were withdrawn, dosed with ammonia, and aerated at
various DO levels. Nitrification rates determined from the data collected are shown in Figure
3-4.36 Fitted to this data is a Monod expression for nitrification rate as a function of DO.
The KQO determined from this data is 2.0 mg/1. Temperature was not specified, but
indicated to be above 20 C.
Several investigations have provided indirect evidence of the importance of the effect of DO
on nitrification rate. A treatment plant operated continuously at a DO near 1 mg/1 gave
lower degrees of nitrification than plants held at 4 and 7 mg/1.3 ' When small scale activated
sludge plants were held at 1,2, 4, and 8 mg/1. British investigators found that the nitrifi-
cation rates at 2.0 mg/1 were about 10 percent lower than at higher levels of DO, although
nitrification was complete. 35 pilot investigations at the Metro Sewer District of Cincinnati,
Ohio, showed that when the DO was held at 2 mg/1, only about 40 percent nitrification
occurred, but when the DO was increased to 4 mg/1, about 80 percent nitrification took
place. 3° Murphy found that in two parallel activated sludge plants, that nitrification was
enhanced in the plant maintaining the DO at 8 mg/1 compared to the plant where the DO
was maintained at 1 mg/1. 3 9
The influence of DO on nitrification rates has been somewhat controversial, as examples of
plants can be found with completely nitrified effluents with operating DO levels of 0.5
mg/1. However, this type of evidence does not indicate that nitrification rate was unaffected,
merely that nitrification could be completed in the presence of a low DO level. Low
nitrification rates, depressed by low DO levels, can still be sufficient to cause complete
nitrification if the aeration tank detention time is large enough.
3-12
-------
FIGURE 3-4
EFFECT OF DISSOLVED OXYGEN ON NITRIFICATION RATE
Ul
^ 0.15
Q:
S Q
9\
* V) 0.10
O tn
lU ^
Q 2
* \
~ ^ 0.05
3
o
° )
.0 + DO /
Data of Nagel and Haworth (Ref. 36)
I I I I
O.5 1.0 1.5 2.O
DO, DISSOLVED OXYGEN, mg/l
2.5
3.0
While the general effect of DO on kinetics is firmly established, there needs to be further
study to determine the factors affecting the values of KQ2- All of the various estimates are
from systems where combined carbon oxidation-nitrification is practiced and no measure-
ments have been made on separate stage nitrification systems. Kc>2 values for separate stage
nitrification systems may very well be different than those for combined carbon
oxidation-nitrification systems. Further refinement of KQ2 values can be expected. For
illustrative use in this manual, a value of KQ2 °f 1 -3 rng/1 has been assumed. This value falls
in the middle of the range of KQ2 observations (0.15 to 2.0 mg/l) and is of a magnitude
such that if the operating DO is 2.0 mg/l or less, the nitrification (or nitrifier) growth rate is
60.6 percent (or less) of the peak rate. This order of reduction in rate could account for
most of the difference in growth rate observed by Downing, et al. (see Section 3.2.5.4)
between river water values at high DO levels and activated sludge operating data at DO levels
of 2.0 mg/l of less.
3.2.5.6 Effect of pH on Kinetics
The hydrogen ion concentration (pH) has been generally found to have a strong effect on
the rate of nitrification. Figure 3-5 presents typical pH relationships from a number of
investigations. The results of other investigations have been summarized in the litera-
ture. 1>44 There is a wide range in reported pH optima; the almost universal finding is that as
the pH moves to the acid range, the rate of ammonia oxidation declines. This has been
found to be true for both unacclimated and acclimated cultures, although acclimation tends
3-13
-------
to moderate pH effects. The findings for an attached growth reactor (Curve E, Figure 3-5)
are very similar to the findings for an activated sludge (Curve C). In neither case were the
cultures acclimated to each pH value prior to determining nitrification rates. When a
three-week acclimation period was provided for the attached growth reactor, it was found
that the rate at pH 6.6 rose to 85 percent of the optimum rate at pH 8.4 to 9.0.34
Various investigators have reported the effects of pH depression on nitrification. For
instance, in an activated sludge with insufficient wastewater alkalinity, pH values of 5 to 5.5
were attained. This high acid concentration resulted in a cessation of nitrification; at the
same time sludge bulking occurred. The point at which the rate of nitrification decreased
was pH 6.3-6.7.1 Parallel investigations on air and high purity oxygen-activated sludge
systems at the Blue Plains Treatment Plant, Washington, D.C., have shown that depressed
pH values in the last oxygen activated sludge stage produced slightly lower nitrification rates
than when the system was operated at higher pH.45 Further information on the EPA
investigations at the Blue Plains Treatment Plant is presented later in Section 4.6.5.
In a study of the effect of abrupt changes in pH, it was found that an abrupt change in
reactor pH from 7.2 to 6.4 caused no adverse effects. However, when the pH was abruptly
changed from 7.2 to 5.8, nitrification performance deteriorated markedly as effluent
ammonia levels rose from approximately zero to 11 mg/1 NH^-N. A return to pH 7.2 caused
rapid improvement indicating that the lowered pH was only inhibitory and not toxic.'^
Haug and McCarty showed that nitrifiers could adapt to nitrify at pH levels as low as 5.5 to
6.0.2 However, since the concentration of biomass in their column was not defined at each
pH level, no conclusions can be drawn from their work as to the effect of pH on the peak
ammonia oxidation rate, qj^f.
For illustrative use in this manual, the equation of Downing, et alA3, showing the effect of
pH on nitrification is adopted. For pH values < 7.2:
MN = £N (1 - 0.833(7.2 - pH)) (3-16)
For pH levels between 7.2 and 8.0, the rate is assumed constant. This expression was
developed for combined carbon oxidation-nitrification systems, but its application to
separate stage nitrification systems is probably conservative.
Because of the effect of pH on nitrification rate, it is especially important that there be
sufficient alkalinity in the wastewater to balance the acid produced by nitrification.
Equation 3-3A (Table 3-1) indicates that 7.14 mg of alkalinity are destroyed per mg of
ammonia nitrogen. Caustic or lime addition may be required to supplement moderately
alkaline wastewaters. Design considerations for pH control are presented in Section 4.9.
3.2.5.7 Combined Kinetic Expressions
In previous sections, the effect of ammonia level, temperature, pH and dissolved oxygen on
nitrification rate has been presented. In all practical systems, these parameters act to affect
3-14
-------
FIGURE 3-5
EFFECT OF pH ON NITRIFICATION RATE
(AFTER SAWYER, ET AL.)
v, ^ PERCENT OF MAXIMUM
i m RATE OF NITRIFICATION
» -<
O f\> •& O> Q> Q
r o» ° O O O O O
i i i i
/ .
- / /
r /
-/ (vf> '
\y 7 '?
^Js?
* \ i i i
/'
j^ •
^ff •
r^f *
f *
f •
•
»
B:
*
i i /i i
^7 "^
*
i i 11
vS>' ' "
\\v
• X .
• Nk \
• \. ^
* >^
_9 ^
*
:B
i i: i i
O 7.O 8.0 9.O 10.0
ENVIRONMENT REFERENCE
A Nitrosomonas- pure culture Engle and Alexander ( 40)
B Nitrosomonas - pure culture Myerhof (41)
C Activated sludge at 20 C Sawyer, et al. (42)
D Activated sludge Downing, et al. (43)
E Attached growth reactor at 22 C Huang and Hopson (34)
the nitrification rate simultaneously. It has been shown that me combined effect of several
limiting factors on biological growth can be introduced as a product of Monod-type
factors: 46
N
K
N
Lv
N
(3-17)
where: L =
N =
P =
KL,KN,andKp =
concentration of growth limiting substance L,
concentration of growth limiting substance N,
concentration of growth limiting substance P, and
half-saturation constants for substances L, N, and P, respectively.
3-15
-------
This concept has also been applied to the analysis of algal growth kinetic data^7 and to
denitrification kinetics.'*"
Taking this approach for nitrification, the combined kinetic expression for nitrifier growth
would take the form:
(1 - 0.833(7.2 - PH)) (3-18)
where:
= maximum nitrifier growth rate at temperature, T, and pH.
Downing, et al. 43 used this procedure to describe nitrifier growth rates, excepting that no
term for DO was included. Using specific values for temperature, pH, ammonia and oxygen,
adopted in this manual in Sections 3.2.5.4, 3.2.5.5., and 3.2.5.6, the following expression
results for pH < 7.2 for Nitrosomonas valid for temperatures between 8 and 30 C:
= 0.47
2 _pR)
N
N + 10
0.051T - 1.158
DO
DO + 1.3
day
-1
(3-19)
The first term in brackets accommodates the effect of temperature. The second term in
brackets considers the effect of pH. For pH > 7.2, the second quantity in brackets is taken
to be unity. The third term in brackets is the Monod expression for the effect of ammonia
nitrogen concentration. Similarly, the fourth term in brackets accounts for the effect of DO
on nitrification rate. Equation 3-19 has been adopted for illustrative use in this manual. As
more reliable data becomes available, Equation 3-19 can be modified to suit particular
circumstances.
An example evaluation of Equation 3-19 at T = 20 C, pH = 7.0, N = 2.5 mg/1, and DO equal
to 2.0 mg/1 is as follows:
MN = 0.47 [1.63] [.833] [0.775] [0.606] = 0.300 day
"1
The ammonia removal rate is defined as done previously (Equation 3-11):
qN =
DO ) 0 - °-833<7-2 -
3-16
-------
where: Q = -~— (3-21)
In the numerical example above, with a yield coefficient of 0.15 Ib VSS per Ib NIfy-N
removed, the nitrification rate is 2 Ib NH^-N oxidized per Ib VSS per day. This rate is
expressed per unit of nitrifiers, assuming that there are no other types of bacteria in the
population. Nitrification rates of a comparable magnitude have been found experimentally
by a number of investigators for laboratory enrichment cultures of nitrifiers.2,23,28 AS wjn
be seen in Section 3.2.7, nitrification rates in mixed cultures are much lower.
3.2.6 Population Dynamics
In previous sections, the kinetics of the growth of nitrifiers have been presented. In all
practical applications in wastewater treatment, nitrifier growth takes place in waste
treatment processes where other types of biological growth occurs. In no case are there
opportunities for pure cultures to develop. This fact has significant implications in process
design for nitrification.
In both combined carbon oxidation-nitrification systems and in separate stage nitrification
systems, there is sufficient organic matter in the wastewater to enable the growth of
heterotrophic bacteria. In this situation, the yield of heterotrophic bacteria growth is greater
than the yield of the autotrophic nitrifying bacteria. Because of this dominance of the culture,
there is the danger that the growth rate of the heterotrophic organisms will be established at
a value exceeding the maximum possible growth rate of the nitrifying organisms. When this
occurs, the slower growing nitrifiers will gradually diminish in proportion to the total
population and be washed out of the system.43
Thus, for consistent nitrification to occur, the following design condition must be satisfied,
assuming pH and DO do not limit nitrifier growth:
A
MN > Mb (3-22)
where: JUN = maximum growth rate of the nitrifying population,
ju, = growth rate of the heterotrophic population.
Reduced DO or pH can act to depress the peak nitrifier growth rate and cause a washout
condition. An expression for this possible reduced rate of growth is:
„ = - DO - ^ (1 _ 0.833(7.2 - pH)) (3-23)
MN MN I K + DO
3-17
-------
where: MXT = maximum possible nitrifier growth rate under environmental
conditions of T, pH, DO, and N ^K.
As before, the last expression in brackets is taken to be unity above a pH of 7.2. The
relationship between actual nitrifier growth rate and maximum possible nitrifier growth rate
can easily be seen from Equations 3-8 and 3-23:
• / N \ ,o ^A-\
"N = *N ^K^TN j (3'24)
A more rigorous condition for prevention of nitrifier washout than Equation 3-22 is:
M ^ M (3-25)
More typically, Equation 3-25 is expressed in reciprocal form in terms of solids retention
time as: 1 1
0d ^ em (3-26)
^
where: 6 = solids retention time of design, days, and
L/
6 = minimum solids retention time, days, for nitrification at given pH,
temperature and DO.
Since n or#c is fixed by the environmental conditions (T, pH and DO), Equations 3-25 or
3-26 is satisfied by modifying/^ or.0c. The various ways of satisfying these relationships can
be established by examining the following growth relationship for the heterotrophic
population ;10,11,12
c
where: Y, = heterotrophic yield coefficient, Ib VSS grown per Ib of substrate
(BOD or COD) removed, and
q, = rate of substrate removal, Ib BOD (or COD) removed/lb VSS/day,
-1
K, = "decay" coefficient, day , and
ju« = net growth rate for heterotrophic population.
3-18
-------
The rate of substrate removal is defined as:
so - si
where: SQ = influent total BOD (or COD), mg/1,
Sj = effluent soluble BOD (or COD), mg/1,
HT = hydraulic detention time, days, and
Xj = MLVSS, mg/1.
d
Since both Yfc and Kd are assumed to be constant, the only way /^ or 6 c can be
manipulated is by altering q^. One way the substrate removal rate can be reduced is
to place an organic carbon removal step ahead of the nitrificaiian stage, creating a "separate
stage" nitrification process. The result of this procedure is to reduce the food available to
the heterotrophic bacteria and to lessen their dominance in controlling the solids retention
time. Separate nitrification stages can have very long solids retention times ( flf. = 15 to 25
days). Another procedure for reducing the substrate removal rate, without separating the
carbon oxidation and nitrification processes, is to increase the biological solids in the
system. This can be done by increasing the concentration of biological solids under aeration
(the MLVSS in the activated sludge system) or by increasing the volume of the oxidation
tank while maintaining the concentration of biological solids at the same concentration.
The level set for biological solids retention time, 6 c , establishes the biological solids
retention time (or growth rate) of the nitrifiers, since selective wasting of the heterotrophic
population is not practical. Therefore, the design solids retention time can be related to the
effluent ammonia level through the Monod relationship and the inverse relationship between
nitrifier growth rate and solids retention time (Equations 3-12 and 3-24). With specific
values of 0 £ > T, pH and DO, Equations 3-12 and 3-24 can be solved for the effluent
ammonia level. Figure 3-6 was developed by such a procedure, with the assumption of T =
15 C, DO = 2 mg/1, and pH >7.2 <8. Also plotted is the nitrification efficiency, assuming
an influent Total Kjeldahl Nitrogen (TKN) concentration of 25 mg/1 and that all of the
influent nitrogen is available for nitrification. As can be seen, significant breakthrough of
ammonia from the system does not occur unless the solids retention time is reduced below 5
days. At that point, ammonia nitrogen breakthrough is very abrupt, rising from 1 mg/1 at
6 c = 4.9 days to 15 mg/1 at 6 c = 3.6 days. The principal cause of the sharpness of the
ammonia breakthrough is due to the low value of KN (0.40 mg/1 NHlj. — N in this case). A
number of investigators have experimentally determined very similar relationships to that
shown in Figure 3-6 J 9, 20, 21
Lawrence and McCarty^l have introduced the concept of a safety factor (SF) in the
• application of biological treatment process kinetics to design. The safety factor was defined
3-19
-------
as the ratio of the design solids retention time to the minimum solids retention time; the
safety factor can also be related to the nitrifier growth rates through Equation 3-12. The
expression for the SF is:
SF =
(3-29)
A conservative safety factor was recommended to minimize process variations caused by pH
extremes, low DO concentrations and toxicants. 11 Also, the SF can be used to ensure that
ammonia breakthrough does not occur during diurnal peaks in load (see Section 4.3.3).
Interestingly, Equation 3-24 can be manipulated to show that the specification of a SF of 2
will establish an ammonia level equal to KN in the effluent of a complete mix activated
sludge plant, if there is a high DO level (DO not limiting). This is because the process will be
operating at one-half its maximum growth rate. Recall that the half-saturation constant, KN,
in this Monod expression is defined as the level of substrate which will cause the organism to
FIGURE 3-6
EFFECT OF SOLIDS RETENTION TIME ON EFFLUENT AMMONIA
CONCENTRATION AND NITRIFICATION EFFICIENCY
25
20
\ IS
o>
6
-
z
i
2 '°
2
Ul
-J
I1- 5
u.
UJ
D
—
—
i-
—
—
—
_
-
1 x*» "" " 1 | 1 I 1
s
^.
PERCENT REMOVAL
— *~ —
—
—
•
—
/-EFFLUENT AMMONIA
\ / ~~
v/
r* — . — ' * ' ' i
IUU
90
2
80 £
tt
iu
0.
70
>?
60 2
0
u.
SO u.
Uj
40 o
30 i
U.
20 t
*
IO
IO 15 2O 25
DESIGN SOLIDS RETENTION TIME, DAYS
3O
35
3-20
-------
grow at half its maximum rate. Many nitrifying activated sludge plants have been observed
to have 1 to 2 mg/1 of ammonia in their effluents, values close to the theoretical value of
KN-
Criteria for establishing the safety factor are presented in Chapter 4. Furthermore, specific
examples of the use of the kinetic expressions developed in this section are presented in
Sections 4.3.3. and 4.3.5.
3.2.7 Nitrification Rates in Activated Sludge
The basic design approach for separate stage nitrification systems (activated sludge type) has
been on a different basis than for combined carbon oxidation-nitrification systems. Rather
than use the sludge growth rate or solids retention time approach described in Section 3.2.6,
the practice frequently has been to base reactor sizing for separate stage systems on the basis
of nitrification rates in terms of Ib NH^-N oxidized/lb MLVSS/day.5>42,49 However, it will
be shown that this parameter is fundamentally related to the nitrification kinetics previously
presented in this chapter.
The nitrification rate can be calculated from the ammonia oxidation rate, q^, by
recognizing that the nitrifiers are only a fraction of the total mass of biological solids in a
nitrification system. The other biological solids in the system result from the growth of the
hetero trophic population. On this basis, the nitrification rate, r^, is as follows:
(3-30)
(3-3D
where: f = nitrifier fraction of the mixed liquor solids, and
rN = nitrification rate, Ib NH* - N oxidized/lb MLVSS/day.
A similar expression for the peak nitrification rates (in activated sludge) is:
/\ A „
where: rN = peak nitrification rate, Ib NH. - N oxidized/lb MLVSS/day.
This latter rate is normally determined experimentally in activated sludge systems, as will be
described in Section 4.6.3.
Specific analytical techniques for determination of the nitrifier fraction have not as yet been
developed. However, f can be estimated from knowledge of the biological yields of the
auto trophic and heterotrophic populations, as follows:
3-21
-------
M
f =
N
(3-32)
where: MN = nitrifiers grown through oxidation of ammonia, and
M = heterotrophs grown through oxidation of organic carbon;
c
MXT and M can be estimated as follows:
N c
MN =
(3-33)
where: N~ =
TKN in the influent, mg/1, and
= NH - N in the effluent, mg/1.
Mc = Vso-V
(3-34)
where: Sn =
Sl =
carbon (BOD5 or COD) in the influent, mg/1
carbon (BOD5 or COD) in the effluent, mg/1, and
net yield of VSS of heterotrophs per unit of carbon (BOD, or
COD) removed.
This procedure neglects the ammonia assimilated by heterotrophic growth and therefore is
approximate. A further approximation is that the net yield of the heterotrophs has been
assumed constant, whereas it is known that it varies with solids retention time.
Several examples can be drawn for separate stage nitrification systems as compared to
combined carbon oxidation-nitrification systems. In a separate stage system, illustrative
values of BOD5 removed and TKN oxidized would be 50 mg/1 and 25 mg/1, respectively. A
reasonable estimate for YN is 0.15 Ib/lb NH^-N rem. (Section 3.2.2) and for Yb (BOD),
0.55 Ib/lb BOD5 removed. Thus, for this separate stage example, f can be calculated to be:
f =
0.15(25)
0.55(50 + 0.15(25)
= 0.12
A similar example can be drawn for combined carbon oxidation-nitrification systems.
Assuming 200 mg/1 of BOD5 removed and 25 mg/1 of TKN oxidized, f can be calculated to
be:
f =
0.15(25)
0.55(200) + 0.15(25)
= 0.033
3-22
-------
Thus, it can be seen that the fraction of nitrifiers is lower in combined carbon
oxidation-nitrification systems that in separate stage nitrification systems.
Equation 3-32 can be reexpressed in terms of the BOD5/TKN ratio in the influent, by
assuming effluent BOD and ammonia are negligibly small:
f =
1
!o_iL
N0 YN
(3-35)
Table 3-5 presents numerical values for the fraction of nitrifiers, using Equation 3-35, and a
condition where YN = 0.15 and Y^ (BOD5) = 0.55. For most separate stage nitrification
systems, the BOD5/TKN ratio is greater than 1.0 and less than 3.0 (Section 4-2). Thus even
in separate stage systems, the fraction of nitrifiers is relatively low. For the assumed yield
values, the fraction is less than 20 percent and greater than 8 percent. It must be emphasized
that the values of nitrifier fraction given in Table 3-5 are estimates only, and not supported
by actual measurements of nitrifier fractions. Table 3-5 does have value in that it shows, at
least qualitatively, the impact of influent BOD5/TKN ratio on the fraction of nitrifiers in a
nitrification system. The influence of this ratio on nitrifier fraction and nitrification rates
was recognized as early as 1940, when Sawyer showed that the BOD5/NH3 ratio correlated
with the nitrifying ability of various activated sludges.^O
The effect of ammonia concentration on the nitrification rate is portrayed in Figure 3-7.
The assumptions made were: f = 0.1, pH > 7.2 <8, and the DO = 2.0 mg/1. Equations 3-20,
3-21, and 3-27 were employed to construct the figure. As can be seen, the nitrification rates
TABLE 3-5
RELATIONSHIP BETWEEN NITRIFIER FRACTION
AND THE BOD5/TKN RATIO
BOD /TKN ratio
o
0.5
1
2
3
4
Nitrifier
•a
fraction
0.35
0.21
0.12
0.083
0.064
BOD /TKN ratio
0
5
6
7
8
9
Nitrifier
fraction
0.054
0.043
0.037
0..033
0.029
Using Equation 3-35 and YN=0.15, Y = 0.55.
3-23
-------
are relatively unaffected by ammonia concentration above 2.5 mg/1 ammonia N. As the
nitrification rates approach their plateau values, nitrification approaches a zero order rate,
uninfluenced by ammonia level. It has been shown that for suspended growth processes, the
rate of removal approximates a zero order reaction.42,45,51 However, in none of these
cases were nitrification rates determined at ammonia concentrations at or below the value of
where non-zero order rates effects would be evident.
FIGURE 3-7
EFFECT OF AMMONIA CONCENTRATION ON NITRIFICATION RATE
O.3
f = 0.1
D0= 2.0 mg/l
pH >7.2 <8
10 2O
NH4-N, mg/l
30
3-24
-------
The fraction of nitrifiers has a marked effect on the nitrification rates. Figure 3-8,
demonstrating this effect, was developed similarly to Figure 3-7, excepting that the effluent
ammonia nitrogen concentration was assumed to be 2.5 mg/1. A principal means of
increasing the nitrification rate is to increase the fraction of nitrifiers. From Table 3-5, it can
be seen that this can be accomplished by lowering the BOD5/TKN ratio. In terms of plant
FIGURE 3-8
EFFECT OF TEMPERATURE AND FRACTION OF
NITRIFIERS ON NITRIFICATION RATE
1.2
CO
00
•J
N
Q
I
•s
QQ
-J
UJ
ft:
o
o
it
ftl
1.O
0.8
0.6
O.4
0.2
DO = 2.0 mg/l
NO =2.5 mg/l NH4-N
pH >7.2 <8
10
I
20
TEMPERATURE, C
30
3-25
-------
design, the BOD5/TKN ratio can be altered by increasing the organic carbon removal ahead
of the nitrification unit.
It must be emphasized that the nitrification rates developed in this section are only
estimated relationships based on theoretical considerations. Actual measured values are
presented in Section 4.6.3.
As a practical example of the effect of the BOD5/TKN ratio on nitrification rates, rate data
for an attached growth system^ are plotted against the BOD5/TKN ratio in Figure 3-9. The
effect of BOD5 in the synthetic waste was to displace nitrifiers with heterotrophic bacteria in
the bacterial film, thereby reducing the nitrifier fraction. and the nitrification rate.
Interestingly, a small amount of BOD5 ^10 mg/1) was found to enhance the nitrification ,
rate.
FIGURE 3-9
EFFECT OF BOD5/TKN RATIO ON NITRIFICATION
RATE - EXPERIMENTAL ATTACHED GROWTH SYSTEM
100
o
II
10
Q
O
03
Uj
u.
o
Uj
o
-------
3.2.8 Nitrification Rates in Trickling Filters and Other Attached Growth Systems
Discussion of kinetic rates in the previous sections has been primarily oriented to
nitrification in activated sludge type (suspended growth) nitrification systems, although
some comparisons have been drawn with attached growth system measurements for
comparative purposes. The growth and oxidation rate relationships presented in Sections
3.2.5.7 and 3.2.6 are directly applicable'to design of suspended growth systems, as will be
shown in Chapter 4. While these relationships are operative in attached growth systems, their
application is complicated by the fact that oxygen mass transfer limitations through the
bacterial slimes may limit reaction rates in some situations. A biofilm model has been
developed which yields insight into which factors are controlling^,53 bu{ jt ^as no{ yet
been extended to the point where it can be applied directly to design applications. As a
consequence, the design relationships presented for attached growth nitrification in Chapter
4 are empirically based and therefore, less theoretically precise than those developed for
suspended growth sytems. Though the design relationships presented are empirical, where
possible the loading relationships are presented on a basis that is at least consistent with the
biofilm model. For instance, ammonia nitrogen oxidation rates are expressed on a surface
area basis when describing separate stage nitrification in trickling filters and the rotating
biological disc process (Sections 4.7.1 and 4.7.2).
Some of the conclusions that can be drawn from the biofilm model are of interest in
considering surface ammonia removal rates in attached growth systems. The biofilm model
shows that the ammonia oxidation rate in attached growth systems should not be decreased
as drastically under adverse environmental conditions as in suspended growth systems.52
This finding is consistent with the observation made in Section 3.2.5.4, namely that
attached growth systems have an advantage over suspended growth systems in withstanding
lower temperatures without as severe losses in nitrification rates.
The biofilm model also shows that the dissolved oxygen concentration must be 2.7 times
the ammonia nitrogen concentration to prevent oxygen transfer from limiting nitrification
rates in attached growth systems.52 Two operational procedures have been suggested to
overcome this limitation: (1) dilution of the ammonia nitrogen through recirculation and
(2) increasing the oxygen transfer through the use of high purity oxygen.5 2 The first
recommendation has been made in this manual in Sections 4.4.1.4 and 4.7.1.3.
3.2.9 Effect of Inhibitors on Nitrification
Certain heavy metals and organic compounds are toxic to nitrifiers. To date, these effects
have not been quantitatively incorporated into the kinetic description of nitrifier growth,
although such approaches have been used to describe toxicity in other biological systems. A
listing of substances toxic to unacclimated nitrifying organisms is presented in Table 3-6,
which is drawn primarily from the review by Painter. 1
3-27
-------
Sawyer, on reviewing the English literature, suggests that 1 0 to 20 mg/1 of heavy metals can
be tolerated due to the low ionic concentrations at high pH values of 7.5 to 8.0.^6 it has
also been pointed out that precipitated metals (such as hydroxides) that concentrate in the
sludge can be disastrous to the sludge if the pH falls and the precipitate dissolves. Such
conditions may occur in the sludge collection zone of the secondary clarifier where
continuing organism activity may cause low pH values. Alternatively, low pH values may
occur when pH control systems fail.
found that silver (Ag) was extremely toxic to nitrification of secondary effluent on
a plastic media trickling filter. Levels of 2 ppb in the influent to the filter were concentrated
to 5 ppm in the biomass on the media. This inhibitory effect was found to severely reduce
allowable loading rates and result in only partial nitrification.
TABLE 3-6
COMPOUNDS TOXIC TO NITRIFIERS
(AFTER PAINTER (1))
Organics
Thiourea
Allyl-thiourea
8-hydroxyquinoline
Salicyladoxine
Histidine
Amino acids
Mercaptobenzthiazole
Perchloroethylene'3
Trichloroethylenec
Abietec acid0
Inorganics
Zn
OCN
-1
-1
Cu
Hg
Cr
Ni
Ag
Also reference 54
Reference 5
Reference 55
The rate and change of magnitude of environmental conditions are nearly as critical to the
biomass as the conditions themselves. It has been shown that nitrifiers can adapt to toxic
substances when they are consistently present at concentrations higher than cause toxic
effects in slug discharges. 1 >$& Unfortunately, slug discharges are often present in municipal
systems and can result from industrial dumps or from urban stormwater inflow.
Under the unusual conditions of discharge of highly concentrated industrial wastes into
municipal systems that contain either nitrite or ammonia, the resulting high concentrations
of ammonia or nitrite in the municipal waste can be temporarily toxic to the nitrifying
3-28
-------
population. 5 9 When these conditions are suspected, the reader is referred to reference 59
which contains charts which allow the identification of the regions of toxicity. Under
normal municipal conditions of pH and concentrations complete nitrification will occur.
When concentrated industrial wastes are present, slug discharges should be avoided; rather,
storage facilities should be provided so that wastes can be metered into the collection
system at a rate sufficient to ensure dilution to safe loads.
In sum, the possibility of toxic inhibition must be recognized in the design of nitrification
systems. Either implementation of source control programs or inclusion of upstream
toxicity removal processes may be required, particularly in those cases where significant
industrial dischargers are tributary to the collection system.
3.3 Denitrification
The biological process of denitrification involves the conversion of nitrate nitrogen to a
gaseous nitrogen species. The gaseous product is primarily nitrogen gas but also may be
nitrous oxide or nitric oxide. Gaseous nitrogen is relatively unavailable for biological
growth, thus denitrification converts nitrogen which may be in an objectionable form to one
which has no significant effect on environmental quality.
As opposed to nitrification, a relatively broad range of bacteria can accomplish
denitrification, including Psuedomonas, Micrococcus, Archromobacter and Bacillus. These
groups accomplish nitrate reduction by what is known as a process of nitrate dissimilation
whereby nitrate or nitrite replaces oxygen in the respiratory processes of the organism under
anoxic conditions. Because of the ability of these organisms to use either nitrate or oxygen
as the terminal electron acceptors while oxidizing organic matter, these organisms are
termed facultative heterotrophic bacteria.
Confusion has arisen in the literature in terminology; the process has been termed anaerobic
denitrification. However, the principal biochemical pathways are not anaerobic, but merely
minor modifications of aerobic biochemical pathways. The term anoxic denitrification is
preferred, since it describes the environmental condition of the absence of oxygen, without
implying the nature of the biochemical pathways.
3.3.1 Biochemical Pathways
Specific information on the specific biochemical reaction intermediates involved in
denitrification are available in the literature^ and only certain concepts are of interest in
process design applications. Denitrification is a two-step process in which the first step is a
conversion of nitrate to nitrite. The second step carries nitrite through two intermediates to
nitrogen gas. This two-step process is normally termed "dissimilation."
Denitrifiers are also capable of an assimilation process whereby nitrate (through nitrite) is
converted to ammonia. Ammonia is then used for the bacterial cell's nitrogen requirements.
3-29
-------
If ammonia is already present, assimilation of nitrate need not occur to satisfy cell
requirements.
As will be shown in Section 3.3.2, electrons pass from the carbon source (the electron
donor) to nitrate or nitrite (the electron acceptor) to promote the conversion to nitrogen
gas. This involves the nitrifiers "electron transport system" and is involved with the release
of energy from the carbon source for use in organism growth. It happens that this electron
transport system is identical to that used for respiration by organisms oxidizing organic
matter aerobically, except for one enzyme. Because of this close relationship, many
facultative bacteria can shift between using oxygen or nitrate (or nitrite) rapidly and
without difficulty.
3.3.2 Energy and Synthesis Relationships
The use of oxygen as the final electron acceptor is more energetically favored than the use
of nitrate. Table 3-7 compares the energy yields per mole of glucose when oxygen and
nitrate are used as electron acceptors.61 The greater free energy released for oxygen favors
its use whenever it is available. Therefore, denitrification must be conducted in an anoxic
environment to ensure that nitrate, rather than oxygen, serves as the final electron acceptor.
TABLE 3-7
COMPARISON OF ENERGY YIELDS OF NITRATE DISSIMILATION
VS OXYGEN RESPIRATION FOR GLUCOSE
Reaction
Energy yield
per mole glucose,
kilocalories
Nitrate dissimilation
5C-H,,O_ + 24KNO,-*- 30 CO0 + 18H0O + 24KOH + 12N
D-l^D *3 / £
570
Oxygen respiration
686
C.H.0OC + 600 -^6CO0 + 6H.O
o 1 Z b £i L &
Reference 61
Methanol, rather than glucose or any other organic, has seen widest use as the electron
donor in the U.S. (see Section 3.3.4). Using methanol as an electron donor and neglecting
synthesis for the moment, denitrification can be represented as a two-step process as shown
in Equations 3-36 and 3-37:
3-30
-------
First Step
NO~ + 0.33CH3OH = NO~ + 0.67 H2O (3-36)
Second Step
NO~ + 0.5CH3OH = 0.5 N2 + 0.5 CO2 + 0.5 H2O + OH~ (3-37)
The overall transformation is obtained by addition of Equations 3-36 and 3-37 to yield
Equation 3-38:
0.833 CH3OH = 0.5 N2 + 0.833 CO2 + 1.167H2O + OH~ (3-38)
In this equation, methanol serves as the electron donor and nitrate as the electron acceptor.
This can be shown by splitting up Equation 3-38 into the following oxidation-reduction half
reactions:
NO" + 6H+ + 5e~ - - 0.5 N2 + 3 H2O (3-39)
(electron acceptor)
0.833 CH3OH + 0.833 H2O •- 0.833 CO2 + 5H+ + 5e (3"40)
(electron donor)
Equation 3-38 can be obtained by adding Equations 3-39 and 3-40 and the following
equation for water:
H2O = H+ + OH"
From Equations 3-39 and 3-40 the meaning of the terms of electron donor and acceptor are
clear. Nitrate gains electrons and is reduced to nitrogen gas, hence it is the electron acceptor.
The carbon source, methanol, loses electrons and is oxidized to carbon dioxide, hence it is
the electron donor.
As mentioned in Section 3.2.2, these reactions take place in the context of the carbonic acid
system. Equations 3-36 and 3-38 have been modified in Table 3-8 to reflect the fact that
hydroxide (OH~) produced reacts with carbonic acid (carbon dioxide) to produce
bicarbonate alkalinity. Also shown in Table 3-8 is the equation of synthesis for those
organisms deriving energy through nitrate respiration.^^
3-31
-------
The equations for energy yielding reactions (Equations 3-36 and 3-37) have been combined.
with the equation for oganism synthesis (Equation 3-41, Table 3-8) through knowledge of
organism yields and are summarized in Table 3-9. Also, shown for completeness is the
combined expression for oxygen respiration (Equation 3-44) since, if any oxygen is present,
it will be used preferentially. Similar expressions can be developed for other organic sources
serving as electron donors if organism yields are known.63
TABLE 3-8
RELATIONSHIPS FOR NITRATE DISSIMILATION AND GROWTH IN
DENITRIFICATION REACTIONS
Reaction
Equation
Equation
No.
Nitrogen dissimilation
Nitrate to nitrite
Nitrite to nitrogen gas
Nitrate to nitrogen gas
Synthesis - denitrifiers
NO, +0.33CH,OH =
«5 o
NO" +0.33H.O +0.33H CO,
Z Z Z - o
NO, +0.5 CH,OH + 0.5HCO, =
Z o Z O
0.5N + HCO~ + HO
Z o Z
NO3 + 0.833
+ 0.167H2CO3
= 0.5 N2 4 1 .33 H2O + HCO~
14CH3OH + 3N03
3C H O N + 20 HO + 3HCO~
O / Z Z O
3-36A
•3-37A
3-38A
3-41
The theoretical methanol requirement for nitrate reduction, neglecting synthesis, is 1.9 mg
methanol per mg nitrate N (Equation 3-36, Table 3-9). Including synthesis (Equation 3-42),
the requirement is increased to 2.47. Similarly, Equations 3-43 and 3-44 in Table 3-9 allow
calculation of methanol requirements for nitrite reduction and deoxygenation to allow a
combined expression to be formulated for the methanol requirement:62
Cm = 2.47NO3-N + 1.53NO~-N + 0.87 DO
(3-45)
3-32
-------
where:
m
NO" - N =
NO" - N =
DO
required methanol concentration, mg/1,
nitrate concentration removed, mg/1,
nitrite concentration removed, mg/1, and
dissolved oxygen removed, mg/1.
Biomass production can be calculated similarly:
C. = 0.53 NO" -N + 0.32 NO" -N + 0.19 DO
where: C. = biomass production, mg/1.
(3-46)
TABLE 3-9
COMBINED DISSIMILATION-SYNTHESIS EQUATIONS
FOR DENITRIFICATION (AFTER MC CARTY, ET AL. (62))
Transformation
Equation
Equation
No.
Overall nitrate removal
Overall nitrite removal
Overall deoxygenatlon
1.08CH3OH
0.056C5H7N02 + 0.47N2 + 1.68H2O + HCO~
0.53H2CO3 + 0.67CH3OH =
1.23H2O + 0.48N2 + HCO~
O, + 0.93CH.OH + 0.056NO, =
0.056C..H NO + 1.04H O + 0.59H CO, + 0.056HCO,
O / £, £, Z i3 O
3-42
3-43
3-44
Most experimental data is expressed in terms of the "M/N ratio," which is the mg of
methanol per mg of initial nitrate nitrogen concentration. The ratio includes the
requirements for nitrite and oxygen, which are usually small relative to the nitrate
requirement. For instance, for a NO§" value of 25 mg/1 of nitrate -N, 0.5 mg/1 nitrite -N and
3.0 mg/1 dissolved oxygen, the methanol requirement can be calculated to be 64.1 mg/1
from Equation 3-45. The M/N ratio is therefore 2.57 (64.4/25), which is only 4 percent
greater than the requirement for nitrate alone (2.47).
Values of the "M/N" ratio required for complete denitrification range from the levels
estimated from Equation 3-45 at 2.5 Ib methanol per Ib of nitrate nitrogen removed up to
3.0 Ib methanol per Ib of nitrate -N removed.5>6>64,65,66 Departures of methanol
3-33
-------
requirements from Equation 3-45 are most likely due to variations in sludge yields among
experimental systems. It has been suggested that column denitrification systems require a
lower M/N ratio than suspended growth systems due to the higher concentration of biomass
maintained in the column systems.67 Higher biomass levels produce longer solids retention
times and reduce organism yields due to increased endogenous metabolism. In turn, this
lower yield would result in less methanol required for synthesis and reduce the "M/N"
ratio. 67
In general, an M/N ratio of 3.0 will enable "complete" denitrification (95 percent removal of
nitrate) and this value may be used for design purposes when methanol is employed as the
carbon source for denitrification.
3.3.3 Alkalinity and pH Relationships
Equations 3-42 and 3-43 (Table 3-9) show that bicarbonate is produced and carbonic acid
concentration is reduced whenever nitrate or nitrite is denitrified to nitrogen gas. The
stoichiometric quantity of alkalinity produced is 3.57 mg alkalinity as CaCC>3 produced per
mg of nitrate or nitrite -N reduced to nitrogen gas.
Since both the alkalinity concentration is increased and the carbonic acid concentration is
reduced, the tendency of denitrification is to at least partially reverse the effects of
nitrification and raise the pH of the biological reaction (Equation 3-9). Denitrification only
partially offsets the alkalinity loss caused by nitrification, since the alkalinity gain per mg of
nitrogen is only one-half the loss caused by nitrification (see Section 3.2.3).
Measured alkalinity production has been reported to be somewhat lower than indicated
theoretically. Experiments with an attached growth process showed that the alkalinity
produced averaged 2.95 mg as CaCC>3 per mg of nitrogen reduced. 65 Similarly, the ratio for
a suspended growth system was 2.89.6 Departures from theory may be due to the fact that
Equations 3-42 and 3-43 (Table 3-9) represent over-simplications of the biological
transformations taking place and do not include all factors affecting alkalinity production.
A value for alkalinity production suitable for engineering calculations would be 3.0 mg
alkalinity as CaCO3 produced per mg nitrogen reduced.
3.3.4 Alternative Electron Donors
Although methanol has found a predominance in U.S. practice as the electron donor of
choice, the significance of the cost of the organic chosen for the process has led to the
consideration of alternative electron donors available, particularly those from waste sources.
Considering alternate commercial sources, methanol seems to continue to be the most
economic choice, because price increases in alternate sources have paralleled those for
methanol.
3-34
-------
A variety of compounds that can substitute for methanol have been experimentally
evaluated, but design data are available only for municipal wastewater organics, volatile
acids, brewery wastes, and molasses. The use of wastewater organics for denitrification is
discussed extensively in Section 5.5.2. Denitrification rates with wastewater oganics are
approximately one-third of those when methanol is employed. Therefore, denitrification
reactors must be proportionately larger. Since using wastewater organics adds ammonia and
organic nitrogen to the wastewater, the sequence of nitrification-denitrification steps must
be modified to ensure that these compounds do not escape from the system. Thus,
wastewater organics are not completely interchangeable with methanol; their attraction,
however, is the possible reduction in operating costs with the elimination of the need for
methanol in the treatment plant.
In studies conducted for the development of the City of Tampa, Florida's treatment plant,
it was shown that brewery wastes could substitute for methanol when used in both
suspended growth and column denitrification systems.68 Bench scale studies exhibited
denitrification rates of 0.25 to 0.22 Ib NC>3 -N rem./lb MLVSS/day with brewery wastes
compared to 0.18 Ib NC>3 -N rem./lb MLVSS/day with methanol at a temperature in the
range of 19 to 24 C. Solids production was found to be greater with brewery wastes than
methanol, but values were not given. Removal efficiencies were similar in a parallel test of
brewery wastes and methanol using columnar denitrification. 68
Volatile acids have also been used as a carbon source for denitrification. In studies of
nitrate reduction in wastewaters generated in the manufacture of nylon intermediates, it was
found that a mixture of C\ to C§ volatile acids was very effective as a carbon source for
denitrification.69 Denitrification rates with this mixture were 0.36 Ib. NO§ -N rem./lb
MLVSS/day at 20 C and 0.10 Ib NO5 -N rem./lb MLVSS/day at 10 C. These rates compare
favorably with those measured for use with methanol (see Section 5.2.1). Volatile acids can
be produced from wastewater organics by anaerobic fermentation or by low temperature wet
oxidation. In either case, the product will contain varying amounts of ammonia nitrogen
which may have to be removed in the process (as described in Section 5.5.2) or removed
prior to use by ammonia stripping.
Molasses was tested at the Central Contra Costa Sanitary District's Advanced Treatment
Test facility as a substitute for methanol. 7® Peak denitrification rates at 16 C in a suspended
growth reaction were found to be only 0.036 Ib NO3 -N rem./lb MLVSS/day. In addition to
having a slower reaction rate with molasses, the sludge tended to bulk to a greater degree
than with methanol, rising from a sludge volume index (SVI) averaging 164 ml/gram to one
having a SVI averaging 257 ml/gram. This caused a decrease in the settling rates of the
sludge when molasses was employed.
Some of the alternatives cause greater sludge production than others. For instance, about
twice as much sludge is produced per mg of nitrogen reduced when saccharose is used than
when methanol is employed. On the other hand, acetone, acetate and ethanol produced
similar quantities of sludge to that produced when methanol is employed.62
3-35
-------
Methanol has certain advantages over wastewater carbon sources. It is free of contaminants,
such as nitrogen, and therefore can be used directly in the process without taking the special
precautions that must be made for use of with a waste carbon source. Second, the product is
of consistent quality while wastewater sources may vary in strength and composition either
daily or seasonally, complicating process control and optimization. Use of wastewater
sources will require regular assaying of the source to check its purity, strength and biological
availability. Methanol also has the advantage of being nationally distributed while suitable
waste carbon sources may not be geographically close to the point of use. Nonetheless, the
significant disadvantage of methanol is its cost and this alone mandates the necessity of
economic comparisons of alternate carbon sources.
3.3.5 Kinetics of Denitrification
Just as in the case of nitrification (Section 3.2.5), environmental factors have a significant
effect on the kinetic rates of denitrifier growth and nitrate removal. Factors considered in
subsequent sections are temperature, pH, carbon concentration and nitrate concentration. A
combined kinetic expression incorporating all these factors is presented.
3.3.5.1 Effect of Nitrate on Kinetics
The absence of significant quantities of nitrite in denitrification systems^.?! has led to the
description of the kinetics of denitrification as a one-step process from nitrate to nitrogen
gas. The Monod expression is employed to describe the influence of nitrate on growth rate:
D D Kp D
where: /UD = growth rate, day"
A _1
Mp = maximum denitrifier growth rate, day
D = concentration of nitrate nitrogen, mg/1, and
Kp = half saturation constant, mg/1 NO- - N.
3.3.5.2 Relationship of Growth Rate to Removal Rate
Denitrification rates can be related to the oganism growth rates by the following
relationship:
qD = "D/YD (3-48)
where: q^ = nitrate removal rate, Ib NOl - N rem./lb VSS/day, and
Y~ = denitrifier gross yield, Ib VSS grown/lb NO~ - N removed.
3-36
-------
Similarly, peak denitrification rates are related to maximum denitrifier growth rates as
follows:
Sn = *VYD (3-49)
3.3.5.3 Solids Retention Time .
Consideration of solids production and solids retention time is an important design
consideration. A mass balance of the biomass in a completely mixed reactor yields the
relationship. 10,11
= YDqD ' Kd <
where: 0£ = solids retention time, days, and
K = decay coefficient, day .
3.3.5.4 Kinetic Constants for Denitrification
The value of the half saturation constant, Kj>, is very low. Investigators at the University of
California at Davis, found Kj) for suspended growth systems to be 0.08 mg/1 NO§ -N
without solids recycle and 0.16 mg/1 NOJ- N with solids recycle at 20 C.71>48 por attached
growth systems the value of Kj) was found to be 0.06 mg/1 NO;} -N at 25 C.^2,73 n can j,e
seen from examination of Equation 3-47 using these values of Kj) that nitrate has almost no
effect on denitrification above 1-2 mg/1 nitrate -N and approaches a zero order rate. The
observations of several investigators suport these low values of Kr_>, as they have reported
zero order rates above 1-2 mg/1 nitrate -N.60,74,75,76
Yield and decay coefficients from data of a number of investigations are shown in Table
3-10. In most cases only net yields are reported or can be calculated from the data reported.
The relationship between the gross yield in Equation 3-50 and the net yield is:
Y a K Y
YDqD Kd YD
D
where: YD = denitrifier net yield, Ib VSS/lb NO3 - N rem.
The data of Stensel, et al. 74 for KJ has been used in some cases to derive calculated Y
values for those cases where none was reported (Table 3-10). Data from another study4°
3-37
-------
allow calculation of both YD and Kj at 20 C. The value of Kj of 0.04 day"1 is consistent
with Stensel's findings at 20 C. Values suitable for use in engineering calculations are YD =
0.6 to 1.20 and Kd = 0.04 day1.
It is notable that when an aerobic stabilization step was incorporated into the process after
anoxic denitrification, net yields reduced by almost an order of magnitude.78,80'T'}1js effect
was attributed to enhanced endogenous metabolism where oxygen is provided as an
electron-acceptor.80 Table 3-10 shows a calculated value for K^ under these conditions of
0.19 days"1, almost five times the rate when nitrate serves as the electron acceptor for
endogenous metabolism. This concept is supported by the results of other investigators
whose data show that the endogenous respiration rates when expressed on an equivalent
basis are significantly greater when oxygen serves as the electron acceptor than when nitrate
1'** 2
Both organism growth rate and nitrate removal rate are significantly affected by
temperature. Only one investigator has reported growth rates, 74 all others have reported
removal rates. To show the effect of temperature on growth and denitrification rates, the
available data have been summarized in Figure 3-10 on a basis that is normalized with
respect to the rate at 20 C. It can be seen that denitrification proceeds at a reduced rate as
TABLE 3- 10
VALUES OF DENITRIFICATION YIELD AND DECAY COEFFICIENTS
FOR VARIOUS INVESTIGATIONS USING METHANOL
Process description
Suspended growth, no solids
recycle, continuous
Suspended growth, batch
Suspended growth, solids
recycle, continuous
Suspended growth, solids
recycle, continuous.
aerated stabilization
Ref.
74
74
74
71
77
7Sa
62
5
48
78
79
78
q
D'
, -1
days
Variable
Variable
Variable
0.12 to 0.32
0.16 to 0.9
0.24 to 3.8
Variable
b
0.131 to 0.347
0.25
e
0.30
Y
D'
Ib VSS
Ib NC>3-N rem.
Variable
Variable
Variable
0.55 to 1.4
0.57 to 0.73
0.45 to 1.43
0.53
0.58
0.542 to 0.703
0.49
0 . 7 to 1 . 4
0.061
Y ,
D'
Ib VSS
Ib NOjj-Nrem
0.57
0.63
0.67
-
-
-
-
0.77C
0.84
0.65C
0.83 to 1.67
0.65d
K,,
d
_1
days
0.05
0.04
0.02
-
-
-
-
0.04d
0.04
0.04°
0.04d
0.19C
Temp, C
10
20
30
20
20
5 to 27
20
10 to 20
20
16 to 18
19 to 20
3 Substrate was sodium citrate
q not given, but Qc = 8.0
c Calculated
Assumed
6 6 = 3 to 6 days
3-38
-------
o
o
CM
t-
«*
UJ
ft
a:
h-
o
or
tt
o
UJ
s
UJ
Q
U.
O
iu
FIGURE 3-10
EFFECT OF TEMPERATURE ON DENITRIFICATION RATE
/SO
Key
Ref.
qD ^ID Electron
days-' days*1 donor
at 20 C
System
160
X X X X X
75,83 -
60
5
• 84
1,07
0.24
/4O
a a a a a a
85
74
86
86
87
0.33
1.7
1.04
Sodium Citrate
Endogenous
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
Methanol
I2O
110
80
60
40
S.G. = Suspended Growth
A.G. = Attached Growth
20
_L
10 15 20
TEMPERATURE, C
25
30
3-39
-------
low as 5 C. Above 20 C, four out of seven sets of data indicate that the denitrification rates
find plateau values at some temperature and do not keep climbing. The parallel systems
study of Murphy et a/. 8 6 js interesting in that it shows attached growth systems to be less
affected by cold temperatures than suspended growth systems. Differences between
attached growth systems and suspended growth systems may reflect differences in the
method of measurement rather than, differences in organism reaction rate. Attached growth
system removal rates were expressed on a unit surface basis while suspended growth systems
were expressed per unit of biomass (MLVSS) in Murphy's study. 86 it is probable that
surface slimes in attached growth systems expand in cold weather and compensate for
reduced reaction rates. If the biomass level could be measured the rate per unit of biomass
may very well be similar. For instance, in one study parallel tests of suspended and attached
growth systems were at 30 C. Biomass measurements were made in both systems and peak
denitrification rates were found to be comparable, 0.38 Ib NO^-N rem./lb MLVSS/day for
the suspended growth system and 0.45 Ib NC>3~ -N rem./lb MLVSS/day for the attached
growth system. ^8
Further information on the effect of temperature on denitrification rates is presented in
Chapter 5.
3.3.5.5 Effect of Carbon Concentration on Kinetics
The effect of carbon concentration on the rate of denitrification has been modeled in terms
of a Monod type of expression. When methanol serves as the carbon source, the expression
is: 74,89
" =
(3'52)
D ^D KM + M
where: M = methanol concentration, mg/1
KM = half saturation constant for methanol, mg/1
of methanol.
The earliest investigators used a nonspecific test for methanol, COD. 74 AS a result the initial
evaluation of KM was somewhat obscured. A later more definitive investigation evaluating
KM used a specific test for methanol.^9 A chemostat system operating at a solids retention
time of five days and a temperature of 20 C was operated in a manner whereby reaction
rates were limited by methanol and not nitrate. The value of KM was found to be very low,
0.1 mg/1 as methanol. The practical implication of this finding is that to achieve 90 percent
of the maximum denitrification rate in a suspended growth reactor, only about 1 mg/1 of
methanol need be in the effluent. In other words, great excesses of methanol above
stoichiometric requirements need not be in the effluent from a suspended growth
denitrification process to achieve nearly the maximum denitrification rates.
3-40
-------
3.3.5.6 Effect of pH on Kinetics
Representative observations of the effect of pH on denitrification rates are shown on Figure
3-11. While there are some anomalies, it is apparent that denitrification rates are depressed
below pH 6.0 and above pH 8.0. There is some disagreement about the pH of the optima,
but the data show the highest rates of denitrification are at least within the range of pH 7.0
to 7.5.
FIGURE 3-11
EFFECT OF pH ON DENITRIFICATION RATE
100
Uj
l-
2 80
O
o
£
ct
Ul
Q
1
u.
o
UJ
O
Ct
UJ
Q.
6O
40
20
6.0
Key
Reference
5
83
60,90
81
I I
b
7.0
pH
8.0
9.0
-------
3.3.5.7 Combined Kinetic Expression
The same approach as employed for nitrification can be used for denitrification to establish
the effects of environmental conditions on the rates of denitrifier growth (and nitrate
removal):
« •
A
where: MT) = peak rate of denitrifier growth at given temperature, T, and
pH, and
MT^ = actual rate of denitrifier growth affected by nitrate, methanol,
T, and pH.
Relationships for temperature, pH, nitrate and methanol established in Sections 3.3.5.3,
3.3.5.4, 3.3.5.5, and 3.3.5.6 can be employed when using this equation to predict growth
rates or removal rates. Ordinarily, the term for methanol can be neglected (Section 3.3.5.5).
Removal rates can be related to growth rates through Equation .3-48.
The safety factor concept presented in Section 3.2.6 can be applied to denitrification as well
as to nitrification, as the concept has general validity for biological systems. Restating the
concept for denitrification:
SF = —£- (3-29)
0m
c
In the case of denitrification, the safety factor can be related to nitrate removal rates
through Equation 3-50 and the following similar equation for the minimum solids retention
time:
—L = Yqn - k, (3-54)
Qm D d
c
The use of these equations for design of suspended growth systems is given in Section 5.2.2
in terms of illustrative examples.
The above equations cannot be directly applied to attached growth denitrification because
the reactions take place in a more complex environment than is present in suspended growth
systems. Rates of nitrate removal in the bacterial films developed in denitrification systems
may be affected by the mass transfer of nitrate or methanol through the bacterial film. A
biofilm model has been developed^2,53 faat may be use(j to describe denitrification in
-------
bacterial slimes,, but its use has not yet been extended to the point where it can be used in
system design. However, the model indicates that removal rates are most usefully expressed
on a unit surface area basis and this is the procedure adopted in Section 5.3.1 to describe
denitrification in the various attached growth systems.
The biofilm model usefully predicts certain properties of attached growth denitrification
that are significant in design. The model shows that the nitrate removal rate in attached
growth sytems should not be drastically affected by adverse environmental conditions
compared to effects in suspended growth systems.52 In Section 3.3.5.4, for instance, it was
shown that attached growth systems are less affected by cold temperatures than suspended
growth systems. Another interesting prediction of the biofilm model is that methanol will
normally be film transfer limiting rather than nitrate, unless the methanol is supplied in
concentrations five times as large as the nitrate concentration^ (an impractical situation).
3.3.6 Effect of DO on Denitrification Inhibition
The role of dissolved oxygen in denitrification is generally to suppress denitrification. This
has been explained on the basis that the rate of dissimilatory nitrate reduction is
considerably slower than the rate of aerobic respiration.71 While it has been observed that
denitrification can occur in the presence of low levels of DO,°>"° the mechanism of
denitrification is attributed to an oxygen gradient in the system whereby some cells are at
zero dissolved oxygen and thus able to reduce nitrate. 1 >6>91
3.4 References
1. Painter, H.A., A Review of Literature on Inorganic Nitrogen Metabolism in
Microorganisms. Water Research, 4, No. 6, pp 393-450 (1970).
2. Haug, R. T., and P. L. McCarty, Nitrification with the Submerged Filter. Report
prepared by the Department of Civil Engineering, Stanford University for the
Environmental Protection Agency, Research Grant No. 17010 EPM, August, 1971.
3. Notes on Water Pollution No. 52, Water Pollution Research Laboratory, Stevenage,
England, 1971.
4. Gujer, W. and D. Jenkins, The Contact Stabilization Process-Oxygen and Nitrogen Mass
Balances. University of California, Sanitary Engineering Research Lab, SERL Report
74-2, February, 1974.
5. Mulbarger, M. C., The Three Sludge System for Nitrogen and Phosphorus Removal.
Presented at the 44th Annual Conference of the Water Pollution Control Federation,
San Francisco, California, October, 1971.
3-43
-------
6. Horstkotte, G. A., Niles, D.G., Parker, D.S., and D.H. Caldwell, Full-Scale Testing of A
Water Reclamation System. JWPCF, 46, No. 1, pp 181-197 (1974).
7. Newton, D., and T.E. Wilson, Oxygen Nitrification Process at Tampa. In Applications
of Commercial Oxygen to Water and Waste-water Systems, Ed. by R.E. Speece and J.F.
Malena, Jr., Austin, Texas: The Center for Research in Water Resources, 1973.
8. Gasser, J.A., Chen, C.L., and R.P. Miele, Fixed-Film Nitrification of Secondary
Effluent. Presented at the EED-ASCE Specialty Conference, Penn. State University,
Pa., July, 1974.
9. Osborn, D.E., Operating Experiences with Double Filtration in Johannesburg. J. Inst.
of Sew. Purif., Part 3, pp 272-281 (1965).
10. Pearson, E.A., Kinetics of Biological Treatment. In Advances in Water Quality
Improvement. Ed. by E.F. Gloyna and W.W. Eckenfielder, Austin, Texas: University of
Texas Press, 1968.
11. Lawrence, A.W., and P.L. McCarty, Unified Basis for Biological Treatment Design and
Operation. JSED, Proc. ASCE, 96, No. SA3, pp 757-778 (1970).
12. Jenkins, D., and W.E. Garrison, Control of Activated Sludge by Mean Cell Residence
Time. JWPCF, 40, No. 11, pp 1904-1919 (1968).
13. Monod, J., Researches sur la croissance des cultures. Hermann and Cie, Paris, 1942.
14. Stankewich, M. J., Biological Nitrification with the High Purity Oxygenation Process.
Presented at 27th Annual Purdue Industrial Waste Conference, Lafayette, Indiana,
May, 1972.
15. Poduska, R.A. and J.F. Andrews, Dynamics of Nitrification in the Activated Sludge
Process. Presented at the 29th Industrial Waste Conference, Purdue University,
Lafayette, Indiana, May 7-9, 1974.
16. Christensen, D.R. and P.L. McCarty, Bio-treat: A Multi-Process Biological Treatment
Model. Presented at the 47th Annual Conference of the Water Pollution Control
Federation, Denver, Colorado, October, 1974.
17. Knowles, G., Downing, A.L., and M.J. Barrett, Determination of Kinetic Constants for
Nitrifying Bacteria in Mixed Culture, with the Aid of An Electronic Computer. J. Gen.
Microbiology, 38, p 263 (1965).
18. Downing, A.L., and A.P. Hopwood, Some Observations on the Kinetics of Nitrifying
Activated Sludge Plants. Schweiz. Zeitsch f Hydrol., 26, No. 2, p 271 (1964).
3-44
-------
19. Wuhrmann, K., Research Developments'in Regard to Concept and Base Values of the
Activated Sludge System. Ed. by E.F. Gloyna and W.W. Eckenfielder, Austin, Texas:
University of Texas Press, 1968.
20. Melamed, A., Saliternik, C., and A.M. Wachs, BOD Removal and Nitrification of
Anaerobic Effluent by Activated Sludge. Presented at the 5th IAWPR Conference, San
Francisco, California, July-August, 1970.
21. Balakrishnan, S., and W.W. Eckenfielder, Nitrogen Relationships in Biological Waste
Treatment Processes — I, Nitrification in the Activated Sludge Process. Water Research,
3, pp 73-81 (1969).
22. Loehr, R.C., Prakasam, T.B.S., Srinath, E.G., and Y.D. Joo, Development and
Demonstration of Nutrient Removal from Animal Wastes. Report prepared for the
Environmental Protection Agency, EPA-R2-73-095, January, 1973.
23. Stratton, F.E., and P.L. McCarty, Prediction of Nitrification Effects on the Dissolved
Oxygen Balance of Streams. Environmental Science and Technology, 1, No. 5, pp
405-410(1967).
24. Lawrence, A.M., and C.G. Brown, Bio kinetic Approach to Optimal Design and Control
of Nitrifying Activated Sludge Systems. Presented at the Annual Meeting of the New
York Water Pollution Control Association, New York City, January 23, 1973.
25. Hoffman, T., and H. Lees, The Biochemistry of Nitrifying Organisms 4, The
Respiration and Intermediate Metabolism of Nitrosomonas. Biochemistry Journal, 54,
575-583(1953).
26. Ulken, A., Die Herkunft des Nitrits in der Elbe. Arch. Hydrobiol., 59, pp 486-501,
1963.
27. Loveless. J.E., and J.E. Painter, The Influent of Metal Ion Concentration and pH Value
on the Growth of a Nitrosomonas Strain Isolated From Activated Sludge. J. Gen.
Microbiology, 52, pp 1-14(1968).
28. Williamson, K.J. and P.L. McCarty, Rapid Measurement of Monod Half-Velocity
Coefficients for Bacterial Kinetics. Unpublished paper, Stanford University, May,
1974.
29. Lees, H., and J.R. Simpson, The Biochemistry of the Nitrifying Organisms. 5, Nitrate
Oxidation by Nitrobacter. Biochemistry Journal, 65, pp 297-305 (1957).
30. Gould, G.W., and H. Lees, The Isolation and Culture of the Nitrifying Organisms. Part
I. Nitrobacter. Can. J. Microbiology, 6, pp 299-307 (1960).
3-45
-------
31. Landelot, H., and L. Van Tichilen, Kinetics of Nitrite Oxidation by Nitrobacter
Winogradskyi, J. Bact., 79, pp 39-42 (1960).
32. Buswell, A.H., Shiota, T., Lawrence, N., and I. Van Meter, Laboratory Studies on the
Kinetics of the Growth of Nitrosomonas with Relation to the Nitrification Phase of the
BOD Test. Applied Microbiology, 2, pp 21-25 (1954).
33. Balakrishnan, S., and W.W. Eckenfielder, Nitrogen Relationships in Biological Waste
Treatment Processes — II, Nitrification in Trickling Filters, Water Research, 3, pp
167-174(1969).
34. Huang, C.S., and N.E. Hopson, Temperature and pH Effect on the Biological
Nitrification Process. Presented at the Annual Winter Meeting, New York Water
Pollution Control Association, New York City, January, 1974.
35. Water Pollution Research, 1964. London: HMSO, 1965.
36. Nagel, C.A. and Haworth, J.G., Operational Factors Affecting Nitrification in the
Activated Sludge Process. Presented at the 42nd Annual Conference of the Water
Polluation Control Federation, Dallas, Texas, October (1969), (available as a reprint
from the County Sanitation Districts of Los Angeles County).
37. Wuhrman, K., Effects of Oxygen Tension on Biochemical Reactions in Sewage
Purification. In Advances in Biological Treatment, Proc. 3rd Conf. Biological Waste
Treatment, Ed. by Eckenfielder, W.W., and J. McCabe. (New York, 1963), Pergamon
Press.
38. Schwer, A.D., Letter communication to D.S. Parker — Metropolitan Sewer District of
Greater Cincinnati, March 9, 1971.
39. Murphy, K.L., Personal communication to D.S. Parker, McMasters University, Ontario,
July, 1974.
40. Engel, M.S., and M. Alexander, Growth and Metabolism of Nitrosomonas europaea.
Journal Bacteriology, 76, pp 217-222 (1958).
41. Meyerhof, O., Untersuchungen uber den Atmungsvorgany Nitrifizierenden Bakterien.
Pflugers Archges Physiol., 166, pp 240-280, 1917.
42. Sawyer, C.N., Wild, H.E., Jr., and T.C. McMahon, Nitrification and Denitrification
Facilities, Wastewater Treatment. Prepared for the EPA Technology Transfer Program,
August, 1973.
3-46
-------
43. Downing, A.L., and G. Knowles, Population Dynamics in Biological Treatment Plants.
Presented at the 3rd Conference of the IAWPR, Munich, 1966.
44. Huang, C.S., Kinetics and Process Factors of Nitrification on a Biological Film Reactor.
Thesis submitted in partial satisfaction of the requirements for the degree of Doctor of
Philosophy, University of New York at Buffalo, 1973.
45. Heidman, J.A., An Experimental Evaluation of Oxygen and Air Activated Sludge
Nitrification Systems With and Without pH Control. EPA report fqr Contract No.
68-03-0349, 1975.
46. Chen, C.W., Concepts and Utilities of Ecological Model. JSED, Proc. ASCE, 96, No.
SA5,pp 1085-1097(1970).
47. Middlebrooks, E.J., and D.B. Porcella, Rational Multivariate Algal Growth Kinetics.
JSED, Proc. ASCE, 97, SA1, pp 135-140 (1971).
48. Engberg, D.J., and E.D. Schroeder, Kinetics and Stoichiometry of Bacterial Denitrifi-
cation as a Function of Cell Residence Time. University of California at Davis,
unpublished paper, 1974.
49. Stamberg, J.B., Hais, A.B., Bishop, D.F., and J. Heidman, Nitrification in Oxygen
Activated Sludge. Unpublished paper, Environmental Protection Agency, 1974.
50. Sawyer, C.N., Activated Sludge Oxidations, V. The Influence of Nutrition in
Determining Activated Sludge Characteristics. Sewage Works Journal, 12, No. 1, pp
3-17(1940).
51. Kiff, R.J., The Ecology of Nitrification IDenitrification in Activated Sludge. Water
Pollution Control, 71, p 475 (1972).
52. Williamson, K.L. and P.L. MeCarty, A Model of Substrate Utilization by Bacterial
Films. Presented at the 46th Annual Conference of the Water Pollution Control
Federation, Cincinnati, Ohio, October, 1973.
53. Williamson, K.L. and P.L. MeCarty, Verification Studies of the Biofilm Model for
Bacterial Substrate Utilization. Presented at the 46th Annual Conference of the Water
Pollution Control Federation, Cincinnati, Ohio, October, 1973.
54. Interaction of Heavy Metals and Biological Sewage Treatment Processes. Division of
Water Supply and Pollution Control, USPHS, May, 1965.
55. Environmental Quality Analysts, Inc., Letter Report to Valley Community Services
District, March, 1974.
341
-------
56. Sawyer, C.N., Letter communication to D.S. Parker, January 24, 1975.
57. Davis, G., Personal communication with D.S. Parker, B.F. Goodrich, Co., August,
1974.
58. Tomlinson, T.G., Boon, A.G., and C.N.A. Trotman, Inhibition of Nitrification in the
Activated Sludge Process of Sewage Disposal. J. Appl. Bact., 29, pp 266-291 (1966).
59. Anthonisen, A.C., Loehr, R.C., Prakasam, T.B.S., and E.G. Srinath, Inhibition of
Nitrification by Un-ionized Ammonia and Un-ionized Nitrous Acid. Presented at the
47th Annual Conference, Water Pollution Control Federation, Denver, Colorado,
October, 1974.
60. Christensen, M.H., and P. Harremoes, Biological Denitrification in Wastewater
Treatment. Report 2-72, Department of Sanitary Engineering, Technical University of
Denmark, 1972.
61. Delwiche, C.C., The Nitrogen Cycle. Scientific American, 223, No. 3, pp 137-146
(1970).
62. McCarty, P.L., Beck, L., and P. St. Amant, Biological Denitrification of Waste-waters by
Addition of Organic Materials. In Proc. of the 24th Industrial Waste Conference, May
6, 7, and 8, 1969. Lafayette, Indiana: Purdue University, 1969.
63. McCarty, P.L., Stoichiometry of Biological Reactions. Presented at the Summer
Institute in Water Pollution Control, Biological Waste Treatment, Manhattan College,
•New York, N.Y., May, 1974.
64. Dholakia, S.G., Stone, J.H., and H.P. Burchfield, Methanol Requirement and
Temperature Effects in Waste-water Denitrification. U.S. Environmental Protection
Agency, Washington, D.C., WPCRS 17010 DHT 09/70, August, 1970.
65. Jeris, John S.5 and R.W. Owens, Pilot Scale High Rate Biological Denitrification at
Nassau County, N. Y. Presented at the Winter Meeting of the New York Water Pollution
Control Association, January, 1974.
66. Smith, J.M., Masse, A.N., Feige, W.A., and L.J. Kamphake, Nitrogen Removal from
Municipal Wastewater by Columnar Denitrification. Environmental Science and
Technology, 6, p 260 (1972).
67. Michael, R.P., and W.J. Jewell, Optimization of the Denitrification Process. Journal of
the Environmental Engineering Division, Proc, ASCE, in press.
3-48
-------
68. Wilson, T.E., and D. Newton, Brewery Wastes as a Carbon Source for Denitrification at
Tampa, Florida. Presented at the 28th Annual Purdue Industrial Waste Conference,
May, 1973.
69. Climenhage, D.C., Biological Denitrification of Nylon Intermediates Waste Water.
Presented at the 2nd Canadian Chemical Engineering Conference, September, 1972.
70. Central Contra Costa Sanitary District, Operating Report, Advanced Treatment Test
Facility. January, 1974.
71. Moore, S.F., and E.D. Schroeder, The Effect of Nitrate Feed Rate on Denitrification.
Water Research, 5, pp 445-452 (1971).
72. Requa, D.A., Kinetics of Packed Bed Denitrification. Thesis submitted in partial
satisfaction of the requirements for the degree of Master of Science in Engineering,
University of California at Davis, 1970.
73. Requa, D.A., and E.D. Schroeder, Kinetics of Packed Bed Denitrification. JWPCF, 45,
No. 8,pp 1696-1707(1973).
74. Stensel, H.D., Loehr, R.C., and A.W. Lawrence, Biological Kinetics of the Suspended
Growth Denitrification. JWPCF, 45, No. 2, pp 249-261 (1973).
75. Murphy, R.L., and R.N. Dawson, The Temperature Dependency of Biological
Denitrification. Water Research, 6, pp 71-83 (1972).
76. Ericsson, et al., Nutrient Reduction at Sewage Treatment Plants. KTH Publication
67:5, Stockholm, 1967.
77. Moore, S., and E.D. Schroeder, An Investigation of the Effects of Residence Time on
Anaerobic Bacterial Denitrification. Water Research, 4, pp 685-694 (1970).
78. Parker, D.S., Zadick, F.J., and K.E. Train, Sludge Processing for Combined
Physical-Chemical-Biological Sludges. Prepared for the EPA, Report No. R2-73-250,
July, 1973.
79. Sutton, P.M., Murphy, K.L., and R.N. Dawson, Low Temperature Biological
Denitrification of Wastewater. JWPCF, 47, No. 1, pp 122-134 (1975).
80. Parker, D.S., Aberley, R.C., and D.H. Caldwell, Development and Implementation of
Biological Denitrification for Two Large Plants. Presented at the Conference on
Nitrogen as a Water Pollutant, sponsored by the IAWPR, Copenhagen, Denmark,
August, 1975.
3-49
-------
81. Clayfield, G.W., Respiration and Denitrification Studies on Laboratory and Works
Activated Sludges. Water Pollution Control, London, 73, No. 1, pp 51-76 (1974).
82. Balakrishnan, S., and W.W. Eckenfielder, Nitrogen Relationships in Biological Waste
Treatment Processes — HI, Denitrification in the Modified Activated Sludge Process.
Water Research, 3, pp 177-188 (1969).
83. Dawson, R.N., and K.L. Murphy, Factors Affecting Biological Denitrification in
Waste-water. In Advances in Water Pollution Research, S.H. Jenkins, Ed., Oxford,
England: Pergamon Press, 1973.
84. Parker, D.S., Case Histories of Nitrification and Denitrification Facilities. Prepared for
the EPA Technology Transfer Program, May, 1974.
85. 'Bishop, D.F., Personal communication to D.S. Parker. Environmental Protection
Agency, Washington, B.C., April, 1974.
86. Murphy, K.L., and P.M. Sutton, Pilot Scale Studies on Biological Denitrification.
Presented at the 7th International Conference on Water Pollution Research, Paris,
September, 1974.
87. Ecotrol, Inc., Biological Denitrification Using Fluidized Bed Technology. August,
1974.
88. Jewell, W.J., and R.J. Cummings, Denitrification of Concentrated Wastewaters.
Presented at the Water Pollution Control Federation, Cleveland, October, 1973.
89. Kaufman, G., and E. Schroeder, Personal communication to D.S. Parker. University of
California at Davis, July, 1974.
90. Renner, Production of Nitric Oxide and Nitrous Oxide During Denitrification by
Comybacterium nephridii. J. Bacterial., 101, pp 821-826 (1970).
91. Skerman, V.B.D., and I.C. MacRae, The Influence of Oxygen Availability on the
Degree of Nitrate Reduction by Pseudomonas Denitrificans. Can. J. Microbiology, 3, pp
505-530(1957).
3-50
-------
CHAPTER 4
BIOLOGICAL NITRIFICATION
4.1 Introduction
The application of biological nitrification in municipal wastewater treatment is particularly
applicable to those cases where an ammonia removal requirement exists, without need for
complete nitrogen removal. Biological nitrification is also the first step of the biological
nitrification-denitrification approach to nitrogen removal.
4.2 Classification of Nitrification Processes
The first means of categorizing nitrification systems concerns the degree of separation of the
carbon removal and nitrification processes. The first nitrification processes developed
combined the functions of carbon oxidation and nitrification in one process. The extended
aeration modification of the activated sludge process is an example of a combined carbon
oxidation-nitrification process. Combined carbon oxidation-nitrification processes generally
have low populations of nitrifiers due to a high ratio of BOD5 to Total Kjeldahl Nitrogen
(TKN) in the influent (see Section 3.2.7 for a discussion of this effect). The bulk of the
oxygen requirement for this process comes from the oxidation of organics.
Separate stage nitrification is the other category of nitrification processes. In this process,
there is a lower BOD5 load relative to the influent ammonia load. As a result, a higher
proportion of nitrifiers is obtained, resulting in higher rates of nitrification. The bulk of the
oxygen requirements in the nitrification stage derive from ammonia oxidation. To obtain
separate stage nitrification, pretreatment is required to lower the organic load or
BOD5/TKN ratio in the influent to the nitrification stage.
Both the combined carbon oxidation-nitrification and separate stage nitrification processes
can be further subdivided into suspended growth and attached growth processes. Suspended
growth processes are those which suspend the biological solids in a mixed liquor by some
mixing mechanism. A subsequent clarification stage is required for returning these solids to
the nitrification stage. Attached growth processes, on the other hand, retain the bulk of the
biomass on the media and therefore do not require a solids separation step for returning the
solids to the nitrification reactor. In separate stage processes operated in the attached
growth mode, a clarification step may not be required since solids synthesis is low and the
sloughed solids are often low in concentration.
There are many different configurations of suspended and attached growth reactors; these
are described in subsequent sections of this manual. For suspended growth reactors refer to
Sections 4.3 and 4.6; for attached growth reactors refer to Sections 4.4 and 4.7.
4-1
-------
Using the classification described above, representative nitrification processes have been
classified in Table 4-1 according to the degree of separation of the carbon removal and
nitrification processes. Many of these facilities are described in further detail in either this
chapter or Chapter 9. Each facility listed has been categorized according to the BOD5/TKN
ratio of the wastewater influent to the nitrification process. Interestingly, the processes
listed can all be categorized according to whether the BOD5/TKN ratio is less than 3.0 or
greater than 5.0. If the BOD5/TKN ratio is less than 3.0, the system can be classified as a
separate stage nitrification process. If the BOD5/TKN is greater than 5.0, the process can be
classed as a combined carbon oxidation-nitrification process. Also shown in Table 4-1 is the
distribution of total oxygen demand in the process between carbonaceous sources (8005)
and nitrogenous sources (NOD). It can be seen that in separate stage processes, the
proportion of nitrogenous oxygen demand is at least 60 percent of the total. In combined
carbon oxidation-nitrification processes, the proportion of nitrogenous oxygen demand is
lower than 50 percent.
There exists a range of BOD5/TKN ratios between 3.0 and 5.0 where no practical examples
currently exist. Facilities in this range could be considered to provide an intermediate degree
of separation of carbon removal and nitrification.
4.3 Combined Carbon Oxidation-Nitrification in Suspended Growth Reactors
The conventional activated sludge process has seen relatively wide application in Great
Britain for use in obtaining effluents low in ammonia nitrogen. Much of the U.S. practice
derives from that experience. Recent U.S. design practice has provided amplifying
information.
General design concepts for the activated sludge process are covered in the Technology
Transfer publication, Process Design Manual for Upgrading Existing Wastewater Treatment
Plants. 25 The following sections provide an extension of these concepts to combined carbon
oxidation-nitrification applications.
4.3.1 Activated Sludge Modifications
Not all of the various modifications of the activated sludge process are appropriate for
nitrification applications, although some see use where only partial ammonia removal is
required. Figure 4-1 gives pictorial representations of four common modifications.
4.3.1.1 Complete Mix Plants
Many plants are designed to operate on the complete mix principle. Shown on Figure 4-1 is
an example of the feed and withdrawal arrangement for a complete mix plant. The complete
mix design provides uniformity of load to all points within the aeration tank, easing the
problems of oxygen transfer presented in the head end of the conventional plants. Complete
mix plants can be designed for complete nitrification at loading rates comparable to
4-2
-------
TABLE 4-1
CLASSIFICATION OF NITRIFICATION FACILITIES
Type and Location
Suspended Growth
Manassas, Va.
Hyperion, Los Angeles, Ca.
Central Contra Costa Sanitary
District, Ca.
Llvermore, Ca.
Flint, Michigan
Valley Community Services District, Ca.
Blue Plains, D.C.
Whittier Narrows, LACSD, Ca.
Jackson, Michigan
Tampa, Florida
South Bend, Indiana
New Market, Ontario, Canada
Cincinnati, Ohio
Fitchburg, Mass.
Marlboro, Mass.
Amherst, N. Y.
Denver, Col.
Attached Growth
Stockton, Ca.
Midland, Mich.
Union City, Ca.
Allentown, Pa.
Lima, Ohio
Scale,
mgda
0.2
46
1.0, design 30
pilot
3.3
34
3.8
pilot, design
309
12
13.5
pilot
design 60
pilot
2.4
pilot
pilot
pilot f
pilot
pilot
pilot
pilot
design 58
pilot
pilot
40
pilot
BOD5/TKN
Ratio
1.2
7.3b
2.4
1.0
2.8
5.5
10. 8b
1.3 to 3.0
6.6
9
3.0
1.8
2.6
7.2
1.0°
i.od
3.6d
* 0.8 to 2.0
2.7
5.3
1.1
1.7
1.9
0.79
Oxygen Demand
Distribution in
percentage
BOD5
20
61
34
18
38
65
70
22 to 39
61
66
40
28
36
61
18°
18
40
22
37
54
19
27
30
15
NOD
80
39b
66
82
62
35b
30b
61 to 78
39
34
60
72
64
39
82°
82
60
78
63
46
81
73
70
85
Bef.
1
2
3
4
5
6
3
7, 8
9
10
11
12
13
14
15
16
17
18
19, 20
21
22
23, 24
25
26, 27
Classification -
Degree of separation
Combined
oxidation -
nitrification
X
X
X
X
X
X
X
Separate
stage
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Pre treatment
Activated sludge
Primary treatment
Lime primary treatment
Activated sludge
Boughing filter
Primary treatment
Primary treatment
Activated sludge
Primary treatment
Primary treatment
Activated sludge
Activated sludge
Lime primary treatment
Primary treatment
Activated sludge
Activated sludge
Trickling filter
Activated sludge
Activated sludge
Primary treatment
Trickling filter
Activated sludge
Trickling filter
Activated sludge
a 1 mgd = 0.044 mVsec
Calculated from effluent
Approximate, calculated from COD data
BOD/NH^"- N ratio; BOD/TKN would be about 3. 0
-------
conventional plants. As will be shown in Section 4.3.3.2, complete mix plants may have
slightly higher effluent ammonia contents than conventional plants due to increased short
circuiting of the influent to the effluent. Design procedures for nitrification with complete
mix plants are presented in Section 4.3.3.
4.3.1.2 Extended Aeration Plants
Extended aeration plants are similar to complete mix plants excepting that hydraulic
retention times range from 24 to 48 hr instead of the 2 to 8 hr used in complete mix plants.
Extended aeration plants are operated to maximize endogenous respiration, consequently
solid retention times of 25 to 35 days are not uncommon. Because of their long aeration
periods, they suffer from unusual heat losses and low temperatures. Extended aeration
plants, because of their low net growth rate, can be expected to nitrify except at the coldest
of temperatures ( < 10 C). Unless the sludge inventory is kept under control via intentional
sludge wasting, solids are periodically lost in the effluent and nitrification efficiency wanes.
Section 4.3.4 includes a discussion of design procedures for extended aeration plants.
4.3.1.3 Conventional or Plug Flow Plants
Conventional plants consist of a series of rectangular tanks or passes with the total tank
length to width ratio of 5 to 50.^5 The hydraulics of the system have been loosely termed a
plug flow configuration, so called 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 from the system. The degree to which the process
actually approaches plug flow is dependent on the amount of longitudinal mixing in the
process. Conventional plants can be designed to dependably nitrify using the design
approach presented in Section 4.3.5.
4.3.1.4 Contact Stabilization Plants
The contact stabilization modification of the activated sludge process derives from the
alteration of the feed pattern to the process. Instead of mixing the influent wastewater with
the return sludge, the return activated sludge is separately aerated in a sludge reaeration tank
prior to mixing with the influent wastewater. Backmixing between the contact tank and the
sludge reaeration tank is prevented by providing overflow weirs or pumps between the
tanks. BOD5 removal can take place in the contact tank which has a relatively short
detention time, 0.5 to 1 hr based on average dry weather flow (ADWF). BOD5 removals can
be fairly high because the bulk of the organics in domestic wastewater are particulate or
colloidal and can be adsorbed to the biological solids for later oxidation in the sludge
reaeration (or stabilization) tank.
The process is not well suited for complete nitrification, even though relatively high solids
retention times can be maintained in the process because of the inventory of solids in the
sludge reaeration tank. Nonetheless, insufficient biological mass is present in the contact
4-4
-------
FIGURE 4-1
MODIFICATIONS OF THE ACTIVATED SLUDGE PROCESS
RAW
WASTEWATER
FINAL \EFFLUENT
CLARIFIER I *
' ' EXCESS SLUDGE
^
CONVENTIONAL ACTIVATED SLUDGE PLANT
RAW
WASTEWATER
OR —
PRIMARY
EFFLUENT
tttt ttt
AERATION TANK
4 * I
FINAL \ EFFLUENT
I CLARIF IER I »•
RETURN SLUDGE
EXCESS SLUDGE
COMPLETE MIX PLANT
RAW
WASTEWATER
SLUDGE
r~^^
i r
SLUDGE
REAERATION
TANK
CONTACT
TANK
RETURN SLUDGE
EXCESS SLUDGE
CONTACT STABILIZATION PLANT
RAW
PRIMARY
^ CLARIFIER ^
•EWATER
JsLUDGE
* *
r
AER
>
ATION
V
r
T A
J
rr t
RETURN
~>
NK
SLUDGE
FIN
CLAR
s
\
AL ^ EFf
IFIER I1
1 EXCESS S
STEP AERATION PLANT
4-5
-------
tank to completely nitrify the ammonia and since ammonia is not adsorbed on the
biological floe, ammonia will bleed through to the effluent. Partial nitrification can be
obtained at levels which can be predicted by methods presented in Section 4.3.6.
4.3.1.5 Step Aeration and Sludge Reaeration Plants
A typical step aeration plant is illustrated on Figure 4-1. Like the conventional plant, the
return sludge is introduced at the head end of the aeration tank. However, the step aeration
plant differs from the conventional plant in that influent wastewater is introduced at several
points along the aeration tank. This distribution of influent flow reduces the initial oxygen
demand usually experienced in the conventional plant. 25
A variation on the step aeration plant that has been popular on the West Coast is to
introduce no feed into the first pass while directing the flow into the remaining downstream
passes. A sludge reaeration zone is established in the first pass and this variation has become
known as a "sludge reaeration plant." Normally, no provision is made to prevent back
mixing between the sludge reaeration pass and the downstream passes.
The ammonia bleedthrough characterizing contact stabilization plants is avoided in a step
aeration plant because of the greater contact times employed and backmixing of influent
occurs. Nonetheless, some bleedthrough of ammonia as well as organic nitrogen can occur.
This breakthrough results from short circuiting of influent to the effluent and insufficient
contact time for complete organic nitrogen hydrolysis (ammonification) and oxidation of
ammonia.
4.3.1.6 High Rate and Modified Activated Sludge
High rate activated sludge processes (high MLSS) and modified activated sludge (low MLSS)
processes are characterized by low solids retention times (0.5 days). Under these conditions,
a nitrifying activated sludge cannot be developed. The high rate and modified activated
sludge processes are acceptable pretreatment techniques for separate stage nitrification
processes (Section 4.5).
4.3.1.7 High Purity Oxygen Activated Sludge Plants
Both covered and uncovered reactors have been used with pure oxygen activated sludge, but
only the former technique has seen actual implementation in full-scale plants. 28 The
covered reactor approach involves the recirculatioh of reactor off-gases to achieve efficient
oxygen utilization. As a consequence, the carbon dioxide which is present in the off-gas is
returned to the liquid. The end result is that high carbon dioxide concentrations build up in
the mixed liquor and recycle gases, depressing the mixed liquor pH. pH levels as low as 6.0
are not uncommon. This effect can have a depressing effect on nitrification rates (cf.
Sections 3.2.5.6 and 4.6.3), resulting in the requirement for somewhat longer solids
retention times for nitrification than would otherwise be the case.
4-6
-------
Virtually all applications of the high purity oxygen activated sludge process to nitrification
have been for separate stage nitrification applications (Section 4.6), rather than for
combined carbon oxidation-nitrification applications.
4.3.2 Utility of Nitrification Kinetic Theory in Design
The nitrification kinetic theory presented in Chapter 3 may be directly applied to the design
of those activated sludge modifications compatible with nitrification. The equations must be
adapted to the hydraulic configuration under consideration, but in all cases this adaptation
is relatively straightforward.
Nitrification kinetic theory can be very usefully applied to define the following parameters:
1. The safety factor required to handle diurnal transients in loading to prevent
significant ammonia bleedthrough under peak load conditions.
2. The design solids retention time under the most adverse conditions of pH, DO and
temperature.
3. The allowable organic loading on the combined carbon oxidation-nitrification
stage.
4. The required hydraulic detention time in the aeration tank at ADWF.
5. The excess sludge wasting schedule.
The following sections present the design procedures in terms of a number of specific
examples. The procedure developed for each case has often been termed the "solids
retention time" design approach.
4.3.3 Complete Mix Activated Sludge Kinetics
As a design example, consider a 1 mgd treatment plant that must achieve complete
nitrification at 15 C. The plant incorporates primary treatment. Primary effluent BOD5 is
150 mg/1, including solids handling return streams to the primary. Total Kjeldahl Nitrogen
(TKN) is 25 mg/1 as N. As a simplifying assumption, neglect that portion of the TKN that is
assimilated into biomass or associated with refractory organics. The wastewater has an
alkalinity of 280 mg/1 as CaCO3- The procedure is as follows:
1. Establish the safety factor, SF. The SF is affected by the desired effluent quality.
Assume a minimum SF of 2.5 is required due to transient loading conditions at
this particular plant (see Section 4.3.3.2).
2. Establish the minimum mixed liquor dissolved oxygen (DO) concentration.
4-7
-------
Consideration of aeration efficiency at the peak hourly load is required (see
Section 4.8). Assume a minimum DO of 2.0 mg/1 is selected as a compromise
between power requirements and a consideration of the depressing effects of low
DO levels on the rate of nitrification as discussed in Section 3.2.5.5.
3. Estimate the process operating pH (see Section 4.9.2). Approximately 7.14 mg/1
of alkalinity as CaCO3 is destroyed per mg/1 of NH4 -N oxidized. Neglecting the
incorporation of nitrogen into biomass, the alkalinity remaining after nitrification
will be at least:
280- [7.14(25)]= 102 mg/1
If a coarse bubble aeration system is chosen, the pH should remain above pH 7.2
and chemical addition is not required for pH control (see Section 4.9.2).
4. Calculate the maximum growth rate of nitrifiers at 15 C, DO = 2 mg/1, and
pH >7.2. The appropriate equation to be used was presented in Section 3.2.6 and
is as follows:
where:
~*
xr
K
_J
maximum possible nitrifier growth rate, day ,
environmental conditions of pH, temperature, and DO,
_1
maximum nitrifier growth rate, day , and
half-saturation constant for oxygen, mg/1.
The last bracketed term is taken as unity at a pH above 7.2. Using the specific
values adopted in Section 3.2.5 for JLI^ and KQ2 leads to the following expres-
sion:
0.098(T-15)
DO
DO + 1.3
- 0.833(7. 2 -pH)
)J
(4-1)
Using the numbers given above:
M = (0.47)(0.61) = 0.285 day
"1
5. Calculate the minimum solids retention time for nitrification. From Equation
3-15, the correct expression is:
4-8
-------
0m= -4- . (4-2)
c , "N
where: d = minimum solids retention time, days, for nitrification at pH,
temperature and DO.
For this example:
6. Calculate the design solids retention time. From Equation 3-29, the correct
expression is:
0J? = SF.0m (4-3)
where: d = solids retention time of design, days.
C/
For this example:
0 =2.5(3.51) = 8.78 days.
C
7. Calculate the design nitrifier growth rate. From Equation 3-12, the correct
expression is:
where: ju = nitrifier growth rate Nitroso mo nas , day" .
For this example:
= -8778 =0-114 day'1
8. Calculate the half-saturation constant for ammonia oxidation at 15 C. The proper
expression is:
KN=100.051T-1.158 (3-13)
where: KN = half-saturation constant for NH, - N, mg/1, and
T = Temperature, C
For this example:
KN=1(T0-3 = 0.405 mg/1
4-9
-------
9. Calculate the steady state ammonia content of the effluent. Equation 3-24 is
directly applicable to complete mix activated sludge systems, where Nj is the
effluent ammonia-nitrogen content:
N.
M =MN - 1 - (3-24)
N IN v-
where: Nj = effluent NH* - N, mg/1
For this case:
Nl
M = 0.1 14 = 0.285
Nj +0.405
=0.27mg/l
Transient loading effects' on effluent quality are presented in Section 4.3.3.2.
10. Calculate -the organic removal rate. The design solids retention time 0 applies to
C
both the nitrifier population and the heterotrophic population. Equation 3-27 can
be applied to determine substrate removal rates:
=-=-K (3'27)
where: Yb = heterotrophic yield coefficient, Ib VSS grown per Ib BOD,
removed,
qb = rate of substrate removal, Ib BOD5 removed/lb VSS/day, and
K, = "decay" coefficient, day .
Assume representative values for Y, and K , :^
o a
Yb = 0.65 Ib VSS/lb BOD rem.
Kd = 0.05 day'1
Therefore:
0.114 = 0.65qb-0.05
qb = 0.252 Ib BOD rem./lb MLVSS/day
4-10
-------
In the above calculation of qj,, it is assumed that the fraction of nitrifiers is low
and can be neglected (see Section 4.6.1 for a discussion of this point).
11. Determine the hydraulic detention time at ADWF. In this analysis, the MLVSS
content and effluent soluble BOD must be known. The effluent soluble BOD5
can be assumed to be very low (say 2 mg/1). The MLVSS content is dependent on
the mixed liquor total suspended solids, which is in turn dependent on the
operation of the nitrification-sedimentation tank (Section 4.10). Assume for the
purposes of this example that the design mixed liquor content at 15 C is
2500 mg/1. At a volatile content of 75 percent, the MLVSS is 0.75 (2500) =
1875 mg/1. From Equation 3-28, the expression for hydraulic detention time is:
S - S
(4_5)
Xlqb
where: HT = hydraulic detention time, days,
X. = mixed liquor volatile suspended solids, MLVSS, mg/1,
SQ = influent total BOD5, mg/1, and
Sj = effluent soluble BOD5, mg/1.
For this example, the hydraulic detention time at ADWF is:
HT°(1875)(0.252)°°-3'3dayS
= 7.5 hours
12. Determine the organic loading per unit volume. The volume required in the
aeration basin for 1 mgd flow is:
Volume = Q • HT = 1(0.313) = 0.313 mil gal = 41,844 cu ft
where: Q = influent flow rate, mgd
The BOD- loading is:
(1)(8.33)(150) = 1249 Ib/day
4-11
-------
The BOD5 load per 1000 cu ft is:
1249
j = 29.9 lbBOD5/l000 cu ft/day
13. Determine the sludge wasting schedule. Sludge is wasted from the system from
two sources: (1) solids contained in the effluent from the secondary sedimenta-
tion tank, and (2) intentional sludge wasting from the return sludge or mixed
liquor. The sludge to be wasted under steady state conditions can be calculated
from the solids retention time. The total sludge wasted per day is:
X + W-X) (4-6)
where: S = total sludge wasted in Ib/day,
W = waste sludge flow rate, mgd
X~ = effluent volatile suspended solids, mg/1, and
X = waste sludge volatile suspended solids, mg/1
The inventory of sludge in the system is:
I = 8.33(Xj • V) (4-7)
where: I = inventory of VSS under aeration, Ib, and
V = volume of aeration tank, mil gal
The solids retention time is defined as:
(4-8)
In this case, application of Equation 4-7 yields:
I = 8.33(1875X0.313) = 4889 Ib VSS
Using Equation 4-8 and a design 6 of 8.78 days, the sludge wasted from the
system is:
S = 4889/8.78 = 557 Ib/VSS day
4-12
-------
The sludge contained in the effluent at 1 mgd can be calculated assuming that the efflu-
ent volatile suspended solids is equal to 12 mg/1:
8.33(1) (12)= lOOlbVSS/day
By difference, the Ib of MLVSS to be wasted from the mixed liquor or return sludge is:
557 - 100 = 457 Ib VSS/day
4.3.3.1 Effect of Temperature and Safety Factor on Design
The design example presented in the previous section provided one solution to a set of
stated conditions. Alteration of the lowest temperature at which nitrification will be
supported, or the design safety factor, or the wastewater strength, or the assumption of
different kinetic constants can materially alter the design.
To give one illustration, Table 4-2 has been prepared using differing safety factors (2.0 to
3.0) and differing minimum wastewater temperatures with design calculations to derive the
computed quantities shown. Assumptions have been made for illustrative purposes as to the
allowable MLSS. Allowable mixed liquor levels are a function of sedimentation tank
operation. The mixed liquor level that can be maintained will be affected by reduced
sedimentation efficiency at lower temperatures. Consideration of aeration tank-secondary
sedimentation tank interactions is presented in Section 4.10.
As can be seen from Table 4-2, low temperature applications (10 C) of combined carbon
oxidation-nitrification in complete mix activated sludge systems require very long hydraulic
residence times to achieve favorable conditions for nitrification. This factor was one of the
reasons for the development of separate stage nitrification systems. As temperatures rise,
required residence times are materially reduced. At 20 C, less than five hours is required for
virtually complete nitrification in the specific case examined. While it is possible to design
for nitrification using the relatively low detention times given in Table 4-2 for 20 C, special
attention must be given to oxygen transfer as a very high oxygen demand is expressed per
unit volume. Considerations for oxygen transfer are given in Section 4.8.
4.3.3.2 Consideration in the Selection of SF
In introducing the safety factor concept to the design of biological treatment systems,
Lawrence and McCarty29 noted that the SF was necessary to achieve high efficiency of
treatment, to insure process stability and to provide resistance to toxic upsets. Excessively
high safety factors resulted in higher operating and capital costs. It was noted that the safety
factor concept had been implicitly incorporated into treatment plant design practice by the
selection of solids retention times in excess of 0m
c
4-13
-------
TABLE 4-2
CALCULATED DESIGN PARAMETERS FOR A 1 MGD
COMPLETE MIX ACTIVATED SLUDGE PLANT
Minimum
temp, for
nitrification,
C
10
15
20
Maximum
possible
nitrlfier.
growth rate,
>»N . day-1
0.175
0.285
0.465
Assumed
allowable
MLSS/MLVSS
mg/I
2,000 s^
//
s' 1, 500
2,500 s'
^<^
^^ 1, 875
3,000 ./
//
s/ 2, 250
Safety
Factor,
SF
2.0
2.5
3.0
2.0
2.5
3.0
2.0
2.5
3.0
Design
solids
retention
time, days
e?
11.5
14.3
17.2
7.0
8.8
10.5
4.3
5.4
6.4
Steady
state
effluent
NH+-N,
mg/1
0.23
0.15
0.11
0.40
0.27
0.20
0.73
0.49
0.36
Organic
removal
rate,
Ib BODrem/
Ib MLVSS-day
0.21
0.19
0.17
0.29
0.25
0.22
0.44
0.36
0.32
Hydraulic
retention
time, a
hours
11.0
12.8
14.0
6.4
7.5
8.5
4.4
5.2
6.0
BOD5
loading
(volumetric) .
lb/1000/cf/day°
20.5
17.5
15.8
34.9
29.9
26.5
51.5
43.0
37.3
* AtADWF
b 62.4 lb/1000 cf/day = kg/m3/day
-------
Because the SF concept is relatively new, there is no plant scale experience with its
application accumulated as yet on which to base broad recommendations. Rather, kinetic
theory itself is used in this section to establish minimum factors of safety considering the
desired degree of nitrification under steady state and transient load conditions. It must be
emphasized that these are minimum values and individual designs may exceed these values
for a variety of reasons. For instance, the presence of industrial wastes may adversely affect
nitrification rates, requiring conservatism in the selection of the SF.
Figure 4-2 provides a wider array of safety factors for the design example presented in Table
4-2. As may be seen, the selection of the SF has a marked effect on the calculated steady
state values of ammonia in the effluent. If relatively complete nitrification is to be obtained
(at steady-state) resulting in 0.5-2 mg/1 of ammonia nitrogen in the effluent, a minimum SF
of 1.5 is appropriate for application to complete mix activated sludge systems. Further,
effluent values for a comparable plug flow system are also shown in Figure 4-2 (see Section
4.3.5 for plug flow data). As may be seen, complete mix systems have higher effluent
ammonia levels than plug flow systems at the same SF.
In all practical applications, waste treatment plants do not operate at "steady state."
Significant diurnal variation in the nitrogen loading on such systems occurs. Figure 4-3
shows the diurnal variations in influent flow and TKN loading experienced at the Chapel
Hill, N.C. treatment plant. The ratio of the maximum TKN loading to the average was 2.17,
while the ratio of the maximum to minimum was 6.72. The Chapel Hill system is a relatively
small system (1.8 mgd) with high peak to average ratios for all constituents.30 The variation
in load for each community will be a function of the unique characteristics of that
community (see Section 4.8), and data must be individually developed for each situation.
TKN load variations have a significant impact on nitrification kinetics, and ammonia
bleedthrough can occur under peak load situations.31>32 Kinetic theory can be applied to
these situations, however, and the safety factor established at levels which will prevent
ammonia bleedthrough from causing significant deterioration of effluent quality.
A mass balance on nitrogen in the organic and ammonia form can be made at any time
during a diurnal cycle which states that the influent TKN load is equal to the effluent
ammonia load plus that nitrified in the complete-mix reactor during any time, At:
NQQAt = qNfX 1 VAt + N {Q At (4-9)
where: N = influent TKN concentration, mg/1,
N. = effluent ammonia nitrogen concentration, mg/1,
Q = influent or effluent flow rate, mgd,
At = time increment,
4-15
-------
V = volume of aeration basin, mil gal,
f = nitrifier fraction of the mixed liquor solids
FIGURE 4-2
EFFECT OF THE SAFETY FACTOR ON STEADY STATE EFFLUENT
AMMONIA LEVELS IN SUSPENDED GROWTH SYSTEMS
2: 3
0
I
I
A. at 20C
COMPLETE MIX
1.0 1.2 1.4 1.6 1.8 -2.0 2.2 2.4 2.6 2.8 3.O
SAFETY FACTOR, SF
2.5
I
t-
3:
1.5
1.0
0.5
0
I I r
B. at 10 C
COMPLETE MIX
1.0 1.2 1.4
1.6 1,8 2.0 2.2 2.4 2.6 2.8 3.0
SAFETY FACTOR, SF
4-16
-------
FIGURE 4-3
DIURNAL VARIATIONS AT THE CHAPEL HILL, N.C.
TREATMENT PLANT (AFTER HANSON, ET AL. (30))
zoo
180
I6O
Ul
v> 120
S
Uj
i. IOO
8O
I
T
Ul
o
(t
UJ
a.
6O
40
I
I
I
I
I
2400
04OO
0600
20OO
1200 1600
TIME
DIURNAL VARIATION IN WASTEWATER FLOW
2400
22O
ZOO
180
160
I4O
IZO
100
Ul
o 60
Ul
40
20
T
INFLUENT
LOAD
INFLUENT
CONCENTRATION
I
_L
I
I
I
2400
0400
0800
I20O
TIME
1600
ZOOO
2400
DIURNAL VARIATION IN NITROGEN LOAD AND CONCENTRATION
4-17
-------
This equation neglects synthesis terms, assumes all influent organic N is hydrolyzed, and
neglects terms relating to the rate of change of ammonia concentration in the reactor.
Numerical solution techniques are available to handle transient load effects more exactly.32
Equation 4-9, however, is useful for approximating the effects of transient loads.
Equation 4-9 may be solved for Nj, by substitution for the terms for nitrification rate, q-^
and the term fXjV, representing the inventory of nitrifying organisms. The inventory of
nitrifying organisms can be related to the solids retention time through the following
equation:
A fX, V
d i - _ (4.10)
where: N = 24 hr-average influent TKN, mg/1
N, = 24 hr-average effluent NH. - N, mg/1
Q = mean flow rate (ADWF), mgd, and
YN = nitrifier yield coefficient, Ib VSS/lb NH* -N removed.
The term C>YN(NO-NI) represents the quantity of nitrifiers grown per day, which must be
wasted each day to establish a steady-stage solids retentiofi time, Q £. The average terms, No
and Nj, are flow weighted averages of nitrogen concentration of an entire day (the
equivalent of composite samples). Q represents the average dry weather flow (ADWF).
The nitrification rate from Equations 3-20 and 3-24 is:
N
Substitution of Equations 4-10, 4-1 1 and Equation 3-29 into Equation 4-9 yields:
Equation 4-12 can be used to solve for Nj over a 24-hr cycle since all other quantities in
Equation 4-12 are known or can be estimated. Initially, Nj can be estimated to be the
calculated steady-state value. Once Equation 4-12 has been applied to generate a 24-hr cycle
of NI values, a new value of Nj may be calculated. If N\ differs significantly from the
initial assumption, the calculation process can be repeated.
4-18
-------
Equation 4-12 has been applied to the variations in load observed at Chapel Hill, and using
the design information used to generate Table 4-2 at a temperature of 15 C. The results of
this analysis are plotted in Figure 4-4 for three different assumed safety factors, 1.5, 2.0,
and 2.5. As may be seen, the assumption of the safety factor has a marked effect on the
average effluent ammonia content, Nj. For this particular case, the ratio of peak to average
TKN loading was 2.2; the SF had to exceed this ratio (2.5) to produce an effluent that had,
on the average, less than 1 mg/1 of ammonia-N.
The application of Equation 4-12 to several other such cases*, showed the same effect;
namely, the minimum safety factor should equal or exceed the ratio of peak ammonia load
to average load to prevent high ammonia bleedthrough at peak loads. This statement may be
used as "a rule of thumb" for designing suspended growth nitrification systems operated in
the complete mix mode.
A flow equalization procedure applicable to reducing diurnal peaking on nitrification
systems is presented in Chapter 3 of the Process Design Manual for Upgrading Existing
Waste-water Treatment Plants.^ By incorporating flow equalization into treatment plants,
FIGURE 4-4
EFFECT OF SF ON DIURNAL VARIATION IN EFFLUENT AMMONIA
o
Ul
o
o
o
I
It
lu
20 r
24 hr average
composite IMH^-N
concentration
04 OO
08OO
I20O
TIME, HR
I6OO
2000
2400
4-19
-------
the safety factor used in kinetic design of the nitrification tanks may be reduced. Case
examples for treatment plants incorporating flow equalization are presented in Sections
9.5.1.1, 9.5.1.2 and 9.5.2.1.
4.3.4 Extended Aeration Activated Sludge Kinetics
The procedure presented in Section 4.3.3 for complete mix activated sludge kinetics is
directly applicable to extended aeration activated sludge. Extended aeration systems are
usually operated at such long solids retention times that except during cold temperatures
(5-10 C) nitrification is usually obtained in properly operated systems.
4.3.5 Conventional Activated Sludge (Plug Flow) Kinetics
The approach for conventional activated sludge plants is similar to that for complete mix
plants with the exception of the equations used to predict effluent quality. The plug flow
model may be applied to approximate the hydraulic regime in these plants. The Monod
expression for substrate removal rate (Equation 3-24) must be integrated over the period of
time an element of liquid remains in the nitrification tank. The following is a solution for
plug flow kinetics that can be adapted to this problem as shown :^9
1 ^N^o'Nl-1
—- = for r < 1
0 No
c (N -N,) + KXTln —-
where: r = recycle ratio (or return sludge ratio).
0
- = - ^— - — forr
-------
A typical DO and nitrification pattern for plug flow tanks where aeration capability is
limited in the front end of the tank is presented in Figure 4-5, where an aeration tank
profile for DO and ammonia nitrogen is plotted. As may be seen, nitrification is inhibited in
the first portion of the tank, because of the DO suppression due to carbon oxidation. Once
the DO rises, the ammonia level falls at a reaction rate that approximates zero order, a
reactor order predicted by kinetic theory (Section 3.2.7). It is notable that if sufficient
aeration capability had been available in the head end of the tank, virtually complete
nitrification probably would have been obtained.
Thus, the first portion of plug flow tanks may be ineffective for nitrification, reducing the
effective contact time for nitrification. 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 degree to which full-scale nitrification tanks approach plug-flow operation can be
examined through reactor diffusion theory.34'35 Reactors can be characterized by an axial
disperson number, D/uL, where D is the axial disperson coefficient in square ft per hr, u is
the mean displacement velocity along the tank length, in feet per hr, and L is the tank
length, ft. In the calculation of the axial disperson number, u and L are known for any
FIGURE 4-5
DO AND AMMONIA NITROGEN PROFILE IN A PLUG-FLOW SYSTEM
(AFTER NAGEL AND HAWORTH (33))
L
AERATION
TANK
-BAFFLES-
£3
20
"\
p l5
~
l 10
_j_~
3;
^ 5
n
b ^ i i i i i
*~~" » """
^^QL
% ^/-^^M* ^^^ ^^^ "~ "
— *^*^ ~~
P\Q C^^ *^^
^•^-^ ^d£.
— T' ^^, —
^r ^^^W ^
s + r *^o,^
_ ^**s N"4 *^-^-
°"* 1 I.I 1 1
f.3
2.0 I
Z
1.5 uj
X
X
1.0 Q
Q
kl
°'5 i
n Q
ISO 200 250
DISTANCE ALONG TANK, FEET
4-21
-------
particular plant design and D must be measured. A valid empirical relationship for D for
both fine and coarse bubble diffused air plants is as follows:-^
D = 3.118 W2(A)°'346 (4-15)
where: W = tank width, ft and
A = air flow per unit tank volume, in
standard cubic feet per minute
per 1,000 cu ft.
The axial disperson coefficient, D/uL is zero for true plug flow plants and infinite (<*=) for
true complete mix plants. Plants with D/uL between 0 to 0.2 are usually classed as plug flow
reactors, while for complete mix systems, D/uL is usually in the range from 4.0 to oo .36
As an example calculation, the Central Contra Costa Sanitary District (CCCSD) plant's
nitrification tanks (Section 9.5.2.1) have the following characteristics:
Air How (av.) = 51.1 SCFM/1000 CF
Width = 35 ft
Area of tank = 525 sf
Length (all 4 passes) = 1080 ft
Flow each tank (4 passes) @ 50% recycle = 22.5 mgd
From the above data, the mean displacement velocity can be calculated to be 239 ft/hr.
From Equation 4-15, the diffusion coefficient is:
D = 3.1 18 (35)2 (51.1)0.346 = 14,939 ft2/hr
Therefore:
D/uL = 14,898/239 (1080) = 0.058
Therefore, the CCCSD nitrification tanks closely approach a plug flow reactor.
Equation 4-15 can be utilized to evaluate mixing in actual plant designs to determine
whether they approach plug flow closely enough to allow use of Equation 4-15 to describe
nitrification. It is probable that most plants operated in the conventional mode do approach
plug flow. For those plants with intermediate values of D/uL, complete mix kinetics can be
employed which yield conservative answers.
The hydraulic configuration of nitrification tanks can also be designed to discourage back
mixing. A series of complete mix tanks can approximate a plug flow reactor. In the case
example for Canberra, Australia (Section 9.5.2.2) complete mix reactors are used in series
for nitrification. Absolute prevention of back mixing is provided by virtue of the mixed
liquor overflowing weirs between reactors. Available head at the site was utilized,
4-22
-------
eliminating the need for mixed liquor pumping.
4.3.5. 1 Considerations in the Selection of the Safety Factor
The factors affecting the choice of the SF for plug flow activated sludge are similar to those
for complete mix applications. Diurnal peaking in load has an important influence on the
choice of the SF, although kinetic models have not been extended to handle diurnal loads in
plug flow systems at the present time. It can be expected that the effects of diurna} loads on
plug flow systems will be similar to those for complete mix systems as when the effluent
ammonia nitrogen level rises to 2 to 3 mg/1 or above, the rate of removal becomes a zero
order reaction (unaffected by ammonia nitrogen concentration). In zero order reaction
situations, differences between plug flow and complete mix kinetics are negligible.
Therefore, the adoption of the criteria advanced for complete mix systems (that the
minimum SF equal or exceed the ratio of peak ammonia load to average daily load), should
prevent high ammonia bleedthrough during diurnal peak loads. The problem of low
dissolved oxygen due to carbonaceous load in the head end of plug flow systems should be
considered in aeration design for combined carbon oxidation-nitrification applications;
indeed, this factor alone would justify a conservative safety factor.
4.3.5.2 Kinetic Design Approach
The kinetic design approach for plug flow (conventional) plants is identical to that
presented in Section 4.3.3, excepting in Step 9, where Equation 4-13 is used instead of
Equation 3-24. If a portion of the nitrification tank is rendered ineffective by DO
suppression at its head end, then only the sludge inventory maintained under adequate DO
conditions should be used in the calculation of 0 r or the SF.
L*
4.3.6 Contact Stabilization Activated Sludge Kinetics
Gujer and Jenkins37,38 have developed the kinetic design procedure for nitrification in
contact-stabilization activated sludge plants. The procedure described herein is a summary
of their approach, and the jeader is referred to their publications for theoretical bases.
The overall nitrifier growth rate in the contact stabilization process is the weighted mean of
their growth rate in the contact tank and in the stablization (sludge reaeration) tank:
where: JUN> ju , p. = growth rate of the nitrifiers in the overall process,
in the contact and stabilization tanks respectively
(day'1).
C, B = the fractions of total sludge in the contact and
stabilization basins respectively.
4-23
-------
Gujer and Jenkins used Equation 4-16, and Monod type expressions for complete mix tanks
to develop a graphical solution for nitrification in contact stabilization (Figure 4-6). In
Figure 4-6, the efficiency of nitrification, 7?m-t, is defined as a fraction by:
(N03}c
n . = — (4-17)
"" OK.?.
where: (NO- ) = NO- - N level in the contact tank, mg/1, and
j C J
(NO-) = NOZ - N level in the stabilization tank, mg/1.
4.3.6.1 Design Example
As a design example, consider a 1 mgd contact stabilization plant operated at a minimum
temperature of 15 C. Influent BOD5 is 150 mg/1, including solids handling returns to the
primary. Total Kjeldahl Nitrogen is 30 mg/1, of which 21 mg/1 is ammonia -N and 9 mg/1 is
organic -N. The wastewater has an alkalinity of 210 mg/1. The effluent requirement is not
more than 10 mg/1 reduced soluble nitrogen (organic and ammonia). The procedure is as
follows:
1. Establish a reasonable safety factor for nitrification, say 2.5 as in Section 4.3.3.
2. Establish the minimum mixed liquor DO; assume 2.0 mg/1 as in Section 4.3.3.
3. Establish the maximum growth rate of nitrifiers, assuming for the moment that
there is sufficient alkalinity in the wastewater to buffer the nitrification pH to
greater than 7.2 (see step 14). Therefore,
AN = 0.285 day'1 as in Section 4.3.3, step 4.
4. Calculate the minimum solids retention time, the design solids retention time and
the actual nitrifier growth rate (as in Section 4.3.3, Steps 5, 6, 7):
0m = 3.51 days
0d = 8.78 days
C
= 0.114day"1
4-24
-------
FIGURE 4-6
NITRIFICATION EFFICIENCY AS A FUNCTION OF
PROCESS PARAMETERS (AFTER GUJER AND JENKINS (37))
Key:
1
/ 2 3
SLUDGE RECYCLE RATIO R/Q
DESIGN EXAMPLE
4-25
-------
5. Calculate the organic removal rate, as in Step 10, Section 4.3.3:
qb = 0.252 Ib BOD rem/lb MLVSS/day
6. Calculate the VSS produced per unit of wastewater treated:
(SQ - Sj) MN/qb = (150 - 2)(0.114)/0.252 = 67.0 mg/1
7. Compute the nitrogen incorporated into VSS, assuming 12 percent nitrogen
incorporated into VSS:
0.12 (67.0) = 8.0 mg/1 N
8. Compute the soluble N content of the effluent, assuming no denitrification. The
effluent soluble N = the total influent N minus N incorporated into VSS:
30 - 8.0 = 22.0 mg/1 soluble N
9. Calculate the soluble organic N in the effluent. Gujer and Jenkins found that 40
percent of the influent organic N appeared in soluble form in the effluent:
0.4 (9) = 3.6 mg/1 organic N
^ 4.0 mg/1 organic N
10. Calculate ammonia nitrogen in the effluent under steady state conditions:
Total reduced N - organic N = 10-4 = 6 mg/1 NH^ -N
11. Calculate nitrate nitrogen in the effluent and in the contact tank under steady
state conditions:
Total soluble N - total reduced N = (NCQ
j C
(N0~)c = 22-10=12mg/l
12. Calculate the required nitrification efficiency from Equation 4-17:
7?nit=l 2/(12 + 6) = 0.667
In this calculation, it is assumed that the concentration of nitrate nitrogen in the
stabilization tank (NO-) totals 18 mg/1, since the contact tank concentration is
12 mg/1 and with the assumption that the 6 mg/1 of ammonia nitrogen is
completely nitrified in the stablization tank.
4-26
-------
13. Calculate the required sludge recycle ratio. Assume the fraction of biomass in the
contact tank is 15 percent (C= 0.15). The dimensionless number A, is used in the
calculation; A is defined as follows:
A =
For this example:
A =
SF
-C
0.15
+ 1
(4-18)
_L -0.15
2.5
+ 1 = 1.60
The required sludge recycle ratio, R/Q, depends on the value of A and the
required nitrification efficiency as follows:
R/Q =
(4-19)
where: R = recycle flow rate, mgd
Q = influent flow rate, mgd
14.
For this example:
R/Q =
0.667
1.60(1-0.667)
•= 1.25
Since Q has been assumed to be 1 mgd, the return activated sludge rate is 1.25
mgd.
This example is also worked graphically in Figure 4-6. The top part of the figure is
used first by entering the abscissa with the value of the SF and rising vertically to
the chosen value of C and then reading the value of A on the ordinate. The
bottom part of Figure 4-6 is then used; the nitrification efficiency, nnit> *s
entered on the ordinate and traveling horizontally to the value of A just
determined and then finding the required recycle ratio on the abscissa. Figure 4-6
also demonstrates a general result; in order to obtain high nitrification efficiency,
a higher than normal sludge recycle ratio must be employed.
Check the buffering of the wastewater. The quantity of ammonia nitrified is
reflected by the level of nitrate in the process effluent. Approximately 7.14 mg/1
of alkalinity as CaCC>3 is destroyed per mg/1 of NH^-N oxidized. The alkalinity
remaining after nitrification would be at least:
4-27
-------
210-7.14(12)= 124mg/l as CaCO3
This should be sufficient residual alkalinity to maintain the pH above 7.2 for
coarse bubble aeration systems. If a fine bubble aeration system were chosen,
chemical addition would be required and the dose estimated from the
procedures discussed in Section 4.9.2. Alternatively, a lower operating pH could
be used with a longer aeration period.
15. Calculate required reactor volumes. As in Section 4.3.3, assume the mixed liquor
content in the contact tank is 2500 mg/1 at a volatile content of 75 percent. The
mixed liquor volatile suspended solids in stabilization can be obtained from the
balance:
(Q + R) X,, = RXc (4-20)
c s
where: X = contact MLVSS, mg/1, and
C
X = stabilization MLVSS, mg/1.
Therefore: X =1875 l + L25 = 3375 mg/1
s 1.25
The total sludge inventory can be calculated from the following equation for
substrate removal rate:
Q(S -Sj)
q = 2 L. (4.21)
b 2XV
where: SXV = total inventory of MLVSS in the contact and
stabilization tanks, Ib
therefore: 2XV = - = 4889 Ib
(.252)
Of this inventory, 15 percent is in the contact tank:
0.15(4889) = 733 Ib MLVSS
The remainder is in the stabilization tank:
4889 - 733 = 4156 Ib MLVSS
4-28
-------
The volume in contact is:
V =733/(8.33)(l875) = 0.047 mil gal
C
The volume in stabilization is:
Vg = 4156/(8.33)(3375) = 0.148 mil gal
The total volume is 0.148 + 0.047 = 0.195 mil gal
16. Calculate the residence time in contact.
0 =V /Q = 0.047 days = 1.1 hr
c c
17. Calculate the sludge wasting schedule. See Section 4.3.3, Step 13.
As can be seen from Figure 4-6, the design of contact stabilization for nitrification is highly
sensitive to the safety factor chosen and the sludge recirculation ratio. A wide range of
alternate designs can be derived from variation in these parameters. Required reactor
volumes are also sensitive to the assumed growth rate of nitrifiers, creating a need for kinetic
data of high accuracy when designing for contact stabilization.
The design procedure described is based on the assumption of steady state operation.
Diurnal variations in load will cause average effluent ammonia levels to exceed those
calculated above. To compensate for this, it would be necessary to use an even higher SF
than assumed in the above example. Regardless of the safety factor chosen, contact
stabilization plants cannot be expected to completely nitrify except under the normally
impractical condition of very high recycle rates (R/Q = 4.0 and above). However, at high
recycle ratios the major advantage of contact stabilization is lost because the sludge in the
stabilization basin becomes more dilute and the overall basin volume requirements approach
those of the conventional process. This limits the application of contact stabilization to
situations where only partial nitrification is required.
A further limitation on nitrification in contact stabilization plants is the incomplete
hydrolysis of organic nitrogen occurring in the short detention time contact tank. As noted
under step 9 above, about 40 percent of the influent organic nitrogen appears in the process
effluent. Conventional or complete mix activated sludge plants, on the other hand, have
sufficient contact time to hydrolyze the bulk of the organic nitrogen to ammonia thus
making the nitrogen available to nitrifiers and leaving very little organic nitrogen in the
effluent.
The short contact time of the contact tank can create problems in the sedimentation tank.
The mixed liquor solids are not well stabilized in the contact tank prior to sedimentation.
4-29
-------
Denitrification activity in the sedimentation tank is therefore greater than in conventional
or complete mix plants and floating sludge may be the result. Procedures for circumventing
the floating sludge problem are discussed in Section 4. 1 0.
4.3.7 Step Aeration Activated Sludge Kinetics
Because of backmixing, the step feed pattern of step aeration plants causes the kinetics of
such plants to more closely approach complete mix than plug flow. As a result, the design
approach developed for complete mix (Section 4.3.3) can usually be employed for step
aeration plants as a reasonable approximation. In those step aeration plants where influent is
fed to the last pass (as in Figure 4-1), there is the danger that there will be insufficient time
for the organic nitrogen to be hydrolyzed prior to discharge, resulting in elevated quantities
of organic nitrogen in the effluent. Further discussion of this effect is presented in Section
4.3.8.2 which is a description of an operating step aeration plant.
4.3.8 Operating Experience with Combined Carbon Oxidation-Nitrification in Suspended
Growth Reactors
While activated sludge-type systems are commonly used in England to obtain dependable
nitrification, their use in the U.S. has not been widespread. Early U.S. activated sludge
plants of the conventional design nitrified in the warmer months of the year or if they were
underloaded. But nitrification became unpopular because of the additional aeration power
cost and the propensity of some sludges to float in the sedimentation tank when nitrifying,
and it was questioned whether the added expense was worth it in many cases.39,40 AS a
consequence, ways and means were sought to prevent nitrification rather than to encourage
it through increasing organic loading or through tapered aeration or by picking
modifications of the process which were less favorable for nitrification. This early
experience with the process may have led to uncertainty about its reliability.
Nonetheless, there have been several plant-scale operations in the U.S. which have
demonstrated the viability of the process. The purpose of this section is to review some of
these cases. Other case examples are presented in Sections 9.5.1 and 9.5.2.
4.3.8.1 Step Aeration Activated Sludge In a Moderate Climate
The Whittier Narrows Water Reclamation Plant is a 1 2 mgd activated sludge plant designed
and operated by the Los Angeles County Sanitation Districts. The basic purpose of the plant
is to reclaim water for groundwater recharge; the entire effluent of the plant is discharged to
spreading basins for recharging groundwater aquifers.
Design data for the plant are summarized in Table 4-3.41 j^g piant operates at either a
constant flow rate or a constant oxygen demand load by pumping wastewater from a trunk
sewer and returning grit, skimmings, primary and waste activated sludge back to the trunk.
No solids handling facilities are provided as the solids returned to the trunk are processed at
4-30
-------
a downstream primary treatment plant. The plant was constructed in 1961 at a cost of
$1,700,000; this cost includes influent pumping, foam fractionation and effluent pipelines
in addition to those items shown in Table 4-3.
Recently, the plant has been operated in a manner promoting nitrification. The three
aeration tanks are operated in a 3 pass series configuration; two-thirds of the primary
effluent is added along the first pass, with the head end of the first pass operating as sludge
reaeration. One-third of the primary effluent is added to the second pass. The plant
TABLE 4-3
DESIGN DATA
WHITTIER NARROWS WATER RECLAMATION PLANT
Plant Flow
Raw Wastewater Loadings
Biochemical Oxygen Demand (BOD)
Suspended Solids '(SS)
Primary Sedimentation Tanks
Number
Overflow Eate
Detention Time
BOD Eemoval
SS Removal
Air Blowers
Number
Discharge Pressure
Capacity - Total
Aeration Tanks
Number
Detention Time (@ 12 mgd)
BODs loading
Final Sedimentation Tanks
Number
Overflow Rate (@ 12 mgd)
Detention Time (@ 12 mgd + 33% return)
Weir Rate (@ 12 mgd)
Chlorine Contact Chambers
Number
Detention Time (@ 12 mgd) including time in Foam
Fractionation Tank & Effluent Pipe
Chlorine
12 mgd (0.53 m3/sec)
270 mg/1
280 mg/1
2 (1 stand-by)
2000 gpd/ft2 (82 m3/m2/day)
1.1 hr
35%
60%
6.5 psig (0.46 kgf/cm2)
29,500 cfm (840 m3/min)
3
6.0 hrs
45 lb/1000 cf/day)
(0.18 kg/m3/day)
790 gpd/ft2 (32.2 m3/m2/day)
1.7 hrs
12,000 gpd/ft (150 m3/m/day)
43 min
600 Ib/day (272 .kg/day)
4-31
-------
performance reflects its very careful control and operation; operating data for a one-year
period are summarized in Table 4-4." While organic nitrogen data are not available, the data
indicate that year-round complete nitrification has been obtained. Climatic conditions for
this California treatment plant are very favorable for nitrification as average monthly
wastewater temperatures did not fall, below 21 C for the year examined.
4.3.8.2 Step Aeration Activated Sludge in a Rigorous Climate
The Flint, Michigan sewage treatment plant is being upgraded to comply with requirements
of the Michigan Water Resources Commission which mandate nitrification for the purpose
of preventing DO depletion in the Flint River. In connection with this upgrading, a large
scale test of combined carbon oxidation-nitrification was conducted with the existing
activated sludge plant over a ten-month period to determine design conditions for the plant
upgrading. The minimum wastewater temperature tested was 7 C.6 During the test, ferric
chloride and polymer were added to the primary treatment stage for phosphorus removal.
This also had the effect of reducing the organic loading to the aeration tank.
The existing plant had three aeration tanks, each with four passes providing a 750,000 cu ft
(21,240 cu m) capacity. With an average design BOD5 loading of 24,500 Ib/day (11,110
kg/day) to the aeration tanks at a 20 mgd (75,700 cu m/day) flow, the aeration tank load
was 32.7 lb/1000 cu ft/day (523 kg/1000 cu m/day). Flows were varied to the facility,
however, to provide variation in loads. Three secondary sedimentation tanks were provided
having a design overflow rate of 678 gal/sq ft/day (27.6m^/m2/day) at ADWF. The plant
was usually operated in a step aeration mode, with one-half the influent directed to the head
ends of the second and third passes.
Average effluent qualities for eight months of the test are shown in Table 4-5. While nitrate
and nitrite are not shown, it is reported that a relatively good balance between ammonia
removal and nitrate production was obtained.6 Nitrite nitrogen was always less than 0.1
mg/1. The appearance of high concentrations of organic nitrogen was attributed to the low
rate of hydrolysis of organic nitrogen compounds.^ It is probable that the provision of
feeding wastewater to the last pass exacerbated the problem by causing insufficient contact
time for that portion of the wastewater to complete the hydrolysis of organic nitrogen to
ammonia.
The effect of temperature and solids retention are considered in Table 4-6. Effluent qualities
deteriorated somewhat with colder temperatures, with only 75 percent ammonia removal
being obtained at IOC. This ammonia bleedthrough may have been due to diurnal peaking
in ammonia at the relatively low solids retention time employed (c.f. Section 4.3.3.2).
4.3.8.3 Conventional Activated Sludge In a Rigorous Climate
The Jackson, Michigan wastewater treatment plant is a 17 mgd conventional activated sludge
plant designed for year-round complete nitrification.^ The existing plant was upgraded in
4-32
-------
TABLE 4-4
NITRIFICATION PERFORMANCE AT THE WHITTIER NARROWS
WATER RECLAMATION PLANT (REFERENCE 9)
Calen-
dar
Month
4
5
6
7
8
9
10
11.
12
1
2
3
Year
1973
1973
1973
1973
1973
1973
1973
1973
1973
1974
1974
1974
Flow,
med
10.4
(0.46)
11.7
(0. 51)
11.9
(0.52)
11.9
(0.52)
12.8
(0. 56)
13.4
(0.59)
13.5
(0. 59)
13.2
(0.58)
11.1
(0.49)
9.9
(0.43)
12.1
(0.53)
12.2
(0.54)
Recycle
ratio
0.39
0.40
0.45
0.46
0.45
0.43
0.41
0.45
0.52
0.60
0.49
0.48
Temp. ,
C
22
24
26
26
27
25
25
24
22
21
21
22
MLVSS,
mg/l
1st pass
2177
2337
2390
2603
2889
2850
2958
2791
2724
2675
2857
2888
3rd pass
1474
1778
2319
8092
2005
1938
2094
2000
1986
1971
2097
1982
SVI,
ml/g
78
64
66
78
77
64
77
62
55
114
88
75
days
13.1
15.4
17.2
37.5
33.4
40.2
20.5
9.4
10.7
16.1
9.8
9.4
HT?
hours
7.0
6.2
6.1
6.1
5.7
5.4
5.4
5.5
6.6
7.4
6.0
6.0
Air
Use,
MCF/day
29.3
(9600)
27.9
(9145)
29.3
(9600)
28.1
(9400)
27.5
(9010)
27.2
(8920)
27.6
(9050)
27.3
(8950)
25.5
(8360)
25.6
(8400)
28.9
(9470)
28.8
(9440)
COD,
mg/l
Primary
effluent
241
243
239
227
223
216
228
235
233
227
242
229
Secondary
effluent
62
47
39
32
30
34
34
33
35
27
31
41
Ammonia-N,
mg/l
Primary
effluent
21.6
23.5
20.8
20.1
18.8
19.2
21.5
21.9
21.8
21.4
22.1
21.2
Secondary
effluent
2.0
3.2
0.8
0.6
0.6
1.4
1.5
2.8b
1.9
0.4
0.9
1.0
Percent
Bemoval
91
86
96
97
97
93
93
87
91
98
96
95
u>
UJ
Based on influent flow and entire aeration tank
Blower trouble this month
-------
1973 in response to an order to improve treatment to a point where a minimum dissolved
oxygen of 4.0 mg/1 could be maintained in the Grand River. An analysis of the assimilative
capacity of the reach indicated that this could only be done if the effluent were completely
nitrified to prevent discharge of NOD to the river.
TABLE 4-5
AVERAGE NITRIFICATION PERFORMANCE AT
FLINT, MICHIGAN FOR 8 MONTHS (REFERENCE 6)
Parameter
(all values in mg/1
except Temp. )
BOD5
SS
Total Kjeldahl nitrogen
Organic nitrogen
Ammonia nitrogen
Phosphorus
Temp. , C
Raw
wastewater
250
300
27.6
13. 3
14.3
15.4
7.2 to 18.3
Settled
wastewater
131
140
23.3
9.9
13.4
2.7
Secondary
effluent
13.6
24.1
7.8
6.1
1.7
2.3
TABLE 4-6
EFFECT OF TEMPERATURE AND SOLIDS RETENTION TIME
ON NITRIFICATION EFFICIENCY AT FLINT, MICHIGAN (REFERENCE 6)
Temperature,
C
18 and greater
13
10
7
Solids retention
time, days
4
4-5
6
10 - 12
NH. removal,
percent
95
87
75
50 (Lab) E
Based on bench scale test results
4-34
-------
Pilot studies 10 indicated that a combined carbon oxidation-nitrification system was more
economical than a two-stage activated sludge system. Design data for this plant are contained
in Section 9.5.1.1. The plant is operated in the conventional (or plug flow) mode. Table 4-7
summarizes the plant operating data; since start-up the plant has obtained complete
nitrification at daily temperatures as low as 8 C.42 A characteristic of this wastewater is that
the primary effluent is weak, both in terms of BOD5 and ammonia -N. The mixed liquor
concentrates very well due to the high inert concentration of the raw wastewater, allowing
high mixed liquor levels under aeration. Both the weak wastewater and the ability to
maintain the mixed liquor at a high concentration allow nitrification to be obtained in
hydraulic retention times of less than eight hours even at temperatures of 8-10 C. Nitrate
and organic nitrogen values are not available.
This case history clearly demonstrates that combined carbon oxidation-nitrification can be
dependably accomplished at temperatures as low as 10 C.
4.4 Combined Carbon Oxidation-Nitrification In Attached Growth Reactors
The two attached growth reactor systems seeing application for combined carbon
oxidation-nitrification in the U.S. are the trickling filter process and the rotating biological
disc process. Procedures for designing nitrification with these two systems are described in
this section.
4.4.1 Nitrification with Trickling Filters in Combined Carbon Oxidation-Nitrification
Applications
Trickling filter design concepts are discussed extensively in the Technology Transfer
publication, Process Design Manual for Upgrading Existing Wastewater Treatment Plants.^
Therefore, the following discussion is limited to the loading ranges that are applicable for
nitrification in trickling filters used in combined carbon oxidation-nitrification applications.
As is the case for the activated sludge system, the development and maintenace of nitrifying
organisms in a trickling filter is dependent on a variety of factors including organic loading,
temperature, pH, dissolved oxygen and the presence of toxicants. However, in the case of
the trickling filter, there has been no comparable development of kinetic theory for
combined carbon oxidation-nitrification that can be directly applied with any degree of
confidence. The approach applied to date has largely been empirical and relied mostly on
specification of an organic loading rate suitable for application to each media type.21
4.4.1.1 Media Selection
The types of media currently available are summarized in Table 4-8. Rock applications are
generally limited to four to ten feet in depth; the plastic and redwood media may be built in
towers of 15 to 25 ft in height due to their lighter weight and greater void space for
ventilation, affording considerable space saving economies. Loading capabilities of trickling
4-35
-------
TABLE 4-7
NITRIFICATION PERFORMANCE AT THE JACKSON, MICHIGAN
WASTEWATER TREATMENT PLANT (REFERENCE 42)
Month
8
9
10
11
12
1
2
3
Year
1973
1973
1973
1973
1973
1974
1974
1974
Flow,
mgd
(m3/sec)
14.5
(0.63)
12.4
(0. 54)
13.2
(0.58)
12.2
(0. 53)
11.4
(0. 50)
14.0
(0.61)
14.2
(0.62)
18.4
(0.81)
Recycle
ratio
.38
.42
.38
.40
.42
.43
.49
.43
Temp. ,
G
21.7
20.0
17.2
15.6
12.2
11.1
10.6
11.1
MLSS,
mg/1
4320
4110
4390
4480
4560
4630
4800
4930
SVI,
ml/g
42
45
47
47
45
43
42
38
Sc,
days
15.6
16.4
16.7
18.6
16.4
10.3
11.0
11.1
HT*
hours
7.5
8.8
8.3
9.0
9.9
7.9
8.2
6.3
Air
use,
MCF/dayb
14
14
14
14
14
14
14
14
BOD5,
mg/1
Primary
effluent
75
82
84
94
85
97
104
90
Secondary
effluent
2.5
2.6
3
4
3
4
5
4
Ammonia-N,
mg/1
Primary
effluent
8.4
9.9
11.6
11.0
11.5
9.2
9.3
7.1
Secondary
effluent
0.6
0.7
1.2
0.8
0.6
0.7
0.6
0.5
Percent
Removal
93
93
90
93
95
92
94
93
o\
Based on influent flow
14 MCF = 4590 I/sec
-------
filter media are known to be related to the available surface area for biological slime growth.
Specific surface, or the amount of media surface contained in a unit volume, is a gross
measure of the available surface for growth of organisms. Plastic media is available in higher
specific surfaces than that shown in Table 4-8. Design practice has been to avoid the use of
media with higher specific surface and lower voids due to the danger of clogging in
combined carbon oxidation-nitrification applications. However, there has been recent
experience which indicates that medias with specific surfaces exceeding 35 sq ft/cu ft (115
m2/m3) have t,een use(} to treat domestic wastewaters without media clogging.43
TABLE 4-8
COMPARATIVE PHYSICAL PROPERTIES OF TRICKLING FILTER MEDIA
a
Media
Plasticb
Redwood*^
Granite
Granite
Blast Furnace Slag
Nominal
Size
(cm)
24 x 24 x 48
(61 x 61 x 122)
47-1/2 x 47-1/2 x 35-3/4
(121 x 121 x 51)
1-3
(2.5 x 6.5)
4
(10.2)
2-3
(5.1 x 7.6)
Unit
Weight
Ib/cu ft
(kg/mj)
2-6
(32-96)
10.3
(165)
90
(1440)
_
68
(1090)
Specific
Surface
Area
sq ft/cu ft
m2/m3
25-35°
(82-115)
14
(46)
19
(62)
13
(47)
20
(67)
Void Space
percent
94-97
76
46
60
49
Reference 25
Currently manufactured in the U.S. by: the Envirotech Corp., Brisbane, Ca.;
the Munters Corp., Fort Meyers, Fla., and the B.F. Goodrich Co., Marietta,
Ohio
.Denser media may be used for separate stage applications, see Section 4.7.1.1
Currently manufactured in the U.S. by Neptune-Microfloc, Corvalis, Or.
4.4.1.2 Organic Loading Criteria
Observations of the effect of organic loading on nitrification efficiency in rock media and
trickling filters are summarized in Figure 4-7. The data are from the following full-scale and
pilot-scale plants: Lakefield, Minn.,25 Allentown, Pa.,25 Gainesville, Fla.,44,45 Corvallis,
Or.,4^ Fitchburg, Mass.,25 Ft. Benjamin Harrison, Ind.,25 Johannesburg, South Africa,4^
and Salford, England.4** Several interesting factors affecting design are evident. First,
organic loading significantly affects nitrification efficiency. This is principally caused by the
fact that the bacterial film in the rock becomes dominated by heterotrophic bacteria. The
relative high bacterial yield when BOD is removed causes displacement of the nitrifiers from
the film by heterotrophic organisms at high organic loadings.
As opposed to nitrification with activated sludge, the breakthrough of ammonia in a
trickling filter is not abruptly affected by loading rate. For rock media, attainment of 75
4-37
-------
percent nitrification or better requires the organic loading to be limited to 10-12 Ib
BOD5/1000 cu ft/day (0.16 to 0.19 kg/m^/day). At higher organic loading rates the degree
of nitrification diminishes, such that above 30 to 40 Ib BODs/lOOO cu ft/day (0.48 to 0.64
kg/m^/day) very little nitrification occurs. These findings are consistent with those of the
National Research Council whose evaluation of World War II military installations indicates
that the organic loading should not exceed 12 Ib BODs/lOOO cu ft/day (0.19 kg/m3/day)
for rock media. 49
The partial nitrification occurring at intermediate loading rates can cause confusion when
attempting to analyze organic carbon removals across trickling filters with the BOD5 test.
Samples from the effluent of these partially nitrifying trickling filters will contain a
FIGURE 4-7
EFFECT OF ORGANIC LOAD ON NITRIFICATION EFFICIENCY
OF ROCK TRICKLING FILTERS
100
s:
UJ
UJ
a.
u.
i
80
iy eo
o
u.
u_
Ul
O 4O
20
i ' i ' I
m
0 NO RECIRCULATION
B RECIRCULATION
Kg/m3/day = 62,4 Ib BOD5/IOOO cu ft/day_
0
O
m
m
J)
0
10 20
BODe; LOAD -
30 40
LB/IOOO CU FT/DAY
50
60
4-38
-------
significant quantity of nitrifiers that could act as seed for promoting nitrification within the
5-day incubation period of the BOD5 test.21>50 About 1.5 mg/1 of ammonia nitrogen is
added to the dilution water in the BOD5 test and will also be nitrified. This will result in
unexpectedly high oxygen demands. If the BOD5 test is to measure organics in effluents
from trickling filters, then nitrification must be suppressed. The same is true for activated
sludge, but to a lesser degree as partial nitrification is less prevalent.
As opposed to the relatively efficient removal of ammonia in trickling filters, it appears that
reductions in organic nitrogen are variable and range between 20 and 80 percent. (Table
4-9). Organic nitrogen reduction can be obtained through employing effluent filtration to
remove particulate organic nitrogen. However, all treatment systems are limited to about 1
to 2 mg/1 of soluble organic nitrogen, contained in refractory organics, and therefore there
are limits to the improvements that can be obtained with effluent filtration.
4.4.1.3 Effect of Media Type on Allowable Organic Loading
The specific surface of media selected has a substantial effect on the allowable organic
loading rate for trickling filters. Greater specific surface in the media allows greater
biological film development and therefore a greater concentration of organisms within a unit
volume. Therefore, the organic loading may be higher in cases where the specific surface of
TABLE 4-9
ORGANIC NITROGEN REDUCTIONS IN NITRIFYING TRICKLING FILTERS
Facility
Location
Gainesville, Fla.a
(pilot)
Johannesburg, S.A.
(full-.scale)
Stockton, Ca.c
(pilot)
Organic
load.
Ib BOD /1 000 cu ft/day
(kg BOI§5/ m3/day)
31.5
(0.50)
31.2
(0.50
19.6
(0.31)
13.1
(0.21)
10.5
(0.17)
6.8
(0.11)
14
(0.22)
22
(0.35)
Depth,
ft
(m)
6
(1.8)
12
(3 . 7)
6
(1.8)
12
(3.7)
21.5
(6.6)
Media
1-1/2 - 2-1/2 in.
(3.8r-c|5.4cm)
2 in.
(5roc$
1-1/2 in.
(3 . 8 cm)
rock
2-3 in.
(5.1ro-c£.6cm)
plastic
27 sf/cu ft
(86 m /in3)
Influent
organic-N,
mg/1
16.6
. 9.8
6.3
8.2
9.9
13.9
11.3
11.4
Effluent
organic-N,
mg/1
7.3
4.7
2.5
3.6
2.2
2.1
8.9
9.0
Organic-N
removal ,
percent
56
52
60
56
78
85
21
21
a
References 44, 45
Reference 47
References 21/51
4-39
-------
the media is increased over that of rock media. An example is the work of Stenquist, et
al. 21 who showed that plastic media (27 sq ft/cu ft) could be loaded at about 25 lb/1000 cu
ft/day and still achieve good nitrification (Table 4-10). The higher allowable loading
attributable to plastic trickling filters was attributed to be at least partly due to the greater
specific surface of plastic media when compared to rock media. Another factor favoring
greater capacity of the plastic media filters may be oxygen supply. Rock filters often have
poor ventilation, particularly when water and air temperatures are close or the same.
4.4.1.4 Effect of Recirculation on Nitrification
The beneficial effects of recirculation on enhancing nitrification in trickling filters is evident
in the data for Salford, England in Table 4-11. The imposition of a 1:1 recycle ratio
consistently improved ammonia removals compared to when no recirculation was the rule.
Trickling filter plants designed for nitrification should incorporate provision for recir-
culation.
The minimum hydraulic application rate for plastic media trickling filters is in the range of
0.5 to 1.0 gpm/sf (0.020 to 0.041 m3/m2/min.). This minimum rate must be supplied to
ensure uniform wetting of the media. Without recirculation, nitrification design loadings
may result in applied hydraulic loads lower than the minimum hydraulic application rate.
Recirculation provides the means for preventing drying out of portions of the media by
ensuring that at least the minimum hydraulic application rate is applied at all times.
4.4.1.5 Effect of Temperature on Nitrification
Available data for nitrification are largely for warm liquid temperatures, and the practical
effects of reduced temperatures (< 20C) on allowable organic loads for combined carbon
oxidation-nitrification applications are not known at this time. However, the kinetic rate
data given in Section 3.2.5.4 would indicate that organic loads would have to be reduced
below those shown in Figure 4-7 for cold weather operation. This reduction in organic load
TABLE 4-10
LOADING CRITERIA FOR NITRIFICATION WITH PLASTIC MEDIA AT STOCKTON
BOD
load.
lb/100£ cu ft/day
(kg/m /day)
14
(0.22)
22
(0.35)
Temp,
C
26
24
Influent
BOD ,
mg/r
155
131
Depth,.
ft
(m)
21.5
(6.6)
21.5
(6.6)
Media
plastic
(Surfpac)3
Same
Recycle
ratiob
5.5
2.25
Influent
NH4-N,
mg/1
16.5
• 17.5
Effluent
NH^-N,
mg/1
1.0
2.0
Percent
nitrification
(or ammonia
removal)
94
89
Reference
21,51
21,51
327 sq ft/cu ft (86 m2/m3)
Recycle ratio is the ratio of recycled effluent to influent. Effluent was recycled prior to sedimentation.
4-40
-------
would reduce the loadings to such low values as to cause capital costs to be higher than
other available ammonia removal techniques, such as separate stage nitrification or a
physical chemical technique.
4.4.1.6 Effect of Diurnal Loading on Performance
While it is known that diurnal variations in nitrogen loading will cause variations in effluent
quality, no information is available which would allow quantitative guidelines to be
formulated. In cases where large peak to average flow ratios are experienced, flow
equalization before the nitrification step may be appropriate.
4.4.2 Nitrification with the Rotating Biological Disc Process in Combined Carbon
Oxidation-Nitrification Applications
The rotating biological disc (RBD) process is beginning to see use in the U.S. in combined
carbon oxidation-nitrification applications. The following discussion is abstracted from
Aritonie.52
The RBD process consists of a series of.large-diameter plastic discs, which are mounted on a
horizontal shaft and placed in a concrete tank. The discs are slowly rotated while
approximately 40 percent of the surface area is immersed in the wastewater to be treated.
TABLE 4-11
EFFECT OF RECIRCULATION ON NITRIFICATION IN ROCK TRICKLING FILTERS
AT SALFORD, ENGLAND (REFERENCE 48, 25)a
BOD5
load
lb/1000 cu ft/day
{kg/m2/day)
22.6
(0.36)
16.3
(0.26)
11.8
(0.19)
9.2
(0.15)
7.7
(0.12)
5.9
(0.095)
4.6
(0.074)
3.2
(0.051)
Influent
BOD5,
mg/1
266
235
191
239
165
192
199
206
Influent
NH4-N
mg/1
33.9
31.3
32.0
43.9
40.5
40.7
38.3
36.6
Effluent
NH^-N,
mg/1
"without
recirculation
19.7
16.9
9.7
125
11.4
5.7
2.8
0.7
with
recirculation
13.6
11.8
4.8
2.2
4.9
2.8
0.9
0.4
Percent
nitrification
without
recirculation
?2
46
70
72
72
86
93
93
with
recirculation
60
62
85
95
88
93
98
99
Media was blast slag, 8 ft (2.4 m) deep. With recirculation a 1:1 ratio was employed.
-------
Shortly after start-up , organisms present in the wastewater begin to adhere to the rotating
surfaces and grow until in about one week, the entire surface area is covered with a layer of
aerobic biomass. In rotation, the discs pick up a thin film of wastewater, which flows down
the surface of the discs and absorbs oxygen from the air. Microorganisms remove both
dissolved oxygen and organic materials from this thin film of wastewater. Shearing forces
exerted on the biomass as it passes through the wastewater strip excess growth from the
discs into the mixed liquor. The mixing action of the rotating discs keeps the sloughed solids
in suspension, and the wastewater flow carries them out of the disc sections into a
secondary clarifier for separation and disposal. The discs also serve to mix the contents of
each treatment stage. Treated wastewater and sloughed solids flow to a secondary clarifier
where the solids settle out and the effluent passes on for further treatment or disinfection.
The settled solids, which can thicken up to 4 percent solids content in the secondary
clarifier, are removed for treatment and disposal. A flow diagram for a typical application of
the RBD process is shown in Figure 4-8.
FIGURE 4-8
A TYPICAL ROTATING BIOLOGICAL DISC PROCESS
(COURTESY OF THE AUTOTROL CORP.)
Primary Treatment
Secondary Clarifier
Raw^A
Waste^
cj> Effluent
Solids Disposal
One current disc design consists of vacuum formed polyethylene sheets formed into
concentric corrugations which provide a high density of surface area. The corrugated sheets
are then welded together to form a stack of discs with approximately \1A in. (3.2 cm)
center-to-center spacing. This type of construction has a surface area density of
approximately 37 ft 2/ft 3 (121 m^/m^). A key feature of this disc design is the provision of
radial passages extending from the shaft to the outer perimeter of the discs. This assures that
wastewater, air, and stripped biomass can pass freely into and out of the disc assembly. In
the twelve-ft (3.7 m) diameter size, radial passages are provided every thirty degrees.
The disc units are normally housed to avoid temperature drops across the process, to
prevent algae growth on the disc surface, and to protect the surface from hail or rain which
can wash the slimes off. Information on disc types and on the general design of these
facilities can be obtained from the various disc manufacturers.
-------
4.4.2.1 Loading Criteria for Nitrification
As the rotating discs operate in series, organic matter is removed in the first disc stages and
subsequent disc stages are used for nitrification. This separation of function occurs without
the need for intermediate clarification. Nitrification does not commence until the bulk of
the BOD5 is oxidized. When low levels of BOD5 are reached, the disc stage is no longer
dominated by heterotrophs, and nitrification can proceed.
Antonie" has summarized the effects of process operating conditions on nitrification. Test
data from a number of locations are shown in Figure 4-9B. The degree of ammonia
oxidation is related to the hydraulic loa'ding on the rotating media as gallons per day per
unit surface of available surface area. Also important, as it affects population dynamics, is
the influent BOD strength. The data in Figure 4-9B was used to arrive at the design criteria
shown in Figure 4-9A. Two changes have been made to allow use of the relationships in
design. One is that there is a maximum ammonia nitrogen concentration for which the data
is considered valid. This concentration is generally 1/5 to 1/10 the influent BOD5
concentration. When the ammonia concentration exceeds these maximum ammonia
nitrogen concentrations on the appropriate BOD5 curve, it has been recommended that the
design curve be used which is rated for that ammonia concentration.53 The second change
made in Figure 4-9A is the identification of a region of unstable nitrification; that is, a
region of hydraulic loading where either a slight change in the hydraulic loading or influent
BOD strength could result in a displacement of the nitrifying population. It is considered
advisable to stay out of that region even, during daily peak flow conditions.53
4.4.2.2 Effect of Temperature
Investigators at Rutgers University^ found that the nitrifying capability of the discs was
relatively constant in a temperature range from 15 to 26 C. Simliar results were obtained by
Antonie53 who has found no effect of temperature above 13 C. Temperature correction
factors derived from pilot data are shown in Figure 4-10.53 These correction factors should
be used to reduce the design hydraulic loading determined in Figure 4-9A for any
waste water temperature lower than 13 C.
4.4.2.3 Effect of Diurnal Load Variations
Precise design criteria for handling load variations in RBD units have not been formulated.
The concern has been that high BOD5 concentrations would break through to the last disc
stage during peak loads and would cause displacement of nitrifiers from that stage.53 While
firm design recommendations have not been made, there are two general approaches
available. One is to arbitrarily derate the surface hydraulic loading to the disc to ensure low
BOD loadings at all times. This will result in a larger amount of rotating surface area. The
other is to install flow equalization to reduce the peak to average flow ratio. 5 3
4-43
-------
FIGURE 4-9
EFFECT OF BOD5 CONCENTRATION AND HYDRAULIC LOAD ON
NITRIFICATION IN THE RBD PROCESS (AFTER ANTONIE (53))
100
95
90
2
o
Ul
It
85
o
Q:
5
2
o
5
80
75
70
100
350 150
120 100 80
Inlet BODg Concentration', mq/t
Maximum
Ammonia Nitrogen
Concentration, mg/l
Temperature,
>t3C
A. Design Relationship
NITRIFICATION OF
PRIMARY EFFLUENT
(l gpd/sq ft = 41 l/mz/day)
0.5 1.0 1.5 2.0 2.5
HYDRAULIC LOADING, gpd/Sq.ft
3.0
3.5
4.0
95 —
9O
O
Ui
85
o
ct
^ 80
1
5
I 75
70
B. Experimental Data
_ Inlet BOD5,mg/l 250 \50
NITRIFICATION OF
PRIMARY EFFLUENT
Inlet BOD$ = 75 to 120 mg/l
O Tallahassee, Flo. 95-120 mg/l
£ Pewaukee, Wis. 75-87 mg/l
ffi Spring City, Pa. 126 mg/l
EB Bloom, Illinois
B Mansfield , Ohio
-- Washington, Pa. 120 mg/l
Inlet BOD5= \50 to 350 mg/l
• Tallahassee , Fla. ISO to 180 mg/l
A Pewaukie.Wis. 220 to 350 mg/l
X Luverne, Minn. 175 mg/l
• Spencer, Iowa 210 to 350 mg/l
I I I I
I I I
(1 gpd/sq ft = 41 l/mz/day)
Temperature,
>I3C
I 0
0.5 1.0 1.5 2.0 2.5
HYDRAULIC LOADING, gpd/sq ft
3.0
3.5
4.0
444
-------
4.5 Pretreatment for Separate Stage Nitrification
Nitrification facilities have been previously classified in Table 4-1. As can be seen from that
table, in order to obtain a separate stage nitrification process, the influent to that process
must be pretreated to remove organic carbon. Further, the pretreatment must achieve a
degree of carbon removal greater than is obtained by primary treatment alone, in order to
reduce the BOD5/TKN ratio to a sufficiently low level to ensure a significant fraction of
nitrifiers in the biomass. Alternatives listed in Table 4-1 include chemical treatment in the
primary, activated sludge, roughing filters, and trickling filters. Not listed, but possible, is
activated carbon treatment in conjunction with primary chemical addition. These
alternatives are summarized in Figure 4-11. This set of pretreatment alternatives is not
meant to be exhaustive, but merely illustrative; other flowsheets are possible. Detailed
design of pretreatment steps is beyond the scope of this manual; however, design procedures
may be found in the publications referenced in Figure 4-11 and the sources cited on Table
4-1.
The pretreatment alternative adopted may have significant effects on the downstream
nitrification stage. In this section, some of these possible effects are considered.
FIGURE 4-10
TEMPERATURE CORRECTION FACTOR FOR NITRIFICATION
IN THE RBD PROCESS (AFTER ANTONIE (53))
it
o
o
tt
Ul
0.
%
Uj
K
3.0
2.5
2.0
1.5
1.0
\
\
I
85N AMMONIA NITROGEN REMOVAL, PERCENT
_90t
95,
99^
I
I
I
I
6 7 8 9 IO
WASTEWATER TEMPERATURE , C
4-45
12
13
-------
4.5.1 Effects of Pretreatment by Chemical Addition
The chemical treatment step may cause significant changes in alkalinity and pH in the
downstream nitrification stage. Several alternate chemicals are available and their results
differ.
FIGURE 4-11
PRETREATMENT ALTERNATIVES FOR SEPARATE STAGE NITRIFICATION
ALTERNATE
CHE
r
RAW *
WASTEWATER
AC
RAW
WASTEWATER
RO
RAW
WASTEWATER
TR
RAW
WASTEWATER
PHI
r
RAW \^
WASTEWATER
:MICAL ADC
EMICAL
PRIMARY
CLARIFIER
PRETREATMENT TECHNIQUES
HTION IN PRIMARY
SEPARATE
NITRIFICATION
(SG or AG)
TIVATEO SLUDGE .. ..
PRIMARY
(OPTIONAL)
ACTIVATED ^/INTERMEDIATE \ SEPARATE
*" SLUDGE — H CLAHIFIER — S 1 AOt ••
y c"-ARIFIERy NITRIFICATION
V^^^/ (SG or AG)
JGHING FILTER
.x— X. /^N.
PRIMARY
CLARIFIER
/ROUGHING \ AERATION AND /WITH IFICATIOW\
*l TniCICLINC V-» NITRIFICATION ...«. IFICATION
VF!LTER J TANK 1 CLARIF.ER 1 »
^ S (SG) \ /
(or RBD Process
CKLING FILTER
PRIMARY
J TRICKLING \ /INTERMEDIATED SE!*"*J'E fc
~^ F.LTER J~*\^ CLAR.FIER^ " N|TRrFICAT,ON
\ / \ / (S3 or AG)
(or RBO Process)
rSICAL-CHEMICAL TREATMENT
EMICAL
PRIMARY
CLARIFIER
MULTIMEDIA ACTIVATED SEPARATE
FILTRATION ADSORPTION NITRIFICATION
(SG or AG)
BOD5 Removal
Prior to
Nitrification
50 to 75
60 to 95
45 to 60
60 to 90
80 to 90
References
25,55
25,36
25,36
25,36
25,55,
56
KEY
SG = Suspended growth
AG = Attached growth
4-46
-------
When alum or ferric chloride is used in the primary treatment stage, both the carbonate and
phosphate components of the alkalinity in wastewater are changed. Table 4-12 summarizes
the changes occurring when alum is added to wastewater. Effects of iron addition are
similar. Since orthophosphate is present in wastewater as HPO4 and H2PO4 between pH 4.5
and 9.3,57 fable 4-12 shows reactions with aluminum for both forms. The compound
HPO4 is measured as part of the alkalinity in wastewaters because of the following
equilibria:
H+ + HPO4 — H2PC>4
This reaction is shifted completely to the right at pH 4.5 which is approximately the
endpoint titration pH for the conversion of bicarbonate (HCC>3) to carbonic acid (H2CO3)
in the standard alkalinity determination. Table 4-12 shows that the same amount of total
alkalinity is lost regardless of the form of the inorganic phosphorus (HPO^ or
TABLE 4-12
EFFECT OF ALUM ADDITION TO WASTEWATER ON ALKALINITY
1. Hydrolysis
AU (SO.) „ +6 HCOl + 6 H0O -*• 2 Al (OH), + 3 SO"? + 6 H-CO,
£ *t O O ^ O ** 6 O
(54 as Al+3) (300 alkalinity
as CaCo3) +g
or 5.6 mg of alkalinity as CaCOg lost per mg Al added
2. Precipitation of inorganic phosphorus
form
A12 (SO4) 3 +2 HCO~ + 2 HPO4 — >- 2
(100 alkalinity
(54 as Al+3) (100 alkalinity as CaCO3)
as CaCOg) (62 as P)
+3
or 3:7 mg of alkalinity as CaCO, lost per mg Al added
O
O
and . 87 mg Al required per mg P
H2PO4
A12 (S04) 3 + 4 ECO' + 2 H2PO4 — *- 2 AlPO4+3 80= + 4
(54 as Al+3) (200 alkalinity (0 alkalinity
as CaCO ) as CaCO3)
(62 as P) (62 as P)
or 3. 7 mg of alkalinity as CaCOg per mg Al+a added
and 0. 87 mg of Al required per mg P
4-47
-------
As an illustrative example of the results of alum addition, consider a case where 10 mg/1 of
inorganic phosphorus is precipitated in the primary stage of treatment. Using a value of 0.64
mg P removed per mg of aluminum added results in the following aluminum use: 1
10mg/1 „ = 15.5 mg/1 Al' " required (total)
0.64 mgj
mgAT
The aluminum used in phosphorus precipitation is (Table 4-12):
10 mg/1 P(0.87 mgA1 ) = 8.7 mg/1 AT" for precipitation
mgP
By difference the aluminum used in hydrolysis is:
i I I
15.5 - 8.7 = 6.8 mg/1 Al .for hydrolysis
The alkalinity loss due to hydrolysis is (Table 4-12):
(5.6mgCaCO~)
6.8 :rrr^- = 37.4 as CaCO-
. ,+TT J
mg Al
The alkalinity loss due to precipitation is (Table 4-12):
(3.7 mg CaCO,)
8.7 — = 32.2 as CaCO-
+++ 3
mgAl
Thus, a total alkalinity loss of 70 mg/1 will occur when 10 mg/1 of P is removed. This loss,
depending on the initial wastewater alkalinity, may have to be made up with downstream
chemical addition to prevent adverse effects on the operating pH of nitrification (see
Section 4.9.2.).
On the other hand, lime addition to the primary may have quite the opposite effect of alum.
The changes in alkalinity occurring with lime addition are dependent on lime dose (or pH)
and the quality of the raw wastewater. Cases of both increases and decreases in alkalinity
with lime treatment may be found.^5
Lime addition may also cause elevation of the operating pH of the nitrification stage.3
Primary effluent after lime treatment will typically have a pH between 9.5 and 11.0,
depending on lime dose and treatment requirements. 55 This pH is higher than can normally
be discharged or introduced into downstream treatment units. To reduce the pH, normal
4-48
-------
practice is to "recarbonate" the high pH primary effluent. Conventionally, this involves the
introduction of gaseous carbon dioxide (CC>2) into the high pH primary effluent in a
reaction basin of at least 20 minutes detention time. Typically the carbon dioxide is either
drawn from refrigerated storage or furnace stack gases containing carbon dioxide are used.
The recarbonation step can be thought of as the conversion of alkalinity in the hydroxide
form (OH~) to that in the bicarbonate form (HCOJ) as follows: 55
Ca(OH)2 + 2CO2 +• Ca(HCO3)2
In many cases there is sufficient carbon dioxide produced from the oxidation of organic
carbon and from nitrification to completely satisfy recarbonation requirements. In a lime
precipitation-nitrification sequence in California^ it was found that external carbon dioxide
requirements were minimal and only occasionally required. In this case, it was calculated
that approximately two-thirds of the carbon dioxide produced was derived from
nitrification, while the remaining one-third was derived from the oxidation of the organic
carbon remaining in the primary effluent. When the same process was tried at the EPA-DC
pilot plant at Blue Plains, it was found that supplemental carbon dioxide was continuously
required to maintain a neutral pH. This is because wastewater in the Washington area is
weaker than in the California case. There is less production of carbon dioxide because there
are lower concentrations of oxidizable substances entering the nitrification stage at Blue
Plains.
The tendency is for the high pH effluent of the lime primary process to elevate the pH in
the nitrification reactor. Often this effect enhances nitrification rates (see Section 3.2.5.6.).
When lime primary treatment is employed, care must be taken not to mix the primary
effluent with the return activated sludge prior to entry into the nitrification tank. The high
pH of the primary effluent return activated sludge mixture would be toxic to both the
nitrifiers and heterotrophic bacteria in the return sludge.
A concern often expressed is that the phosphorus removal obtained in a chemical primary
treatment step will be so great as to starve the downstream nitrifying biomass for
phosphorus as a nutrient for growth. Actually, the requirements for phosphorus in a
separate stage nitrification system are very low. Typically, organism biomass contains about
2.6 percent phosphorus. This number can be used to calculate phosphorus requirements, as
in the following example. Assume a case where 36 mg/1 of TKN are nitrified and 60 mg/1 of
BOD5 are removed in a separate nitrification stage. The quantity of biomass grown can be
conservatively estimated as follows: (see Section 3.2.7)
Nitrifiers: 0.15(36) = 5.4 mg/1 VSS
Heterotrophs: 0.55(60) = 33.0 mg/1 VSS
Total biomass: 38.4 mg/1 VSS
4-49
-------
The phosphorus required is 2.6 percent of the VSS grown:
0.026(38.4)= 1.0 mg/1 as P
Thus, the phosphorus requirement in this example is only 1.0 mg/1 P. In most cases, the
BOD5 and TKN levels will be less than given in the example, so that the 1 mg/1 requirement
can be considered a maximum requirement for separate stage nitrification.
It is a relatively simple matter to manipulate chemical doses in the primary so that
phosphorus residuals are sufficient to support biological growth. 3 It may be that low levels
of phosphorus or some trace micronutrient removed by the primary may cause "bulking" in
a suspended growth type nitrification stage. 58 Very high SVI measurements were observed
in the nitrification stage following lime treatment at the Central Contra Costa Sanitary
District's Advanced Treatment Test Facility.^ However, it has been found that this bulking
could be very effectively controlled by a continuous low dose of chlorine to the return
sludge, without impairment of nitrification efficiency. A dose of 2 to 3 Ib. per 1,000 Ib.
MLVSS per day (2 to 3 g/kg/day) in the inventory reduced the SVI from 160-200 to 45-80
ml/gram. Overdosing at 5 to 7 Ib./1,000 Ib./day (5 to 7 g/kg/day) caused impairment of
nitrification and higher effluent solids.59
4.5.2 Effects of Degree of Organic Carbon Removal
The pretreatment alternatives arrayed in Figure 4-11 provide varying degrees of organic
carbon removal ahead of the nitrification step. Not only is there variation among the
alternatives, but there is possible range of intermediate effluent qualities within each
alternative. It was shown in Chapter 3 that high degrees of organic carbon removals led to
the highest nitrification rates. This implies that reactor requirements diminish with
increasing degrees of carbon removal in the pretreatment stage.
Very low levels of organic carbon in the nitrification process influent has differing
implications for suspended growth and attached growth nitrification reactors. The effluent
solids from the sedimentation step in a suspended growth system can exceed the solids
synthesized in the process when the level of organics is low. This leads to a requirement of
continuously wasting solids from other suspended growth stages (carbon removal or
denitrification) to the nitrification stage or to increase the BOD of the influent in order to
maintain the inventory of biological solids in the system. This point is discussed further in
Section 4.6.4.
Low levels of organics in the influent to attached growth reactors can be advantageous. The
synthesis of solids occurring with low levels of influent organics results in very low levels of
solids in the effluent from an attached growth reactor. In some cases this can eliminate the
need for a clarification step, especially if multimedia filtration follows or if there is some
other downstream treatment unit such as denitrification.
4-50
-------
4.5.3 Protection Against Toxicants
All of the pretreatment alternatives portrayed in Figure 4-11 provide a degree of removal of
the toxicants present in raw wastewater. However, the types of toxicants removed by each
pretreatment stage vary among the alternatives. Chemical primary treatment can be used
where toxicity from heavy metals is the major problem. Lime primary treatment is one of
the most effective processes for removal of a wide range of metals. 5 5 Chemical treatment is
usually not effective for removal of organic toxicants, unless it is coupled with a carbon
adsorption step such as in the physical-chemical treatment sequence. The biological
pretreatment alternatives (activated sludge, trickling filters, and roughing filters) provide a
degree of protection against both organic and heavy metal toxicants. An exception would be
organics resistant to biological oxidation, such as the solvents perchloroethylene and
trichloroethylene which have been identified as toxicants which can upset nitrification. 1 >60
When materials toxic to nitrifiers are present in the influent raw wastewater on a regular
basis, the pretreatment technique most suitable for their removal can be used in the plant
design to safeguard the nitrifying population. The determination of the most suitable system
configuration need not involve an extensive sampling or elaborate pilot program. A recently
developed bench top analysis can be used to screen alternates. 61 The test procedure involves
batch oxygen uptake tests using a respirometer to measure oxygen utilization. Composite
wastewater samples are subjected 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 Allythiourea
(ATU).61 The other respirometer is inoculated with a small amount of mixed liquor from a
nitrifying activated sludge plant. 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 sample, known to contain ammonia and organics, but no toxicants.
The batch nitrification rates, determined by the batch procedure, can be examined to
determine the pretreatment technique apparently most suitable among the options
examined. Often, some of the pretreatment techniques will result in little or no nitrification
in the inoculated sample, indicating inadequate removal of the toxicant(s). In other cases,
the pretreatment techniques will allow vigorous nitrification in the sample indicating good
removal of the toxicant(s).61 The particular pretreatment technique that is effective may
also indicate the type of toxicant that is interfering with nitrification and may permit
identification and elimination of the source tributary to the system. For instance, if lime
treatment is effective, the problem may be a heavy metal that can be precipitated by the
lime. Alternatively, if biological oxidation is ineffective but activated carbon treatment
allows nitrification to proceed, then a nonbiodegradable organic is suspect. Subsequent
specific analyses can then be run in the identified category of compounds. If the toxicants
4-51
-------
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 is not capable of doing.
4.6 Separate Stage Nitrification with Suspended Growth Processes
There are many examples of separate stage nitrification processes in the U.S. (Table 4-1).
The initial development of the suspended growth process in separate stage application was
oriented to the isolation of the operation of the carbonaceous removal and nitrification
processes so that each could be separately controlled and optimized. 15 By placing a carbon
removal system (originally conceived as a high rate activated sludge system) ahead of the
separate nitrification stage, the sludge would be enriched with nitrifiers as opposed to the
marginal population present in combined carbon oxidation-nitrification systems. By this
enrichment process, nitrification would be expected to be less temperature sensitive than in
a combined carbon oxidation-nitrification system.
First applications of the process were in the northern portion of the U.S. where low ( < 10
C) liquid temperatures were obtained in the wintertime. Subsequently, the system has been
applied in moderate climates such as Florida and California where some of the other
advantages of the process have made its application desirable.
4.6.1 Application of Nitrification Kinetic Theory to Design
The kinetic approach for design of separate stage nitrification in suspended growth systems
is fundamentally related to the kinetic design approach used for combined carbon
oxidation-nitrification, as was shown in Section 3.2.7. To date, however, the practice has
been to adopt the "solids retention time" design approach for combined carbon
oxidation-nitrification applications and to use the nitrification rate approach for separate
stage design. The nitrification rate approach has been based on experimentally measured
rates, ^ >62,63 rather than attempting to relate the rates to fundamental kinetic theory (see
Section 3.2.7). The theoretically determined nitrification rates are limited in their
applicability chiefly because the nitrifier fraction of the mixed liquor cannot be accurately
assessed. Nonetheless, the concepts developed from kinetic theory are applicable even in the
absence of information about the nitrifier fraction.
The solids retention time design approach is also directly applicable to many separate stage
nitrification design problems. The limit of its applicability is related to the difficulties with
the nitrification rate approach, that is,the yield of nitrifiers grown through nitrification is
not accurately known. However, for designs where the BOD5 level in the nitrification
influent is 30 to 60 mg/1 (or BODs/TKN ratio is 2 to 3) the growth of nitrifiers will
generally be a small fraction relative to the heterotrophic population. In this case, the errors
in assumption of the yield of nitrifiers will be masked by the growth of heterotrophs. For
these cases, the contribution of the nitrifiers to the overall process growth rate may be
4-52
-------
neglected with the assurance that their contribution is less than 12 percent (Section 3.2.7).
A model which explicitly considers the nitrifier contribution to the system growth rate is
available which can be used when accurate yields for nitrification become known.™
In situations where the organic carbon is low in the influent (BOD5/TKN ratio 0.5 to 2.0)
to the nitrification stage, both the assumptions for heterotrophic and nitrifier yields become
uncertain. These low BOD5/TKN ratio situations are usually cases where a well stabilized
secondary effluent is being nitrified. Residual BOD5 in these effluents is often biological
solids rather than residual raw wastewater organic matter. The biomass yield in the
nitrification stage for these cases is less well defined. This leads to uncertainty in estimating
design sludge inventories and wasting schedules when using the solids retention time design
approach.
In this section, both the solids retention time approach and the nitrification rate approach
are presented. The direct theoretical interrelationships of the approaches, as developed in
Section 3.2.7, should not be overlooked.
4.6.2 Solids Retention Time Approach
The design procedures developed in Section 4.3 are directly applicable to separate stage
nitrification design when the BOD5/TKN ratio equals or exceeds 2.0. A summary of these
design concepts follows.
4.6.2.1 Choice of Process Configuration
In general, the favored system for separate stage nitrification is the plug flow system. It was
already shown in Section 4.3.5 that the plug flow process results in lower effluent ammonia
than a complete mix process at the same SF, or alternately, the same ammonia level at a
lower SF. The only disadvantage of the process in combined carbon oxidation-nitrification
applications is the difficulty in supplying adequate DO in the head end of the system,
rendering that zone ineffective for nitrification in cases of low DO. With a carbon removal
step ahead of the separate nitrification stage, the high oxygen demand at the head end of
the system is minimized, and less difficulty is found in designing aeration systems that
ensure adequate DO levels throughout the nitrification tanks. With adequate DO levels
throughout the tanks, the full advantage of plug flow kinetics over the kinetics of any other
configuration is obtained.
In cases where a lime precipitation step precedes the nitrification step, it is often desirable
to use the carbon dioxide (CO2) produced by the nitrification process for recarbonation
(c.f. Section 4.5.1). Since the process of nitrification will be spread throughout the plug
flow reactor, and recarbonation is desired at its head end to avoid pH toxicity, only a
portion of the carbon dioxide produced in the process is available for recarbonation at the
head end of the process. A solution to this problem is to increase the use of external carbon
dioxide. On the other hand, external carbon dioxide usage may be minimized by adopting
4-53
-------
one of the other process configurations, such as complete mix, that more evenly spread out
the load throughout the aeration tank, thereby taking full advantage of the in-process
carbon dioxide generation for recarbonation.
4.6.2.2 Choice of the Safety Factor
The same considerations discussed in Section 4.3.3.2 are applicable to separate sludge
applications. Diurnal variations in nitrogen load will effect effluent quality in the same
manner as for combined carbon oxidation-nitrification applications. However, upstream
treatment steps may moderate the fluctuations in nitrogen load experienced by separate
stage nitrification processes. While primary treatment and roughing filters have so little
liquid holdup that little or no nitrogen load equalization is provided, this is not true of
either the activated sludge or trickling filter pretreatment alternatives. Both of these
alternatives provide a degree of nitrogen load equalization.
A typical ammonia load curve for the secondary effluent from the Rancho Cordova,
California treatment plant is shown in Figure 4-12. The plant provides activated sludge
treatment. On the date monitored, the average flow was 1.9 mgd with a peak hourly flow of
FIGURE 4- 12
RANCHO CORDOVA WASTEWATER TREATMENT FACILITY
EFFLUENT AMMONIA CHARACTERISTICS, MARCH 19-20, 1974
50
— 5OO
5>
o
30
Uj
o
o
t>
20
to
\
CD
— 300
— 2OO
too
I— O
i i r
i i i
NH4-N MASS FLUX
PLANT FLOW
NH^-N CONCENTRATION
= .454 Kg
I I | I I
I i
oeoo
1200
I6OO 20OO
TIME OF DAY
Z4OO
III III
04OO
oaoo
4-54
-------
3.0 mgd, for a flow peaking factor of 1.6. The ammonia loading behaved similarly with an
average load of 252 Ib /day and a peak hourly load of 379 Ib /day for a nitrogen load
peaking factor of 1.6. Using the concept developed in Section 4.3.3.2, the minimum safety
factor adopted for a separate stage nitrification process serving this plant should be 1.6, to
prevent significant ammonia leakage during the peak hour.
4.6.3 Nitrification Rate Approach
The kinetic design approach using nitrification rates places reliance on experimentally
determined rates obtained from pilot studies. Available data are summarized in Figure 4-13,
plotted against temperature. Also shown are two other principal variables that can effect
nitrification rates, the BODs/TKN ratio and the mixed liquor pH.
In general, the nitrification rates follow the predictions of the theory (Section 3.2.7). As the
temperature rises, nitrification rates increase. The BOD5/TKN ratio strongly influences the
nitrification rates. Comparing the Manassas, Blue Plains and Marlborough data, it can be
seen that the lower the BOD5/TKN ratio (and the higher the nitrifier fraction) the higher
the nitrification rates. Also the effects of pH depression on nitrification rates is apparent.
Particularly interesting is the data from Blue Plains for the air and oxygen system run at
approximately the same pH, but at differing BOD5/TKN ratios. The air nitrification system,
running at a lower influent BOD5/TKN ratio exhibited higher nitrification rates than the
oxygen system running at a relatively higher BOD5/TKN ratio. It is notable that when
oxygen and air nitrification systems were run in parallel at the same pH and influent
BOD5/TKN ratios, the same nitrification rates were obtained.^ A further comparison of
oxygen and air nitrification is presented in Section 4.6.5.
In the absence of pilot "data specific to a particular situation, Figure 4-13 can be used to
approximate design nitrification rates. What must be known or estimated is the BODs/TKN
ratio in the influent, the minimum temperature for nitrification and the mixed liquor pH. In
essence, Fig. 4-13 is a plot of experimentally determined values of the peak nitrification
rate, as defined in Section 3.2.7 as follows:
* = f (3-3D
where: TN = peak nitrification rate, Ib NH4 - N oxidized/lb MLVSS/day,
f = nitrifier fraction, and
q^ = peak ammonia oxidation rate, Ib NH. - N rem/lb VSS/day.
In other words, these rates are determined at values where the DO and the ammonia level
are not limiting the rate of nitrification. However, to be useful for design purposes, the
effects of ammonia content and DO should be considered. The effect of operating DO can
4-55
-------
ON
0.7
0.6
X
Q
Q
\ 0.5
<0
t/i
^
-J
§
-Q 0.4
^
5.5
TS
IS
S=>
-V
I!
->e -~
o
\;f>-y; >^
^^^K^<:;:;^
V ^, '^\^*^i^ ^, ^< /' - ^ *\-J>
"-'- -fvi ->^.*>>x;-;-?
" ' *' ^*C>^V"'-""'"'''*-'''"''""°v
Marlborough, Mass.
(ref. 62) BOD5/TKN = 3.
+
/
Key to Individual Data Points
• South Bend, Indiana
(ref. 12), pH 7.5, BOD5/TKN=t.8
• Denver, Colorado
(ref. 19,20), BOD5/TKN = 2.7
-Blue Plains, pH 6.8 to 7.2 (Air System)
(ref. 7 ) BOD5/TKN = 1.3
10 II 12 13 14 15 16 • 17 18 19 20 21
TEMPERATURE, C
Blue Plains, pH 7.0 (Oxygen System)
( ref. 8) BOD5/TKN = 3.0
+
22 23 24 25 26 27 28 29 30
-------
be incorporated through the Monod expression for DO. The effect of desired effluent
ammonia content can be considered through the safety factor concept. The use of
Equations 3-31, 3-20, and 3-29 yields the following expression for either a complete mix, or
plug flow system: ^ / nn \
rxi =— — 1 (4-22)
N SF I KQ2 + DO j
Effluent ammonia nitrogen content can be estimated for a complete mix system at steady
state from the equation:
i Ni
IF = TV^T (4-23)
For plug flow reactors, the effluent ammonia content can be estimated from Equation 4-14.
Criteria for establishing the safety factor are discussed in Sections 4.3.3.1, 4.3.3.2, and
4.6.2.2.
Once the design value of the nitrification rate is established, the design can proceed in a
manner similar to the F/M design approach adopted for activated sludge design. The total
nitrogen load per day and the nitrification rate are used to establish the mass of solids that
must be maintained in the nitrification reactor. The volume of reactor is determined from
the allowable mixed liquor solids level and the inventory of solids required. The allowable
mixed liquor level is influenced primarily by the efficiency of solids-liquid separation
(Section 4.10).
4.6.4 Effect of the BOD5/TKN Ratio on Sludge Inventory Control
At Manassas, Virginia, Jackson, Michigan, and Contra Costa, California, difficulty was
experienced in maintaining a nitrifying sludge inventory when the influent contained low
amounts of organics (low BODs/TKN ratios).64,10,4 Typically,effluent solids fluctuated
between 10 and 50 mg/1 and the effluents contained a good deal of dispersed solids that
were not captured in the secondary clarifier. It has been suggested that a high fraction of the
mixed liquor must be heterotrophic to maintain good bioflocculation in a separate stage
nitrification system. * *• Since the synthesis of solids in these systems is often less than the
solids appearing in the effluent from the system, an unstable condition can result. Several
remedies are available. At the three locations mentioned, solids from the upstream activated
sludge process were periodically transferred to the separate stage nitrification process to
maintain the solids inventory.
In other cases, pretreatment steps have been purposely chosen which do not provide as high
a degree of carbon removal as activated sludge and therefore cause greater synthesis of
heterotrophic biomass in the nitrification stage. Examples are lime precipitation in the
primary^ and modified aeration activated sludge with alum addition. 65 In still another case,
10 percent of the primary effluent was bypassed around the activated sludge carbon removal
step to a separate stage nitrification step. 11 The amount of primary effluent bypassed can
4-57
-------
be varied to precisely control the solids retention time and details of a recommended
procedure for accomplishing this control can be found in reference 66.
4.6.5 Comparison of the Use of Conventional Aeration to the Use of High Purity
Oxygen
A comprehensive study of the effect of pH on covered high purity oxygen nitrification
systems with comparisons to conventionally aerated systems has recently been completed.^
High purity oxygen systems typically operate at somewhat lower pH levels than
conventionally aerated systems, and it has been intimated that this will result in lower
nitrification rates and efficiency than conventionally aerated systems. The reason for the
lower pH is because of the method of oxygenation in covered high purity oxygen systems.
As shown in Figure 4-14, the system uses a covered and staged oxygenation basin for
contact of gases and mixed liquor. High purity oxygen (90+ percent purity) enters the first
stage and flows concurrently with the wastewater being treated. The gas is reused in
successive stages, resulting in the buildup of carbon dioxide released by biological activity in
the gas and in the liquid. This results in a depression of pH. While the pH is also depressed
by carbon dioxide release in conventionally aerated systems (c.f. Sec. 4.9), the pH
depression is less than occurs in the high purity oxygen system because evolved carbon
dioxide is continually stripped from the system by the aeration air.
The work at the EPA-DC Blue Plains treatment plant consisted of two carefully controlled
pilot investigations as follows: (1) separate stage nitrification with high purity oxygen with
and without pH control and (2) separate stage nitrification by conventional aeration and
FIGURE 4-14
COVERED HIGH PURITY OXYGEN REACTOR WITH
THREE STAGES AND MECHANICAL AERATORS
OXYGEN
FEED GAS
WASTE
LIQUOR
FEED
RECYCLE
SLUDGE'
EXHAUST
"GAS
MIXED LIQUOR
EFFLUENT TO
CLARIFIER
4-58
-------
high purity oxygenation controlled at the same pH level. The purpose of this section is to
review the salient features of this work as it affects nitrification design with high purity
oxygen. No attempt will be made to present general high purity oxygen design concepts as
these are covered in the EPA Office of Technology Transfer's publication, Oxygen Activated
Sludge Wastewater Treatment Systems Design Criteria and Operating Experience. 67
4.6.5.1 High Purity Oxygen Nitrification With and Without pH Control
Operational results are shown in Table 4-13 for a high purity oxygen pilot plant without pH
control and a plant with the last of four stages held at pH 7.0 by lime addition to the first
stage. As may be seen from the table, the system with pH control provided an effluent with
somewhat better quality. Lime addition to the pH controlled reactor caused buildup of
inerts which resulted in better thickening sludge and better clarification efficiency at the
expense of greater sludge production.^ The effluent ammonia level was somewhat lower
with pH control, but the difference is not significant when it is considered that a chlorine
dose as little as 10 mg/1 would be sufficient to remove all traces of ammonia in both
wastewaters (c.f. Chapter 6). It should be noted that both systems were operated at high
solids retention times and therefore had high SF values. Therefore, differences in effluent
ammonia levels would be expected to be small.
Comparisons of nitrification rates did not present a clear picture of differences between the
two systems. However, it was shown that when the pH drops below 6.0 nitrification
rates did decline. However, the pH generally did not drop below 6.0 in the system without
pH control until the last reactor stage, but there was so little ammonia remaining to be
oxidized there that essentially no effect of pH on nitrification performance could be
discerned.
In sum, the work at Blue Plains demonstrates that the pH of the nitrification reactor can
drop as low as 6.0 and allow acclimation of the nitrification organisms with attainment of
complete nitrification. The pH should not be allowed to drop below 6.0, except perhaps in
the last reactor stage, and in those cases where the carbon dioxide evolution is sufficient to
cause the pH to drop below 6.0, pH control should be implemented. From an organics
standpoint, effluent qualities are superior in the pH controlled system because of lime
addition, and this should be kept in mind when designing for stringent effluent
requirements.
4.6.5.2 Comparison of Conventional Aeration and
High Purity Oxygen at the Same pH
Table 4-14 shows the results of a parallel study of a conventional aerated nitrification
system with a high purity oxygen system; both systems were held at pH 7.0 in the last stage
of a four-stage system.^ As may be seen from the table, the concentration of organics and
nitrogen species were virtually identical in the two systems. Greater lime was required in the
high purity oxygen system than in the conventional system to maintain the same pH level.
4-59
-------
This greater lime dose resulted in greater sludge production in the oxygen system but also
improved the sludge thickening properties, allowing the same MLVSS level to be maintained
in both systems at a higher MLSS level in the oxygen system. Nitrification kinetic rates were
found to be the same in both systems. Choice between the two pH controlled systems
should be based on economic considerations as the systems are equivalent in other respects.
TABLE 4-13
COMPARISON OF PROCESS CHARACTERISTICS FOR
OXYGEN NITRIFICATION SYSTEMS WITH AND WITHOUT
pH CONTROL AT BLUE PLAINS, WASHINGTON, D.C.
With pH
control,
days 51-150a
Without pH
control,
days 26-8Sa
Operating Parameters'3
Oxygenation time, hrs°
F/M ratiod
Solids retention time, days
MLSS, mg/1
MLVSS, mg/1
Return sludge SS, mg/1
Sedimentation tank overflow rate ,
gpd/sq ft
m /m2/day
Sludge production, mg/1
Lime dose (CaO) , mg/1
4.0
0.15
13
5,660
2,620
38,650
620
25.3
69
126
3.9
0.16
17
3,520
2,780
13,640
645
26.3
35.4
0
Effluent Qualities13
Concentration, mg/1
BOD 5
COD
SS
VSS
TKN
NHj-N
NO^NOj-N
mean
67
152
77
58
21.7
14.9
-
Effluent
mean
6.3e
26.4
10
6.0
1.3
0.15
13.4
standard
deviation
3.1e
6.2
4.8
4.2
0.6
0.11
1.1
mean
70
152
75
58
22.7
150
-
Effluent
mean
8.6e
42.5
24.0
17.1
2.4
0.64
12.8
standard
deviation
3.06
8.2
10.2
8.6
0.7
0.64
12
Day 1 = January 1, 1974
Data from reference 8 ; while the data are not from the same time period, data from a common time
period (days 56-85) showed the same trends.
GBased on influent flow
F/M ratio is the ratio of the Ib BODj in the influent to the activated sludge process and the Ib of
MLVSS inventory under aeration or oxygenation.
eNitriftcation inhibited
4-60
-------
4.7 Separate Stage Nitrification with Attached Growth Processes
Three types of attached growth processes have been employed for separate stage
nitrification. The differences lie in the type of medium provided for biological growth. The
three types of processes are the trickling filter, the rotating biological disc and the packed
bed reactor.
TABLE 4-14
COMPARISON OF PROCESS CHARACTERISTICS OF CONVENTIONALLY
AERATED AND HIGH PURITY OXYGEN SYSTEMS WITH pH
CONTROL AT BLUE PLAINS, WASHINGTON, D.C.
High purity
oxygen,
days 186-2653
Conventional
diffused aeration,
days 186-265a
Operating Parameters
Oxygenatlon time, hr
F/M ratiod
Solids retention time, days
MLSS, mg/1
MLVSS, mg/1
Return sludge SS, mg/1
Sedimentation tank overflow rate,
gpd/sq ft
3 2
m /m /day
Sludge production, mg/1
Lime dose (CaO) , mg/1
3.8
0.16
10.0
6,355
2,260
40,850
656
26.7
97.6
128
3.7
0.16
9.9
3,890
2,240
16,460
678
27.6
59.2
47
Effluent Qualities
Concentration, mg/1
BOD5
COD
SS
VSS
TKN
NH*-N
NO"+ NO"-N
Influent
mean
56
130
65
49
19.5
15.2
-
Effluent
mean
3. .5°
20.6
8.1
4.8
0.94
0.19
13.6
standard
deviation
1 . 1 '
2.6
2.6
2.1
0.23
0.20
1.2
Influent
mean
56
130
65
49
19.5
15.2
-
Effluent
mean
3.4
20.6
6.2
4.2
1.0
0.19
12.5
standard
deviation
1.3e
3.4
3.4
2.5
0.3
0.16
1.2
Day 1 = January 1, 1974
b
Data from reference 8
Based on influent flow
F/M ratio is the ratio of the Ib BOD in the Influent to the activated sludge process
o
and the Ib of MLVSS Inventory under aeration or oxygenation.
eNitrlflcation inhibited
4-61
-------
4.7.1 Nitrification with Trickling Filters
The development of two-stage trickling filtration or double filtration preceded the
development of two-stage suspended growth systems for nitrification. In fact, two-stage
filtration was in operation at several military installations during World War 11.4° initially,
the two-stage trickling filtration process was developed to increase the removal of organics
in the effluents from the high rate trickling filters. Later, it was observed that under some
operating conditions, the second stage produced a well nitrified effluent.68
In separate stage nitrification application, trickling filters can follow a high rate trickling
filter plant with intermediate clarification, or an activated sludge process or any of the other
alternatives listed in Fig. 4-11.
4.7.1.1 Media Type and Specific Surface
It has been shown that in separate stage nitrification applications, the rate of nitrification is
proportional to the surface area exposed to the liquid being nitrified.69,70 In other words,
when all other factors are held constant, the allowable loading rates can be expected to be
related to the media surface area, rather than to the media volume.
Very little biological film development has been observed in separate stage applica-
tions.71>72 AS a consequence, pluggage of voids in the media and ponding becomes of less
concern than in combined carbon oxidation-nitrification applications. Media of higher
specific surface than normally employed may be used. Plastic media is characterized by
having very high specific surface available while maintaining a high void ratio ( > 90
percent). The high specific surface area of plastic media allows the trickling filter volume to
be reduced, significantly reducing the cost of the distributor arms and the structure.
Available types of plastic media are summarized in Table 4-15. Most experience in the U.S.
has been with the corrugated sheet module type, rather than with the dumped media which
has just become available. Media applicable to nitrification applications is commercially
available in specific surfaces ranging from 27 to 68 sq ft/cu ft (89 to 223 m^/m^).
4.7.1.2 Loading Criteria
As previously stated, nitrification rates in trickling filters are related to the wetted surface
area of the media. Therefore, the most rational criterion would be in terms of surface area.
Unfortunately, information on specific surface is not always available, and volumetric
loading criteria must occasionally be resorted to.
The pilot study at the Midland, Michigan wastewater treatment plant provides the most
comprehensive set of data currently available on nitrification with trickling filters.22,71 jj-ie
influent to the pilot plant was well treated trickling filter effluent with BOD5, SS and
ammonia-N values ranging from 15-20, 15-20 and 8-18 respectively. The BODs/TKN ratio
4-62
-------
TABLE 4-15
COMMERCIAL TYPES OF PLASTIC MEDIA FOR
SEPARATE STAGE NITRIFICATION APPLICATIONS
Manufacturer
Envlrotech Corp . , Brisbane ,
Ca.*
B.F. Goodrich, Marietta, Ohio
Enviro Development Co. , Inc.
Palo Alto, Ca.b
Mass Transfer, Ltd. , Houston,
Texas
Norton Co. , Akron, Ohio
Munters Corp., Ft. Meyers, Fla.
Trade
Name
Surfpac
Vinyl Core
Flocor
Filterpack
Actifil
PLASdek
Type
Corrugated shcat
.modules
Corrugated sheet
modules
Corrugated sheet
modules
Dumped rings
Dumped rings
Corrugated sheet
module s
Specific surface available
sf/cu ft (m / m )
27
30.5
45
27
40
36
57
27
42
42
68
(89)
(100)
(148)
(39)
(131)
(118)
(189)
(89)
(138)
(138)
(223)
Formerly available from the Dow Chemical Co., Midland, Mich.
Under license from ICI, Great Britain; formerly available from the Ethyl Corp.,
Baton Rouge, La.
was 1.1, indicating a high degree of BOD removal in the pretreatment stage. The pilot unit
was a 21.5 ft (6.55 m) unit filled with Surfpac media. During the 18 month project period, a
variety of climatic conditions were experienced with wastewater temperatures in the pilot
unit as low as 7 C and as high as 19 C encountered.
The data from the various operating periods for the project have been reexpressed in Figure
4-15 in terms of the surface area required for nitrification and the desired effluent ammonia
nitrogen content. As may be seen, greater surface area is required at low temperature (7 to
11 C) than high temperatures (13 to 1-9 C). Further, to obtain ammonia-N contents below
2.5 to 3.0 mg/1, greater surface area is required than for effluent ammonia contents above
2.5 to 3.0 mg/1 ammonia-N.
When effluent ammonia-N levels less than 2.5 mg/1 are desired, consideration should be
given to using breakpoint chlorination (Chapter 6) for removing ammonia residuals rather
than increasing the surface area of the filter, as the cost of removing the last 1-3 mg/1 of
ammonia becomes very high because of the very much larger trickling filters required.
Figure 4-15 was developed solely from the Midland, Michigan data. Data are available from
two other locations which allow calculation of surface requirements for nitrification.2',73
Figure 4-16 shows surface reaction rates for Lima, Ohio data^7 compared with the trend
lines developed from the Midland, Michigan data. Lesser surface area is required at the Lima
4-63
-------
location for the same degree of nitrification. This lower surface requirement is chiefly due
to the higher wastewater temperature, but the fact that the influent 6005 levels were lower
may also have caused a higher proportion of nitrifiers to be present in the trickling filter's
surface film. In the case of Lima, Ohio, the influent to the nitrification stage is produced by
a step aeration activated sludge plant.
Surface requirements for nitrification of oxidation pond effluent at Sunnyvale, California
are shown in Figure 4-17.'^ In this case, large quantities of algae were present in the
trickling filter influent. While the bulk of the algae passed through the unit unaffected, at
FIGURE 4-15
e
5
3
o
Ui
Uj
tt
UJ
SURFACE AREA REQUIREMENTS FOR NITRIFICATION
MIDLAND MICHIGAN
12,000
X
a" 10,000
kl
N
i 8,000
f
QQ
6,000
4,000
2,000
©
1 SF/lb/day = 0.2 m2/ kg/day
Influent Data (mean)
BOD5 !5-20mg/l
SS 15-20 mg/l
Organic N 1-4 mg/l
IMH^-N 8-18 mg/l
BOD5/TKN ~ 1,1
7 to II C
•f
\?> \o 19 C
© —
©
Key :
H T = 7 to 11C
© T = 13 to I9C
I.O 2.0 3.0 4.0
EFFLUENT AMMONIA-N, mg/l
5.0
6.0
4-64
-------
least 20 to 40 percent were trapped and eventually oxidized. This would have affected the
proportion of heterotrophic bacteria in the bacterial film, causing higher surface requirements
for nitrification at Sunnyvale, California than at Midland, Michigan.
Available data with rock media is more sparse and is summarized in Table 4-16 in terms of
ammonia oxidized per unit volume. Rock media is capable of ammonia oxidation at only 15
to 50 percent of the plastic media rate, on a volumetric load basis. The principal reason for
this is undoubtedly the rock media's lower specific surface, although the lower depth of the
typical rock filter may also have a role to play.
FIGURE 4-16
SURFACE AREA REQUIREMENTS FOR NITRIFICATION -
LIMA, OHIO
12,000
10,000
X
o
S:
-------
TABLE 4-16
NITRIFICATION IN SEPARATE STAGE ROCK TRICKLING FILTERS
Facility location
Johannesburg, S.A.
(full-scale)
North Hampton,
England
(pilot-scale)
Ref.
47
72
Depth,
ft
(m)
12
(3.7)
12
(3.7)
9
(2.7)
6
(1.8)
Media
2-3 in.
(5.1 to 7.6 cm)
rock
1.5 in.
(3 . 8 cm)
rock
1 in-
(2 . 5 cm)
rock
1.5 in.
(3 . 8 cm)
rock
Influent
BOD
mg/1
28
32
23
80
+.
NH4-N,
mg/1
23.9
25.2
22
33
Effluent
BOD ,
mg/1
14
13
10
10
NHT-N
4
mg/1
8.3
4.4
9.1
11.2
Percent
removed
65
83
59
66
Ammonia - N
oxidized
Ib/lOM cu ft/day
(kg/mVday)
3.5
(0.055)
2.2
(0.035)
2.4
(0.038)
1.0
(0.016)
4.7.1.3 Effect of Recirculation
An analysis of the Midland, Michigan data and Lima, Ohio, data has led to the conclusion
that while recirculation improved nitrification efficiency only marginally on an average
basis, the periods with recirculation demonstrated greater consistency (less fluctuations)
than when no recirculation was employed. '2' This conclusion, together with the
improvements seen with recirculation in combined carbon oxidation-nitrification applica-
tions (Section 4.4.1.4), leads to a general recommendation for the provision of recirculation.
A 1:1 recirculation ratio is considered adequate at average dry weather flow for most
applications.
4.7.1.4 Effluent Clarification
Since the organisms are attached to the media in attached growth systems, effluent
clarification steps are not required in all cases. In the case of Midland, Michigan it was found
that the effluent solids were approximately equal to the influent solids at 9 to 28 mg/1.22
This is because influent BODs levels were low (15 to 20 mg/1). When influent BODs loads
were increased above previous low levels, trickling filter effluent solids rose to 58 mg/1. The
insertion of clarifier allowed this to be reduced to 19 mg/1. Subsequent multimedia filtration
allowed further reduction to about 4 mg/1.
4.7.1.5 Effect of Diurnal Load Variations
Trickling filters used for nitrification, like any other nitrification process, are affected by
diurnal variations in nitrogen load. The rule of thumb developed in Section 4.3.3.2 can
likely be applied to trickling filters to prevent high ammonia bleed through during diurnal
peaks in load. Thus, the amount of surface area determined from Figures 4-15, 4-16 or 4-17
4-66
-------
under average daily loading conditions should be multiplied by the ratio of peak ammonia
load to average load to establish design surface area. An alternative would be to provide flow
equalization.
FIGURE 4-1 7
Q /2,000
\
Q
UJ
N
10,000
I
m
CD
-J
U.
CO
UJ
tt
UJ
tt
CO
8,000
. 6,000
4,000
2,000
SURFACE AREA REQUIREMENTS FOR NITRIFICATION -
SUNNYVALE, CALIFORNIA
1
18, 20O
1
1
1
Influent Data (mean)
COD =210 mg/l
SS =104 mg/l
Organic N = 10.0 mg/l
NHj-N = 16.7 mg/l
0
©
Sunnyvale, California Data
13 C to 19 C
0
O-"f
18
Midland, Michigan Data
13 C to 19 C
Key:
© Test result from
Sunnyvale work
1 1
468
EFFLUENT AMMONIA-N, mg/l
10
12
4-67
-------
4.7.1.6 Design Example
As an example consider a 10 mgd conventional activated sludge plant that must be upgraded
to meet effluent requirements of 4 mg/1 ammonia nitrogen and 10 mg/1 suspended solids on
an average basis. The plant is located in a temperate zone, and the minimum wastewater
temperature is 15 C. Present average effluent qualities are 1 mg/1 organic nitrogen, 20 mg/1
ammonia nitrogen, 15 mg/1 of suspended solids, and a BOD5 of 25 mg/1. The peak to
average nitrogen load ratio is 1.9. Consider as one alternative a plastic media trickling filter.
1. Calculate the BODs/TKN ratio:
BOD5/TKN = 25/21 = 1.19
2. The closest set of data based on BODs/TKN ratio and temperature is that for
Midland, Michigan (Figure 4-15). For an effluent ammonia nitrogen concentration
of 4 mg/1 at 15 C, the unit surface area requirement is about 3,800 sf/lb Nlfy-N
oxidized/day.
3. Calculate the ammonia nitrogen oxidized daily. The following equation is
appropriate:
NT = 8.33 • Q(NQ - Nj) (4-24)
where: NT = ammonia nitrogen oxidized, Ib per day
Q = average daily flow, mgd
NQ = influent NH* - N, mg/1
Nj = effluent NH+ - N, mg/1
For this example, assuming no change in the organic nitrogen level, the result is:
NT = 8.33 (10) (20-4) = 1,332 Ib/day
4. Find the total surface area requirement under average load conditions. Multiply-
ing the nitrogen oxidized per day (step 3) by the unit surface area requirement
(step 2) results in:
(1,332) (3,800) = 5,061,000 sf of media
5. Consider diurnal peak loading. One approach would be to provide flow
equalization. Assume that in this case site restrictions prevent this. Therefore,
increase the surface requirement by the peak to average nitrogen load ratio as
follows (Section 4.7.1.5):
4-68
-------
1.5 (5,061,000) = 7,592,000 sf
6. Choose a media type and establish media volume requirements. Effluent BOD5
and SS are low enough so that fairly high density media can be employed. In this
instance, a corrugated sheet module media having a specific surface of 42 sf/cu ft
is chosen. Media volume requirements are determined by dividing the total surface
requirement by the specific surface as follows:
7,592,000/42 = 180,750 cu ft
This volume could be provided by a variety of configurations; for instance two 75
ft diameter trickling filters with a media height of 21 ft would have the necessary
volume of media. Whatever configuration is chosen, the filter shouldn't be less
than about 12 to 15 ft in height because of the danger of short circuiting. Usual
practice is to consult with the media manufacturers) prior to final selection of
media configuration.
7. Establish recirculation rate. At 10 mgd ADWF, a 1:1 recycle is adequate (Section
4.7.1.3); therefore 10 mgd of recirculation capacity is recommended.
8. Establish clarification requirements. Effluent solids in the nitrification process
effluent will be approximately at the influent SS level, 15 mg/1. Therefore, to
meet a 10 mg/1 requirement, some form of effluent clarification is required such
as dual or multimedia filtration.
4.7.2 Nitrification with the Rotating Biological Disc Process
The rotating biological disc (RED) process, discussed in Section 4.4.2 for combined carbon
oxidation-nitrification applications may also be applied to nitrifying secondary effluents.
The process is constructed as shown in Fig. 4-8, excepting that it may be possible to
eliminate the secondary clarifier when the secondary effluent being treated has a BOD5 and
suspended solids less than about 20 mg/1.74 Under this circumstance the very low net
growth occurring in the nitrification stage causes the RBD process effluent suspended solids
to approximately equal the influent solids level. If lower levels of suspended solids are
required, the RBD process could be followed directly by tertiary filtration without the need
for intermediate clarification.7'*
One manufacturer has announced the availability of media especially adapted to
nitrification. The minimal biomass film development in separate stage nitrification
applications has allowed a 50 percent increase in surface area of the corrugated polyethylene
media. Standard shafts were 100,000 sq ft (9300 m^) of available surface area; the new
media is available at 150,000 sq ft (13,900 m^) of surface per shaft. This results in a
reduction of one-third in the number of shaft assemblies required for nitrification with the
RBD process.74
4-69
-------
4.7.2.1 Kinetics
The reaction rates occurring in each stage of the RED process treating secondary effluents
have been analyzed; the correlation between surface reaction rates and stage (effluent)
concentration is shown in Figure 4-18.74 jne trend line does not reach a plateau value but
keeps gradually rising because the biomass developed per unit surface is not constant.
Antonie found that the amount of culture developed on the rotating surface increased with
increasing ammonia nitrogen concentration. 74
A stage-by-stage application of Fig. 4-18 allowed the construction of Fig. 4-19 to be used
for design of 4 stage nitrification systems, the most commonly employed configuration.74
It can also be employed for other numbers of stages using the relative capacities shown in
the figure. The relative capacity factor should be applied to the hydraulic loading to obtain
design values for situations where other than 4 stages are employed.
Very little test data is available for temperatures below 13 C. For applications below 13 C,
the provisional recommendation has been made that the temperature correction factors
FIGURE 4-18
NITRIFICATION RATES AS A FUNCTION OF STAGE EFFLUENT
CONCENTRATION (AFTER ANTONIE (74))
1-
u.
o
CO
Q
\
1.0
0.8
Q
iu
lu
O
I
O.6
0.4
"
T
T
I
© Phoenix, Arizona, 22.8 to 27.8 C
A Madison, Wisconsin , 14.5 to 17.3 C
Q Broward County, Fla., 21 C
<•> Mansfield, Ohio
• Tiffin, Ohio, 13.9 C
Conditions: BOD5<20mg/l
Lb/day/iOOO sq. ft =
4.85 kg/1000 sq. m/day
I I 1 I
2 4 6 8 IO 12 14
EFFLUENT AMMONIA NITROGEN CONCENTRATION, tng/l
16
IB
4-70
-------
developed for combined carbon oxidation-nitrification (Section 4.4.2.2) be applied for
separate stage nitrification. '4
Figure 4-19 may also be used for hour-by-hour analysis of the effects of diurnal variations in
flow on effluent quality.74 This may tend to overestimate effluent quality during peaking
periods, however. To ensure that severe ammonia bleedthrough does not occur during peak
load periods, it would appear prudent to adopt the rule formulated in Section 4.7.1.5 for
trickling filters. Namely, the surface area determined from Figure 4-19 should be multiplied
by the ammonia nitrogen peaking ratio to establish the design surface area.
FIGURE 4-19
DESIGN RELATIONSHIPS FOR A 4-STAGE RED PROCESS
TREATING SECONDARY EFFLUENT (AFTER ANTONIE (74))
too
l-
e
LU
o
I3C
BOD5 <20 mg/l
Inlet Ammonia Nitrogen
Concentration, mg/l
40
30
1 gpd/sq ft = 41 Jt/mz/day
I
2345
HYDRAULIC LOADING, GPD/SQ FT
4-71
-------
4.7.3 Nitrification with Packed-Bed Reactors
Packed bed reactors (PER) for nitrification are a comparatively recent development, having
progressed from the laboratory stage to pilot-scale and commercial availability in a period of
only 5 years.23,24,75,76,77,78,79
Figure 4-20 shows one design.77,80 A pgR consists of a bed of media upon which biological
growth occurs overlaying an inlet chamber, much as in an upflow carbon column or filter.
Wastewater is distributed evenly across the floor of the PBR by baffles, nozzles or strainers,
similar to the way backwash water is distributed in down flow rapid sand filters. The
wastewater flow is upward, and a nitrifying biological mass is developed on the large surface
area of the media.
FIGURE 4-20
SCHEMATIC DIAGRAM OF A PACKED-BED
REACTOR (PBR). (AFTER YOUNG, ET AL., REF 77)
EFFLUENT
SUPPORT MEDIA
INLET CHAMBER
WASTEWATER INFLUENT
4-72
-------
4.7.3.1 Oxygenation Techniques
Several means have been employed for supplying the necessary oxygen for nitrification. The
earliest work used injection of air into the feed line entering the chamber.79 A subsequent
pilot-scale investigation used a similar procedure, excepting that the air was distributed
across the PBR floor, as shown in Figure 4-20.77 High purity oxygen has been used in two
alternative procedures.23,75,76 in one the oxygen was bubbled directly into the PBR. In
the second procedure, the liquid was preoxygenated in a reaction chamber prior to entry
into the PBR. Since preoxygenation is limited to satisfying the oxidation of about 10 mg/1
NH^-N due to the solubility of oxygen in water, effluent was recycled at a 2 to 3:1 ratio to
provide sufficient oxygen for nitrification.
4.7.3.2 Media Type, Backwashing and Loading Criteria
Several types of media have successfully performed in the PBR including 1-1.5 in. (2.5 to
3.8 cm) stones, 0.5 cm gravel, 1.8 mm (effective size) anthracite and 9 cm "Maspac," a
plastic dumped media manufactured .by the Dow Chemical Company, Midland, Michi-
gan.76,75,23,81,77
In the studies using the relatively light density anthracite and Maspac where ah- was injected
directly into the PBR, no backwashing was found to be necessary due to the turbulence
developed in the bed.77 Despite this, the General Filter Company recommends that when
anthracite is used, provision be made for increasing the hydraulic loading in surges for 1 to 2
hours to about four times the average ra'te, with air at a rate of 0.5 to 1.0 scfm/sq ft (2.54 to
5.08 l/s/m^). The frequency of the surging will vary, depending on influent quality and
flow rate. With the plastic media, the frequency of the surging can be reduced considerably
because of the high void volume, and in most cases excess solids can be withdrawn simply
by draining or backflushing the unit on a monthly or less frequent schedule.^
With the gravel media, standard practice was to backwash the reactor at 25 gpm/ft^ (127
1/s/m^) at least three times per week and in some cases daily.** 1 In the studies with the
stone media, backwashing was required with both direct and pre-oxygenation. Gravity
draining at 6 to 20 gpm/sq ft (30 to 102 l/s/m^) once or twice per week was sufficient to
prevent clogging. •*>^4
Data available for formulation of design criteria for PBR units are summarized in Table
4-17. Oxidation rates fall in the range of 4 to 27 Ib NH^-N oxidized per 1000 cu ft/day
(0.06 to 0.43 kg/m^/day). Factors affecting the oxidation rate are the influent quality
(BODj, TKN and NH^-N), temperature, and the type of media selected as a biological
growth surface. Oxidation rates at Pomona, Ca. were much greater than those at Ames, Iowa
at the same temperature, which is very probably due to the higher BOD5 and lower
ammonia content of the Ames secondary effluent. Very likely, there was a higher fraction
of nitrifiers in the Pomona biofilm. Interestingly, chemically clarified raw sewage (BOD = 93
mg/1) was compared to secondary effluent (COD = 46) at Pomona and only 60 percent
4-73
-------
nitrification was achieved with the chemically clarified feed at the detention time sufficient
to produce virtually complete nitrification of the secondary effluent.** 1 It is very probable
that this reduced efficiency was caused by a higher fraction of heterotrophs being present
when chemically clarified wastewater was being treated.
Temperature has a strong effect on the PER process. Figure 4-21 shows the detention time
required for relatively complete nitrification (< 2 mg/1 NH^-N) at steady state as a function
of temperature. If Figure 4-21 is used for sizing a PBR, attention must also be given to
diurnal variation in nitrogen loads. It would be prudent to multiply the detention time
determined from Figure 4-21 by the peak to average nitrogen load ratio to establish the
design detention time. This should present extremes in ammonia bleedthrough during
diurnal peak conditions.
The media type chosen affects the amount of surface available for nitrifier growth. For
instance, Table 4-17 shows that anthracite was superior to Maspac media in terms of
oxidation rate at Ames, Iowa; this may be due to the higher surface area of anthracite media
compared to Maspac.
FIGURE 4-21
TEMPERATURE DEPENDENCE OF DETENTION TIME FOR COMPLETE
NITRIFICATION, (<2 mg/1 NH*-N) AT STEADY STATE IN THE PBR
30
25
Uj
tt
ft:
LU
Q.
§
LU
20
15
O
L- O
A
KEY:
Ref.
S 81
81
81
23
77
77
77
Tiiiir~
•TREND LINE FOR POMONA DATA (Ref. 81)
Location
Media
Comment
One Column, 02
Two Columns, 02
One Column, Air
Pomona, Ca. Gravel
Pomona, Co. Gravel
Pomona, Co. Gravel
Union City, Ca. Stone
Ames, Iowa Anthracite
Ames, Iowa Maspac
Ames, Iowa Series Maspac-Anthracite
I
40
60 SO IOO I2O
EMPTY BED DETENTION TIME, MIN
I4O
160 180 2OO
4-74
-------
TABLE 4-17
PACKED BED REACTOR PERFORMANCE WHEN TREATING SECONDARY EFFLUENTS
Location
and type
of aeration
Union City, Ca.
Preoxygenatlon
(oxygen)
Bubble
(oxygen)
Ames, Iowa
Bubble
(air)
fc
J
n
Pomona, Co.
Bubble
(oxygen or
air)
Ref.
23,
24
23,
24
77,
83
81
Media
depth ,
ft
(m)
3.0
(.91)
3.0
(.91)
5.0
(1.5)
e.o
(2.4)
13
(4.0)
5.5
(1.7)
11.0
(3.4)
Media
type
1 to 1.5 In
(2 . 5 to 3 . 8 cm)
stone
1 to 1.5 in
(2.5 to 3.8cm)
stone
1 .8 mm (Dio)
U.C. = 1.7
anthracite
Maspac
Series
anthracite
Maspacb
5 mm
gravel
5 mm.
gravel ,
two columns
in series
Surface
loading
gpm/sf
(m3/m2/day)
.15
(8.8)
.21
(12)
.29
(17)
.15
(8.8)
.29
(17)
. 1.0
(59)
0.4
(23)
1.0
(59)
0.4
(23)
0.5
(29) b
0.75
(44)
0.75
(44)
0.59
(35)
0.41
(24)
0.46
(27)
0.39
(23)
1.49
(87)
0.75
(44)
Empty
bed
hydraulic
detention
a
time , min
154
103
77
1S4
77
" 37
94
60
ISO
195
130
55
70
100
90
105
55
110
Ammonia-N
oxidation
rate
lb/1000 cu ft/day
(kg/m3/day)
7.7
(.12)
12.1
(.19)
9.8
(.15)
9.7
(.15)
-
9.4
(.15)
5.9
(0.09)
5.1
(0.08)
3.5
(0.06)
6.1
(0.10)
4.6
(0.07)
26.5
(0.42)
20.1
(0.32)
13.3
(0.21)
14.7
(0.24)
16.4
(0.26)
26.7
(0.43)
13.8
(0.22)
Temp. ,
C
21 to 27
21 to 27
21 to 27
16 to 30
IE to 30
21 to 23
21 to 23
9 to 14
12
27 to 28
25 to 26
19 to 22
20 to 25
20 to 22
26 to 28
22 to 25
Influent Quality, mg/1
BOD5
35
38
37
37
43
39 ;
20
39
20
26
37
7C
SS
27
38
25
30
48
.13
26
43
26
47
63
9c
Organic-N
3.6
-
5.7
5
-
'_
-
-
_
_
-
_
-
-
-
-
_
-
Ammonla-N
14.3
19.6
15.2
18.3
-
8.4
6.8
8.4
6.8
14.4
11.2
18.1
17.6
16.8
16.4
20.7
17.6
18.9
Effluent Quality, mg/1
BOD5
5
3
10
10
25
19'
5
19
8
8
16
_
-
-
-
-
_
-
SS
4
7
7
16
51
;
_
_
_
_
_
.
-
-
-
-
_
-
Organic-N
1.5
-
2.5
2.2 - 4.7
-
;
•
_
_
_
-
_
-
-
-
-
_
-
Ammonia-N
1.0
5.6
6.9
1.8
-
5
1
5
1
1
5
1.9
1.4
2.0
1.7
1.5
1.3
1.9
Nltrite-N
0.6
4.1
1.7
0.4
-
• .. •
_
-
_
.
_
0.6
0.6
0.9
0.4
0.5
0.4
0.2
Nitrate -N
15.9
6.9
6.0
18.3
-
.
_
_
_
_
_
16.6
16.3
16.9
15.6
20.7
17.1
18.2
Removals,
percent
BOD
86
91
74
74
41
51
77
51
59
68
56
~
-
-
SS
87
83
74
48
-6
20
32
42
40-
49
39
_
_
Organic -N
58
-
56
_
-
_
-
-
-
-
_
_
_
NH+-N
93
71
55
91
-
46
90
40
87
92
59
90
92
88
90
93
93
90
Basis: Influent flow
A product of the Dow Chemical Co., Midland, Mich.
cAverage for test series treating" activated sludge effluent
-------
Effluent BOD and SS levels are affected by the type of aeration (see Table 4-17).
Preoxygenation allows the PBR to produce effluents of similar quality to tertiary
multimedia, filtration. Bubble aeration, however, causes continuous shearing of the
biological film from the media, resulting in lower reductions of BOD and suspended solids.
When very low levels of effluent solids are required, effluent filtration may be required
when bubble aeration is used in the PBR.
4.8 Aeration Requirements
Care must be exercised in designing aeration systems for nitrification. Unlike BOD,
ammonia is not adsorbed to the biological floe for later oxidation, and therefore the ammonia
must be oxidized during the relatively short period it is in the nitrification reactor.
Therefore, sufficient oxygen must be provided to handle the load impressed on the
nitrification process at all times. This problem is particularly critical when either activated
sludge or a packed bed reactor system is used for nitrification; in the other attached growth
systems the aeration is provided as the liquid spills over the media and the design
considerations relate to proper ventilation rather than oxygen transfer.
Very significant diurnal changes in nitrogen load have been observed. Load variations at the
Chapel Hill treatment plant are shown in Fig. 4-3 and the load pattern would be
representative of systems provided with no significant in-process flow equalization. In this
case, peak to average nitrogen load rose to nearly 2.2, considerably above the peak to
average flow ratio of 1.44. As an example of a plant with some in-process flow equalization,
Fig. 4-12 shows the load variations observed in the activated sludge effluent of the Rancho
Cordova Wastewater Treatment Plant. In this case the nitrogen peaking is moderated by the
equalization in quality provided by the activated sludge aeration tanks and secondary
clarifiers. The ammonia peak to average ratio at 1.63 approximates the flow peak to average
ratio of 1.57.
In addition to being affected by in-process flow equalization, the diurnal variation in
nitrogen loading is also very significantly affected by equalization in the wastewater
collection system. Large collection systems serving spreadout urban areas have high built-in
storage providing unintentional flow and quality equalization. This relationship is indicated
in Fig. 4-22 where the nitrogen load peaking (expressed as the ratio of the maximum hourly
load to average load) is plotted for eight treatment plants having no significant in-process
equalization. There is an interesting relationship between flow peaking and ammonia load
peaking shown in Figure 4-22. In large plants such as Blue Plains plant at Washington, D.C.
and Sacramento, California a spread out collection system causes moderation of both flow
and nitrogen load peaking. In the smaller systems, however, without such "flow
equalization," ammonia load peaking can be substantial; for example at the Central Contra
Costa Sanitary District's (CCCSD) plant an hourly peaking factor of 2.4 has been measured.
The aeration system must accommodate these changes in loads to avoid ammonia bleeding
through during the peak load period. The diurnal variations in load can be quite extreme; in
Figure 4-23 the peak to minimum hourly loads are plotted against the flow peaking factor.
Ratios as high as 10:1 have been observed.
4-76
-------
FIGURE 4-22
RELATION BETWEEN AMMONIA PEAKING AND HYDRAULIC PEAKING
LOADS FOR TREATMENT PLANTS WITH NO IN-PROCESS EQUALIZATION
2.5
8,
o
o
o
-J
c
o
E
5
I
o
o
o
-J
c
o
I.O
T
Y = 1.457 X -0.217
r =0.823
1 rr\Z/Sec = 43.8 mgd
KEY
Plant
Sample
Lebanon, Ohio
Livermore, Ca
CCCSD, Co
Sacramento, Ca
Blue Plains, DC
Chapel Hill, NC
Canberra, Australia-
Weston Creek Raw
Belconnen Raw
I
Primary
Roughing Filter
Primary
Primary
Primary
Raw
ADWF,
mgd
\.2
3.4
2!
45
274
1.8
Ret.
84
85
63
86
87
30
1.0 88
10 88
(X)
I.O 1.5 2.0
Maximum Hourly Flow, mgd/Average Daily Flow, mgd
2.5
An early decision must be made during the design process as to what level of peaking of
oxygen demanding substances will be designed for. In addition to peaking of ammonia or
organic nitrogen, a concurrent peak may also occur in the loading of organic substances. If
very low levels of ammonia nitrogen are required at all times care must be used to develop a
statistical base whereby the frequency of peak oxygen loads can be identified. Not only
should daily peaks be identified, but possibly those occurring on weekly or monthly bases.
Table 4-18 presents an example of such an analysis for the primary effluent from two plants
in St. Louis, Missouri using COD as a measure of oxygen demanding substances since in that
4-77
-------
particular instance nitrification was not required. As may be seen, significant departures
from average conditions occur on a fairly frequent basis. Similar analyses may be justified
when designing for nitrification.
The extra aeration capacity required for handling diurnal variations in nitrogen load, coupled
with the extra tankage and equipment required, may dictate in-plant flow equalization in
many instances. The reductions in capital and operating cost of aeration tankage and
aeration facilities must be compared with the cost of flow equalization to determine
applicability to specific cases. Design procedures for flow equalization are contained in
Chapter 3 of the Process Design Manual for Upgrading Existing Wastewater Treatment
Plants.25
FIGURE 4-23
^ RELATIONSHIP OF MAXIMUM/MINIMUM NITROGEN
£ LOAD RATIO TO MAXIMUM/AVERAGE FLOWS
*" 13
o
-j
•J
.o
c
o
6
6
I
Y = 9.461 X- 8.28
r = 0.871
See Fig. 4-22
for Symbols
I.O 1.5 2.0
Maximum Hourly Flow, mgd/Average Daily Flow.mgd
4-78
-------
TABLE 4-18
PEAKING FACTORS VERSUS FREQUENCY OF OCCURRENCE
FOR PRIMARY TREATMENT PLANT EFFLUENT
•,.
Frequency of Occurrence
4 hours /day
4 hours /week.
4 hours/month-*
4 hours/3 months
4 hours/6 months
Bissel Point
Treatment Plant
COD load
peaking
factor b
1.30
1.72
2.02
2.25
2.40
Flow
peaking^
factor15
1.18
1.40
1.60
1.72
1.80
Lemay
Treatment Planta
COD load
peaking,
factor
1.40
1.85
2.35
2.62
2.80
Flow
peaking
factor D
1. 20 '
1.44
1.7.0
1.88
1.96
Data is from reference 89; both plants serve the St. Louis area in Missouri and
process about 100 mgd each.
Peaking factor is defined in each case as the ratio of the 4 hour load listed to the
average daily^load.
4.8.1 Adaptability of Alternative Aeration Systems to Diurnal Variations in Load
Careful consideration should be given to maximizing oxygen utilization per unit power
input. In the face of significant load variation, the aeration system should be designed to
match the load variation while economizing on power input. Obviously, designing the
aeration system ft) provide for the maximum hourly demand 24 hours a day would provide
over aeration the majority of the time with wasteful losses of power.
The available means for aeration are summarized in Section 5.3.4 of the Process Design
Manual for Upgrading Existing Wastewater Treatment Plants, a Technology Transfer
publication. Of the available aeration devices, the mechanical surface aerator is least well
suited to nitrification applications. This is because they are normally designed to operate at
fixed speed and^therefore must overaerate the majority of the day to satisfy the peak
oxygen demands. Even when equipped with variable submergence of the blade, the units are
limited to matching less than a 2:1 variation in load, at best. Therefore, unless flow
equalization is provided somewhere in the system, mechanical surface aerators are not
capable of matching variations in nitrogen loads without overaerating the mixed liquor
during a significant portion of the day.
4-79
-------
Using diffused air aeration, air rates can be easily modulated to closely match the load,
merely by turning down or shutting off individual blowers. Thus, the diurnal load variations
can be matched without the necessity of over aerating the mixed liquor and wasting power.
Fine bubble diffusers can be arranged across the tank floor,90 allowing fairly even
distribution of energy input. Gentler mixing is provided than with mechanical aeration
plants, providing less tendency for floe breakup.
Submerged turbine aeration systems are intermediate in terms of their responsiveness to the
problem of aeration in nitrification systems. Because of their capability to vary the air rate
to the sparger, they may be designed to match the load variation in oxygen demand. A
drawback, however, is that the impeller normally operates at fixed speed, imparting no turn
down capability for a significant part of the power draw. In fact, in some impeller designs,
the power draw of the ungassed impeller is actually greater than when gas is fed to the unit.
4.8.2 Oxygen Transfer Requirements
Oxygen requirements for nitrification alone were discussed in Section 3.2.4. Oxygen
requirements in all practical cases are compounded by the oxygen required for stabilization
of organics.
Reasonably exact expressions for oxygen requirements for heterotrophic organisms and
nitrifiers have been developed.38 The approach, however, requires pilot plant data to
provide COD balances and sludge yields. In general, this information is not available and a
simpler approach may be adopted.
In normal activated sludge treatment when nitrification is not required, the amount of
oxygen needed to oxidize the BOD5 can be calculated by the following equation:
B = X(BOD5) (4-25)
where: B = oxygen required for carbonaceous oxidation, mg/1
X = a coefficient
The coefficient X relates to the amount of endogenous respiration taking place and to the
type of waste being treated. For normal municipal wastewater, the X value would range
from .5-.V for high rate activated sludge systems to 1.5 for extended aeration. For
conventional activated sludge systems X can be taken as 1.0.
In the case of nitrification, the oxygen requirement for oxidizing ammonia must be added
to the requirement for BOD removal. The coefficient for nitrogen to be oxidized can be
conservatively taken as 4.6 times the TKN content of the influent (Section 3.2.4) to obtain
the nitrogen oxygen demand (NOD) and the value of X in Equation 4-25 can be assumed to
be approximately 1.0. In actual fact, some of the influent nitrogen will be assimilated into
4-80
-------
the biomass or is associated with refractory organics and will not be oxidized. These
assumptions lead to the following oxygen requirement:
W = BOD5 + NOD (4-26)
where: W = the total oxygen demand mg/1, and
NOD = oxygen required to oxidize a unit of TKN taken as
4.6 times the TKN
Since aeration devices are rated using tap water at standard conditions, the rated
performance of the aerator must be converted to actual process conditions by the
application of temperature corrections and by factors designated c*and |3 which relate
waste characteristics to tap water characteristics.
Temperature corrections are made by the relationship:
(T-20)
1 .024 where T = process temperature in degrees C.
The PC factor is the ratio of oxygen transfer in wastewater to that in tap water and is
represented by the following:
KT a (process conditions)
CX = -L - (4-27)
KT a (standard conditions)
Values of * can vary widely in industrial waste treatment applications, but for most
municipal plants, it will range from 0.40 to 0.90.
The /3 factor is the ratio of oxygen saturation in waste to that in tap water at the same
temperature. A value of 0.95 is commonly used. Thus, the actual amount of oxygen
required to be transferred (W) can be determined from the amount transferred under test
conditions (WQ) by the equation:
W0 -_ (4_28)
where: W = oxygen transferred at process conditions, Ib/day
WQ = oxygen transferred at standard conditions
(T = 20 C, DO = .01 mg/1, tap water), Ib/day
T - process temperature, C.
C = oxygen saturation in water at temperature T, mg/1
o
C. = process dissolved oxygen level, mg/1
4-81
-------
The process dissolved oxygen level, Cj, must be set high enough to prevent inhibition of
nitrification rates (Section 3.2.5.5). For this purpose, a minimum value of 2.0 mg/1 is
recommended. This value is also applicable under peak diurnal load conditions, and the
practice of allowing the DO to drop below 2.0 mg/1 under peak load is not recommended. If
the DO were to be allowed to drop below 2.0 mg/1 during peak load conditions, excessive
bleedthrough of ammonia could be expected.
Using example values for domestic sewage (
-------
FIGURE 4-24
RELATIONSHIP OF AERATION AIR REQUIREMENTS FOR
OXIDATION OF CARBONACEOUS BOD AND NITROGEN
DIFFUSER EFFICIENCY, PERCENT
RATED IN TAP WATER -STD CONDITIONS
^coSivlSoJcoc
v
N
Terr
\
iperat
X
ure 2
x
0 C
kg/m3/day = 62.4 Ib/
\
**X_
^^^
^^
1
1000
— —
C F/day
500
IOOO I5OO
AIR REQUIRED, CU FT/LB BOD5 + NOD
I TOO
theoretical considerations for coarse bubble aeration systems. It should be noted that none
of the plants listed in Table 4-20 have automated blower control systems linked to DO
probes. When the Livermore, California plant staff manually adjusted the blower output
according to the DO probe reading, they were able to reduce the air requirement to an
average for the day of 680 CF/lb BODs + Ib NOD (42 m3/kg) based on hourly variations in
TABLE4-19
RELATION OF OXYGEN TRANSFER EFFICIENCY TO
AERATOR POWER EFFICIENCY (AFTER NOGAJ (92))
Diffuser oxygen
transfer efficiency,
percent
(at standard conditions)
Aerator power
efficiency
Ib 0_/bhp/hr
Lt
4
6
8
10
12
1.23
1.85
2.46
3.08
3.70
4-83
-------
Thus, it is possible that automated blower operation may reduce aeration
requirements.
In addition to determining the total air requirement, attention should also be given to air
distribution within the aeration tanks. If the conventional (or plug flow) mode of operation
is established as the normal operating procedure, the air requirements will be greatest at the
head end of the aeration tanks (see Section 4.3.5) .
TABLE 4-20
AIR REQUIREMENTS FOR NITRIFICATION ACTIVATED SLUDGE PLANTS
Treatment
plant
Medford, Oregon
Flint, Michigan
Livermore, Calif.
Central Contra Costa
Sanitary District, Calif.
Jackson, Mich.
Hyperion Treatment Plant
West Battery,
Los Angeles, Ca.
Whittier Narrows ,
County Sanitation
Districts of Los Angeles
County, Ca.
San Pablo Sanitary District
Treatment Plant, Ca.
Configuration
Plug flow or
step aeration
Plug flow or
step aeration
Separate stage:
roughing bio-
filter followed
by nitrification
Separate stage:
lime followed
by nitrification
Plug flow
Plug flow
Step aeration
Plug flow;
roughing filter
followed by
nitrification
Diffuser type
Coarse
bubble
—
Coarse
bubble
Coarse
bubble
Coarse
bubble
Coarse
bubble
Coarse
bubble
Coarse
bubble
Oxygen demand
distribution, %
BOD
b
73
65
50
21
66
61
59b
59
NOD
27
35
50
79
34
39
41
41
cu ft/lb of BOD
and NOD3
1390
1280
1250
1700
910
1160
1180b
1410C
Data reference
93
6
5
3
42
2
9
94
al cu ft/lb = .062 m3/kg
Assuming primary effluent BOD5/COD ratio of 0.6
CDuring June 1974 when nitrification run at design loads
4.8.3 Example Sizing of Aeration Capacity
As a design example, consider a 10 mgd plant with lime clarification in the primary
sedimentation stage:
4-84
-------
Given the following primary effluent properties:
Design flow = 10 mgd
Average BOD 5 load = 4,170 Ib/day
Average TKN load = 2,100 Ib/day
Peak/average TKN load ratio = 1.8
Peak/average BODs load ratio = 1.5 (coincident with peak TKN load)
Average load condition:
BOD5= 4,170 Ib/day
NOD = 4.6 x 2,100 = 9,660 Ib/day
BOD5 + NOD = 13,830 Ib/day
Peak hourly load condition:
BOD5 = 4,170 x 1.5 = 6,260 Ib/day
NOD = 9,660 x 1.8 = 17,390 Ib/day
BOD5 + NOD = 23,550 Ib/day
Ratio peak hour to average load: 1.7
For the purposes of this example assume a fine bubble diffused air aeration with a
design figure of 725 cf/lb BOD5 + NOD.
Average aeration requirement:
725 x 13,830/1,440 = 6,963 CFM
Peak aeration requirement:
725 x 23,55O/1,440 = 11,860 CFM
In sizing plant air requirements, separate provision must be made for preaeration, air in
mixed liquor and return sludge channels, and air requirements for aerated stabilization in
downstream denitrification processes.
4.9 pH Control
The implications of adverse operating pH and its causes have been discussed previously in
Sections 3.2.3 and 3.2.5.6. In cases where the alkalinity of the wastewater will be depleted
by the acid produced by nitrification, the natural alkalinity of the wastewater will have to
be supplemented by chemical addition. As discussed in Section 4.5.1, the effects of
operation of upstream processes on alkalinity and pH must be considered. Alum or iron
4-85
-------
addition tend to deplete wastewater alkalinity, whereas in some instances lime addition
increases the alkalinity.
4.9.1 Chemical Addition and Dose Control
Two alternative chemicals, caustic (NaOH) and lime (CaO or Ca(OH)2) are in predominate
use for pH control. As lime is less expensive than caustic for the same change in alkalinity,
lime will generally be favored, except in smaller plants. Procedures for the feeding and
storage of these chemicals are described in two EPA Technology Transfer publications,
Process Design Manual for Phosphorus Removal^ and Process Design Manual for
Suspended Solids Removal. 96
The need for pH adjustment may vary diurnally. In one case it was found that an alkalinity
deficit occurred daily only for several hours at the peak nitrogen load condition. At other
times, sufficient alkalinity was available. This condition caused a cyclic variation in pH. In
situations where diurnal variation in the pH depression may be encountered, continuous
on-line monitoring of pH for the purpose of controlling chemical addition seems justified. In
the case of suspended growth systems operated in the plug flow mode, probes may be
positioned at several points in the aeration basin, with provision for addition of chemical at
several points.62 jn the case of attached growth sytems and suspended growth systems
operated in the complete mix mode, effluent monitoring of pH would be the usual choice
for controlling chemical addition to the influent.
4.9.2 Effect of Aeration Method on Chemical Requirements
The type of oxygen transfer device chosen can have a marked effect on the chemical dose
required for pH control. As an example, the differences between coarse bubble and fine
bubble air diffusion systems will be examined. The following equation from bicarbonate-
carbonic acid equilibrium is useful for estimating the pH level in aeration tanks:
CO)
(3'9)
At 20 C, the value of pKj is approximately 6.38. In using the equation, the alkalinity in the
aeration tank can be used to estimate the bicarbonate level (HCC>3) and the value used
should reflect any alkalinity depletion resulting from nitrification. The level of H2CO3
(carbon dioxide in solution) can be estimated from Henry's Law as follows:
Ceq = H Pgas (4-30)
where: C = concentration of gas dissolved in liquid at equilibrium, mg/1,
eq
H = Henry's Law constant, mg/1/atmosphere
P = partial pressure of gas in equilibrium with the liquid,
atmosphere
4-86
-------
The Henry's law constant for carbon dioxide at 20 C is 1688 mg/1/atmosphere. To establish
Ceq for use in Equation 3-9, the level of carbon dioxide in the aeration air must be
estimated. In unpolluted atmospheric air, the partial pressure of carbon dioxide is only
about 0.0003 atm. However in aeration air, carbon dioxide is generated from the oxidation
of organics and from nitrification. For the mix of organics in municipal wastewaters, about
one mole of carbon dioxide is produced per mole of oxygen consumed. Similarly for
nitrification, about one mole of carbon dioxide is produced per mole of oxygen consumed
(c.f. Section 3.2.2). On a weight basis, for every Ib of oxygen consumed about 1.38 Ib of
carbon dioxide is produced. Equations 3-9 and 4-30 can be solved simultaneously to
determine the amount of dissolved carbon dioxide in the liquid and the amount present in
the gas as well as the process operating pH.
The following example is illustrative of the procedure. Assume a residual alkalinity of 150
mg/1 as CaCC>3 (3 x 10"3 moles/liter of HCO~3) and an oxygen transfer efficiency of 12
percent, under standard conditions. Equation 4-29 would indicate that 725 cu ft/lb BOD5 +
NOD is required. Assume also that the example conditions given in Section 4.8.3 are used,
namely that 13,830 Ib of oxygen demand are contained in 10 million gallons of wastewater;
this allows the calculation that 1 Ib of oxygen demand is contained in 725 gallons.
Therefore the aeration rate is 1 cu ft/gal of wastewater nitrified (725 cu ft/125 gal).
Assuming a pH of 7.1 (by trial and error), Equation 3-9 allows calculation of the dissolved
carbon dioxide concentration, and for this case it is found to be 25 mg/1. Therefore at this
concentration, 725 gallons contain 0.15 Ib of carbon dioxide (CO2) of the total evolved
(1.38 Ib of CO2 evolved/lb of oxygen). Equation 4-30 allows calculation of the partial
pressure of CO2 in the off gases; the calculated result is 0.0149 atm. This concentration,
plus the volume of gas required allows calculation that the gas contains 1.23 Ib of carbon
dioxide. The total is then 1.38 Ib which checks with the expected result. If the total had not
been 1.38 Ib, a new pH value level would be assumed and the procedure tried again until a
balance is obtained.
Table 4-21 presents the results for a number of oxygen transfer efficiency values, using the
procedure outlined above. The trends shown are valid for municipal wastewaters, although
the absolute value of the pH may differ slightly from those indicated due to variation from
the assumed conditions.
Examining Table 4-21, it can be seen that fine bubble diffuser systems, at high oxygen
transfer efficiency, will operate at lower operating pH levels than coarse bubble diffusers
operated at lower oxygen transfer efficiencies. If the same operating pH is to be
maintained, this will translate into higher chemical doses for fine bubble aeration systems.
For instance, Table 4-21 shows that if the residual alkalinity is 100 mg/1, a pH of 7.0 can be
maintained with a coarse bubble aeration system rated at 9 percent efficiency under
standard conditions. For a comparable fine bubble system rated at 18 percent under
standard conditions, the residual alkalinity would have to be raised to 150 mg/1 as CaCO3 to
maintain the pH at 7.0. This would require an extra lime dose of at least 28 mg/1 as CaO.
4-87
-------
TABLE 4-21
EFFECT OF OXYGEN TRANSFER EFFICIENCY AND
RESIDUAL ALKALINITY ON OPERATING pH
Residual
pH at stated
Operating transfer efficiency, percent3
as CaCO , mg/1
O
50
75
100
125
150
175
200
Coarse bubble range
6
6.9
7.1
7.2
7.3
7.4
7.4
7.5
9
6.7
6.9
7.0
7.1
7.2
7.3
7.3
Fine bubble range
12
6.6
6.8
6.9
7.0
7.1
7.2
7.2
18
6.5
6.7
6.8
6.9
7.0
7.0
7.1
at standard conditions
As an example of the use of Table 4-21, presume that it is desired to maintain the operating
pH level at 7.2. Assume that an aeration system has been chosen which has an oxygen
transfer efficiency, eo, of 12 percent and residual alkalinity of 75 mg/1 as CaCO3- If it is
desired to maintain the process pH at 7.2 to prevent inhibition of nitrification rates, Table
4-21 indicates that a residual alkalinity of 175 mg/1 is required. The difference between the
available alkalinity and the required alkalinity is 100 mg/1 as CaCO3 and this could be
supplied by a lime dose of 56 mg/1 (as CaO).
In plug flow systems, the pH will steadily decline from the influent end to the effluent end,
following the similar decline in alkalinity occurring due to nitrification. The most severe pH
depression will be at the effluent end, after the bulk of the nitrogen has been oxidized.
Complete mix systems, on the other hand, will be uniformly depressed in pH throughout
the tank because of the uniformity of aeration tank contents. This is a disadvantage for
nitrification with complete mix activated sludge compared to plug flow as the pH of the
entire complete mix tank will be the same as the pH at the effluent end of the plug flow
tank at the same oxygen transfer efficiency.
The preceding discussions about the effect of carbon dioxide evolution in operating pH in
the aeration tank are not applicable to cases where nitrification follows lime treatment. In
4-88
-------
these cases, the carbon dioxide that is evolved is used in recarbonation reactions and very
little enters the aeration air (c.f. Section 4.5.1).
4.10 Solids-Liquid Separation
In all suspended growth systems and in most attached growth systems, the nitrification stage
must be followed with a solids removal stage. Because of the complexity of the solids-liquid
separation problem, full consideration cannot be given to it within the scope of this manual.
Rather, a brief review of the problem is given with reference to the pertinent literature.
In some attached growth nitrification systems, multimedia filtration is provided following
the nitrification stage as the level of effluent solids is low. Design criteria for multimedia
filtration are presented in the EPA Technology Transfer Publication, Waste-water Filtration,
Design Considerations. General design criteria for secondary sedimentation as well as
multimedia filtration applicable to nitrification processes are presented in the Process Design
Manual for Upgrading Existing Wastewater Treatment Plants. 25
In suspended growth systems, there is a strong interrelationship between operation of the
secondary clarifier and operation of the aeration basin. The degree to which the return
sludge can be thickened will affect the allowable mixed liquor in the aeration tank and
therefore the size of the aeration tank. Dick and coworkers have presented design
relationships which are useful for analyzing clarifier-aeration tank interactions.98,99,100
Consideration of several factors is required when designing secondary sedimentation tanks.
Mechanical design of the tank is very important. Inlet turbulence must not upset the tank;
an energy dissipating inlet well serves the purposes of distributing the flow equally across
the tank and of providing for some mild turbulence to help aggregate the finely divided
solids into floe and increase the separation efficiency. Effluent launders must be sized and
placed so that currents are not created that will upset the tank and cause short-circuiting of
the solids to the effluent. The tank must also have sufficient depth to allow a sludge blanket
to form under all conditions, especially under those conditions which occur when the
system operation is limited by the ability to obtain a specific return activated sludge
concentration. Overflow rates must not be so great that sludge is suspended in the "upflow"
and carried over the weirs. Lastly, the sludge must be quickly removed from the tank to
minimize the ocurrence and duration of anoxic (zero DO) conditions.
The nature of the biological solids developed in suspended growth systems plays an
important role in the operation and design of the secondary sedimentation tank. First, the
ability of the mixed liquor to be clarified will affect the size of the required sedimentation
tank. This property is dependent on the extent of disperson of the biological solids. If a
large proportion of the biological solids are finely divided, the separation efficiency of these
solids will be poor. If, however, only small portions of the biological solids are dispersed and
the bulk is incorporated into floe, the separation efficiency of the mixed liquor should be
high. Consolidation of the small dispersed particles into the large floe to improve effluent
4-89
-------
clarity can be encouraged by a number of techniques. Chemical coagulation of the dispersed
solids has been successfully employed to enhance the flocculation of biological solids J01
Physical conditioning of the activated sludge floe prior to secondary sedimentation is
another technique which can be used with or without chemicals. It has been found that by
incorporating mild turbulence beweeen the oxidation tank and the secondary sedimentation
tank, effluent clarification can be enhanced. ^2 Mild turbulence can be conveniently
accomplished by using mild aeration in transfer channels and by incorporating energy-
dissipating inlet wells in circular secondary sedimentation tanks.
The thickening qualities of the sludge will determine the required area for sludge thickening
in the tank, depending on the desired or optimal MLSS or Return Activated Sludge (RAS)
levels. The limiting design consideration for thickening is usually at peak wet weather flow
(PWWF), for it is under these conditions that the solids loading on the secondary sedimenta-
tion tanks are the greatest. Secondary sedimentation tanks must be sized such that mixed
liquor solids are not lost in the effluent during PWWF conditions. Polymer feed facilities
are appropriate to allow short-term polymer dosing during wet weak peak load conditions to
enhance the thickening qualities of the sludge.
Thickening qualities will decrease and sedimentation tank area requirements will increase
when wastewater temperatures decline. Figures 4-25 and 4-26 show the effect of
temperature on three different oxygen activated sludges and the temperature trends would
be expected to be similar for air activated sludges. In sedimentation tank design, solids
loadings should reflect the minimum wastewater temperatures expected. When temperatures
decline the mixed liquor levels that can be maintained under aeration also decline due to the
decreased level of thickening that can be obtained in the sedimentation tank.
Rising sludge caused by denitrification in secondary clarifiers has occasionally plagued
nitrification operations. Denitrification occurs because the organisms in the biological sludge
in the secondary clarifier can utilize nitrate and nitrite as electron acceptors to oxidize
organic compounds in the sludge layer. The formation of bubbles from nife-ogen gas evolved
in denitrification and other gasses, such as carbon dioxide, can then cause flotation. Early
experimental work by Sawyer and Bradney^4 an(} subsequent experimental and theoretical
work by Clayfield 105 have identified the important parameters affecting denitrification. A
factor of foremost importance is the presence of adequate quantities of nitrites or nitrates in
the wastewater to cause bubble formation. In essence, the dissolved gases in the wastewater
(nitrogen, carbon dioxide, and oxygen) must be in sufficient concentration so that the sum
of their partial pressures equals or exceeds the ambient liquid pressure. In the case of
Sawyer and Bradney's work, as little as 4-6 mg/1 of nitrate-N or nitrfJTe-N would cause
flotation in a graduated cylinder, while Clayfield found that 16-18 mg/1 nitrate-N were
needed to cause flotation in full scale sedimentation tanks. Differences between investiga-
tions are in part due to the greater liquid pressures in deep sedimentation tanks compared to
graduated cylinders and may also reflect differing levels of concentration of other gases,
such as carbon dioxide.
4-90
-------
The degree of stabilization of the sludge also has a profound effect on denitrification. It
has been shown that sludges incorporating unoxidized feed organics float more readily than
well oxidized sludges. 104 Temperature is also important as it affects the rate of
denitrification and therefore affects the rate of gas and bubble formation. 104, 105
FIGURE 4-25
EFFECT OF TEMPERATURE ON THICKENING PROPERTIES
OF OXYGEN ACTIVATED SLUDGE AT MLSS = 4000 mg/1
(REFERENCES 86 AND 103)
30
28
6 26
O 24
•*. 22
O
CO
CO 20
5
«.
O
IB
16
£ *
-J
Uj /O
to
5
-J
K
K-
kl
CO
LU
S
o
N
2 —
WASHINGTON, D.C
A
T
CITY "A"
SACRAMENTO, CA.
KEY
0 SACRAMENTO,CA.
& WASHINGTON, D.C.
Q CITY "A"
IO 15 2O 25
TEMPERATURE, C
3O
4-91
-------
Oi
E
I
CO
CO
-J
O
O
-J
Uj
Uj
CO
Uj
O
N
FIGURE 4-26
EFFECT OF TEMPERATURE ON THICKENING PROPERTIES
OF OXYGEN ACTIVATED SLUDGE AT MLSS = 7000 mg/1
(REFERENCES 86 AND 103)
© SACRAMENTO, CA.
WASHINGTON, O.C.
GJ CITY "A"
WASHINGTON, D.C
SACRAMENTO,CA
10 15 20 25
TEMPERATURE, C
The following conclusions can be drawn about denitrification in secondary clarifiers, based
on work done to date;33,104,105 (j) Rapid sludge removal can prevent sufficient time
being available for nitrogen bubble formation; (2) sludges with low SVI values are preferable
as they can be withdrawn faster; (3) since the saturation level of nitrogen is greater in deep
tanks than laboratory cylinders, bubbles will form and sludges will float faster in the
4-92
-------
laboratory than in the field; (4) there is a minimum concentration of nitrate nitrogen below
which there is insufficient nitrogen to cause flotation. In weak wastewaters or for those
plants in which nitrification is suppressed, sludge flotation will not occur; (5) a drop in
temperature will reduce denitrification rates and may render rising sludge a problem only
under warm conditions; (6) an equivalent amount of nitrite will produce flotation faster
than nitrate, because denitrification is more rapid when nitrite serves as the electron
acceptor rather than nitrate; (7) a sludge with low activity or low rate of denitrification
should be less susceptible to flotation. Sludge activity will vary among the types of
suspended growth processes. For instance, separate stage nitrification sludge is less
susceptible to flotation than combined carbon oxidation-nitrification sludges due to the low
organic loading on the former process and resultant lower activity with respect to carbon
oxidation and denitrification in sedimentation tanks. Among combined carbon oxidation-
nitrification alternatives, the contact stabilization and step aeration alternatives are most
susceptible to sludge flotation due to the enhanced possibility in those modifications of
influent organics being incorporated into the mixed liquor without being oxidized prior to
the clarification step. To a lesser degree, complete mix systems are also prone to sludge
flotation for the same reason.
Control measures for preventing floating sludge should be incorporated into the initial plant
design. Provision of rapid sludge removal (vacuum pickup type) in sedimentation tank
design can prevent there being sufficient contact time for bubble formation to occur and
cause flotation. Flexibility in influent feed points (e.g., allowing switching from step
aeration to plug flow in warm weather periods) can provide the operator with options in
process operation that allow him to get out of difficult operating situations. Provision for
chlorination of the return activated sludge is recommended for all suspended growth
applications. Recent work done in California has shown that continuous low dose
chlorination can be used for controlling sludge bulking and reducing the sludge volume
index, apparently without impairing nitrification (see Section 4.5.1).^9 This allows more
rapid sludge withdrawal from the sedimentation tanks. When nitrification is not required,
higher doses of chlorine can be used to suppress nitrification and thus avoid flotation.
Overdosing chlorine on a slug dose or continuous dose basis should be avoided, however, as
it can cause increases in the level of organics in the process effluent. 104
Control of solids retention time to values below that which will support nitrification has
been a procedure that has been occasionally recommended for cases where nitrification is
not required and sludge flotation is to be prevented. However, when the DO level or pH is
not limiting the nitrification rate, nitrification will proceed at solids detention times as low
as 1.30 days at temperatures equal to or greater than 20 C (Section 3.2.5.4). This renders
the control of nitrification impractical with solids retention time control in warm weather as
stable operation at such a low value for the solids retention time would not be possible. It
may be possible to suppress nitrification through control of DO to levels < 0.5 mg/1 and
thereby limit the nitrifier growth rate to levels which result in washout of the nitrifying
organisms, but this will require accurate around-the-clock control of DO. Further,
maintenance of low DO can cause another operational problem, sludge bulking.
4-93
-------
4.11 Considerations for Process Selection
In selecting nitrification as the process for ammonia removal, two kinds of comparisons can
be made. First, the process can be compared to the physical-chemical alternatives. Second,
alternative nitrification processes can be compared. It is emphasized that no single
alternative will be the best choice for all situations.
4.11.1 Comparison to Physical-Chemical Alternatives
Several factors dictate the choice between biological and physical-chemical techniques for
ammonia removal. Cost is often the single most influential factor in process choice.
Ammonia removal via the nitrification process is widely recognized to be the least costly
ammonia removal alternative. Unless phosphorus removal is also required, the combined
cost of lime precipitation-air stripping is normally greater than the cost for nitrification.
Likewise, breakpoint chlorination and ion exchange are normally more costly than
nitrification. 106
In the majority of situations existing facilities are utilized when treatment is upgraded,
rather than construction of wholly new facilities. The layout of the existing facility may be
more adaptable to one specific alternative or another. In many instances, it has been found
that biological nitrification has been the process most easily incorporated into the upgraded
treatment system.
Very low temperature operation ( < 10 C) may favor a physical-chemical process rather
than a biological process as reaction rates become very low, requiring very large reactors.
Physical chemical processes are also affected by low temperatures, but to a lesser degree.
The presence of compounds toxic to nitrifiers may also dictate against the choice of
nitrification. Some toxicants are resistant (e.g., nonbiodegradable solvents) to most forms of
pretreatment. Unless a very effective source control program can be implemented for these
compounds, dependable operation of nitrification may become impractical.
4.11.2 Choice Among Alternative Nitrification Systems
All of the factors described in Section 4.11.1 are also factors to be considered in selection
among nitrification systems. Other factors affecting choice among nitrification alternatives
are summarized in Table 4-22 as a guide for process selection. Each of these factors is
considered earlier in this chapter.
Higher effluent ammonia (1-3 mg/1) in the attached growth effluents than suspended growth
effluent is cited as a disadvantage of attached growth systems in Table 4-22. However,
breakpoint chlorination' is easily appended to attached growth systems, as the chlorine dose
for breakpoint is low ( < 30 mg/1). The addition of breakpoint chlorination puts attached
growth systems on an equal footing with suspended growth systems with respect to ammonia
control.
4-94
-------
TABLE 4-22
COMPARISON OF NITRIFICATION ALTERNATIVES
System Type
Advantages
Disadvantages
Combined carbon
oxidation - nitrification
Suspended growth
Attached growth
Combined treatment of carbon and
ammonia in a single stage
Very low effluent ammonia
possible
Inventory control of mixed liquor
stable due to high BOD /TKN
ratio
Combined treatment of carbon and
ammonia in a single stage
Stability not linked to secondary
clarlfler as organisms on media
No protection against toxicants
Only moderate stability of operation
Stability linked to operation of
secondary clarifier for biomass return
Large reactors required in cold weather
No protection against toxicants
Only moderate stability of
operation
Effluent ammonia normally 1-3 mg/1
(except RBD)
Cold weather operation impractical
in most cases
Separate stage
nitrification
Suspended growth
Attached growth
Good protection against most
toxicants
Stable operation
Very low effluent ammonia possible
Good protection against most
toxicants
Stable operation
Less sensitive to low temperatures
Stability not linked to secondary
clarifier as organisms on media
Sludge inventory requires careful
control when low BOD.AKN ratio
Stability of operation linked to operation
of secondary clarifier for biomass return
Greater number of unit processes required
than for combined carbon oxidation -
nitrification
Effluent ammonia normally 1-3 mg/1
Greater number of unit processes
required than for combined
carbon oxidation - nitrification
Refinement of process choice may require pilot studies. This is particularly true where
wastewater toxicity may affect the efficacy of nitrification (see Section 4.5.3).
A common issue faced by the engineer when dealing with suspended growth system design is
whether to separate the carbon oxidation stage from the nitrification stage or whether to
provide a combined carbon oxidation-nitrification system. In a recently conducted pilot
study using these systems in parallel at two wastewater temperatures, 8 C and 20C, for the
town of Cheektowaga, New York, it was shown that when the separate nitrification stage
system was operated at the same solids retention time as the combined carbon
oxidation-nitrification system and the same temperature, nitrification effluents of essentially
identical quality were produced. 1 "7
4-95
-------
The investigators concluded that in most cases the combined carbon-oxidation nitrification
system should be chosen for the following reasons: 107
1. Use of the biological solids retention time concept and controlled sludge wasting make
the combined carbon oxidation-nitrification system as controllable as a two stage
suspended growth system. The investigators did not agree with the often stated
concept that separating the stages leads to more positive control of the carbon
oxidation and nitrification functions, as their experimental study demonstrated quite
the opposite.
2. The use of combined carbon oxidation-nitrification results in lower sludge quantities to
be wasted than in a two stage suspended growth system. This is because the first stage
(carbon oxidation) operates at a low solids retention time (say 8 = 2 days) which
L>
results in less solids destruction than when 10 or 20 days are used in the combined
carbon oxidation-nitrification system. This phenomenon has also been observed by
others. 108,109
3. The longer sludge retention times employed in separate stage nitrification systems ( 8
c
= 10 to 20 days) results in improved sludge settling characteristics as compared to high
rate activated sludge systems (at 6 —,2 days).
C
4. A two stage suspended growth system appears to be more prone to control problems
relating to sludge inventory control. The separate stage reactor's sludge inventory must
be maintained by shifting inventory from the first stage, or by some other means
(Section 4.5.2). Further, two sets of sedimentation tanks are required. Sedimentation
tanks are often the most vulnerable components of the activated sludge system. "Thus
it is difficult to envision that the path to increased controllability of nitrifying
activated sludge should lead to a doubling of the least stable element in the process
configuration, i.e.; the clarifier. Rather, if a two stage nitrification system is required, it
appears more reasonable to explore the capabilities of a fixed film nitrification
reactor. . ."107
5. Toxicants affecting nitrifiers present in the raw sewage or primary effluent are often
cited as a reason to provide a high rate activated sludge effluent to act as a toxicant
removal step ahead of a separate stage nitrification unit. First, a detailed evaluation
may show that in fact toxicity is not a problem. "Secondly, toxic materials might
better be excluded from wastewater systems by regulation rather than relying on
'sacrificial' biosystems to protect the nitrifying capability of the system. Thirdly, in
most cases it may become attractive to remove phosphorus in primary treatment by
the addition of coagulants." 107
These conclusions are presented in this manual as an excellent basis for consideration of the
reasons for process selection. In many cases effective counter arguments can be presented.
In the last analysis, the process choice must be made by the local agency and its engineering
4-96
-------
consultant or staff. As an example, counter-arguments are listed in the same order as the
arguments previously presented:
1. The parallel study of carbon oxidation-nitrification and two sludge systems at
Cheektowaga was based primarily on municipal sewage, with only 10 percent industrial
load. 107 Situations do exist \vhere significant industrial contributors of organic load
are tributary to the municipal system. For instance, seasonal canning industries
tributary to California municipalities treatment plants in some instances cause 3 to 4
times the non-canning season organic load at the peak of the canning season. Further,
industrial waste production may vary from year to year depending on factors beyond
control or accurate prediction. In the face of this unpredictable variation it may be
difficult or uneconomic to design a system for combined carbon oxidation-nitrification
due to the sensitivity of nitrification to solids retention time. Less difficulty is
experienced in designing for high rate activated sludge in a two stage system, as
production of effluent of a quality suitable for separate stage nitrification is not as
sensitive to load or solids retention time as is nitrification.
2. The phenomenon of lower sludge production from a two stage suspended growth
system compared to a combined carbon oxidation-nitrification system is well
established and cannot be contested. However, from an energy standpoint, the
greater sludge production from the two stage system may be an advantage for two
reasons. First, the lower sludge production of the combined carbon oxidation-
nitrification system is obtained at the expense of greater power requirements because
greater amounts of oxygen must be supplied for the oxidation of solids. Second, if
anaerobic digestion is employed and the gas recovered for useful energy purposes, less
energy is available from the digestion process when less solids are produced. However,
these two factors may be outweighed by the increased cost of ultimate disposal in
some cases.
3. It has occasionally been found that the longer solids retention times (10 to 20 days)
also result in sludge settling difficulties.
4. With careful monitoring, the separate nitrification stage's inventory requirements can
be managed. In some situations where industries are tributary, the inventory control
problem may be easier with a two stage system (see counter-argument 1 above).
Two sets of sedimentation tanks do present more control requirements than one set.
However, if the conditions for tank upset are present when two sets are provided, they
also can be present for one set. Careful control of sedimentation tank operation is
mandatory in either case.
5. In the case of the parallel study at Cheektowaga, there was no evidence of significant
toxicants in the primarily domestic sewage processed. 107 So perhaps it is natural to
discount the difficulty in dealing with this problem.
4-97
-------
The unfortunate aspect of the presence of toxicity in wastewater is that it is often
ephemeral in nature; i.e., it's there and then it's gone. Toxicants discharged on a
continuous basis are handled with relative ease compared to the occasional dump.
Unless the problem is recognized in its earliest stage, the causative agent may not even
be sampled by plant personnel, rendering it impossible to trace. Even if a sample is
caught, tracing it back through the system is not always possible until the next
occurrence. Another aspect of the problem is that the dumps offer no opportunity for
the biomass to adapt to the toxicant, whereas if it were continuously present,
adaptation of the nitrifiers might be possible (see Section 3.2.9).
In "bases where toxicants are occasionally present, the issue boils down to the need for
plant reliability. In cases of discharge to an estuary or groundwater where mixing in the
environment causes dilution of the effluent, occasional process failures may be
accepted. However, where stringent regulatory requirements exist or where the water is
reused and the water user demands the consistent performance expected of a water
utility, some compensation must be made to handle the problem of toxic upsets. This
may be done by any number of means, one of which is to provide pretreatment via
chemical addition or by a biological treatment stage (Section 4.5.3). Another is to
provide supplemental breakpoint chlorination at the end of the system. The cost of the
latter facility is very much affected by the degree of upset in nitrification expected.
In the last analysis, the parallel study at Cheektowaga showed that combined carbon
oxidation-nitrification could be just as reliable as separate stage nitrification at low
temperatures (8 C) with a primarily domestic wastewater.107 The study provides further
proof dispelling the poor reputation that the combined carbon oxidation-nitrification
system has acquired in the U.S. Its wide application in England where it is coupled with a
toxicity source control program offers additional testimony to the efficacy of the process.
Similar lengthy discussions could be prepared which present the pros and cons of other
systems: e.g., rotating biological discs versus plastic media trickling filters. These
comparisons not only are beyond the scope of this manual but would not have general
validity. It is a hazardous task to make such an attempt as the specifics of individual
circumstances affect the decisions to a large degree. There is no universally best nitrification
approach. Rather, the broad variety of alternatives should be viewed as a positive situation.
The fact that there are many alternatives makes the task of adapting nitrification into waste
treatment easier, not harder. A myriad of flowsheets incorporating nitrification are not only
possible, but economically feasible and with proper design and operation, quite reliable.
4-98
-------
4.12 References
1. Mulbarger, M.C., The Three Sludge System for Nitrogen and Phosphorus Removal.
Presented at the 44th Annual Conference of the Water Pollution Control Federation,
San Francisco, California, October, 1971.
2. City of Los Angeles, California, Hyperion Treatment Plant West Battery Operating
Reports. January through March, 1969.
3. Horstkotte, G.A., Niles, D.G., Parker, D.S., and D.H. Caldwell, Full-Scale Testing of a
Water Reclamation System. JWPCF, 46, No. 1, pp 181-197 (1974).
4. Weddle, C.L., Niles, D.G., Goldman, E., and J.W. Porter, Studies of Municipal
Wastewater Renovation for Industrial Water. Presented at the 44th Annual Conference
of the Water Pollution Control Federation, San Francisco, California, October, 1971.
5. Loftin, W.E., Annual Report, Livermore Water Reclamation Plant, 1970. City of
Livermore, California, March, 1971.
6. Beckman, W.J., Avendt, R.J., Mulligan, T.J., and G.J. Kehrberger, Combined Carbon
Oxidation-Nitrification. JWPCF, 44, No. 10, pp 1916-1931 (1972).
7. Stamberg, J.B., Hais, A.B., Bishop, D.F., and J.A. Heidman, Nitrification in Oxygen
Activated Sludge. Unpublished paper, Environmental Protection Agency, 1974.
8. Heidman, J.A., An Experimental Evaluation of Oxygen and Air Activated Sludge
Nitrification Systems With and Without pH Control. EPA report for Contract No.
68-03-0349, 1975.
9. County Sanitation Districts of Los Angeles County, Monthly Operating Reports,
Whittier Narrows Water Reclamation Plant. April, 1973 to March, 1974.
10. Greene, R.A., Complete Nitrification by Single Stage Activated Sludge. Presented at
the 46th Annual Conference of the Water Pollution Control Federation, Cleveland,
Ohio, October, 1973.
11. Newton, D., and T.E. Wilson, Oxygen Nitrification Process at Tampa. In Applications
of Commercial Oxygen to Water and Wastewater Systems, Ed. by R.E. Speece and J.F.
Malena, Jr., Austin, Texas: The Center for Research in Water Resources, 1973.
12. Tenney, M.W., and W.F. Echelberger, Removal of Organic and Eutrophying Pollutants
by Chemical-Biological Treatment. Prepared for the EPA, Report No. R2-72-076
(NTISPB-214628), April, 1972.
4-99
-------
13. Black, S.A., Lime Treatment for Phosphorus Removal at the New Market/East
Gwillimbury WPCF. Paper No. W3032, Ontario Ministry of the Environment, Research
Branch, May, 1972.
14. Schwer, A.D., Letter communication to D.S. Parker — Metropolitan Sewer District of
Greater Cincinnati, March 9, 1971.
15. Barth, E.F., Brenner, R.C., and R.F. Lewis, Chemical-Biological Control of Nitrogen in
Wastewater Effluent. JWPCF, 40, No. 12, pp 2040 - 2054 (1968).
16. Rimer, A.E.> and R.L. Woodward, Two Stage Activated Sludge Pilot Plant Operations,
Fitchburg, Massachusetts. JWPCF, 44, No. 1, pp 101-116 (1972).
17. Wild, H.E., Jr., Letter communication to D.S. Parker. Briley, Wild and Associates,
Ormond Beach, Florida, September 9, 1974.
18. Union Carbide Corporation, "UNOX" System Study at Town ofAmherst, New York.
1972.
19. Linstedt, K.D., and E.R. Bennett, Evaluation of Treatment for Urban Wastewater
Reuse. Report prepared for the Environmental Protection Agency, EPA-R2-73-122,
July, 1973.
20. Linstedt, K.D., Letter communication to D.S. Parker. University of Colorado, Boulder,
Colorado, August 5, 1974.
21. Stenquist, R.J., Parker, D.S., and T.J. Dosh, Carbon Oxidation-Nitrification In
Synthetic Media Trickling Filters. JWPCF, 46, No. 10, pp 2327-2339 (1974).
22. Duddles, G.A., Richardson, S.E., and E.F. Barth, Plastic Medium Trickling Filters for
Biological Nitrogen Control. JWPCF, 46, No. 5, pp 937-946 (1974).
23. McHarness, D.D., Haug, R.T., and P.L. McCarty, Field Studies of Nitrification with
Submerged Filters. JWPCF, 47, No. 2, pp 291-309 (1975).
24. McHarness, D.D., and P.L. McCarty, Field Study of Nitrification with Submerged
Filter. Report prepared for the EPA, EPA-R2-73-158, February, 1973.
25. Process Design Manual for Upgrading Existing Wastewater Treatment Plants. U.S. EPA,
Office of Technology Transfer, Washington, D.C. (1974).
26. Richardson, S.E., Pilot Plants Define Parameters for Plastic Media Trickling Filter
Nitrification. Presented at the 46th Annual Conference of the Water Pollution Control
Federation, Cleveland, Ohio, October, 1973.
4-100
-------
27. Sampayo, Felix F., The Use of Nitrification Towers at Lima, Ohio. Presented at the
Second Annual Conference Water Management Association of Ohio, Columbus, Ohio,
October, 1973.
28. Brenner, R.C., EPA Experiences in Oxygen Activated Sludge. Prepared for the EPA
Technology Transfer Program, October, 1973.
29. Lawrence, A.W., and P.L. McCarty, Unified Basis for Biological Treatment Design and
Operation. JSED, Proc. ASCE, 96, No. SA3, pp 757-778 (1970).
30. Hanson, R.L., Walker, W.C., and J.C. Brown, Variations in Characteristics of
Wastewater Influent at the Mason Farm Wastewater Treatment Plant, Chapel Hill, No.
Carolina. Report No. 13, UNC Wastewater Research Center, Chapel Hill, N.C.,
February, 1971.
31. Lijklema, L., A Model for Nitrification in the Activated Sludge Process. ESE
Publication No. 303, Department of Environmental Sciences and Engineering,
University of North Carolina, June, 1972.
32. Poduska, R.A. and J.F. Andrews, Dynamics of Nitrification in the Activated Sludge
Process. Presented at the 29th Industrial Waste Conference, Purdue University,
Lafayette, Indiana, May 7-9, 1974.
33. Nagel, C.A. and J.G. Haworth, Operational Factors Affecting Nitrification in the
Activated Sludge Process. Presented at the 42nd Annual Conference of the Water
Pollution Control Federation, Dallas, Texas, October (1969), (available as a reprint
from the County Sanitation Districts of Los Angeles County).
34. Murphy, K.L., and P.L. Timpany, Design and Analysis of Mixing for an Aeration Tank.
JSED, Proc. ASCE, 93, No. SA5, pp 1-15 (1967).
35. Murphy, K.D., and B.I. Boyko, Longitudinal Mixing in Spiral Flow Aeration Tanks;
JSED, Proc. ASCE, 96, No. SA2, pp 211-221 (1970).
36. Metcalf and Eddy, Inc., Wastewater Engineering. New York, McGraw Hill Book Co.,
1972.
37. Gujer, W. and D. Jenkins, A Nitrification Model for the Contact Stabilization Activated
Sludge Process. Water Research, in press.
38. Gujer, W. and D. Jenkins, The Contact Stabilization Process-Oxygen and Nitrogen Mass
Balances. University of California, Sanitary Engineering Research Lab, SERL Report
74-2, February, 1974.
4-101
-------
39. Sawyer, C.N., Letter communication to D.S. Parker; January 24, 1975.
40. Sawyer, C.N., Activated Sludge Oxidations, HI. Factors Involved in Prolonging the
High Initial Rate of Oxygen Utilization by Activated Sludge-Mixtures. Sewage Works
Journal, 11, No. 4, pp 595-606 (1939).
41. County Sanitation Districts of Los Angeles County, A Plan for Water Reuse. Report
prepared for the members of the Boards of Directors, July, 1963.
42. City of Jackson, Michigan, Sewage Treatment Plant Operating Reports. August, 1973
to March, 1974.
43. Bruce, A.M., and J.C. Merkins, Further Studies of Partial Treatment of Sewage by High
Rate Biological Filtration. Water Pollution Control (London), pp 499-527, 1973.
44. Grantham, G.R., Phelps, E.B., Calaway, W.T., and D.L. Emerson, Progress Report on
Trickling Filter Studies. Sewage and Industrial Wastes, 22, No. 7, pp 867-874 (1950).
45. Grantham, G.R., Trickling Filter Performance at Intermediate Loading Rates. Sewage
and Industrial Wastes, 23, No. 10, pp 127-1234 (1951).
46. Burgess F.J., Gilmour, C.M., Merryfield, F., and J.K. Carswell, Evaluation Criteria for
Deep Trickling Filters. JWPCF, 33, No. 8, pp 787-799 (1961).
47. Osbom, D.E., Operating Experiences with Double Filtration in Johannesburg. J. Inst.
of Sew. Purif., Part 3, pp 272-281 (1965).
48. Stones, T., Investigation on Biological Filtration at Salford. Journal of the Institute of
Sewage Purification, No. 5, pp 406-417 (1961).
49. Mohlman, F.W., Norgaard, J.T., Fair, G.M., Fuhrman, R.E., Gilbert, J.J., Heacox, R.E.,
and C.C. Ruchoft, Sewage Treatment at Military Installations. Sewage Works Journal,
18, No. 4, pp 789-1028 (1946).
50. Heukelekian, H., The Relationship Between Accumulation, Biochemical and Biological
Characteristics of Film and Purification Capacity of a Biofllter and a Standard Filter.
Sewage Works Journal, 17, No. 3, pp 516-524(1945).
51. Brown and Caldwell, Report on Pilot Trickling Filter Studies at the Main Water Quality
Control Plant. Prepared for the City of Stockon, California, March, 1973.
52. Antonie, R.L., Three Step Biological Treatment with the Bio-Disc Process. Presented at
the New York Water Pollution Control Association, Spring Meeting, Mantauk, New
York, June, 1972.
4-102
-------
53. Antonie, R.L., Nitrification and Denitrification with the Bio-Surf Process. Presented at
the Annual Meeting of the New England W.P.C. Association in Kennebunkport, Maine,
June 10-12, 1974.
54. Rotating Biological Disk Wastewater Treatment Process — Pilot Plant Evaluation.
Report by the Department of Environmental Sciences, Rutgers University, prepared
for the Environmental Protection Agency, Project No. 17010 EBM, August, 1972.
55. Brown and Caldwell, Consulting Engineers, Lime Use in Wastewater Treatment: Design
and Cost Data. Report submitted to the U.S. Environmental Protection Agency, 1975.
56. Process Design Manual for Carbon Adsorption. U.S. EPA, Office of Technology
Transfer, Washington, D.C. (1974). *
57. Rainwater, F.H. and L.L. Thatcher, Methods for Collection .and Analysis of Water
Samples. Geological Survey Water-Supply Paper 1454, USGPO, 1960.
58. Wood, O.K. and G. Tchbanoglous, Trace Elements in Biological Waste Treatment with
Specific Reference to the Activated Sludge Process. Presented at the 29th Industrial
Waste Conference, Purdue University, May, 1974. ,
59. Eisenhauer, D.L., Sieger, R.B. and D.S. Parker, Design of an Integrated Approach to
Nutrient Removal. Presented at the BED- ASCE Specialty Conference, Penn. State
University, Pa., July, 1974.
60. Environmental Quality Analysts, Inc., Letter Report to Valley Community Services
District, March, 1974.
61. Stensil, H.D., Oases Wastewater Characterization Study, Chattanooga Moccasin Bend
Wastewater Treatment Plant. Report prepared by Air Products and Chemicals, Inc.,
1975.
62. Sawyer, C.N., Wild, H.E., Jr., and T.C. McMahon, Nitrification and Denitrification
Facilities, Wastewater Treatment. Prepared for the EPA Technology Transfer Program,
August, 1973.
63. Parker, D.S., Case Histories of Nitrification and Denitrification Facilities. Prepared for
the EPA Technology Transfer Program, May, 1974.
64. Mulbarger, M.C., Private communication to D.H. Caldwell, 1971.
65. Schwinn, D.E., Treatment Plant Designed for Anticipated Standards. Public Works,
104, No. l,pp 54-57 (1973).
4-103
-------
66. Wilson, T.E., and M.D.R. Riddel, Nitrogen Removal: Where Do We Stand? Water and
Wastes Engineering, 11, No. 10, pp 56-61 (1974).
67. Wilcox, E.A., and A.A. Thomas, Oxygen Activated Sludge Wastewater Treatment
Systems, Design Criteria and Operating Experience. Prepared for the EPA Technology
Transfer Program, August, 1973.
«.
68. Sorrels, J.H., and P.J.A. Zeller, Two-Stage Trickling Filter Performance. Sewage and
Industrial Wastes, 18, No. 8, pp 943-954 (1956).
69. Huang, C.S., Kinetics and Process Factors of Nitrification On a Biological Film
Reactor. Thesis submitted in partial satisfaction of the requirements for the degree of
Doctor of Philosophy, University of New York at Buffalo, 1973.
70. Williamson, K.L. and P.L. McCarty, A Model of Substrate Utilization by Bacterial
Films. Presented at the 46th Annual Conference of the Water Pollution Control
Federation, Cincinnati, Ohio, October, 1973.
71. Duddles, G.A. and S.E. Richardson, Application of Plastic Media Trickling Filters for
Biological Nitrification. Report prepared for the Environmental Protection Agency,
EPA- R2-73-199, June, 1973.
72. Bruce, A.M., and J.C. Merkens, Recent Studies of High-Rate Biological Filtration.
Water Pollution Control, pp 113-148 (1970).
73. Brown and Caldwell, Report on Tertiary Treatment Pilot Plant Studies. Prepared for
the City of Sunnyvale, California, February, 1975.
74. Antonie, R.L., Nitrification of Activated Sludge Effluent with the Bio-Surf Process.
Presented at the Annual Conference of the Ohio Water Pollution Control Association,
Toledo, Ohio, June 7-13, 1974.
75. Haug, R.T., and P.L. McCarty, Nitrification with the Submerged Filter. JWPCF, 44, p
2086(1972).
76. Haug, R.T., and P.L. McCarty, Nitrification with the Submerged Filter. Report
prepared by the Department of Civil Engineering, Stanford University, for the
Environmental Protection Agency, Research Grant No. 17010 EPM, August, 1971.
77. Young, J.C., Baumann, E.R., and D.J. Wall, Packed-Bed Reactors for Secondary
Effluent BOD and Ammonia Removal. JWPCF, 47, No. 1, pp 46-56 (1975).
78. General Filter Co., Paktor, Packed Bed Reactor, Bulletin No. 7305, 1974.
4-104
-------
79. Mechalas, B.J., Allen, P.M. and W.W. Matyskiela, A Study of Nitrification and
Dehitrification. A report prepared for the Federal Water Quality Administration,
WPCRS 17010 DRD 07/70, July, 1970.
80. Young, J.C. and M.C. Stewart, Advanced Wastewater Treatment with Packed Bed
Reactors. Report of the Engineering Research Institute, Iowa State University, No.
ERI-73108,May, 1973.
81. Gasser, J.A., Chen, C.L., and R.P. Miele, Fixed-Film Nitrification of Secondary
Effluent. Presented at the EED-ASCE Specialty Conference, Penn. State University,
Pa., July, 1974.
82. Kenney, F.R., Letter communication to D.S. Parker. General Filter Co., Ames, Iowa,
September, 1974.
83. Young, J.C., Unpublished data, September 27, 1974.
84. Smith, J., Personal communication to D.S. Parker. Environmental Protection Agency,
Cincinnati, Ohio, March, 1974.
85. Loftin, W.E., Personal communication to D.S. Parker. City of Livermore, California,
April, 1972.
86. Sacramento Area Consultants, Oxygen Activated Sludge Pilot Studies for the
Sacramento Regional Treatment Plant. Report prepared for the Sacramento Regional
County Sanitation District of Sacramento County, California, July, 1974.
87. Bishop, D.F., Personal communication to D.S. Parker. Environmental Protection
Agency, Washington, D.C., April, 1974.
88. Commonwealth Department of Works, Australian Capital Territory, letter communi-
cation to R.C. Aberley, dated June 1 and July 20, 1972.
89. Metropolitan St. Louis Sewer District and Havens and Emerson, Ltd., Cost-Effective
Design of Wastewater Treatment Facilities Based on Field Derived Parameters.
Prepared for the EPA, Report No. EPA- 670/2-74-062 (PB-234 356), July, 1974.
90. Aberley, R.C., Rattray, G.B. and P.P. Dougas, Air Diffusion Unit. JWPCF, 46, No. 5,
pp 895-910 (1974).
91. Leary, R.D., Ernest, L.A., and W.J. Katz, Full Scale Oxygen Transfer Studies of Seven
Diffuser Systems. JWPCF, 41, No. 3, pp 459^73 (1969).
4-105
-------
92. Nogaj, R.J., Selecting Wastewater Aeration Equipment. Chemical Engineering, April,
1972.
93. City of Medford, Oregon, Sewage Treatment Plant Operating Reports. July, 1974.
94. San Pablo Sanitary District, California, Wastewater Treatment Plant Operating Reports.
June, 1973, to July, 1974.
95. Process Design Manual for Phosphorus Removal. U.S. EPA, Office of Technology
Transfer, Washington, D.C. (1971).
96. Process Design Manual for Suspended Solids Removal. U.S. EPA, Office of Technology
Transfer, Washington, D.C., January, 1975.
97. Cleasby, J.L., and E.R. Baumann, Wastewater Filtration, Design Considerations.
Prepared for the EPA Technology Transfer Program, July, 1974.
98. Dick, R.I., Role of Activated Sludge Final Settling Tanks. JSED, Proc. ASCE, 96, No.
SA2,pp 423-436 (1970).
f
99. Dick, R.I. and A.R. Javaheri, Discussion of Unified Basis for Biological Treatment
Design and Operation by A.W. Lawrence and P.L. McCarty. JSED, Proc. ASCE, 97,
SA2, pp 234-238 (1971).
100. Dick, R.I. and K.W. Young, Analysis of Thickening Performance of Final Settling
Tanks. Presented at the 27th Industrial Waste Conference, Purdue University,
Lafayette, Indiana, May 7-9, 1974.
101. Tenney, M.W. and W. Stumm, Chemical Flocculation of Microorganisms in Biological
Waste Treatment. JWPCF, 37, p. 1370 (1965).
102. Parker, D.S., Kaufman, W.J., and D. Jenkins, Physical Conditioning of Activated
Sludge Floe. JWPCF, 43, No. 9, pp 1817-1833 (1971).
103. Stamberg, J.B., Bishop, D.F., Hais, A.B., and S.M. Bennet, System Alternatives in
Oxygen Activated Sludge. Presented at the 45th Annual Conference of the WPCF,
Atlanta, Ga., 1972.
104. Sawyer, C.N., and L. Bradney, Rising of Activated Sludge in Final Settling Tanks.
Sewage Works Journal, 17, No. 6, pp. 1191-1209 (1945).
105. Clayfield, G.W., Respiration and Denitrification Studies on Laboratory and Works
Activated Sludges. Water Pollution Control, London, 73, No. 1, pp 51-76 (1974).
4-106
-------
106. Stone, R.W., Parker, D.S., and J.A. Cotteral, Upgrading Lagoon Effluent to Meet Best
Practicable Treatment. Presented at the 47th Annual Conference of the Water
Pollution Control Federation, Denver, Colorado, October, 1974.
107. Lawrence, A.M., and C.G. Brown, Biokinetic Approach to Optimal Design and Control
of Nitrifying Activated Sludge Systems. Presented at the Annual Meeting of the New
York Water Pollution Control Association, New York City, January 23, 1973.
108. Stall, T.R., and R.H. Sherwood, One Sludge or Two Sludge? Water and Wastes
Engineering, p 41-44, April, 1974.
109. Sutton, P.M., Murphy, K.L., and B.E. Tank, Biological Nitrogen Removal - The
Efficacy of the Nitrification Step. Presented at the WPCF Conference, Denver,
October, 1974.
4-107
-------
CHAPTER 5
BIOLOGICAL DENITRIFICATION
5.1 Introduction
The process of biological denitrification is applicable to the removal of nitrogen from
wastewater when the nitrogen is predominately in the nitrate or nitrite form. In municipal
applications, the nitrogen in the raw wastewater is primarily present as organic and
ammonia-nitrogen and first mus,t be converted to an oxidized form (nitrite or nitrate) prior
to biological denitrification. The biological oxidation process used for this conversion,
nitrification, was described in Chapters 3 and 4.
This chapter presents design criteria for several alternative denitrification systems including
suspended growth and attached growth systems using methanol as the carbon source and
combined carbon oxidation-nitrification-denitrification systems using wastewater or endo-
genous carbon sources. The basic chemistry of denitrification was described in Section 3.3.
5.2 Denitrification in Suspended Growth Reactors Using Methanol as the Carbon Source
The suspended growth denitrification process is a form of the activated sludge process.
There are several differences between its typical application for organic carbon removal and
in its use for denitrification. In common is the provision of a reactor, in which the biomass is
kept in suspension in the liquid by mixing. Also provided in both applications is a
sedimentation tank for separation of the mixed liquor solids from the effluent, allowing the
biomass to be recycled in the system and also allowing the production of a clear effluent for
discharge or subsequent treatment. Two typical suspended growth denitrification systems
are illustrated in Figure 5-1.
There are other analogies'between suspended growth systems used for denitrification and
organic carbon removal. In organic carbon removal applications, dissolved oxygen is
introduced into the reactor by aeration so that biological oxidation of the organic matter
can take place. In the process of carbon oxidation, oxygen is consumed as the electron.
acceptor in the oxidation process. In the process of denitrification, carbon (usually
methanol) is oxidized with nitrate or nitrite serving as the electron acceptor (see Section
3.3.2). In denitrification as opposed to organics removal, it is the nitrate that is the
pollutant that is to be removed and the carbon source that is added. In organics removal, it
is the carbon that is the pollutant that is to be removed and the oxygen that is added.
•Needless to say, only sufficient carbon (such as methanol) is added in denitrification to
accomplish the nitrate removal, as excess dosing causes organics to appear in the effluent
unless control measures are undertaken. These residual organics, if left in the effluent,
would exert BOD5 and might cause violation of effluent requirements.
5-1
-------
FIGURE 5-1
SUSPENDED GROWTH DENITRIFICATION SYSTEMS USING METHANOL
A. ORIGINAL DENITRIFICATION SYSTEM (Reference))
METHANOL
NITRIFIED i
EFFLUENT
ANOXIC MIXED
DENITRIFICATION REACTOR
CLARIFIER /DENITRIFIED
EFFLUENT
-AERATED
NITROGEN STRIPPING
CHANNEL T = 5 min.
RETURN DENITRIFIED SLUDGE
B. MODIFIED DENITRIFICATION SYSTEM (Reference 2,3)
METHANOL
i
NITRIFIED I
EFFLUENT
i
A NO* 1C M 1 y FTJ
nFMlTDIFIPATIf^M D IT A f* T A D
AERATED
O T A R 1 1 1 7 A T 1 /"* M
T A M Vf
T = 50 min.
^\
, ^ _ ^/nFwiTRiFirATinNJ ^
\ Cl ARIFIFR /PFNITRIFIED
/ \ / EFFLUENT
^-MILDLY AERATED
PHYSICAL CONDITIONING
CHANNEL i'
RETURN DENITRIFIED SLUDGE
-------
5.2.1 Denitrification Rates
Currently used denitrification rate data for design of denitrification systems are based
upon work described in references 4, 5, 6, and 7. Data from these investigations are
summarized in Figure 5-2. Rather than show individual data points, or trend lines, Figure
5-2 shows boundaries for the data so that the range in variation of denitrification rates can
be inspected. Earliest available measurements were those from Manassas, Va.4 which have
been found to be considerably higher than subsequent observations at three other locations.
The data from CCCSD, Ca., Blue Plains, D.C., and Burlington, Canada, all are in reasonable
agreement with each other, and are all well below the Manassas rates. A possible reason for
FIGURE 5-2
OBSERVED DENITRIFICATION RATES FOR SUSPENDED
GROWTH SYSTEMS USING METHANOL
CO
to
-J
5
.a
\
o
0)
k.
a
•o
N
X
O
O
ft:
.o
*.
O
-------
the higher measurements at Manassas is that an acid solubilization procedure was employed
prior to the volatile solids determination, which may have acid hydrolized organic matter
resulting in low measurements of volatile solids (and higher apparent denitrification rates).
The earlier work at Manassas^ prior to implementation of the" acid step showed
denitrification rates closer to the observations of other locations.
Laboratory studies on synthetic nitrate containing wastes have shown much higher
denitrification rates than are found in Figure 5-2.10,11 However, the biological solids
developed in such laboratory systems do not contain the levels of refractory solids that
build up in practical systems operated under field conditions. Therefore, the data developed
from such laboratory studies are not directly useful in establishing accurate design
parameters.
Conditions maintained during the field studies may influence field measured denitrification
rates considerably. When the denitrification reactor is continuously operated close to the
maximum growth rate of the denitrifying organisms, it is probable that the denitrifying
activity of the biomass is higher than when the system is operated with a high safety factor.
For instance, in studies in Ontario it was found that measured peak denitrification rates
were approximately 30 percent greater at a solids retention time of 3 days than rates
measured at a solids retention time of 6 days. 12 Thus, differences in measured rates may
reflect variations in operating conditions among the various locations.
Observed rates in Figure 5-2 are essentially the experimentally determined values of the
term qp using the notation presented in Section 3.3.5.2. The peak nitrate removal rate, qrj,
is the reaction rate when neither methanol nor nitrate is limiting the reaction rate. Sub-
sequent sections show how these peak nitrate removal rates are used in design calculations.
5.2.2 Complete Mix Denitrification Kinetics
The equations presented in Section 3.3.5 are directly applicable to the design of complete
mix 'denitrification systems. The design procedure for denitrification uses the safety factor
concept to relate peak nitrate removal rates, qj), to design nitrate removal rates, qp.
Expressed in terms of solids retention time, the safety factor concept is:
- (3-29)
c
where: SF = safety factor,
6 = solids retention time of design, days, and
C
6 = minimum solids retention time, days, for
denitrification,
5-4
-------
The design nitrate removal rate, qp, can be related to the safety factor and the peak nitrate
removal rate by using the following equations in conjunction with Equation 3-53:
where: YD = denitrifier gross yield, Ib VSS grown/lb NO" - N
removed,
K , = decay coefficient, day" ,
qD = nitrate removal rate, Ib NOl -N rem./lb VSS-day, and
q~ = peak nitrate removal rate, Ib NOl -N rem./lb VSS-day.
In evaluating these equations in design calculations, the specific values of Y and Kj given in
Section 3.3.5.4 may be used. Figure 5-2 may be used to arrive at estimated values of qry
Considering the range in the data in Figure 5-2, conservative practice in the absence of pilot
data would be to pick the lowest denitrification rates observed for qrj, e.g. at 10 C qrj =
0.05, at 15 C qD = 0.08, at 20 C qD = 0.15, and at 25 C qD = 0.20. Use of these minimum
values of qjj> will result in very conservative reactor designs. Pilot plant studies may be useful
to define applicable values of qpj, as the potential for establishing higher denitrification rates
for a particular location is good; evidence of this is the range of denitrification rates among
the various locations shown in Figure 5-2.
/\ _
As a design example consider a case where the temperature is 25 C, qp_) = 0.2 Ib NC>3 rem./lb
MLVSS/day, Y = 0.9 Ib VSS/lb NO^ rem., Kj = 0.04 day"1, and KD = 0.15 mg/1. Assume
that due to diurnal variations in load (Section 5.2.2.2), a minimum safety factor of 2.0 is
adopted. Consider a 30 mgd treatment plant, where 25 mg/1 of nitrate-N must be removed.
1. Using Equation 3-54, calculate the minimum solids retention time for denitrifica-
tion:
-i- =0.9(0.2) -0.04 = 0.1 4
0m = 7.14 days
2. Calculate the design solids retention time (Equation 3-29):
5-5
-------
0 = 2.0(7.14) = 14.3 days
c
3. Calculate the design nitrate removal rate (Equation 3-50):
,-43=^0-0.04
••• qD = 0. 1 2 Ib NO~ -N rem./lb MLVSS/day
4. Calculate the steady state nitrate content of the effluent. The expression relating
removal rates to nitrate level, from Equations 3-47, 3-48 and 3-49, is as follows:
where: K^ = half saturation constant, mg/1 NO- -N, and
D, = effluent concentration of nitrate nitrogen mg/1.
Evaluation of this equation for this example yields:
Dl
0.12 = 0.20
0.15+ DJ
D] =0.23 mg/1 NO3-N
5. Determine the hydraulic detention time at average dry weather flow. The
equation for nitrate removal rate is useful in this calculation.
D-D,
where: D = influent NOl -N, mg/1
D. = effluent NO--N, mg/1
Xj = MLVSS, mg/1, and
HT = hydraulic detention time, days.
5-6
-------
The mixed liquor volatile suspended solids (MLVSS) level is dependent on the
mixed liquor total suspended solids, which is in turn dependent on the operation
of the denitrification sedimentation tank (see Sections 4.10 and 5.6). Assume for
the purposes of this example that the design mixed liquor content at 25 C is 3000
mg/1. At a volatile content of 80 percent, the MLVSS is 2400 mg/1. From
Equation 5-2, the hydraulic detention time is:
= 1.99hr
6. Determine the sludge wasting schedule. The equations developed for wasting
in the nitrification system are directly applicable here. The necessary sludge
inventory is:
I = 8.33(X1 • V) (4-7)
where: I = inventory of VSS in the anoxic denitrification
reactor, Ib,
X. = MLVSS in the reactor, mg/1, and
V = volume of the reactor, mil gal.
In the example at hand:
I = 8.33(2400)(0.083)(30) = 49,780 Ib VSS
From Equation 4-8, the sludge wasting from all sources is defined by the
following equation:
S= (5-3,
C
where: S = total sludge wasted in Ib/day
For this example:
S = 49,780/14.3 = 3,481 Ib VSS/day
The total sludge to be wasted each day is made up of two components, as shown:
S = 8.33(Q-X2+W-Xw) (4-6)
5-7
-------
where: Q = influent (or effluent) flow rate, mgd,
W = waste sludge flow rate, mgd,
X^ = effluent volatile suspended solids, mg/1, and
X = waste sludge volatile suspended solids, mg/1.
iV
The sludge contained in the effluent (the term Q-X2 above) can be calculated
assuming that the effluent VSS is 10 mg/1:
8.33(10)(30) = 2,499 Ib VSS/day
By difference, the Ib of MLVSS to be wasted from the mixed liquor or return
sludge is:
3,481 - 2,499 = 982 Ib VSS/day
7. Methanol requirement. From Section 3.3.2, an estimate of 3.0 Ib per Ib of nitrate
N removed is reasonable. The methanol requirement is:
3.0(25 - 0.23)(8.33)(30) = 18,570 lb/day
The sludge yield and decay values used above are for a case where only a short aeration
period is used prior to clarification. When an aerated stabilization step is employed, very
much lower sludge wasting is required than presented in the above example. In cases where
an aerated stabilization tank is employed, only the sludge inventory under anoxic conditions
should be considered in the sizing of the anoxic reactor for denitrification (Step 5).
5.2.2.1 Effect of Safety Factor on Steady-State Effluent Quality
In the design example previously presented, the safety factor was assumed to be 2.0, based
on considerations presented in Section 5.2.2.2. The effect of alternative assumptions on the
effluent nitrate level are presented in Figure 5-3. As may be seen, the assumption of the
safety factor has a marked effect on the effluent quality of complete mix denitrification
systems.
5'.2.2.2 Effect of Diurnal Load Variations on Effluent Quality
As is the case for nitrification, load variations have a significant effect on effluent quality.
Since upstream treatment units to some extent equalize load variations, the peak to average
load ratio is generally lower than for the nitrification stage. The effect of load
variations can be analyzed in a similar manner to that used for nitrification (Section
4.3.3.2). By analogy to the nitrification case, the mags balance yields the following
expression for nitrate at any time:
5-8
-------
FIGURE 5-3
EFFECT OF SAFETY FACTOR ON EFFLUENT NITRATE
LEVEL IN SUSPENDED GROWTH SYSTEM
COMPLETE MIX
SAFETY FACTOR, SF
3.0
(5-4)
where: D = influent nitrate -N level at any time, mg/1,
D = mass average influent nitrate -N level over
24 hours, mg/1,
D^ = mass average effluent nitrate -N level at any time, mg/1,
D , = mass average effluent nitrate -N level over 24 hours, mg/1,
1 o
Q = influent flow rate at any time, °
Q = average daily influent flow rate, 0
jz = design denitrifier growth rate, day" , and
jx = maximum denitrifier growth rate, day
5-9
"
-------
Equation 5-4 has been used to evaluate the effect of the diumal load variations shown on
Figure 5-4 using the design example conditions given in Section 5.2.2.1. The influent
nitrate-nitrogen concentration was assumed constant at 25 mg/1, and the load variation was
assumed to be due to variation in flow only. Several trial calculations using Equation 5-4
over a 24 hour cycle were necessary to derive values of Dj. Results of these calculations for
several values of the SF are also shown on Figure 5-4. As may be seen, the safety factor, SF,
has a marked effect on the nitrate bleedthrough occurring during peak load conditions. In
the case examined, a safety factor of 2.0 was sufficient to prevent excessive nitrate leakage.
The peak to average load for this case was 1.5. In summary, it would appear that as a
minimum, the safety factor should exceed the peak to average load ratio to prevent
excessive nitrate leakage during peak load conditions.
5.2.3 Plug Flow Denitrification Kinetics
The design approach for plug flow denitrification reactors is similar to the approach
developed for complete mix reactors, with the exception of the equations used to predict
effluent quality. Lawrence and McCarty's^ solution for plug flow kinetics is applicable:
YDqD(D -D)
— D -Kd <5-5>
All these terms are as defined previously in Section 5.2.2.
The kinetic design approach for plug flow follows that used for complete mix systems in
Section 5.2.2, excepting that at step 4, Equation 5-5 is used instead of Equation 5-1 to find
the effluent nitrate level.
Equation 5-5 can also be used to find the safety factor required to obtain any desired nitrate
level. This has been done for the example presented in Section 5.2.2.1 and plotted in Figure
5-3. A comparison of the safety factor required to obtain the same nitrate level in a
complete mix system yields the conclusion that plug flow systems can be designed with
considerably lower safety factors while obtaining the same effluent quality.
While kinetic models have not been extended to the point where they can be expected to
describe the effect of diurnal variations on plug flow systems, it can be expected that the
effects of these loads will be similar to those experienced in complete mix systems. This is a
result of the fact that once the effluent level rises above 1 mg/1 nitrate -N, the
denitrification rate becomes essentially zero order. For zero order reactions, there is little
difference between plug 'flow and complete mix reaction kinetics. Therefore, the nitrate
bleedthrough in a plug flow reactor can be expected to closely approach that in a complete
mi-: reactor under diurnal peak load conditions. To prevent excessive nitrate leakage during
peak load conditions, the recommendation in Section 5.2.2.2 should be adopted; as a
minimum, the safety factor should exceed the peak to average load ratio.
5-10
-------
The plug flow hydraulic regime can be approximated by a series of complete mix
denitrification tanks, in which backmixing is prevented. An example is provided by the case
history of the CCCSD Water Reclamation Plant, presented in Section 9.5.2.1.
FIGURE 5-4
EFFECT OF DIURNAL VARIATION IN LOAD ON EFFLUENT NITRATE
LEVEL IN COMPLETE MIX SUSPENDED GROWTH SYSTEM
Q
s
UJ
55
O
(fc ly
Ul
Uj
Uj
Uj
I
kl
ISO
100
50
T
I
I
I
I
I
I
I
2400 0200 0400 0600 0800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, HR
A. ASSUMED VARIATION IN INFLUENT NITROGEN LOAD
2400 0200 0400 0600 0800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, HR
B. CALCULATED EFFLUENT NITRATE LEVEL
5-11
-------
5.2.4 Effluent Quality from Suspended Growth Denitrification Processes
Since denitrification technology is new, there is a concern on the part of some design
engineers that biological nitrification-denitrification systems are unstable and produce
results of high variability. However, large scale tests of biological nitrogen removal have
demonstrated over relatively long periods that a consistently low nitrogen level can be
obtained.
5.2.4.1 Experience at Manassas, Va.
The EPA conducted a 0.2 mgd (760 cu m/day) test of the "three sludge" system for eight
months at Manassas, Va. The system consisted of primary treatment and three separate
suspended growth stages for organic carbon oxidation, nitrification and denitrification
followed by filtration, as shown in Figure 5-5. Alum was added to the first and third
suspended growth stages for phosphorus removal. A dose of polymer was added to the third
reactor effluent. Performance data presented in the form of frequency distribution diagrams
show that the performance of the closely monitored system was very stable.^ A tabular
summary of denitrification effluent quality is shown in Table 5-1 for the last four months of
operation. As may be seen, an effluent very low in total phosphorus and total nitrogen was
obtained from the denitrification system and filtration provided further reductions. Further
details are available in the papers produced from the project. 4,9
FIGURE 5-5
THE THREE SLUDGE SYSTEM AS TESTED AT MANASSAS, VA (REFERENCE 4)
VI
1
Ett
BACKWASH
ALUM METHAN
LI
,, .,— JL
(2) © f (2) © f '
xxi Y '
RETURN SLUDGE] 1 RETURN SLUDGE 1 ( R
I 1 II
i
A
OL ACID
1.
LUM
POLYMER
I
(C) ® ^
ETURN SLUDGE |
1
CI2
CD ©
I
i
i
t !
BACKWASH 1
1
1
^ 1
TYPICAL WASTE SLUDGE LINE
7
WASTE TO SOLIDS HANDLING SYSTEM AND ULTIMATE DISPOSAL
PRIMARY
TREATMENT
©SEDIMENTATION
TANK
HIGH RATE
ACTIVATED SLUDGE .
(5)AERATION TANK
(3)SEDIMENTATION
^-^ TANK
NITRIFYING
ACTIVATED SLUDGE
(?)AERATION TANK
(JT) SEDIMENTATION
^ TANK
DENITRIFYING
ACTIVATED SLUDGE
§ANOXIC REACTORS
AERATED CHANNEL
(5)S£DIMENTATION TANK
POST
TREATMENT
§ MIXED MEDIA FILTERS
CHLORINE CONTACT
(JT) POST AERATION
5-12
-------
TABLE 5-1
DENITRIFICATION PERFORMANCE: FINAL FOUR MONTHS
OF OPERATION AT MANASASS, VIRGINIA (REFERENCE 4)
Parameter
S3
COD
BOD5
Total P
>
Organic N
NH+-N
N02-N
NOg-N
rf
T
O
T
r A
L
N
After final
clarification,
mg/1
2
21
4.0
0.6
1.0
0.0
0.0
0.8
+*
> 1.8
After mixed
media
filtration,
0
16
0.8
0.3
0.8
0.0
0.0
0.7
*i
mg/1
^1.5
5.2.4.2 Experience at the CCCSD's Advanced Treatment Test Facility
In November, 1971, the Central Contra Costa Sanitary District (CCCSD) began the
operation of a full-scale Advanced Treatment Test Facility (ATTF) at its existing
wastewater treatment plant in California. 3 Operation of the facility ultimately extended
over 23 months. A purpose of the test facility was to obtain data on the ATTF System
sequence of processes (Figure 5-6) that had been proposed for the CCCSD Water
Reclamation Plant. ^ Another purpose was to dispel the notion that the nitrification and
denitrification processes were unstable due to the erratic nitrification-denitrification results
previously obtained in a small-scale pilot study.3>15,16
The ATTF process units had capacities ranging from 2.5 mgd (9,464 cu m/day) in the
primary step to 0.5 mgd (1,990 cu m/day) in the denitrification facilities. Primary
clarification follows lime addition and preaeration and is coupled with a separate stage
nitrification step. The use of lime in the primary stage removes much of the organic carbon
load from the nitrification stage, thus allowing stable oxidation of ammonia to nitrate.
Addition of lime also enhances the removal of phosphorus, heavy metals and viruses.
Addition of lime in the initial stage of treatment, in contrast to the use of lime after
conventional secondary treatment, enables the achievement of better stability in succeeding
5-13
-------
FIGURE 5-6
ATTF SYSTEM FOR NITROGEN AND PHOSPHORUS REMOVAL
RAW WASTEWATER
LIME
""" POLYMER OR
FERRIC CHLORIDE
LIME REACTOR
(PREAERATION)
I
PRIMARY
SEDIMENTATION TANK
CHEMICAL
PRIMARY
EFFLUENT A
C02
OX IDAT ION -
NITRIFICATION TANK
I
SECONDARY
SEDIMENTATION TANK
N2
E:
SLUDGE TO
* SOLIDS
PROCESSING
AIR
RETURN
SLUDGE
WASTE SLUDGE
*• TO
RAW WASTEWATER
METHANOL
DENITRIFICAT ION
TANK
I
MIXING
AERATED
STABILIZATION TANK
I
FINAL
SEDIMENTATION TANK
I
RETURN
SLUDGE
WASTE SLUDGE
-*- TO
RAW WASTEWATER
CHLORINE CONTACT
I
ADDITIONAL TREATMENT
FOR INDUSTRY
5-14
-------
treatment processes and also allows the elimination of the need for a biological treatment
step for organic carbon removal ahead of the nitrification stage. Biological denitrification
follows nitrification, converting nitrate to nitrogen gas.
Performance data for a representative three months of operation of the ATTF are shown in
Table 5-2.3 of particular interest is the fact that the 90 percentile performance level did not
vary widely from the median performance level for the various constituents. This provides
statistical confirmation that the nitrification and denitrification performance of the ATTF
system was quite stable. The concentration of organics in the nitrified and denitrified
effluents was low, as measured by BOD and organic carbon. Operation for complete
nitrification also resulted in high organic removals. Similarly, suspended matter in the
nitrified and denitrified effluents were also exceptionally low (Table 5-2). Nutrients are
effectively removed in the ATTF system. Total nitrogen in the denitrified effluent averages
less than 2 mg/1. Total phosphorus averaged 0.5 mg/1.
5.3 Denitrification in Attached Growth Reactors Using Methanol as the Carbon Source
Denitrification in attached growth reactors has been accomplished in a wide variety of
denitrification column configurations using various media to support the growth of
denitrifiers. In part because of this variability among systems, it is difficult to set forth
generally useful design criteria at the present time. Several useful approaches are suggested
for characterizing denitrification in attached growth systems and presently available data
(1975) are analyzed by these procedures.
5.3.1 Kinetic Design of Attached Growth Denitrification Systems
In order to size an attached denitrification reactor, knowledge is required of the reaction
rates taking place in the reactor volume. In estimating reaction rates, the level of biomass
effective in denitrification must also be known. One approach is to estimate the level of
biomass on the media surface and thenuse measured reaction rates per unit of biomass to
obtain the nitrogen removal capability of a column containing the estimated amount of
biomass. 17,18,19 ffas approach is limited in its usefulness in design applications because
there is insufficient data available at the present time to predict in advance the level of
biomass that will develop on the media. Biomass development will be dependent on
hydraulic regime, type of media, loading, means for promoting sloughing and possibly the
temperature of operation.
An indirect procedure for consideration of denitrification rates in design is to adopt the
approach in which denitrification rates are expressed in terms of surface nitrate removal
rates, e.g. Ib nitrate -N removed per sq ft per day.7> 18,20 Qn this basis, high surface
removal rates would reflect extensive biological film development, whereas low surface
removal rates would reflect minimal surface film development. The surface denitrification
rate varies considerably among the various denitrification column configurations and is
affected by the loadings under which the process is operated.
5-15
-------
TABLE 5-2
ATTF PERFORMANCE SUMMARY, APRIL 16 TO JULY 15, 1972
(CENTRAL CONTRA COSTA SANITARY DISTRICT, CA., REF. 3)
Constituent
BOD5
TOCc
SOC
SS
Turbidity (JTU)
Settleable solids (ml/1)
Organic N
NH4- N
NOjj-N
NO§-N
Total P
Ortho P
TDS
Conductivity (105 mho/cm)
Alkalinityd
Cad
Mgd
Raw wastewater, mg/1
mean
203
122
22
214
-
13.5
_
-
_
_
9.86
9.74
583
1230
215
67.9
84.0
median
199
120
23
212
r
13.0
_
-
-
_
9.57
9.90
561
1210
218
66.0
84.0
90%a
235
152
25
295
-
16.5
-
_
_
11.19
ll.:24
750
1366
237
78.5
95.6
Chemical primary,0 mg/1
mean
57
42
27
26
12.8
.084
_
24.0
-
_
.86
.61
_
-
254
155
38.6
median
54
41
27
23
12.0
0
—
23.8
-
_
.59
.46
_
-
249
150
34.0
90%a
79
55
31
45
23
.37
_
28.5
-
-
1.85
1.75
_
-
293
184
74.0
Nitrified effluent, mg/1
mean
3.6
8.9
5.6
4.5
1.4
-
.26
.48
27
»OI5
1.04
1.00
634
1226
105
159
57.3
median
3.5
8.5
5.5
4.0
1.3
-
0
.30
27
.013
.72
.71
636
1223
106
161
61.0
90%a
6.8
11.5
6.7
7.8
2.2
-
.67
.58
32
.027
2.11
1.80
724
1394
127
187
73.0
Denitrified effluent, mg/1
mean
3.2
9.5
5.9
4.5
1.4
-
1.1
.31
.48
.009
.50
.52
537
1160
183
174
30.1
median
3.0
9.0
6.0
4.0
1.3
-
1.1
.30
0
.008
.36
.36
551
1147
187
172
30.0
90%a
4.8
11
6.6
7,7
1.8
-
2.5
.40
.79
.018
.98
.90
616
1318
217
190
48.4
a90% of observations are equal to or less than stated value.
pH 10.2 operation to June 1, pH 11.0 thereafter.
°roC = total organic carbon, SOC = soluble organic carbon
das CaC00
-------
5.3.2 Classification of Column Configurations
The various types of denitrification columns currently available are summarized in Table 5-3
along with calculated peak surface denitrification rates. The first type of categorization is
with respect to the condition of the void space in the column. Until very recently, all
denitrification work has been conducted on submerged columns wherein the voids were
filled with the fluid being denitrified. Very recently, a new type of column has been
developed in which the voids are filled with nitrogen gas, a product of denitrification.6>21
The submerged columns can be further subdivided into packed bed and fluidized bed
operations.
The varieties of media being employed for commercial application are also shown in Table
5-3; the listing is not meant to exclude commercial products which may be equivalent to
those listed. For instance, other vendors of plastic media are listed in Table 4-15.
Details of design construction and operation of each column type are presented in the
following sections. Also included is a comparison of column systems.
5.3.2.1 Nitrogen Gas Filled Denitrification Columns — Packed Bed
The nitrogen gas filled column was recently developed for use at the Lower Molonglo Water
Quality Control' Centre (LMWQCC), currently under construction.6»21 Details of the
column design used for this Canberra, Australia installation are shown in Figure 5-7. The
column media consists of corrugated plastic sheet modules of the same type used in plastic
media trickling filters (Table 4-15). As opposed to previously developed attached growth
processes, the present column system is not submerged with liquid; rather, the column's
void spaces are filled with nitrogen gas, a reaction product of the denitrification process.
In the denitrification column the influent wastewater is spread out over the top of the
media and then the liquid flows in a thin film over the media on which the organisms grow.
These organisms maintain a balance so that an active biological film develops. The balance is
maintained by sloughing of biomass from the media, either by death or by hydraulic erosion
or both. Sufficient voids are present in the media to prevent clogging and ponding. The
denitrification column must be followed by a clarification step to remove sloughed solids.
Pilot studies for the LMWQCC facility indicated that effluent solids should be sufficiently
low so that the effluent can go directly to a tertiary multi-media filter.
Oxygen must be excluded from the denitrification column since its presence would prevent
nitrate or nitrite in the applied liquor from serving as the electron acceptor in the biological
oxidation of the applied carbon (methanol). Therefore, the denitrification column is sealed
to prevent intrusion of air. The units are vented to allow outflow of nitrogen gas while
preventing inflow of atmospheric air. Soon after start-up, the nitrogen gas displaces the air
or other gases initially present, leaving a nitrogen gas atmosphere in the voids. This ensures
the anoxic environment that is required for denitrification.
5-17
-------
TABLE 5-3
TYPES OF DENITRIFICATION COLUMNS AND MEASURED DENITRIFICATION RATES
Void
Space
Nitrogen
gas
Liquid
1
*
1
Type
Packed bed
Packed bed
Fluldlzed
bed
Nature of
Surface
High porosity
corrugated sheet
modules
High porosity
corrugated sheet
modules or
dumped media
Low porosity
fine media
High porosity
fine media,
sand
Activated carbon
Ref.
No.
6,
21
6,
21
22
23
24
7.
20
17,
25
18
26
24
27,
28
29
30
Media
trade name.
Munters
Plasdek
Koch
Flexlrlngs
Envlrotech
Surfpac
Koch
Flexlrlngs
Intalox
saddles
Raschlg
rings
Gravel
d50=3. 4 to
14. 5 mm
Sand
d50=0. 9mmf
Sand
d50=3 to
4 mm
Sand
d10»2.9mm
U.C.= 1.13
d50=0.8Smmf
d10=0.65mm
Specific
surface
sf/cu ft
(mW)
68
(223)
42
(138)
65
(213)
27
(89)
105
(344)
142 to 274
466 to 899)
79
(259)
245 to 85
804 to 279)
450'
(1,476)
250
(825)
270'
(886) .
130f
(426)
3908
(1,279)
Voids,
percent
~95
~95
96
94
92
70 to 78
80
28 to 37
_
40
-
_
Surface or volume denltrlflcatlon rate at stated temperature, C
Ib N rem/sf/day x 10* a (ft N rein/1000 cf/day b)
5
.32
(4.5
to
8.8)
10
.37
(5.3
to
10.1)
11
12
13
.43
(2.8)
0.6
(16)
14
15
.50
(3.3)
.26
(3,7
to
7.1)
16
_
(257)
17
6.2
(42)
.53
(3.4
18
55
(720)
19
18.6
(126)
20
1.3
(8.5)
.95
(13.5
to
26)
_
(308)
21
1.1,
(7.2)
22
98
(1275)
-
(292)
(388)
23
12.1
(82)
.59
(3.8)
24
-
(424)
25
1.1
(15.6
to
30.1)
2.4d
(19)
1.46=
(12)
26
27
2.02
(21)
2.3
-
2.9
(73)
4.6
(124)
Other
.93°
(2.5)
3.9
(176)
1 kg /m2/day = 0.205 Ib/ei/day
b 1 fcgytnVday = 62.4 Ib/1000 CF/day
c 10 - 23C
1 hour detention
6 2 hour detention
Estimated from data In publication
8 Prior to bed expansion
Note: 10 = effective size, 50 = median
and U.C, = uniformity coefficient
size,
O
= particle size at which 60 percent of the material Is smaller
-------
982m 12.25 m
TYP CELL TYR CELL
SEAL WftU.
3OOmm DUMPED MEDIA
WEIR
METHANOL FLOW RATE
CONTROLLER
ADJUSTABLE HANGER RODS
DENITRIFIED EFFLUENT
CHANNELS
DENITRIFIED EFFLUENT
COLLECTION CHAMBER
FIGURE 5-7
DESIGN DETAILS OF NITROGEN GAS FILLED
DENITRIFICATION COLUMN (REFERENCE 21)
SPLASH PLATES. APPROX
2OOmm ABOVE MEDIA
-------
Media specific surface and configuration must be carefully selected to ensure that clogging
does not occur. Clogging can occur under high loadings if the restrictions in the media are
too small. Pilot studies may be warranted for media selection when design conditions depart
from those previously tested. Once the proper media is selected, the process has the
advantage that backwashing is not required, considerably simplifying design, construction,
and operation. Further, the column construction is simplified since it need not be a pressure
vessel as the column is at approximately atmospheric pressure.
The denitrification column was first tested at the Central Contra Costa Sanitary District's
Advanced Treatment Test Facility (ATTF)A21 The test program had three primary
objectives: the first was to develop confidence in the ability of the attached growth reactor
to function consistently and predictably; second, to determine the optimum specific area of
the plastic media; third, to develop criteria upon which design of the prototype columns for
the LMWQCC could be based.
The test denitrification column consisted of a sealed vertical 24-inch (610 mm) diameter
reinforced concrete pipe 12 ft (3.66 m) in height and filled with 10 ft (3.05 m) of media.
Plastic media consisted of PVC corrugated sheet modules, supplied by Munters Corporation.
The top of the column was sealed by a gasketed cover and the nitrified liquor was applied to
the top of the tower by a round pattern nozzle. After passing through the column, the
denitrified effluent was collected in a sump from where it was discharged. Provision was
made to allow pumping of column effluent back to the influent of the column to test the
merits of recirculation. A brief description of the test program and a summary of the results
is presented on Table 5-4. The initial media tested had a specific surface of 68 sq ft per cu ft
(223 m^/m^). This media clogged at the high application rates applied in May and June
1972 and was subsequently replaced with media having a specific surface of 42 sq ft per cu
ft (138 m^/m^). Recirculation of effluent was not found to be required and was dropped
from the test program because of the high energy costs that would be incurred if
recirculation was used in the full-scale plant.
Experience with the pilot column indicated that an application rate of 5 gpm appeared to be
a reasonable flow that could be applied continuously without an objectionable buildup of
growth on the media. That value is equivalent to a rate of 245 gallons per cubic foot per day
(33,300 l/m^/day). Because the Contra Costa wastewater has a lower nitrogen concentration
than the wastewater at the LMWQCC, the hydraulic loading rate was adjusted to ensure an
equivalent nitrogen load was applied. The value was then further reduced to permit lower
loadings at low wastewater temperature. Taking these factors into account, an ADWF
loading rate of 144 gallons per cubic foot per day (12,562 l/m^/day) was selected for design
purposes. In terms of nitrate removal rate on the basis of media surface, this corresponds to
9.9 x 10-4 Ib NC-3-N rem/sq ft/day (4.83 x 10~3 kg/m2/day).
While media with a specific surface of 42 sq ft per cu ft (12.8 m^/m^) is believed to be not
susceptible to clogging in municipal applications based on the limited experience to date, it
would appear prudent to provide chlorine addition capability to aid sloughing should
clogging occur.
5-20
-------
TABLE 5-4
SUMMARY OF OPERATION - NITROGEN GAS FILLED DENITRIFICATION COLUMN
IS)
to
Dclt6»
1972
April 24
April 27
May 2
May8
May 11
May 16
May 18
May 23
May 26
May 31
June 1
June 5
June 9
June 12
June 15
June 16
July 27
July 28
August 1
August 4
August 9
August 15
August 21
August 25
August 28
August 30
September 1
September 7
September 12
September 19
September 27
October 3
October 25
Effluent
flow rate,
gpma
4.7
5.3
2.3
2.1
2.0
2.5
2. 5/5. 0
5.0
10.0
10.0
15.0
15.0
10.0
10.0
7.0
_
-
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
Recirculation
rate
gpma
9.4
9.4
9.4
9.4
8.0
0
0
0
0
0
0
0
0
0
0
_
-
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NOs-N, mg/1
In
27
22
28
30
24.5
28
27
27
-
22
25.5
-
27
27
27
_
-
15.3
13.5
11.8
11.8
9.8
13.2
12.5
23
30
18
27
17
24.5
18
19
20
Out
4.5
1.6
<0.2
<0.4
Nil
1.6
<0. 1
0.6
-
<1
3.4
-
1
5
6.8
_
-
5.4
2.5
2.7
1.0
Nil
4.0
5.0
17
22
3.2
0.5
Nil
Nil
0.8
. Nil
1.4
TOC, mg/1
In
26
29
37
36
40
10
9
8
-
9
_
-
7
8
7.5
_
-
46
10
73
11
-
8
8
8
8
8
-
8
8.5
8
8
6.5
Out
15.5
19
25
11
16
13
57
14
-
16
_
-
15
17
12
_
-
38
10
45
14
-
18
13
12.5
13
13
-
18.5
12
12
15.5
11
Comment
Media installed had 68 sq ft/cu ft. (20. 7 m2/m3)
Eliminate recirculation and increase flow rate.
Column being overloaded.
Severe foaming. Feed rate decreased.
Mechanical breakdowns and operation difficulties.
Test results unreliable.
Media removed.
New Media used was 42 sq. ft/cu ft.
(12. 8 m2/m3)
Denitrification reestablished.
al gpm - 0.065 1/s
-------
The design of the denitrification column for the LMWQCC is portrayed in Figure 5-7 and
design data are given in Section 9.5.2.2. Operation will be by gravity flow, as the LMWQCC
is located on a very steep site. Fixed distribution troughs with splash plates were used to
handle the wide range in flows expected and to minimize the restrictions in the distribution
system which might be clogged by the growth of denitrifiers. Further, a top layer of
dumped media is placed over the corrugated media to ensure good distribution and growth
at the top of the column.
5.3.2.2. Submerged High Porosity Media Columns — Packed Bed
Submerged denitrification columns packed with high porosity media have been piloted at
several locations?. 17,20,22,23 ^d tested full-scale at El Lago, Texas. 24
A typical schematic illustrating essential process elements is shown on Figure 5-8. The media
is normally contained in pressure vessels. To obtain sufficient contact time, a series
configuration of 2 or 3 vessels is employed. 7,20,24 Either an up flow or downflow column
operation may be used. While a variety of media types have been used (Table 5-3), a
FIGURE 5-8
TYPICAL PROCESS SCHEMATIC FOR SUBMERGED
HIGH POROSITY MEDIA COLUMNS
TO STORAGE
M WAItK
XGE
.•NITRIFICATION
COLUMN
IPED
)IA
^
s
'
»
h
H
tFFLU
CLARI
FILTR
DENITRIFIC
COLUMN
METHANOL-
INFLUENT
PUMP
CLARIFIED
BACKWASH
WATER
BACKWASH
PUMP
5-22
-------
common characteristic is that a high void volume is maintained in the unit. As a
consequence, biomass is allowed to continuously slough from the media, minimizing the
requirements for backwashing. A corollary is that the media does not build up the layers of
biomass that would develop if the void fraction were smaller such as with sand or pebble
media.23 The lower surface denitrification rates for this type of media compared to sand or
rock (Table 5-3) reflect this difference in attached biomass development. Most often
dumped media have been used, though there is one instance when corrugated sheet media
has been tried. 23
Backwashing, though infrequently used, is still required. At El Lago, Texas, where the media
was Koch Flexirings, the water backwash rate was 10 gpm/sf (13.5 l/s/m^) coupled with an
air backwash rate of 10 cfm/sf (3.6 m3/m2/min). Backwashing was routinely done every
four weeks.24 Backwashing in this type of column is not required due to excessive head
losses in the column; rather, it is required to prevent the accumulated solids in the column
from continuously sloughing into the effluent and causing high effluent suspended solids.
Others have used backwash rates up to 44 gpm/sf (29.7 l/s/m^) but did not use an air
backwash procedure.7.20 The El Lago air-water backwash procedure is the recommended
approach for design purposes. As opposed to the situation with respect to other column
designs, a fairly broad data base exists for this column type. Surface removal rates observed
at various locations are summarized in Figure 5-9. As may be seen, most data points fit the
data correlation of Sutton, et al. 20fOr Intalox saddles.
Figure 5-9 may be used to size the denitrification column. First, peak diurnal nitrate loading
and minimum wastewater temperature must be known. From Figure 5-9, surface removal
rate can be determined. Then from the loading, the media surface area can be calculated.
Finally, a specific media is selected and column volume requirements are calculated.
5.3.2.3 Submerged Low Porosity Fine Media Columns — Packed Bed Configuration
The submerged low porosity column using fine media (Table 5-3) is the column system
seeing widest commercial application at the present time. One manufacturer's concept
(Dravo) of how to incorporate this type of column into a treatment plant is shown in Figure
5-10.27 jn this flowsheet combined carbon oxidation-nitrification is accomplished in an
activated sludge step. Of course, other nitrification processes could be employed for
producing a nitrified effluent. Clarified nitrified effluent then flows to the denitrifica-
tion column. The concept employed in this flowsheet is that the column combines two
functions in one. First, it serves the purpose of denitrifying the wastewater; second, the
column serves the purpose of effluent filtration that normally would be required in many
plants anyway.27 A discussion of the cost-effectiveness of combining the denitrification and
filtration functions is presented at the end of this section.
The units manufactured by the Dravo Corp. typically consist of 6 ft (1.83 m) of uniformly
graded sand 2 to 4 mm in size. Filtration rates normally recommended by the Dravo Corp.
when removing 20 mg/1 NO3 —N from municipal wastewaters are 2.5 and 1.0 gpm/sf for
5-23
-------
FIGURE 5-9
I
i
ki
5
CD
Uj"
K,
$
-j
o
Uj
Uj
<->
Cfc
^
CO
SURFACE DENITRIFICATION RATE FOR SUBMERGED
HIGH POROSITY MEDIA COLUMNS
2.5
2.0
1.5
1.0
0.5
KEY
I
I
SYMBOL
V
~ •
A
•
LOCATION
Davis, Ca.
Hamilton,
Ontario
Firebaugh,
Ca.
El Logo,
Texas
REF.
17,
25
7,
20
22
24
MEDIA
Raschig
rings
Intalox
saddles
Koch
Flexirings
Koch
Flexirings
SPECIFIC
SURFACE
SF/CU FT*
79
142 to
274
65
105
VOIDS,
PERCENT
80
70 to 78
96
92
'* 1 SF/Cu Ft = 3.28 m2/™3
<35 % CONFIDENCE LEVEL-
FOR ONTARIO DATA
WEIGHTED LEAST SQUARES FIT-
FOR ONTARIO DATA (Refs. 7 and 20)
K= 90.46
I kg/m2/day =0.205 Ib/sf/day
I I I
O 5 10 15 20 25 3O
TEMPERATURE, C
minimum wastewater temperatures of 21 C and 10 C respectively. 31
The procedure for backwashing the Dravo filter begins with one or two minutes of air
agitation followed by 10 to 15 minutes of air and water scouring and finally five minutes of
water rinse. Air and water backwash rates recommended by Dravo are 6 cfm/sf (1.83
m^/m^/min) and 8 gpm/sf(5.41 l/s/m^), respectively.-* 1 In addition, it has been found that
nitrogen gas accumulates in the filter during a filter run. This imposes a loss of head on the
5-24
-------
FIGURE 5-10
NITRIFICATION-DENITRIFICATION FLOW SHEET UTILIZING
LOW POROSITY FINE MEDIA IN COLUMNS (REFERENCE 27)
ORGANIC
CARBON
FEED PUMPS
COMMINUTOR
INFLUENT
0-n
BACKWASH RETURN
-cn
BLOWER
00 OOOpO
oo
O O AERATION
O O O 0TANK
°n ° o°
O °
I \
CLARIFIER
ACTIVATED
SLUDGE RETURN
WASTE SLUDGE
Q-o n O O .
O DIGESTER OR>
w SLUDGE o
O o HOLDING H
TANK f
-^. SUPERNATANT
TO INFLUENT
WASTE
SLUDGE
BACKWASH
WELL
-QH
n
1 BACKWASH
RETURN PUMP
DENITRI-
FICA.T.IQN
COLUMN
COLUMN
BACKWASH
CHLORINE
° CHLORINE
i° CONTACT
"' TANK
EFFLUENT
REAERATION
AIR
BACKWASH BACKWASH
BLOWER PUMPS
filter and requires periodic removal of the trapped gas bubbles in the media. A "bumping"
procedure was evolved whereby a filter was taken out of service for what amounts to a short
backwash cycle. In the Dravo design, "bumping" backwash rates of 8 to 16 gpm/sf (5.4 to
10.8 1/s/m^) for one or two minutes is required every four to twelve hours.27
The period between regular backwashes in the Dravo filter is dependent on the rate of
headloss buildup. At El Lago, Texas it was found that daily backwashing was required
whereas in a pilot study in Tampa, Florida the time between regular backwashes ranged
from 4 to 40 days.24,32 Durmg the Tampa study, the air backwashing was found to cause a
temporary partial inhibition of denitrification that was not present when only a water
backwash was used. For instance, with an influent nitrate-nitrogen level of 15 mg/1, the
effluent nitrate nitrogen level was 10 mg/1 one-half hour after the air-water backwashing and
reached 0 mg/1 seven and one-half hours after back washing. 3 2 in multiple filter installations,
the effluents from the recently backwashed filter would be blended with other normally
operating filters, so the impact of this nitrate leakage would be expected to be moderated.
Generally speaking, even the smallest of plants will require multiple filters so that an
effluent can be continously produced, otherwise a filter influent storage basin will be
required.
5-25
-------
Neptune-Microfloc, Inc. has made available suggested design guidelines for their media
designs.33 Four media designs were tested on a nitrified effluent containing 20-30 mg/1 of
NOj-N from an extended aeration plant. Best overall performance was obtained from the
two media designs shown in Table 5-5. Basic conclusions of the study were as follows:
"Utilizing a 36-inch (0.92m) mixed-media filter (F-III), essentially complete
denitrification of a highly nitfified wastewater can be achieved at filtration rates
of 1.5 gpm/sq ft (l.Ol/s/m^) for temperatures of 10 C, and at 3 gpm/sf
(2.01/s/m2) at temperatures of 20 C. The methanol to nitrate nitrogen ratio was
found to be between 2.0 and 2.5. At applied nitrate nitrogen concentrations of 10
mg/1, filter run times between 16 and 24 hours to 8 feet of headless were realized
at a filtration rate of 3 gpm/sq ft (2.01/s/m2). At higher applied nitrogen levels,
filter runs were reduced in direct relation to nitrogen concentration."^
In another study of a Neptune-Microfloc filter, a fully nitrified effluent from a trickling
filter was denitrified. No attempt was made to determine limiting filter loadings, however.
TABLE 5-5
NEPTUNE-MICROFLOC MEDIA DESIGNS
FOR DENITRIFICATION (REFERENCE 33)
Filter Material
Layer depths, inches (cm)
F-n
F-III
Garnet Sand
d!0 = 0.27 mma
Silica Sand
d!0 ~ 0.5 mma
Anthracite Coal
d!0 = 1.05 mma
Anthracite Coal
d!0 = 1.75 mma
3
(7.6)
9
(22. 9)
18
(45.1)
3
(7.6)
9
(22. 9)
8
(20. 3)
16
(40.6)
10 = effective size
5-26
-------
The filter had three layers of media: anthracite, silica sand and garnet sand with sizes
ranging from 1.2 mm at the top to 0.2 mm on the bottom. At a surface application rate of
2.3.gpm/sf (1.61/s/m2) and influent nitrate levels averaging 8 to 9 mg/1, better than 95
percent removals were obtained. Operating temperature ranged from 16 to 18 C during this
study. 34,35
Media size is important in establishing denitrification column requirements. The relationship
between specific surface and column size was established in a pilot study at Lebanon, Ohio
using the following media sizes for three columns: 3.4 mm, 5.9 mm, and 14.5 mmJ^
Biological film development per unit surface area was shown to be approximately the same
for each size media. Therefore, the smaller the media, the higher the media surface per unit
volume and the smaller the column as shown in Figure 5-1 1 .
FIGURE 5-1 1
COLUMN DEPTH VS SPECIFIC SURFACE AREA (REFERENCE 18)
Q
iu o
- §
Z3 LU
O
-------
Care should be exercised in the design of the column underdrain system. Neptune-Microfloc
recommends that an extremely open underdrain system be employed (pipe lateral, Leopold
tile, etc.) to avoid the very real possibility that an overfeed of methanol will cause denitrifier
growth to clog the underdrain as was experienced in one test with a nozzle-type
underdrain.33
Since it has been proposed to use this type of column for both filtration of suspended solids
and nitrate removal, it would be well to examine the performance of the columnar system
for suspended solids removal. Various observations of suspended solids removal are shown in
Table 5-6. With the exception of the El Lago, Texas data, the performance of these columns
as tertiary filters falls within the range normally expected for tertiary filtration (for
comparative data see Section 9.3.2.3 of the Process Design Manual for Suspended Solids
Removal, an EPA Technology Transfer publication).36 Suspended solids removals will be
affected by filter design and the fact that the filter is operating as a biological treatment
system as opposed to a purely physical separation process.
In considering the use of this process as both an effluent filter and a denitrification system,
an important design factor should be borne in mind that has considerable implications on
TABLE 5-6
COMPARISON OF SUSPENDED SOLIDS REMOVAL EFFICIENCY
FOR SUBMERGED FINE MEDIA DENITRIFICATION COLUMNS
Location
El Lago, Texas
North Huntington
Township, Pa
Tampa, Fla.
Lebanon, Ohio
Corvallis, Or.
Midland, Ml.
Media type
Dravo
dgQ = 3 to 4mma
Dravo
d10 = 2.9mm
Dravo
d!0 = 2.9 mm
dgo = 3. 4mma
d50 = 5. 9mma
d50 = 14. 5mma
Neptune
Media F-H b
Media F-III b
Neptune
Reference
24
27,28,31
27,31
18
18
18
33
34,35
Surface
loading
gal/min/sf
(l/s/m2)
6.27
(4.23)
0.72
(0.49)
2.5
(1.70)
7.0
(4.75)
7.0
(4.75)
7.0
(4. 75)
3.0
(2. 04)
3.0
(2. 04)
*j.5
(1.70)
Depth,
ft
(m)
13
(4)
6.0
(1.8)
a
10
(3.1)
20
(6.1)
20
(6.1)
2.5
(0.76)
3.0
(0.91)
5.0
(1.50)
Influent
SS,
mg/1
37
16
20
13
13
13
25-65
25-65
13-30
Effluent
• SS,
mg/1
17
7
5
4
2
1
8
4
2-10
SS removal
efficiency,
percent
54
56
75
69
85
92
68-88
84-93
67-93
uniformly graded
J Table 5-5
5-28
-------
cost. It has been claimed that combining the functions of filtration and denitrification
reduces tankage and equipment requirements and therefore yields cost savings in plants
requiring filtration.27 However, it should be recognized that the column loading criteria are
different for the functions of filtration and nitrogen removal. For effluent filtration, fairly
high hydraulic loadings can be applied (4 to 6 gpm/sf or 2.7 to 4.1 l/s/m^). However, for
filters 3 to 6 ft (0.9 to 1.8 m) deep acting as denitrification columns, available data indicates
that hydraulic loading should be between 0.5 to 1.5 gpm/sf (0.34 to 1.02 l/s/m-2) at a
wastewater temperature of 10 C. Thus, to accomplish denitrification at IOC, it would be
necessary to have column surface areas five times as large as required for filtration alone.
Thus, an economic analysis must be done in each case to determine the most economic
process configuration.
5.3.2.4 Submerged High Porosity Fine Media Columns — Fluidized Bed
The introduction of fluidized bed technology into the field of columnar denitrification is a
comparatively recent development.29,30,37,38 Figure 5-12 depicts a typical fluidized bed
reactor with its ancillary facilities. In the fluidized bed unit wastewater passes upwards
vertically through a bed of small media such as activated carbon or sand at a sufficient
velocity to cause motion or fluidization of the media. The small media provides a large
surface for growth of denitrifiers.
High surface application rates were recently employed in a pilot study of the process at
Nassau County, New York (15 gpm/sf or 10.2 l/s/m2).38 Tne column had a fluidized bed
depth of 12 ft (3.7 m). The bed settled to about 6 ft (1.8 m) when the influent was shut off
FIGURE 5-12
FLUIDIZED BED DENITRIFICATION SYSTEM
FLUIDIZED SAND
BED SEPARATION
REACTOR TANK
BIOMASS
SEDIMENTATION
TANK
FLUIDIZED
MEDIA
PEA GRAVEL^
METHANOL
y_u I
SAND
CLEANI-NG
AND RETURN
PUMP
J
INFLUENT
PUMP
WASTE SLUDGE
TO DISPOSAL
/ V
u
DENITRIFIED
EFFLUENT
5-29
-------
so bed expansion during operation was 100 percent. Initially the clean bed in the column
contained 1.5 in. (38 mm) of pea gravel and 3 ft (0.91 m) of silica sand with an effective
size of 0.6 mm and a uniformity coefficient of 1.5. During operation, the media became
completely covered with denitrifier growth and the individual particles grew in size.
During the initial lab-scale test for this process, the media grew from 0.65 mm particles, to
particles 3 to 4 mm in size.™ The attached growth accounts for the greater depth of media
in the non-fluidized bed after the column had been in operation.
In a packed bed, this growth of particle size would result in high headloss, channeling, and a
loss in efficiency. In an expanded bed, however, there are sufficient voids between the
denitrifier-sand particles to provide good liquid contact at modest headlosses. This greater
biological film development allows higher surface reaction rates (expressed per unit of media
surface) than for any other type of column configuration as shown in Table 5-3. Since the
surface contained in a unit volume is high, higher volumetric loadings are also possible as
compared to any other column configuration (Table 5-3). Empty bed detention time during
the recent pilot test at Nassau County was only 6.5 min. Maximum nitrate removal rates as a
function of temperature are shown in Figure 5-13 and are based on the Nassau County
data.39 if diurnal variations in nitrogen load are to be accommodated by the column
without nitrate bleedthrough, then column volume requirements will be greater than that
determined in Figure 5-13. A provisional recommendation would be to increase the reactor
requirements by the ratio of the peak to average nitrogen loads.
The process has been shown to be responsive to both diurnal load variations and to cold
temperature operation.29 Methanol feed was not automatically controlled, so periods of
nitrate bleedthrough occurred.29 When methanol feed was under control, 99 percent
removal of influent nitrate and nitrite was demonstrated. Total nitrogen reductions were not
given. 29,38
While backwashing facilities are not required in this type of column, facilities must be
provided for managing the column media inventory. During operation, the denitrification
column increases in depth due to biological growth causing a continuous small loss of media
from the system. Further, diurnal flow variations cause height variations which may
contribute to media loss. This loss can be minimized by provision of flow equalization
facilities (see Chapter 3, Flow Equalization, Process Design Manual for Upgrading Existing
Wastewater Treatment Plants, an EPA Technology Transfer Publication).40
The manufacturer of the system suggests that for most plants subject to diurnal flow
variations, media losses and effluent solids levels can be controlled by placing two tanks in
series with the column as shown in Figure 5-12. The first tank would be a sand separation
tank followed by a biomass sedimentation tank for biological solids removal. The sand
separation tank might be very small, as a tank with an overflow rate of 13,600 gpd/sf (554
m-Vm^/day) served satisfactorily during the pilot study. It has been suggested by Ecolotrol,
the manufacturer of the fluidized bed system, that the swirl concentrator which was
developed for grit removal from combined stormwater and wastewaters could serve as the
5-30
-------
FIGURE 5-13
VOLUME DENITRIFICATION RATE FOR SUBMERGED
HIGH POROSITY FINE MEDIA COLUMNS (REFERENCE 39)
24OO
Q
I-
U.
§
O
\
^
co
2000
I6OO
Ui
t-
2 I2OO
2
O
O
u!
(t
Ul
Q
Ul
5
BOO
4OO
I
I
I
I
10
15 2O
TEMPERATURE, C
25
30
sand separation device.^^'^2 The media settling in the sand separation tank would be
pumped back to the column with the:pumping action shearing the denitrifiers from the
media. This sheared biomass would pass through the column and sand separating tank and
then settle in the biomass sedimentation tank.29 if very low levels of suspended solids are
required, the system would have to be followed with tertiary filtration.
5-31
-------
The manufacturer's cost estimate indicates that the fluidized bed system is competitive with
suspended growth or packed bed columnar systems, but it was also stated that these must be
confirmed by larger scale tests on the fluidized bed system.29
5.3.2.5 Comparison of Attached Growth Denitrification Systems
The three column systems previously described in Sections 5.3.2.1, 5.3.2.2, and 5.3.2.3 are
seeing commercial applications at this time and the fluidized bed system described in
Section 5.3.2.4 will likely see application in the near future. Therefore, the design engineer
has four alternative column systems to consider.
Where low treatment plant effluent solids are required, tertiary filtration will have to follow
all column systems except the low porosity fine media system described in Section 5.3.2.3.
In small plants, the elimination of a unit process may favor combining the denitrification
and filtration functions. In larger plants the cost trade-offs between alternatives need to be
considered.
Where space restrictions exist at a plant site, there is an incentive to pick those systems
requiring the least land area possible. While from Table 5-3 it might appear that the fluidized
bed had the distinct advantage because of highest volumetric loading rates, the ancillary
sand separation and biomass sedimentation tanks diminish its advantage over the submerged
fine media column and the nitrogen gas filled column.
The submerged high porosity column configuration appears to offer the least attractive
alternative. Both surface and volumetric removal rates are low, requiring comparatively large
reactors (Table 5-3). Further, the unit must incorporate the design features of a filter,
without having the advantage of low effluent suspended solids of the submerged low
porosity columns. The system has an advantage for small treatment plants in that
backwashing is only infrequently required and can be scheduled to coincide with plant staff
availability.
The advantages of the nitrogen gas filled column are: (1) similar space requirements to low
porosity submerged columns (2) column walls need not be designed to handle hydrostatic
loads and (3) with proper media selection, sloughing occurs naturally and backwashing is
not required.
5.4 Methanol Handling, Storage, Feed Control, and Excess Methanol Removal
Methanol is a chemical not normally dealt with in wastewater treatment plant operation and
care must be exerted in the design and operation of methanol handling, storage and feeding
facilities to ensure its safe and proper use.
5-32
-------
5.4.1 Properties of Methanol
Methanol, CH3OH, has a variety of names such as methyl alcohol, carbinol and wood
alcohol and is normally supplied pure (99.90 percent). It is a colorless liquid, noncorrosive
(except to aluminum and lead) at normal atmospheric temperature. Some important
properties of methanol are shown in Table 5-7. Additional data is available in references 43
and 44 and manufacturer's information.
TABLE 5-7
PROPERTIES OF METHANOL
Property
Value
Density
Vapor Density (air = 1. 00)
Vapor Pressure 0 C
10 C
20 C
30 C
40 C
50 C
Solubility
Viscosity @ 20 C
Combustible Limits, percent
by volume in air at STP
Flash Point Tag Open Cup
Tag Closed Cup
0. 7913 g/ml @ 20C (6. 59 Ib/gal)
1.11
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
7.3 to 36
16 C (61 F)
12 C (54 F)
Taken internally, methanol is highly toxic. It is harmful if the vapors are inhaled or if skin
contact by liquid or vapors is prolonged or repeated. Fire and explosion are primary dangers
of methanol. Persons involved in handling methanol should be aware of these hazards.
Federal, state and local regulations for safety should be posted along with information from
references 43, 45 and manufacturer's data.
5.4.2 Standards for Shipping, Unloading, Storage and Handling
The shipping, unloading, storage and handling of any flammable chemical including
methanol is governed by a multitude of stringent regulations which include: Federal, such as
5-33
-------
the Department of Transportation (DOT) and the Occupational Safety and Health Act
(OSHA); State, which has various safety orders and codes; municipal ordinances;
independent associations such as the National Fire Protection Association (NFPA) and the
Manufacturing Chemists Association (MCA); and insurance requirements. It is necessary that
all of these regulations be reviewed and studied before the design of any methanol facilities,
and all such regulations must be followed.
5.4.3 Methanol Delivery and Unloading
Methanol may be received in 55 gallon (208 1) metal drums, tank wagon, tank truck or tank
cars. Other methods of shipping, not discussed herein, are by barge, metal drums (less than
55 gal) and glass and metal cans. Tank wagons are normally 1,000 to 4,000 gal (3,785 to
15,142 1) in size, tank trucks range from 4,000 to 9,000 gal (15,142 to 34,069 1) while tank
cars are shipped in 6,000, 8,000 and 10,000 gal (22,713, 30,283, 37,854 1) capacities. Tank
cars and tank trucks are the most economical shipping mode for most plants. However, for
pilot work and small plants, 55 gallon metal drums may be feasible. Since methanol is
classified as a flammable liquid by the DOT, all shipping containers must be approved and
labeled in accordance with applicable DOT regulations.
The recommended method of unloading methanol from any container is by pumping. Some
barges and tank wagons have their own pumps for unloading. Tank cars and trucks can be
unloaded from the top or bottom and be pumped or conveyed by gravity or syphoning. The
preferred method of unloading is pumping from the top via an eductor tube. Syphoning and
gravity unloading are only permitted when the top of the storage vessel is below the bottom
of the shipping container. Due to the increased spillage probability using bottom unloading,
it is only permitted on cars or trucks approved for bottom unloading which include valving
approved by the Association of American Railroads (AAR) and in agreement with the DOT
requirements. This valving helps contain the 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 never be used for methanol unloading.
However, top unloading using the water displacement method or inert gas padding, i.e.
carbon dioxide, nitrogen, etc., may be used if the exact unloading procedures are as provided
by the chemical manufacturer. Unloading procedures can be found in references 43 and 45
and ,the supplier's data.
General requirements for the design of unloading facilities for methanol are applicable to
both tank car and truck. The unloading area should be arranged to avoid traffic areas. Also,
all facilities should be outside due to the fire hazard and all equipment in the vapor area
must be explosion-proof, Class I, Group D, Division 1 or 2 per the National Electrical Code.
Tools should be "non-sparking." Unloading should occur during daylight hours since the
safety and lighting requirements for night operation are very extensive. Ample fire
extinguishers, safety blankets, deluge showers, eye washes, no smoking signs and unloading
signs are also required.
5-34
-------
If top unloading will be practiced, approach platforms are required for access to the top of
the tank.46 The approach platforms used at the CCCSD Water Reclamation Plant are steel,
swinging type with pneumatic operated drawbridges. These drawbridges provide for three
feet (0.9 m) of horizontal adjustment, 30 in. (76 cm) of vertical adjustment and a 45 degree
pivot to either side for ease in unloading both tank cars and trucks. The drawbridges fold to
the platforms when not in use providing the required railroad clearances.
In all unloading setups, all equipment must be grounded. This includes the shipping vessel,
interconnecting piping, pumps, approach platforms, etc. Also, bonding jumpers must be
used to provide a good continuous system. Periodic checks of the grounding system must be
made.
Static electricity buildup must be minimized since it is not dangerous until at the spark
discharge level. A spark discharge can easily start a fire or cause an explosion. Refer to
reference 47 for static electricity accumulation prevention.
Facilities for truck unloading must provide for ease in truck maneuvering, both entering and
leaving the area. Consideration must be given to the number of individual unloading spots
regarding frequency of use, space and simultaneous unloading. For ease in truck traffic, it is
best to have parallel unloading areas for straight-through driving.
Rail unloading creates additional considerations. Unloading areas must have derails or a
closed switch a minimum of one car length from the car. A primary concern is who spots
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 spot cars at night, over weekends or holidays is high and because the
railroad cannot guarantee time of shipment, safety provisions and lighting must be provided
for night operations. Two sidings, one for empty cars and one for full cars should be
provided. The cars can be spotted by the plant personnel with rail car movers which operate
both on rail or streets or in the case of short distances, car spotters (winches) can be used.
By paving the railroad yard area, both truck and rail unloading can be practiced. This is
advantageous because major strikes affecting either kind of transport cannot cripple the
plant. At the CCCSD Water Reclamation Plant (described in Section 9.5.2.1), two sidings
are provided with space for ten cars on each track for storage. Unloading sidings with two
platforms and two bottom stations will be used for simultaneous spotting and unloading. A
rail car mover with a capacity of two cars is also provided.
Unloading equipment is normally steel but many other materials, except aluminum, are
acceptable providing they can withstand the pressure and are completely grounded. Pumps must
be "non-sparking" such as bronze fitted steel pumps with bronze impellers. Many materials
are compatible with methanol so seals and gaskets can be common materials. Pumps may be
either centrifugal or positive displacement gear type. However, positive displacement pumps
must have relief valves. Because of the widely varying heads encountered during
5-35
-------
unloading, care must be taken in pump selection. Piping should have as few joints as
possible and should be Schedule 40 minimum. Splash guards at joints may be desired in traffic
areas. Valves may be gate, plug or diaphragm, iron or steel with bronze trim, and neoprene
plugs in the plug valve or a neoprene disc in the diaphragm valves. Refineries on the West
Coast have adopted a standard of cast steel valves on all flammable materials to prevent
damage during a fire. Couplings must be leak tight and 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 storage equipment.
Care must be exercised to not overflow the storage vessel. A high level alarm and pump
shutoff should be utilized. Due to the increasing cost of methanol, it may be desirable to
have a flowmeter in the unloading piping. All vessels must be vented during unloading or
loading.
5.4.4 Methanol Storage
In order to provide for possible delays in methanol delivery, a storage capability of two to
four weeks supply is recommended. The volume of storage will be determined by various
site and cost requirements; however, storage of less than two weeks is too short for expected
delivery delays and strikes. Tank truck deliveries require in-plant storage. However, with rail
deliveries, the rail cars can be used for storage, but charges (demurrage) are levied by the
carriers for time on site in excess of a fixed time. For small plants, demurrage may become
cost effective. However, carriers may have a time limit on cars or have excessive demurrage
charges.
Methanol may be stored in vertical or horizontal tanks above ground inside or outside, or
buried. It is strongly recommended that all methanol equipment and tanks be located
outside. If interior storage is required refer to reference 43 and 44 for detailed requirements.
An exception to this rule is drum storage which, if not stored indoors, must be shaded from
direct sunlight or constantly sprinkled with water.
Layout of methanol tanks should be in accordance with reference 48. There should also be a
dike around each aboveground tank or group of tanks to contain 125 percent of the largest
tank volume in case of rupture or fire. If the tanks are not of steel, care must be taken so that a
fire will not cause a rupture in the group of tanks thereby overflowing the dike. Fire
protection is very critical, especially when the tanks are near other structures. For large
volumes of methanol storage, low expansion alcohol-type foam is used for fire extinguish-
ing. For very small fires, dry chemical or carbon dioxide extinguishers can be used.
Rate-of-rise or ultraviolet detectors may be used for sensing of fire and initiating automatic
foam release. Water should not be used, but may be used for plant area fire control.
Storage tanks are normally of steel, but most materials are satisfactory except for aluminum.
Tank size is only dependent upon the capacities required and any size limitations imposed
5-36
-------
by the tank material. Piping, valves, etc. should be as described in Section 5.4.3. Ta'nk
fittings should include the following: (1) an inlet with dip tube to prevent splash and static
electricity, (2) an anti-siphon valve or hole on the inlet to prevent back siphonage, (3) vent
pipe with pressure-vacuum relief valve^S with flame arrester, (4) an outlet connection, (5) a
drain connection, and (6) various openings for depth gauges, sample points, level switches,
etc. Manholes for access should also be provided. Also, extreme corrosion will take place if
the tank is drained dry. The tank must also be grounded. Due to increasing air pollution
requirements, venting must be controlled by conservation type vents or by maintaining a
slight negative pressure in the tank using a small ejector.
To maintain a correct inventory of tank contents, a diaphragm level sensor or float should
be used. Low and high level alarms are needed for protection against overfill and settled
material at the bottom of the tank. The high level alarm should be separate from the tank
sensors for a fail-safe design.
5.4.5 Transfer and Feed
Methanol must be transferred and controlled from the storage vessel to the point of feed.
Methanol is removed from the tank and is fed by gravity or pumps. Normally, pumping will
be utilized for ease of control. The transfer pumps should always have positive suction
pressure and should be protected by a strainer. As with all methanol situations, it is
desirable to mount all equipment outside. There are three basic pumping arrangements
which can be used: (1) diaphragm chemical feed pumps using an adjustable stroke for
volume control; (2) positive displacement pumps with variable speed drives controlled by
either counting revolutions to obtain flow or using a flow meter; (3) centrifugal or
regenerative turbine pumps with variable speed drives controlled by a flow meter. Each
arrangement has its own particular problems and must be studied for each installation. To
cover the widest range of feed rates, arrangements 1 and 2 are used due to the limited
accuracy range of flow meters.
All pumps, piping, etc., should be the same as noted in Section 5.4.3. All piping should be
tested for 1.5 times the maximum system pressure for 30 minutes with zero leakage.
Methanol addition to the denitrification process is relatively simple. In the CCCSD's
Advanced Treatment Test Facility, methanol was pumped into the influent line ahead of the
denitrification reactor and the stirring of the reactor was sufficient mixing. In the CCCSD
Water Reclamation Plant, a multi-orifice diffuser is used to evenly distribute methanol in the
channel ahead of denitrification.
5.4.6 Methanol Feed Control
Because methanol is expensive and a methanol overdose can result in a high effluent BOD5,
it is essential to accurately pace methanol with oxidized nitrogen load. Simply pacing
methanol dose against plant flow is inaccurate as it does not account for daily and diurnal
variations in the nitrate concentration. Feed forward control utilizing plant flow and
5-37
-------
nitrification effluent nitrate nitrogen is shown in Figure 5-14. Feed ratio is approximately
three parts methanol per part of nitrate nitrogen by weight (see Section 3.3.2). This method
requires continuous on-line measurement of nitrate utilizing an automated wet chemistry
analyzer. The wet chemistry analyzer (AIT) output is proportional to nitrate concentration
in the nitrification effluent. 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 required methanol flow ratio. This signal
may then be fed to a chemical proportioning pump, as shown, or may be the setpoint of a
flow control loop. FFIK provides 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.
In a research laboratory environment methanol was paced with the use of a Technicon Auto
Analyzer for one week periods between maintenance checks. ^
FIGURE 5-14
FEEDFORWARD CONTROL OF METHANOL
BASED ON FLOW AND NITRATE NITROGEN
NITRIFICATION
EFFLUENT
KEY
FT = Flow Transmitter
FFIK = Ratio Station
AIT = Wet Chemical
Analyzer
HIK = Auto/Manual
Control Station
X = Analog Multiplier
CH3OH
TO
DENITRIFICATION
TANKS
5.4.7 Excess Methanol Removal
Unless specific measures are taken to provide for methanol removal, methanol addition
above stoichiometric requirements (Section 3.3.2) will cause methanol to appear in the
denitrification process effluent.3>4,7, 15,16,27,29,32,49,50 in one instance, a methanol
overdose caused an effluent BODs of 106 mg/1.3 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 good
methanol control system and a methanol removal system as backup should allow nearly
fail-safe operation in terms of preventing effluents containing high levels of organics.
5-38
-------
A recent modification of the suspended growth denitrification process,that has as one ofits
objectives preventing methanol bleedthrough,is shown in Part B of Figure 5-1.2>->l After
denitrification, mixed liquor passes to an aerated stabilization tank. In this tank, facultative
organisms "switch over" from using nitrate to dissolved oxygen and oxidize any remaining
methanol. While refinements undoubtedly could be made in determining the length of time
required for the facultative bacteria to switch over and complete methanol oxidation, it is
known that 30 minutes 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.^2 A period of about 48 minutes has been found to be sufficient for methanol
oxidation.3 Therefore, a period of one hour aeration is recommended based on experience
to date. Further details of this modification concerning solids-liquid separation are
presented in Section 5.6.
In attached growth denitrification systems, the provision of an aerated basin after the
denitrification column would not ensure oxidation of excess methanol. This is because there
is an insufficient mass of facultative organisms in the column effluent to accomplish the
biological oxidation of the carbon, since the denitrifying organisms are retained on the
media and only, a few pass into the column effluent. To date, excess methanol removal
systems applicable to attached growth systems have not been developed.
5.5 Combined Carbon Oxidation-Nitrification-Denitrification Systems with Wastewater and
Endogenous Carbon Sources
The methanol price increases experienced during late 1973 have caused renewed interest in
alternative carbon sources. Alternatives having the least chemical cost for nitrate reduction
are the organics present in domestic wastewater or endogenous respiration of the biological
sludge. The problems experienced in the past with these sources are lower denitrification
rates and contamination of the effluent with the ammonia released when wastewater
organics or biological sludge serve as the carbon source for denitrification. The former
problem has been mitigated by increasing reactor detention time. The latter problem has
been addressed by adapting the suspended growth process configuration in specific ways to
expose the sludge to alternating aerobic and anoxic environments so that released ammonia
is subjected to nitrification and subsequent denitrification. The alternatives which achieve
higher than 80 percent nitrogen removal, while avoiding the use of methanol, combine the
carbon oxidation, nitrification and denitrification processes in single sludge systems with no
intervening clarification steps.
5.5.1 Systems Using Endogenous Respiration in a Sequential Carbon Oxidation-Nitrifi-
cation-Denitrification System
When the process uses the endogenous decay of the organisms for denitrification int the
system, the rate limiting step becomes the organism's endogenous decay rate in an
oxygen-free environment. The process flowsheet for this system is shown on Figure 5-15; it
was developed and first tested in Switzerland^ and subsequently evaluated in other
5-39
-------
locations.54'55-56'57 In the initial tests, fairly high mixed liquor solids (5200-5300 mg/1)
were employed and detention times of 2.2 to 2.8 hours were used in the denitrification
section. Results of the various studies are shown in Table 5-8. Deviations at the other
locations from the Swiss results are explained on the basis of insufficient reaction times in
either the nitrification and denitrification stages.5 9
FIGURE 5-15
SEQUENTIAL CARBON OXIDATION-NITRIFICATION-DENITRIFICATION
RAW
WASTEWATER
[•»
1
COMBINED CARBON
OXIDATION-
NITRIFICATION
-*•
ANOXIC
DENITRIFICATION
_( SECONDARY
"^(SEDIMENTATION
SLUDGE RECYCLE
TABLE 5-8
PILOT TESTS OF WUHRMAN'S SEQUENTIAL CARBON OXIDATION-
NITRIFICATION DENITRIFICATION SYSTEM (AFTER
CHRISTENSEN & HARREMOES, REF. 59)
Location
Switzerland
Germany
Germany
Germany
Seattle, Wa.
New York
Reference
53a
54a
55a
56b
57a
58C
Temperature,
C
13.6
17.1
16
16
12-16
20
20
-
"£)]-,, peak
denitrification rate
Ib NOs rem/Ib MLVSS/day
or kg/k.g/day
0.0168
0.041
0.022
0.026
0.038
0.048
0.026
-
Effluent
NO~-N.
2
-
3
7.8
-
1. 5 to 3. 1
Range of
total nitrogen removal,
percent
82-90
40-60
36-88
46 (average)
31-65
84-89
, pilot scale
° lab scale ,
full scale, 95,000 gpd (360 mVday)
5-40
-------
A modification of the concept tested in New York state involved placing a short anoxic cell
prior to the combined carbon oxidation-nitrification tank.58 An aerobic cell of 0.5 hr
detention time was also placed downstream of the anoxic denitrification, presumably to
improve settling. Total nitrogen removals of 84 to 89 percent were obtained, which are
comparable to the Swiss results.
Denitrification rates determined at the various locations are shown in Figure 5-16. Data for
Blue Plains and Pretoria, S.A. are shown also. In these two pilot plants wastewater was used
as the principal carbon source, but treatment stages using endogenous respiration were
employed also. The data shown for these two pilot plants are for those treatment stages
where endogenous respiration was predominant. When data were reported in terms of MLSS
only, volatile content was assumed to be 70 percent to allow reporting on a basis consistent
with denitrification rates given in Section 5.2.
An envelope has been drawn around all data points in Figure 5-16 to emphasize the
variation in measured denitrification rates. In the absence of pilot data, it would be prudent
to establish reactor sizes on the basis of the lower envelope line shown in Figure 5-16.
FIGURE 5-16
DENITRIFICATION RATES USING ENDOGENOUS CARBON SOURCES
O.IO
C
o
o
'c
o
a,
CO
CO
O.05
£
X)
KEY
SYMBOL
A
T
D
o
A
LOCATION
Switzerland
Germany
Germany
Germany
Seattle,Wa
Blue Plains,DC
Pretoria, S.A.
ENVELOPE FOR ALL DATA POINTS
I I I
10
15 20
TEMPERATURE, C
5-41
25
3O
-------
Further, rates shown in Figure 5-16 are peak nitrate removal rates, qrj, and a safety factor
must be employed in design to ensure low nitrate levels in the effluent, as is done in Section
5.2. It is notable that these rates fall considerably below those shown for methanol as the
carbon source in Figure 5-2. For instance, median values at 20 C for methanol and
endogenous carbon are about 0.25 Ib NO^ -N rem./lb MLVSS/day (0.25 g/g/day) and 0.04
Ib NO^ —N rem./lb MLVSS/day (0.04 g/g/day). Therefore, a denitrification reactor using an
endogenous carbon source would have to be about six times larger than a reactor with a
methanol carbon source at 20 C.
The combined carbon oxidation-nitrification step can be designed with the criteria set forth
in Sections 4.3.3 and 4.3.5, because it has been found that the anoxic denitrification step
has no impact on sludge activity, if the length of the anoxic period is below 5 hours.59,62
In the calculations for the carbon oxidation-nitrification function, only the inventory under
aeration should be included in the solids retention time, growth rate, and removal rate
calculations.
5.5.2 Systems Using Wastewater Carbon in Alternating Aerobic/Anoxic Modes
All of the systems using wastewater as the chief organic carbon source for denitrification use
an alternating aerobic-anoxic sequence of stages, without intermediate clarification, to
effect total nitrogen removal while attempting to avoid ammonia nitrogen bleedthrough.
Some of the demonstrations of these systems have shown that removals of 90 percent were
possible with the alternating mode concept.
5.5.2.1 Aerobic/Anoxic Sequences in Oxidation Ditches
Pasveer was the first to investigate denitrification in oxidation ditches and reported total
nitrogen removals of 90 percent, but few details have been reported.59,63 Subsequent
investigators have confirmed Pasveer's results and defined the conditions required for
denitrification in oxidation ditches.
Ditch design has varied among investigators principally in the means used for aeration. Even
though aeration devices have ranged from Kessner brushes and cage aerators to vertical
turbine mechanical aerators, all the variations can be termed oxidation ditches because all
use the concept of a channel with aeration devices placed at localized points as shown in
Figure 5-17. By controlling the level of aeration, the mixed liquor is exposed to alternating
aerobic and anoxic zones. The channel operates as an "endless channel" since only a portion
of the mixed liquor passing the channel outlet is withdrawn, with the bulk of the flow
remaining in the channel. In this way the mixed liquor is recirculated many times through
aerobic and anoxic zones prior to discharge from the channel.
The largest scale test of the oxidation ditch employed for nitrogen removal has taken place
at the Vienna-B lumen thai plant in Vienna, Austria. 64,65 -p^g plant flowsheet is shown in
Figure 5-18 and design data are given in Table 5-9. The plant consists of a pumping station,
5-42
-------
FIGURE 5-17
PASVEER DITCH OR ENDLESS CHANNEL SYSTEM FOR NITROGEN REMOVAL
RAW WASTEWATER
MIXED
LIQUOR
RETURN SLUDGE
SECONDARY \ EFFLUENT
^SEDIMENTATION
TANK
ROTOR OR OTHER
\ERATION SYSTEM
screens, aerated grit chamber, two aeration tanks, two secondary sedimentation tanks and a
return sludge pumping station. The plant does not incorporate either primary sedimentation
or sludge handling facilities. It was constructed in 1968 at a cost of $ly623,000 (1968 U.S.
dollars).
Since the Vienna-Blumenthal plant is currently operating below design capacity, it has been
possible to operate it in a manner encouraging nitrogen removal. The two aeration tanks
have been connected in series and the number of operational cage aerators varied to
encourage nitrogen removal. It has been found that dissolved oxygen could be measured in
the mixed liquor immediately after the rotor, however the oxygen demand of the
microorganisms caused the oxygen to be depleted prior to contact with the next rotor. This
resulted in an alternating contact of the mixed liquor with aerobic and anoxic zones.
Nitrification took place in the aerobic zones and denitrification occurred in the anoxic
zones. 65
5-43
-------
FIGURE 5-18
VIENNA-BLUMENTHAL WASTEWATER TREATMENT PLANT
EFFLUENT
r
\
v
\
-X
^_
r.
*
v
>>
/
t
PAIR OF MAMMC
AERATION
TANKS
RETURN SLUDGE
PUMP STATION
EXCESS
SLUDGE
TO SEWER
/i
GRIT CHAMBER
SETTLING
TANKS
INFLUENT
SCREEN
PUMP STATION
\
r
OPERATION
BUILDING
5-44
-------
TABLE 5-9
DESIGN DATA FOR THE VIENNA-BLUMENTHAL
TREATMENT PLANT (REFERENCE 65)
Population equivalents 150,000a
SL 3
Average dry weather flow (ADWF) 22.8 mgd (1.0 m /sec)
Aerated grit removal tanks
Number 1
Volume . 9,500 cu ft (270 m )
Detention time, ADWF 4.5. min
Aeration tanks
Number 2
Passes/tank 2
Length/pass 492 ft (150m)
Width/pass 29 ft (8. 5m)
Depth 8. 2 ft (2.5m)
Volume - total 420, 000 cu ft (12, 000m )
Detention time, ADWF 3.3hr
Cage aerator pairs/tank 6
Horsepower each, rotor, pair 56
Secondary sedimentation tanks
Number 2
Diameter 148 ft (45m)
Average depth 10 ft (3. Om)
Overflow rate, ADWF 670 gpd/sf (27.2m3/m2/day)
Estimate current connections, ultimate capacity estimated to be 300,000 population equivalents
The results of special 24-hr tests at the Vienna-Blumenthal plant are shown in Table 5-10.
While in some cases the available data on total nitrogen are incomplete, the ammonia and
nitrate nitrogen data indicate that nitrogen removals of 80 to 90 percent are obtainable.
Special attention should be given to the operating conditions for these tests. While aeration
detention time was only 5.5 to 8.4 hr, the mixed liquor solids were maintained at relatively
high values, 5,300 to 6,700 mg/1. These high MLSS levels were made possible due to the low
overflow rates maintained in the sedimentation tanks (280 to 410 gpd/sf or 11.4 to 16.6
m^/m^/day). U.S. practice would be to use higher overflow rates in sedimentation, reducing
the ability to maintain high MLSS levels under aeration and consequently requiring higher
aeration detention times (see Sections 4.10 and 5.6).
Alternative aeration patterns were evaluated in a recent South African Study.66 Disc-type
aerators were arranged in a series of four concentrically arranged channels. Effluent was
monitored over a full year of weather conditions in the South African study.
One problem that was eventually overcome in the South African work was floating sludge in
the sedimentation tank. This condition was especially prevalent when an anoxic channel
section preceded the sedimentation tank. The condition was corrected by ensuring that each
channel had sufficient aeration to maintain DO in the channel after the disc. Before the disc
5-45
-------
TABLE 5-10
OPERATION AND PERFORMANCE OF THE VIENNA-BLUMENTHAL PLANT
24-HOUR INVESTIGATIONS (REFERENCE 65)
Q
Influent flow rate, mgd (m /sec)
Aeration detention time, hr
MLSS, mg/1
Return sludge SS, mg/1
Average sedimentation tank
Overflow rate, gfid/sL
mVmVday
Aeration tank temperature, C
Operating rotors, tank 1, tank 2
Sludge loading,
Ib BODg/lb MLSS/day
Influent concentrations, mg/1
BOD.
0 •
COD
TOC
Total - N
Ammonia - N
Effluent concentrations, mg/1
BOD.
o
COD
TOC
Total - N
Ammonia - N
Removal efficiencies, percent
BODC
5
COD
TOC
Total - N
Ammonia - N
Sept. 2,
1971
9.6(0.42)
8.0
6,700
10, 000
280
11.4
18
4,2
0.11
268
475
155
36
21.8
13
49
14
4
3.8
95
90
91
88
83
Feb. 17,
1972
14.0(0.61)
5..S
6,200
10,300
410
16.6
12
3, 3
0.13
200
384
126
24
9.0
13
50
13
4
2.7
93
87
90
82
70
June 6,
1974
12.4(0.54)
6.1
5,300
8,000
360
14.7
18
3, 3
0.20
268
463
128
a
21.8
10
39
12
a
6.3
96
92
91
a
71
July 16,
1974
10.9(0.48)
7.0
5,900
9,000
320
13.0
20
4, 2
0.19
a
574
143
a
16.6
a
39
14
a
3.9
a
93
90
b
76
July 31,
1974
9.5(0.42)
8.0
5,400
7,900
280
11.4
19
4, 3
0.16
a
515
134
a
17.2
a
35
13
a
2.4
a
93
90
c
86
Aug. 7,
1974
9.0(0.40)
8.4
5,400
8,400
270
10.9
20
4, 4
0.24
a
778
208
a
21.2
a
35
15
a
2.7
a
92
93
d
87
aNot available
Not available, but no nitrate - N In effluent
°Not available, but only 0.3 mg/1 Nitrate - N In
dNot available, but only 2.4 mg/1 Nitrate - N In
effluent
effluent
aerator, DO was allowed to drop to zero. Since mixed liquor was withdrawn after the disc in
a fresh condition, a sparkling clear effluent was produced.
Table 5-11 summarizes the performance data for the successful configurations of aeration in
the South African study. Nitrogen removal averages 79 percent. A detailed 4-day analysis
during a period when 86 percent nitrogen removal was observed showed that 40 percent of
the influent nitrogen was incorporated in the sludge, 45 percent was nitrogen gas lost to the
atmosphere and 14 percent appeared in the effluent.
5-46
-------
TABLE 5-11
OPERATION AND PERFORMANCE OF OXIDATION DITCH
OPERATED FOR NITROGEN REMOVAL IN SOUTH AFRICA (REFERENCE 66)
Feed flow rate, gpm (1/h)
Retention time, hr
Beturn-sludge ratio
MLSS, mg/1
Water temperature, C
Number of discs used, total
Dissolved oxygen: mg/1
Before aeration discs
After aeration discs
SVI, ml/gram
COD, Ib COD/lb MLVSS/day
(g/g/day)
Influent
COD, mg/1
Kjeldahl-N, mg/1
Ammonia-N, mg/1
Effluent
COD, mg/1
Kjeldahl-N, mg/1
Ammonia-N, mg/1
Nitrite-N, mg/1
Nltrate-N, mg/1
COD removed, percent
N removed, percent
Bun 1
21.4 (4800)
23; 2
2:1
4330
21-22.5
5 (on single shaft)
Virtually always 0
Banging 0. 3-0. 8
213
0.155
749
39.5
21.3
28.2
4.5
3.4
Trace
96; 2
85.2
Bun 2
35.2 (8000)
13.9
1.5:1
4280
20-22
11 (on two shafts)
Positive DO with
only Infrequent
zero readings
Higher by only 0. 2-
0. 3 mg 1 than
before the discs
212
0.28
791
39.4
20.7
30i9
4.05
3.3
0,3
2.65
96,0
81.6
Bun 3
22 (5000)
22; 2
2:1
3720
14-17
5 (on single shaft)
Usually zero or slightly
positive. 0-0- 3 in
channel 1
Oi3-0;6 in chs. 2 and
3. 0.6-1.0 in chs.
1 and 4
263
0:20
678
43; 1
28:2
34.4
6;4
314
Trace
94;6
77.3
Bun 4
26.5-28.6 (6000-6500)
17.1-18.5
3660
14-15. 5
7 (on single shaft,
3-2-1-1)
0-0-4 in first 3 chs.
0-6-2-0 inch. 4.
0.3-1.3 in chs. 1,2 and
3. 1.2-2.3 inch. 4.
268
0.26
723
45.0
29.2
37.4
5.0
2.1 1
0.6
7.1
94.5
72.0
-------
The sludge in the South African study was always in a bulking condition, with SVI levels
ranging from 212 to 268 ml/g. This bulking tendency may be a general property of
alternating aerobic-anoxic processes, as bulking sludges have developed in other alternating
aerobic/anoxic systems (see Sections 5.5.2.3 and 5.5.2.4).
It has been found that the cage aerators which are typically employed in the oxidation ditch
are not well suited to nitrogen removal applications.67 The cage aerator is not capable of
simultaneously mixing and maintaining DO control; too much oxygen is imparted to allow
development of alternating aerobic and anoxic zones while maintaining sufficient ditch
velocities (1 fps or 0.30 m/s) for prevention of settling of solids in the ditch. In one case, the
problem was solved by providing separate submerged propellers for mixing which allowed
the cage aerators to be managed for DO control alone. 67 An aeration system has been
developed that can both control the mixing level and the level of aeration simultaneously.
Vertical shaft aerators are placed at the channel turning points in restricted areas so that the
fluid rotation generated by the aerator is marshalled to move the flow around a semicircular
turn. Power input and the number of on-line aerators can be varied to control DO level
while maintaining sufficient mixing in the system. Nitrogen removals of 80 to 87 percent are
reported.68
5.5.2.2 Denitrification in an Alternating Contact Process
The novel alternating contact process shown in Figure 5-19 has been tested in Denmark in
both lab-scale and full-scale tests at wastewater flows up to 1.5 mgd (0.067 m-Vsec).69 The
process is similar to the oxidation ditch approach in that alternating aerobic and anoxic
residence periods are provided in contact with raw wastewater. The means of accomplishing
this alternating contact are very different from an oxidation ditch, however.
The operational sequence consists of the four phases shown in Figure 5-19. Aeration air is
controlled to provide alternating aerobic and anoxic conditions. When anoxic conditions
were required, the aeration air was merely turned down to a level sufficient to keep the
sludge anoxic but in suspension. Raw wastewater is alternately directed to one or the other
of the two tanks, according to a predetermined cycle. The cycle that an individual tank
passes through is depicted on Figure 5-20. The operational phasing of wastewater addition
and aeration are shown in addition to the phasing of nitrification and denitrification.
Multiple liquid exposure to anoxic and aerobic zones is done by limiting the sequence time
so that an average of 6 to 15 cycles are completed prior to liquid discharge. The effluent is
discharged from the tank to sedimentation only when the tank is under aeration.
The stoichiometric carbon to nitrogen ratio for denitrification was defined in terms of a 7
day BOD and found to be BODy/NO^ —N = 5.2. Denitrification rates are summarized in
Section 5.2.6. Effluent nitrate levels of 2.0 to 5.0 mg/1 appeared obtainable with proper
selection of design and operating mode. A full description of the mathematical model used
for process design can be found in reference 69.
5-48
-------
FIGURE 5-19
ALTERNATING CONTACT PROCESS
Phase 1 - Denitrif ication/Nitrification Phase 3- Nitrification/Denitrification
RAW
WASTEWATER
RAW
TEWATER_
J
I
SL
AN
^
UDGE RETURN
Phase 2 - Intermediate Aeration
RAW
WASTEWATEF^
Phase 4 - Intermediate Aeration
RAW
WASTEWATER
1
SI
1
A
s
-UDGE RETURN
KEY
AN = Anoxic conditions , A = Aerobic conditions, S = Sedimentation
5.5.2.3 The Bardenpho Process
A recent South African development for nitrogen removal using both wastewater and
endogenous carbon for denitrification is shown in Figure 5-21. Termed the "Bardenpho"
process by its developer, the system is a combination of two previously developed
processes.7® Mixed liquor containing nitrate is recycled from the second (aerobic) tank to
the initial (anoxic) tank for denitrification.71 Appended to the first two tanks is a third
(anoxic) tank for removing the nitrate remaining in the effluent from the second (aerobic)
tank. In this third tank endogenous respiration is used for denitrification in the manner
described in Section 5.5.1. Finally, a period of aeration is provided to improve sedimenta-
tion.
Initial lab scale tests of the Bardenpho process showed that nitrogen removals of 93 percent
were possible at recycle rates of 4:1 and 1:1 of mixed liquor and return sludge, respectively.
5-49
-------
FIGURE 5-20
OPERATIONAL SEQUENCING OF ONE OF TWO
AERATION TANKS IN ALTERNATING CONTACT PROCESS
TSEQ
-H
0.5 (TSEQ)
0.5 (TSEQ)
TLP-r
IA
TLP+
IA
Wastewater Add i t ion No Wastewater Add i tion
Effluent to other
Reaction Tank
No Aeration
Eff I. to Sedimentation
Tank
Aeration
i>
Aerobic
Anoxic
Aerob
ic
Anoxic
^^
Denitrification
Nitrification
KEY TO SEQUENCING TIMES
TSEQ = Overall Anoxic-Aerobic Sequence Time
TLP-S. = Lag phase of Denitrification
TLp+ = Lag phase of Nitrification
TIA = Intermediate Aeration Time
Under this condition, 5 to 7 mg/1 of total nitrogen appeared in the effluent.70 A pilot test
of 9 months duration was then conducted to determine the long-term performance of the
system on a scale of 26,000 gpd (100 m3/day). During the last three months of the study,
nitrogen removals were in the range of 80 to 90 percent.61 Observations of denitrification
rates for the first (anoxic) rank are summarized in Section 5.5.2.5 and denitrification rates
for the third (anoxic) tank where endogenous carbon is employed are presented in Section
5.5.1.
5-50
-------
FIGURE 5-21
THE BARDENPHO SYSTEM - SEQUENTIAL UTILIZATION
OF WASTEWATER CARBON AND ENDOGENOUS CARBON
MIXED LIQUOR RETURN
RAW
WASTEWATER'
PRIMARY
EFFLUENT
ANOXIC
DENITRIFI-
CATION
TANK
AEROBIC COMBINED
OXIDATION-
NITRIFICATION
TANK
RETURN
S
-^
LUC
ANOXIC
DENITRIFI-
CATION
TANK
-*
AEROBIC
TANK
>GE
/ SECON
— ^SEDIME
EFFLUENT
The sludge tended to be in bulking condition, settling very little in the standard one-liter
cylinder SVI test. Even with stirring a relatively high SVI was obtained (150 ml/g). This
bulking sludge condition appears to be typical for alternating aerobic/anoxic systems, and
care must be employed in design and operation to deal with this problem.
The Bardenpho process has been tested in a second 26,000 gpd (100 m^/day) pilot plant at
Pretoria, South Africa.72 While full test results were not available, test data from the initial
few weeks of operation are shown in Table 5-12. These data demonstrate that relatively high
nitrogen removals are obtainable with the Bardenpho process.
TABLE 5-12
PERFORMANCE OF THE "BARDENPHO" PROCESS
AT PRETORIA, SOUTH AFRICA (REFERENCE 72)
Parameter
Influent COD, mg/1
Effluent COD, mg/1
Percent COD removal
Influent TKNa, mg/1
Effluent TKN, mg/1
Percent TKN removal
Effluent nitrate - N, mg/1
Influent total nitrogen , mg/1
Effluent total nitrogen , mg/1
Percent total nitrogen removal
Period
Jan. 7 to Jan. 31,
1975
226
46
79
21.1
1.9
91
2.6
21.1
4.5
79
Feb. 2 to Feb. 14,
1975
176
48
73
15.9
1.4
91
1.7
15.9
3.1
81
TKN = total Kjeldahl nitrogen
Assuming influent oxidized nitrogen is zero
Assuming effluent nitrite nitrogen is zero
5-51
-------
5.5.2.4 Alternating Aerobic/Anoxic System Without Internal Recycle
Investigators at the EPA Blue Plains pilot plant conceived of still another way to achieve
alternating aerobic and anoxic environments with the system shown in"Figure 5-22.60 A
two pass aeration tank was provided with separate aeration and mixing facilities. In each
basin, 2 mechanical mixers were employed to keep the mixed liquor in suspension,
independent of the aeration system. Air was supplied alternately to each basin; first to one
basin and then to the other in a 30 minute cycle.6® Dissolved oxygen level in the pass under
aeration was controlled between 2 and 3 mg/1, while the anoxic pass decreased to zero
rapidly after cessation of aeration. The pilot process was typically operated at a flow rate of
50,000 gpd( 189 m3/day).
FIGURE 5-22
BLUE PLAINS ALTERNATING ANOXIC AEROBIC SYSTEM (REF. 60)
D C. RAW WASTEWATER
FeCI3-
PRIMARY
AIR
i
C\
j U
o fx, Q
A I
NITRIFICATION DENITRIFICATION
REACTOR
WASTE —
RETURN SLUDGE
SECONDARY
—I I—
CLARIFIER
ALUM
FILTERS
FINAL EFFLUENT
5-52
-------
A summary of operating and performance data for the 9 months of test work is shown in
Table 5-13. During 9 months of operation, the plant was operated at a F/M of
approximately 0.1 Ib BODs/lb MLVSS/day (0.1 g/g/day), expressed on the basis of the total
inventory in the reactor. This F/M was sufficiently low to permit the development of a
mixed culture of organisms for carbon oxidation, nitrification and denitrification.
Nitrogen removals during the study varied from 54 to 84 percent, but operational problems
contributed to the lower reported removals. These problems may be avoidable in full-scale
operation. For instance, in July and August, ferric chloride addition in the primary
treatment stage reduced the COD/TKN ratio from 10 to about 7.5 to 8.0, and resulted in a
situation in which the lack of a sufficient organic carbon source limited the degree of
denitrification obtainable. The lower removals experienced in April and May were due to
the fact that during a portion of each month, the alternate aerobic/anoxic regime was
altered to a full aerobic mode to rid the system of filamentous growth.
During September, performance deteriorated for an unexpected reason. The flow sheet in
Figure 5-22 was modified to include two more reactor stages prior to the clarifier. In the
first added stage, an anoxic one, methanol was added to cause denitrification of residual
nitrite and nitrate. An aerobic stabilization step was added as the last stage. It was found
that methanol addition caused an immediate ammonia increase in the process. Subsequent
studies showed that methanol is toxic to nitrifiers.
Since all operating problems can be explained, it can be concluded that the system is
capable of 84 percent nitrogen removal in the summer (23 C) and 75 percent nitrogen
removal in the winter (14 C).
Like in other alternating aerobic/anoxic studies, it was found that a severe filamentous
bulking condition developed in the sludge, limiting wintertime clarifier overflow rates to
about 300 gpd/sf.60 Bulking sludge has been observed at low temperatures at Blue Plains
and at low F/M operation at Blue Plains and at other plants. Whatever the cause of the
bulking problems, it appears that clarifier operation will limit operation of this
aerobic/anoxic system, as it will limit the other aerobic/anoxic systems.
Kinetic data was obtained during the study. On and off aeration of samples of the pilot
plant mixed liquor was employed over several cycles to simulate operation of the pilot unit.
Nitrification and denitrification rates determined by this procedure are shown in Table 5-14.
Nitrification rates were similar through all cycles, while denitrification rates decreased as the
batch reaction continued. Denitrification rates measured in the first cycle represent peak
rates possible when a readily available carbon source is available during denitrification. By
the third or fourth cycle, the rates represent denitrification when the readily available
carbon is depleted and an endogenous carbon source is used. Denitrification rates for cycle 1
are also presented in Section 5.5.2.5 with measurements of other observers, while rates for
cycle 4 are presented in Section 5.5.1.
5-53
-------
TABLE 5-13
SUMMARY OF OPERATION AND PERFORMANCE FOR THE BLUE PLAINS
ALTERNATING AEROBIC/ANOXIC SYSTEM (REFERENCE 60)
Month
1973
Jan
Feb
March
April
May
.Tune
July
August
Sept
Det.
time
hr
12i3
12.3
12.3
12.4
10.5
8.8
6.8
6.6
8.7
F/M
lb BOD applied/
tb MLVSS/day
or g/g/day
0.072
0.066
0.10
0.081
0.089
0.105
0.093
0.089
0.11
MLSS
mg/1
(%
volatile)
3510
(74)
3980
(73)
2950
(73)
3540
(67)
4170
(69)
4010
(69)
3040
(64)
3200
(57)
3700
(65)
SVI,
sl
g
245
250
330
277
227
188
133
134
-
COD/
TKN
ratio
9.6
9.9
10.5
10.6
10.0
10.3
7.9
7.5
10.0
Temp. ,
C
14.0
14.2
15.5
-.
-
23.0
25.0
25.5
26.0
Influent quality mg/1
BOD5
96.5
99
110
98.8
115
107
51
44.2
99
SS
110
108
128
120
109
112
153
197
110
Total
Kjeldahl
Nitrogen
25.7
23.2
24.8
21.7
23.3
24.0
15.0
14.9
22.6
Effluent quality, mg/1 a
BOD,.
o
20.4
14.0
L
6.5b
5.3b
3.3b
3.2b
t.
3.8b
U
2.6b
7.2b
SS
15.4
14.3
15.0
13.0
11.8
7.8
9.0
10.0
16.0
Total
Kjeldahl
Nitrogen
2.28
1.52
4.20
5.20
1.36
1.51
2.14
1.23
10.2°
NOg and
NOJ - N
3.99
4.41
2.30
6.03
8.25
2.30
2.72
3.74
0.22
Removals, percent
BOD5
79
86
94b
95b
97b
97b
93b
94b
93b
SS
85
87
88
89
89
93
94
95
85
Total
Nitrogen
76
75
74
49
59
84
68
67
54
Prior to filtration
bNltrificatlon inhibited
Ammonia level was 9.4 mg/1 as N (see text)
-------
TABLE 5-14
OBSERVED NITRIFICATION AND DENITRIFICATION RATES FOR
BLUE PLAINS ALTERNATING ANOXIC/AEROBIC SYSTEM
Mode
Aerobic-
Nitrification
Anoxic-
Denitrification
Temp. ,
C
15.5
25.0
27.0
26.5
15.5
25.0
27.0
26.5
Peak nitrification or denitrification
removal rate, Ib N/lb MLVSS/day
Cycle 1
0.032
0.083
0.11
0.12
0.032
0.055
0.042
0.026
Cycle 2
0.042
0.095
0.11
-
0.029
0.030
-
0.0075
Cycle 3
0.016
-
-
-
0.021
0.033
-
-
Cycle 4
0.026
-
-
-
0.019
0.030
-
-
Cycle 5
0.035
-
-
-
-
-
-
5.5.2.5 Kinetic Design of Alternating Aerobic/Anoxic Systems
The four factors which can limit denitrification process efficiency in alternating aerobic/
anoxic systems using wastewater as the carbon source are as follows:^
1. Nitrification
2. Denitrification
3. Carbon-nitrogen ratio
4. Operational mode (process hydraulics)
The third factor has been evaluated by several investigators (see Sections 5.5.2.2 and
5.5.2.4) and further discussion here is unnecessary.
5-55
-------
To evaluate nitrification limitations on the system, nitrogen loads and nitrification rates
must be taken into account. Most investigators agree that the design of the combined carbon
oxidation-nitrification functions of the aerobic phase can be separated from the anoxic
phase. 61,69,73 jj has been found that anoxic periods up to 5 hours have no impact on
aerobic sludge activity. 62,5 9 Therefore, the carbon oxidation and nitrification calculations
for the aerobic periods can be virtually identical to those advanced for combined carbon
oxidation-nitrification in Sections 4.3.3 and 4.3.5. In the calculations of nitrifier solids
retention time, nitrifier growth rate,and removal rates in the aerobic residence periods, only
the solids inventory under aeration is employed. This is because the environment must be
aerobic for nitrifier growth to occur. As always, a safety factor must be employed.
Sizing of denitrification steps must consider nitrite load and nitrate removal rates and
consideration of the safety factor concept in design. Of the two models formulated for these
systems, the safety factor concept is used only for the nitrification step in the Bardenpho
design^ but for some unapparent reason not for denitrification. 61 The safety factor concept
was not used for nitrification or denitrification in the alternating contact process design
(Section 5.5.2.2).69 it must be emphasized that unless a safety factor is incorporated in the
design, nitrogen removal will deteriorate under peak load conditions.
Most of the alternating processes employ both wastewater carbon and endogenous carbon
for denitrification at some point in the system. Observed denitrification rates for
endogenous carbon have been summarized in Section 5.5.1. Experimentally determined
denitrification rates in alternating aerobic/anoxic systems with wastewater as the carbon
source are shown in Figure 5-23. When data were reported in terms of MLSS only, volatile
content was assumed to be 70 percent to allow reporting on a basis consistent with the
denitrification rates shown in Section 5.2. These rates are peak nitrate removal rates, and are
expressed as q^, using the terminology developed in Section 3.3.5.2. As can be seen from
Figure 5-23, there is a wide variation in measured denitrification rates in systems using
wastewater as the organic carbon source. As a result, it may not be a conservative practice to
use the denitrification rates given in Figure 5-23; rather, it would appear prudent to conduct
pilot investigations to verify design parameters for denitrification when wastewater is the
carbon source for denitrification.
The rates for denitrification with wastewater as the carbon source fall below those found for
methanol as the carbon source shown in Figure 5-2. Median rates at 20 C for methanol and
wastewater carbon are about 0.25 Ib NO^ -N removed/lb MLVSS/day (0.25 g/g/day) and
0.07 Ib NO3 -N removed/lb MLVSS/day (0.07 g/g/day), respectively. A denitrification
reactor using a wastewater carbon source would have to be about three and one-half times
larger than a denitrification reactor using methanol as the carbon source.
One cause of the difference in reaction rates between methanol and wastewater carbon
relates to biological availability. Methanol is a simple, easily degraded compound, whereas
wastewater contains a mix of easily degraded and hard to degrade compounds. Wastewaters
may vary in the relative distribution of easily degraded and hard to degrade compounds,
thus causing variations in denitrification rates between locations.
5-56
-------
FIGURE 5-23
EFFECT OF TEMPERATURE ON PEAK DENITRIFICATION
RATES WITH WASTEWATER AS CARBON SOURCE
U. 10
O
\
to
to
^
5
e
u o.io
2»
Qj
N
•^
O
•Q
Qj
^.
O
tfc
c 0.05
.0
o
O
lj
M •
U
H—
c
0)
Cl
>0*
n
i 1 1 1
KEY
SYMBOL LOCATION REF.
• Blue Plains 60
• Austin, Texas 74
D Pretoria, S. A. 61
A Denmark 69 D
~ D ~
a
a
a
•
•
— —
A •
•
•
1 1 1 1
IO
15 20
TEMPERATURE, C
25
30
The hydraulic mode of operation significantly affects the kinetic design procedure. In all of
these systems, relatively complex and lengthy mass balances are necessary to describe the
system. None-the-less, such descriptions are possible and have been developed for two of the
alternating aerobic/anoxic systems.61.69 These models are presented in sufficient detail in
the literature to allow their modification for use in design. Iterative solution of equations is
required, and the digital computer has proven a useful design tool.69 The limitation of these
models is that generally applicable kinetic rate data are not yet available.
5-57
-------
5.6 Solids-Liquid Separation
The considerations for design of sedimentation tanks for denitrification systems are the
same as those discussed for nitrification systems in Section 4.10 and those points common
to both will not be repeated herein.
Rising sludge has occasionally plagued denitrification systems, depending on system
design.^A?, 15,49,50 jo remedy this and other problems, the original suspended growth
denitrification system (using methanol) was modified by placing an aerated stabilization
step between the anoxic denitrification reactor and the denitrification clarifier (Figure
5-lB).2>3,21,51 This step was taken because it was found that the "conventional" design
was basically an unstable process. In the conventional system, the methanol: nitrogen ratio
(M:N ratio) had to be kept at precisely the optimum level (2.5 to 3.0). When methanol was
overfed, the effluent BOD5 would rise. When methanol was underfed, nitrate would bleed
through to the clarifier, and denitrification would proceed in the clarifier using the sludge as
the carbon source. Floating sludge, buoyed up by nitrogen gas bubbles, caused a severe
deterioration in effluent quality.
The coupling of an anoxic residence period and an aerobic residence period in the modified
system (Figure 5-1) is based on the recognition that dissimilatory denitrification is
accomplished by facultative bacteria using biochemical pathways that are almost identical to
aerobic biochemical pathways. The main difference in the biochemical pathways lies in the
electron transport system where the terminal enzyme is changed and nitrate replaces oxygen
as the final electron acceptor. 75 These facultative bacteria can shift rapidly from using
nitrate to using oxygen and vice versa. In the aerobic tank, the excess methanol is oxidized
and the mixed liquor solids are aerobically stabilized. The aerobic tank also serves the
purpose of stripping supersaturated nitrogen gas from solution so that nitrogen gas bubbles
will not form during sedimentation.
A mildly aerated physical conditioning channel transfers the denitrification mixed liquor to
the final clarifier. Recognizing that the very turbulent conditions in the aerated stabilization
tank causes floe breakup and dispersed fines in suspension, the purpose of the channel is to
allow these dispersed particles to be incorporated into floe under mild turbulence conditions
that favor aggregation over breakup. 76
Tests of the original and modified systems were conducted at the Central Contra Costa
Sanitary District's Advanced Treatment Test Facility.2>3,21,51 Comparing performance
with stabilization to performance without it (Tables 5-15 and 5-16), indicates the
substantial merits of the aerated stabilization tank. The test period without stabilization was
one in which daily adjustments were made in the methanol feed rate, hence there was little
if any methanol bled into the effluent. This is reflected in the low soluble BODs of 5 mg/l.
During the test period with aerated stabilization, less careful control was exerted in
methanol feed which resulted in a fairly high M/N ratio of 3.3. Despite this overfeeding of
methanol, the effluent soluble BODs remained 5 mg/l. In other words, the aerated
5-58
-------
stabilization tank formed a favorable environment for the oxidation of the excess methanol.
One significant factor contributing to less suspended solids and turbidity in the effluent was
the development of a ciliate and rotifer population in the culture. Previously these
organisms were not abundant. With a significant aerobic residence period, these organisms
TABLE 5-15
EFFECT OF STABILIZATION TANK ON DENITRIFIED EFFLUENT AT THE
CENTRAL CONTRA COSTA SANITARY DISTRICT'S
ADVANCED TREATMENT TEST FACILITY (REFERENCE 3)
Constituent
Nitrate as N
Total BODg
Filtered BOD,
o
Suspended Solids
Turbidity
Total organic carbon
Soluble organic carbon
Temperature
Mean effluent quality
without stabilization, mg/1
(Feb. 13 to Mar. 13, 1972)
0.5
37
5
14
5a
18
7
16 to 17b
Mean effluent quality
with stabilization, mg/1
(Mar. 28 to April 20, 1972)
0.7
6
5
4
1.3a
8.6
5
16 to 19b
JTU
b
degrees C
TABLE 5-16
DENITRIFICATION PROCESS PARAMETERS AT THE CENTRAL CONTRA COSTA
SANITARY DISTRICT'S ADVANCED TREATMENT TEST FACILITY (REF. 3)
Parameter
Flow, mgd
Residence time, hr
reactor
stabilization tank
MLSS, mg/1
SVI, ml/g
Nitrogen rem Ib/lb MLVSS/day
Methanol/ nitrate - H ratio
Without stabilization
Feb. 13 to
Mar. 13, 1972
.46
.85
0
3000
143
.23
2.8
With stabilization
Mar. 28 to
April 20, 1972
.47
.82
.79
2500
242
.18
3.3
5-59
-------
could flourish and clarify the liquid. Sulfide odors in the sludge were eliminated by the
modification.
In addition to stabilizing the liquid, the sludge is stabilized as well. Without the stabilization
tank, solids appearing in the effluent contained about 2.3 Ib BOD5 per Ib of SS. With
stabilization, effluent solids contain less BOD5 with a BOD5/SS ratio of 0.25. The net effect
of this is to reduce the effluent BOD5 from 37 to 6 mg/1. This reduction in BOD5 value of
the solids is very likely due to enhanced endogenous respiration in the stabilization tank.
Another indication of sludge stabilization in the aerated stabilization tank is the reduction
in denitrification rates observed in the modified system compared to the conventional
system. Prior to the modification, denitrification rates were as high as 0.3 to 0.56 Ib Nitrate
-N rem./lb MLVSS/day at 16 to 18 C and these rates were below the peak limiting rate, qpj,
as effluent nitrate was always low. This range in rates is roughly twice the range in rates
shown for the system employing aerated stabilization at CCCSD as shown in Figure 5-2.
While one impact of aerated stabilization is to decrease denitrification rates and therefore
increase anoxic reactor requirements, the other effect of this impact is to render the sludge
less active and more resistant to the rising sludge problem in the denitrification clarifier flue
to denitrification. This is in agreement with the conclusions drawn in Section 4.10, where it
is shown that the tendency for rising sludge was related to denitrification rates and sludge
residence time in the clarifier. Rapid sludge removal equipment, such as the vacuum pickup
type, should be provided in all sedimentation tanks to minimize sludge residence time and
reduce the likelihood of rising sludge.
Similar to the findings with methanol based denitrification, an aerobic step has been
usefully employed in combined carbon oxidation-nitrification-denitrification systems.
Generally, a 1 to 2 hr residence period is provided prior to mixed liquor separation in the
sedimentation step.
Reactor-clarifier interactions should not be ignored. For instance design examples are
presented in the literature for alternating aerobic/anoxic systems where the mixed liquor
solids are assumed to be 5000 mg/1 (at 14 C) and 7000 mg/1 (at 10 C).61>69 Current U.S.
practice is to limit mixed liquor levels below 3000 mg/1 unless clarifier overflow rates are
reduced to account for the need to thicken and return the sludge.40 Operation at 5000 to
7000 mg/1 would require very large clarifiers to ensure that solids are not lost at peak wet
weather flow conditions. Bulking sludge tends to occur in the combined carbon-oxidation-
nitrification-denitrification systems which mandates even greater conservatism in allowable
mixed liquor levels and clarifier design than would normally be the case.
5.7 Considerations for Process Selection
The choice of denitrification for nitrogen removal mandates the process sequence of
nitrification-denitrification for nitrogen removal. Two kinds of comparisons in denitrifica-
tion process selection can be made. First, the nitrification-denitrification sequence can be
5-60
-------
compared to the physical-chemical alternatives. Second, if nitrification-denitrification is
chosen, comparisons must be made among the denitrification alternatives to make the
process selection.
5.7.1 Comparison to Physical-Chemical Alternatives
The comparison made between the nitrification portion of the sequence and the
physical-chemical alternatives in Section 4.11.1 need not be repeated here.
Total dissolved solids (TDS) increment in the nitrogen removal system will have a bearing on
process selection in many situations. Nitrification-denitrification leads to a net reduction in
wastewater alkalinity and no change in the total mineral content of the water. Both the
breakpoint chlorination and the selective ion exchange process lead to TDS increases.
5.7.2 Choice Among Alternative Denitrification Systems
Many of the considerations presented in Section 4.11 are applicable to denitrification
system selection and are not repeated here. Other factors affecting process choice are
summarized in Table 5-17. Most of these factors were considered earlier in this chapter.
In some treatment plants, both nitrogen and phosphorus removal has been mandated. The
combined carbon oxidation-nitrification-denitrification systems are somewhat restricted in
the sequencing of the phosphorus removal step. Chemical addition in the primary treatment
stage cannot be employed, as this would not leave sufficient organic carbon influent to the
process to complete denitrification.
Another factor listed in Table 5-17 is stability of operation and degree of nitrogen removal.
Several long-term tests of denitrification systems using methanol have successfully
demonstrated consistently high levels of nitrogen removal. Equivalent operational exper-
ience with the combined carbon oxidation-nitrification-denitrification systems will soon be
obtained, as large-scale experimental work is currently underway in the U.S., Denmark, and
South Africa. When the results of this work are available, fair comparisons can be made
between the two types of systems. Based upon presently available data, it appears that the
combined carbon oxidation-nitrification-denitrification systems are capable of 75 to 90
percent nitrogen removal; in comparison, methanol based systems can achieve 90 to 95
percent nitrogen removal.
One other kind of comparison can be made between the systems using methanol and the
combined systems. Since the rate of denitrification with wastewater as the carbon source in
the combined system is lower than with methanol as the carbon source, greater
denitrification reactor sizes are required for the combined system. This issue can best be
analyzed with an example comparing the alternative systems. A design example for the
Bardenpho process has been presented in the literature that can be usefully employed for
the comparison.61 Specific loading criteria are not important in this example; for these the
5-61
-------
TABLE 5-17
COMPARISON OF DENITRIFICATION ALTERNATIVES
System Type
Advantages
Disadvantages
Suspended growth using
methanol following a
nitrification stage
Denltrification rapid, small structures required
Demonstrated stability of operation
Few limitations in treatment sequence options
Excess methanol oxidation step can be easily
incorporated
Each process in the system can be separately
optimized
High degree of nitrogen removal possible
Methanol required
Stability of operation linked
to clarifier for biomass return
Greater number of unit processes
required for nitrification-de-
nitrification than in combined
systems
Attached growth (column)
using methanol following
a nitrification stage
Denitrification rapid, small structures required
Demonstrated stability of operation
Stability not linked to clarifier as organisms
on media
Few limitations in treatment sequence options
High degree of nitrogen removal possible
Each process in the system can be separately
optimized
Methanol required
Excess methanol oxidation process
not easily incorporated
Greater number of unit processes
required for nitrification-
denitrification than in combined
system
Combined carbon oxi-
dation-nitrification-
denitrlfication in sus-
pended growth reactor
using endogenous
carbon source
No methanol required
Lesser number of unit processes required
Denitrification rates very low;
very large structures required
Lower nitrogen removal than in
methanol based system
Stability of operation linked to
clarifier for biomass return
Treatment sequence options
limited when both N and P
removal required
No protection provided for
nitrifiers against toxicants
Difficult to optimize nitrification
and denltrification separately
Combined carbon oxi-
dalion-nitrification-
denltrlflcation in
suspended growth
reactor using wastewater
carbon source
No methanol required
Lesser number of unit processes required
Denltrification rates low; large
structures required
Lower nitrogen removal than In
methanol based system
Stability of operation linked to
clarifier for biomass return
Tendency for development of
sludge bulking
Treatment sequence options
limited when both N and P
removal required
No protection provided for
nitrifiers against toxicants
Difficult to optimize nitrification
and denitrlficatlon separately
reader is referred to reference 61. Reactor residence times are as provided in reference 61 at
a temperature of 14 C excepting that the MLSS value has been downwardly adjusted from
5000 to 3000 mg/1, according to U.S. practice. The effect of this is to increase by the ratio
of 5/3 the detention times in the reactors in the design example. The adjusted residence
times are shown in Figure 5-24. A methanol-based system useful for comparison purposes is
also shown in Figure 5-24. In this sytem, a combined carbon oxidation-nitrification step is
chosen, since if a combined operation provides acceptable treatment for the Bardenpho
5-62
-------
INFLUENT
WASTEWATER
FIGURE 5-24
COMPARISON OF DENITRIFICATION SYSTEMS
RETURN SLUDGE
ANOXIC
DENITRIFICATION
TANK (4HR)
I
AEROBIC
COMBINED CARBON
OXIDATION-
NITRIFICATION
TANK
(10 HR)
I
UJ
oc
(£.
O
ID
O
ANOXIC
DENITRIFICATION
TANK (5HR)
I
AEROBIC TANK (2 HR)
I
SEDIMENTATION
TANK
(8 HR)
CO
INFLUENT
WASTEWATER
COMBINED CARBON
OXIDATION-
NITRIFICATION
TANK
(10 HR)
UJ
o
o
ID
_l
V)
UJ
QL
INTERMEDIATE
SEDIMENTATION
TANK (4 HR)
METHANOL
DENITRIFICATION
TANK (2 HR)
AEROBIC TANK (l HR)
I
FINAL
SEDIMENTATION
TANK (4 HR)
UJ
o
o
3
_l
CO
UJ
cc
DENITRIFIED
EFFLUENT
A. Bardenpho Process (29 hr)
5-63
DENITRIFIED
EFFLUENT
B. Alternative Methanol
Based System (21 hr)
-------
process it would also work effectively for the comparative case. Solids retention time (and
hydraulic detention time) of the nitrification step would be the same as for the nitrification
tank in the Bardenpho process.
Since denitrification rates are more rapid, the denitrification tank in the methanol based
system can be proportionately smaller than in the Bardenpho process. Interpolating
denitrification rates from Figures 5-2 and 5-23, about 2 hours would be required. A
residence time of 4 hours is assumed for sedimentation tanks in the methanol-based system,
whereas 8 hours is assumed for the Bardenpho process due to the bulking tendency of the
sludge. Comparing the two alternative systems, it can beseen that the alternating aerobic/anoxic
system requires greater tankage than the methanol-based system (29 hours compared to 21
hours). In terms of economics, differences in the systems can be seen as a trade-off of
capital cost (tankage) with operating cost (methanol).
5.8 References
1. Barth, E.F., Brenner, R.C., and R.F. Lewis, Chemical-Biological Control of Nitrogen in
Wastewater Effluent. JWPCF, 40, No. 12, pp 2040-2054 (1968).
2. Parker D.S., Zadick, F.J., and K.E. Train, Sludge Processing for Combined Physical-
Chemical-Biological Sludges. Prepared for the EPA, Report No. R2-73-250, July, 1973.
3. Horstkotte, G.A., Niles, D.G., Parker, D.S., and D.H. Caldwell, Full-Scale Testing of a
Water Reclamation System. JWPCF, 46 No. 1, pp 181-197 (1974).
4. Mulbarger, M.C., The Three Sludge Systems for Nitrogen and Phosphorus Removal.
Presented at the 44th Annual Conference of the Water Pollution Control Federation,
San Francisco, California, October, 1971.
5. Sawyer, C.N., Wild, H.E., Jr., and T.C. McMahon, Nitrification and Denitrification
Facilities, Wastewater Treatment. Prepared for the EPA Technology Transfer Program,
August, 1973.
6. Parker D.S., Case Histories of Nitrification and Denitrification Facilities. Prepared for
the EPA Technology Transfer Program, May, 1974.
7. Murphy, K.L., and P.M. Sutton, Pilot Scale Studies on Biological Denitrification.
Presented at the 7th International Conference on Water Pollution Research, Paris,
September, 1974.
8. Bishop, D.F., Personal communication to D.S. Parker. Environmental Protection
Agency, Washington, D.C., April, 1974.
9. Mulbarger, M.C., Nitrification and Denitrification in Activated Sludge Systems.
JWPCF, 43, No. 10, pp 2059-2070 (1971).
5-64
-------
10. Dawson, R.N., and K.L. Murphy, Factors Affecting Biological Denitriflcation in
Wastewater. In Advances in Water Pollution Research, S.H. Jenkins, Ed., Oxford,
England: Pergamon Press, 1973.
11. Stensel, H.D., Loehr, R.C., and A.W. Lawrence, Biological Kinetics of Suspended
Growth Denitriflcation. JWPCF, 43, No. 2, pp 249-261 (1973).
12. Sutton, P.M., Murphy, K.L., and R.N. Dawson, Continuous Biological Denitriflcation
of Wastewater. Environmental Protection Service (Canada), Report EPS 4-WP-74-6,
August, 1974.
13. Lawrence, A.W., and P.L. McCarty, Unified Basis for Biological Treatment Design and
Operation. JSED, Proc. ASCE, 96, No. SA3, pp 757-778 (1970).
14. Brown and Caldwell, Project Report for the Water Reclamation Plant. Report to the
Central Contra Costa Sanitary District, November, 1971.
15. Weddle, C.L., Niles, D.G., Goldman, E., and J.W. Porter, Studies of Municipal
Wastewater Renovation for Industrial Water. Presented at the 44th Annual Conference
of the Water Pollution Control Federation, San Francisco, California, October, 1971.
16. Horstkotte, G.A., Jr., Pilot Demonstration Project for Industrial Reuse of Renovated
Municipal Wastewater. Prepared for the EPA, EPA-670/2-72-064, August, 1973.
17. Requa, D.A., and E.D. Schroeder, Kinetics of Packed Bed Denitriflcation. JWPCF, 45,
No. 8, pp 1696-1707(1973).
18. Smith, J.M., Masse, A.N., Feige, W.A., and L.J. Kamphake, Nitrogen Removal from
Municipal Wastewater by Columnar Denitriflcation. Environmental Science and Tech-
nology, 6, p 260 (1972).
19. Jewell, W.J., and R.J. Cummings, Denitriflcation of Concentrated Wastewaters.
Presented at the Water Pollution Control Federation, Cleveland, October, 1973.
20. Sutton, P.M., Murphy, K.L., and R.N. Dawson, Low Temperature Biological
Denitriflcation of Wastewater. JWPCF, 47, No. 1, pp 122-134 (1975).
21. Parker, D.S., Aberley, R.C., and D.H. Caldwell, Development and Implementation of
Biological Denitriflcation for Two Large Plants. Presented at the Conference on
Nitrogen as a Water Pollutant, sponsored by the IAWPR, Copenhagen, Denmark,
August, 1975.
22. Denitriflcation for Anaerobic Filters and Ponds, Phase II. Robert S. Kerr Water
Research Center, EPA WPCRS 13030 ELY 06/71-14, June, 1971.
5-65
-------
23. Denitrification for Anaerobic Filters and Ponds. Robert S. Kerr Water Research Center,
EPAWPCRS 13030 ELY 04171-8, April, 1971.
24. Description of the El Logo, Texas. Advanced Wastewater Treatment Plant. Seabrook,
.Texas: Harris County Water Control and Improvement District Number 50, March,
1974.
25. Requa, D.A., Kinetics of Packed Bed Denitrification. Thesis submitted in partial
satisfaction of the requirements for the degree of Master of Science in Engineering,
University of California at Davis, 1970.
26. English, J.N., Carry, C.W., Masse, A.M., Pitkin, J.B., and F.D. Dryden, Denitrification
in Granular Carbon and Sand Columns. JWPCF, 46, No. 1, pp 28-42 (1974).
27. Savage, E.S., and J.J. Chen, Operating Experiences with Columnar Denitrification.
Pittsburgh, Pennsylvania: Dravo Corporation, 1973.
28. Lamb, G.L., Nitrogen Removal Utilizing Single Stage Activated Sludge and Deep Bed
Filtration. Presented at the Joint Meeting of the Sanitary Engineering Section, ASCE,
and the Metropolitan Section, NYWPCA, New York, N.Y., March, 1972.
29. Ecotrol, Inc., Biological Denitrification Using Fluidized Bed Technology. August,
1974.
30. Jeris, J., Beer, C., and J.A. Mueller, High Rate Biological Denitrification Using a
Granular Fluidized Bed. JWPCF, 46, No. 9, pp 2118-2128 (1974).
31. Chen, J.J., Letter communication to D.S. Parker, Dravo Corporation, November, 1974.
32. Kapoor, S.A., and T.E. Wilson, Biological Denitrification on Deep Bed Filters at
Tampa, Florida. Unpublished paper, Greeley and Hansen, Engineers, Chicago, 111., (no
date).
33. Hansen, S.E., Letter communication to D.S. Parker. Neptune MicroFloc Inc., Corvallis,
Oregon, September 11, 1974.
34. Duddles, G.A., Richardson, S.E., and E.F. Earth, Plastic Medium Trickling Filters for
Biological Nitrogen Control. JWPCF, 46, No. 5, pp 937-946 (1974).
35. Duddles, G.A. and S.E. Richardson, Application of Plastic Media Trickling Filters for
Biological Nitrification. Report prepared for the Environmental Protection Agency,
EPA-R2-73-199, June, 1973.
36. Process Design Manual for Suspended Solids Removal. U.S. EPA, Office of Technology
Transfer, Washington, D.C., January, 1975.
5-66
-------
37. Jens, J.S., and F. J. Flood, Plant Gets New Process. Water and Wastewater Engineering,
pp 45-48, March, 1974.
38. Jeris, John S., and R.W. Owens, .Pilot Scale High Rate Biological Denitrification at
Nassau County, N. Y. Presented at the Winter Meeting of the New York Water Pollution
Control Association, January, 1974.
39. Owens, R., Letter communication to D.S. Parker, Ecolotrol Corp., October, 1974.
40. Process Design Manual for Upgrading Existing Wastewater Treatment Plants. U.S. EPA,
Office of Technology Transfer, Washington, D.C. (1974).
41. Ecolotrol, Inc., Hy-Flo Fluidized Bed Denitrification, Ecolotrol Technical Bulletin, No.
123-A, November, 1974.
42. Sullivan, R.H., Cohen, M.M., Ure, J.E., and F. Parkinson, The Swirl Concentrator as a
Grit Separator Device. Prepared for the EPA by the American Public Works
Association, EPA-670/2-74-026, June, 1974.
43. Manufacturing Chemists Association, Properties and Essential Information for Safe
Handling and Use ofMethanol, Chemical Safety Data Sheet SD-221, 1970.
44. Austin, George T., Industrially Significant Organic Chemicals — Part 7. Chemical
Engineering, 81, No. 13, pp 152-153 (1974).
45. Manufacturing Chemists Association, Recommended Practice-Unloading Flammable
Liquids from Tank Cars, Manual Sheet TC-4. 1969.
46. Manufacturing Chemists Association, Tank Car Approach Platforms, Manual Sheet
TC- 7. 1965.
47. National Fire Protection Association, Static Electricity 1966, NFPA No. 77. 1966.
48. National Fire Protection Association, Flammable and Combustible Liquids Code 1972,
NFPA No. 30, 1972.
49. Smith, A.G., Denitrification Reactor Studies in a Lime Treated Sewage Plant. Paper
No. 2029, Ontario Ministry of the Environment, Research Branch, July, 1972.
50. Tenney, M.W., and W.F. Echelberger, Removal of Organic and Eutrophying Pollutants
by Chemical-Biological Treatment. Prepared for the EPA, Report No. R2-72-076
(NTISPB-214628), April, 1972.
5-67
-------
51. Parker, D.S., and D.G. Niles, Full-Scale Test Plant at Contra Costa Turns Out Valuable
Data On Advanced Treatment. The Bulletin (of the California Water Pollution Control
Association), 9, No. 1, pp 25-27 (1972).
52. Sutton, P.M., Murphy, K.L., and B.E. Sank, Nitrogen Control: A Basis for Design With
Activated Sludge Systems. Presented at the Conference on Nitrogen as a Water
Pollutant, sponsored by the IAWPR, Copenhagen, Denmark, August, 1975.
53. Wuhrmann, K., Nitrogen Removal In Sewage Treatment Processes. Verh. Int. Ver.
Limnol., 15, pp 580-596 (1964).
54. Hunerberg and Sarfert, Experiments on the Elimination of Nitrogen from Berlin
Sewage. Gas u. Wasserfach, Wasser-Abwasser, 108, pp 966-969 and 1197-1205 (1967).
55. Hamm, The Influence on Denitrification of Phosphate Precipitants. Z. Wasser and
Abwasserforschung, 2, pp 180-182 (1969).
56. Schuster, Laboratory Experiments on Removal of Nitrogen Compounds from
Domestic Sewage. Fortschr. Wasserchem. Greng., 12, pp 139-148 (1970).
57. Carlson, D., Nitrate Removal from Activated Sludge Systems. Report for OWRR,
Project AO26, July, 1970.
58. Beer, C., Hetling, L.J., and R.E. McKinney, Nitrogen Removal and Phosphorus
Precipitation in a Compartmentalized Aeration Tank. Technical Paper 32, New York
State Department of Environmental Conservation, February, 1974.
59. Christensen, M.H., and P. Harremoes, Biological Denitrification in Wastewater
Treatment. Report 2-72, Department of Sanitary Engineering, Technical University of
Denmark, 1972.
60. Bishop, D.F., Heidman, J.A., and J.B. Stamberg, Single Stage Nitrification-
Denitrification. Presented at the 47th Annual Conference of the Water Pollution
Control Federation, Denver, Colorado, October 6-11, 1974.
61. Barnard, J.L., Cut P and N Without Chemicals. Water and Wastes Engineering, 11, No.
7, pp 33-36 (1974) and No. 8, pp 41-94 (1974).
62. Wuhrmann, K., Effect of Oxygen Tension on Biological Treatment Processes. Proc.
Third Conference Biological Waste Treatment, Manhattan College, N.Y., 1960, pp
27-38.
63. Pasveer, A., Contribution on Nitrogen Removal from Sewage. Muncher Bertrage zur
Abwasser-Fisherei-and Flussbiologie, Bd 12, pp 197-200 (1965).
5-68
-------
64. Matsche, N., The Elimination of Nitrogen in the Treatment Plant of Vienna-
Blumenthal. Water Research, 6, No. 4/5, pp 485-486 (1972).
65. Matsche, N.F., and G. Spatzierer, Austrian Plant Knocks Out Nitrogen. Water and
Wastewater Engineering, 12, No. 1, pp 18-24 (1975).
66. Drews, R.J.L.C., and A.M. Greeff, Nitrogen Elimination by Rapid Alternation of
Aerobic/"Anoxic" Conditions in "Orbal" Activated Sludge Plants. Water Research, 7,
pp 1183-1194(1973).
67. Halvorson, H.O., Irgens, R., and H. Bauer, Channel Aeration Activated Sludge
Treatment at Glenwood, Minnesota. JWPCF, 44, No. 12, pp 2266-2276 (1972).
68. Zemaitis, W.L., Letter communication to D.S. Parker. Envirobic Systems, Inc., New
York, New York, September 27, 1974.
69. Christensen, M.H., Denitrification of Sewage by Alternating Process Operation.
Presented at the IAWPR, Paris, September, 1974.
70. Barnard, J.L., Biological Denitrification. Presented at the monthly meeting of the
South African Branch of the Institute of Water Pollution Control, August, 1972.
71. Ludzack, F.J., and M.B. Ettinger, Controlling Operation to Minimize Activated Sludge
Effluent Nitrogen. JWPCF, 34, No. 9, pp 920-931 (1962).
72. van Vuuren, L.R.J., Letter communication to D.S. Parker, February, 1975.
73. Gujer, W. and D. Jenkins, The Contact Stabilization Process-Oxygen and Nitrogen Mass
Balances. University of California, Sanitary Engineering Research Lab, SERL Report
74-2, February, 1974.
74. Balakrishnan, S., and W.W. Eckenfielder, Nitrogen Relationships in Biological Waste
Treatment Processes — III, Denitrification in the Modified Activated Sludge Process.
Water Research, 3, pp 177-188 (1969).
75. Schroeder, E.D., and A.W. Busch, The Role of Nitrate Nitrogen In Bioxidation.
JWPCF, 40, No. 11, pp R445-R457 (1968).
76. Parker, D.S., Kaufman, W.J., and D. Jenkins, Physical Conditioning of Activated
Sludge Floe. JWPCF, 43, No. 9, pp 1817-1833 (1971).
5-69
-------
CHAPTER 6
BREAKPOINT CHLORINATION
6.1 Process Chemistry
When chlorine is added to dilute aqueous solutions containing ammonia nitrogen, reactions
occur which may lead ultimately to oxidation of the ammonium ion to end products
composed predominantly of nitrogen gas. When such chemical processes are performed in
water and wastewater treatment for the purpose of ammonia nitrogen removal, the
procedure is termed breakpoint chlorination. This chapter discusses the theoretical
stoichiometry of breakpoint chlorination, presents the practical process considerations
which influence actual chemical consumption, reaction end products and rate of the
reaction, and presents process design criteria.
Recent work at the Blue Plains wastewater treatment pilot plant in Washington, D.C.I>2>3
has confirmed breakpoint chlorination reaction products. Gas chromatography was used at
Blue Plains to identify breakpoint reaction products from buffered aqueous wastewater
samples in laboratory tests. Further confirmation of breakpoint reaction end products was
obtained in pilot scale investigations with wastewater effluent of different qualities.
Breakpoint chlorination tests on domestic wastewaters at the Blue Plains pilot plant showed
that 95 to 99 percent of the ammonia nitrogen in solution is converted to nitrogen gas.2>3
No breakpoint reaction intermediate compounds of N2O, NO or NO2 were detected. The
oxidized ammonia nitrogen fraction which did not appear as nitrogen gas was found to be
made up of nitrate and nitrogen trichloride.
6.1.1 Chemical Stoichiometry
When chlorine gas is dissolved in water, hydrolysis of the chlorine molecule occurs according
to the following relationship:
C12 + H20 ?=Z HOC1 + H+ + Cf (6-1)
(hypochlorous acid)
The active (oxidizing) forms of chlorine in solution are hypochlorous acid and its
dissociation product, hypochlorite ion.
HOC1 5^=r OC1~ + H+ K = 3.3 x 10~8 (6-2)
at20C
where: K = dissociation constant
6-1
-------
The fraction of the total chlorine residual in a sample which is made up of hypochlorous
acid and hypochlorite ion is termed the "fiee available" chlorine residual. The rate of
dissociation of HOC 1 is very rapid and equilibrium proportions are maintained even when
HOC1 is continuously being reacted. The equilibrium relationship between HOC1 and OC1~
in relation to solution pH is shown in Figure 6-1.
FIGURE 6-1
RELATIVE AMOUNTS OF HOC1 AND OCf AT
VARIOUS pH LEVELS (REFERENCE 4)
too
too
10 It 12
The oxidizing capability of the free available chlorine residual is manifest in the chemical
transformation of hypochlorous acid to chloride ion (Cl~). This transformation involves a
gain of two electrons and a valence change of the chlorine atom from "+1" to "-1".
Ammonia nitrogen concentrations of 10 mg/1 to 40 mg/1 may be found in typical
municipal wastewater treatment plant effluents. The source of ammonia nitrogen typically
includes direct discharge from industrial processes and release following hydrolysis of urea
and biological degradation of amino acids and other organic derivatives of ammonia
nitrogen. The actual chemical form of ammonia nitrogen in solution is pH and temperature
6-2
-------
dependent. The relative distribution of ammonia nitrogen and ammonium ion may be
defined according to the equation below:
NH,
K = 5.0x10
at20C
-10
(6-3)
This relationship is indicated graphically in Figure 6-2 in relation to pH of the solution.
Reactions between chlorine and ammonium in dilute aqueous solution can proceed
according to a number of competing pathways. Formation of chloramines, termed
"combined residual" and nitrate can occur in the following manner:
+ HOC1
NH2C1 (monochloramine) + F^O + H
(6-4)
NH2C1 + HOC1
NHCL, (dichloramine) + HO
^ 2
(6-5)
FIGURE 6-2
EFFECTS OF pH AND TEMPERATURE ON DISTRIBUTION
OF AMMONIA AND AMMONIUM ION IN WATER
IOO
10 II 12
IOO
6-3
-------
NHC12 + HOC1 - - NC13 (nitrogen trichloride) + H2O (6-6)
and
NH* + 4HOC1 - - HN03 + 5H+ + 4Cf + H2O (6-7)
The reactions are dependent upon certain process variables, including pH, temperature,
contact time, and the initial chlorine to ammonia nitrogen ratio
Breakpoint chlorination occurs when sufficient chlorine has been added to a water or
wastewater sample to cause the chemical oxidation of the ammonium in solution to
nitrogen gas and other end products.
Significant aspects of breakpoint chlorination process chemistry which were studied during
the Blue Plains pilot work included identification of the predominant end products of the
breakpoint reaction. Tests on municipal wastewater and wastewater treatment plant effluent
at Blue Plains indicated that 95 to 99 percent of the ammonia nitrogen in solution was
converted to nitrogen gas. Nitrate and nitrogen trichloride account for the remaining
fraction. The overall reaction between the ammonium ion and chlorine leading to formation
of nitrogen gas may be expressed in terms of the simplified equations below:
NH* + HOC1 - - NH2C1 + H2O + H+ (6-4)
NH2C1 + 0.5 HOC1 - - 0.5 N2 + 0.5 H2O + 1 .5 H+ + 1 ,5 Cf (6-8)
NH* + 1.5 HOC1 ^ 0.5 N2 + 1.5 H2O + 2.5 H+ + 1.5 Cl (6-9)
Stoichiometrically, the breakpoint reaction of Equation 6-9 requires a weight ratio of
chlorine to ammonia nitrogen at the breakpoint of 7.6:1, as shown below:
v
Molecular Weight HOC1 = 70.9 (expressed as CU)
Molecular Weight NH. = 14.0 (expressed as N)
Therefore, weight ratio of CLrNH, -N at breakpoint:
C12:NH+-N = (1.5X70.9): (0(14.0) = 7.6:1
6-4
-------
Therefore 7.6 parts of chlorine are theoretically required to chemically oxidize one part of
ammonia nitrogen in aqueous solution. In practice, the actual weight ratio of chlorine to
ammonia nitrogen at breakpoint has ranged from about 8:1 to 10:1. Many of the process
variables which are known to affect the total chemical requirement for this process have
been identified and those factors are discussed in subsequent sections.
6.1.2 The Breakpoint Curve
The breakpoint chlorination curve is a graphical representation of chemical relationships
which exist as varying amounts of chlorine are added to dilute solutions of ammonia
nitrogen. An investigation in 1939 led to the discovery that increasing the chlorine dose in
certain waters resulted in an overall reduction in the chlorine residual measured in the water
sample following contact. ^ The point of maximum reduction of chlorine residual was
termed the "breakpoint".
The theoretical breakpoint curve shown in Figure 6-3 has several characteristic features. The
characteristics of the breakpoint curve shown in Zone 1 include principally the reaction
between chlorine and ammonium indicated in Equation 6-4. The hump of the breakpoint
curve occurs, theoretically, at a chlorine to ammonia nitrogen weight ratio of 5:1 (molar
ratio of 1:1). That ratio corresponds to the point at which the reacting molecules are
present in solution in equal numbers. "
The chemical equilibria of Zone 2 favor the formation of dichloramine (Equation 6-5) and
the oxidation of ammonium according to Equation 6-9. These reactions proceed in
competition to, theoretically, a Cl2:NH4-N weight ratio of 7.6:1. At the breakpoint, the
ammonium concentration is minimized.
To the right of breakpoint, Zone 3 chemical equilibria include the build-up of free chlorine
residual as well as the presence of small quantities of dichloramine (Equation 6-5), nitrogen
trichloride (Equation 6-6), and nitrate (Equation 6-7)1 The free chlorine residuals which
result from dosages beyond breakpoint are known to be considerably more bactericidally
potent than the combined residuals found at lower chlorine dosages (See Section 6.2.7).
6.2 Process Application Considerations
The basic theoretical background chemistry must be combined with application funda-
mentals if proper process designs and operations are to be achieved in full scale wastewater
treatment practice. Much background work on breakpoint chlorination has been done on a
laboratory "pure system" basis and in potable water. This information is useful, but not
always applicable to wastewater treatment considerations. Generally, the presence of high
concentrations of ammonia nitrogen and other amino substances as well as the presence of
other chemical constituents in wastewater effluents contribute to the discrepancy between
some data available from laboratory testing and that collected in actual wastewater
treatment applications.
6-5
-------
FIGURE 6-3
THEORETICAL BREAKPOINT CHLORINATION CURVE
TC TAL
CHLORINE
APPLIED
MEASURED
CHLORINE
RESIDUAL
BREAKPOINT-
IRREDUCIBLE
MINIMUM CHLORIN
RESIDUAL
1
5 7.6
CI2:NH^-N WEIGHT RATIO
6.2.1 Chlorine Dosage Requirement
The total amount of chlorine which must be added to wastewater to achieve breakpoint is
affected by the chemical nature of the wastewater and by the conditions which exist in the
zone where the reacting species come into contact. Several important factors which should
be considered in breakpoint chlorination process application and design are discussed below.
6.2.1.1 Effect of Pretreatment
The degree of treatment which a wastewater stream receives prior to breakpoint
chlorination effects both the chlorine dosage and end-product distribution. The relationship
between pretreatment and nitrate and nitrogen trichloride formation is discussed in Section
6.2.2.
6-6
-------
The chlorine demand of a wastewater treatment plant effluent sample is the total chlorine
oxidative capacity consumed during a given period of time by substances in solution which
do not result in measurable chlorine residual or chemical oxidation of ammonia nitrogen.
This is the chlorine oxidative capacity which is essentially "lost" from participation in the
desired breakpoint reaction. Chlorine demand may be exerted by a number of substances
commonly present in wastewater, including S^— ,HS~SC)2~ ,NO~ Fe2+, phenols, amino
acids, proteins and carbohydrates. 9
Generally, as the degree of pretreatment of wastewater is increased, the chlorine demand
exerted by the substances noted above is reduced. One particularly important factor is that
if a treatment process employing anoxic conditions precedes a chlorination facility,
substances in solution may be converted from an oxidized to a reduced form and the
chlorine demand may be substantially increased.
Laboratory and pilot plant studies of breakpoint chlorination at the Blue Plains pilot
plant ^>2 and at Sunnyvale 10»H have shown that increasing levels of pretreatment decrease
the amount of chlorine required to achieve breakpoint. Table 6-1 shows that laboratory
tests on buffered distilled water containing only ammonia nitrogen reached breakpoint at a
Cl2:NH4-N ratio of 8:1, a level near that predicted by chemical stoichiometry (7.6:1). In
comparison, raw wastewater required a Cl2:NH4-N of 9:1 - 10:1 to reach breakpoint.
Ammonia nitrogen concentrations in the samples following breakpoint were found to be
consistently in the range of 0.2 mg/1 or less. Pilot plant scale testing of breakpoint
chlorination processes has confirmed chlorine dosages predicted through laboratory work.
6.2.1.2 Effect of pH and Temperature
1 9
Laboratory studies at Blue Plains > in which buffered distilled ammonia nitrogen solutions
of 20 mg/1 concentration were subjected to breakpoint chlorination dosages showed a
definite optimum pH for breakpoint in the range of pH 6 to 7. The chlorine dosage at
optimum pH levels was found to be Cl2:NH4-N of 8:1. Breakpoint tests conducted outside
the apparent optimum range of pH 6 to 7 showed an appreciably higher chlorine
requirement for breakpoint and slower reaction rates.
Comparable tests carried out with filtered secondary effluent did not show a clearly defined
relationship between pH and Cl2:NH4-N to reach breakpoint. Formation of other
nitrogenous residuals (NO~3 and NC13) was considered to be the controlling criteria in
selection of the optimum pH operating range of pH 6 to 7 (See Section 6.2.2).
There is no evidence that ordinary variations in the temperature of wastewater effluents
affect the C^NHiJ-N to reach breakpoint.
6.2.1.3 Initial Mixing of Chlorine
The significance of initial mixing in certain unit processes of sanitary engineering has been
6-7
-------
TABLE 6-1
EFFECT OF PRETREATMENT ON Cl^NHj-N BREAKPOINT RATIO
Sample
Buffered water
Raw wastewater
Lime clarified
raw wastewater
Secondary effluent
Lime clarified
secondary effluent
Ferric chloride
clarified raw
wastewater-
carbon adsorption
Filtered secondary
effluent
Lime clarified raw
wastewater-filtered
Alum clarified
oxidation pond
effluent-filtered
Breakpoint
PH
6-7
6.5 - 7.5
6.5 - 7.5
6.5 - 7.5
6.5 - 7.5
3.2
6-8
7.0 - 7.3
6.6
Initial
NH . -N
(mg/D
Final
NH4 -N
(mg/1)
Irreducible
minimum
residual
(mg/1 as C12)
Laboratory Tests
20
15
11.2
8.1
9.2
10.2
0.1
0.2
0.1
0.2
0.1
0.1
0.6
7
7
3
4
20
Pilot Plant Tests
12.9 - 21.0
9.7 - 12.5
20.6
0.1
0.4 - 1.2
0.1
2 - 8.5
-
7.6
Breakpoint
ratio
C12:NH+ -N
(weight basis)
8:1
9:1 - 10:1
8:1-9:1
8:1-9:1
8:1
8.2:1
8.4:1 - 9.2:1
9:1
9.6:1
Ref
2
2
2
2
2
12
2
2
11
oo
-------
amply demonstrated. Tests 13 have shown significantly improved alum coagulation
efficiency as a direct result of increasing the level of turbulence of mixing in the zone of
alum application to a water sample. Improved disinfection efficiency in laboratory tests was
also noted when the application of chlorine to a wastewater effluent was accomplished at
increased levels of turbulent mixing.
Recent data from Blue Plains^ have shown the total chlorine dose required to reach
breakpoint is not affected by initial mixing conditions to the degree which was first
reported. A number of pilot plant tests with secondary effluent in which the chlorine was
dosed into a mechanical mixer showed no difference in the efficiency of chlorine utilization
whether or not the mixer was operating. Other laboratory studies have shown the
Cl2:NH4-N ratio at breakpoint to be unaffected by the degree of mixing in the reaction
zone.
In plant-scale design of breakpoint chlorination facilities, a quantity of hydraulic or
mechanical energy sufficient to facilitate rapid and thorough blending of the chlorine
solution, pH adjustment chemical and process influent should be provided. The blending of
chemicals with the process influent initiates the breakpoint reactions and allows completion
of the breakpoint reaction within the contact zone, provides the basis for containment of
process odors and assures process consistency, a necessary prerequisite for the feedback
element of process control (Section 6.3).
6.2.2 Residual Nitrogenous Materials
Nitrate (NO^) and nitrogen trichloride (NC13) are occasionally found in the effluent from
the breakpoint chlorination process. Both compounds can be found in varying con-
centrations, depending upon the degree of pretreatment and pH in the reaction zone. The
total concentration of these residuals seldom exceeds 10 percent of the influent ammonia
nitrogen concentration.2
Nitrogen trichloride is a particularly volatile compound which exhibits a very strong
chlorinous odor. It is an extremely strong-oxidizing agent, having been used for many years
in the bleaching of flour. Formation of NC13 in breakpoint chlorination, even in fairly small
concentrations, is undesirable because of the obnoxious and dangerous characteristics of the
compound in the gaseous form. Nitrate formation in breakpoint chlorination should be
avoided since it represents a reaction product which has consumed considerable amounts of
chlorine (Equation 6-7) and because it reduces the overall nitrogen removal capability of the
breakpoint process.
Table 6-2 presents data on the effect of wastewater pretreatment on the formation of
residual nitrogenous materials at breakpoint. * >2 Although the investigators concluded that
NC13 formation decreased with decreasing pretreatment, it appears that very highly treated
effluent may be breakpointed with the production of negligible levels of NC13. Nitrate
production was similar at each treatment level.
6-9
-------
TABLE 6-2
EFFECT OF PRETREATMENT ON FORMATION OF
NITROGENOUS RESIDUALS AT BREAKPOINT (REFERENCE 1, 2)
Sample
Raw Wastewater
Lime Clarified and
Filtered Raw Wastewater
Secondary Effluent
Lime Clarified and
Filtered Secondary
Effluent
Initial
Ammonia-N Cone .
(mg/1)
15.0
11.2
8.1
9.2
NC13 Cone, at
Breakpoint
(mg/1 as N)a
0.0
0.25
0.13
0.0
NO^ Cone . at
Breakpoint
(mg/1 as N)a
0.3
0.45
0.24
0.2
pH Range = 6.5 to 7.5
Laboratory and pilot-scale tests have confirmed the pH sensitivity of NC13 formation in
breakpoint chlorination. Pilot-scale tests of filtered secondary effluent showed 0.33 mg/1
NC13 (as N) after breakpoint chlorination at pH 6; at pH 7 and above, NC13 concentration
was reduced to about 0.05 mg/1. Nitrate formation was found to be slightly affected by pH
at breakpoint. Nitrate concentration in pilot tests of filtered secondary effluent ranged from
about 0.7 mg/1 (as N) at pH 6 to 1.0 mg/1 (as N) at pH 8.
Figure 6-4 shows the consequences of chlorine dosages beyond breakpoint on the formation
of nitrogenous residuals. Nitrate production following breakpoint chlorination of lime
clarified filtered secondary effluent (breakpoint at Cl2:NH4-N of 8:1) showed a linear
increase in concentration at increased chlorine dosages. Nitrogen trichloride formation was
noted beginning at C^Ntfy-N ratios exceeding 9:1. Close control of chlorine dosage levels
in breakpoint is shown to be an important factor in minimizing production of these
residuals.
Pilot testing of breakpoint chlorination at Blue Plains showed that variations in temperature
from 5 to 40 C did not affect the final distribution of nitrogenous residuals in the effluent.
Awareness of the pH sensitivity of formation of NO3 and NC13 has led to formulation of
two specific recommendations for design of plant-scale breakpoint chlorination facilities.
First, the pH adjustment chemical (Section 6.2.3) should be added to the chlorine solution
prior to application of the chlorine solution to the breakpoint chlorination process influent.
Without premixing of chemicals, disproportionate mixing of chlorine solution and pH
6-10
-------
FIGURE 6-4
EFFECT OF Cl^NH^ -N ON NITROGEN RESIDUALS IN
LIME CLARIFIED FILTERED SECONDARY EFFLUENT (REF. 1)
O.8
0.7
^ °6
* 0.5
o
« O4
\ 0-3
Uj
%
AMMONIA-N CONC. = 9.2 mg/t
pH RANGE 6.5-7.5
Cl2:NHj-N DOSAGE AT BREAKPOINT = 7:1 to 8:1
O.2
O.I
O.O £* &-
NITRATE
-o-
-^
NITROGEN
TRICHLORIDE
8
IO II
12
CL2--NH4~N WT. RATIO
adjustment chemical in the breakpoint reaction zone can result in "pockets" of liquid in the
breakpoint reaction zone not at the desired pH level. Occurrence of the breakpoint reaction
in such "pockets" could lead to formation of excessive concentrations of NO^ and NC13.
A second design recommendation is that the operating pH for breakpoint chlorination
should be maintained at pH 7. This pH allows optimal Cl2:NH4-N ratios and reaction rates
and also reduces the NC13 production and attendant odor production to about 0.2 mg/1 or
less, as long as careful control over chlorine dosage is maintained.
In summary, the design engineer should be cautioned that a poorly designed or maintained
system for breakpoint chlorination pH and chlorine dosage control can result in the
generation of chlorinous odors due to the formation of nitrogen trichloride. However, it
appears that if appropriate care is exercised in the design and maintenance of pH and
chlorine dosage systems, the breakpoint chlorination process can function satisfactorily
without the need for odor control facilities.
6-11
-------
6.2.3 Alkalinity Supplementation
In the addition of chlorine to a wastewater sample for the purposes of breakpoint
chlorination, acidity is generated through the hydrolysis of chlorine gas (Equation 6-1) and
oxidation of ammonium (Equation 6-9). Theoretically, four moles of hydrogen ions are
generated for every mole of ammonia nitrogen which is oxidized with chlorine gas.
l.SCU+l.SILO - +• 1.5HOC1+1.5 H++ 1.5 Cf (6-1)
NH* + 1 .5 HOC1 - - 0.5 N2 + 1 .5 H2O + 2.5 H+ + 1 .5 Cf (6-9)
1.5 C12 + NH* - 0.5 N2 + 4H+ + 3C1 (6-10)
The buffering capacity (alkalinity) of the breakpoint process influent is consumed by the
acidity of Equation 6-10. Stoichiometrically, 14.3 mg/1 of alkalinity (as CaCO3) is
consumed for each 1.0 mg/1 of ammonia nitrogen which is oxidized in breakpoint
chlorination. In actual practice, around 15 mg/1 alkalinity is consumed due to the hydrolysis
of chlorine needed beyond that predicted by stoichiometry.
It is apparent that if the ammonia nitrogen concentration in the breakpoint process influent
is high, or if the wastewater alkalinity is relatively low, insufficient buffering capacity will
be available to maintain the process pH at a reasonable level. Any alkaline substance may be
used to supplement the natural alkalinity for pH control but sodium hydroxide (NaOH) and
lime (CaO) are the compounds most commonly used. If all of the acidity generated in
breakpoint must be neutralized through chemical addition, 1.50 parts of sodium hydroxide
(NaOH) would be needed for each part of chlorine, or 1.05 parts of lime (CaO) would be
needed for each part of chlorine. In practice, the alkalinity of the process influent may
supply a portion of the total buffering capacity needed under breakpoint conditions. If a
highly alkaline stream is to be treated (such as might result from a lime precipitation
process), alkalinity supplementation would not be needed and, perhaps, acid addition would
be needed to achieve the recommended operating pH of 7. In situations where breakpoint
chlorination is used for removing low ammonia nitrogen residuals (< 3 mg/1) from
nitrificaton process effluents, there may be sufficient remaining alkalinity to buffer the
water without the need for addition of neutralizing chemicals.
6.2.4 Reaction Rates
The reaction rate for breakpoint chlorination has not been measured quantitatively, but
several investigations^! 1 have noted that the reaction is very rapid. According to reaction
6-12
-------
rates established by Morris,6>7 the reaction between ammonium and hypochlorous acid
(Equation 6-4) to form monochloramine is 99 percent complete in 0.2 seconds at pH 7 and
of 0.2:1. The optimum reaction pH was shown to be pH 8.3, allowing 99
percent conversion to monochloramine in 0.069 seconds. Obviously, the rate of this
reaction is important in breakpoint since it is the first of several sequential reactions.
Laboratory studies have shown the breakpoint reaction rate to vary, depending upon pH. At
pH levels between 6 and 7, the breakpoint reaction was found to proceed to completion in
less than 15 seconds when secondary effluents were tested. When reaction rates at pH levels
outside the optimum range of pH 6-7 were tested, the rate was found to be slowed
considerably. At pH 3.5, for example, the breakpoint chlorination reactions were not
complete following two hours of contact.
The rapid rates noted for breakpoint reactions conducted in the pH range of 6-7 lead to the
design recommendation that a breakpoint chlorination contact period of one minute is
sufficient for plant-scale applications. The design of the contact basin should provide, as
closely as possible, a plug flow contact regime. The blending of breakpoint chemicals with
the process influent should be carried out as indicated in Section 6.2. 1.3.
6.2.5 Effect on Total Dissolved Solids
In many instances where high levels of wastewater treatment are required, total dissolved
solids (TDS) limitations may be a controlling criterion in the selection of alternative
treatment processes. Breakpoint chlorination can involve the addition of large quantities of
chemicals to solution. The TDS increment attributable to each of several chemicals which
may be utilized in breakpoint chlorination is summarized in Table 6-3.
TABLE 6-3
EFFECTS OF CHEMICAL ADDITION ON TOTAL
DISSOLVED SOLIDS IN BREAKPOINT CHLORINATION
Chemical Addition
TDS Increase : NH4~N Consumed
Breakpoint with chlorine gas
Breakpoint with sodium
hypochlorite
Breakpoint with chlorine gas -
Neutralization of all
acidity with lime (CaO)
Breakpoint with chlorine gas -
Neutralization of all acidity
with sodium hydroxide (NaOH)
6.2 : 1
7.1 : 1
12.2 : 1
14.8 : 1
6-13
-------
If breakpoint chlorination is contemplated on a wastewater effluent stream which contains
20 mg/1 ammonia nitrogen, the increase of TDS in solution following addition of chlorine in
the gaseous form would amount to 124 mg/1. If all of the acidity generated in the hydrolysis
of chlorine and the oxidation of ammonium is neutralized with lime (CaO), the total
increase in TDS would amount to 244 mg/1.
6.2.6 Reactions with Organic Nitrogen
Studies at Blue Plains^ found only "slight reduction in organic nitrogen within the two hour
contact time." The data of Lawrence, et al. ^ showed a decrease in organic nitrogen from
3.2 mg/1-3.5 mg/1 to levels of 0.2 mg/1 - 0.4 mg/1. More recent data^ collected in
Sunnyvale, California also show an apparent decrease in soluble organic nitrogen following
breakpoint (Table 6-4).
TABLE 6-4
EFFECT OF BREAKPOINT CHLORINATION
ON SOLUBLE ORGANIC NITROGEN (REFERENCE 1 l)a
Date, 1973
8/30
9/4
9/11
9/12
Soluble Organic Nitrogen
before Breakpoint
(mg/1 as N)
2.7
2.8
4.6
4.7
Soluble Organic Nitrogen
after Breakpoint
(mg/1 as N)
1.0
1.7
1.3
2.0
Soluble organic nitrogen determination conducted on filtrate
from 0.45/x membrane filter. Wastewater treated was tertiary
treated oxidation pond effluent
reported in 1953 that the concentration of unsubstituted ammo nitrogen of many
common amino acids was reduced slowly by reaction with chlorine. Organic nitrogen in the
more complex form of proteins was practically unaffected by chlorine over a period of
several days. The degree of organic nitrogen reduction through breakpoint chlorination is
likely a function of the relative proportion of proteins to the simpler hydrolytic products
(including amino acids) of the protein molecules. In summary, the true reductions in organic
nitrogen with breakpoint chlorination are difficult to predict.
The presence of organic nitrogen in solution at breakpoint has also been shown to effect the
shape of the breakpoint curve. ^ It has been noted that waters containing a mixture of
ammonia nitrogen and organic nitrogen did not display the classic "dip" of the breakpoint
curve. 17,18,19 The irreducible minimum residual was found to be appreciably greater when
6-14
-------
organic nitrogen was present than when the sample contained only the inorganic ammonia
nitrogen form.
6.2.7 Disinfection
Under normal conditions, disinfection of a non-nitrified wastewater effluent stream with
chlorine is mainly accomplished by the form of combined chlorine residual known as
monochloramine (NH2C1 at pH 7.0). When contact for disinfection is provided downstream
from the breakpoint chlorination process, the free chlorine residual in solution following
breakpoint will provide much higher bactericidal potential than with monochloramine
alone.
Figure 6-5 is a comparison of the germicidal efficiency of hypochlorous acid, hypochlorite
ion and monochloramine. Figure 6-1 shows the ionization characteristics of the hypo-
chlorous acid at various pH levels. The data of Figure 6-5 show that hypochlorous acid is a
far more effective germicidal agent than either hypochlorite ion or monochloramine.
Formation of hypochlorous acid following breakpoint chlorination will, therefore,
considerably enhance the capability of a wastewater disinfection system if an efficient
contacting system is .provided.
6.3 Process Control Instrumentation
Both South African^! and American researchers2>22 have reported that if a continuously
functioning breakpoint chlorination process is to be a consistent and reliable environmental
engineering unit process, the system must be capable of responding rapidly to changes in
ammonia nitrogen concentration, chlorine demand, pH, alkalinity and flow.
6.3.1 Process Control System
Failure of the chlorine dosage control system to respond adequately to changes in process
conditions can result in a substantial loss in nitrogen removal capability as well as
potentially adding significant overdoses of chlorine to the process. Overdoses of chlorine to
the system are a direct waste of this chemical, they result in increased difficulty of pacing
the dechlorination equipment,and can cause the direct discharge of high concentrations of
chlorine residuals to the receiving water.
6.3.1.1 Chlorine Dosage Control
A function diagram of the breakpoint chlorination process control system is shown in
Figure 6-6. This system is the same as that used in the pilot plant testing of breakpoint
chlorination at the Blue Plains pilot plant in Washington, D.C. The control of chlorine
dosage is accomplished by a combination feed foreward and feedback control loop. The
feed foreward component utilizes ammonia concentration and flow signals together with a
manually selected multiplier to establish an approximate chlorine dosage to achieve
6-15
-------
FIGURE 6-5
COMPARISON OF GERMICIDAL EFFICIENCY OF
HYPOCHLOROUS ACID, HYPOCHLORITE ION, AND
MONOCHLORAMINE FOR 99 PERCENT DESTRUCTION
OF E. COLI AT 2-6 C (REF. 20)
MONOCHLORAMINE
HYPOCHLORITE
ION
HYPOCHLOROUS ACID
O.OOI
IO IOO
TIME, MIN.
IOOO
6-16
-------
FIGURE 6-6
BREAKPOINT CHLORINATION CONTROL - FUNCTIONAL SCHEMATIC
FEED -
FOREWARD
CONTROL
COMPONENT
AMMONIA -N
CONCENTRATION
(PROCESS INFLUENT)
PROCESS
FLOWRATE
(MULTIPLY)
i
SIGNAL FOR AMMONIA MASS FLUX
PRESELECTED
MULTIPLIER
I
(MULTIPLY)
I
(CORRESPONDS TO CI2:NH4 - N)
SIGNAL FOR COMPUTED
CHLORINE REQUIREMENT
FEED-
BACK
CONTROL
COMPONENT
MEASURED
FREE CHLORINE
RESIDUAL CONCENTRATION
(PROCESS EFFLUENT)
PRESELECTED
FREE CHLORINE
RESIDUAL CONCENTRATION
(SUBTRACT)
..J
ELECTRONIC
FUNCTION
GENERATOR
1
i
1
^-CHLORINATION
PACING
SIGNAL
SIGNAL INDICATING DEPARTURE
FROM DESIRED FREE CHLORINE
RESIDUAL IN PROCESS EFFLUENT
-------
breakpoint. The manually preselected multiplier corresponds to the C^.'NH^-N ratio
required for breakpoint and varies from values of about 8 to 10.
The feedback control loop involves measurement of the free chlorine residual in the process
effluent. The level of measured free chlorine residual is compared to a setpoint value
(usually 2-4 mg/1) using a standard process controller and a signal is generated which
provides a "trimming" of the chlorine dosage to the system.
6.3.1.2 pH Control
Under most circumstances, a base is added to the breakpoint process to neutralize a portion
of the acidity generated in the chlorine addition (Equations 6-1 and 6-9). The base
requirements at a particular installation are related to wastewater alkalinity, individual
treatment processes employed prior to breakpoint chlorination as well as pH or alkalinity
restrictions which might be imposed upon the breakpoint effluent.
pH control in the breakpoint chlorination process may be effectively achieved using a
combination feed foreward and feedback system. The feed foreward component of the
system involves pacing the base addition directly on the chlorine application rate. This
accomplishes neutralization of a preselected portion of the chlorine acidity. A feedback
loop is employed to "trim" the system to a designated pH level. Accurate pH control of the
process promotes efficient utilization of chlorine and can reduce or eliminate undesirable
end products of the reaction.
6.3.2 Process Control Components
Effective control over breakpoint chlorination requires utilization of accurate and reliable
sensory equipment. In this regard two components deserve special attention; namely the
ammonia nitrogen and free chlorine monitoring elements.
The ammonia nitrogen and free chlorine monitoring device which was used during the Blue
Plains pilot testing of breakpoint chlorination was an automated wet chemical analyzer
using colorimetric techniques to measure the ammonia nitrogen and free chlorine
concentrations in small individual samples of the process influent. A continuous-duty free
chlorine residual analyzer was also successfully used at Blue Plains to provide a feedback
signal for breakpoint chlorination control.
6.4 Dechlorination Techniques
Dechlorination is a unit process which eliminates the active chlorine residual in solution
prior to effluent discharge or additional treatment steps. Current emphasis on elimination of
chlorine residual from wastewater treatment effluents has resulted in dechlorination
requirements in many cases. Frequently, dechlorination will be required as a companion
process to breakpoint chlorination. Dechlorination techniques using sulphur dioxide and
activated carbon may be used.
6-18
-------
6.4. 1 Sulphur Dioxide Dechlorination
Sulphur dioxide is a colorless, odorless gas which hydrolyzes in aqueous solution to form
sulfurous acid, a strong reducing agent. When added to a sample containing active chlorine
residual (oxidizing agent), the chlorine residual is converted to a non-toxic form, normally
chloride ion.
6.4.1.1 Stoichiometry
Sulphur dioxide reacts with both free and combined forms of chlorine residuals according to
the expressions noted below:
a. Reaction with free chlorine residual (i.e. HOC1)
SO2 + H2O - ~ HSO~ + H (6-11)
HOC1 + HSO~ - - Cf + SO^ + 2H+ (6-12)
HOC1 + HO Cf + SO4 + 3H+ (6-13)
h. Reaction with combined chlorine residual (i.e. NH2C1)
SO2 + H2O ^ HSO3 + H+ (6-11)
NH2C1 + HSO3 + H2O ^ Cf + SO4 + NH* + H+ (6-14)
S02 + NH2C1 + 2H20 Cf + S04 + NH+ + 2H+ (6-15)
Several important observations can be made from the chemical reactions presented in
Equations 6-11 to 6-15. First, both combined and free chlorine residuals are reduced to
chloride ion following the reaction. The sulfite ion is correspondingly oxidized to.the sulfate
ion. Both chloride and sulfate ions are usually present in abundance in wastewater,. so that
the additional increment added in dechlorination is usually insignificant.
6-19
-------
The predictions of sulphur dioxide dosages required for dechlorination (Equations 6-13 and
6-15) suggest an SO2:Cl2 molar ratio of 1:1. This corresponds to a weight ratio of 0.9:1. In
practice, 0.9 to 1.0 parts of SO2 are required for dechlorination of 1.0 part of chlorine
residual (expressed as <
The acidity generated in dechlorination, while appearing to be significant on a molar basis, is
seldom an important factor in practice. Roughly 2 mg/1 of alkalinity are consumed for each
1.0 mg/1 of sulphur dioxide applied. The chlorine residual remaining in solution following
breakpoint chlorination and disinfection contact is usually low enough (approximate range
of 1 mg/1 to 8 mg/1) so that the alkalinity consumption and pH change resulting from
sulphur dioxide addition may usually be neglected.
6.4.1.2 Reaction Rates
The rate of reaction between SC>2 in solution and chlorine residual is practically
instantaneous. The high specificity of SC>2 in solution for the chlorine residual and the rapid
reaction rate tend to reduce to very low levels competing side reactions which could result
in wastage of the sulphur dioxide.
6.4.1.3 Significance of Sulphur Dioxide Overdose
Sulphur dioxide dechlorination is most often accomplished to eliminate the chlorine induced
toxicity in a wastewater effluent prior to discharge. Studies on primary and secondary
effluents23,24 have shown that effluent toxicity following chlorination-dechlorination, as
measured by fish bioassay, is reduced to a level slightly below that measured on comparable
unchlorinated effluent. These studies indicate that sulphur dioxide dechlorination can
eliminate acute toxicity related to chlorine residual. Tests in an aerated continuous-flow fish
bioassay apparatus at sulphur dioxide overdoses of about 70 mg/1 (as 862) beyond that
necessary for dechlorination showed no adverse effects on the test fish.
Large overdoses of sulphur dioxide should be avoided in plant scale dechlorination systems
because of chemical wastage and the oxygen demand exerted by the excess sulphur dioxide
in solution. Excess sulphur dioxide reacts slowly with dissolved oxygen in solution
according to the following equation:
HSO^ + 0.5O2 —-SO4+H+ (6-16)
The net result of this reaction can be a reduction in wastewater effluent dissolved oxygen
and an increase in the measured effluent COD and BOD levels. Proper control of the
dechlorination system is essential to minimize these adverse effects.
6.4.1.4 Process Application and Control
Chemical dechlorination using sulphur dioxide can be a rather inexpensive process for
6-20
-------
reducing effluent toxicity related to chlorine residual. Much of the process equipment,
chemical handling and chemical addition components of a sulphur dioxide dechlorination
system are identical to those used in a chlorination system. For example, gaseous
chlorinators are used to feed sulphur dioxide, evaporaters may be used practically
interchangeably between chlorine and sulphur dioxide, and piping,valves and gaseous
injectors are identical in the two systems. The speed of the reaction precludes the necessity
for providing a separate contact basin as in chlorine contact for disinfection. The dosage of
sulphur dioxide required is only that amount needed to dechlorinate the effluent following
the chlorine contact period, a level substantially less than the total chlorine dose.
There are no analytical instruments currently on the market which have demonstrated a
capability to reliably and accurately measure the excess sulphur dioxide in solution following
dechlorination. This is why the process control schemes which have been devised do not
have the feedback self-corrective aspects which provide a high degree of consistency to
effluent quality. Control is usually achieved by pacing the dechlorination equipment on
signals of flow and chlorine residual measured directly upstream from the point of sulphur
dioxide application. Operator surveillance and frequent manual adjustments to the system
are needed to maintain a slight overdose of sulphur dioxide, say 0.5 mg/1.
A more complicated control scheme which does incorporate positive control features
involves application of a preselected chlorine dosage to a small side stream of dechlorinated
effluent. If, for example, a preselected chlorine dose of 1 mg/1 is applied and dechlorination
is paced to maintain a chlorine residual of 0.8 mg/1 in the "biased" side stream, the net
effect of the control system would be to provide a controlled overdose of 0.2 mg/1
(expressed as Cty of sulphur dioxide.
With either control scheme, the sulphur dioxide overdose in the final effluent may be
controlled to 0.5 mg/1 (as Cty or less. Given that the reaction rate of Equation 6-16 is very
slow, the effect of the overdose on plant effluent dissolved oxygen concentration would be
minimal, probably less than 0.2 mg/1 DO depletion. In such a case, aeration following
sulphur dioxide dechlorination would seldom be warranted.
6.4.2 Activated Carbon Dechlorination
Dechlorination of wastewater effluent streams using activated carbon can serve several
functions other than removal of chlorine residuals. In the case of breakpoint chlorination,
the activated carbon can effectively catalyze the chemical reactions, serving as a "reaction
bed." Some removal of soluble organics is also accomplished through adsorption.
6.4.2.1 Stoichiometry
Activated carbon (C*) reacts with both free and combined chlorine residuals in the
following manner:
6-21
-------
a. Reaction with free chlorine residual
C* + 2HOC1 - CO2 + 2H+ + 2Cf (6-17)
b. Reactions with combined chlorine residual
C* + 2NH2C1 + 2H2O ^ CO2 + 2NH* + 2C1 (6-18)
C* + 4NHC12 + 2H20 ^ C02 + 2N2 + 8H+ + 8Cf (6-19)
Carbon dioxide is formed in each case following reaction with chlorine residual. Ammonia is
returned to solution following reaction of carbon with monochloramine, but dichloramine
has been observed to decompose to nitrogen gas following contact with activated carbon.
These chemical pathways have been confirmed by several recent studies.25,26
6.4.2.2 Process Application
Studies by Stasiuk et al. 25 showed complete dechlorination of both free and combined
chlorine residuals following carbon contact times of 10 minutes. Studies on the
dechlorination characteristics of granular activated carbon have been conducted27,28,29
which show a variation in the dechlorination capacity of activated carbon as a function of
the hydraulic application rate and particle size. Those data suggest the formation of a
dechlorination intermediary compound, nascent oxygen, on the surface of the carbon
which builds up and causes a gradual loss of dechlorination efficiency. Regeneration of the
carbon was accomplished by heating in the absence of air at 400 C or higher. Smaller
activated carbon particles were found to give enhanced dechlorination capacity. One cubic
foot of activated carbon was found to dechlorinate 0.55 million gallons at 2 gpm/sq ft
hydraulic application rate and 5 mg/1 free chlorine residual. Approximately 3.6 million
gallons of the same water could be dechlorinated by the activated carbon when applied at a
rate of 1 gpm/sq ft.
6.5 Design Example
As an example, consider a 10 mgd conventional activated sludge plant that must be
upgraded to meet an effluent nitrogen limitation of 2 mg/1. Present effluent quality is 1 mg/1
organic nitrogen, 20 mg/1 ammonia nitrogen, 15 mg/1 of suspended solids, BOD5 of 25 mg/1
and pH of 7.0. The peak to average nitrogen load ratio is 1.9 The breakpoint chlorination
design should provide capacity for removal of all ammonia nitrogen in the plant effluent.
6-22
-------
1 . Calculate the average ammonia nitrogen oxidized daily.
= 8.33-Q(NQ-N1) (4-24)
where: NT = ammonia nitrogen oxidized, Ib per day
Q = average daily flow, mgd
N = influent ammonia nitrogen, mg/1
N, = effluent ammonia nitrogen, mg/1
For this example, the result is:
NT = 8.33(10)(20-0) = 1,666 Ib/day
2. To calculate the average daily chlorine consumption, a Cl2:NH4-N of approxi-
mately 9:1 would be appropriate. The average daily chlorine consumption would
be:
Average daily O, = (9)(1 ,666) = 14,994 Ib/day
3. The peak rate of chlorine utilization, that rate used to size the chlorine feed
system, may be computed by multiplying the peak to average nitrogen load ratio
by the average daily chlorine consumption as follows:
Peak C12 = (1.9)(14,994) = 28,489 Ib/day
4. To calculate the average daily quantity of alkalinity supplementation, assume that
sodium hydroxide is to be used. Since the effluent pH is at pH 7 prior to
breakpoint chlorination,and the breakpoint process itself should be conducted at
pH 7, all acidity generated in breakpoint chlorination must be neutralized by the
sodium hydroxide added. Since 1.50 Ib. of NaOH are required per Ib of C\2
added in breakpoint, the average daily NaOH added per day would be:
Average daily NaOH = (1.50)(14,994) = 22,491 Ib/day
5. The TDS increment added to the plant effluent as a result of breakpoint
chlorination can be computed from data of Table 6-3. For breakpoint
chlorination with chlorine gas and neutralization of all acidity with sodium
6-23
-------
hydroxide, the TDS increase per mg/1 of ammonia nitrogen consumed is 14.8
mg/1. The total TDS increase in this example would be:
TDS increase = (14.8)(20) = 296 mg/1.
6. To compute the average daily consumption of sulphur dioxide needed to
completely dechlorinate the plant effluent prior to discharge, assume that the
average chlorine residual following breakpoint chlorination and disinfection is 5
mg/1. The average daily sulphur dioxide consumption would be:
Average daily SO2 = (8.33)(10)(5)(1.0) = 416 Ib/day.
6.6 Considerations for Process Selection
The breakpoint chlorination process offers a number of advantages which should be
considered when evaluating alternative nitrogen removal processes:
1. Ammonia nitrogen removal may be accomplished in a one-step process to
concentrations less than 0.1 mg/1 (as N). The major end product of the breakpoint
reaction is nitrogen gas which is evolved to the atmosphere.
2. Breakpoint chlorination is free from the toxic upsets and acclimation periods
which can affect biological nitrification and denitrification processes. Ammonia
nitrogen may be successfully removed from solution regardless of upstream
treatment processes. Breakpoint chlorination is also rather insensitive to changes
in process temperature.
3. The breakpoint process may be operated intermittently or in split-stream
arrangements as needed to meet individual receiving water nitrogen limitations.
The process can be used as a polishing step for other ammonia removal processes
such as nitrification to provide low ammonia nitrogen concentrations.
4. Breakpoint chlorination is reliable and consistent in terms of process perfor-
mance.
5. Disinfection of a wastewater effluent is enhanced following breakpoint chlorina-
tion due to the presence of free available chlorine residuals (HOC1 and OC1~).
6. The low spacial requirement of the breakpoint process makes it particularly
suitable for certain applications, including addition to an existing facility, where
nitrogen removal is required but where space constraints exist.
6-24
-------
7. The cost of physical facilities for breakpoint chlorination is much less than for
biological nitrification-denitrification facilities.
A list of potential disadvantages of the breakpoint chlorination process includes:
1. The breakpoint chlorination process is usually quite high in operating costs.
2. The addition of chorine, pH adjustment chemicals and sulphur dioxide all
contribute to the level of total dissolved solids in the wastewater effluent. Some
treatment plants have TDS limitations which would limit the applicability of
breakpoint chlorination.
3. Dechlorination will be required in many cases to remove the potentially toxic
chlorine residual. It should be noted, however, that this is a disadvantage of any
chlorination process.
6.7 References
1. Pressley, T.A., Bishop, D.F., and S.G. Roan, Nitrogen Removal By Breakpoint
Chlorination. Report prepared for the Environmental Protection Agency, September,
1970.
2. Pressley, T.A., Bishop, D.F., Pinto, A.P., and A.F. Cassel, Ammonia-Nitrogen Removal
by Breakpoint Chlorination. Report prepared for the Environmental Protection
Agency, Contract Number 14-12-818, February, 1973.
3. Pressley, T.A., et al, Ammonia Removal by Breakpoint Chlorination. Environmental
Science and Technology, 6, No. 7, 662, July, 1972.
4. White, G.C., Handbook of Chlorination. New York, Van Nostrand Reinhold Company,
1972.
5. Morris, J.C., and I. Weil, Chlorine-Ammonia Breakpoint Reactions: Model Mechanism
and Computer Simulation. Paper, annual meeting Am. Chem. Soc., Minneapolis, Minn.,
April 15, 1969.
6. Morris, J.C., Weil, I., and R.H. Culver, Kinetic Studies on the Break-Point with
Ammonia and Glycine. Unpublished copy from senior author, Harvard Univ. (1952).
7. Morris, J.C., Kinetic Reactions between Aqueous Chlorine and Nitrogen Compounds.
Fourth Rudolphs Res. Conf., Rutgers Univ. (June 15-18, 1965).
8. Griffin, A.E., and N.S. Chamberlin, Some Chemical Aspects of Breakpoint Chlori-
nation. J. NEWWA, 55, pp 371, 1941.
6-25
-------
9. Weber, W.J., Jr., Physio chemical Processes for Water Quality Control. New York,
Wiley-Interscience, 1972.
10. Stone, R.W., Parker, D.S., and J.A. Cotteral, Upgrading Lagoon Effluent to Meet Best
Practicable Treatment. Presented at the 47th Annual Conference of the Water
Pollution Control Federation, Denver, Colorado, October, 1974.
11. Brown and Caldwell, Report on Tertiary Treatment Pilot Plant Studies. Prepared for
the City of Sunnyvale, California, February, 1975.
12. Stearns and Wheeler, Waste-water Facilities Report. Cortland, New York, May, 1973.
13. Stenquist, R.J., and WJ. Kaufman, Initial Mixing in Coagulation Processes. Prepared
for EPA, Project EPA-R2-72-053, November, 1972.
14. Schuk, W., personal communication with D.S. Parker. EPA-DC Blue Plains Pilot Plant,
September 9, 1974.
15. Lawrence, A.W., et al, Ammonia Nitrogen Removal from Wastewater Effluents by
Chlorination. Presented at the 4th Mid-Atlantic Industrial Waste Conference,
University of Delaware, November, 1970.
16. Taras, M.J., Effect of Free Residual Chlorine on Nitrogen Compounds in Water.
JAWWA,45,47(1953).
17. Griffin, A.E., and N.S. Chamberlin, Some Chemical Aspects of Break-Point
Chlorination. J. NEWWA, 55, pp 371 (1941).
18. Griffin, A.E., Chlorine for Ammonia Removal. Fifth Annual Water Conf. Proc. Engrs.
Soc. Western Penn., pp 27 (1944).
19. William, D.B., private communication with G.C. White, Brantford, Ontario, Canada
(1967).
20. Clarke, N.A., et al, Human Enteric Viruses in Water: Source, Survival and
Removability. International Conf. Water Poll. Research, Pergamon Press, London
(September, 1962).
21. van Vuuren, L.R.J., et al, Slander Water Reclamation Plant: Chlorination Unit
Process. Project Report 21, Pretoria, So. Africa, November, 1972.
22. Pollution: Engineering and Scientific Solutions. E.S. Barrekette Ed., New York,
Plenum Publishing Corp., pp 522-547.
6-26
-------
23. Esvelt, L.A., Kaufman, W.J., and R.E. Selleck, Toxicity Removal from Municipal
Wastewater. University of California, Sanitary Engineering Research Laboratory
Report 71-7, 1971.
24. Stone, R.W., Kaufman, W.J., and A.J. Home, Long-Term Effects of Toxicants and
Biostimulants on the Waters of Central San Francisco Bay. University of California,
Sanitary Engineering Research Laboratory Report 73-1, May, 1973.
25. Stasuik, W.N., Hetling, L.J., and W.W. Shuster, Removal of Ammonia Nitrogen by
Breakpoint Chlorination Using an Activated Carbon Catalyst. New York State
Department of Environmental Conservation Technical Paper No. 26, April, 1973.
27. Hagar, D.G., and M.E. Flentje, Removal of Organic Contaminants by Granular Carbon
Filtration. JAWWA, 57, 1440 (November, 1965).
28. Kovach, J.L., Activated Carbon Dechlorination. Industrial Water Engineering, pp
30-32, October/November, 1973.
29. Calgon Corporation, Bulletin 20-44.
6-27
-------
CHAPTER 7
SELECTIVE ION EXCHANGE FOR AMMONIUM REMOVAL
7.1 Chemistry .and Engineering Principles
The basic concepts of the ion exchange process apply to its use for ammonium removal and
these concepts are discussed in the following paragraphs.
7.1.1 Basic Concept
The use of conventional ion exchange resin for removal of nitrogenous material from
wastewaters has not proven attractive because of the preference of these exchangers for ions
other than ammonium or nitrate ions. In addition, the regeneration of conventional ion
exchange resins results in regenerant wastes which are difficult to handle. The non-selective
nature of conventional resins is unfortunate because as a unit process, ion exchange is easily
controlled to achieve almost any desired product quality. The efficiency of the process is
not significantly impaired at temperatures usually encountered and ion exchange equipment
can be automatically controlled, requiring only occasional monitoring, inspection, and
maintenance.
The above limitations of conventional resins may be largely overcome by using an exchanger
which is selective for ammonium. The exchanger currently favored for this use is
clinoptilolite, a zeolite, which occurs naturally in several extensive deposits in the western
United States. It is selective for ammonium relative to calcium, magnesium and sodium. The
removal of the ammonium from the spent regenerant permits regenerant reuse. The
ammonium may be removed from the regenerant and released to the atmosphere as
ammonia (in certain situations) or nitrogen gas or it may be recovered as an ammonium
solution for use as a fertilizer. Figure 7-1 is a simplified schematic of the process.
The wastewater is passed downward through a bed of clinoptilolite (typically 4-5 ft or 1.2
to 1.5 m of 20 x 50 mesh particles) during the normal service cycle. When the effluent
ammonium concentration increases to an objectionable level, the clinoptilolite is regen-
erated by passing a concentrated salt solution through the exchange bed. By removing the
ammonia from the spent regenerant, the regenerant may be reused, eliminating the difficult
problem of brine disposal associated with conventional exchange resins. Some of the
regenerant recovery techniques, as discussed later in detail, remove the ammonia as nitrogen
gas which is discharged to the atmosphere while others remove and recover the ammonia in
solution form for potential use as a fertilizer.
7-1
-------
FIGURE 7-1
SELECTIVE ION EXCHANGE PROCESS
INFLUENT
SPENT REGENERANT
"I
cLiNOPTiLOLiTE
BED
REGENERANT
RECOVERY
AMMONIA OR
NITROGEN GAS
'OR RECOVERED
AMMONIA
SOLUTION
________
FRESH REGENERANT
EFFLUENT
7.1.2 Ion Exchange Principles
Ion exchange is a process in which ions held by electrostatic forces to charged functional
groups on the surface of a solid are exchanged for ions of similar charge in a solution in
which the solid is immersed. It is a stoichiometric, reversible exchange of ions between a
liquid and solid which produces no significant changes in the structure of the solid. The
mass action equilibria expression provides a useful model for ion exchange behavior. In a
binary system, the reaction,
bA+a + aBZ,
bAZ +aB
a
+b
(7-1)
expresses the reversible equilibrium where a and b are the valences of ions A and B,
respectively, and Z is the exchange site in the solid. * This reaction may be expressed as the
equilibrium constant,
(a)
K =
AZ
(a)
B
(a)
(a):
(7-2)
B
Zb
7-2
-------
in which (a)^, (a)AZa> etc- are the activities of the various spc-cies. Because of the difficulty
in measuring activities, especially in the solid phase, it is convenient to use concentrations
uncorrected for activity. In doing so, the equilibrium constant in Equation 7-2 varies with
concentration and has been termed the "selectivity coefficient,"
where q is the solid phase ionic concentration in milliequivalents per gram (meq/g) and c is
the solution phase concentration in meq/1. Alternatively, the selectivity coefficient can be
expressed in terms of dimensionless concentration,
a-b / \b / \a
=(4(f)B
These variables are expressed in terms of the total solution concentration, Co, in meq/1 and
the total exchange capacity, Q, in meq/g. Thus, x = c/Co and y = q/Q.
The preference of an ion exchange for one ion relative to another in binary systems is often
expressed as the "separation factor,"
ocA = /-9-U-} =(!-} (+-} (7'5)
B KH ^'AH
Because the numerical value of the separation factor is not affected by the choice of
concentration units, equilibrium data are often expressed in this way. If the equivalent
fraction of ion A in the solid phase, y^, is plotted against the equivalent fraction of A in the
solution, XA, three cases can be identified corresponding to oc<.l, ot = 1, and ex. > 1 as
shown in Figure 7-2. Isotherms which are concave upward, l, are referred to as "favorable" isotherms since the solid prefers ion A to ion
B. Ion exchange operations almost always are concerned. with systems in which the ion of
concern has a separation factor greater than unity during the service cycle.
The basic principles of ion exchange can be used to determine the capacity of clinoptilolite
for ammonium and an excellent example is available in Reference 2. However, such
calculations are lengthy and rather complex and a later section (7.2.9) presents a simplified
technique for quickly approximating the ammonium capacity of clinoptilolite for varying
concentrations of competing cations.
7.1.3 Properties of Clinoptilolite
7.1.3.1 Selectivity
Isotherms demonstrating the selectivity of clinoptilolite for ammonium over other cations
have been reported in the literature. 3 An example is the comparison of Hector
7-3
-------
FIGURE 7-2
GENERALIZED ION EXCHANGE ISOTHERMS
0.0 O.I 0.2 0.3 0.4 0.5 0.6 O.7 0,8 0.9 t.O
EQUIVALENT ION FRACTION IN SOLUTION PHASE, XA
Clinoptilolite and a strong acid polystyrene resin, IR 120, shown in Figure 7-3. The total
equilibrium solution normality was constant at O.IN. The terms (Ca)z, (NH4)Z = equivalent
fraction of calcium or ammonium on the zeolite. The terms (Ca)^, (NH|>N = normality of
calcium or ammonium in the equilibrium solution. The ammonium capacity of IR 120 was
4.29 meq/g of air-dried resin. A comparison of the generalized isotherm presented in Figure
7-2 with Figure 7-3 clearly shows that the IR 120 prefers calcium to ammonium ions.
Clinoptilolite, on the other hand, prefers ammonium to calcium ion and is of greater utility
for ammonium ion removal from wastewaters containing calcium.
4
The ion exchange equilibria for the systems NHj-Na+, NH4-K+, NH4-Ca+2 and NH|-Mg+2,
with Clinoptilolite and other zeolites, are also available in the literature.2 Plots of the NH^
selectivity coefficients vs. the solution concentration ratios of the cations are shown in
Figures 7-4 and 7-5. .Using these data in conjunction with published calculation
techniques, 2 it is possible to accurately predict the ammonium capacity of Clinoptilolite in
the presence of various concentrations of other cations.
7-4
-------
FIGURE 7-3
THE 23 C ISOTHERMS FOR THE REACTION, (Ca)z + 2(NH*)N = 2(NH^)z + (Ca)N
WITH HECTOR CLINOPTILOLITE AND IR 120
1,0
Hector
Clinoptilolite
0 0.2
EQUIVALENT
0.4 0.6
ION FRACTION IN
0.8 1.0
SOLUTION PHASE
Section 7.2.9 presents a curve useful in approximating the ammonium exchange capacity
which can be employed in sizing the ion exchange beds. The equilibrium isotherms for
ammonia and other cations which are present as macrocomponents in wastewaters are
shown in Figure 7-6J These isotherms illustrate that clinoptilolite is selective for
ammonium relative to all of the listed ions except potassium.
7.1.3.2 Mineralogical Classification
The zeolites are classified as a family in the silicate group. They are hydrated
alumino-silicates of univalent and bivalent bases which can be reversibly dehydrated to varying
7-5
-------
FIGURE 7-4
SELECTIVITY COEFFICIENTS VS. CONCENTRATION RATIOS OF SODIUM
OR POTASSIUM AND AMMONIUM IN THE EQUILIBRIUM SOLUTION WITH
HECTOR CLINOPTILOLITE AT 23 C FOR THE REACTION
-*•<*•
UJ
u,
Lu
o
o
O
UJ
-J
UJ
CO
(Y)
(Y)
N
IOOO
100
10
O.I
I Hill I I I ll Illl I 1 I I I I II I I I I Illll
O.OI
O.I
10
IOO
IOOO
CONCENTRATION RATIO,
degrees without undergoing a change in crystal structure and are capable of undergoing
cation exchange. The general composition of zeolites is given by the formula
(M,N2)O-Al2O3-nSiO2'mH2O where M and N are, respectively, the alkali metal and
alkaline earth counter ions present in the zeolite cavities.
Clinoptilolite is a common material found in bentonite deposits in the western United
States. The largest known deposit of clinoptilolite in the United States is found in southern
California within a deposit of bentonite called Hectorite because of its proximity to Hector,
California. The U.S. deposits are predominantly in the sodium form. Although a widely
occurring material, not all deposits produce a clinoptilolite of adequate structural strength
to withstand the handling which occurs in a columnar operation.
7-6
-------
FIGURE 7-5
SELECTIVITY COEFFICIENTS VS. CONCENTRATION RATIOS OF CALCIUM
OR MAGNESIUM AND AMMONIUM IN THE EQUILIBRIUM SOLUTION WITH
HECTOR CLINOPTILOLITE AT 23 C FOR THE REACTION
(X)z + 2(NH+)N = 2(NH+) + (X)
xl
u.
u.
iu
o
o
I-
o
Lu
10''
10'
10'
10
I . ill...I I I . I....I I I lill.J I 1 llll.j I I II.1..I I I .(...I II ll.l.J I I lllll.i
/ 10 10'
I05 IO6 I07 I08
CONCENTRATION RATIO,
(X)N
7.1.3.3 Total Exchange Capacity
Although the total ion exchange capacity of a material is by no means a complete
description of its ion exchange properties, it is an indication of the applicability of the
substance for process use. For example, New Jersey greensand, which was widely used in
water softening before the development of organic exchangers, has a total exchange capacity
of 0.17 meq/g. 1 In comparison, the exchange capacities of strong acid cation exchangers are
usually 4 to 5 meq/g. The total exchange capacity of clinoptilolite as measured by several
different investigators ranges from 1.6 to 2.0 meq/g and is slightly lower than the average
for zeolites. With typical cation concentrations encountered in municipal wastewaters, the
capacity for NH$ is typically 0.4 meq/g (see Section 7.2.9).
7.1.3.4 Chemical Stability
The instability of natural clays and zeolites toward acids and alkalis is known as these
materials are widely used in water softening. However, clinoptilolite is considerably more
7-7
-------
FIGURE 7-6
ISOTHERMS FOR EXCHANGE OF NH* FOR K+, Na+, Ca**, AND Mg"1
ON CLINOPTILOLITE
' I I
_ Magnesium
Potassium
Total Solution
Concentration 0.1 N-,
Temperature 23 C
EQUIVALENT FRACTION OF NH$~N IN SOLUTION PHASE,
*NH+-N
acid resistant than other zeolites. ^ Very high strength (20% NaOH) caustic solutions
produce significant chemical attack on the clinoptilolite. However, at the lower solution
strengths encountered in systems which use a caustic regenerant, physical attrition is more
significant than chemical attack. This will be discussed later in this section.
7.1.3.5 Physical Stability
When crushed, sieved, and thoroughly washed with agitation to remove fines, clay, and
other impurities, 20 x 50 mesh Hector clinoptilolite gives a wet attrition test of 3 percent. 4
7-8
-------
The wet attrition test determines the amount of fines (less than 100 mesh) generated by 25
grams of the granular zeolite during rapid mixing with 75 milliliters of water on a paint
shaker for 5 minutes. Commercial zeolites, such as erionite and chabazite, which are
powdered, mixed with clay binder, extruded, and fined, will generally give a wet attrition
test of about 6 percent or twice that of the Hector clinoptilolite. Low wet attrition is
important to minimize losses of clinoptilolite in an ion exchange column operation.
7.1.3.6 Density
Clinoptilolite (20 x 50 mesh) has been reported to have a wet particle specific gravity of
1.5 9 and a bulk density of 0.74-0.79 g/cc1.
7.2 Major Service Cycle Variables
The factors which have a major effect on process efficiency include: pH, hydraulic loading
rate, clinoptilolite size, pretreatment, wastewater composition, and bed depth.
7.2. IpH.
Within an influent pH range of 4 to 8, optimum ammonium exchange occurs. 1 As the pH
drops below this range, hydrogen ions begin to compete with ammonium for the available
exchange capacity. As the pH values increase above 8, a shift in the NH3-NH4 equilibrium
towards NH3 begins. Operation outside of the pH range of 4 to 8 results in a rapid decrease
of exchange capacity and increased ammonium leakage.
7.2.2 Hydraulic Loading Rate
Variations in column loading rates within the range of 7.5-20 Bed Volumes (BV)/hour (7.5
BV/hr is equivalent to 0.95 gpm/cu ft or 2.15 l/m3/sec) have been shown to produce no
significant effects on the ammonium removal efficiency of 20 x 50 mesh clinoptilolite.3
Ammonium concentrations in the clinoptilolite effluent of 0.22-0.26 mg/1 were produced
throughout the above range in one set of tests. 1 When rates exceed 20 BV/hour, the
exchange kinetics suffer as demonstrated by a significant leakage of ammonium early in the
loading cycle. The effects of loading rate as a function of clinoptilolite size are discussed in
the next section.
7.2.3 Clinoptilolite Size
Mine-run clinoptilolite is typically 1-2 inch (25 to 51 mm) chunks which must be ground
and screened to the size desired for column operation. As would be expected, the smaller
the clinoptilolite size, the better the kinetics of the exchange reaction. This effect is
illustrated by data that show that 20 x 50 mesh clinoptilolite kinetics begin to suffer (see
Section 7.2.2) at rates of 20-30 BV/hour while 50 x 80 mesh kinetics do not suffer until
rates of 40 BV/hour are reached. However, the improved rate of exchange is accompanied
7-9
-------
by the disadvantage of higher head loss. It appears that 20 x 50 mesh clinoptilolite (about
the size of typical filter sand) offers an adequate compromise between acceptable headless
and exchange kinetics. At a loading of 15 BV/hour in a 3 ft (0.9m) deep bed (5.6 gpm/sq ft
or 3.8 l/m^/sec), the headloss is 2.1 ft (0.64m) with 20 x 50 mesh clinoptilolite. Lower
headlosses could be obtained by lower rates. Use of deeper beds would result in greater
headlosSji.e., 6 ft (1.8 m) depth would have a headloss of 4.2 ft (1.28 m) at 5.6 gpm/sq ft
(3.8 l/m^/sec). These headloss values do not include losses in inlet and outlet piping or in
the underdrain system.
7.2.4 Pre treatment
To avoid excessive headloss, the clinoptilolite influent must be relatively free of suspended
solids — preferably less than 35 mg/1. Tests with clarified and filtered raw wastewater
indicate no problems with organic fouling. Biological growths which occurred were
adequately removed in the regeneration cycle.2 Additional data on pretreatment effects are
presented in the next section.
7.2.5 Wastewater Composition
As noted earlier, although clinoptilolite prefers ammonium ions to other cations, it is not
absolutely selective and other cations do compete for the available exchange capacity. Pilot
tests conducted at several locations illustrate the effects of wastewater composition on the
useful capacity of the clinoptilolite.^ Tests at three locales that span a wide range of
wastewater compositions are shown in Table 7-1.
The equilibrium Nffy-N bed loading computed for each of the wastewaters listed in Table
7-1 was 4.1 g/1, 3.9 g/1, and 4.3 g/1, respectively, forTahoe, Pomona, and Blue Plains. Figure
7-7 presents equilibrium bed loading in an alternate way. The minimum bed volumes
required to attain equilibrium Nlfy-N loading are expressed as a function of the Nlfy-N
concentration in the influent wastewater. The equilibrium bed volume values given in Figure
7-7 normally represent the 50 percent breakthrough point (the effluent concentration is 50
percent of the feed concentration). The water lowest in competing cations (Blue Plains) had
the greatest ammonium removal capacity. In the 10-20 mg/1 influent Nlfy-N range, the
lower competing ion concentrations at Blue Plains resulted in the useful ammonium
exchange capacity being about 33 percent greater than that for Pomona with its higher TDS
water. The lower degree of pretreatment at Blue Plains (i.e., no biological pretreatment) did
not impair the effectiveness of the clinoptilolite for ammonium removal.
7.2.6 Length of Service Cycle
To illustrate the determination of a permissable length service cycle for a given wastewater,
ammonia breakthrough curves for a single 6 ft deep (1.8m) bed of clinoptilolite are
illustrated in Figure 7-8 for Tahoe tertiary effluent with flow rates varying from 6.5 to 9.7
7-10
-------
TABLE 7-1
INFLUENT COMPOSITION FOR SELECTIVE ION EXCHANGE
PILOT TESTS AT DIFFERENT LOCALES
Parameter, mg/1
NH*-N
Na
K
Mg
Ca
PH
Range
COD
IDS
Activated sludge plant effluent
Tahoe
Carbon treated
15
44
10
1
51
7-8b
11
325
a
Pomona
16
120
18
20
43
6.5-8.2b
10
700
Clarified raw wastewater
Blue Plains
(Washington, D
12
35
9
0.2
30
7-9b
50
250
.C.)
Approximately half of the runs at Pomona were made with carbon treated
secondary effluent and the others with alum coagulated secondary effluent.
pH units
BV/hour (bed volumes per hour) with 15 to 17 mg/1 NH^-N in the feed stream. These curves
indicate a throughput value of 150 bed volumes should be used for this wastewater for
design for effluents requiring a high degree of ammonia removal. Although the effluent
concentration had reached 2-3 mg/1 at 150 BV, the average concentration produced to this
point in the cycle was less than 1 mg/1. Breakpoint chlorination would be more economical
for removing the 1 mg/1 residual, if required, than would provision of greater ion exchange
column capacity. The average ammonium concentration for a breakthrough curve is
obtained by integrating the area under the breakthrough curve and dividing by the total
flow. For example, integrating the area under the curve for 8.1 BV/hr in Figure 7-8 indicates
an average NH^-N concentration of 0.67 mg/1 for 150 BV.
7.2.7 Bed Depth
The effect of bed depth on ammonia breakthrough at 9.7 BV/hr is illustrated in Figure 7-9.
In general, the 3 ft (0.9 m) bed of clinoptilolite was not as effective for ammonium removal
as the 6 ft (1.8 m) bed at the same bed volume rate. The shallow bed has a lower flow
velocity because a 9.7 BV/hr flow in a 3 ft (0.9 m) deep bed corresponds to 3.6 gpm/sq ft
(2.5 l/m2/sec) while in a 6 ft (1.8 m) deep bed it corresponds to 7.2 gpm/sq ft (4.9
7-11
-------
FIGURE 7-7
MINIMUM BED VOLUMES AS A FUNCTION OF INFLUENT
NH*-N CONCENTRATION TO REACH 50 PERCENT
BREAKTHROUGH OF AMMONIUM (REFERENCE 2)
401
'a:
LU
Cj
Uj
-J
u.
30
20
10
D TAHOE
X POMONA
O BLUE PLAINS
100 ZOO 300
MINIMUM BED VOLUMES
400
500
l/m^/sec). The lower velocity might increase the likelihood of plugging of portions of the
bed. Plugging would cause poor flow distribution and lower bed efficiency. As discussed in
the design examples in Chapter 9, full-scale designs are using bed depths of 4-ft (1.2 m) with
a high degree of pretreatment (coagulation and filtration) which will minimize plugging of
the clinoptilolite bed.
7.2.8 One Column vs. Series Column Operation
Operation to the 150 bed volume throughput value (Figure 7-9) to maintain an average
NH4—N concentration at or below 1 mg/1 uses only 55 to 58 percent of the zeolite's
equilibrium capacity, The number of bed volumes throughput per bed can be increased
while maintaining low NH4—N effluent concentrations with semi-countercurrent operation,
using two beds in series. Semi-countercurrent operation is achieved by first operating the
7-12
-------
FIGURE 7-8
AMMONIUM BREAKTHROUGH CURVES FOR A 6 FT CLINOPTILOLITE
BED AT VARIOUS FLOW RATES
¥
Ul
-J
u.
u.
10
Q
UJ
00
OPERATING CONDITIONS:
ZEOLITE GRAIN SIZE: 20x50 MESH
BED VOLUME: 50 FT3 (l.4l'6<)
FEED: TAHOE TERTIARY EFFLUENT
Avg. Influent
- Symbol Flow Rate NH^-N.mg/J
_L
20 4O 60 80 IOO
BED VOLUMES
120
140
160
180
columns in a 1-2 sequence and then placing column 2 into the lead position, after the first
regeneration, with column 1 becoming the polishing column. A column is removed from the
influent end when it becomes loaded while simultaneously adding a regenerated column, at
the effluent end. This procedure in effect moves the beds counter-current to liquid flow by
continually shifting the more saturated beds closer to the higher influent concentrations.
Beds can be loaded nearer to capacity with this procedure than with single column or
parallel feed multi-column operation. The most highly loaded column is always at the
influent end backed up by one (if two in series) or more columns having decreasing loadings
and NH4-N concentrations at locations progressively nearer the end of the series. Removal
of a column is not decided by applying a breakthrough criterion to the column's own effluent
but by breakthrough at the end of the series. Tests have indicated that the ammonium
loadings could be increased from 55-58 percent of the equilibrium capacity to 85 percent by
using two columns in series. 2 Average throughputs for the Tahoe example discussed earlier
increased from 150 to 250 BV/cycle. However, such a two column operation requires three
columns (two on stream while the third is being regenerated) and more complicated valving
and piping than a parallel column operation. Because of the added capital costs involved in a
7-13
-------
FIGURE 7-9
EFFECT OF BED DEPTH ON AMMONIUM BREAKTHROUGH AT 9.7 BV/HR
6 i T 1 1 1 1 1 1 1
Oi
6
i»
f
K-
Uj
3
-J
U.
U.
UJ
Q
UJ
OJ
OPERATING CONDITIONS:
ZEOLITE GRAIN SIZE: 20x50 MESH
BED VOLUMES: 3 FT DEPTH = 25 FT3, 6 FT DEPTH = 50 FT3
(0.9 m =708 «) (l.8m = 14164)
AVG INFLUENT NH^-IM: 17 mg/l
LOCATION: TAHOE
RATE: 9.7 BV/Hr
6 FT DEPTH
_L
20 4O 60 8O IOO
BED VOLUMES
I2O
140
I6O
ISO
series system, all of the full-scale systems currently under design or in operation utilize
parallel single beds. By blending the effluents from several parallel columns, each of which is
in a different stage of exhaustion, improved utilization of the available exchange capacity is
also achieved. That is, if equal volumes of effluent containing 2 mg/l NH4—N from one
column are blended with effluent containing 0.6 mg/l from another, some added throughput
through the more heavily loaded column could be achieved while still meeting an overall
standard of no more than 2 mg/l NH4-N.
7.2.9 Determination of Ion Exchanger Size
In order to calculate the size of the ion exchange unit needed, the ammonium capacity of
the clinoptilolite must be determined from the characteristics of the influent water. The
ammonium capacity of clinoptilolite can be estimated from Figure 7-10 if the cationic
strength of the wastewater is known. The data used to plot Figure 7-10 were determined in
several experimental runs where the influent ammonium nitrogen concentration was
16.4-19.0 mg/jl. Although the curve is empirical and is a simplification of the complex
effect of competing cation concentrations on ammonium capacity, it illustrates this effect
and serves as a useful tool in sizing the exchange bed.
7-14
-------
FIGURE 7-10
VARIATION OF AMMONIUM EXCHANGE CAPACITY WITH
COMPETING CATION CONCENTRATION FOR A
3 FT DEEP CLINOPTILOLITE BED (REFERENCE 1)
»*
Ul
o
o
o
o
UJ
V)
<*
a:
Q.
O
CO
0.7
0.6
- O.5
O.4
0.3
O.2
0.1
0.0
\
I
Total Ammonium Exchange
Capacity
•Effective Ammonium Exchange Capacity-
(to 1 mg/£ NHj-N in effluent)
I
0.015
0.02
m,zf,
0.005 0.01
CATIONIC STRENGTH, 1/
Where-, m = concentration of the cation species i
z = valence of the cation species i
7-15
-------
Assuming that the influent water has a cationic strength of 0.006 moles/1, the breakthrough
ammonium capacity of the clinoptilolite will be approximately 0.25 meq/g for a 3-ft (0.9
m) bed; the capacity to saturation will be approximately 0.44 meq/g. A greater effective
ammonium capacity can be realized by increasing the depth of the zeolite bed. The use of a
6-ft (1.8 m) bed would result in greater ammonia capacity per unit of exchanger and while
requiring a deeper structure, the additional cost would be nominal. Assuming that 3 ft (0.9
m) of the zeolite bed will have an ammonium exchange capacity equal to 0.25 meq/g and
that the remaining 3 ft (0.9 m) will have a capacity equal to 90 percent of the saturation
capacity or 0.40 meq/g, the 6 ft (1.8 m) bed will have an effective capacity of 0.32 meq/g
[equivalent to 6.6 eq/cu ft (236 eq/m3) and 5.1 kgr/cu ft (182 kgr/m3)].
The zeolite volume required to treat a 10 mgd (0.44 m3/sec) waste flow at 15 BV/hr(1.9
gpm/cu ft or 4.3 1/sec/m3) is 3650 cu ft (102 m3). Assuming complete removal of
ammonium, the throughput to ammonium breakthrough would then be 165 BV with a run
length of 11 hr. Allowing 2 hr down time per cycle for regeneration and rinsing, the zeolite
volume would be increased proportionately to 4300 cu ft (120.4m3) to accommodate the
total design flow. Using four units, each having the dimensions 12 ft x 15 ft x 6 ft deep
(3.66 m x 4.6 m x 1.8 m), the total zeolite volume would be 4320 cu ft (121 m3).1
7.3 Regeneration Alternatives
The key to the applicability of this process is the method of handling the spent regenerant.
The following paragraphs discuss available alternates.
7.3.1 Basic Concepts
As noted earlier, after about 150-200 bed volumes of normal strength municipal waste have
passed through the bed, the capacity of the clinoptilolite has been used to the point that
ammonium begins to leak through the bed. At this point, the clinoptilolite must be
regenerated so that its capacity to remove ammonium is restored. The resin is regenerated
by passing concentrated salt solutions through the exchange bed when the ammonium
concentration in the solid phase has reached the maximum desirable level. The ammonium-
laden spent regenerant volume is about 2.5 to 5 percent of the throughput treated prior to
regeneration. By removing the ammonium from the spent regenerant, the regenerant may be
reused. The alternative approaches available for regenerant recovery are:
• air stripping
• steam stripping
• electrolytic treatment
These alternatives for regenerant recovery will be discussed following a discussion of the
regeneration process.
7-16
-------
7.3.2 Regeneration Process
The ammonium retained on the clinoptilolite exchange sites may be eluted by either sodium
or calcium ions contained in a regenerant solution. While the normal service cycle is
downflow, regeneration is carried out by passing the regenerant up through the clinoptilolite
bed.
7.3.2.1 High pH Regeneration
The approach originally studied for wastewater applications was to use a lime slurry (5 gm/1)
as the regenerant so that the ammonium stripped from the bed during regeneration would
be converted to gaseous ammonia which could then be removed from the regenerant by air
stripping. 3 It was found that elution with lime could be speeded up by the addition of
sufficient NaCl to render the regenerant 0.1./V with respect to NaCl. 3
In addition to converting the ammonium ion to ammonia so it can readily be removed from
the regenerant, the volume of regenerant required for complete regeneration has been found
to decrease with increasing regenerant pH. 1 However, high pH regeneration was found to be
accompanied by an operational problem of major significance.^ Precipitation of
magnesium hydroxide and calcium carbonate occurs within the exchanger during the
regeneration cycle. This leads to plugging of the exchanger inlets and outlets, as well as
coating of the clinoptilolite particles. Violent backwashing of the clinoptilolite was found to
be necessary to remove these precipitants from the clinoptilolite particles, which resulted in
increased mechanical attrition of the clinoptilolite. Chemical attrition also increases at
elevated pH values. 1
Substantial data have been collected on high pH regeneration and are available in references
1, 2 and 3 if this approach is considered. However, the practical problems of scale control
are major limitations which can be overcome by using neutral regenerants. The use of closed
loop regenerant recovery processes negates the disadvantage of higher regenerant volumes
required at lower regenerant pH values (See Section 7.3.2.2).
7.3.2.2 Neutral pH Regeneration
Two of the largest municipal wastewater installations under design which will use
clinoptilolite are the Upper Occoquan (Virginia) Regional Plant (15 mgd or 0.66 m-^/sec) as
described in Section 9.5.4.1 and the Tahoe-Truckee (California) Sanitation Agency plant (6
mgd or 0.26 m^/sec), both of which will utilize a regenerant with a pH near neutral. The
active portion of the regenerant will be a 2 percent sodium chloride solution. Calcium and
potassium will be eluted as well as ammonia and will build up in the regenerant until they
reach equilibrium: The typical elution curve for ammonium with a neutral pH regenerant is
shown in Figure 7-11. Approximately 25-30 BV were required before the ammonium
concentration reached equilibrium.^ Although greater regenerant volumes are required than
7-17
-------
with a high pH regenerant (10-30 BV), this is not a major disadvantage if the regenerant is
recovered and reused in a closed loop system.
Variations in regenerant flow rates of 4-20 BV/hr do not affect regenerant performance.
Higher rates result in less ammonia removed per volume of regenerant. Typical design values
are 10 BV/hr which insures efficient use of the regenerant while keeping headless values at
low levels. Provisions should be made for backwash at rates of 8 gpm/ sq ft (3.9 l/m^/sec)
and surface wash of the contactor prior to initiation of the regenerant flow. Additional
details on neutral pH regeneration are contained in Section 9.5.4.
FIGURE 7-11
AMMONIUM ELUTION WITH 2 PERCENT SODIUM
CHLORIDE REGENERANT (REFERENCE 5)
1000
BED
7-18
-------
7.3.2.3 Effects on Effluent TDS
Effluent Total Dissolved Solids (TDS) is an important consideration in many plants. When a
2-3 percent solution of salt is used for regeneration, elution of this salt remaining in the bed
after regeneration at the start of the service cycle may result in an increase in TDS of about
50 mg/1. The increment would be greater with stronger regenerants. The TDS effect is much
less than for the breakpoint process, however.
7.3.3 Regenerant Recovery Systems
Ammonia may be removed from the regenerant so that the regenerant may be reused. Air
stripping of a high pH regenerant, air stripping of a neutral regenerant, steam stripping, or
electrolytic treatment may be used.
7.3.3.1 Air Stripping of High pH Regenerant
In the original pilot work on this approach to regenerant recovery, a stripping tower packed
with 1 inch (2.54 cm) polypropylene saddles was used.^3 Because the regenerant volume is
only a small portion of the total wastewater flow, it becomes feasible to heat the air used in
the stripping process. The regenerant was normally recycled upflow through the zeolite bed
at a flow rate of 4.8-7.1 gpm/sq ft (3.3-4.8 l/m^/sec) until the NH3—N approached a
maximum concentration. The regenerant was then recycled through both the zeolite bed
and the air stripper until the NH3—N was reduced to about 10 mg/1. The liquid flow rate to
the stripper was normally 2 gpm/sq ft (1.36 l/m^/sec) with an air/liquid ratio of 150
cu ft/gal (1.1 m^/l). Ammonia removal in the air stripper generally averaged about 40
percent per cycle at 25 C. Calcium carbonate scaling occurred on the polypropylene saddles,
but the scale could generally be removed by water sprays. The headloss through the 1 inch
(2.54 cm) pilot plant saddles caused the power requirements for the air stripping to be
excessive. It was suggested for a full-scale design that the ammonia stripping tower be sized
to treat the contents of an elutriant tank in 8 hours, using two passes through the tower at
85 percent removal per pass at an air-to-water ratio of 300 cfm/gpm (2.2 m^/1), and a
loading of 3.5 gpm/ft^(0.63 l/m2/sec)..2 The tower would be a modified cooling tower
with low differential pressure across the tower as discussed in Chapter 8.
An example design of a 7.5 mgd (0.33 m^/sec) system illustrates how the air stripping
system can be integrated into an overall system.^ A schematic diagram of the ion exchange
beds, lime elutriant system, and ammonia air stripping system is shown in Figure 7-12. For
design flows, nine beds (12 ft or 3.65 m diameter and 8 ft or 2.4 m deep) would be in
service and three beds in regeneration. The direction of flow for the beds in service would be
downflow. All beds would operate in parallel. When a given volume of wastewater has
passed through a set of three beds, for example, beds 1, 2, and 3, the set of beds would be
taken off line for regeneration. At this time elutriant tank A would contain elutriant water
from a previous regeneration with a very high ammonia nitrogen content (say 600 mg/1);
tank B would contain elutriant water with a low ammonia nitrogen content (say 100 mg/1);
7-19
-------
FIGURE 7-12
EXAMPLE ION EXCHANGE - AIR STRIPPING SYSTEM FOR HIGH pH REGENERANT
INFLUENT HEAOtR ,
INFLUENT FEED
ION
EXCHANGE
BEDS
9 x
ri
O
r
O x
Ts Tf TJ TT T*
j T FINAL j T TBEATCP J T HOOUCT j T MIAOtR ^ T
10
N)
O
'REGfNCRANT SOLUTION
l*ltT MCADEft
KIGCNEKANT
REOIMEHANT SOLUTION SUCTION LINC
BACKWASH Sumv LINI
TOWK INFLUfNT
_L
REGENERANT STORAGE TANKS
AMMONIA
STRIDING
T01KER
<
)
TOWtR
STORAGE
TANK
i) (£> 1 ' "
fS ^rS TOWIH INFLUCNT LINE
® r
\ TOWtR IFFLUINT LINE __ :
f 1
^— TOMER
HECVCLI
LINE
NoTes
1 BEOS 1.943 ARE SHOWN «M REGENERATION CYCLE
2 BEDS 4 . 12 ARE SHOWN IN SERVICE CVCLE
3 TOWER SHOWN STRIPPING FROM REGENERANT
TANK A. SINGLE CVCLE
TANKS B 4 C ARE IN REGENERATION CVCLE.
* ® INDICATES VALVE CLOScD
5 X INDICATES VALVE OPEN
-------
and tank C would contain nearly ammonia-free elutriant water (say 10 mg/1). The contents
of tank A would be air stripped during the regeneration of exchange beds 1, 2, and 3. The
regeneration would proceed as follows:
1. Exchange beds 1, 2 and 3 would be drained to the final effluent.
2. Low ammonia content elutriant water from tank B (100 mg/1) would be
recirculated upflow through the three exchange beds and back through tank B to
the exchange beds until the concentration of ammonia in the elutriant began to
approach a maximum value (say 600 mg/1). Throughout the recirculation,
make-up lime and salt would be added. A pH of about 11.5 would be maintained.
About 4 BV will elute 75 percent of the NH^-N.
3. After an allotted time (long enough for elutriant from tank B to approach a
maximum ammonia concentration), the elutriant would be changed to recircula-
tion to and from tank C through beds 1, 2, and 3. Tank C with its ammonia-free
elutriant water would be recirculated for an allotted time (long enough for
elutriant from tank C to reach about 100 mg/1) which will bring the elution up to
more than 90 percent. About 4 BV are required. At this stage of the elution, the
small amount of ammonia left on the zeolite would be distributed uniformly
throughout the bed. Tank A with nearly ammonia-free water (10 mg/1 Nrfy-N,
water stripped during the regeneration of beds 1, 2, 3) would be pumped once
upflow through the bed to further polish the lower portion of the bed and
prevent leakage of ammonia during the downflow service cycle.
The elutriant tank B (600 mg/1 — Nrfy—N) would be held for air stripping during
the regeneration of the next set (say beds 4, 5, and 6) of ion exchange beds. Tank
C with 100 mg/1 elution water would become the lead tank for this next set of ion
exchange beds. Tank A with ammonia-free elution water (10 mg/l-NH^-N,
water stripped during the regeneration of beds 1, 2, 3) would be used at the
polishing tank for beds 4, 5, and 6.
4. Once the elution of beds 1, 2, and 3 was completed, the three beds would be
drained back to tank A.
5. Beds 1, 2, and 3 would then be filled slowly from the bottom to remove trapped
air with product water from the other nine beds in service.
6. After the beds were filled with product water, more product water would be
pumped at a high rate through beds 1, 2, and 3 in sequence. The backwash water
would be returned to the wastewater treatment plant.
7. After backwashing was completed, ion exchange beds 1, 2, and 3 would be placed
in service and beds 4, 5, and 6 would be taken off line for regeneration.
7-21
-------
Ammonia in the elutriant solution would be removed by air stripping at a pH of about 11.5.
In the preceding example, during the regeneration of beds 1, 2, and 3, the very high
ammonia nitrogen content (600 mg/1) in the elutriant solution of tank A was to be air
stripped. The following procedure would be used:
1. The contents of tank A would pass through the tower down into the recycle basin
below the tower.
2. The contents of the recycle basin would then be pumped back up through the
tower once again. This time, however, the effluent from the tower would flow
back to tank A.
3. The contents of tank A would now contain about 10 mg/1 of ammonia, and
would be ready to serve as the polishing volume during the regeneration of ion
exchange beds 4, 5, and 6.
By using the above batch countercurrent recycle technique, it is possible to achieve
complete regeneration with only about 4 BV requiring stripping per cycle. This is a key to
making steam stripping (as discussed later) practical.
7.3.3.2 Air Stripping of Neutral pH Regenerant
As previously discussed, the use of high pH regenerant is accompanied by scab'ng problems
within the ion exchange beds. Thus, as discussed in Section 7.3.2.2, a regenerant with 2
percent sodium chloride as the active agent and pH nearer neutral has been used to
overcome the scaling problem. This regenerant may also be recovered for reuse by air
stripping. Figure 7-13 is a schematic diagram of such a system. 6 In this system, the stripping
tower off-gases are not discharged to the atmosphere but are instead passed through an
absorption tower where the ammonia in the off-gases is absorbed in sulfuric acid. The
stripping gases are recycled to the tower. This approach eliminates discharge of ammonia to
the atmosphere and recovers the ammonia in a form suitable for fertilizer usage. The
stripping-absorption approach is applicable to high pH regeneration systems as well. It also
reduces scaling problems in the stripping tower by limiting the CO 2 content of the stripping
air. This system (the Ammonia Removal and Recovery Process — "ARRP") is also discussed
in Chapter 8 (see Section 8.4.1 and Figure 8-7).
In the system shown in Figure 7-13, batch-countercurrent regenerant flow similar to that
described above for the high pH regenerant is practiced to reduce the amount of regenerant
which must be stripped per cycle. The first 11 BV of spent regenerant are discharged to the
spent regenerant tank for stripping. The second and third 11 BV batches are stored and used
as the first 22 BV of regenerant flow in the next regenerant cycle. The last 6-11 BV batch of
regenerant is mixed with the 11 BV of stripped regenerant for use as the final regenerant
flow in the next cycle. Thus, although 40-44 BV of regenerant are passed through the
exchanger per cycle, only 11 BV are actually renovated by air stripping per cycle.
7-22
-------
FIGURE 7-1 3
FLOW DIAGRAM OF NEUTRAL pH REGENERATION
SYSTEM USING AIR STRIPPING
FIRST 11 BED VOLUMES
11 BED VOLUMES
"~[SEC
VOL
CLINO
BED
'
J
1
SPENT
REGENERANT
TANK
DND 11 BED
UMES~~J
INTERMEDIATE
REGENERANT
TANK
TIRST 11 BED
[VOLUMES J
i LT.HI
•
•
[LA;
tlED VOLUMES
RD 11 BED
lURCS"^
INTERMEDIATE
REGENERANT
TANK
>T 6-11
7 \
'
RECOVERED
REGENERANT
TANK
1
PER R
EGENEf
I
NATION
^ 3 .... NoOH
^1 .. MAKEUP
NaCI
i
CLARIFIER
Y
SLUDGE
Mg(OH)2
+
X*^ _ H,S04
f ^ —
HI
STRIPPER ABSORBER "
o
V J *"
AHHK
LAb I 1 / — if. I
"BED"~VOLUMET
7-23
-------
Regenerant stored in the tanks shown in Figure 7-13 varies in ammonium-nitrogen
concentration from about 250 mg/1 in the spent tank to 50 mg/1 in the recovered regenerant
tank. The intermediate tanks have intermediate concentrations. The regenerant pH varies
from about 9.5 in the recovered regenerant tank to about 7.0 in the spent tank. As discussed
earlier, higher pH values produce more efficient regeneration but near-neutral pH levels
avoid problems with magnesium hydroxide precipitation in the bed during regeneration and
attrition of the clinoptilolite caused by high pH. Media attrition has been insignificant in
pilot studies under these pH conditions."
A typical ammonium elution curve is shown on Figure 7-14 with the background
concentrations in each regenerant storage tank also shown. At the end of the cycle, the last
portion of spent regenerant is discharged to the recovered regenerant tank. This has the
effect of neutralizing the alkaline pH from the ARRP process. ARRP effluent is normally at
a pH of 10.7 to 11.0. This is reduced to 9.5± by recycling the last portion of spent
regenerant. In this manner, pH is controlled without use of acid.
When spent regenerant is accumulated to a predetermined amount, the recovery portion of
the process is activated. This system operates at a flow rate of approximately 1/13 of
average plant flow since the regenerant concentration is about 13 times as concentrated as
plant waste. Initially, sodium hydroxide is added to the spent regenerant to achieve a pH of
about 11. Sodium chloride is also added because of some salt loss from the regenerant
solution in sludge removal and bed rinsing. Following pH adjustment, the regenerant is
clarified and any magnesium hydroxide formed is removed, the clarified regenerant at pH =
11 is then pumped to the ARRP process for ammonia removal and recovery. The ARRP
effluent flows to the recovered regenerant tank where it is mixed with the last 6-11 BV of
spent regenerant for pH adjustment prior to reuse.
This system is being used in the design of plants for the Tahoe-Truckee Sanitation Agency
(California) and for the Upper Occoquan (Virginia) Regional Plant discussed in Section
9.5.4.1.
7.3.3.3 Steam Stripping
Steam Stripping of regenerant is being practiced at the 0.6 mgd (0.026 m^/sec) Rosemount,
Minnesota physical-chemical plant. 7)8 This process is economically feasible only with high
pH regenerant.
The higher regenerant volumes resulting from the neutral regenerant approach are not
economically treated by this approach. This process is feasible only if the regenerant volume
requiring stripping is held to 4 BV per cycle which is achievable with the high pH regenerant
batch recycle system discussed in Section 7.3.3.1.9 In this case, the necessary portion of the
spent regenerant is stripped in a distillation tower in which steam is injected countercurrent
with the regenerant. An air cooled plate-and-tube condenser condenses the vapor for
collection in a covered tank as a one percent aqueous ammonia solution which could be
7-24
-------
FIGURE 7-14
TYPICAL ELUTION CURVE
500
40O
300
I
200
IOO
CONCENTRATION BEING
ELUTED FROM BED
BACKGROUND CONCENTRATION
IN FIRST REGENERANT TANK
SECOND REGENERANT
TANK
RECOVERED
REGENERANT
TANK
10 20 30
TOTAL BED VOLUMES
40
50
used as a fertilizer. A stripping tower depth of 24 feet (7.3 m) and a loading of 7 gpm/ sq ft
(21 l/m^/sec) are being used at Rosemount. Ceramic saddles are used rather than wooden
slat packing because wood is not a suitable packing in a high pH- steam environment.
Heat exchangers are used to transfer heat from the stripped regenerant to the incoming, cold
regenerant. Heat transfer to the incoming regenerant from the condenser used to condense
the stripped regenerant may also be attractive. Provisions for scale control in the heat
exchangers should be provided. The steam requirements have been estimated to be 15
pounds per 1,000 gallons (1.8 g/1). Added information may be found in the Rosemount
design example, Section 9.5.4.2.
7-25
-------
7.3.3.4 Electrolytic Treatment of Neutral pH Regenerant
In this approach, ammonium in the regenerant solution is converted to nitrogen gas by
reaction with chlorine which is generated electrolytically from the chlorides already present
in the neutral pH regenerant solution. The regenerant solution is rich in NaCl and CaCl2
which provide the chlorine produced at the anode of the electrolysis cell. A diagram of the
regeneration system is presented in Figure 7-15. The regeneration of the clinoptilolite beds
is accomplished with a two percent sodium chloride solution. The spent regenerant is
collected in a large holding tank and then subjected to soda ash treatment for calcium
removal. After the soda ash addition, the regenerant is clarified and transferred to another
holding tank where the regenerant is recirculated through electrolysis cells for ammonia
destruction.
FIGURE 7- 15
SIMPLIFIED FLOW DIAGRAM OF ELECTROLYTIC
REGENERANT TREATMENT SYSTEM
Na2C03
NaOH
ZEOLITE
BED
I
H2,N2
SPENT
REGENERANT
HOLDING
TANK
CLARIFIER
t
RENOVATED
REGENERANT
HOLDING
TANK
SLUDGE
REGENERANT
REG
IN
ANODE
REG .1 I
OUT
CO
CO
> CO
CC QJ
•- o
o
UJ
-I
UJ
CATHODE
RECTIFIER
7-26
-------
During the regeneration of the ion exchange bed, a large amount of calcium is eluted from
the zeolite along with the ammonia. This calcium tends to scale the cathode of the
electrolysis cell, greatly reducing its life. Calcium may be removed from the spent regenerant
solution by a soda ash softening process prior to passing the spent regenerant through the
electrolytic cells. High flow velocities through the electrolysis cells are required in addition
to a low concentration of MgCl2 to minimize scaling of the cathode by calcium hydroxide
and calcium carbonate. The effects of flow rate are well illustrated by pilot test data. 5 Using
the system shown in Figure 7-14, the flow rate through the cell was initially set at velocities
of 0.13 to 0.16 ft per second (0.04-0.05 m/sec) and a thin buildup of scale was observed on
the cathode at the bottom cell inlet end after 160 hours of operation. After 230 hours of
operation, the flow velocity was reduced to 0.06 ft per second (0.018 m/sec) and very light
scale buildup was observed depositing over the entire cathode area. Scale was removed from
a one-square inch (6.45 cm2) area of the cathode and the flow velocity through the cell was
increased to 0.21 ft per second (0.064 m/sec) to determine the effect of scaling at higher
cell velocities. At this increased flow which was maintained for most of the period of this
study, no new scale was deposited on the cathode. Visually it appeared that from 25 to 50
percent of the previously deposited scale was removed. These observations suggest that
scaling within the cell can be controlled by sufficient flow velocities. Acid flushing of the
cells is necessary to remove this scale when the cell resistance becomes too high for
economical operation.
In pilot tests of the electrolytic treatment of the regenerant at Blue Plains, about 50
watt-hours of power were required to destroy one gram of ammonia nitrogen.2 When
related to the treatment of water containing 25 mg/1 Nlfy—N, the energy consumed would
be 4.7 kwh/1,000 gallons (1.2 watt-hrs/1). Tests at South Tahoe also indicated that a value
of 50 watt-hours per gram is reasonable for design.5
The electrolytic process also results in about 56 cu ft (1586 1) of hydrogen gas being evolved
per pound of ammonia nitrogen destroyed. Provisions must be made to vent, burn, or
otherwise adequately control the hydrogen gas evolved in the electrolytic process.
The major disadvantage of the electrolytic approach is the substantial amount of electrical
energy required. The electrical requirements of the air stripping (ARRP) system described in
the preceding section are only about 10 percent of that required by the electrolytic process.
7.4 Considerations in Process Selection
The selective ion exchange process has the advantages of high efficiency, insensitivity to
temperature fluctuations, and removal of ammonium with a minimal addition of dissolved
solids. It may also be used with regenerant recovery systems which enable the recovery of
the nitrogen removed from the wastewater in a reusable form. Its major disadvantage is its
relatively complex operation. The process should be controlled by a system which will
automatically initiate and program the regeneration cycle and return the ion exchangers to
normal service.
7-27
-------
The process is particularly attractive for those cases requiring year-round high level removal
of nitrogen and where effluent TDS is of major concern. Although the effluent TDS is
increased by the process (see Section 7.3.2.3), the overall increase is much less than for the
breakpoint chlorination process. It must be recognized in the sizing of the upstream process
capacities that there will be backwash wastes returned from the ion exchange process. The
capacity of the clinoptilolite may be predicted accurately, based on the concentrations of
ions present in the wastewater, minimizing the need for pilot tests for defining ion exchange
capacity. Pilot tests of the overall ion exchange-regenerant recovery system may be useful,
however, in evaluating physical and economic aspects of the proposed system design as
applied to a specific wastewater.
7.5 References
1. Koon, J.H. and W.J. Kaufman, Optimization of Ammonia Removal by Ion Exchange
Using Clinoptilolite. Environmental Protection Agency Water Pollution Control
Research Series No. 17080 DAR 09/71.
2. Battelle Northwest and the South Tahoe Public Utility District, Wastewater Ammonia
Removal by Ion Exchange. Environmental Protection Agency Water Pollution Control
Research Series No. 17010 ECZ 02/71, February, 1971.
3. Battelle Northwest, Ammonia Removal from Agricultural Runoff and Secondary
Effluents by Selective Ion Exchange. Robert A. Taft Water Research Center Report
No. TWRC-5, March, 1969.
4. Mercer, B.W., Clinoptilolite in Water Pollution Control. The Ore Bin, published by
Oregon Dept. of Geology and Mineral Industries, p. 209, November, 1969.
5. Prettyman, R., et al, Ammonia Removal by Ion Exchange and Electrolytic
Regeneration. Unpublished report, CH2M/Hill Engineers, December, 1973.
6. Suhr, L.G., and L. Kepple, Design of a Selective Ion Exchange System for Ammonia
Removal. Presented at the ASCE Environmental Engineering Division Conference,
Pennsylvania State University, July, 1974.
7. Physical-Chemical Plant Treats Sewage Near the Twin Cities. Water and Sewage Works,
p. 86, September, 1973.
8. Larkman, D., Physical-Chemical Treatment. Chemical Engineering, Deskbook Issue, p.
87, June 18, 1973.
9. Personal communication, B.W. Mercer, Battelle Northwest, December 14, 1973.
7-28
-------
CHAPTER 8
AIR STRIPPING FOR NITROGEN REMOVAL
8.1 Chemistry and Engineering Principles
The ammonia stripping concept is based on very simple principles. Because of its simplicity,
it offers a reliable means of ammonia removal when applied under appropriate conditions.
The following section describes the basic concept.
8.1.1 Basic Concept
The equilibrium equation for ammonia in water is represented by:
NH* i " NH3* + H+ (8-1)
At ambient temperatures and pH 7, the reaction is nearly complete to the left and only
ammonium ions are present. As the pH,is increased above 7, the reaction is driven to the
right, and the fraction of dissolved ammonia gas increases until at pH values of 10.5-11.5,
essentially all of the ammonium is converted to NH3 gas (see Figure 6-2). The gaseous .form
may be removed by stripping.
The ammonia stripping process itself (Figure 8-1) consists of: (1) raising the pH of the water
to values in the range of 10.8 to 11.5 generally with the lime used for phosphorus removal,
(2) formation and reformation of water droplets in a stripping tower, and (3) providing
air-water contact and droplet agitation by circulation of large quantities of air through the
tower. The towers used for ammonia stripping of municipal wastewaters closely resemble
conventional cooling towersJ Countercurrent towers, as opposed to cross-flow towers,
appear best suited to ammonia stripping applications.
Detailed discussions of mass and enthalpy relationships and theoretical mathematical models
of the stripping process are available in references 2, 3 and 4. However, these models are not
normally used for stripping tower design, and an empirical design procedure is used.
Before addressing detailed design considerations, the general environmental impacts of the
stripping process must be evaluated. It is obvious from Figure 8-1 that ammonia is being
discharged into the atmosphere. Does the process solve a water pollution problem while
creating an air pollution problem? What is the fate of the ammonia in the atmosphere?
These questions must be satisfactorily addressed prior to the selection of air stripping for
ammonia nitrogen removal.
8.2 Environmental Considerations
There are three major potential environmental impacts which must be evaluated if use of the
8-1
-------
FIGURE 8-1
AMMONIA STRIPPING PROCESS
AIR
^^ OUT s&
JOL
HIGHpH
INFLUENT
PACKED
TOWER
/ AIR IN X
EFFLUENT TO pH ADJUSTMENT
AND OTHER AWT PROCESSES
ammonia stripping process is proposed: air pollution, washout of ammonia from the
atmosphere, and noise. If these three-concerns cannot be favorably resolved for any given
situation, then the potential process advantages of simplicity and low cost may become only
academic.
8.2.1 Air Pollution
At an air flow of 500 cu ft per gallon (3.7 m^/1) and at an ammonia concentration of 23
mg/1 in the tower influent, the concentration of ammonia in the stripping tower discharge is
about 6 mg/m^. As the odor threshold of ammonia is 35 mg/m^, the process does not
present a pollution problem in this respect. Concentrations of 280-490 mg/m^ have been
reported to cause eye, nose, and throat irritation. 5 Concentrations of 700 mg/m^ can have
adverse effects on plants. Concentrations of 1,700-4,500 mg/m^ must be reached before
8-2
-------
human or animal toxicities begin to occur. Ammonia discharged to the atmosphere is a
stable material that is not oxidized to nitrogen oxides in the atmosphere. 6 Ammonia can
react with sulfur dioxide and water to form an ammonium sulfate aerosol. However, for this
consideration to be a limitation, the stripping tower would have to be located adjacent to a
point source of sulfur dioxide.
The production and release of ammonia as part of the natural nitrogen cycle is about
50,000,000,000 tons per year. Roughly 99.9 percent of the atmosphere's ammonia
concentration is produced by natural biological processes, primarily the bacterial breakdown
of amino acids. 5.6 Although they are relatively insignificant sources, burning of coal and oil
produces measurable quantities of ammonia. 6 The background levels of ammonia in the
atmosphere have been observed to vary from .001 mg/m-* to 0.02 mg/m^ with a value of
0.006 mg/m^ being typical.6
Available diffusion technology can be used to estimate the atmospheric concentration of
ammonia at any point downwind of the stripping tower. 7 Calculations were made for the
Orange County, California stripping tower for low mixing conditions (wind speed 1 m/sec).
The resulting surface concentrations at the center of the downwind discharge zone including
natural background levels were as follows:
Distance from Tower Surface Air Concentration
ft. m of Ammonia, mg/m^
300 91 5.2
1,000 305 1.6
1,600 488 0.6
3,200 975 0.2
16,000 4,877 0.0006
Background levels of ammonia are reached within 3 miles (4.8km). No U.S. ammonia
emission standards have been established by regulatory agencies because there are no known
public health implications at concentrations normally encountered.
The American Conference of Governmental Industrial Hygenists recommended in 1967 an
occupational threshold limit of 35 mg/m^.S The permissible limit for ammonia in a
submarine during a 60 day dive is 18 mg/m^.S The Navy's Bureau of Medicine and Surgery
has recommended an ammonia threshold limit for 1 hour of 280 mg/m^. All of these values
are above the 6 mg/m^ which will typically occur at the tower discharge. As noted above,
no ambient air quality standards for ammonia exist for the United States. However, such
ambient air standards exist for Czechoslovakia, the U.S.S.R., and Ontario, Canada, as shown
below: 5
8-3
-------
Basic Standard Permissible
mg/mr Averaging mg/rn^ Averaging
Location Time Time
Czechoslovakia 0.1 24 hr 0.3 30 min
U.S.S.R. 0.2 24 hr 0.2 20 min
Ontario, Canada 3.5 30 min
8.2.2 Washout of Ammonia from the Atmosphere
There is a large turnover of ammonia in the atmosphere with the total ammonia content
being displaced once per week on the average. Ammonia is returned to the earth through
gaseous deposition (60 percent), aerosol deposition (22 percent), and precipitation (18
percent). 5 Although not the most significant mechanism for removal of ammonia from the
atmosphere, precipitation does provide one pathway for the return of atmospheric ammonia
to bodies of water and to soil. In rainfall, the natural background ranges from 0.01 to 1 mg/1
with the most frequently reported values of 0.1 to 0.2 mg/1. The amount of ammonia in
rainfall is directly related to the concentration of ammonia in the atmosphere. Thus, an
increase in the ammonia in rainfall would occur only in that area where the stripping tower
discharge increases the natural background ammonia concentration in the atmosphere.
Calculations for the ammonia washout in the rainfall rate of 3 mm/hr (0.12 in./hr) have been
made for the Orange County, California project with the following results:
Peak Rainfall Ammonia
Distance From Tower Concentration, mg/1
ft m
60
18
11
5
0.5
The concentrations of ammonia in the rainfall would approach natural background levels
within 16,000 ft (4.8 km) of the tower. The ultimate fate of the ammonia which is washed
out by rainfall within this 16,000 ft (4.8 km) downwind distance depends on the nature of
the surface upon which it falls. Most soils will retain the ammonia. That portion which lands
on paved areas or directly on a stream will appear in the runoff from that area. Unless the
stripping tower is located upwind in close proximity to a lake or reservoir, the direct return
of ammonia to the aquatic environment by atmospheric washout should not make a
significant contribution to the total ammonia discharged to the aquatic environment.
However, this is a factor which must be carefully evaluated for each potential application.
300
1,000
1,600
3,200
16,000
91
305
488
975
4,877
-------
8.2.3 Noise
There are three potentially significant noise sources in an ammonia stripping tower: (1)
motors and fan drive equipment; (2) fans; and (3) water splashing. The following control
measures are available:
• Motors - proper installation, maintenance and insulation
• Fans - reduction in tip speed and exhaust silencers
• Water — shielding of the tower packing and air inlet plenum
Based on sound level measurements from the tower at Lake Tahoe, the expected noise level at
the tower is calculated to be about 64 decibels (dBA). This noise level can be reduced to 46
dBA at 600 ft (183 m) from the towers by control measures. The Orange County project (Sec.'
9.5.5.2) includes several specific noise control measures. Before construction of the plant,
ambient nighttime noise levels in the residential neighborhood around the Orange County
plant were 40-45 dBA.
8.3 Stripping Tower System Design Considerations
The major factors affecting design and process performance include the tower configuration,
pH, temperature, hydraulic loading, tower packing depth and spacing, air flow, and control
of calcium carbonate scaling.
8.3.1 Type of Stripping Tower
There are two basic types of stripping towers now being used in full-scale applications:
countercurrent towers and cross-flow towers (see Figure 8-2). Countercurrent towers (the
entire airflow enters at the bottom of the tower while the water enters the top of the tower
and falls to the bottom) have been found to be the most efficient. In the crossflow towers,
the air is pulled into the tower through its sides throughout the height of the packing. This
type of tower has been found to be more prone to scaling problems (see Section 8.3.7).
8.3.2 pH
The pH of the water has a major effect on the efficiency of the process. The pH must be
raised to the point that all of the ammonium ion is converted to ammonia gas (see Section
8.1.1). If phosphorus removal is required, the use of lime as the coagulant will generally
enable the necessary pH elevation to be achieved concurrent with phosphorus removal. If
pH elevation does not occur in some upstream processes, then the economics of the
stripping process are adversely affected since the costs of pH elevation must then be
incurred solely for ammonia stripping.
8-5
-------
FIGURE 8-2
TYPES OF STRIPPING TOWERS (REFERENCE 8)
^-COLLECTION BASIN
CROSS-FLOW TOWER
FAN
1 AIR
[OUTLET
AIR INLET
DRIFT
ELIMINATORS
DISTRIBUTION
SYSTEM
-AIR INLET
WATER
COLLECTING BASIN
COUNTERCURRENT TOWER
8.3.3 Temperature
A critical factor is the air temperature. The water temperature reaches equilibrium at a value
near the air temperature in the top few inches of the stripping tower. As the water
temperature decreases, the solubility of ammonia in water increases and it becomes more
difficult to remove the ammonia by stripping. The amount of air per gallon must be
increased to maintain a given degree of removal as temperature decreases. However, it is not
practical to supply enough air to fully offset major temperature decreases. For example, at
20 C, 90-95 percent removal of ammonia is typically achieved. At 10 C, the maximum
practical removal efficiency drops to about 75 percent. Data collected in pilot tests by EPA
at the Blue Plains plant in Washington, D.C. well illustrate the temperature effects and are
shown in Figure 8-3.9 in warm weather tests at pH 11.5, with inlet air and water
temperatures averaging 25.5 C and 26 C respectively, air stripping cooled the outlet water
temperature by evaporation of the liquid within the tower to an average of 22.2 C. In a
8-6
-------
FIGURE 8-3
EFFECT OF TEMPERATURE ON AMMONIA REMOVAL EFFICIENCY
OBSERVED AT BLUE PLAINS PILOT PLANT (REFERENCE 9)
too
80
UJ
o
UJ
a. 6O
uj 4Q
oc
Z
O
s 20
«t
TOWER EFFLUENT WA.TER
TEMPERATURE = 22.2 C
/' • LIQUID RATES
>/'* gpm/sf
/ TOWER EFFLUENT WATER B 2 44
/ TEMPERATURE =5C A 2 QQ
• 1.50
Metric Conversion: * '-O0
1 gpm/sf = 0.68 4/m2/sec
1 cu ft/gal = 7.48 m3/m3
0 200 400 6OO 8OO
CU FT AIR /GAL LIQUID
similar test with the inlet air temperature averaging 6 C and the inlet water temperature
averaging 16 C, the air stripping cooled the outlet water to an average temperature of 5 C.
Data from both the 22.2 C and 5 C conditions are shown in Figure 8-3. The decrease in
efficiency from the warm to cold temperatures was approximately 30 percent over a wide
range of air to water flows.
When air temperatures reach freezing (or when the wet bulb temperature of the air within
the tower reaches 0 C), the tower operation must generally be shut down due to icing
problems. The very large volumes of air required for the stripping process make it
impractical to heat the air in cold climates. Waste heat from potential on-site sources such as
sludge incinerators is typically only a small percentage of that needed.
8.3.4 Hydraulic Loading
The hydraulic loading rate of the tower is an important factor. This is typically expressed in
terms of gallons/minute applied to each square foot (or l/m^/sec) of the plan area of the
8-7
-------
tower packing. When the hydraulic loading rates become too high, the good droplet
formation needed for efficient stripping is disrupted and the water begins to flow in sheets.
If the rate is too low, the packing may not be properly wetted resulting in poor performance
and scale accumulation. Data collected in pilot tests at South Tahoe illustrate this
relationship and are shown in Figure 8-4.^ In optimum summer conditions, tthe pilot data
indicate that a flow rate of 2 gpm/sf (1.4 l/m^/sec) is compatible with efficient tower
operation at 20-24 ft (6.1-7.3 m) packing depths. Adequate flow distribution over the entire
packing area is a critical factor. Full-scale towers at Orange County and Pretoria, South
Africa are based on tower loadings of 1-1.13 gpm/sf (0.68-0.771/m^/sec).
FIGURE 8-4
PERCENT AMMONIA REMOVAL VS. SURFACE LOADING RATE
FOR VARIOUS DEPTHS OF PACKING (REFERENCE 10)
100
Metric Conversion:
1 ft = 0.305m
1 gpm/sf = 0.68 ^/m2/sec
I I I
Z
UJ
o
cc
UJ
Q.
- 60 -
5
O
5
UJ
Q:
5
5
2O-
2.O 3.0 4.0 5.0
SURFACE LOADING RATE, gpm/sf
8-8
-------
8.3.5 Tower Packing
8.3.5.1 Packing Depth
The depth of tower packing required for maximum ammonia removal will depend on the
tower packing selected. Most stripping tower designs are based on the use of an open,
cooling tower-type packing (horizontal packing members spaced about 2 in (5.1 cm) apart
both horizontally and vertically) to minimize the power required to move adequate air
quantities through the tower (see Sec. 8.3.6). If maximum removals are desired, tower
packing depth should be at least 24 ft (7.3 m) with this type of packing, unless pilot plant
data indicate that a lesser depth of a specific packing will accomplish the required removal.
Packing with members spaced more than 2 inches (5.1 cm) apart may require greater depths,
and pilot tests should be run to determine the required depth if greater spacings are
proposed.
8.3.5.2 Packing Material and Shape
Both wood (Lake Tahoe) and plastic packings (Orange County) have been used in full-scale
towers. The smooth plastic surfaces appear to be one factor accounting for reduced calcium
carbonate scaling at the Orange County facility. Plastic packing has an advantage in that it
does not suffer from the delignification that occurs with wood at elevated pH values.
Pilot studies at Orange County evaluated three different types of packing: Vi inch (1.27 cm)
diameter PVC pipe, triangular shaped splash bars, and vertical film packing like that used in
cooling towers. * 1 With vertical packing the water moves in a thin film down vertical sheets
of packing rather than moving as droplets as occurs in packing composed of horizontal
splash bars. The film packing was found to provide only 50 percent or less ammonia removal
and was eliminated from consideration early in the tests. Film packing fails to provide the
repeated droplet formation and rupture needed for efficient stripping.
Since repeated splashing and droplet formation is a key parameter in ammonia stripping, a
triangular shaped splash bar was tested. It was thought that it might provide two points of
droplet formation compared to only one for a round splash bar. It was observed that droplet
formation throughout the tower still occurred at only one point. The water flowed down
the sides of the triangle and around the corners, where it collected on the base and dripped
from a single point. Air flow and pressure drop measurements were made on both the
circular and triangular packing. The static pressure drop (24 ft or 7.3 m packing depth) was
0.40-0.44 in (1-1.1 cm) of water when the triangular packing was used, compared to
0.36-0.40 in (0.9-1 cm) of water when the circular packing was used. No significant
differences in ammonia removals were noted between the round and triangular shaped
packings.
8-9
-------
8.3.5.3 Packing Spacing and Configuration
Figure 8-5 is a sideview of a typical packing configuration using wood or plastic slats. In this
example, the slats are spaced 2 in. (5.1 cm) apart (center to center) on the horizontal and
1.5 in. (3.8 cm) apart vertically. This spacing is referred to as 1.5 x 2 in.(3.8 x 5.1 cm).
Figure 8-6 shows that the spacing of the tower packing is important in determining the air
requirements for ammonia stripping. The 1.5 x 2 in.(3.8 x 5.1 cm) packing has 2.66 more
slats for droplet formation and coalescing than does a 4 x 4 inch (10.2 x 10.2 cm) packing.
Although spacing the packing members closer than 1.5 x 2 inches (3.8 x 5.1 cm) would
improve performance, the increased pressure drop would greatly increase power costs (see
Sec. 8.3.6).
Tests at Orange County indicate that packings in which alternate layers of packing are
placed at right angles, rather than the parallel position shown in Figure 8-5, maintains better
flow distribution and may be less susceptible to scale accumulation.
FIGURE 8-5
ILLUSTRATIVE PACKING CONFIGURATION
5.1 cm
3.8 cm
f.5
I
8-10
-------
FIGURE 8-6
EFFECT OF PACKING SPACE ON AIR REQUIREMENTS
AND EFFICIENCY OF AMMONIA STRIPPING (REF. 1)
I 'I ' I I
'/2 x 2QIN. PACKING (REDWOOD SLATS) o
4x4 IN. PACKING (PLASTIC TRUSS BARS)
NOTE: 24 FT PACKING DEPTH
500 1000 /500 2000 2500 3OOO 3500 40OO
CU FT AIR/QALLON TREATED
8.3.6 Air Flow
Gas transfer relationships indicate that an increase in ammonia removal'can be achieved by
increasing the air flow for a given tower height (see Figures 8-3 and 8-6). However, there is a
practical limit on air flow rate due to the increase in air pressure drop with increasing flow
rate. This results in higher capital investment for fans and increased power costs. The air
pressure drop in a countercurrent tower is given as: 2
where:
P = f-z-Q .
Pressure drop, in. of water
Fanning friction factor
Air flow rate, cu ft/min/sq ft (m /min/m )
(8-2)
air
z = Packing height, ft
8-11
-------
Pressure drop increases exponentially with air flow rate. In general, air velocities of 550 cu
ft/min/sq ft (1600 m^/min/m^) are considered to be the practical upper limit for
countercurrent towers. The friction factor should be obtained from the packing manufac-
turer. General guides for wood grids are available in reference 3.
Figure 8-3 reflects the effects of the ratio of air to wastewater as observed at the Blue Plains
pilot plant. These data are in general agreement with similar pilot data collected at South
Lake Tahoe (Figure 8-6). For warm weather conditions, typical air requirements are about
300 cu ft/gal (2240 m3/m3) for 90 percent removal and 500 cu ft/gal 3740 m3/m3 for 95
percent removal. In cold weather conditions, the air requirements to achieve maximum
tower efficiency increase substantially. Full-scale data at Tahoe indicate that, for their
packing design, air flows of about 800 cu ft/gal (5980 m3/m3) would be needed to achieve
90 percent removal at an air temperature of 4 C. However, reliance solely on the stripping
process in cold weather conditions is usually not practical, and most designs are based on
moderate to warm weather conditions. Typical air design quantities for 90 percent removal
are as follows: Orange County, California - 400 cf/gal (2990 m3/m3); South Lake Tahoe -
390 cf/gal (2920 m3/m3); Pretoria, South Africa - 338 cf/gal (2530 m3/m3).
The required air quantities are usually provided by a fan located on top of the tower. Two
speed fan motors may be used to better match air supplied to the actual requirements.
Because of the low pressure drops associated with the types of packings typically used (less
than 1 inch or 2.5 cm water), the horsepower requirements for the fans are not great for
these large quantities of air. For the 15 mgd (0.65 m3/sec) Orange County plant, the total
installed fan brake horsepower is 1380 HP.
8.3.7 Scale Control
A factor which may have an adverse effect on tower efficiency is scaling of the tower
packing resulting from deposition of calcium carbonate from the unstable, high pH water
flowing through the tower. Scaling potential can be minimized by maximizing the extent of
completion of the calcium carbonate reaction in the lime treatment step. Using a high level
of solids recycle in the clarification step will ensure more complete reaction. Another
approach is to eliminate CO2 from the air (see Section 8.4).
The original crossflow tower at the South Lake Tahoe plant has suffered a severe scaling
problem. The severity of the scaling problem was not anticipated from the pilot studies in
which a countercurrent tower was used. As a result, the full-scale cross-flow tower packing
was not designed with access for scale removal in mind. Thus, portions of the tower packing
are inaccessible for cleaning. Those portions which were accessible were readily cleaned by
high pressure hosing.
The severity of the scaling problem has varied widely. Perhaps the most severe case is that
reported at the Blue Plains pilot plant. 9 When operating at a pH of 11.5, a heavy scale of
8-12
-------
calcium carbonate formed on the crossflow tower packing (polypropylene grids). The scale
was crystalline, hard, and could not be removed by a high pressure water hose. In contrast,
several months of operation of the countercurrent pilot tower at Orange County at pH
values above 11 resulted in only a thin coat of calcium carbonate scale on the 0.5 in.(l-3
cm) diameter PVC pipe packing. The thin coat of scale stabilized and did not continue to
accumulate. The scale was friable and easily removed by water hosing. The Orange County
pilot tower was later moved to South Tahoe where several months of operation indicated no
scale buildup. The differences in operation between the pilot tower at Tahoe and the
full-scale Tahoe tower were as follows: packing of plastic rather than wood, packing shape
round rather than rectangular, countercurrent tower rather than parallel. The relative
importance of these factors in eliminating the scaling noted in the full-scale Tahoe tower is
uncertain. The experience with the full-scale (1 mgd) countercurrent tower at Pretoria,
South Africa is similar to that at the full-scale Tahoe tower in that the scale formed can be
readily flushed from the packing by a water jet. ^ The feeding of a scale-inhibiting polymer
to the tower influent may also offer a means of scale control and such provisions are being
made in the Orange County facility.
In light of the as-yet unpredictable nature of factors contributing to scaling of tower
packing, it is prudent to conduct pilot tests for 3-6 months on the specific wastewater
involved with the specific tower configuration proposed. The pilot vs. full-scale experience
at Tahoe and the independent pilot tests at Orange County indicate that the use of
countercurrent rather than crossflow towers will reduce scaling problems. Also, access
should be provided to all of the tower packing for cleaning (see Sec. 9.5.5.2 for a discussion
of the Orange County plant).
8.4 Ammonia Recovery or Removal From Off-Gases
As noted earlier, the calcium carbonate scaling problem can be minimized or eliminated by
removing carbon dioxide (CO2) from the stripping air. This section describes two
approaches which accomplish this goal by removing ammonia and CO2 from the off-gas
from the tower and recycling the air.
8.4.1 Acid Systems
An approach to overcoming the limitations of the stripping process is currently being
developed. 13 it appears that the process overcomes many limitations of the stripping
process and has the advantage of recovery of ammonia as a byproduct.
The improved process is shown diagramatically on Figure 8-7. The process includes an
ammonia stripping unit and an ammonia absorption unit. Both of these units are sealed
from the outside air but are connected by appropriate ducting. The stripping gas, which
initially is air, is maintained in a closed cycle. The stripping unit operates in the same
manner as is now being or has been done in a number of systems with the exception that the
gas stream is recycled rather than outside air being used in a single pass manner.
8-13
-------
FIGURE 8-7
PROCESS FOR AMMONIA REMOVAL AND RECOVERY
WASTEWATER
CONTAINING
DISSOLVED
AMMONIA (NH3)
FAN (TYPICAL)
RECYCLE , .
ALTERNATE 11
PUMP
VS STREAM WITH
/IMONIA INCREASED
A, AA A
STRIPPING
UNIT
Vj^TS ^V
/TVi/ \
DUCTING (TYPICAL)
V
A
ABSORPTION
UNIT
RECYCLED
ABSORBENT
LIQUID
GAS STREAM-AMMONIA
REDUCED BY ABSORPTION
ACID AND
WATER MAKEUP
AMMONIUM SALT
SLOWDOWN (LIQUID
OR SOLID), OR
DISCHARGE TO STEAM
STRIPPER FOR AMMONIA
GAS REMOVAL AND
RECOVERY
WASTEWATER STRIPPED OF NEARLY
ALL OR PART OF AMMONIA (NH3)
Most of the ammonia discharged to the gas stream from the stripping unit is removed in the
absorption unit. Because of the favorable kinetics of the absorption reaction, the absorption
unit may be reduced in size by about one third from that required for the stripping unit.
The absorbing liquid is maintained at a low pH to convert absorbed and dissolved ammonia
gas to ammonium ion. This effectively traps the ammonia and also has the effect of
maintaining the full driving force for absorbing the ammonia since dissolved ammonia gas
does not build up in the absorbent liquid. The absorption unit can be a slat tower or packed
tower using sprays similar to the stripping unit, but will usually be smaller due to the
kinetics of the absorption process.
The absorbent liquid initially consists of water with acid to obtain low pH (usually below
7). In the simplest case, as ammonia gas is dissolved in the absorbent and converted to
ammonium ions, acid is added to maintain the desired pH. If sulfuric acid is added, as an
8-14
-------
example, an ammonium sulfate salt solution is formed. This salt solution continues to build
up in concentration and the ammonia is finally discharged from the absorption device as a
liquid or solid (precipitate) blowdown of the absorbent. With shortages of ammonia based
fertilizers, a saleable byproduct may result. Ammonia sulfate concentrations of 50 percent
are obtainable.
Mist eliminators are necessary between the absorber and stripper to prevent carryover of the
ammonia laden moisture from the absorber to the stripper effluent. Because of the headless
in the mist eliminators and absorber packing, total headless for the air approaches 2 inches
(5.1 cm). It is believed that the usual scaling problem associated with ammonia stripping
towers will be reduced by the improved process, since the carbon dioxide which normally
reacts with the calcium and hydroxide ions in the water to form the calcium carbonate scale
is eliminated from the stripping air during the first few passes. The freezing problem is
eliminated due to the exclusion of nearly all outside air. The treatment system will normally
operate at the temperature of the wastewater.
As discussed in Chapter 7, this approach is being used for stripping of ammonia from
selective ion exchange process regenerant. Because of the higher power requirements (as
compared to single-pass stripping), the use of this process may be limited to regeneration of
the brines from the selective ion exchange process. Full-scale designs of this application are
underway for a 22.5 mgd (1.0 m^/sec) plant serving the Upper Occoquan Sewage Authority,
Virginia and for a 6 mgd (0.26 m^/sec) plant for the Tahoe-Truckee Sanitation Agency,
California. Design criteria for this application are presented in Sec. 9.5.4.1 which describes
the Upper Occoquan plant.
8.4.2 Nitrification-Denitrification
Another approach to achieving many of the same objectives of the acid system described
above is also being developed. 14 The system, termed the Ammonia Elimination System
(AES), is shown in Figure 8-8. Basic elements of the process are as follows:
1. Ammonia Stripping Tower. Ammonia is transferred in this tower from the liquid
stream to the gas stream at high pH.
2. Ammonia Absorption-Oxidation Tower. Ammonia is transferred from the gas
stream to the liquid stream at about pH 8. Ammonia is oxidized to nitrate by
nitrifying bacteria in this tower. The following reaction summarizes the reactions
in this tower:
(8-3)
3. Denitrifying Reactor. Nitrate is reduced to nitrogen gas in this reactor (methanol
or a waste carbon source is added to the reactor) as shown by the reaction:
8-15
-------
N0~ + 1 16
5/6 CH3(OH)
1 /2
8/6
HCO
(8-4)
4. Solids Separation. Denitrification organisms are settled from the process stream
and are returned to the denitriflcation reactor or wasted.
5. The overall reaction in the AES system can be found by adding equations 8-3 and
8-4:
2O2 + 5/6 CH3(OH) — •- 1/2 N2 f + 7/3 H2
-------
1. Removal of nitrogen from the main flow stream by a physical process (which has
advantages over biological systems in terms of ease of process control).
2. Isolation of the nitrification stage from most agents in the plant influent which
are toxic to nitrifying organisms.
3. Isolation of the ammonium oxidation stage from the carbon oxidation or removal
stage.
4. Nitrogen is eliminated from the system as inoffensive nitrogen gas.
In addition to the above advantages, the AES has the following potential advantages not
shared by either ammonia stripping or the three sludge system:
1. The insulation and heating of the liquid recycle streams, oxidation column,
denitrification reactor, and clarifier(s) to increase process efficiency becomes
economically feasible.
2. A waste carbon source may be used in place of me than ol, since denitrification is
accomplished on a side stream.
The AES shares the following advantages of the acid system described in Sec. 8.4.1:
1. Free ammonia gas is not discharged to the atmosphere.
2. Water vapor is not discharged to the atmosphere in large volumes.
3. The heating of air to prevent tower freezing and increase process efficiency
becomes economically feasible due to gas recycle.
4. Gas recycle reduces scaling problems in the stripping tower as recycle gas will be
very much lower in CC>2 content than atmospheric air.
From Equation 8-4, it can be seen that there is no net requirement for acid, as in the acid
system described in Section 8.4.1.
8.5 Stripping Ponds
The South Tahoe system has been modified to reduce the impact of temperature and scaling
limitations encountered at the plant. ^ Basically, the modified process consists of three
steps (See Figure 8-9): (1) holding in high pH, surface agitated ponds, (2) stripping in a
modified, crossflow forced draft tower through water sprays installed in the tower, and (3)
breakpoint chlorination (see Sec. 9.5.5.1 for pilot plant results). This system was inspired by
observations in Israel of ammonia nitrogen losses from high pH holding ponds. 1°
8-17
-------
FIGURE 8-9
AMMONIA STRIPPING POND SYSTEM
AlR SPRAYING OF RECYCLED POND WATER
IN THE SECOND OF TWO PONDS
IN SECOND POND. TWO
RECYCLE PUMPS
34 mgd CAPACITY
414 TO 13» RECYCLES
CLARIFIED LIME
TREATED WASTEWATER.
pH- 11.0
TWO HIGH pH PONDS IN SERIES
7 TO 18 HOURS DETENTION TIME
FLOW VARIES. 2.5 TO 7.5 mgd
New High pH Flow Equalization Ponds
EXISTING CROSS FLOW AMMONIA
STRIPPING TOWER
Stripping Tower Modified with New Sprays
COZ OR
••
•
/ EXISTING 2 STAGE
/ RECARBONATION
^^/ BASIN
i
^^N-
.^-*— -
TO FILTERS AND
CARBON COLUMN
NEW BREAKPOINT EXISTING 1 MG
CHlORINATION BALLAST POND
CHAMBER FOR CHLORINE CONTACT
Breakpoint Chlorination
8-18
-------
Pilot tests at South Tahoe indicated that the release of ammonia from high pH ponds could
be accelerated by agitation of the pond surface. In the modified Tahoe system, the high pH
effluent from the lime clarification process flows to holding ponds. Holding pond detention
times of 7-18 hours are used in the modified South Tahoe plant. The pond contents are
agitated and recycled 4-13 times by pumping the pond contents through vertical spray
nozzles into the air above the ponds. The holding pond detention time and number of
recycles vary with plant flow with the time and cycles decreasing as plant flow increases. At
least 37 percent ammonia removal in the ponds is anticipated, even in cold weather
conditions. The pond contents are then sprayed into the forced draft tower. The packing
has been removed from the tower and the entire area of the tower is equipped with water
sprays. At least 42 percent removal of the ammonia in the pond effluent is anticipated,
based on pilot tests, from this added spray in cold weather that includes recycling of the
pond effluent through the tower to achieve 2-5 spraying cycles. The ammonia escaping this
process is removed by downstream breakpoint chlorination. It appears that stripping ponds
offer an approach that takes advantage of the low cost and simplicity of the stripping
process for removal of the bulk of the nitrogen, making breakpoint chlorination more
attractive for complete removal of ammonia.
8.6 Considerations in Process Selection
One of the great advantages of air stripping is its extreme simplicity. Water is merely
pumped to the top of the tower at a high pH, air is drawn through the fill, and the ammonia
is stripped from the water droplets. The only control required is the proper pH in the
influent water. This simplicity of operation enhances the reliability of the process. The
process costs are also significantly less than any alternate method of nitrogen removal,
assuming that the needed pH elevation occurs in conjunction with upstream phosphorus
removal.
The major engineering limitations on application of the process result from its sensitivity to
temperature variations and from potential scale accumulation on the tower packing. The
first of these limitations is that of temperature as discussed in Section 8.3.3. Ammonia
removals decrease as air and water temperatures decrease. Although increased air flows can
offset temperature effects to some degree, it is not practical to supply enough air to offset
major temperature drops. It has not been practical to operate stripping towers at ambient air
temperatures below 0 C. Of course, this is not a limitation in climates where freezing
temperatures do not occur or for plants where nitrogen removal is not required during cold
or freezing weather. Modifications (Section 8.4) of the ammonia stripping process are being
developed which may eliminate temperature limitations.
For applications in cold weather where a high degree of nitrogen removal is required, the
stripping process itself will generally not be adequate. It is usually not practical to heat the
large quantities of air required for the stripping process unless one is so fortunate as to have a
large source of waste heat in proximity to the stripping tower (see Section 9.5.5.2 for a
description of the use of waste heat from a desalting plant to heat stripping tower air while
8-19
-------
providing needed cooling of desalting plant water and wastes). The use of the stripping
process supplemented as needed in cold weather by breakpoint chlorination is a process
combination that may be attractive in some cases where efficient cold weather operation is
needed and cold weather conditions do not persist for prolonged portions'of the year.
A potentially serious problem is the formation and accumulation of calcium carbonate scale
on the tower packing. Designs should anticipate this problem and provide for easy access for
cleaning the packing as has been done in the Orange County tower (Sec. 9.5.5.2). It appears
that scaling is not as severe in a countercurrent tower as in a crossflow tower and that scale
does not adhere as tightly to the smooth, hard surface of plastic packing as it does to the
rough, soft wood surface. The more open "criss-cross" type of packing developed at Orange
County may also be more resistant to scale accumulation than the parallel packing
arrangement used at Tahoe. Experience to date indicates that if adequate provisions are
made in the design of stripping towers for access to packing, in many cases scale removal
can be accomplished by water sprays without the use of chemicals or mechanical means.
However, this is a factor that deserves special consideration and investigation at each
location, since the scaling characteristics of different wastewaters may differ markedly. In
light of the as-yet unpredictable nature of causes contributing to scaling of tower packing, it
would be prudent to conduct pilot tests for 3-6 months on the specific wastewater involved
with the specific tower configuration proposed.
8.7 References
1. Slechta, A. F., and G. L. Gulp, Water Reclamation Studies at the South Tahoe Public
Utility District. JWPCF, 39, No. 5, pp 787-814 (1967).
2. Roesler, J. F., Smith, R., and R. G. Eilers, Mathematical Simulation of Ammonia
Stripping Towers for Wastewater Treatment. U.S. Department of Interior — FWPCA,
Cincinnati, Ohio, January, 1970.
3. Roesler, J.F., Smith, R., and R.G. Eilers, Simulation of Ammonia Stripping From
Wastewater. JSED, Proc. ASCE, 97, No. SA 3, pp 269-286 (1971).
4. Perry, J.H., Chemical Engineering Handbook. McGraw Hill Book Co., New York, NY.
5. Miner, S., Preliminary Air Pollution Survey of Ammonia. U.S. Public Health Service,
Contract No. PH22-68-25, October, 1969.
6. Nitrogenous Compounds in the Environment. EPA Report SAB-73-001, December,
1973.
7. Pasquill, F., Atmospheric Diffusion. D. Nostrand Co., Ltd., London, 1962.
8. Gulp, R.L., and G.L. Gulp, Advanced Wastewater Treatment. Van Nostrand Reinhold,
New York, 1971.
8-20
-------
9. O'Farrell, T.P., Bishop, D. F., and A.F. Cassel, Nitrogen Removal by Ammonia
Stripping. EPA Report 670/2-73-040, September, 1973.
10. South Tahoe Public Utility District, Advanced Wastewater Treatment as Practiced at
South Tahoe. EPA Report 1701OELQ 08/71, August, 1971.
11. Wesner, G.M., and D.G. Argo, Report on Pilot Waste Water Reclamation Study. Orange
County Water District Report, July, 1973.
12. van Vuuren, L.R., personal communication, May 30, 1972.
13. Kepple, L.G., Ammonia Removal and Recovery Becomes Feasible. Water and Sewage
Works, p. 42, April, 1974.
14. Brown and Caldwell, Application for a. Research Contract for the Ammonia
Elimination System. Submitted to the Office of Research and Development, Water
Quality Office, EPA, March, 1971.
15. Gonzales, J.G., and R.L. Gulp. New Developments in Ammonia Stripping. Public Works,
104, No. 5, p 78 (1973) and No. 6, p 82 (1973).
16. Folkman, Y., and A.M. Wachs, Nitrogen Removal Through Ammonia Release From
Ponds. Proceedings, 6th Annual International Water Pollution Research Conference,
Tel Aviv, Israel, June 18-23, 1972.
17. Gulp, G.L., Physical-Chemical Techniques for Nitrogen Removal. Prepared for the EPA
Technology Transfer Program, July, 1974.
8-21
-------
CHAPTER 9
TOTAL SYSTEM DESIGN
9.1 Introduction
With a defined set of effluent quality objectives, the environmental engineer must develop a
cost-effective treatment system that is suited to the local situation. A variety of nitrogen
removal options are presented in this manual that may form a part of the total treatment
system. No one combination of processes is universally applicable.
The main thrust of this chapter is to present specific examples of treatment systems that
have been selected and implemented. Design concepts that have evolved to suit local
circumstances are emphasized.
9.2 Influence of Effluent Quality Objectives on Total System Design
The effluent quality obtainable from a treatment system has the most significant impact on
its selection or rejection. When only ammonia removal (or conversion) is required, as
opposed to total nitrogen removal, cost considerations would dictate selection of
nitrification over any other method of ammonia removal in most cases. Exceptions would
be low temperature ( < 5 C) situations or where toxicants which cannot be effectively
removed by a source control program are present in the wastewater. Other possible
exceptions are cases wherein influent nitrogen levels are so low that they can be
economically removed by breakpoint chlorination and where the carbon to nitrogen ratio is
such that all the nitrogen can be assimilated into biomass during biological oxidation of
organics.
In Section 2.5.5 the effectiveness of each type of process in removing the various nitrogen
species was summarized. Tables 9-1, 9-2 and 9-3 contain representative nitrogen data for
final effluents from systems incorporating nitrification-denitrification, ion exchange, and
breakpoint chlorination. Ammonia stripping has not been included, since it usually must be
backed up by breakpoint chlorination to achieve consistently low effluent nitrogen levels.
These systems are capable of producing average effluent nitrogen levels of 2.0 to 3.0 mg/1,
whether the nitrogen removing processes are biological or physical-chemical in nature.
Each nitrogen removal system should be assessed from the standpoint of its reliability in
meeting effluent objectives. Factors resulting in the failure of a system to achieve nitrogen
removal objectives are as follows:
1. Toxicity upsets
2. Overloading
9-1
-------
3. Design deficiencies
4. Poor operation
5. Mechanical failures
6. Changes in influent quality
TABLE 9-1
EFFLUENT NITROGEN CONCENTRATIONS IN TREATMENT
SYSTEMS INCORPORATING NITRIFICATION - DENITRIFICATION
Type and process sequence
Lime treatment of raw sewage.
nitrification,
denitrification
Primary treatment.
high rate activated
sludge, nitrification.
denltrlfication.
filtration0
Primary treatment.
roughing filters ,
nitrification.
denitrification ,
filtration
Location
CCCSD.Ca.
Manassas,
Va.
El Lago,
Texas
Ref.
1
2
3
Scale
mgd
(mVsec)
0.5
(0.022)
0.2
(0.0088)
0.3
(0.0132)
Period,
days
90
120
55
Average effluent nitrogen, mg/1
Organic-N
1.1
0.8
0.8
4
0.3
0.0
0.9
NO~-N
0.5
0.7
0.6
NO~-N
0.0
0.0
0.0
Total
N
1.9
1.5
2.3
CCCSD = Central Contra Costa Sanitary District
bSysteni 1, Fig. 9-1
CSystem 3B, Fig. 9-2
Coarse media
TABLE 9-2
EFFLUENT NITROGEN CONCENTRATIONS IN
TREATMENT SYSTEMS INCORPORATING ION EXCHANGE
Type and process sequence
Lime treatment of
raw wastewater,
two-stage recarbonation.
filtration, activated
carbon, ion exchange
Lime treatment of
raw wastewater,
recarbonation ,
filtration, activated
carbon, ion exchange
Location
Blue Plains,
D.C.
East Bay
Municipal
Utilities
District,
Ca.
Ref.
4
5
Scale,
mgd
(mVsec)
0.05
(0.0022)
Pilot
scale
Period
days
90
na
Average effluent nitrogen, mg/1
Organic-N
a
na
2.4
NH4-N
3.6b
0.5
NO'-N
na
d
NO~-N
na
d
Total
N
4.5°'
2.96
na = not available
Intermittent operator attention only
°System 2A, Fig. 9-1
Assumed negligible
Estimated
9-2
-------
TABLE 9-3
EFFLUENT NITROGEN CONCENTRATIONS IN TREATMENT SYSTEMS
INCORPORATING BREAKPOINT CHLORINATION
Type and process sequence
Lime treatment of
raw waste water,
two-stage recarbonation.
filtration, breakpoint
chlorination , activated
carbon"
Lime treatment of
raw wastewater.
filtration, activated
carbon, breakpoint chlorin-
ation, and dechlorlnation
by activated carbonb
Primary treatment.
oxidation ponds.
algae removal by
alum-flotation.
filtration, breakpoint
chlorlnatlon
Location
Blue Plains,
D.C.
Same,
digital
control
Owosso,
Michigan
Sunnyvale,
Ca.
Ref.
4
6
7
8
Scale
mgd
(mVsec)
0.05
(0.0022)
0.05
(0.0022)
0.02
(0.0009)
0.01
(0.0004)
Period,
days
120
9
11
2
Average effluent nitrogen, mg/1
Organlc-N
naa
na
0,58
2.6
NH+-N
na
na
1.42
0.2
NO~-N
na
na
c
0.4
NO~-N
na
na
c
0.0
Total
N
3.3
1.6
2.0d
3.2
na = not available from publication
System 2A, Figure 9-1
Assumed to be negligible
Estimated
Toxicity upsets affect only biological nitrogen removal processes. A degree of protection
against toxicity upset can be provided for the nitrification-denitrification system by
providing pretreatment processes as described in Section 4.5.3. Several case histories are
presented in this manual which show a very high stability for the biological processes, due in
part to pretreatment. There are classes of toxicants, such as nonbiodegradable solvents,
which are not effectively removed by pretreatment. Reliability under this circumstance is
dependent on source control. Under some circumstances, the reliability of the source
control program may be not effective enough to allow dependable nitrification.
The other factors listed affect both physical-chemical and biological systems to varying
degrees. Overloading can be defined as operation which exceeds design conditions.
Obviously both physical-chemical and biological processes can be expected to lose
effectiveness when overloaded.
Theoretically sound processes can fail to meet objectives due to design deficiencies,
mechanical breakdown or poor operation. This is true for both physical-chemical processes
and biological processes.
Some of the nitrogen removal systems are more sensitive to the form of nitrogen in the
influent than others. Generally, the physical-chemical processes are geared to a specific
9-3
-------
chemical form of nitrogen; for instance, urea cannot be removed by air stripping or
breakpoint chlorination. Biological systems have the inherent capability to transform a
multitude of nitrogenous compounds to ammonia for subsequent conversion to nitrate, a
form suitable for denitrification. Thus, if changes in the distribution of influent nitrogen
compounds occur with time, the biological processes may be more able to adapt to
treatment of the new compounds than the physical-chemical processes.
In sum, the issue of relative reliability of the various approaches is mixed, and it cannot be
claimed that some specific approach has a clear advantage over the others for general
application. When stringent regulations require enhanced reliability, it is relatively simple to
provide breakpoint chlorination for effluent polishing. Since breakpoint chlorination has no
effect on nitrate or nitrite, it cannot make up for deficiencies in the denitrification process.
However, in the nitrification-denitrification system, it is the nitrification step that is most
susceptible to upset, and the breakpoint process provides full backup for it.
9.3 Other Considerations in Process Selection
Costs of the alternative nitrogen removal systems are specific to each situation and
time-frame and generalizations about the alternatives are difficult to make. Long-run
operating costs are of interest, but the long-term prices of chemicals and energy are
particularly difficult to estimate.
Total dissolved solids (TDS) in the process effluent is sometimes a consideration. Biological
nitrification-denitrification results in little change in TDS, whereas both ion exchange and
breakpoint chlorination result in an increase in TDS.
Low liquid temperatures ( < 10 C) often favor physical-chemical systems because the
tankage requirements for biological nitrification-denitrification become very large. Biologi-
cal nitrification-denitrification becomes less cost-effective below 5 C.
Receiving water standards or effluent requirements may dictate intermittent nitrogen
removal. This requirement may favor breakpoint chlorination. An example is the nitrogen
removal facility for the Sacramento Regional Treatment Plant, described in Section 9.5.3.1.
In this case, breakpoint chlorination was chosen because its relatively low capital cost
avoided the higher fixed costs of the other alternatives.
9.4 Interrelationships with Phosphorus Removal
Phosphorus removal is the subject of another publication of the EPA Office of Technology
Transfer.9 However, experience indicates that the majority of treatment plants being
designed for nitrogen removal also have a requirement for phosphorus removal. As these two
requirements have very different influences on treatment plant design, some consideration
needs to be given to how nitrogen and phosphorus removal are interrelated in treatment
system design.
9-4
-------
9.4.1 Alternative Systems
Figures 9-1 and 9-2 are summaries of the five general approaches currently being considered
or implemented for cases where high degrees of phosphorus and nitrogen removal are
required. There are few exceptions to the listing; one exception concerns nitrogen removal
from oxidation pond effluent, but these approaches tend to be very case specific and are
difficult to generalize.
All flow diagrams in Figures 9-1 and 9-2 are capable of achieving low effluent levels of nitrogen
and phosphorus. These levels are taken as averaging 2-3 mg/1 of total nitrogen and 0.1 to 0.3
mg/1 of total phosphorus. Multipoint chemical addition and filtration are shown to achieve
low phosphorus levels in the final effluent. If lesser degrees of phosphorus removal are
required, then some of these steps may be eliminated. Also, in each case, it is assumed that
effluent BOD5 objective is on the order of 10-15 mg/1. If further organic reduction is
required, supplemental treatment is also required. Figures 9-1 and 9-2 also show variations
within the five flowsheets where substitute processes are possible. The flowsheets are general
process arrangements; for example, a block showing denitrification could mean either an
attached growth process or a suspended growth process with a sedimentation tank.
System No. 1 is an integrated chemical-biological system using lime or other metal salt in
the primary treatment stage to reduce phosphorus and organic loads followed by
nitrification and denitrification stages and filtration. By moving lime treatment to the
primary treatment stage, as opposed to tertiary applications, several advantages are gained.
First, moving lime treatment to the primary stage causes enough organic reduction in the
primary tanks to eliminate the need for a separate carbon removal step. Second, lime dose
can be adjusted to elevate the pH in the nitrification step to the optimum range for
nitrification as well as to compensate for any alkalinity depletion due to nitrification.
Lastly, protection for the nitrifiers against most toxic heavy metals is provided.
Systems 2A and 2B are the independent physical-chemical treatment sequences incorporat-
ing physical-chemical nitrogen removal. A coagulant such as lime, or a metal salt and
polymer, is used in the primary step for organics and phosphorus reduction. Activated
carbon is provided for further organics reduction. In System 2A, either breakpoint
chlorination or ion exchange is usual for nitrogen removal. In System 2B, ammonia stripping
is used for nitrogen removal. Filtration is placed ahead of carbon adsorption in both
variations of System 2; however, it may be placed after carbon adsorption in certain
instances. For considerations in arrangement of the adsorption component of physical-
chemical systems, the reader is referred to the Process Design Manual for Carbon
Adsorption, a publication of the EPA Office of Technology Transfer. 10
Another approach to integration of biological and physical-chemical treatment is provided
by Systems 3A and 3B. System 3A takes advantage of the favorable effect of alkalinity
depletion in nitrification on reducing lime dose in the chemical precipitation step as lime
dose is directly affected by alkalinity. ^ > 12 System 3B shows a slightly different way of
9-5
-------
FIGURE 9-1
ALTERNATE PROCESS SEQUENCING FOR SYSTEMS YIELDING COMBINED
NITROGEN AND PHOSPHORUS REMOVAL - SYSTEMS
WITH COAGULANT ADDITION TO PRIMARY SEDIMENTATION
SCREENED $ DEGRITTED
RAW WASTEWATER
LIME and or
METAL SALT
PRIMARY
SEDIMENTATION
C02(WITH LIME
r2 EFFLUENT)
NITRIFICATION
•C02 (WITH LIME
EFFLUENT)
FILTRATION
METHANOL
DENITRIFICATION
(LIME
EFFLUENT
ONLY)
AIR STRIPPING
METAL
SALT "
BREAKPOINT OR
ION EXCHANGE
•METAL
SALT
FILTRATION
(OPTIONAL)
DISINFECTION
I
•C02
FILTRATION
I
BREAKPOINT FOR
RESIDUAL AMMONIA
ACTIVATED CARBON
ADSORPTION
FINAL EFFLUENT
I
DISINFECTION
T
SYSTEM 1
INTEGRATED
CHEMICAL-BIOLOGICAL
TREATMENT
EXAMPLES I CONTRA COSTA
CANBERRA
FINAL EFFLUENT
SYSTEM 2A
SYSTEM 2B
INDEPENDENT
PHYSICAL -CHEMICAL
TREATMENT
EXAMPLE: ROSEMONT
9-6
-------
FIGURE 9-2
ALTERNATIVE PROCESS SEQUENCING FOR SYSTEMS YIELDING COMBINED
NITROGEN AND PHOSPHORUS REMOVAL - SYSTEMS WITH
COAGULANT ADDITION AFTER PRIMARY TREATMENT
CARBON
1 (eg ACTIV
LIJe. Moy be I
SALT 1 «t«P
I -
CHEMICAL V MITR
PRECIPITATION NITR
DENITRIFICATION DENITF
FIL
DISIN
SCREENED 1 DEGRITTEO
RAW WASTEWATER
1
PRIMARY
SEDIMENTATION
1
OXIDATION SALT COMBINED CARBON CARBON OXIDATION
.TED SLUDOE, "^ 'J^^KlVVo. <" »™"» *""»*
3ti only)
LIME and /or LIME and/or
METAL SALT . METAL SALT
CHEMICAL CHEMICAL
FICATION PRECIPITATION PRECIPITATION
— METHANOL METHANOL
* FOR RESIDUAL,
NITRATE
IFICATION *i DENITRIF
* fSALT
• ^ 1 («. TeBUATFs}
1 C02 (L,ME EFFLUENT ONLY) « COZ(*ITH
< METAL SALT 1 « METAL SALT
i°AT*NN) A'R STRIPPING FILTRATION
. — co2
1 METAL SALT
TRATION DISINFECTION FILTRATION ""EXCHANGE
«^S02 FOR
OECHLORINATION IF
< BREAKPOINT EFFLUENT
M«.L t^LUtN, BREAKPO.NT FOR DIS.NFECTION
FECTION RESIDUAL AMMONIA (IF NO BREAKPOINT)
I U S02 1
& 1 for Dechlorination 1
FINAL EFFLUENT FINAL EFFLUENT FINAL EFFLUENT
SYSTEM 3A SYSTEM 38 SYSTEM 4 SYSTEM 5A SYSTEM 5B
INTEGRATED
BIOLOGICAL-CHEMICAL
COMBINED SYSTEM BIOLOGICAL TREATMENT
WITH TERTIARY WITH TERTIARY PHOSPHOROUS AND
PHOSPHORUS AND PHYSICAL - CHEMICAL NITROGEN REMOVAL
RESIDUAL NITRATE
REMOVAL
EXAMPLES: EL LAOO
BLUE PLAINS
EXAMPLES: NONE
EXAMPLES: SOUTH LAKE TAHOE
ORANGE COUNTY
MONTGOMERY COUNTY
UPPER OCCOQUAN
-------
accomplishing the same objective. In this case metal salt is added to treatment units such as
the carbon oxidation step and the filtration step, avoiding the need for the separate
precipitation step of System 3A.
System 4 uses a combined carbon oxidation-nitrification-denitrification sequence. The
organic carbon in the primary effluent serves as the carbon source for the removal of the
bulk of the nitrogen in the influent wastewater. This effluent is polished with a tertiary
phosphorus removal step. Residual nitrates are removed in the wastewater filter. System 4
has lower operating costs than System 1 or 3 because of the elimination of the bulk of the
methanol costs incurred in the latter systems. On the other hand, the phosphorus removal
step must be placed after the biological treatment step in System 4 for two reasons. First,
precipitation of phosphorus in the primary step would reduce the organics in the primary
effluent to the point where denitrification could not be supported. Second, metal salt
addition to the biological step would add so many solids as to render unmanageable the
simultaneous operation of carbon oxidation, nitrification and denitrification.
System 5 was the first system implemented in the U.S. and was used at South Lake Tahoe
(in the System 5A configuration). It consists of conventional biological treatment followed
by tertiary steps for phosphorus removal and physical-chemical nitrogen removal. System
5A employs air stripping with polishing by breakpoint chlorination while System 5B uses
either breakpoint chlorination or ion exchange for nitrogen removal.
In conjunction with the system descriptions in Figures 9-1 and 9-2 are listed the case
examples presented in this chapter which generally fit the system description. In most cases
the examples do not precisely follow the system description because local requirements have
dictated lesser or greater degrees of treatment. However, the case examples are close enough
to be fitted into system categories.
9.4.2 Considerations in System Selection
Each of the systems outlined in Section 9.4.1 has its inherent advantages and disadvantages
that need to be considered by the treatment plant designer in each individual situation.
Some of these considerations are described in this section.
9.4.2.1 Phosphorus Removals Obtainable
Perhaps because of the long experience accumulated with System 5A (Figure 9-2) at South
Lake Tahoe and the more recent development of alternative systems, System 5A or SB were
thought to have had a decided advantage in terms of low phosphorus residual over any of
the integrated systems (System 1, 3A and 3B). It has been suggested that the integrated
system, even with proper coagulant dosage, is limited to reduction of effluent phosphorus
levels to 0.5 mg/1 total phosphorus and that when lower phosphorus levels are required,
System 5A or 5B should be employed, l^
9-8
-------
In actuality, the degree of phosphorus removal is not primarily affected by the system
selected but by the pattern of chemical addition, the nature and doses of the chemicals used
and the sophistication of process control. Regardless of the system selected, low effluent
phosphorus levels are made possible by multiple phosphorus removal steps; this may be
achieved by multipoint chemical addition or chemical addition in conjunction with other
phosphorus removal methods such as tertiary filtration.
An example is provided by the well documented operation of the South Lake Tahoe
plant. 13 A flowsheet for the plant as it existed in 1970 is presented in Figure 9-3.
Multipoint chemical addition was practiced with lime treatment of the secondary effluent
plus alum treatment at the tertiary filtration step. Phosphorus removal resulted at both these
steps. In addition, phosphorus uptake occurred in the activated sludge step and possibly
some removal occurred in the carbon adsorption step. The lime treatment alone reduces the
phosphorus level to about 0.6 mg/1 entering the filtration step. About 30 percent phosphorus
removal occurred in the filter without alum addition. With alum addition, the effluent
phosphorus level was reduced to 0.1 mg/1 in a special one month test. 13 The phosphorus
residuals fora one year period averaged 0.17 mg/1 in the plant effluent as shown in Table 9-4.
Alum dosage required to boost phosphorus removals by the filters from 30 to 90 percent was
only 10 mg/1.13
From the Tahoe example it can be seen that the key to obtaining a low phosphorus residual
is multiple removal steps. Multiple phosphorus removal steps have been included in all of the
treatment systems portrayed in Figures 9-1 and 9-2. For instance, the option of metal salt
addition to tertiary filters is available to all of the systems, not just System 5A or 5B, and
comparable performance can be expected in each application. System performance is given
in Table 9-4 for all systems except System 4. Favorable examples were chosen for each case
in Table 9-4 and other cases for each system could be found with higher effluent
phosphorus values. These examples are illustrative of what good design, operation, and
control can produce.
9.4.2.2 Impacts on Sludge Handling
In Systems 3A, 4, 5A and 5B, (Figure 9-2).-chemical precipitates can be kept separate from
organic sludges. In Systems 1, 2A, 2B and 3B, (Figures 9-1 and 9-2) chemical sludges are
combined with organic sludges. Separation of sludges allows the plant designers more
options in sludge handling. For instance, chemical sludges can be subjected to coagulant
recovery operations while processing of organic sludges can proceed without the hindering
effects of the inert chemical sludges.
It has also been suggested that tertiary phosphorus removal (System 5A or 5B) may have an
advantage in that lime recovery can be practiced.^ However, System 1, 2, 3A and 4 all
possess an advantage in common with 5A or 5B; namely, the ability to recover lime if it is
the chosen coagulant.
9-9
-------
FIGURE 9-3
SCHEMATIC FLOW DIAGRAM - SOUTH LAKE TAHOE, CALIFORNIA PLANT (1970)
SECONDARY
TREATMENT
PRIMARY
TREATMENT
CHEMICAL TREATMENT
AND
PHOSPHATE REMOVAL
MAJOR TYPES
OF TREATMENT
PROVIDED
NITROGEN
REMOVAL
(BIOLOGICAL
TREATMENT)
(SOLIDS
SEPARATION)
60 MG EMERGENCY
HOLDING POND
RECLAIMED
WATER ro
INDIAN
CREEK
RESERVOIR
PARSHALL FLUMES-
FLOW MEASUREMENT
AND DIVISION
BMMINUTORS
CHLORINE
APPLICATION
RETURN TO
SECONDARY BALLAST
POND AT PLANT
RECARBONATION
(STAND an
AMMONIA
STRIPPING
TOWER
LUTHER PISS .
BOC/STER PUMP
STATION
WASTE WATER
FLOW
THROUGH
PLANT
SECONDARY
CLARIFIERS
RE-
CARBONATION
BASIN
TERTIARY
PUMP
STATION
PLANT INFLUENT
FORCE MAINS
PRIMARY
SLUDGE PUMPS
SCCONDARYX
SLUDGE PUMPS
SLUDGE FLOW
DIVISION BOX
WASTE ACTIVATED
SLUOCC
SOLIDS
HANDLING
LIME AND
CARBON
RECLAMATION
BIOLOGICAL
SLUDGE
FURKACt
RCCAUWCO LIME
TORE-USE
JTEHILE ISM
TO DISPOSAL
REGENERATED
CARBON TO
RE-USE
-------
TABLE 9-4
EFFLUENT PHOSPHOROUS CONCENTRATION FROM
ALTERNATIVE SYSTEMS
Type and process sequence
System 1
Lime (with recalcination and
recycle), nitrification.
denitrification, without
filtration
System 2A
Lime with iron in second . .
settling tank, filtration.
ion exchange or break-
point, filtration
System 3B
Carbon oxldation-
nitrif ication-denitrif ication ,
filtration with alum addition
to carbon oxidation and
denitrification
System 5 A
Carbon oxidation, lime.
ammonia stripping',
recarbonation , filtration
with alum addition, carbon
ad sorption
Location
CCCSD, Ca.a
Blue Plains,
b.c. •
Manassas,
Va.
South Lake
Tahoe, Ca.
Ref.
14
. ,4
2
13
Scale,
mgd
(mVsec)
0.5 to
(0.022)
0.05 .
(0.0022)
0.02
(0.0009)
2.4
(0.11)
Period,
days
W
10b
480
120
365C
Average effluent phosphorus,
mg/1
Total P
0.04
0.14
0.3
/
/
' /
0.17
PO"-P
0.01
na
/x
X
./ na
na
Average effluent
total suspended
solids, mg/1
3.0
4.0
0
0
aCCCSD = Central Contra Costa Sanitary District
bAugust 1 to 10, 1973
C1970; representative year, the plant has been operational since 1968.
-------
9.4.2.3 Reliability
Factors affecting the reliability of nitrogen removal processes have already been described in
Section 9.2. Most of these same factors affect phosphorus removal and will not be repeated
here.
The longest record of reliable experience in obtaining low phosphorus residuals is at the
South Lake Tahoe Plant, operational since 1968 (Table 9-4). Low values have also been
obtained consistently for long periods with physical-chemical systems (System 2A). In the
latter case, iron has been used in a second stage settler after lime treatment and
recarbonation, achieving further phosphorus removals. Less experience is available with
Systems 1, 3A or 3B, but the limited testing to date indicates very low phosphorus residuals
can be consistently obtained.
9.4.2.4 Flexibility of Operation in Multipurpose Treatment Units
The systems portrayed in Figures 9-1 and 9-2 incorporate varying degrees of integration of
process function into the various treatment units. Systems 4, 5A and 5B represent extremes
in terms of combining functions; in System 5A or 5B the tendency is for individual steps to
perform a minimum of purposes while in System 4 many functions are accomplished in
parallel in each step.
The argument can be made that the level of integration in a plant can affect its flexibility of
operation in terms of adjustability of the system to meet varying loads or in terms of
providing redundancy for possible process failures. The degree of integration possible is best
studied by examination of pilot or full-test results. There have been full-scale tests that have
shown that nothing has been lost in terms of flexibility or performance with a degree of
process integration. Examples are provided by the Manassas and CCCSD experience
(Systems 1 and 3B) described in Section 5.2.4. No long-term test results are available for
System 4, which is unfortunate, since a high degree of integration of function is provided in
the combined carbon oxidation-nitrification-denitrification step.
9.4.2.5 Cost
Cost is an essential factor in process selection. It is widely recognized that the integrated
approaches hold a potential of cost savings over the biological-tertiary approach (System
5) 1,12,15 jhe reality of this cost saving will be determined in individual situations by local
factors and must be specifically evaluated in each case by cost-effective analyses of the
alternative systems.
9.5 Case Examples
Fourteen case examples of nitrogen control are presented, each showing how the various
nitrogen removal systems described in this chapter have been applied. These include: four
9-12
-------
examples of nitrification for ammonia reduction, four examples of nitrification-
denitrification for nitrogen removal, and two examples each of breakpoint chlorination, ion
exchange and air stripping for nitrogen removal.
9.5.1 Case Examples of Nitrification for Ammonia Reduction
Four examples of how biological nitrification has been implemented are presented in the
following discussion. The Jackson, Michigan plant design was oriented to reducing the
nitrogenous oxygen demand (NOD) of the plant effluent in the receiving waters. The designs
of the Valley Community Services District plant and the City of Livermore's plant were
oriented to satisfying the very low coliform requirements set by the State of California. In
these cases, ammonia reduction allowed efficient disinfection after breakpoint chlorination.
In the design of another California plant, operated by the San Pablo Sanitary District,
nitrification was included so that effluent toxicity requirements could be met.
Other case examples of nitrification were presented previously in Section 4.3.8. These
included the Whittier Narrows Reclamation Plant in California which is oriented to
groundwater recharge and the Flint, Michigan plant, designed for NOD removal.
9.5.1.1 Jackson, Michigan
The City of Jackson is operating a 17 mgd (0.74 m-^/sec) activated sludge plant that is
designed to nitrify year-round. Nitrification is provided for removal of nitrogenous oxygen
demand so that the receiving water, the Grand River, can be maintained at dissolved oxygen
levels of about 4.0 mg/1. Since the implementation of full nitrification at this plant, this
requirement has been consistently met in the zone of influence of the plant's discharge.
The City of Jackson, Michigan, with an equivalent population of 60,000 has been served by
a conventional activated sludge treatment plant since 1936.16 The current upgrading of the
plant was completed in 1973 and resulted in the plant depicted in Figure 9-4 with the design
data shown in Table 9-5. During this upgrading, the following facilities were added: (1) new
primary effluent pumps, (2) a stormwater retention basin, (3) stormwater pumps for filling
the retention basin, (4) additions to the aeration tanks, (5) additions to the secondary
sedimentation tanks, (6) new return activated sludge pumps, (7) three new blowers and a
blower building for the secondary treatment additions, and (8) a new plant electrical and
control system.17 This work was bid in December, 1970 and totaled $3,200,000 including
legal, engineering and contingency costs. Operation and maintenance costs for the entire
plant for fiscal year 1973/1974 totaled $464,159 for 5255 million gallons treated or
$88/milgal.18
Several features have been incorporated into this plant that have been stressed in this
manual. First, the activated sludge system is operated in the conventional or plug flow
manner to gain highest efficiency of nitrification even at the coldest temperature conditions
(as low as 8 C). Coarse bubble aeration is utilized. Another feature of the plant is the
9-13
-------
TABLE 9-5
DESIGN DATA
JACKSON, MICHIGAN WASTEWATER TREATMENT PLANT
Average dry weather flow (ADWF)
Peak dry weather flow (PDWF)
Peak wet weather flow (PWWF)
Raw wastewater quality at ADWF
BOD5
SS
17 mgd (0.74 m3/sec)
22 mgd (0.96 mVsec)
30 mgd (1.31 m3/sec)
145 mg/1
200 mg/1
Primary sedimentation tanks
Number
Length
Width
Depth
Overflow rate at ADWF
Detention time at ADWF
Retention basin
Volume
Air blowers
Number
Discharge pressure
Capacity - total
Aeration tanks
Old tanks
Number tanks
Passes per tank
Width
Length/pass
Depth
Volume (4 tanks)
New tanks
Number of tanks
Passes per tank
Width/pass
Length/pass '
Depth
Volume (2 tanks)
Detention time at ADWFa
Final sedimentation tanks
Old tanks
Number
Diameter
Sldewater depth
60 ft (18.3 m)
24 ft (7.3 m)
11 ft (3.4 m)
,970 gpd/sf (80.2 m3/m2/day)
1 hr
12 mil gal (45,430 m3)
6.5 psig (0.46 kgf/cm2)
33,000 cfm (940 m3/min)
4
1
25.5 ft (7.8 m)
240 ft (73.2 m)
14.5 ft (4.4 m)
355,000 cu ft (10,000 m3)
2
2
25 ft (7.6 m)
150 ft (45.7 m)
14.5 ft (4.4 m)
217,500 cu ft (6,159 m3)
t
6.0 hr
70 ft (21.3 m)
11 ft (3.4 m)
9-14
-------
TABLE 9-5
DESIGN DATA
JACKSON, MICHIGAN WASTEWATER TREATMENT PLANT
(CONTINUED)
New tanks
Number
Diameter
Sidewater. depth
Overflow rate at ADWF a
at PDWF a
Detention time at ADWFa
Chlorine contact chamber
Number of passes
Detention time at ADWF
Sludge digestion
Number digesters
Volume
Sludge drying beds
Surface area, sf
80 ft (24.4 m)
12 ft (3.7 m)
670 gpd/sf (27.3 m3/m2/day)
865 gpd/sf (35.2 m3/m2/day)
3.0 hr
4
47 min
297,410 cu ft (8,422 m3)
328,000 sf (30,489 m2)
Including both old and new tanks
FIGURE 9A
JACKSON, MICHIGAN WASTEWATER TREATMENT PLANT FLOW DIAGRAM
WASTE ACTIVATED SLUDGE
REENEO
tw
kSTEWATER
GRIT
REMOVAL
DRYING BEOS
1
'
SUPER-
NATANT
RETURN
r PRIMARY
RAW ond
WASTE
SLUDGE
SLUDGE \
DIGESTION]
^^
\
1
RETURN SLUDGE
/^ ^\
AERATION- /SECONDARY \ CHLORINE
TANKS V TANKS / EFFLUENT
\ / ' " TO
\ ' RECEIVING
WATER
LINED
RETENTION
BASIN
9-15
-------
primary effluent retention basin. Rather than designing the secondary facilities for the full
wet weather flows, flows above about 22 mgd are pumped to the retention basin. This flow
is brought back by gravity to the secondary facilities when storm flows subside. It is also
anticipated that the retention basin will serve as flow equalization storage during dry
weather, once plant flows reach design capacity. This will be done to prevent ammonia
bleedthrough during peak flow periods. 1 ^
The retention basin is lined, but has no provision for mixing. Actual operation indicates very
satisfactory performance without odor development during storage. Except for completely
draining the basin, cleaning is limited to once per year. ^
Performance of the plant has been exceptionally stable and was previously summarized in
Section 4.3.8.3. It should be noted that the plant is not yet being operated at its design
flow.
9.5.1.2 Valley Community Services District, California
The Valley Community Services District (VCSD) Wastewater Treatment Plant at Dublin,
California, is treating an average daily flow of 3.7 mgd (0.16 m^/sec) from a largely
residential service' area. The original plant, consisting of raw wastewater screening, grit
removal, primary sedimentation, activated sludge aeration, secondary sedimentation,
digestion and sludge lagooning, was constructed in 1960 for approximately $1.5 million. In
1972, an additional primary sedimentation tank and appurtenances, an additional secondary
clarifier, a digester, dual media filters, and chlorination and dechlorination facilities were
constructed for about $3 million. These facilities raised the average dry weather flow
capacity of the plant to 4 mgd (0.17 m^/sec). The design data for the plant are shown in
Table 9-6.19
Waste discharge requirements mandate effluent filtration, as the State of California requires
that any effluent that may be used for water contact sports must be coagulated and filtered.
Requirements also dictated that the median coliform content must not exceed an MPN of
2.2 per 100 ml. This plant incorporates nitrification for ammonia reduction so that residual
ammonia can be economically breakpointed. This allows disinfection to proceed with a free
chlorine residual so that the stringent bacteriological requirements may be met.
A flow diagram of the existing facilities (Fig. 9-5) shows a holding basin which is routinely
used for flow equalization after primary treatment. Figure 9-6 is a photo of the holding
basin. The holding basin is asphalt-lined and is equipped with a sprinkler system for washing
out accumulated solids when the basin is drained. The sprinkler system was added by plant
staff and was found to be very effective for odor control. The basin is emptied daily. A
peak-to-average flow ratio of 3.4:1 is equalized in the holding basin to maintain stable
biological treatment conditions. It was found prior to the 1972 additions to the plant that
operation without flow equalization resulted in ammonia bleedthrough and eventual
complete loss of nitrifying capability. The aeration tank is generally operated with the first
9-16
-------
TABLE 9-6
VCSD WASTEWATER TREATMENT PLANT DESIGN DATA
Population
Average dry weather flow (ADWF)
Peak dry weather flow (PDWF)
Peak wet weather flow (PWWF)
Raw wastewater quality at ADWF
BOD5
Suspended solids
Primary treatment
Preaeratlon and grit removal tanks
Number
Detention time at ADWF
Primary sedimentation tanks
New tanks
Number
Length
Width
Old tanks
Number
Length
Width
Average water depth
Detention time at ADWFa
Overflow rate at ADWF a
Assumed removals
BOD5
Suspended solids
Activated sludge
Aeration tanks
Number
Passes per tank
Width each pass
Length each pass
Average water depth
Detention time, based on ADWF
BODs loading
Aeration blowers
Number
Capacity per blower
Discharge pressure
Secondary sedimentation tank
Old tanks
Number
Diameter
Sidewater depth
New tanks
Number
Diameter
Sidewater depth
Detention time, based on ADWF8
Overflow rate at ADWF a
Dual media filters
Number
Area each filter
Filtration rate at PWWF
Anthracite media
Depth
Effective size
48,000
4 mgd (0.17 m3/sec)
8 mgd (0.33 mVsec)
12 mgd (0.50 m3/sec)
330 mg/1
330 mg/1
2
24 min
100 ft (30.5 m)
19 ft (5.8 m)
110 ft (33.5 m)
19 ft (5.8 m)
9 ft (2 .74 m)
1.7 hr
960 gpd/sf (38.9 m3/m2/day)
40 percent
70 percent
1
2
30 ft (9.2 m)
210 ft (64.0 m)
15 ft (4.6 m)
8.5 hrs
35 lb/1000 cf/day
2,900 cfm(82 m3/min)
7.5 psig (0.53 kgf/cm2)
65 ft (19.8 m)
9 ft (2.7 m)
1
90 ft (27.4 m)
14 ft (4.3 m)
5.3 hrs
410 gpd/sf (16.6 m3/m2/day)
728 sf (67.66 m2)
3.8 gpm/sf (1.36 l/m2/sec)
36 inches (0.91 m)
2.4 - 4.8 mm
9-17
-------
TABLE 9-6
VCSD WASTEWATER TREATMENT PLANT DESIGN DATA
(CONTINUED)
Sand media
Depth
Effective size
Pea gravel size
Depth
Backwash water rate (max.)
Chlorine contact tanks
Number
Volume
Detention time @ ADWF
Sludge digestion
Digester loading
Primary solids
Biological solids
Total solids
Digesters
Number
Diameter
Sldewater depth
Total volume
Sludge disposal
Sludge lagoons
Number
Volume
18 inches (0.45 m)
0.8 - 1.0 mm
8 Inches (203 mm)
20 gpm/sf (13.6 ]/m2/sec)
22,500 cu ft (637 m )
1.0 hr
7,700 Ib/day (3,500 kg/day)
3,800 Ib/day (1,730 kg/day)
11,500 Ib/day (5,230 kg/day)
55 ft (16.8 m)
33 ft (10.1 m)
157,POO cu ft (4,500 rn )
350,000 cu ft (9,.920 m )
Including both old and new tanks
half of the first pass being used for reaeration of the return sludge, and with the primary
effluent being step fed in increments to the remainder of the first pass. Fine bubble aeration
is employed.
The VCSD. plant has consistently nitrified the influent ammonia as shown by the
performance data in Table 9-7.20 j^g nitrogen figures are from once monthly 24-hour
composite samples, while the BOD and suspended solids data is the average of daily
composite samples. For the first ten months of 1974, the VCSD waste water treatment plant
averaged 98.6 percent BOD removal and 99.3 suspended solids removal. The ammonia
nitrogen concentration in the effluent was typically less than 1 mg/1 and has been at this
level since August; 1 973, apart from a notable process upset caused by an industrial spill in
March, 1974. The nitrate-nitrogen concentration in the effluent has generally been around
24 mg/1 and this is about 99 percent of the nitrogen in the effluent. For several months
before August, 1973, the aeration capability was limited to two on-line blowers which were
not enough to sustain complete nitrification. The data given in Table 9-8 shows the change
in process performance after mechanical difficulties were overcome and a third blower was
started up. 21 Before the aeration capacity was increased, the average ammonia nitrogen
concentration in the effluent composite samples was 3.9 mg/1 with the nitrate nitrogen level
averaging 13.9 mg/1. A dissolved oxygen level of 2 to 4 mg/1 is now maintained in the last
portion of the aeration tank, whereas before August, 1973, the level was often- less than 2
mg/1.
9-18
-------
FIGURE 9-5
VALLEY COMMUNITY SERVICES DISTRICT (CALIF.)
WASTEWATER TREATMENT PLANT FLOW DIAGRAM
RAW
WASTEWATEF
TO
ALAMO CANAL
9-19
-------
TABLE 9-7
NITRIFICATION PERFORMANCE AT VALLEY COMMUNITY SERVICES DISTRICT
WASTEWATER TREATMENT PLANT, CALIFORNIA
Month
1974
January
February
March
April
May
Tune
July
August
September
October
Raw
waste-
water
flow,
mgd
(m /sec)
4.09
(0.179)
3.64
(0.159)
3.99
(0.175)
4.12
(0.181)
3.64
(0.159)
3.76
(0.165)
3.49
(0.153)
3.41
(0.149)
3.56
(0.156)
3.55
(0.156)
Mixed
liquor
recycle
ratio
0.49
0.41
0.46
0.42
0.42
na
0.55
0.43
0.45
0.47
Temp.
C
nab
na
16
19
22
25
28
29
27
26
MLSS,
mg/1
na
5,512
6,130
1,533
4,759
7,696
4,811
5,132
4,771
4,872
Sludge
volume
Index
119
119
126
73
81
82
73
77
85
103
Solids
retention
time,
days
10.5
11.9
15.2
8.8
9.6
10.9
11
10.9
8.1
8.3
BOD,., mg/1
Primary
effluent
na
231
167
140
147
142
106
116
164
183
Secondary
effluent
na
10.0
21.0
14.7
16.0
8.8
6.5
8.0
7.7
10.0
Final
effluent
2.5
5.6
3.7
3.8
5.6
3.0
3.0
2.5
2.0
1.6
Percent
removal
na
98.6
98.5
98.8
96.0
98.9
98.4
98.4
99.4
99.4
Suspended solids, mg/1
Primary
effluent
91
87
99
86
88
84
90
84
93
91.7
Secondary
effluent
19.2
9.6
8.2
9.2
9.0
10.3
9.4
12.2
15.7
10.6
Final
effluent
2.5
1.3
1.7
1.9
2.8
1.6
2.5
l.S
1.4
1.3
Percent
removal
99.0
99.5
99.3
99.1
98.7
99.4
99.0
99.4
99.6
99.7
Nitrogen,8 mg/1
Effluent
NH4-N
0.23
2.4
16
0.17
0.06
0.84
0.06
0.22
0.78
0.11
Effluent
NO~-N
17.0
21.8
6.8
24.4
26.6
24.9
23.1
21.9
28.9
21.9
Results of one 24-hour composite sample per month.
na = not available.
-------
FIGURE 9-6
HOLDING BASIN AT THE VALLEY COMMUNITY SERVICES
DISTRICT (CALIFORNIA) WASTEWATER TREATMENT PLANT
TABLE 9-8
NITROGEN ANALYSES ON 24 HOUR COMPOSITE EFFLUENT SAMPLES AT
THE VALLEY COMMUNITY SERVICES DISTRICT TREATMENT PLANT
Date
sampled,
1973
June
July3
August
September*3
October13
November*3
December'3
Nitrate nitrogen
as N, mg/1
9.9
16.5
9.5
14.0
26.7
26.0
27.1
Ammonia nitrogen
as N, mg/1
6.3
2.0
4.7
0.39
<0.06
0.30
<0.06
aTwo blowers on-line.
Three blowers on-line.
9-21
-------
The VCSD staff feels that this type of nitrification system is particularly subject to upset
due to toxicants in the primary effluent. One recurring loss of nitrifying ability was
ultimately traced to the periodic discharge of a solvent, trichlorethylene.22 Once the
industrial discharger was located and the spills ceased, the problem disappeared.
Operation and maintenance costs have averaged $350 per million gallons since the 1972
additions.
9.5.1.3 Livermore, California
Since 1967, the City of Livermore has operated the 5 mgd (0.22 m^/sec) Water Reclamation
Plant whose flow diagram is shown in Figure 9-7. Before 1967, the original plant included
primary treatment, trickling filters, secondary sedimentation, polishing treatment with
oxidation ponds and sludge digestion. During the 1967 enlargement, existing structures were
rearranged in the flowsheet and additional facilities were added.23 After the enlargement,
the plant consisted of preliminary treatment and primary sedimentation, roughing filters,
activated sludge for nitrification, and chlorination. The existing oxidation ponds were
converted to emergency holding basins. Sludge disposal is by digestion with digested sludge
being applied to sludge lagoons. Drying beds are used intermittently.
The plant layout was oriented to meeting discharge requirements mandating an effluent that
contained no more than 20 mg/1 of BOD5, 20 mg/1 of SS, and a five-day-median total
coliform of 5 MPN per 100 ml. The low bacteriological requirements dictated that the plant
FIGURE 9-7
CITY OF LIVERMORE WATER RECLAMATION PLANT (CALIF.) FLOW DIAGRAM
9-22
-------
be designed to dependably nitrify on a year-round basis. A free chlorine residual was
thought to be required for disinfecting to such low levels and that the ammonia level must
be minimized in the secondary effluent so that disinfection through breakpoint chlorination
could be economically practiced.
The trickling filters were retained in the plant layout to act as pretreatment for the
nitrification step, to reduce organic loads and to moderate any shock loads that might upset
nitrification. The sloughed solids from the roughing filter pass directly to the aeration tanks
without any intermediate clarification.
Plant design data are shown in Table 9-9 and operating data for the year 1971 in Table
9-10.23,24 The plant has consistently met effluent requirements and demonstrated stable
year-round operation. Wastewater temperatures are favorable for nitrification, with a range
of 15 to 24 C. Even when ammonia breakthrough has occurred in the secondary effluent, it
has been effectively reduced by breakpoint chlorination. Recently, the plant has not had
sufficient aeration capability to completely nitrify during peak nitrogen loan conditions, but
this condition is being rectified in plant modifications currently underway. Coarse bubble
aeration is used in the plant. A photograph of the aeration tank is shown in Figure 9-8. The
activated sludge exhibits very good settling properties with sludge volume index (SVI)
measurements consistently below 100 ml/g.
On the few occasions when plant upsets occur, making it likely that bacteriological
requirements will not be met, the effluent is directed to the emergency holding basins. The
chlorine contact tank has also served as a supplementary settling tank and yielded further
reductions of BOD5 and SS. The chlorine contact tank must be occasionally drained to
remove accumulated solids.
Twenty to twenty-five percent of the effluent is used for irrigation purposes at a nearby golf
course and on agricultural land. Effluent is also used at the golf course to fill several small
lakes.
The initial plant had a construction cost of $900,000 (1957 dollars), which included land
purchase, while the cost of the 1967 plant expansion was $1,300,000 (1968 dollars).
Operational expenditures for 1971 were approximately $275,000 which, when expressed on
a unit basis, is $224 per million gallons treated.
9.5.1.4 San Pablo Sanitary District, California
The San Pablo Sanitary District, California, operates a 12.5 mgd (0.55 m^/sec) wastewater
treatment plant designed for year-round complete nitrification. The original plant consisted
of a primary treatment plant with effluent chlorination and digestion for solids processing.
Additions completed in 1972 resulted in the plant flow diagram shown in Figure 9-9 and
included additional treatment facilities, a new roughing trickling filter, new aeration-
nitrification tanks, new secondary clarifiers, an additional chlorine contact tank, new
9-23
-------
TABLE 9-9
DESIGN DATA - LIVERMORE WATER RECLAMATION PLANT
Population
Average dry weather flow (ADWF)
Peak dry weather flow (PDWF)
Peak wet weather flow (PWWF)
Raw wastewater loadings
BOD5
Suspended solids
Primary treatment
Preaeratlon and grit removal tanks
Number
Detention time at ADWF
Primary sedimentation tanks
Number
Length
Width
Average water depth
Detention time at ADWF
Overflow rate at ADWF
Emergency holding basin
Volume
Roughing filters
Number
Diameter
Depth of media
Total volume of media
Media type
Media size
Reclrculation ratio at ADWF
BODs load
Air blowers
Number
Discharge pressure
Capacity - total
Aeration-nitrification tank
Number
Passes per tank
Length/pass
Width/pass
Depth
Detention time at ADWF
BODs load
Secondary sedimentation tank
Number
Diameter
Sidewater depth
Overflow rate at ADWF
Chlorine contact tank (Breakpoint chlorination)
Number
Passes per tank
Detention time at ADWF
Anaerobic digestion
Number
Volume
Loading, total solids
Volatile matter, percent
Sludge disposal
Digested sludge lagoons
Number
Total volume
Sludge drying beds
Number
Total area
62,500
5 mgd (0.22 mVsec)
10 mgd (0.44 m3/sec)
18 mgd (0.79 m3/sec)
12,500 Ib/day (5,670 kg/day)
12,500 Ib/day (5,670 kg/day)
2
36 mln
124 ft (37.8 m)
19 ft (5.8 m)
9 ft (2 .7 m)
1.5 hr
1,050 gpd/sf (42.8 m3/m2/day)
31 mil gal (117,000 m3)
110 ft (33.5 m)
4.25 ft (1.3 m)
80,152 cf (2,270 m3)
Rock
2 to 4 in.
3.0 to 1.0
100 Ib BODs/1,000 cf/day (1.62 kg/m3/day)
3
7.5 psi (0.53 kgf/cm2)
6,000 cfm (170 m3/min)
1
2
160 ft (48.8 m)
30 ft (9.2 m)
15 ft (4.6 m)
5.2 hr
28 Ib BOD5/1,000 cf/day (0.45 kg/m3/day)
1
90 ft (27.4 m)
12 ft (3.7 m)
767 gpd/sf (31.2 m3/m2/day)
1
2
1 hr
27,500 cf (779 m3)
11,800 Ib/day (5,350 kg/day)
75
320,000 cf (9,060 m3)
22,400 sf (2,080 m2)
9-24
-------
dissolved air flotation thickeners, and two new digesters. This flowsheet is very similar to
that used at the Livermore plant, described in Section 9.5.1.3, except that plastic media is
used in the roughing filter in place of rock media. Design data for the plant are shown in
Table 9-11.25'26
FIGURE 9-8
AERATION TANK AT THE LIVERMORE WATER RECLAMATION PLANT
(CALIFORNIA) WITH ROUGHING TRICKLING FILTERS IN BACKGROUND
9-25
-------
TABLE 9-10
NITRIFICATION PERFORMANCE AT
THE LIVERMORE WATER RECLAMATION PLANT
Month
1971
January
February
March
April
May
Jane
July
August
September
October
November
December
Flow,
mgd
(mVsec)
3.3
(0.14)
3.3
(0. 14)
3.2
(0.14)
3.2
(0.15)
3.4
(0.15)
3.5
(0.15)
3.3
(0.14)
3.4
(0.15)
3.6
(0.18)
3.5
(0.15)
3.4
(0.15)
3.3
(0.14)
Recycle*
ratio
0.35
0.34
0.35
0.34
0.35
0.38
0.44
0.43
0.44
0.42
0.42
0.44
MLSS,
mg/1
1803
1756
1702
1748
1743
2112
2189
2160
2123
2157
2275
2316
SVI.
ml/g
99
66
79
82
91
79
78
79
74
64
58
52
9c.
days
4.3
4.7
4.2
4.3
5.0
5.4
7.7
7.2
8.0
9.6
5.8
5.0
HT,
hr
7.8
7.8
8.1
8.1
7.6
7.4
7.8
7.6
7.2
7.4
7.6
7.8
Air use
MCF/day
(1/eec)
6.4
(2100)
6.4
(2100)
5.9
(1900)
8.0
(2100)
6.3
(2100)
6.8
(2200)
6.9
(2300)
7.2
(2400)
7.3
(2400)
7.3
(2400)
7.1
(2300)
6.5
(2100)
Primary effluent
BOD«.
mg/r
140
128
156
125
125
110
118
108
114
106
146
147
ss.
mg/1
80
85
84
74
98
81
71
81
60
75
63
94
Roughing filter effluent
BOD*,
mg/T
129
79
121
107
112
74
81
46
82
66
115
104
SS.
mg/1
137
122
128
96
118
110
102
118
115
74
89
107
NHd-N
mg/r
33.5
41.0
41.9
33.3
31.0
27.1
23.3
25.2
22.5
32.6
35.6
44.4
Secondary effluent
BOD.,
mg/1
8.9
11.8
16.9
8.7
11.6
6.6
8.1
5.2
5.4
10.6
11.3
12.6
SS,
mg/1
20
17
24
15
19
24
19
31
19
18
31
27
NHi-N,
mg/1
0.86
1.07
6.73
0.94
0.76
0.48
0.88
0.88
1.75
1.03
0.46
4.32
NO£-N,
mg/1
0.07
0.32
0.78
0.02
0.05
0.02
0.04
0.12
0.17
0.03
0.10
0.13
NC£-N,
mg/1
17.8
17.6
19.3
18.8
20.2
16.5
18.3
19.4
18.2
18.2
18.9
16.6
Final effluent
BODg.
mg/1
9.3
6.3
5.2
3.1
8.5
6.8
7.3
8.1
6.1
7.8
12.4
7.0
SS.
mg/1
19
13
16
8
12
12
9
17
11
15
12
17
NHj-N.
mg/1
<0.1
<0.1
1.3
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
0.4
Organic N,
mg/1
2.0
3.1
1.9
2.0
<0.1
7.8
1.0
1.3
0.4
1.7
4.0
1.3
Cotlform
MPNper
100ml
1.5
0.6
3.4
7.0
3.0
3.8
0.4
0.5
3.3
1.9
3.4
1.7
o\
"Return activated sludge
-------
Current discharge requirements are essentially those defined by EPA for municipal
secondary treatment plants with additional requirements set on effluent toxicity.26 \n
addition, the effluent pH is restricted to the range of 6.5 to 8.5. The acid production from
nitrification and subsequent chlorination normally forces the effluent pH below 6.5. This
has necessitated the addition of caustic to the final effluent to raise the pH to or above 6.5.
Toxicity requirements state that fish bioassays must be run on the undiluted effluent and
that 90 percent of a series of 10 consecutive tests must show 70 percent fish survival for
96 hr. Experience at this plant has indicated that the requirements cannot be met without
removal of ammonia through nitrification.26
A primary design consideration in laying out the plant for nitrification was the presence of a
significant volume fraction (11 to 13 percent) of potentially toxic industrial wastes in the
influent wastewater. Tank truck washing residues and the waste from a manufacturer of
organic peroxide and phenol formaldehyde are the major industrial waste sources. The
roughing filter is used in the treatment plant to protect the nitrifying organisms from
influent wastewater toxicity. Toxic dumps have caused severe sloughing and loss of growth
on the media in the roughing filter, but nitrification remained unaffected.
FIGURE 9-9
SAN PABLO SANITARY DISTRICT TREATMENT PLANT
(CALIFORNIA) FLOW DIAGRAM
CAUSTIC
ADDITION
SU8NATANT /DISSOLVED
TO HEADWORKS / AIR 1_ WASTE ACTIVATED SLUDGE
FLOTATION
THICKEN WO
m SLUDGE TO DRYING BEOS
^^ CENTRIFUGES (FUTURE)
9-27
-------
TABLE 9-11
DESIGN DATA, SAN PABLO SANITARY DISTRICT TREATMENT PLANT
Average dry weather flow (ADWF)
Peak wet weather flow (PWWF)
Raw wastewater quality
BODs
SS
Primary sedimentation tanks
New (Dry and wet weather use)
Number
Diameter
Sidewater depth
Detention time (ADWF)
Overflow rate (ADWF)
Existing (Wet weather use only)
Number
Diameter
Sidewater depth
Roughing filter
Number
Diameter
Media type
Depth of media
Volume of media
Media specific surface
Recirculation ratio
BOD5 load
Air blowers
Number
Discharge pressure
Capacity - total
Aeration-nitrification tanks
Number
Passes/tank
Length/pass
Width/pass
Depth
Volume (total)
Detention time (ADWF)
Secondary sedimentation tanks
Number
Length
Width
Depth
Overflow rate at ADWF
at PDWF
Detention time at ADWF
Chlorine contact chambers
Number
Passes/tank
Length/pass
Width/pass
Depth
Detention ADWF
Caustic addition (Na(OH))
Dose
Waste activated sludge thickening
Number
Diameter
Sidewater depth
Solids loading
Anaerobic Digestion
Number
Volume - total
Sludge drying bed
Surface area
12.5 mgd (0.55m3/sec)
30 mgd (1.32m3/sec)
340 mg/1
300 mg/1
70 ft (21.3m)
10 ft (3.1m)
2 hr
1624 gpd/sf(66/m3/m2/day)
1
100 ft (30.5m)
10 ft (3.1m)
52 ft (15.6 m)
Plastic corrugated sheet modules
18 ft (5.5 m)
38,200 cf (1080 m3)
29 sf/cu ft (95 m2/m3)
2.4 to 1.0
350 lb/1000 cf/day (5.6 kg/m3/day)
6.5 pslg (0.46 kgf/cm2)
24,000 cfm (629 m3/min)
2
1
252 ft (76.9m)
50 ft (15.2m)
15 ft (4.6m)
378,000 cu ft(10,700m3)
5.4 hr
. 2
180 ft (54.9m)
60 ft (18.2m)
8 ft (2.4m)
580 gpd/sf(23.6 m3/m2/ day)
1390 gpd/sf(S6.6 m3/m2/day)
2.5 hr
2
2
110 ft (33.5m)
15 ft (4.6m)
9 ft (2.7m)
0.85 hr
20 mg/1
1
35ft (10.7m)
8 ft (2.4m)
48 Ib ds/sf/day (235 kg/m2/day)
367,000 cu ft (10,400m3)
158,000 sf (14,700m2)
9-28
-------
Consistent year-round nitrification is obtained in this plant as is shown in Table 9-12. While
only once-monthly analyses of nitrogen are shown in Table 9-12, the consistency of
nitrification in the plant is confirmed by daily ammonia nitrogen analyses of grab samples
which normally show less than 0.2 mg/1 ammonia nitrogen. Wastewater temperatures are
favorable for nitrification, typically dropping to only 17 C, the average monthly
temperature in January.
The treatment plant is currently operating at only one-half of its design capacity. To test the
nitrification portion of the system at close to its design condition, one of the two aeration
tanks was taken out of service during May, June and July of 1974. Both secondary
sedimentation tanks remained in service during this period. Operating conditions and plant
performance for this period are summarized in Tables 9-13 and 9-14 respectively.2? Full
nitrification was maintained throughout this special test period.
Examination of Table 9-14 provides some insight into the operation of the roughing filter.
Total BODs and total COD remained relatively unaffected by the roughing filter operation.
Evidence of treatment, however, is provided by the soluble BOD and soluble COD tests and
the total suspended solids (TSS) determinations. A reduction in soluble BODs an(* soluble
COD occurred coincidentally with an increase in TSS. This indicates organic removal with
associated growth of biomass. In this plant's operation, the roughing filter converts influent
organic matter to biological organisms. Subsequent treatment in the aeration-nitrification
tank provides further oxidation and nitrification.
Construction cost of the added facilities totaled $4,900,000, including legal, engineering and
contract administration. The contract was awarded in October 1970 and construction
essentially completed by October, 1972. Treatment plant operating costs totaled $397,500
for fiscal year 1973/1974, during which time a total of 2,600 million gallons (9.8 million
m^) were processed through the secondary treatment facilities. On a unit basis, treatment
plant O & M costs total $153/mil gal for fiscal year 1973/1974.26
9.5.2 Case Examples of Nitrification-Denitrification for Nitrogen Removal
Four examples of the incorporation of biological nitrification-denitrification in treatment
plants for nitrogen removal are presented in this section. The Central Contra Costa Sanitary
District's plant design is oriented towards reuse of the reclaimed water by nearby industries
as well as meeting discharge requirements. The designs of the Canberra, Australia,
Washington, D.C., and El Lago, Texas plants are laid out so that nitrogen and phosphorus
are removed to protect the receiving waters.
9.5.2.1 Central Contra Costa Sanitary District, California
The Central Contra Costa Sanitary District (CCCSD) has under construction a new 30 mgd
(1.31 m^/sec) Water Reclamation Plant near Concord, California. Due to go on-line in 1976,
the plant is designed to produce water for sale to the Contra Costa County Water District
9-29
-------
TABLE 9-12
NITRIFICATION PERFORMANCE AT THE
SAN PABLO SANITARY DISTRICT TREATMENT PLANT
Month
7
8
9
10
11
12
1
2
3
4
5
6
Year
1973
1973
1973
1973
1973
1973
1974
1974
1974
1974
1974
1974
Flow,
mgd
( mVsec)
6.3
(0.28)
5.6
(0.25)
5.7
(0.25)
5.9
(0.26)
9.7
(0.43)
8.4
(0.37)
9.1
(0.40)
6.6
(0.29)
8.3
(0.36)
7.4
(0.32)
6.2
(0.27)
6.2
(0.27)
Return
sludge
ratio
0.30
0.34
0.25
0.31
0.24
0.29
0.22
0.29
0.19
0.23
0.44
0.44
Temp.
C
22.2
23.6
24.4
23.2
20.3
18.5
17.0
17.8
18.2
19.0
22.0
22.0
MLSS
(% vol-
atile)
1403
(79)
1497
(80)
1758
(76)
1765
(76)
1609
(72)
1545
(73)
1481
(74)
1415
(78)
1522
(74)
1581
(70)
e
2833
(78)
SVI,
ml/
gram
73
68
48
66
73
85
87.
96
101
69
82
77
ec,
days
12.3
13.9
15.4
11.8
9.1
12.1
7.4
9.2
7.5
6.8
e
6.0
HT, -
hr
10.8
12.1
11.9
11.5
7.0
8.1
7.5
10.3
8.2
9.2
e
5.5
Air
use
tfCF/day
(I/sec)
naa
na
20.0
19.4
19.4
20.0
na
na
18.3
na
na
15.3
Roughing filter effluent
BODSb
mg/1
121
125
134
131
88
81
91
92
107
111
140
123
CODb
mg/1
279
306
283
281
212
214
249
240
314
247
332
327
ssb.
mg/1
8.2
97
67
95
59
73
96
79
135
95
130
118
Secondary effluent
BODs°,
mg/1
16
4
6
6
7
4
3
4
4
5
3
2
CODC,
mg/1
78
56
55
62
54
59
53
S3
61
54
50
46
SSC,
mg/1
8
7
8
6
3
7
4
4
6
7
4
4
Organic-ltf1,
mg/1
7.8
3.8
3.1
3.4
3.4
4.8
4.8
2.2
2.8
4.8
5.1
3.6
NH}-NC,
mg/1
<0.2
<0.2
<0.2
<0.2
0.7
0.1
<0.2
<0.2
<0.2
<0.2
<0.2
0.3
NO3~-Nd;
mg/1
18.6
19.8
18.2
20.4
24.8
17.6
17.0
IS. 8
11.8
11.2
18.7
19.0
NO2-NC,
mg/1
0.05
0.02
0.02
0.02
0.26
0.05
0.22
0.02
0.09
0.10
0.06
0.05
na = not available
grab sample at peak flow each week day
composite sample, once per week
composite sample, once per month
e on May 7, one aeration tank taken out of service
-------
TABLE 9-13
AVERAGE PROCESS LOADING CONDITIONS AT THE SAN PABLO SANITARY
DISTRICT TREATMENT PLANT DURING SPECIAL TEST,
MAY 19TH TO JULY 8TH, 1974
Flow, mgd (mVsec)
Temperature, C
Roughing filter
BOD loading,
Ib BOD5/1,000 cf/day (kg/m 3/day)
Hydraulic loading
gpm/sf (mVmVmin)
Aeration-nitrification tanks
MLSS, mg/1
Percent .volatile
SVI, ml/gram
Average detention time, hr
BOD load
Ib BOD5/1,000 cf/day (kg/m3/day)
Ib BOD5/lb MLVSS/day (kg/kg/day)
Solids retention time , days
Secondary sedimentation tanks
Average overflow rate, gpd/sf (mS/
Average return ratio
Average solids load, Ib/sf/day (kg/m2/day)
Return activated sludge, mg/1
6.30(0.28)
23.0
199 (3.19)
2.1 (0.086)
3070
78.2
80.2
5.4
35.8 (0.57)
0.24 (0.24)
6.6
292 (11.89)
0.48
11.0 (53.7)
6835
TABLE 9-14
PERFORMANCE SUMMARY FOR THE SAN PABLO SANITARY DISTRICT
TREATMENT PLANT DURING SPECIAL TESTING, MAY 19TH TO JULY 8TH, 1974
Characteristic
Total BODs, mg/1
Soluble BODs, mg/1
Total COD, mg/1
Soluble COD, mg/1
TSS, mg/1
Ammonia-N
Nitrate -N
Nitrite-N
Raw
waste-
water
220
nab
na
na
191
na
na
na
Primary
effluent
145
97.5
322
191
86.8
na
na
na
Roughing
filter
effluent a
129
52.8
334
137
121
19.8
na
na
Secondary
effluent3
3.3
na
47.5
40.7
4.9
<0.2
19.0
0.04°
aComposite sample each weekday, except as indicated
na = not available
c
Grab sample at peak flow
9-31
-------
which will resell the water to five large industries for cooling and process water. The
reclamation contract calls for production of a water containing less than 10 mg/1 of BOD5,
1 mg/1 total phosphorus and 5 mg/1 total nitrogen. 1
The liquid processing flowsheet for the CCCSD Water Reclamation Plant is shown in Figure
9-10 and design data are in Table 9-15.29 primary treatment follows lime addition and
preaeration and is followed with a separate stage nitrification step. The use of lime in the
primary treatment stage removes the bulk of the organic carbon before nitrification,
resulting in a very stable oxidation of ammonia to nitrate. Addition of lime also enhances
the removal of organic nitrogen, phosphorus, heavy metals and viruses. Biological
denitrification follows nitrification, converting nitrate to nitrogen gas. Multimedia filtration
will also be provided prior to distribution of reclaimed water to industry. Not shown in
Figure 9-10 is a 140 million gallon (530,000 m3) storage basin that can be used to store
primary effluent to reduce peak wet weather loads on the nitrification and denitrification
units. Also, the filtration facility has been provided with a 5 million gallon (18,900 m3)
storage basin for equalizing filter influent and a 30 million gallon (114,000 irP) clear well
for storage of filtered water before pumping it into the distribution system.
There is no intervening recarbonation stage between the primary clarification and
nitrification stages. External carbon dioxide (CO2) is added directly to the first pass of the
aeration-nitrification tanks as needed. External requirements are minimal as the chief source
of CC>2 is not the external source, but the CC>2 generated in the process. Carbon dioxide is
derived from the oxidation of both organic carbon and ammonia. The in-process CC>2
generation capability is an example of lime clarification-nitrification compatibility. The
nitrification pH is also kept in the 7.0 to 8.5 range, which is optimal for nitrification.
Based upon tests by the City of Milwaukee in the 1960's^0,31 ancj testing performed at the
South Eastern Purification Plant in Melbourne, Australia,32 the decision was made to use
flat porous plates arranged uniformly over the bottom of the aeration-nitrification tanks.
This method of fine bubble aeration gives an oxygen transfer efficiency of between 14 and
20 percent under standard conditions. The porous plates are 14 in. (36 cm) in diameter by
\1A in. (3.2 cm) thick and are secured in polypropylene holders as shown in Figure 9-11.
Thirty plates are arranged in a single precast panel; there are 40 panels per pass. Each panel
has an inverted channel shape and is grouted to the tank. By using an inverted channel
shape, the channel forms an air duct with the bottom slab of the aeration-nitrification tanks.
Each channel has a manual drain so that each pass may be drained during start-up and
shutdown operations. Four panels are fed by each downcomer pipe, thereby allowing
throttling of the downcomer pipes and a tapered aeration operation. Air for the nitrification
tanks, channel aeration, and preaeration is provided by two 60,000 scfm (1,698 m^/min)
steam turbine driven single-stage centrifugal blowers and is filtered in large compartment-
type bag filters, to achieve a particulate concentration of less than 0.09 mg/1,000 cu ft
(0.32 mg/100m3) of air.
9-32
-------
FIGURE 9-10
LIQUID PROCESS FLOW SHEET - CCCSD WATER RECLAMATION PLANT (CALIF.)
RAW WASTEWATER
•PRECHLORINATION
SCREENING
INFLUENT
PUMPING
WASTE
BIOLOGICAL'
SLUDGES
LIME REACTOR
AND PREAERATION
PRIMARY
SEDIMENTATION
I
-POLYMER
and/or FeCI3
SLUDGE _
RECLAIMED LIME
PRIMARY EFFLUENT
PUMPING
MAKEUP LIME
.ASH TO
DISPOSAL
NITROGEN GAS
TO
AERATION-
NITRIFICATION
|
SECONDARY
SEDIMENTATION
AIR
RETURN
SLUDGE
SLUDGE
TO PREAERATION
METHANOL
ATMOSPHERE|_
STABILIZATION
POSTAERATIO
POSTCHLORINA
DENITRIFICATION
REACTOR
AERATED
STABILIZATION
r.
1
i
FINAL
SEDIMENTATION
N AIR——*.
1
t
FINAL EFFLUENT
PUMPING
^
-MIXING
RETURN
.SLUDGE
WASTE SLUDGE
*TO PREAERATION
EFFLUENT TO
" SUISUN BAY
CHLORINATION
INDUSTRIAL SYSTEM
PUMPING
T
RECLAIMED WATER
TO INDUSTRY
9-33
-------
TABLE 9-15
CENTRAL CONTRA COSTA SANITARY DISTRICT
WATER RECLAMATION PLANT - DESIGN DATA
Population
Average dry weather flow (ADWF)
Peak dry weather flow (PDWF)
Peak wet weather flow (PWWF)
Raw wastewater quality at ADWF
BODs
SS
TKN
Total phosphorus as P
Primary solids separation
Chemical addition
Lime dose as CaO
Ferric chloride dose as Fed,
PH 3
Preaeration, flocculation, grit removal tanks
Number (one existing)
Volume, ea.
Detention time, ADWF
Primary sedimentation tanks
Number (two existing)
Length
Width
Average depth
Overflow rate at ADWF
Detention time at ADWF
Air blowers - steam turbine driven
Number
Discharge pressure
Capacity, ea.
Horsepower/ea. turbine
Aeration-nitrification tanks
Number
Passes per tank
Length/pass
Width/pass
Depth
Detention time, ADWF
Secondary sedimentation tanks
Number
Diameter
Sidewater depth
Overflow rate, ADWFa
Denitrification tanks (anoxic contact)
and aerobic stabilization)
Number
Length/tank
Width/tank
Depth
Detention time, ADWF
Reactors/tank
Reactors used for stabilization (max)
Methanol to Nitrate - N ratio
310,000
30 mgd (1.31 m3/sec)
48 mgd (2.10 m3/sec)
140 mgd (6.13 m3/sec)
216mg/l
240 mg/1
30 mg/1
11 mg/1
303 mg/1
14 mg/1
11.0
27,700 cu ft (785 m3)
30 min
254 ft (77.4 m)
38 ft (11.6 m)
9.5 ft (2.9 m)
780 gpd/sf (31.8 m3/m2/day)
2.2 hr
8.0 psig (0.56 kgf/cm2)
60,000 cfm
2,750
2
4
270 ft (82.3 m)
35 ft (10.7 m)
15 ft (4.6m)
6.8 hr
115 ft (35.1 m)
16 ft (4.9 m)
720 gpd/sf (29.3 m3/m2/day)
315 ft (96.0 m)
30 ft (9.1 m)
15 ft (4.6m)
102 min
9
4
3.0
9-34
-------
TABLE 9-15
CENTRAL CONTRA COSTA SANITARY DISTRICT WATER
RECLAMATION PLANT - DESIGN DATA (CONTINUED)
Final sedimentation tanks
Number
Diameter
Sidewater depth
Overflow rate, ADWFa
Effluent filtration13
Number
Total rated capacity
Total hydraulic capacity
Media depth
Anthracite
Sand
Filtration rate at rated capacity
Water backwash rate, max
Surface wash rate
Solids disposal and Lime reclamation
Sludge thickening
Primary sludge thickener (converted digester)
Number
Diameter
Sludge to thickener
Solids loading
Centrate thickener (converted digester)
Number
Diameter
Solids loading
Centrifugation
Number
Type
Max feed rate
Max g force, G
Cake solids cone., first stage
Cake solids cone., second stage
Furnaces
Number
Diameter of hearth
Number of hearths
Rated capacity
Sludge burning duty
Rec ale ination-duty
Recycled lime fraction
115 ft (35.1 m)
20 ft (6.1 m)
720 gpd/sf (29.3 m3/m2/day)
36 mgd (1.46 m3/sec)
54 mgd (2.20 m3/sec)
2 ft (0.61 m)
1 ft (0.30 m)
4.0 gpm/sf (2.72 VmVsec)
25 gpm/sf (17.0 l/m2/sec)
0.75 gpm/sf (0.51 l/m2/sec)
62 ft (18.9 m)
243,000 Ib/day (110,500 kg/day)
80 Ib DS/sf/day (390 kg/m2/day)
62 ft (18.9 m)
42 Ib DS/sf/day (205 kg/m2/day)
t
vertical, solid bowl
260 gpm (0.016 m3/sec)
3,100
55 percent
14 percent
22 ft (6.7 m)
11
70,000 Ib DS/day (32,000 kg/day)
150.,000 Ib DS/day (68., 000 kg/day)
60 - 65 percent
Loading applied to sedimentation tanks at PWWF can be equalized by a primary effluent holding basin
Data on filtration from reference 28
Nitrified mixed liquor flows to four circular sedimentation tanks. The tanks have a
center-feed, peripheral discharge arrangement with sludge removal by vacuum-type sludge
collectors. Sludge return rate is controlled by the blanket level in each secondary
sedimentation tank. Solids retention time is controlled by use of waste activated sludge flow
meters and return sludge concentration.
9-35
-------
FIGURE 9-11
NITRIFICATION-DENITRIFICATION SYSTEM AT THE
CCCSD WATER RECLAMATION PLANT (CALIFORNIA)
REnjRJTNITRIFIED
DOE (RNS) PUMPS
(TYP.)
NITRIFIED MIXED
\LIQUOR CHANNEL
\ VRNS CHANNEL
PASS WIDTH 35'-O
AIR PLENUM CHANNEL
TYP OF 40 PER PASS
31'-6
DENITRIFICAT|ION
TANKS
EFFLUENT FRO»
PRIMARY
SEDIMENTATION
Nl TRIP ICATION
-jL^j^U^-:-.-
FINAL
SEDIMENTATION
TANKS
NITRIFIE
(RDS) PUMPS
SECTIONt-1
<—POROUS PLATE
POLYPROPYLENE HOLDER
-POSTAERATION CHANNELS
PLAN AND FLOW DIAGRAM
9-36
-------
Suspended growth denitrification is used for nitrate removal. Uncovered reactors are
employed, as initial testing showed them to be acceptableJ Two parallel denitrification
tanks consist of nine completely mixed reactors in series. This arrangement allows
approximation of a plug flow hydraulic regime. Each reactor is equipped with a low-shear
turbine type mixer to keep the mixed liquor solids in suspension. The last four reactors in
the series are equipped with spargers for aeration. With this arrangement, the first five cells
in the series can be used for anoxic denitrification reactors and the last four cells can be
used for denitrification or aerated stabilization depending on whether or not aeration is
used. The volume devoted to each function can be varied to meet seasonal loads and
temperatures. The anoxic reactors are designed to operate at 0.1 Ib NC>3 — N rem./lb
MLVSS/day at MLSS levels ranging from 3000 to 4000 mg/1. The gates between the last five
reactors allow positive prevention of backmixing of oxygen between the reactors selected
for denitrification and the reactors selected for aerated stabilization. The gates are open at
the top to prevent trapping of floating solids in any reactor. The denitrified mixed liquor
channel between the denitrification tanks and the final sedimentation tanks is aerated for
further stabilization. Final sedimentation tanks are similar to the secondary sedimentation
tanks.
All waste sludge produced in the CCCSD Water Reclamation Plant is eventually cycled into
the primary sedimentation tanks and appears in the primary underflow. Sources of sludge
include the suspended solids associated with the raw wastewater, the solids that are wasted
to the primary sedimentation tanks from the subsequent biological treatment stages, and the
inorganic sludges that are precipitated due to chemical action.
To maintain a pH of 11.0 in the primary sedimentation tank, large quantities of lime,
approximately 400 mg/1 as Ca(OH>2, must be used. The need for such a large dosage
predicates the economical recovery and reuse of the lime. The solids flowsheet is shown in
Figure 9-12. The heart of the system is a two-stage centrifuge process using vertical
solid-bowl centrifuges, where primary sludge is separated into two components, sludge cake
rich in calcium carbonate (CaCO3) and centrate containing most of the organic material,
magnesium and phosphorus. The first stage centrifuge cake is approximately 70 percent
CaCO3- This cake is recalcined in a multiple hearth furnace, subsequently dry classified and
the lime returned to storage. The lime recovery is expected to be approximately 60-65
percent of the lime used.33
The first stage centrate is thickened before being clarified in centrifuges identical to those
used for primary sludge classification. The resulting cake is reduced in a multiple-hearth
furnace (MHF) to a sterile ash which is used for landfill. The MHF is identical to that used
for re calcining.
Energy in the hot off-gases from each MHF is reclaimed via waste heater boilers. Recovered
steam is used to run the turbines which power the aeration blowers.
9-37
-------
FIGURE 9-12
SOLIDS FLOW DIAGRAM AT THE CCCSD WATER RECLAMATION PLANT (CALIF.)
MAKEUP
LIME.
RECYCLE LIME I
1 1
LIME
STORAGE
EQUIP.
M
EXHAUST
GAS TO
ATMOSPHERE
WET
SCRUBBER
STEAM 1
I r~ *-^yWASTE
1 — ^ JHEAT
BOILER
1 U -i
DRY! P
SCRU8BER\/
1 RECALCINE
FURNACE
WASTI
BIOLOG
LIME SOLIC
FEED
EQUIP. OV
4 R
T
SCREENED
WASTEWATEft
r
ICAL POLYMER and/or F»CI3
ERF LOW
ETURN
\ i
P0IUAP
REACTOR PRIMAK
tfPREAE*' SEDIMENTA
__» ATION TANKS
T.ON EFFLUENT
_-— J
f THICKENER
SLUDGE
THICKENER)
THICKENER UNDERFLOW |
J
PIO«?T ^TAffF CE»
CENTRIFUGE
(WET
V-CLASSIFICAJ
XTION)/'
|CoC03
CAKE
RECALCINEO
ASH
\
_ RECYCLE LIME
DRY LIME
CLASSIFICATION
' REJECTS
UTRATE
T THICKENER
OVERFLOW
CfcNTRATE ~~~ ~~*
THICKENER
S^
(THICKENED
ICENTRATE
LJ SECOND U POLYME
STAGE
ICENTRIFUGEJ Hg
' » m
SLUDGE
IASH
1 TO PRIMARY
INFLUENT
EXHAUST
GAS TO
ATMOSPHERE
WET
CRUBBER
(
STEAM
/ ^ f
c \_j
WASTEV ^
AT BOILER
-GRIT L /DRY
YSCRUBBER
-SCUM
^ ASH TO
" DISPOSAL
9-38
-------
Operational control of the CCCSD Water Reclamation Plant will be divided into four areas,
each of which will be manned by a senior operator, operator, and various maintenance men
during each shift. The four areas are: (1) primary treatment, (2) biological treatment, (3)
filtration, and (4) solids handling and conditioning. Because of the highly automated control
system, cathode ray tubes (CRT) are provided in each operator control room, eliminating
the need for lighted control panels. In addition to the main CRT's in the computer room,
CRT's are located throughout the plant for data monitoring. Because of the very exacting
standards required of industrial water, this plant is designed to prevent potential plant
upsets by incorporating a direct digital control (DDC) dual-computer system to monitor and
control all process functions.34 Operation of the on-line computer is by a "management by
exception" basis which means that as long as the process status is within normal limits,
information is not printed, displayed, or needlessly alarmed.
While this plant will not go on-line until 1976, the process was rigorously monitored in a
full-scale test at the existing CCCSD plant for a 23-month period. Portions of the test data
are summarized in Section 5.2.4.2.
The construction contract for the first phase of work, excluding effluent filtration, totaled
$47,000,000 and was let in mid-1972. The construction contract for the effluent filtration
and appurtenances totaled approximately $14,000,000 and was let in the fall of 1974.
Operation and maintenance costs were estimated for fiscal year 1976/1977 at $300/mil gal
based on an average dry weather flow of 30 mgd.
9.5.2.2 Canberra, Australia
The Lower Molonglo Water Quality Control Centre (LMWQCC), an advanced wastewater
treatment plant under construction, is designed to serve the City of Canberra, Australia's
national capital. Present discharges from Canberra's existing plants are causing algal growth
problems in the receiving water and in a downstream reservoir. To circumvent these
problems, it was established that the LMWQCC should produce an effluent that contains:35
1. Substantially no settleable solids, turbidity, color or odor.
2. BOD5 and suspended solids concentrations of less than 5.0 mg/1.
3. A median fecal coliform content less than 50 per 100 ml and a 90 percentile value
of less than 400 per 100 ml.
4. Total nitrogen not exceeding 2.0 mg/1 as N and total phosphorus not exceeding
0.15 mg/1 as P.
5. Detergent concentrations less than 0.5 mg/1 of MBAS.
6. No substances toxic to aquatic biota.
9-39
-------
Unit processes employed at the LMWQCC include raw wastewater screening, lime addition,
grit removal, flocculation, primary sedimentation, nitrification, secondary solids separation,
denitrification, effluent filtration, effluent disinfection by chlorination, and dechlorination
prior to discharge. Figure 9-13 is the process flow diagram for the treatment of the liquid
fraction. Design data adopted and used for the LMWQCC is presented in Table 9-16.36 The
solids processing flowsheet is very similar to that shown in the case history for the Central
Contra Costa Sanitary District's Water Reclamation Plant in Section 9.5.2.1 and will not be
duplicated here.
Very steep site conditions and confinement of the plant to a limited area mandated an
unusual arrangement of treatment structures. An example is the nitrification tanks which
step down the hillside as shown in Figure 9-14. A total of 4 parallel tanks are provided. Each
tank is subdivided into 8 compartments. A very close approach to plug flow is provided by
these tanks since backmixing is prevented because the only way mixed liquor can pass along
the tank is by overflowing weirs between compartments. The plug flow arrangement is the
recommended process configuration for separate stage nitrification when very low residual
ammonia nitrogen levels are required. Aeration air is provided to each compartment through
porous plate diffusers spread across the tank floor. Diffuser arrangement is very similar to
that described in Section 9.5.2.1 for the CCCSD plant. Carbon dioxide (in furnace stack gas)
is added to the first two compartments on a continuous basis according to pH level in the
nitrification tanks.
Due to site restrictions, an attached growth reactor system was chosen for the
denitrification unit. The reactor chosen was specifically developed for this project and is the
nitrogen gas filled denitrification column described in Section 5.3.2.1. Design details of the
column are shown in Figure 5-7.
Bids for the LMWQCC were received in February, 1974 and the winning tender was $A
27,000,000 ($US 35,600,000). The plant is expected to be completed by late 1976. O&M
costs calculated in February 1974, exclusive of amortization, were estimated at the
equivalent of $US 266/mil gal.36
9.5.2.3 Washington, D.C.
In 1969, regulatory agencies established stringent effluent standards for treatment plants
that discharge into the Potomac River in the vicinity of Washington, D.C. These standards
required upgrading the Washington, D.C. Blue Plains Plant to provide phosphorus and
nitrogen removal as well as improved BOD and SS removals.37,38,39,40 jn early 1975 the
EPA announced that the construction of the denitrification portion of the Blue Plains plant
would be delayed for two years. This decision came after study of the energy and
construction costs for the facility. During the postponement period, water quality
improvement due to phosphorus removal and other treatment will be evaluated to
determine if denitrification is necessary to achieve eutrophication control goals.41 -phe
9-40
-------
FIGURE 9-13
PROCESS FLOW DIAGRAM FOR THE LOWER MOLONGLO WATER QUALITY
CONTROL CENTRE (CANBERRA, AUSTRALIA)
CI2
PREAERATION
PRIMARY SOLIDS AND GRIT
SEPARATION TANKS REMOVAL TANKS
LOME
BYPASS IN EXCESS
OFSxADWF
RAW
WASTEWATER
BIOLOGICAL
NITRIFICATION
T.
FINAL CLARIFIERS
BIOLOGICAL
DENITRIFICATION
COLUMNS
EFFLUENT FILTER
CI2 Clg DE-CI2
CONTACT
TANKS
TO NITRIFICATION TANKS
-------
TABLE 9-16
LOWER MOLONGLO WATER QUALITY CONTROL
CENTRE (AUSTRALIA), DESIGN DATA
Population
Average dry weather flow (ADWF)
Peak dry weather flow (PDWF)
Peak wet weather flow (PWWF)
Raw wastewater quality at ADWF
BOD,.
SS 5
NH^-N
TKN
Primary solids separation
Chemical addition
Lime dose as CaO
Ferric chloride dose
PH
Flocculatlon and grit removal tanks
Number
Volume (each)
Detention time (ADWF)
Primary sedimentation tanks
Number
Length
Width
Depth
Overflow rate at ADWF
Detention time at ADWF
Air blowers
Number
Discharge pressure
Capacity - total
Biological nitrification tanks
Number of tanks
Compartments per tank
Width
Length/compartment
Depth
Volume (4 tanks)
CO- required
Detention time at ADWF
BOD load
Final Sedimentation tanks
.Number
Diameter
Sldewater depth
Overflow rate at 3 x ADWF
Biological denitriflcatlon columns
Number of cells
Width/cell
Length/cell
Media depth
Total media volume
Application rate
Methanol to nitrate - N ratio
Effluent filtration
Number
Width
Length
Media depth
Anthracite
Sand
Filtration rate at 3 x ADWF
Water backwash rate
269,000
28.8 mgd (1.26 mVsec)
43.3 mgd (1.90 mVsec)
144.0 mgd (6.30 mVsec)
221 mg/1
242 mg/1
35 mg/1
59 mg/1
230 mg/1
20 mg/1
11.0
27,700 cu ft (785 m )
20 mln
212 ft (64.7 m)
38.4 ft (11.7 m)
8.9 ft (2.7m)
880 gpd/sf (35.7 m3/ni /day)
1.8 hr
3 7
7.5 psig (0.53 kgf/cm )
90,000 cfm (2,550 m3/min)
4
8
35 ft (10.5 m)
39.4 ft (12 m)
16.2 ft (4.95 m) ,
565 cu ft (19,960 m )
3,600 Ib/day (13,600 kg/day)
4.4 hr
0.18 Ib BOD /lb MLVSS/day
120 ft (36.6 m)
20.6 ft (3.28 m)
1,635 gpd/sf (66.6 nvVmVday)
8
32.2 ft (9.8 m)
40.2 ft (12.3 m)
20 ft (5.8m) .
196,750 cu ft (5,572 m )
146 gal/cu ft/day (19.0 m3/m /day)
2.8
23.75 ft (7.24 m)
105 ft (32.00 m)
3.5 ft (1.06 m)
1.5 ft (0.45 mi
6.0 gpm/sf (4.1 1/mVsec)
18.6 gpm/sf (12.65 l/m2/sec)
9-42
-------
TABLE 9-16
LOWER MOLONGLO WATER QUALITY CONTROL
CENTRE (AUSTRALIA), DESIGN DATA (CONTINUED)
Effluent chlorlna tion-dechlorination
Number of tanks
Contact time at ADWF
Solids disposal and lime reclamation
Primary underflow solids
First stage centrifugallon (classification)
Number of centrifuges
Calcium carbonate recovery
Cake solids contraction
Total solids capture
Second stage centrlfugation (clarification)
Number of centrifuges
Cake solids concentration
Total solids capture, minimum
Furnaces
Number
Diameter of hearth
Number of hearths
Rated capacity
Sludge burning duty
Recalclnatlon duty
Reclaimed lime to storage
Recycled lime fraction
1
50 min
198,450 Ib/day (90,000 kg/day)
90 percent
50-60 percent
60 percent
12-18 percent
70 percent
22 ft (6.7m)
9
70,400 Ib DS/day (32,000 kg/day)
237,600 Ib DS/day (108,000 kg/day)
47,960 Ib/day (21,800 kg/day)
68 percent
Assumed influent nitrate nitrogen: 28 mg/1
discussion which follows includes the design of the denitrification facilities for this plant as
it provides an example of how denitrification can be incorporated in a. large plant. Most of
the discussion in this section is drawn from the Process Design Manual for Upgrading
Wastewater Treatment Plants, an EPA Technology Transfer publication.37
In 1969, very little performance data were available on the alternative phosphorus and
nitrogen removal methods that might be used in this situation. Through the cooperation of
the Joint EPA-DC Pilot Plant, it was possible to pilot and evaluate several alternative
nutrient removal treatment sequences. Based on these studies, two-point addition of a metal
salt was selected for phosphorus removal, and it was determined that nitrogen removal
would be best achieved through biological nitrification and denitrification processes. The
pilot studies also indicated that to consistently meet the established effluent standards,
multimedia filtration was required. Anticipated performance data for the upgraded plant are
presented in Table 9-17. Figures 9-15, 9-16 and 9-17 show, respectively, the flow diagrams
for the primary and secondary systems, nitrification and denitrification systems and
filtration and disinfection systems of the plant.
The existing secondary system consists of four aeration tanks and 12 sedimentation units.
To handle the anticipated increase in plant design flow from 240 mgd to 309 mgd, (10.5
to 13.6 m^/sec) the existing secondary system will be enlarged with two additional
9-43
-------
FIGURE 9-14
SECTION THROUGH NITRIFICATION TANKS AT THE LMWQCC,
CANBERRA, AUSTRALIA
LEVEL
CONTROL PRIMARY SOLIDS
GATE SEPARATION TANK
EFFLUENT
COLLECTOR PIPE
MIXED ±
LIQUOR
CLARIFIERS
COMPARTMENT (TYPICAL)
RETURN ACTIVATED SLUDGE
DISTRIBUTION CHANNEL
CARBON DIOXIDE
(STACK GAS)
•AIR (TYPICAL)
PLUG FLOW NITRIFICATION TANK (ONE OF FOUR)
-------
TABLE 9-17
ANTICIPATED PERFORMANCE DATA AND EFFLUENT STANDARDS
BLUE PLAINS PLANT (REFERENCE 39)
BOD, mg/1
Total phosphorus , mg/1
Nitrogen:
Organlc-N, mg/1
NHj-N, mg/1
NOj + NO J-N , mg/1
Total N, mg/1
Influent
206
8.4
8.6
13.7
0
22.3
Secondary
Effluent
35
2.0
3.0
14.8
0.2
18.0
Nitrification
Effluent
10
1.0
1.0
1.5
11.1
13.6
Denltrification
Effluent
6
0.5
1.0
1.0
1.0
3.0
Filtration
Effluent
4
0.2
0.5
1.0
0.5
2.0
Effluent
Standard
5
0.22
-
-
-
2.4
aeration tanks and 12 additional final sedimentation tanks. The aeration tanks are designed
for a volumetric loading of 120 Ib BOD5/1,000 cu ft/day (1.92 kg/m3/day), an organic
loading of 2.4 Ib BOD/lb MLSS/day and a MLSS concentration of 1,300 mg/1. Since the
increased design loadings require more air per unit volume than the existing aeration system
can deliver, the existing aeration capability will be increased. This system will be modified
from a coarse-bubble, spiral-roll system to a coarse-bubble, spread-pattern system to
improve oxygen transfer efficiency. The secondary system air capacity has been designed to
provide 0.54 Ib C>2/lb BODs removed.
Alum or ferric chloride will be added to the mixed liquor of the secondary system and is
expected to remove approximately 70 percent of the phosphorus contained in the plant
influent. The addition of metallic salts to the secondary system is also expected to improve
the BOD removal in the secondary system from 75 percent to 85 percent. This will ensure a
secondary effluent BOD concentration of less than 40 mg/1, which was found during pilot
testing to be desirable for successful nitrification in the second stage. To remove most of the
remaining phosphorus, metallic salts will also be added to the nitrogen release tanks.
Biological nitrification facilities are designed for oxidation of 0.066 Ib NH^-N/lb MLVSS/day
at minimum wastewater temperatures and a MLVSS concentration of 1,700 mg/1.
At the stoichiometric oxygen requirement of 4.6 Ib O2/lb NH^-N oxidized, one hundred
and twenty - 75 hp turbine aerators are required. Maximum air supply to the turbines will
be 88,000 cfm (2500 m /min). The turbines were selected in this instance because, due to
the limitations of the site, the nitrification tanks are designed to have depths of 30 feet
(9.15 m) to obtain the required volume. The turbines will provide adequate mixing to this
depth and are capable of supplying a range of oxygen to the system as required by varying
ammonia-nitrogen influent loads and varying wastewater temperatures. Lime will be added
to the nitrification reactor to maintain a minimum wintertime pH of 7.5.
9-45
-------
FIGURE 9-1 5
ON
WASHINGTON, D.C. BLUE PLAINS TREATMENT PLANT
FLOW DIAGRAM OF PRIMARY AND SECONDARY SYSTEMS
|CL2
RAW ' '
WASTEWATER
[£E
RAW 1 '
WASTEWATER
CG
SLUDGE -
WASTEWATER _
CHEMICALS _
(INC. AIR)
S3
ANACOSTIA ^1 '
FORCE MAIN
ruup 1 AERATED
^ rump i^^^ r. m rHAM
~ MAIIONl^^^* "K" "•"*""
o
s| PSl
-• n + G><
< Z "* THICK
2 o
O u
Ul
EXIST 1 EXIST
fc, PUMP 1 kr.,V™Il?.
* STATION 1 "-BKII «.M«WI!
LEGEND
NTINUOUS INTERMITTENT
FLOW FLOW
ALT.
SPENT
WASHWATER
I FERRIC OR ALUM!
-
-. \r^\
\\ '**/ PRIMARY V
lERlT T%l SEDIMENTATION V
| *A TANKS r
r
i
d
TO
kVITY
ENERS
r;
ERl??f\s
SLUDGE
PROCESSING »
RECYCLE 1
GRAVITY '
THICKENING ,
OVERFLOW 1
EXIST. \
PRIMARY 1
EDIMENTATIOM |__
TANKS 1
L
r
i r
12
_^ DISINFECTION | EXCESS FLOW TO RIVER
SECONDARY| ^ SECONDARY |
1
t
RSI
A
EXIST. |T
SECONDARY 11
REACTORS H
|AIR|
-*
li^/ •
[^•2] rQ5tI*!!!iJ
f FERRIC OR ALUM J
t ^ 1
• •
! POLYMER |
FERRIC" oR~AiuM]
WSl
EXIST. | EXIST. |
SECONDARY! SECONDARY!
k
RSL
1 fyMLl
rpbl'YMETI
FERRlC~6T Al1T«r]
. . *Sl
1
W
^mm^fim
t
•HB^^H
N
DE
r
NITRIFICATION
| INFLUENT*
1
i
ALT. TO
FILTERS
OR
OUTFALL
ITC'IT"
£__
ALT. TO
ITRIFICATION
OR
NITRIFICATION
r
SL
-*•
TO FLOTATION THICKENERS
PSL
PRIMARY SLUDGE
RSL
RETURN SLUDGE
WSL
WASTE SLUDGE
-------
FIGURE 9-16
WASHINGTON, D.C. BLUE PLAINS TREATMENT PLANT
FLOW DIAGRAM OF NITRIFICATION AND DENITRIFICATION SYSTEMS
LIME
1
SECONDARY
EFFLUENT |
i
i
i_
POLYMER AIR METHANOl
SPENT
WASHWATER
it t
NITRIFICATION! !
NITRIFICATION! 1 ...... | I
SED. BASINS |-^$TPAUTT0PNU
1 t !'
•» 1 ! 1
i
ALT. WSL FROM
SECONDARY
OR
DENITRIFICATION
M
DENITRIFICATIONli
REACTORS U
FERRIC OR ALUMJ
AIR 1 ' '
i
NITROGEN
. RELEASE
r TANKS
POLYMER
"\
DENITRIFICATIONI TO MULTI-MEDIA.
^ 1 FILTERS
t -"
; RSI
rwSl 1 [ALTj i , |WSL
1 i
ALT. WSl FROM
SECONDARY
OR
NITRIFICATION
r
Cl
SL
WASTEW
LEGEND
CONTINUOUS INTERMITTENT
FLOW FLOW
TO
FLOTATION THICKENERS
CHEMICALS
(INC. AIR)
RSL
WSL
RETURN SLUDGE
WASTE SLUDGE
-------
FIGURE 9-17
WASHINGTON, D.C. BLUE PLAINS TREATMENT PLANT
FLOW DIAGRAM OF FILTRATION AND DISINFECTION SYSTEMS
SPENT WASHWATER
TO NITROGEN RELEASE
OO
LEGEND
CONTINUOUS INTERMITTENT
FLOW FLOW
FLUSHING. SERVICE AND
DILUTION WATER
WASTEWATER
CHEMICALS
-------
The nitrification sedimentation tanks are designed for average and peak hydraulic and solids
loadings of 580 and 1,210 gpd/sq ft (23.6 m3/m2/day and 49.3 m3/m2/day) and 17.4 and
36.6 Ib/sq ft/day (85 and 177 kg/m2/day), respectively. The sludge return system is
designed to provide return of 40 percent of peak flow. However, the system will normally
be operated to return 30 percent, of the average flow. Continuous monitoring of the DO
content of the nitrification effluent will be provided to ensure that the influent DO to the
denitrification system is minimized.
The biological denitrification system is laid out to include reactors, nitrogen release tanks
and sedimentation tanks. The reactors have been designed for removal of 0.0425 Ib NO^
-N/day/lb MLVSS at a design MLVSS concentration of 2,100 mg/1, with up to 4.5 Ib
methanol added/lb NO^ —N applied. The reactors will be 44 ft (13.4 m) deep and be
equipped with forty-eight - 75 hp mixers, and will be covered but not airtight.
The nitrogen release tanks were designed to serve three functions: (1) to strip supersaturated
nitrogen gas, (2) to provide mixing for second-stage metal salt addition for residual
phosphorus removal and ,(3) to provide an. aerobic zone for removal of excess methanol.
These tanks will furnish a 20-minute detention period at average flow and 12 minutes at
peak flow.40
The denitrification sedimentation tanks are designed for hydraulic loadings of 670 gpd/sq ft
(27.3 m3/m2/day) at average flow, and 1,410 gpd/sq ft (57.4 m3/m2/day) at peak flow.
Solids loadings are 25.6 and 54.0 Ib/sq ft/day (125 and 264 kg/m2/day) at average and peak
flows, respectively.
The 36 multimedia filters are designed for filtration rates of 3.0 gpm/sq ft (2 l/m2/sec) at
average flow and 6.2 gpm/sq ft (4.2 l/m2/sec) at peak flow. Backwashing will be done at
intervals of 24 hours at a rate of 25 gpm/sq ft (17 l/m2/sec). The backwash water will be
equalized in conduits and may be returned upstream of either the secondary reactors or the
nitrogen release tanks.
Provision has been made to chlorinate either upstream or downstream of the filters with 24
minutes detention provided in contact tanks following the filters.
Sludge processing facilities will include gravity thickening of primary sludge, flotation
thickening of secondary and advanced treatment sludges, vacuum filtration and sludge
incineration.
9.5.2.4 El Lago, Texas
El Lago, Texas, is a small suburban community of 3,000 persons located near the Lyndon B.
Johnson Space Center. The operating agency for wastewater treatment is the Harris County
Water Control and Improvement District #50. This district currently operates a 0.3 mgd
(0.013 m3/sec) treatment plant. In 1969, the District received an order from the Texas
9-49
-------
Water Quality Board that mandated protection of Clear Lake from excessive eutrophication.
Two means were available for compliance with this order at that time; export of wastewater
or providing nutrient removal prior to discharge to Clear Lake. The second option was
chosen and the District obtained a grant from the EPA to demonstrate full-scale nitrogen
and phosphorus removal.
The original plant consisted of a rock trickling filter plant with anaerobic sludge digestion
for solids processing. The modified flowsheet incorporating nutrient removal is shown in
Figure 9-18. Added facilities are identified by asterisks and include new aeration-nitrifi-
cation tanks, new denitrification columns, new tertiary filtration, and facilities for
metal salt, polymer and methanol addition. All existing structures were incorporated into
the upgraded plant. Design criteria for the modified plant are shown in Table 9-18. Two
separate types of denitrification columns were provided so that alternative designs could be
compared. One set of columns was of the submerged high porosity media type described in
Section 5.3.2.2 and was filled with Koch Flexirings (coarse media column in Table 9-18).
The other type was the submerged low porosity media type described in Section 5.3.2.3 and
as supplied by the Dravo Corp. (Fine media column in Table 9-18.) Both column types are
shown on Figure 9-19.
Tables 9-19 and 9-20 are tabular summaries of the initial performance with the fine media
and coarse media respectively. Phosphorus removal during the fine media evaluation was
erratic due to the cessation of iron addition during storm conditions. The fine media
columns produced an effluent containing 17 mg/1 of suspended solids because the columns
are backwashed with nitrified effluent rather than clear tertiary filter effluent. Since tertiary
filtration is also provided, this has not been a problem at El Lago.
Table 9-21 reproduces a month of operating data for the fine media denitrification
columns. This data shows the improvements in all parameters of effluent quality obtained
after one year of experience in plant operation.
The coarse media denitrification columns also performed well during the investigation.
While the fine media columns required at least daily backwashing to prevent excessive
headlosses, the high void volume of the coarse media allowed operation without frequent
backwash. The routine procedure was to backwash every 4 weeks. This proved to be an
important difference to the plant operators, who found that the coarse media units required
much less attention than the fine media units.43
The capital costs for the modifications to the El Lago Facility were incurred over a two-year
period (1971-1973) and totaled $312,365 including change orders. This cost includes the
provision of dual denitrification facilities; had only one type of denitrification system been
included it is estimated that the total would be about $75,000 less. The only increase in
o
operating costs has been for chemicals and power and this has totaled $96/mil gal. It was
found that the existing plant operators could adapt to advanced waste treatment processes
and no increase in staff was required.
9-50
-------
FIGURE 9-18
EL LAGO, TEXAS WASTEWATER TREATMENT PLANT, FLOW DIAGRAM
Fed
t+polymer* ^
RAW WASTEWATER
m Digester Supernatant
INFLUENT PUMPING
1
PRIMARY
SEDIMENTATION
SLUDGE
ROUGHING
TRICKLING FILTERS
ANAEROBIC
DIGESTION
WASTE
FeCI,
Methanol
r
TION-
CATION
NKS*
1
JDARY
JTATION
NKS
1
i
SLUDGE
RETURN
SLUDGE
I
SAND
DRYING
BEDS
BACKWASH
TO
HEADWORKS
DENITRI-FICATION
COLUMNS
(LARGE MEDIA)
DEN IT RI.F 1C AT ION
COLUMNS
(FINE MEDIA)
J I TERT
"^ FILTR
I ARY
ATION*
KEY
*New facility
CHLORINE CONTACT
FINAL EFFLUENT
TO RECEIVING WATERS
9-51
-------
TABLE 9-18
DESIGN DATA, EL LAGO, TEXAS WASTEWATER TREATMENT PLANT
Population
Average dry weather flow (ADWF)
Peak dry weather flow (PDWF)
Peak wet weather flow (PWWF)
Raw wastewater quality
BOD,.
SS 5
Total Kjeldahl nitrogen
Primary sedimentation tanks (existing)
Number
Detention time (ADWF)
Overflow rate (ADWF)
Roughing trickling filter (rock, existing)
Number
Depth
Rock size
Volume of media
Organic load (ADWF)
Recirculation rate
Air blowers (new)
Number
Discharge pressure
Capacity - total
Aeration-nitrification tanks (new)
Number
Arrangement
Volume - Total
Detention time (ADWF)
MLVSS
Solids retention time
Secondary sedimentation tanks (existing)
Number
Detention time (ADWF)
Overflow, rate at ADWF
Overflow rate at PWWF
Denltrification columns
Coarse media
Number (series)
Type
Diameter
Media depth
Media type
Specific surface
Voids
Empty bed contact time
Surface application rate at ADWF
Air backwash rate
Water backwash rate
Fine media
Number (series)
Type
Diameter
Media height
Media type
Specific surface
Voids
Empty bed contact time
Surface application rated at ADWF
Air backwash rate
Water backwash rate
3,000
0.3 mgd (0.013 mVsec)
0.5 mgd (0.022 m3/sec)
1.0 mgd (0.044 m /sec)
161 mg/1
195 mg/1
37.5 mg/1
2
1.6hr
440 gpd/sf (30 mVmVday)
6.5 ft (2.0 m)
4 in (100 mm)
20,900 cu ft (590 m3)
12 Ib BOD /1,000 cf/day (0.192 kg/rn/day)
t>.3 mgd (0.013 mVsec)
6.5 psig (0.46 kgf/cm2)
900 cfm (25.2 mVmin)
2
Series _
10,100 cu ft (302 m )
6.1 hr
1,000 mg/1
10 days
2
5.4 hr
320 gpd/sf (13 m3/m2/day)
1,060 gpd/sf (42 m3/m2/day)
£,
50 psig (3.5 kgf/cm2) pressure vessel
10 ft (3 m)
10 ft (3 m)
Koch Flexirings
105 sf/cu ft (346 m2/m3)
92 percent
1 hr
2 . 5 gpm/sf (1.6 1/m /sec)
10 cfm/sf (3.1 mVmVmin)
20 gpm/sf (13.5 l/m2/sec)
50 psig (3.5 kgf/cm2) pressure vessel
6 f t (1. 8 m)
6.5 ft (2.0 m)
3 to 4 mm uniform sand
250 sf/cu ft (825 m2/m3)
40 percent
0.25hr „
7.4 gpm/sf (5 1/m /sec)
8 cfm/sf (2.4 mVmVmin)
20 gpm/sf (13.5 l/m2/sec)
9-52
-------
TABLE 9-18
DESIGN DATA, EL LAGO, TEXAS WASTEWATER TREATMENT PLANT (CONTINUED)
Tertiary filtration
Number (parallel)
Type
Height
Diameter
Media height and type
Media support base
Surface application rate at ADWF
Water backwash rate
Chlorine contact tank
Detention time (ADWF)
Anaerobic digestion
Volume
Sand drying bed-
Area
o
30 pslg (2.1 kgf/cm ) pressure vessel
8 ft (2.4 m)
3.5 ft (1.15 m)
3 f t (1.0 m) of 0 .3 to 0 . 8 mm sand
0.5 ft (0.15 m) gravel
2.2 gpm/sf (1.5 l/m2/sec)
15 gpm/sf (10 l/m2/sec)
1 hr
8,830 cf (2,472 m3)
6,300 sf (580 m2)
FIGURE 9-19
EL LAGO, TEXAS DENITRIFICATION COLUMNS, COARSE MEDIA
TYPE ON RIGHT AND FINE MEDIA TYPE ON LEFT
9-53
-------
TABLE 9-19
INITIAL PERFORMANCE OF FINE MEDIA DENITRIFICATION COLUMNS
AT EL LAGO, TEXAS - JUNE 4 TO JULY 6, 1973
Constituent
Total P
Soluble P
SS
NH+-N
TKN
NO~-N6
BOD5
COD
Temperature
Mean value, mg/1 at indicated sample location
RaW a,b
wastewater
12.8
10.3
113
18.7
42.6
175
297
26.5
Primary
Influent
15.4
4.7
289
21.7
38.6
222
488
-
Primary
effluent
8.4
4.1
72
21.5
30.2
181
-
Nitrified
effluent
7.3
3.4
37
0.9
3.7
15.2
65*
121
-
Denitrified
effluent0'
6.6
5.5
17
0.8
2.4
2.6
9
72
-
Final
effluent
4.8
3.6
3
0.6
3.3
2.3
9
51
-
Average flow to plant: 0.307 mgd (0.013 m /sec)
Peak dally flow to plant: 1.0 mgd (0.044 m /sec) 3
Average flow to denltrlflcatlon columns: 0.254 mgd (0.011 m /s,ec)
Peak dally flow to denltrlficatlon columns: 0.420 mgd (0.018 m /sec)
Nitrite - N always less than 0.2 mg/1
Includes methanol
Degrees C
TABLE 9-20
INITIAL PERFORMANCE OF COARSE MEDIA DENITRIFICATION COLUMNS
AT EL LAGO, TEXAS - JULY 8 TO AUGUST 31, 1973
Mean value, mg/1 at Indicated sample location
Constituent
Total P
Soluble P
SS
NH4-N
TKN
N0~-Ne
BOD
O
COD
Tempera tureg
Raw
wastewater '
12.3
10.3
102
16.3
29.7
143
248
27.2
Primary
Influent
13.1
3.1
231
14.6
31.8
156
336
-
Primary
effluent
6.7
2.4
63
14.4
26.7
87
167
-
Nitrified
effluent
-
-
43
0.9
2.6
13.6
43 f
107f
-
Denitrified
effluent0'
-
-
19
1.2
2.5
0.9
15
52
-
Final
effluent
2.8
2.3
4.5
0.9
1.7
0.6
8
38
-
!| Average flow to plant: 0.320 mgd (0.014 m /s.ec)
Peak daily flow to plant: 0.900 mgd (0.039 m /sec) 3
° Average flow to denitrlflcatlon columns: 0.315 mgd (0.014 m /s^ec)
Peak dally flow to denltrlflcatlon columns: 0.632 mgd (0.028 m /sec)
6 %T I i.—I f. _ TlT _l...«..n lAnn ^Vi^r\ n O rvirv/l
e Nitrite - N always less than 0.2 mg/1
* Includes methanol
9 Degrees C
9-54
-------
TABLE 9-21
SUBSEQUENT PERFORMANCE OF FINE MEDIA DENITRIFICATION COLUMNS
AT EL LAGO, TEXAS - OCTOBER 1 THROUGH OCTOBER 31, 1974
Constituent
Total P9
Soluble P
SS
NH|-N
TKN
NO3-N
BOD5
COD
Temperature
Mean value, mg/1 at indicated sample location
Primary .
Influent3'
12
1.8
295
-
-
-
~
-
Primary
effluent
3.6
1.0
51
18
24
-
~
-
Nitrified
effluent
3.1
0.41
81
0.4
2.6
15
62e
113e
21
Denitrified
effluent0 -d
Column No. 1
-
-
51
-
-
1.9
16
44
-
Denitrified
effluentc-d
Column No. 2
-
-
44
0.4
1.5
0.9
12
36
-
Final
effluent
0.41
0.40
1
0.3
0.9
0.6
3
19
-
3 Average daily flow to plant: 0.301 mgd (0.013 mVsec)
Peak daily flow to plant: 0.47 mgd (0.021 m3/sec)
^Average flow to denitrification columns: 0.282 mgd (0.012 m3/sec)
Peak daily flow to denitrification columns: 0.470 mgd (0.021 m /sec)
, Includes methanol
Degrees C
gReduction of P due to forric chloride addition to primary and nitrification step (37 mg/1 as Fe), Also polymer,
DOW A-23, added to primary at 0.23 mg/1 and to tertiary filter at 0.17 mg/1.
9.5.3 Case Examples of Breakpoint Chlorination for Nitrogen Removal
Two examples of the use of breakpoint chlorination for nitrogen removal are presented in
this section. The Sacramento Regional County Sanitation District's plant will incorporate
breakpoint chlorination of approximately one-half of the plant effluent to achieve the
partial nitrogen removal dictated by plant effluent requirements. The Montgomery County,
Maryland facility is designed for breakpoint chlorination of the entire flow to meet rigid
limitations on total nitrogen set on the plant's effluent.
9.5.3.1 Sacramento, California
The proposed Sacramento Regional Wastewater Treatment Plant will be owned and
operated by the Sacramento Regional County Sanitation District to serve the City and
County of Sacramento. The Regional Plant is designed for 125 mgd (5.43 m /sec) average
seasonal dry weather flow. It consolidates 23 existing plants presently discharging to the
Sacramento and American Rivers above Sacramento and provides for discharge to the
Sacramento River, downstream. At minimum river flows maintained by upstream dam
development, the 125 mgd average daily flow will be about 2.7 percent of the total river
flow at the plant discharge site.
9-55
-------
Effluent from the plant will comply with waste discharge requirements adopted by the
California Regional Water Quality Control Board on October 25, 1974. Effluent quality
requirements require BOD and suspended solids to average less than 30 mg/1 on a monthly
basis. Total nitrogen is limited to 15 mg/1 when Sacramento River flow is below 12,000 cfs
(340 m /sec) at a specified gauging station. The 15 mg/1 total nitrogen requirement under
low flow conditions in the receiving water is to reduce algae blooms. Based on long-term
rainfall and river flow data, it is anticipated that nitrogen removal will be required for an
average of 67 days per year. »4" One consideration in developing design criteria for the
regional plant was to find the most cost-effective solution for intermittent nitrogen removal.
In the regional facility mixed municipal and cannery waste will undergo treatment steps
including prechlorination, preaeration, grit removal, primary sedimentation, oxygen
activated sludge treatment, secondary sedimentation, post aeration to strip carbon dioxide
and raise pH, chlorine disinfection, and sulfur dioxide dechlorination. For intermittent
nitrogen removal at Sacramento, breakpoint chlorination was most cost-effective. In
comparison to biological nitrification-denitrification, breakpoint chlorination was cost-
effective if the total nitrogen limitation did not exceed a duration of about 300 days per
year. Breakpoint via on-site production of hypochlorite solution from direct mixing of
liquid chlorine and caustic was cost-effective in comparison to electrolytic generation of
hypochlorite if the duration of the nitrogen limitation did not exceed about 200 days per
year.46
Most significant in the cost analysis is the capital expenditure to meet the maximum
chlorination capacity. Calculations indicate the maximum required chlorination capacity for
breakpoint will be 170 tons per day (154,000 kg/day). This figure represents treating the
entire plant flow. The average required chlorination capacity for breakpoint is 78 tons of
chlorine per day (70,700 kg/day). One alternative for meeting maximum required capacity
would be to provide 42 standard 8,000 Ib/day (3,600 kg/day) vacuum chlorinators.
Alternatively, generation of hypochlorite using the electrolytic process would require 51
units of 3.3 ton/day (2,990 kg/day) capacity as well as brine and salt storage. Another
alternative was reviewed, that employed by the Los Angeles County system which utilizes
liquid chlorine and base mixed in a water stream. This system appeared to provide the
required maximum operating flexibility with minimum investment cost. Thus, it was decided
to feed liquid chlorine and caustic soda into a recirculated effluent water stream directly
forming hypochlorite.
Liquid chlorine is available in rail tank cars from San Francisco Bay area producers as well as
Washington state. The rail facilities allow flexibility of plant deliveries and eliminate the
need for permanent on-site chlorine storage tanks.
Design criteria for the breakpoint facilities at Sacramento are summarized in Table 9-22.
The breakpoint system is designed as two complete and separate units each capable of
chlorinating half of the plant flow. This allows breakpoint chlorination of only the portion
of the plant flow required to meet the 15 mg/1 total nitrogen limitation. It is estimated that
9-56
-------
TABLE 9-22
DESIGN CRITERIA FOR HYPOCHLORITE PRODUCTION FACILITY
SACRAMENTO REGIONAL WASTEWATER TREATMENT PLANT
Plant Loading
Flows
Maximum hourly
Average dry weather
Average seasonal dry weather
Total nitrogen (Influent)
Maximum hourly
Average dry weather
Chemical requirements
Chemical ratios
Ammonia nitrogen to total nitrogen
Chlorine to ammonia nitrogen ratio
Maximum
Average
Caustic to chlorine
Maximum
Average
Chemical feed rates
Chlorine
Maximum
Average8
Caustic
Maximum
Average a
Chemical storage and delivery
Chlorine storage
Number single unit railroad tank car spots
Maximum capacity, each unit
On-line storage capacity
In-plant storage capacity
Sodium hydroxide storage
Number tanks
Size, dla. x height
Capacity, each tank (25% cone.)
Number single unit railroad tank car spots
Maximum capacity, each unit (50% cone.)
Sodium hydroxide feed pumps
Number
Capacity, each unit (25% cone.)
Padding air compressors
Number
Capacity, each unit
Discharge pressure
Sodium Hypochlorlte Generators
Number
Maximum capacity, each unit
Channel mixers
Number
Capacity
Ammonia analyzers
Chlorine analyzers, total
Chlorine analyzers, free
Water supply pumps
Number
Capacity, each unit
150 mgd (6.57 rnVs)
110 mgd (4.82 m3/s)
125 mgd (5.48 mVs)
37 mg/1
33 mg/1
0.75 : 1.00
10.0 : 1.0
9.0 : 1.0
1.3 : 1.0
1.0 : 1.0
170 tons/day (154,195 kg/day)
78 tons/day ( 70,750 kg/day)
221 tons/day (200,450 kg/day)
78 tons/day ( 70,750 kg/day)
12
90 tons (81,630 kg)
5 days
7 days
40 ft x 20 ft (12.2m x 6.1 m)
188,000 gal (711,580 1)
2
90 tons (81,630 kg)
75 gpm (4.73 1/s)
2 3
130 scfm (61.3 m /s)
220 psig (15.2 bar)
85 tons/day (77,100 kg/day)
40 hp (29.8 kW)
2
2
2
800-2100 gpm (50.5-132.5 1/s)
Represents breakpoint chlorlnatlon of three-quarters of average plant flow.
9-57
-------
both units will be required to be operational at certain times to maintain effluent
requirements. Figure 9-20 is a simplified schematic of the breakpoint chlorination system
and its control instrumentation. The process consists of producing sodium hypochlorite by
inline mixing of liquid chlorine, water, and caustic at a chlorine concentration of
7,000-8,500 ppm and a pH of .7.5. Maintenance of proper pressures in the chlorine feed
system is the key to successful operation of the hypochlorite generation system. Experience
at Los Angeles County White Point plant indicates a necessary pressure of 40 psig (2.8
kgf/cm2) in the mixing tee and greater than 110 psig (7.73 kgf/cm ) at the chlorine feed
valve. These pressures require pressurizing ("padding") the chlorine railcars with 175 psig
(12.30 kgf/cm2) dry air.
A chlorine reserve tank is used to provide chlorine during periods when empty chlorine tank
cars are being replaced with full cars and can provide about a one-hour supply of chlorine.
The feed rate of liquid chlorine is measured with a steel tube rotameter and controlled with
an automatic valve. Pressure drop from the chlorine feed line to the mixing tee and initial
mixing of chlorine, caustic and water is partially accomplished using a plug injector. To
improve mixing immediately downstream of the mixing tee, an inline powered mechanical
mixer is provided. The hypochlorite solution line then includes three taps for pH metering
and a back pressure valve to maintain 40 psig at the mixing tee. The pH meters are
duplicated for each unit. The duplication is for improved reliability of the pH system and
the multiple taps are for additional flexibility in selecting the point which is monitored for
control purposes.
The related feeds of caustic and water are provided by pumping from the chemical storage
area and the effluent channel. Caustic is fed from storage tanks through centrifugal pumps,
automatic control valves, and magnetic flow meters. The control signal is proportional to
the chlorine feed with feedback control provided by the pH meters. Manual caustic feed is
provided for breakpoint start-up and for pH adjustment of final effluent.
Plant effluent is used for sodium hypochlorite solution water and is provided by two
variable speed pumps located as shown in Figures 9-20 and 9-21. The control of the solution
flow (via pump speed) is proportional to the chlorine feed. The solution water flow is
measured by magnetic flow meters just ahead of the mixing tee.
The plan view of Figure 9-21 also shows the application points for the breakpoint
chlorination located at the end of each battery of secondary sedimentation tanks. To insure
rapid mixing, two spargers are installed with orifices that insure a minimum exit velocity of
about 10 fps (3.1 m/sec). Automatic valving provides one sparger for lower feed rates and
two when flows cause back pressure to reach the control limit of the hypochlorite back
pressure valve. Immediately downstream of the breakpoint spargers is a mechanical mixer
followed by a submerged overflow weir. The mixing channel and the post-aeration channel
following are covered. The exhaust can be passed through mobile activated carbon filters to
remove odors, such as nitrogen trichloride, if necessary. The activated carbon filters are
9-58
-------
FIGURE 9-20
HYPOCHLORITE GENERATION SCHEMATIC - SACRAMENTO
REGIONAL WASTEWATER TREATMENT PLANT
CONTROL PRESSURE INDICATOR ^
f TRANSMITTER Q-H
TOTAL Clj RESIDUAL
BREAKPOINT
MONITOR ANALYZER
CHLORINATED
EFFLUENT
SECONDARY
tFFLUEN
BREAKPOINT
MIXER
DISINFECTION
MIXER
-------
FIGURE 9-21
PLAN AND SECTION OF THE BREAKPOINT FACILITY AT THE
SACRAMENTO REGIONAL WASTEWATER TREATMENT PLANT
BATTERY JL
SECONDARY
EFFLUENT
BATTERY I
SECONDARY
EFFLUENT
VO
ON
o
_._.
-BREAKPOINT MIXER
KEY
-BREAKPOINT MIXER
EXHAUST GAS SYSTEM
DISINFECTION MIXER
/-WATER SUPPLY
/ PUMPS
o
7
CHLORINATED
EFFLUENT
[A] CI2 RESIDUAL ANALYZER
TO AMMONIA-
NITROGEN 40 HP
ANALYZER BREAKPOINT
MIXER
PLAN VIEW-POST AERATION CHANNEL
NO SCALE
DISINFECTION
TOTAL CI2
RESIDUAL ANALYZERS-
BOO-2100 GPM
WATER SUPPLY
PUMP
33/75 HP DISINFECTION
MIXER
MOBILE
VENTILATION^^.
BREAKPOINT
TOTAL f FREE CI2
RESIDUAL ANALYZERS
SECONDARY
EFFLUENT BREAKPOINT
SPARGER
-DETENTION. TIME = 25 Sec
kAT 62.5 MGO
DISINFECTION/ ZDETENT|ON TIME , 20 S.C*
SCHEMATIC SECTION- POST AERATION CHANNEL
NO SCALE
-------
mobile units that are identical to those used throughout the plant in various applications.
When the carbon is exhausted, the mobile units are replaced with duplicate units and taken
to a central carbon handling facility for bed replacement. Breakpoint chlorination control is
closed loop. Feed forward control includes ammonia nitrogen concentration, effluent flow,
and a ratio proportioner of chlorine to ammonia nitrogen. Feedback control is achieved by
free chlorine residual measured after the breakpoint reaction. Ammonia nitrogen samples
are pumped continuously from each battery upstream of the hypochlorite sparger to
ammonia nitrogen analyzers. Chlorine residual analyzers are of the amperometric type and
are in duplicate. One analyzer measures free chlorine residual and is used for control; the
other analyzer measures total chlorine residual and is used as a monitor. Separate provision
is made for feeding chlorine for disinfection when breakpointing is not being practiced. This
was done because different magnitudes of chemicals are involved when dealing with
breakpoint than with disinfection. Estimates for capital and operating costs for the
breakpoint chlorination facility are in Tables 9-23 and 9-24.
TABLE 9-23
CAPITAL COST BREAKDOWN FOR BREAKPOINT CHLORINATION AT
THE SACRAMENTO REGIONAL WASTEWATER TREATMENT PLANT
Item
Breakpoint generation equipment
Outside piping
Chlorine unloading facilities
Caustic storage and pumping
Railroad
Air padding facilities
Subtotal
Engineering and contingencies
Total
Estimated cost9
$ 95,000
240,000
110,000
162,000
244,000
55,000
$ 906,000
272,000
$1,178,000
aCost basis: October, 1974
TABLE 9-24
TOTAL ANNUAL COST BREAKDOWN FOR BREAKPOINT CHLORINATION AT
THE SACRAMENTO REGIONAL WASTEWATER TREATMENT PLANT
Chemical cost '
Labor cost
Subtotal
Amortization of capital :
Total annual costc
1,000/yr
1,265
30
1,295
101
1,396
$/mil gal treated
172.00
4.00
176.00
13.50
189.50
$/mil gal - annual average
31.50
3.00
34.50
2.50
37.00
Costs are based on chlorine @ $144/ton and caustic soda @ $168/ton, October 1974 prices;
average plant flow of 110 mgd; and'breakpoint chlorination required 67 days/aver, yr. at
an average of three quarters of the plant flow.
Power costs negligible
Q
Capital recovery at 7 percent and 25 years
9-61
-------
9.5.3.2 Montgomery County, Maryland
This new 60 mgd (2.6 m3/sec) facility is now being designed. ° It will be owned and
operated by the Washington Suburban Sanitary Commission to serve the Maryland suburbs
of Washington, D.C. The plant will discharge to the Potomac River above the raw water
intakes for the Washington, D.C. water treatment plants. " At minimum river flows, the
effluent will make up about 15 percent of the water volume at the water plant intakes.
Because of the critical nature of the downstream water use, the following effluent goals have
been established:
Parameter Value
BOD5, mg/1 1.0
Suspended solids, mg/1 0
Total nitrogen (as N), mg/1 2.0
Total phosphorus (as P), mg/1 0.1
Chloride, mg/1 200-350
Total dissolved solids, mg/1 850-1,120
Coliform bacteria - MPN/100ml 2.2
Fecal coliform bacteria - MPN/100ml 2.2
Figure 9-22 is a flow diagram of the treatment process. The final effluent will be stored in a
reservoir with a 10 day holding capacity before discharge to the Potomac. The lime sludges
will be recalcined and reused and the granular carbon will be regenerated on-site for reuse.
The breakpoint process was selected over alternate approaches for several reasons. It was
felt that the results of the first scale-up of the selective ion exchange process underway at
the nearby Upper Occoquan (see Sec. 9.5.4.1) plant in Virginia should be available before a
60 mgd (2.6 irr/sec) selective ion exchange facility was attempted.
The prolonged cold weather periods made ammonia stripping inadequate for substantial
portions of the year. The biological approach seemed to offer no significant cost savings and
was prone to occasional inefficiencies which would be particularly significant because of the
effluent quality goals. The effluent TDS additions from the breakpoint process were not a
limitation in this case because the effluent does not enter a reservoir or confined watershed
where the solids would be recycled and continue to build up but enters the Potomac River
shortly before it flows into the Potomac Estuary.
Concern over the hazards of transporting and storing large quantities of chlorine gas led to
the decision to use sodium hypochlorite for the breakpoint process. An investigation of the
availability and cost of sodium hypochlorite led to a decision to install an on-site
hypochlorite generation facility of the membrane cell type.51>52 For this electrolytic
process of sodium hypochlorite generation, the raw materials required are electrical power,
salt, and water. The membrane cells and the overall system are shown schematically in
9-62
-------
FIGURE 9-22
FLOW DIAGRAM OF THE MONTGOMERY COUNTY, MARYLAND PLANT
1
1
1
1
Primary
Settling
Watte Organic
Sludge
Incineration
T"
1
1
t
Activated
Sludge
"t
Lime
Recovery
Secondary Litr" Coagulation
Settling f Phosphor ut'Removol
1
— — Recvcle —
Recarbonat ion
* to pH 9,3
1
t
Plant
Effluent
Retervoir
(Optional)
Supplemental
Chlorination
(at needed)
Granular
Carbon
Adtorption
-
BP Chlorination
for
NrlJ-N Removal
Oitchorge
to Potomac
River
LI
Carbon
Regeneration
Figures 9-23 and 9-24. Saturated brine is fed to the anode compartment of the cells. At the
anode surface, chlorine gas is generated. The effluent from the cathode compartment is sent
to a gas-liquid separator in which the hydrogen is removed from the caustic solution.
Chlorine and caustic are fed to the reactor, with the caustic in slight excess, to produce
sodium hypochlorite. The reactor is water cooled to avoid decomposition of hypochlorite
and formation of oxygen. Insoluble gases, mainly oxygen, are removed in a gas-liquid
separator before the hypochlorite is sent to product storage. Brine feed for the system
comes from a salt storage tank which is used for salt storage and production of saturated
brine. The cell typically produces hypochlorite with 8 percent available chlorine at 1.7
kwh/lb-(0.77-kwh/gm) available chlorine.
The design criteria for the hypochlorite production facility at Montgomery County are
summarized in Table 9-25. Concentrated brine is created by combining softened water
(softened to minimize potential for scaling of the membrane with calcium) and solid salt to
achieve a saturated brine solution (approximately 26 percent salt by weight). Both the solid
salt storage and salt solution are contained in lixators which are fitted with a brine
collection system and a brine level control system. Delivery of the solar-type salt to the
lixators is by 22 ton (19.95 metric tons) dump trucks. The brine is pumped from the
lixators to a brine treatment system, consisting of successive addition and mixing of caustic
soda (NaOH) and soda ash (Na2CO3) followed by settling to precipitate calcium and
magnesium. The brine is then filtered through a rapid sand filter and a cartridge filter to
remove suspended solids and pumped to storage. Storage for a one-day supply of treated
9-63
-------
FIGURE 9-23
MEMBRANE CELL USED FOR HYPOCHLORITE PRODUCTION
CHLORINE*
T BRINE
ANODE »»
MEMB
i
BRIMF ——
RANE'^
i
,
HYDRC
^CATHODE
N
i
t
WATER
OR
DILUTE BRINE
brine is provided. The direct current source required by the membrane cells is provided by a
solid state rectifier, with side-mounted controls. Safety features in the control panel include
automatic shutdown in case of high reactor temperature, high DC voltages, blower failure,
or high or low DC current. The hydrogen from the electrolytic cells can be either (1)
diluted with air as it is formed to maintain a hydrogen concentration of less than 0.25
percent by volume, the explosive limit, or (2) compressed and piped to the solids processing
building for use as a fuel. Sodium hypochlorite leaves the reactors by gravity flow to pumps
from which it is sent to storage.
The breakpoint reaction is accomplished by adding a sodium hypochlorite solution to the
wastewater at a dosage slightly in excess of the stoichiometric requirement for oxidation of
the ammonia nitrogen to gaseous nitrogen. Figure 9-25 illustrates the design of the system
and Table 9-26 presents the design criteria. Filter effluent flows by gravity to the breakpoint
chlorination reactors.
9-64
-------
FIGURE 9-24
OVERALL SYSTEM USING MEMBRANE CELLS FOR HYPOCHLORITE PRODUCTION
CHLORINE
BRINE-
WATER-
CELL
STACK
SEPARATOR
SPENT BRINE
HYDROGEN
VENT
TlNERTS
REACTOR
r
WASTE
VENT
SEPARATOR
CAUSTIC
COOLER
PRODUCT
The wastewater first passes through two in-line mechanical mixers in series. The sodium
hypochlorite is added to the first mixer along with sodium hydroxide if needed for pH
adjustment. The second mixer is provided for protection if the first mixer malfunctions. The
second mixer is normally operated to provide thorough mixing. The mixer was sized so that
a single mixer would provide violent mixing (G = 1,000 sec~l) to insure instant and
complete mixing of the hypochlorite and the wastewater.
Wastewater then flows to breakpoint reactors (closed concrete tanks) where it is air mixed
to complete the chemical reaction and where 30 minutes contact time for disinfection is
provided. Flow is distributed over the first half of the length of each basin through a
multiport header. Distribution of flow in this manner minimizes the decrease in pH caused
by the reaction of sodium hypochlorite with ammonia nitrogen (by avoiding a single point
of injection into the basin of the wastewater-hypochlorite mixture) and thereby minimizes
the formation of nitrogen trichloride. Air is diffused into the wastewater over the bottom of
9-65
-------
TABLE 9-25
DESIGN CRITERIA FOR HYPOCHLORITE PRODUCTION
FACILITY AT THE MONTGOMERY COUNTY FACILITY
Average dry weather flow
Sodium chloride storage and treatment
1. NaCl required @ 60 ton/day C12 production8
2. Salt storage
Liquid level
Storage capacity
Maximum saturated brine production
3. Type of salt used
4. Brine treatment system
A. Soda ash requirements
3 minute rapid mix
B. NaOH requirements
3 minute rapid mix, 3 hour settling
C. Rapid sand filters and cartridge filtration
follows settling
5. Treated brine storage, fiberglass tanksb
6. Truck delivery
7 trucks/day/5 day week
Sodium hypochlorite generators
1. Generating capacity
2. Modules required @ 3.3 tons each
On line
Redundant
3. Rectifiers required
4. Power required
5. Building required
Builder provides for housing, modules, and rectifiers.
Monorail with 5 ton capacity provided for cell stack
maintenance.
Rectifiers separated from modules by a glass wall to
prevent corrosion.
6. Generator module dimensions
7. Rectifier dimensions
Sodium hypochlorite storage
1. Storage sized to provide 1 day storage for power outage
plus storage for 7 days @ 90 mgd (3.94 m3/sec)
2. Storage capacity
60 mgd (2.63 m3/sec)
105 tons/day (95.2 metric tons/day)
4 @ 22 ft x 33 ft x 18 ft (6.7 m x 10.1 m x 5.5 m)
1,470 tons (1333 metric tons)
60 gpm (3.8 I/sec) each (min.)
Solar
450 Ib/day (204 Kg/day)
3 ft x 3 ft x 3 ft swd (0.91 m x 0.91 m x 0.91 m)
360 Ib/day (163 Kg/day)
3 ft x 3 ft x 3 ft swd (0.91 m x 0.91 m x 0.91 m)
6 @ 12 ft x 12 ft (3.66 m x 3.66 m)
16,000 gal each (60,560 1)
22 tons/load (19.95 metric tons)
60 tons/day (54.4 metric tons)
20
18
2
7
10,000 KVA
87 ft x 96 ft x 20 ft (26.5 m x 39.2 m x 6.1 m)
4 ft x 16 ft x 8 ft (1.2 m x 4.9 m x 2.4 m)
4 ft x 6 ft x 8 ft (1.2 m x 1.8 m x 2.4 m)
700,000 gal (2650 m3)
230 tons equivalent (209 metric tons)
6-117,000 gal ground level tanks (443 m3)
6-30 ft x 22 ft tanks (9 .1 m x 6.7 m)
Assumes 90 percent utilization
Each tank will provide 4 hours storage at full production
the second half of the basin to strip any nitrogen trichloride from solution. Additional air
may be introduced in the space between the liquid surface and the basin cover to further
dilute any nitrogen trichloride that might be present to prevent the development of an
explosive concentration that occurs at 0.5 percent by volume. The diffusion of air into the
breakpoint reactor contents also strips gaseous nitrogen and carbon dioxide from solution.
The latter will result in a desirable increase in pH.
Exhaust gases from the reactors are recirculated to provide mixing with some of the gas bled
off to the recalcining furnace for thermal decomposition of the nitrogen trichloride. The
alkaline environment in the recalcining furnace will avoid discharge of hydrochloric acid
(HC1) to the atmosphere that might otherwise occur.
9-66
-------
FIGURE 9-25
SCHEMATIC OF MONTGOMERY COUNTY, MARYLAND
BREAKPOINT CHLORINATION PROCESS
o\
ISOLATION VALVE
METER
BALL VALVE
FROM
FILTERS /
(One of Six/
lines shown)
C|RCULA|ON
GAS
^BLEED-OFF TO
RECALCINERS
EXHAUST
GAS /--EFFLUENT
^ WEIR
INLET DIFFUSER
AIR HEADERS
ISOLATION
INLET GATES
n
N >*1t
-^"^
^-
One of Six Basins shown
UJ
z
0
" J
•3f
u. L»-
UJ
TO
CARBON
PUMP
STATION
-------
TABLE 9-26
BREAKPOINT CHLORINATION DESIGN CRITERIA
FOR THE MONTGOMERY COUNTY FACILITY
Average dry weather flow
In-line mixers
1 . Lines normally in service
2 . Lines normally on standby
3. In-line mixers
Size
Mixers per line
HP
G
4 . Flow rate through mixer with 5 in service
Average
Maximum
Peak instant
Chemical feed
NaOCl
Feed rate @ 8% available
service
Total feed rate per basin
and with 5 units in
NaOHa
Feed @ 20% NaOH with 5 units in service (total)
3.
Feed rates based on:
Influent NHj-N
C12 : NHj-N
Reaction basins and air mixing
1.
2.
3.
4.
Basins normally in service
Basins normally on standby
Basin dimensions
Volume
Theoretical mix time
Diffuser air requirements
Total
Per basin
Air headers and diffusers
Main headers
Cross headers
Diffusers
60 mgd (2.63 m3/sec)
5, 3 ft dia. (91.4 cm) influent lines
1
36 in. (91.4 cm)
2 (in series)
3 HP/Mixer (2.23 Kw), 6 HP/Line (4.46 Kw)
1000 sec'1
12.81 mgd (0.56 m3/sec)
19.21 mgd (0.84 mVsec)
21 .34 mgd (0.93 mVsec)
210,000 gpd @ 90 mgd (552 1/min @ 3.94 mVsec)
32.5 gpm @ 100 mgd (123 1/min @ 4.38 m3/sec)
30.0 gpm @ 90 mgd (113 1/min @ 3.94 m3/sec)
20.0 gpm @ 60 mgd (75.7 1/min @ 2.63 mVsec)
15 mg/1 average
52 mg/1 max. w/no alkalinity
5500 gpd @ 90 mgd @ avg. (14.5 1/min @ 3.94 mVsec)
21,000 gpd @ 100 mgd @ maximum (55 1/min @ 4.38 mVsec)
18.7 mg/1
10:1 by weight
5
1
20 ft x 120 ft x 15 ft swd (6.1mx36.6mx4.6m)
36,000 gal (136 m3)
5 basins in service
30 min @ average flow
20 min @ max. flow
30 scfm/1,000 cu ft (30.3 1/m3)
5,400 scfm (150 m /min) - 5 basins in service
1,080 scfm (30 mVmin)
2 (one at each end) 8 in. dia. (20.3 cm)
540 scfm normal (15 m3/min)
1,080 scfm max. (30 m3/min)
24, 2.5 in. dia. (6.4 cm) 455 scfm (12.6 m3/min) normal
90 scfm max. (2.5 mVmin)
11/header, 264/basin, 2 ft (0.61 m) O.C. on cross headers,
4.1 scfm (0.11 m3/min) normal, 8.2 scfm max. (0.22 mVmin)
7 . Blowers
Normally in service
Normally on standby
Capacity
HP
8. Inlet diffuser
Diameter
Material
Length
Inlet ports
2
1
3,000 cfm (85 mVmin) @ 8 psi (0.60 kgf/cm2) ea.
150 HP each (112 Kw)
48 in. (121.9 cm)
FRP
42 ft 6 in. (12.95 m)
8 @ 3 ft 0 in. (0.9 m) O.C.
NaOH needed only when alkalinity is 150 mg/1 and no air stripping.
Basins covered to contain product gases and lined with corrosion protection membrane to 1 ft (30.5 cm) below water surface.
9-68
-------
The breakpoint process is controlled by pacing the sodium hypochlorite feed rate to the
influent flow and influent ammonia nitrogen concentration. The pH is also monitored and
controls the addition, when necessary, of sodium hydroxide to maintain an optimum pH for
the breakpoint reaction.
Ammonia nitrogen in the breakpoint effluent is monitored to determine efficiency of the
process. Free and combined chlorine residuals and pH are also continuously monitored in
breakpoint reactor effluent.
The estimated costs of the Montgomery County facility are shown in Table 9-27. The
hypochlorite generation facility will also provide hypochlorite for uses other than
breakpoint although its entire cost has been shown for the breakpoint process. As noted in
Table 9-27, the hydrogen liberated in the breakpoint process has a potential value of
$10.00/mil gal ($0.0026/m3) if it is collected and used as a fuel.
TABLE 9-27
ESTIMATED COSTS OF BREAKPOINT CHLORINATION
AT THE MONTGOMERY COUNTY PLANT
Capital
a,b
Breakpoint reaction basins
Operations building (including off-gas
treatment and blowers)
Hypochlorite pla nt
Storage
Generation
Salt dissolution
Total
$2,300,000
565,000
817,000
3,780,000
705,000
$8,167,000
Operation and maintenance
Power
Hypochlorite production
Mixing, stripping
Salt
Labor
Subtotal
Amortized capital, $8,167,000, 20 years
@ 7% @ 60 mgd
$ 26.20/mil gal
18.00
25.20
12.60
$ 82.00 ($0.022/m3)
35.20 ($0.009/m3)
$ 177.20 ($0.031/m3)
Not including contingencies and engineering
December, 1974 cost levels
9-69
-------
9.5.4 Case Examples of Selective Ion Exchange for Nitrogen Removal
Two examples of the incorporation of ion exchange into wastewater treatment plant layouts
are presented in this section. In the case example for Upper Occoquan Sewage Authority,
ion exchange is used in a tertiary treatment step following biological treatment. The
objective of this plant is to meet an effluent limitation of 1 mg/1 total nitrogen. In the
Rosemont case, ion exchange is used in a physical-chemical flowsheet to meet an effluent
limitation of 1 mg/1 ammonia nitrogen.
9.5.4.1 Upper Occoquan Sewage Authority, Va.
This new regional plant now under construction will replace 11 small secondary plants
which discharge into tributaries of a water supply reservoir which serves as the raw water
source for water treatment plants serving about 500,000 people in the Virginia suburbs of
Washington, D.C. Information about the water reuse aspects of the project is available in
reference 49. The effluent eventually reaches a water supply reservoir in which nitrogen is
believed to be one of the principal eutrophication factors. The effluent standards are shown
below:
Parameter Value
BOD5, mg/1 1.0
COD, mg/1 10.0
Suspended solids Unmeasurable
Phosphorous, mg/1 0.1
Methylene Blue Active Substances
(MBAS), mg/1 0.1
Turbidity, JTU 0.4
Coliforms, total/100 ml 2
Nitrogen, total 1.0 mg/1
The main processes which are included in this plant are shown in Figure 9-26.
The initial plant capacity will be 22.5 mgd (0.99 m^/sec), although an initial daily capacity
of 10.9 mgd (0.48 m3/sec) is all that will be initially certified by the State of Virginia so as
to provide 100 percent complete backup facilities in the initial operation. Provisions are
included in the Virginia State Water Control Board regulatory policy to increase the rated
capacity to 15.0 mgd (0.66 m^/sec) after a one year satisfactory demonstration period.
Backup facilities would then constitute approximately 50 percent of the rated capacity.
The methods of nitrogen removal of biological, ammonia stripping, and breakpoint
chlorination were evaluated before selection of the selective ion exchange process. This
process was selected primarily because of its inherent reliability and efficiency and the
9-70
-------
FIGURE 9-26
FLOW DIAGRAM - UPPER OCCOQUAN SEWAGE AUTHORITY PLANT (VIRGINIA)
, -1
1
LT-T
1 1
1
Preliminary Primary ^
Treatment *" Settling
Plant
Effluent «
Reservoir
1
Discharge to
Bull Run
^ Sludge
Digestion
— Recycle •
1
Activated Secondary L'm"
' Sludge fc Settling f * ^
1
1
Chlorination Selective
NHj Removal Exchange
i
Regenerant
Ns^""™
' Recovery
Coagulation Two-Stage
tiling for » Recarbonation
torus Removal
1
- Recycle *
Ballast
Pond
i
Granular
Adsorption
Carbon
Regeneration
minimal effect on total dissolved solids (TDS). The selective ion exchange process was
located after the carbon columns to take advantage of the available head remaining after
pumping through the pressure carbon columns. Also, as an incidental benefit of this process
sequence, the clinoptilolite will serve as a final polishing filter to remove the small amount
of carbon fines or other suspended solids which may be present in the carbon column
effluent.
Carbon fines and other solids trapped in the clinoptilolite bed are removed by backwashing
before the regeneration cycle. These backwash wastes will be returned to the treatment
process, typically to the chemical coagulation process where these solids will be removed in
the precipitation process and be trapped in the chemical sludge.
The selective ion exchange regenerant recovery was initially planned to be accomplished by
breakpoint chlorination of the ammonium using electrolytic cells (see Chapter 7). A delay in
the project of about one year occurred following final design while awaiting the project
funding to develop. During this period, the ammonia removal and recovery process (ARRP)
described in Sections 7.3.3.2 and 8.4.1 was developed by the design engineers. Based on
pilot plant results, it was concluded that the annual operating cost for the plant could be
9-71
-------
reduced by $375,000 at a flow of 15 mgd (0.66 m3/sec) ($0.07/1,000 gal or $0.18/m3). In
addition, the electrical energy requirement would be only 10 percent of the electrolytic cell
breakpoint chlorination process needs. Also, a byproduct would be obtained in the form of
ammonium sulfate, a common chemical fertilizer.^3 Based on this information, the
Authority authorized a redesign of the regeneration facilities to incorporate the ARRP
process. The following paragraphs describe the full-scale ion exchange and regenerant
recovery facilities.
Eight ion exchange beds operate in parallel and are separated into two independent trains,
each with four beds and a common manifold. Each bed is a horizontal steel pressure vessel,
10 ft (3.05 m) in diameter by 50 ft (15.24 m) long containing a four ft (1.2 m) deep bed
of clinoptilolite (see Figures 9-27 and 9-28).^4 Each parallel train is completely
independent, including piping, instrumentation and control, and electrical supply. In
addition, certain backup facilities are available in each train such as key instrumentation and
control. Such measures were necessary to conform to the design policy for reliability
established by the Virginia State Water Control Board.
Table 9-28 is a summary of the design criteria for the ion exchange process at the future
anticipated rating of 15 mgd (0.66 m3/sec). The system will be entirely automated using
automatic valves in a manner similar to most larger water treatment plant filtration facilities.
Regeneration will be initiated either on a run time basis, volume throughput basis, or
manually. Backwashing will be done before each regeneration. Backwash water will be'
returned to the wastewater process, typically to the chemical coagulation process, or to the
plant headworks.
TABLE 9-28
DESIGN CRITERIA SELECTIVE ION EXCHANGE PROCESS FOR AMMONIUM
REMOVAL AT THE UPPER OCCOQUAN PLANT (VIRGINIA)
Flow rate
Beds in service
Beds in regeneration
Beds - backup capacity
Flow per bed
Bed loading rate
Backwash rate
Bed volumes to exhaustion
Average ammonia removal efficiency
Average influent ammonia nitrogen concentration
Average effluent ammonia nitrogen concentration
Normal concentration of ammonia nitrogen at initiation
of regeneration
Clinoptilolite size
Clinoptilolite depth
15 mgd (0.66 m3/sec)
4
2
2
3.75 mgd (0.16 mVsec)
10.82 bed volume s/hr
5.25 gpm/sf (3.6 l/m2/sec)
8 gpm/sf (5.4 l/m2/sec)
145
95%
20 mg/1
1 mg/1
2.5 mg/1
20 x 50 mesh
4 ft (1.22 m)
9-72
-------
FIGURE 9-27
PLAN AND SECTION OF ION EXCHANGE BEDS AT UPPER OCCOQUAN PLANT
SURFACE WAJSH SURFACE
f HEADER SUPPORT y-CLINO SURFACE
/" 'J
/ '
PVC SURFACE
, HEADER
I I I I
WASH
i I
QUICK OPENING MANHOLE
\
ii i i i , i i i i
: LATERAL SUPPORTS
-3T.
?r-E
18" WSP-
PLAN
-£
•TYPICAL LATERAL - 49 REQ'D
PER CLINO BED
COMBINATION
AIR
18 WSP INFLUENT HEADER
1
H_ja~i 1 1 i 1 <^ i 1 1_^ =jt,
v SURFACE WASH HEADER !
n\8 8 8 8888888888888888888 8,8 8 8 ¥8 888
II lit Ie"
.11 « - 1 j 1
25'-0" J
WSP EFFLUENT HEADER t=^
1 Ik \/
^PIPE SU
SUPPORT
-------
FIGURE 9-28
ADDED DETAILS - ION EXCHANGE BEDS AT UPPER OCCOQUAN PLANT
3" PVC
8 HOLES-7 SPACES @ 5'/2"
8-'/4" OIA. HOLES IN LATERAL
FACING DIRECTLY DOWNWARD
3" PVC
THREADED TEE
• — 3" EXTRA HEAVY
STEEL PIPE
I8"WSP
SYMMETRICAL ABOUT
TYPICAL UNDERDRAIN LATERAL
Z ROWS OF l^2" OIA. HOLES
@, 6 l/2" O.C. FOR INFLUENT
DISTRIBUTION, EACH ROW 30°
OFF VERTICAL, 86 HOLES EACH
SIDE , 172 HOLES TOTAL
18 WSP
SUPPORTING HIGH
DENSITY GRAVEL
GRADED GRAVEL
QUICK OPENING
MANHOLES
10* DIA. STEEL
CLINO BED
UNDERDRAIN LATERAL
CONCRETE FILL
9-74
-------
The beds will be regenerated with a 2 percent sodium chloride solution. The regeneration
process will be that shown in Figure 7-13 and as described in Section 7.3.3.2. The
regenerant recovery system consists of four 375,000 gallon (1420 m3) tanks, a regenerant
pumping system and associated automatic valves, two 35-ft (10.7 m) diameter clarifiers for
magnesium hydroxide removal, and 18 ARRP modules. Figures 9-29 and 9-30 illustrate the
ARRP module design. The ARRP units will be shop fabricated. The basic tower units are
12-ft (3.66 m) diameter fiberglass tanks. All materials will be fiberglass or PVC. Each tower
has a 25 HP (18.6 kw) fan. The air rate is approximately 34,000 cfm/tower (952 m3/min)
at an air to liquid ratio of 566 ft3/gal (4.15 m3/!). Tower air velocities are 300 fpm (91.4
m/min). Knitted mesh mist eliminators prevent moisture carryover from tower to tower.
The total system head loss is about 2.5-3.0 inches (6.4 cm - 7.6 cm) of which about 1.5
inches (3.8 cm) is in the media. The media is 2-inch (5.1 cm) diameter polypropylene plastic
packing (Tellerette). A summary of the regeneration system design criteria is shown in Table
9-29.
FIGURE 9-29
PLAN VIEW OF ARRP MODULE - UPPER OCCOQUAN PLANT
BOTTOM ACCESS
HATCH
4'-6" DIA
FRP DUCTING (TYP)
LADDER
SUPPORT
PLATFORM
ABOVE
SEE
FIGURE 9-30
FOR
SECTION A-A
12' DIA. FRP
TOWER (TYP)
9-75
-------
FIGURE 9-30
SECTION OF ARRP MODULE - UPPER OCCOQUAN PLANT
DUCT
SUPPORTS
TYP
KNITTED MESH
MIST ELIMINATOR
V / HATCHES
SERVICE AND
SUPPORT
PLATFORMS
ACCESS HATCH
BEYOND
PIPING DUCT
BEYOND
REGENERATION
BASIN BELOW
9-76
-------
TABLE 9-29
REGENERATION AND REGENERANT RECOVERY SYSTEM DESIGN
CRITERIA AT THE UPPER OCCOQUAN PLANT (VIRGINIA)
Regeneration system
Number of regenerant tanks
Size of each tank
Number of beds regenerated at once
Number of regeneration cycles per day
Regeneration bed volumes
Regenerant recovery system3
Recovery system flow rate
Operation time per day
Clarifiers
Number of units
Diameter
Overflow rate
Ammonia removal and recovery process3
Number of ARRP modules
Liquid loading rate
Air to liquid loading rate
Media height
Removal efficiency at 10 C
at 20 C
375,000 gals (1420 mj)
2
3.58
39-44
1 ,080 gpm (68 I/sec)
16 hr
35 ft (10.66 m)
800 gpd/sf (32.6 m3/m2/day)
18
760 gpd/sf (31.1 m3/m2/day)
566 cf/gal(4234 m3/m3)
7.S ft (2.29 m)
90%
95%
At 15 mgd flow rate
The average ammonia-nitrogen concentration from ion exchange beds will be about one
mg/1. The organic nitrogen is expected to be 0.5-0.8 mg/1 and the nitrate nitrogen is
expected to be from 0.1-0.2 mg/1. Thus, the total nitrogen leaving the ion exchange process
will be about 1.6-2.0 mg/1. Since the discharge standard is 1.0 mg/1, additional nitrogen
removal is necessary. This will be accomplished by breakpoint chlorination of the ion
exchange effluent. A dosage of approximately 8-10 mg/1 will result in nearly complete
removal of ammonia nitrogen. The final effluent is then expected to have a total nitrogen
concentration of less than 1 mg/1.
The estimated costs are shown in Table 9-30. Since the initial constructed capacity of 22.5
mgd (0.99 m^/sec) may be operated at no more than 15 mgd (0.66 m^/sec) because of state
requirements for backup capacity, costs are shown on the basis of operation of two-thirds of
the constructed capacity and at the full constructed capacity, which would be the more
generally applicable circumstance. The income from sale of ammonium sulfate is based on
the lowest wholesale price in effect at the time of this writing,$43/ton ($39/metric ton).
In some areas, the wholesale price is as high as $65/ton ($59/metric ton).
9.5.4.2 Rosemount, Minnesota
This new 0.6 mgd (2271 cu.m/day) plant operated by the Metropolitan Sewer Board of the
Twin Cities area (Minneapolis-St. Paul) provides independent physical-chemical treatment of
a municipal wastewater.55,56 This plant is the first full-scale physical-chemical plant to be
3-77
-------
TABLE 9-30
ESTIMATED COSTS OF SELECTIVE ION EXCHANGE
AT THE UPPER OCCOQUAN PLANT (VIRGINIA)
Item
a
Operating and maintenance
Chemicals
NaOH
NaCl
H2SO4
Income from sale of (NH^ SO4 @ $43/ton
Net chemical cost
Power, 18 HP/mil gal @ $0.0192/kwh
Labor
Total, O & M
Capital3
$4,470,000, 20 years @ 7%
Total annual costa
Estimated costs, $/mil gal
at 15 mgd
(0.66m3/sec)
$ 26.80
7.10
9.80
$ 43.70
$(12.60)
31.10
6.90
17.70
$ 55.70
$ 77.22
$132.92
($0.035/m3)
at 22.5 mgd
(0.99 m3/sec)
$ 26.80
7.10
9.80
$ 43.70
$ (12.60)
31.10
6.90
17.70
$ 55.70
$ 51.59
$107.29
($0.027/m3)
August, 1974 costs
placed in operation in the U.S. A schematic of the process is shown in Figure 9-31. The
entire plant is enclosed in a single 14,500 sq ft (1347 m^) steel building. Final effluent
standards are as follows:
Parameter
, mg/1
Suspended solids, mg/1
COD, mg/1
Ammonia — Nitrogen, mg/1
Phosphorus, total (P), mg/1
PH
Value
10
10
10
1
1
8.5
9-78
-------
FIGURE 9-31
SCHEMATIC OF ROSEMOUNT, MINNESOTA PLANT
CHEMICALS
RAW
WASTEWATER
SOLIDS
CONTACT
CLARIFIER
DUAL
MEDIA
FILTER
BAR
SCREEN
SETTLED
SOLIDS
< a:
DU
MED
FIL
1
AL
1 A
TER
I
I
AM
EX(
SLUDGE
CONCENTRATOR
SLUDGE
DISPOSAL
SPENT
CARBON
REGENER-
ATED
CARBON
STORAGE
Selective ion exchange is accomplished by clinoptilolite in 2 of 3 downflow columns in
series, each containing a 6-ft (1.82 m) depth of clinoptilolite. When the ammonia-nitrogen
reaches 1 mg/1 in the effluent from the polishing column, the lead column is removed from
service and regenerated (upflow) and the third column placed on line. The steam process
discussed in Section 7.3.3.3 is used for clinoptilolite regeneration and ammonia recovery.
Brine is stored at 77 C and is cooled to 27 C while passing through a heat exchanger on the
9-79
-------
way to the column. The waste brine leaving the column is passed through the other side of
the heat exchanger to elevate its temperature to 71 C before entering the stripping process.
Brine temperature in the storage tanks is controlled by steam supplied to internally mounted
coils. Waste brine is collected in a mixed storage tank before being stripped. Soda ash is
added to the storage tank to elevate the pH to 12, which results in the precipitation of
calcium carbonate and magnesium hydroxide. Mixing is discontinued 20 minutes after the
pH has reached 12 to allow these precipitates to settle. The sludge is then pumped from the
bottom of the waste brine storage tank. The waste brine is then pumped to the steam
stripper at a rate of 53 gpm (3.34 I/sec) and a steam flow of 3000 Ib/hr (1498 kg/hr). The
stripper operates at an equilibrium temperature of 104 C and the condensor at 38 C. The
time required for a stripping/reclaiming operation is about 5 hr. Design criteria are
summarized in Table 9-31. The plant was in the start-up phase at the time of this writing
and no operating data were available.
TABLE 9-31
ROSEMOUNT (MINNESOTA) ION EXCHANGE DESIGN CRITERIA
Ammonium exchange columns (Two trains of 3)
Loading rate8
Clinoptilolite capacity
per unit volume
per column
Ammonia nitrogen loading rate
Ammonia removal
Clinoptilolite depth per column
Clinoptilolite size
Normal operation
Backwash rate
Regeneration system
Brine solution to columns'3
Hydraulic application rate
Volume
Strength
Temperature
PHC
Brine solution regeneration
Regeneration cycle length
Hydraulic loading rate to steam stripping tower
Tower depth
Caustic soda added
Bed rinse3
Rinse rate
Time
Ammonia recovery
Aqueous ammonia strength
Aqueous ammonia volume
Ammonia stripper
Steam @ 10 psig
Throughput
Size - diameter
height
4.2 gpm/sf (2.85 l/m2/sec), 5.6 BVAr
0.3 Ib/cu ft (4.86 kg/m3)
90 Ib (40.8 kg)
50 Ib/day (22.7 kg/day)
95%
6 ft (1.83 m)
20 x 50 mesh
2 columns in series, 250 BV/cycle
8 gpm/sf (5.43 l/m2/sec)
2.0 gpm/sf (1 .36 l/m2/sec)
4.5 BV
6% NaCl
71 C
11
5 hr
7 gpm/sf (4.75 l/m2/sec)
24 ft (7.3 m)
3 Ib/lb NH^-N
300 gpm (18.9 I/sec)
70 min
1%
1,000 gpd (3,785 Ipd)
3,300 IbAr (1,498 kg/hr)
53 gpm (3.34 I/sec)
3 ft (0.91 m)
18 ft (5.5 m)
Downflow
Upflow
Elevated with caustic soda
9-80
-------
9.5.5 Case Examples of Air Stripping for Nitrogen Removal
Two plants which utilize air stripping for nitrogen removal are described in this section.
Both of these, South Lake Tahoe and Orange County, use the biological-tertiary approach in
which ammonia stripping is used after biological treatment. At Tahoe, nitrogen removal is
incorporated on an experimental basis, as no nitrogen removal requirement exists. At the
Orange County Water District plant, nitrogen is removed to permit recharge of groundwater.
9.5.5.1 South Lake Tahoe, California
At South Tahoe an experimental full-scale ammonia stripping tower built to handle half of
the total plant design flow of 7.5 mgd (0.33 m^/sec) has been operated on an intermittent
basis since 1969.13,57,58,59 jhe piant flow sheet is shown in Figure 9-3. The stripping
process was installed at South Lake Tahoe as an EPA research and demonstration
installation .and not as the result of any requirement to remove nitrogen from the plant
effluent. In the absence of any need for full time operation, the purposes of this stripping
tower have included: (1) demonstration of full-scale tower efficiencies as compared to pilot
plant test results; (2) determination of cold weather operating limitations and other
operating problems with investigation of solutions to these problems and (3) collection of
data for design purposes for future expansion of this process to full plant design capacity as
well as for use in planning similar facilities at other locations. The large tower capability to
remove ammonia almost exactly duplicates the results of pilot plant operation, reaching 95
percent removal in warm weather. The cold weather operating limitations and recommended
tower design improvements have been determined as discussed in Chapter 8. The new design
criteria obtained from operating the tower have already been used to prepare plans for
stripping towers to treat wastewaters at Orange County, California*^ and the operating
experience has also provided the direction and the basis for expansion of the Tahoe nitrogen
removal facilities to meet anticipated future requirements for nitrogen removal. As discussed
later in this section and in Section 8.5, the original packed tower system has recently been
modified to provide year round, full-scale nitrogen removal.
The design data for the original packed tower are given in Table 9-32. It is a crossflow
cooling type tower modified for ammonia stripping. The overall dimensions of the tower are
32 ft (9.75 m) x 64 ft (19.5 m) x 47 ft (14.3 m) high. Water, at pH 11, was pumped to the
top of the tower by either or both of two constant-speed pumps. These pumps were
backflushed two or three times daily to minimize buildup of calcium carbonate scale in the
pump units. When the plant inflow was less than the rate at which the pumps were delivering
water to the tower, some water was recycled from the tower effluent back to the pump
suction well. This avoided the need for variable speed pump control, and at the same time
provided some .recirculation through the tower, which improved ammonia removal. At the
top of the tower, the influent water entered a covered distribution box and overflowed to a
distribution basin. The distribution basin is a flat deck with a series of holes fitted with
plastic nozzles. Further distribution of the inflow was provided by diffusion decks
9-81
-------
immediately below the distribution basin. Three other diffusion decks were provided at 6 ft
(1.82 m) vertical intervals in the fill. The tower fill provided, theoretically, 215 successive
droplet formations as the water passed down through the tower. The tower effluent fell into
a concrete collection basin which also formed the base for the tower structure. From the
collection basin, the tower effluent passed through a Parshall measuring flume into the
first-stage recarbonation chamber, where excess pumpage returned through a flap gate into
the tower pump sump to recirculate through the tower.
TABLE 9-32
DESIGN DATA, AMMONIA STRIPPING TOWER
AT SOUTH LAKE TAHOE, CALIFORNIA
Nominal capacity: 3.75 mgd (0.164 m3/sec)
Type:
Fill:
Air flow:
Cross flow with central air plenum and vertical air discharge
through fan cylinder at top of tower
Plan area, 900 sq ft (83.6 m2)
Height, 24 ft (7.3 m)
Splash bars:
Material, rough sawn treated hemlock,size,
3/8 in. x 1-1/2 in. (0.95 cm x 3.8 cm)
spacing, vertical 1.33 in. (3.37 cm)
horizontal 2 in. (5.08 cm)
Fan, two-speed, reversible, 24-ft diameter, horizontal
Water
gpm
(I/sec)
1,350
(85)
1,800
(114)
2,700
(170)
rate
gpm/sf
(l/m2/sec)
1.0
(0.68)
2.0
(1.36)
3.0
(2.04)
Air
cfm
(mVmin)
750,000
(21,000)
700,000
(19,600)
625,000
(17,500)
rate
cf/gal
(m3/m3)
550
(4,115)
390
(2,918)
230
(1,720)
Tower structure:
redwood
Tower enclosure: corrugated cement asbestos
Air pressure drop: 1/2 in. of water at 1 gpm/sf
(1.27 cm @ 0.68 l/m2/sec)
9-82
-------
Air entered the tower through side louvers, passed horizontally through the tower fill and
drift eliminators (or air flow equalizers), and entered a central plenum. At the top center of
the plenum is a 24 ft (7.3 m) diameter, six-bladed, horizontal fan. Fan blades and fan
cylinder are both made of glass-reinforced polyester. The fan takes suction from the plenum
and discharges to the atmosphere through the fan cylinder. The fan has a maximum capacity
of about 750,000 cfm (21,000 m^/min). It is equipped with a two-speed reversible 75 HP
(56 kw) motor.
The limitations of the packed tower at Tahoe are caused by the cold winter temperatures
and the scaling of the tower packing. Because the scaling problem was not anticipated from
the pilot studies, the access to the tower packing needed to remove the scale with water jets
was not provided. As a result, portions of the original hemlock packing became hopelessly
fouled with calcium carbonate scale, thereby decreasing effective tower packing area.
The costs of the operation of the packed tower at Tahoe at a 2 gpm/sq ft (1.36 l/
rate as estimated for a 7.5 mgd (0.33 m^/sec) scale for continuous operation are listed in
Table 9-33. Operating labor included backflushing of tower pumps and cleaning the
distribution deck to remove CaCC>3 precipitation, process inspection, lubrication, and daily
determination of tower ammonia removal efficiency. Maintenance costs reflect the cost of
cleaning tower fill.
Because of the cold weather limitations of packed tower inherent to the Tahoe location and
the reduced efficiency of the existing tower due to scaling of the tower packing, a research
program was initiated at the South Lake Tahoe plant to develop alternate, low-cost
techniques which could be applied to the full-scale plant while using as much of the existing
facilities as possible. Although there are no current regulatory agency requirements for
nitrogen removal for exported wastes, future requirements for effluent reuse or disposal in
the Lake Tahoe basin will probably include nitrogen removal. At the current time, the
States of California and Nevada require that all wastewaters, regardless of the degree of
treatment be exported from the Lake Tahoe basin. The South Tahoe effluent is exported to
Alpine County, California where it is used to form the 1 billion gallon (3.785 billion liters)
Indian Creek Reservoir.59 An excellent trout fishery has been established in this recreation
reservoir and it is necessary to control pH and ammonia concentrations to prevent fish
toxicity. Thus, anticipated future regulatory agency requirements and current effluent reuse
practices required that a system for full-scale nitrogen removal be developed for the South
Lake Tahoe plant.
Inspired by observations of ammonia release from holding ponds in Israel," 1 research was
undertaken at Tahoe to improve the release of ammonia from the high pH effluent. The
results led to the following three steps being applied to a full-scale modification of the
69
Tahoe system.
1 . Holding in high pH ponds (with surface agitation in one pond).
9-83
-------
2. Stripping in a modified crossflow forced draft tower through air sprays installed
in the tower.
3. Breakpoint chlorination.
Design data are given in Table 9-34 and the entire system is depicted in Figure 8-9. Because
ammonia removals by stripping vary so much with temperature, and since low temperatures,
high ammonia concentrations, and high flows never occur simultaneously at this location,
design data are presented for two sets of conditions which are expected to occur: low flow
(2.5 mgd or 0.11 m^/sec), low water temperature (3 C), and low ammonia content; and high
flow (7.5 mgd or 0.33 m^/sec), high water temperature (22 C), and high ammonia content.
TABLE 9-33
OPERATING COSTS FOR AMMONIA STRIPPING FOR
CONTINUOUS OPERATION OF TAHOE AIR STRIPPING TOWER AT 7.5 MGD
Operating cost per day
Electricityb
Operating labor
Maintenance labor
Repair material
Instrument maintenance
Total operating cost
Total cost per mil gala
Operating
Capital
Total
$/Dgy
60.78
4.63
5.17
.78
.94
72.30
$/mil gal
9.64
8.00
17.64
1970 dollars
Average cost per day at 7.5 mgd from 10 months of continuous operation, i.e.,
May, 1969, through September, 1970.
9-84
-------
TABLE 9-34
DESIGN DATA AND ESTIMATED NITROGEN REMOVALS
FOR ALL-WEATHER AMMONIA STRIPPING AT SOUTH TAHOE, CALIFORNIA
Description
Influent, pH = 11.0
Holding ponds
Detention = 7 hr
Detention = 18 hr
Air spraying in second pond
Turnovers = 4-j
Turnovers = 13|
Stripping tower, air spraying with
forced draft
Recycle turnovers = 1.6
Recycle turnovers = 5
Overall removal
NH3~N remaining to be removed by
by breakpoint chlorination, mg/1
Chlorine required for breakpoint
chlorination, Ib/day
Flow 2 . 5 mgd
(0.11 m3/sec)
Water Temp. 3 C
NH3-N
Estimated
reduction,
percent
10
30
42
64
Remaining,
mg/1
15
13.5
9.5
5.4
5.4
1,130
Flow 7 . 5 mgd
(0.33 mVsec)
Water Temp. 22 C
NH3-N
Estimated
reduction,
percent
15
28
23
40
Remaining,
mg/1
35 •
30
21.5
16
16
10,000
It is difficult to predict accurately the performance of the full-scale ponds and natural
wind-vented sprays from the results of laboratory and pilot plant data, because the
efficiency of ammonia removal by these methods depends to a great extent on the
immediate removal of the released ammonia gas from the vicinity of the air-water interface.
The efficiency of air sweeping in removing ammonia declines as the area of the ponds
increases. In the absence of full-scale plant data, the removal efficiencies obtained in the
laboratory and pilot plant*>2 were heavily discounted in estimating the performance of the
full-scale plant shown in Table 9-34. The following paragraphs describe the modified Tahoe
stripping process.
9-85
-------
The two high pH ponds also serve to equalize flow to the modified stripping tower and
breakpoint facilities. This reduces design capacity requirements for these two processes and
improves their operating characteristics. The total surface area of these ponds is about
64,000 sq ft (5946 m2). The water depth varies from 3.0 to 7.25 ft (0.9 m - 2.2 m).
Detention time ranges from 7 hr, at 7.5 mgd (0.33 m^/sec) to 18 hr at 2.5 mgd (0.11
m^/sec). The second pond in the series is provided with surface agitation in order to increase
the ammonia removal. The system of surface agitation consists of sprinkling with 34 mgd
(1.5 m^/sec) of recycled pond water through 2 pumps and a spray system consisting of
about 75 vertical nozzles each delivering about 320 gpm (20.1 I/sec). At the 2.5 mgd (0.11
m^/sec) plant flow, the spray system recycles the pond water 13 times, and at the 7.5 mgd
(0.33 m-Vsec) flow, 4 recycles occur. The spray nozzles are 4 in. x 2^ in. (10.2 cm x 6.35
cm) female pipe reducers each fitted with a 2l/2 in. x \Vz in. (6.35 cm x 3.8 cm) bushing.
Each spray orofice is about 1-7/8 in. (4.8 cm) in diameter. Nozzles having interval vanes or
other obstructions introduce air containing carbon dioxide to the water in the nozzle. This
causes deposition and rapid buildup of calcium carbonate within the nozzle. Such nozzles
are unsatisfactory because of the resulting plugging and flow restriction problems. Only
nozzles with unobstructed clear opening, and without internal vanes, should be used.
A major modification has been made to the existing stripping tower. The existing packing
has been removed, and the entire area of the tower equipped with water sprays. The existing
trays at the top of the tower distribute part of the flow, and 4 nozzle equipped headers in
the bottom of the tower spray water upward into the tower. The pump capacity to the
tower is 11.8 mgd (0.52 m^/sec). At cold weather plant flow rates of 2.5 mgd (0.11
m^/sec), this flow will provide a recycle rate of 5 through the sprays in the tower. The
capacity and type of nozzle used in the tower is similar to the nozzles used in the ponds.
Based on plant scale tests of this spray system, with the induced draft fan operating at high
speed, it is anticipated that at least 42 percent of the ammonia in the pond effluent will be
removed in the tower under cold weather operating conditions.
Chlorine may be added at two points in the process (Figure 8-9). The first point of
application is in the primary recarbonation chamber at a pH of 11.0. Only enough chlorine
is added to reduce the pH to about 9.6, thus eliminating the need for addition of carbon
dioxide (CO2) at this point. About 65 mg/1 is required to reduce the pH from 11.0 to 9.6.
The balance of the chlorine needed to reach the breakpoint for complete ammonia removal
is added in a chamber immediately downstream from the secondary recarbonation chamber.
At times sufficient chlorine is added at this point to reduce the pH to 7.0 or less, so that no
CO2 will be required. A dose of approximately 160 mg/1 of chlorine is required to reduce
the pH from 11.0 to 7.0. When the breakpoint is reached with a lesser dose of chlorine, then
some CO2 must be added in the secondary stage of recarbonation to produce a pH of 7.0.
About 10 mg/1 of chlorine is required for each mg/1 of NIfy —N to reach breakpoint. After
breakpoint chlorination treatment, water from the ballast pond is pumped to the existing
filters and carbon columns. The carbon columns remove any excess chlorine. The modified
Tahoe process was being placed in operation at the time of this writing and full-scale data
are not yet available.
9-86
-------
9.5.5.2 Orange County Water District, California
The Orange County Water District (OCWD) at Santa Ana, California has under construction
a 15 mgd (0.66 m-Vsec) wastewater reclamation plant and a 3 mgd (0.13 m^/sec) seawater
desalting plant at the same site.60 The water from the two plants will be blended and
pumped into a line of injection wells which are located on the land side of a line of seawater
pumping wells to form a barrier against seawater intrusion into the fresh water aquifer.63
The injection system also serves to replenish the supply of groundwater available for use.
The OCWD water reclamation plant will take 15 mgd (0.66 m^/sec) of trickling filter
effluent from the secondary treatment plant of the Orange County Sanitation District. It
will be subjected to high lime treatment at pH of 11.0 and clarification in a basin equipped
with settling tubes. The clarified high pH water will be pumped to two countercurrent
ammonia stripping towers. In the climate at this location, freezing temperatures are not
experienced, and waste heat from the desalting plant operation will be used to heat the inlet
air to the stripping towers for increased ammonia removal efficiency. The design also fully
realizes scaling problems encountered elsewhere and incorporates provisions for scale
control even though scaling was not a problem in pilot tests at this site.
The plant is designed with two ammonia stripping/cooling towers each equipped with six 18
ft (5.49 m) diameter fans. An end section view of a tower is shown in Figure 9-32 and an
overall view is shown in Figure 9-33. The stripping sections are designed for a hydraulic
loading of 1.0 gpm/ft2 (0.68 l/m2/sec) and an air flow of 400 cu ft/gal (2,990 m3/m3). The
cooling sections are designed to cool the desalting process water and brines from 46-49 C to
27-29 C and will raise the air temperature to the stripping section to 31-33 C. Splash-bar
packing will be used for ammonia stripping and a film packing located in the air inlet plenum
will be used for cooling.
A prime design criterion was that the ammonia stripping packing be accessible and
removable for cleaning because scaling of the packing might reduce air flow and ammonia
removal efficiency. Provisions have been made to feed a scale inhibiting polymer, if needed,
to the tower influent. The warm saturated air exhausted from the cooling sections
theoretically would permit the tower packing to be less than the design height of 25 ft (7.6
m). Because during the first few years of operations the desalting plant will operate on an
intermittent basis, it was not considered prudent to reduce the packing depth.
The tower fill or packing is made from l/i in. (1.27 cm) diameter Schedule 80 PVC pipe at 3
in. (7.6 cm) centers horizontally and with alternate layers placed at right angles and at 2 in.
(5.1 cm) centers vertically (see Figure 9-34). The fill was factory prefabricated in modules
which are about 6 ft (1.82 m) by 6 ft (1.82 m) by 4 ft (1.22 m) high. Each module is
supported within its own steel or fiberglass frame, so that it is easily removable if necessary
for cleaning. An overhead hoist and moveable dolly are provided to assist in packing removal
for cleaning. However, the access corridors and the removable air baffle panels in the tower
should make it possible to reach all of the fill in its normal operating position within the
tower for hosing down to remove any excessive calcium carbonate scale which may form.
9-87
-------
Based on the results of extensive pilot plant tests at the OCWD, it is expected that the
towers will remove more than 90 percent of the ammonia from the wastewater. The
ammonia remaining in the tower effluent will be removed by breakpoint chlorination using
about 10 mg/1 of chlorine for each mg/1 of ammonia nitrogen still present. No difficulty is
anticipated in meeting the limits set by the regulatory agency of 1.0 mg/1 of ammonia-
nitrogen in the injection water. The Orange County design offers an example of a design
with adequate scale control provisons incorporated.
FIGURE 9-32
ORANGE CO. AMMONIA STRIPPING/COOLING TOWER SECTION
AIR OUTLET
2 SPEED MOTOR
FAN
:\
FAN STACK
WATER INLET FLOW
CONTROL VALVE
i j
s
K>
.1
' ]
<: — r
* I !
1 •
>!
^
-WHI
r
t
j
i
o
,4if 3
J
J
'•
I—I .
r- V
AIR
el
.
OUT
!_„
_ll_
_JU
nn
Hi
H
_^AIR IN
fs"
i
1 V\
» HOIST^
INTERIOR
ACCESS
AREA^
>
AMMONIA
REMOVAL
FILL
DOLLY-j
r—L
S
1
r
a.--
T
i
1
1
1
1
1
t
1
1
1
I
j
i
i
^WASTE
DRAIN
j
XWATER
DRIFT
ELIMINATOR -j
o1 '
r ^
13" NOZZLE'
WATER
•-•- -!-.-
IN
-"
!— '-—
WATER 0
UT
R f-
i
I
.,
j
^-AMMONIA
REMOVAL
FILL BUNDLES
^-EXTERIOR
ACCESS
AREA
\\ AIR
i'i i''i "' V 1
6
POOLING FILL
AND AIR INLET
WASTEWATER-
COLLECTION
CHANNEL
COOLED PROCESS
WATER OR BRINE
COLLECTION CHANNEL
9-88
-------
The estimated costs (1974 costs) for the Orange County ammonia stripping towers are as
follows:
Capital $42/mil gal
Operating $29/mil gal
Total $71/mil gal
These costs are substantially higher than those reported for the Tahoe facility. The difference
results from the following factors: Orange County towers were designed at 1 gpm/ft^ (0.68
1/m^/sec) loading rate while Tahoe towers were designed at 2 gpm/ft^ (1.36 l/m^/sec);
extensive provisions for tower cleaning were included at Orange County; heat exchange
facilities were included in the Orange County design but no credit taken in the tower design;
Orange County costs were for 1974 vs the 1970 basis for the Tahoe costs.
FIGURE 9-33
OVERALL VIEW OF THE ORANGE COUNTY WATER DISTRICT (CALIF.)
(DESALTING PLANT IS IN RIGHT BACKGROUND, CHEMICAL CLARIFIER IN
RIGHT FOREGROUND. NOTE WALLS AT BASE OF TOWERS TO SHIELD
NOISE AT AIR INLET.)
9-89
-------
FIGURE 9-34
STRIPPING TOWER PACKING MODULE AT THE
ORANGE COUNTY WATER DISTRICT PLANT (CALIF.)
9-90
-------
9.6 References
1. Horstkotte, G.A., Niles, D.G., Parker, D.S., and D.H. Caldwell, Full-Scale Testing of a
Water Reclamation System. JWPCF, 46, No. 1, pp 181-197 (1974).
2. Mulbarger, M.C., The Three Sludge System for Nitrogen and Phosphorus Removal.
Presented at the 44th Annual Conference of the Water Pollution Control Federation,
San Francisco, California, October, 1971.
3. Description of the El Lago, Texas, Advanced Wastewater Treatment Plant. Harris
County Water Control and Improvement District Number 50, Seabrook, Texas, March,
1974.
4. Bishop, D.F., O'Farrell, T.P., Cassel, A.F., and A.P. Pinto, Physical-Chemical
Treatment of Raw Municipal Wastewater. Report prepared for the Environmental
Protection Agency, Contract Number 14-12-818, February, 1973.
5. Koon, J.H., and W.J. Kaufman, Optimization of Ammonia Removal by Ion Exchange
Using Clinoptilolite. Environmental Protection Agency Water Pollution Control
Research.Series No. 1708009/71, 1971.
6. Bishop, D.F., Schuk, W., Yarrington, R., Bowers, J.F., Fein, E.D., and H.W. Treupel,
Physical-Chemical Wastewater Treatment Under Direct Digital Control. Presented at
the International Workshop: Instrumentation, Control and Automation for Wastewater
Treatment Plants, London, September 15-20, 1973.
7. Barnes, R.A., Atkins, P.F., Jr., and D.A. Scherger, Ammonia Removal in a
Physical-Chemical Wastewater Treatment Process: Prepared for the EPA, Report No.
R2-72-123, November, 1972.
8. Brown and Caldwell, Report on Tertiary Treatment Pilot Plant Studies. Prepared for
the City of Sunnyvale, California, February, 1975.
9. Process Design Manual for Phosphorus Removal. U.S. EPA, Office of Technology
Transfer, Washington, D.C. (1971).
10. Process Design Manual for Carbon Adsorption. U.S. EPA, Office of Technology
Transfer, Washington, D.C. (1973).
11. Brown and Caldwell, Consulting Engineers, Lime Use in Wastewater Treatment. Report
submitted to the U.S. Environmental Protection Agency, 1975.
12. Culp, G.L., Wesner, G.M., and R.L. Gulp, 7974 Lake Tahoe Advanced Wastewater
Treatment Seminar Manual. Clean Water Consultants, El Dorado Hills, California,
1974.
9-91
-------
13. South Tahoe Public Utility District, Advanced Waste Treatment as Practiced at South
Tahoe. EPA-WPCRS 17010ELQ8/71, August, 1971.
14. Brown and Caldwell, Lime Sludge Recycling Study. Report to the Central Contra
Costa Sanitary District, June, 1974.
15. Battelle Memorial Institute, Pacific Northwest Laboratories, Evaluation of Municipal
Sewage Treatment Alternatives. Prepared for the Council on Environmental Quality
and the Environmental Protection Agency, Contract EQC 316, February, 1974.
16. Greene, R.A., Complete Nitrification by Single Stage Activated Sludge. Presented at
the 46th Annual Conference of the Water Pollution Control Federation, Cleveland,
Ohio, October, 1973.
17. McNamee, Porter and Seeley, Consulting Engineers, Plans for Sewage Treatment Plant
Additions, City of Jackson, Michigan. Submitted to the Michigan Water Resources
Commission, October, 1970.
18. Greene, R.A., Personal Communication to Parker, D.S., City of Jackson, Michigan,
September, 1974.
19. Brown and Caldwell, Project Report for the Wastewater Treatment Plant Stage 3
Improvements. Valley Community Services District, September, 1972.
20. Valley Community Services District, Wastewater Treatment Plant Operating Reports.
21. Johnson, James L., Personal Communication to G.A. Carthew. Valley Community
Services District, November, 1974.
22. Environmental Quality Analysts, Inc., Letter Report to Valley Community Services
District, March, 1974.
23. Brown and Caldwell, Contract for Treatment Plant Enlargement — 1965. Prepared for
the City of Livermore, December, 1965.
24. City of Livermore, California, Water Reclamation Plant Operating Reports, 1971.
25. John Carollo Engineers, Contract Documents for San Pablo Sanitary District
Wastewater Treatment Plants - 1970 Additions. 1970.
26. Kennedy, Bill, personal communication with D.S. Parker. San Pablo Sanitary District,
November, 1974.
27. San Pablo Sanitary District, California, Wastewater Treatment Plant Operating Reports.
June, 1973 to July, 1974.
9-92
-------
28. Weddle, C.L., Personal communication to D.S. Parker, Bechtel Corp. October, 1974.
29. Eisenhauer, D.L., Sieger, R.B. and D.S. Parker, Design of an Integrated Approach to
Nutrient Removal. Presented at the EED-ASCE Specialty Conference, Penn. State
University, Pa., July, 1974.
30. Leary, R.D., et al, Effect of Oxygen-Transfer Capabilities on Wastewater Treatment
Plant Performance. Journal of the Water Pollution Control Federation, 40, p 1978
(1968).
31. Leary, R.D., et al. Full Scale Oxygen Transfer Studies of Seven Diffuser Systems.
Journal of the Water Pollution Control Federation, 41, p 459 (1969).
32. Aberley, R.C., Rattray, G.B. and P.P. Dougas, Air Diffusion Unit. JWPCF, 46, No. 5,
pp 895-910 (1974).
33. Parker, D.S., Carthew, G.A. and G.A. Horstkotte, Lime Recovery and Reuse in Primary
Treatment. BED of the ASCE, in press.
34. Flanagan, M.J., Direct Digital Control of Central Contra Costa Sanitary District Water
Reclamation Plant. Presented at IAWPR Specialized Conference, London, England,
September, 1973.
35. Caldwell Connell Engineers, Design Report, Lower Molonglo Water Quality Control
Centre. Report to the National Capital Development Commission, April, 1971.
36. Caldwell Connell Engineers, Revisions to Design Report, Lower Molonglo Water
Quality Control Centre. Report to the National Capital Development Commission,
May, 1972.
37. Process Design Manual for Upgrading Existing Wastewater Treatment Plants. U.S. EPA,
Office of Technology Transfer, Washington, D-C. (1974).
38. Schwinn, D.E., Design Features of the District of Columbia's Water Pollution Control
Plant. Presented at the Sanitary Engineering Specialty Conference, ASCE, Sanitary
Engineering Division, Rochester, New York, June, 1972.
39. Schwinn, D.E., and G.K. Tozer, Largest Advanced Waste Treatment Plant in the U.S.
and in the World. Environmental Science and Technology, 8, No. 10, (1974).
40. Sawyer, C.N., Supplementary Comments on Nitrification and Denitrification Facilities.
Prepared for the EPA Technology Transfer Seminar, Denver, Colorado, November 13,
1974.
9-93
-------
41. EPA Postpones Denitrification. Reporter, Interstate Commerce Commission on the
Potomac River Basin, 32, No. 2, February, 1975.
42. Harris County Water Control and Improvement District No. 50, Monthly Progress
Report for October 1974, on Project 11010 GNM, prepared for the EPA.
43. Barth, E.F., EPA Technology Transfer Design Seminar on Nitrogen Control Tech-
nology. Presented at Denver, Colorado, November 13, 1974.
44. Brown and Caldwell/Dewante and Stowell, Feasibility Study for the Northeast-Central
Sewage Service Area. Prepared for the County of Sacramento, November, 1974.
45. Sacramento Area Consultants, Outfall Project and Environmental Impact Report.
Prepared for the Sacramento Regional County Sanitation District, December, 1974.
46. Sacramento Area Consultants, Chemical Report, Sacramento Regional Wastewater
Treatment Plant. January, 1975.
47. White, G.C., Personal Communication to E. Appel, June, 1974.
48. CH2M/Hill, Design of Montgomery County, Maryland Plant for the Washington
Suburban Sanitary Commission, 1915.
49. Gulp, G.L., Gulp, R.L. and C.L. Hamann, Water Resource Preservation by Planned
Recycling of Treated Wastewater. JAWWA, 65, No. 10, pp 641-647 (1973).
50. CH2M/Hill, Wastewater Treatment Study, Montgomery County, Maryland. November,
1972.
51. Cloromat. Ionics, Inc. Brochure
52. Michalek, S.A., and F.B. Leitz, On-Site Generation of Hypochlorite. JWPCF, 44, No. 9,
pp 1697-1712(1972).
53. Suhr, L.G., and L. Kepple, Design of a Selective Ion Exchange System for Ammonia
Removal. Presented at the ASCE Environmental Engineering Division Conference,
Pennsylvania State University, July 1974.
54. CH2M/Hill, Design of the Upper Occoquan Sewage A uthority. (1974).
55. Physical-Chemical Plant Treats Sewage Near the Twin Cities. Water and Sewage Works,
p. 86, September, 1973.
9-94
-------
56. Larkman, D., Physical-Chemical Treatment: Chemical Engineering, Deskbook Issue, p.
87, June 18, 1973.
57. Gulp, R.L., Nitrogen Removal by Air Stripping. Proceedings, Wastewater Reclamation
and Reuse Workshop, SERL, Univ. of California, LakeTahoe, Ca., June 25-27, p. 128,
1970.
58. Gulp, R.L. and G.L. Gulp, Advanced Wastewater Treatment. Van Nostrand Reinhold
Co., New York, p. 51, 1971.
59. Gulp, R.L., and H.E. Moyer, Wastewater Reclamation and Export at South Tahoe. Civil
Engineering, p. 38, June, 1969.
60. Wesner, G.M., and R.L. Gulp, Wastewater Reclamation and Seawater Desalination.
JWPCF, 44, No. 10, pp 1932-1939 (1972).
61. Folkman, Y., and A.M. Wachs, Nitrogen Removal Through Ammonia Release From
Holding Ponds. Proceedings, 6th Annual International Water Pollution Research
Conference, Tel Aviv, Israel, June 18-23, 1972.
62. Gonzales, John G., and R.L. Culp, New Developments in Ammonia Stripping, Public
Works, 104, No. 5, p. 78 (1973) and No. 6, p. 82 (1973).
63. Wesner, G.M., and D.C. Baier, Injection of Reclaimed Wastewater Into Confined
Aquifers. JAWWA, 62, No. 3, pp 203-210 (1970).
9-95
-------
APPENDIX A
GLOSSARY OF SYMBOLS
Symbol
A
ADWF
(a)
8
BHP
C
Cl
Cb
m
D
D
D
Definition
air flow per unit of tank volume, standard
cu ft per min per 1000 cu ft; dimensionless
number in contact stabilization calculations
average dry weather flow
activity of an ion
fraction of sludge in stabilization tank; oxygen
required for carbonaceous oxidation, mg/1
brake horsepower
fraction of total sludge in contact tank
process dissolved oxygen level, mg/1
denitrifier biomass production, mg/1
required methanol concentration, mg/1
oxygen saturation in water at temperature T,
mg/1
concentration of nitrate nitrogen, mg/1; axial
dispersion coefficient, sq ft/hr
influent NO^-N, mg/1
effluent NO~ -N, mg/1
mass average influent NO~-N level over 24 hr,
mg/1
A-l
-------
Symbol Definition
Dj mass average effluent NOl-N level over 24 hr,
mg/1
DO dissolved oxygen, mg/1
•
e aerator rated oxygen transfer efficiency at
standard conditions, percent
F/M food to microorganism ratio
f nitrifier fraction of the mixed liquor solids;
fanning friction factor
HT hydraulic detention time
I inventory of VSS under aeration, Ib
A
KJJ selectivity coefficient for ion exchange equili-
bria.
half saturation constant for nitrate, mg/1
NO~-N
K, "decay" coefficient, day
K^ half saturation constant for methanol, mg/1
of methanol
Kj^ half saturation constant for oxidation of
ammonia nitrogen, mg/1
Kn half saturation constant for oxygen, mg/1
2
K half saturation constant = substrate concentra-
tion, mg/1 at half the maximum growth rate
L tank length, ft
M methanol concentration, mg/1
M mass of heterotrophs grown through oxidation
of organic carbon
A-2
-------
Symbol
Definition
MLSS
MLVSS
MPN
N
N
N
1
N
N,
NOD
NT
PDWF
PWWF
Q
Q
mass of nitrifiers grown through oxidation of
ammonia
mixed liquor suspended solids, mg/1
mixed liquor volatile suspended solids, mg/1
most probable number
NH. -N concentration, mg/1
TKN in the influent, mg/1
NH£ -N in the effluent, mg/1
24 hr-average influent TKN, mg/1
24 hr-average effluent NH.-N, mg/1
NOl -N level in the contact tank, mg/1
NO--N level in the stabilization tank, mg/1
nitrogenous oxygen demand, mg/1
ammonia nitrogen oxidized, Ib/day
peak dry weather flow
peak wet weather flow
influent flow rate, mgd
mean flow rate (ADWF), mgd
air flow, cfm
rate of substrate removal, Ib BOD (or COD)
removed/lb VSS/day
A-3
-------
Symbol
°lr>
D
A
qN
JN
A
rN
R
S
SCFM
S
o
Sl
SF
SS
SVI
T
TKN
At
Definition
nitrate removal rate, Ib NOu-N rem./lb VSS/
day
peak nitrate removal rate, Ib NO~-N rem./lb
VSS/day
ammonia oxidation rate, Ib NH.-N oxidized/
Ib VSS/day
peak ammonia oxidation rate, Ib NtL-N
oxidized/lb VSS/day
nitrification rate, Ib NH.-N oxidized/lb
MLVSS/day
peak nitrification rate, Ib NH .-N oxidized/
Ib MLVSS/day
recycle flow rate, mgd
growth limiting substrate concentration, mg/1;
total sludge wasted in Ib/day
standard cubic ft per minute
influent total BOD (or COD), mg/1
effluent soluble BOD (or COD), mg/1
safety factor
suspended solids, mg/1
sludge volume index, ml/g
temperature, degrees C
Total Kjeldahl Nitrogen
time increment
A-4
-------
Symbol Definition
u mean displacement velocity, ft per hr
V volume of aeration tank or reactor, mil gal
V volume of contact tank, mil gal
c
V volume of stabilization tank, mil gal
o
VSS volatile suspended solids
W waste sludge flow rate, mgd; tank width, ft;
-
wase suge ow ra, mg; w,
total oxygen demand, mg/1; oxygen trans-
ferred under process conditions, Ib/day
oxygen transferred under standard conditions,
Ib/day
a coefficient in oxygen transfer calculations
MLVSS, mg/1
effluent VSS, mg/1
X contact tank MLVSS, mg/1
C
X stabilization tank MLVSS, mg/1
X waste sludge VSS, mg/1
Y. heterotrophic yield coefficient, Ib VSS grown
per Ib of substrate removed
Y. net yield of VSS of heterotrophs per unit of
carbon (BOD5 or COD) removed
YD denitrifier gross yield, Ib VSS grown/lb NO~-N
rem.
YD denitrifier net yield, Ib VSS grown/lb NO~-N
rem.
A-5
-------
Symbol Definition
organism yield coefficient, Ib Nitrosomonas
grown (VSS) per Ib NH^-N removed
CX „ selectivity coefficient for ion exchange equili-
bria
9 solids retention time, days
c
9 solids retention time of design, days
C
.0 minimum solids retention time, days, for
nitrification at given pH, T and DO
M growth rate of microorganism, day
A
IJL maximum growth rate of microorganism,
day"1
M, net growth rate of heterotrophic population
^c growth rate of nitrifiers in contact tank, day"
MT-V denitrifier growth rate, day
* 1
D maximum denitrifier growth rate, day
Mpv design denitrifier growth rate, day
Mxr Nitrosomonas growth rate, day"
A _1
MXT peak Nitrosomonas growth rate, day
•
MN maximum possible nitrifier growth rate under
environmental conditions of T, pH and DO,
and N >> KN
n
s growth rate of nitrifiers in stabilization tank,
day"1
A-6
-------
APPENDIX B
METRIC EQUIVALENTS
METRIC CONVERSION TABLES
Recommended Units
Description
Length
Area
Volume
Mass
Time
Force
Unit
meter
kilometer
millimeter
centimeter
micrometer
square meter
square kilometer
square centimeter
square millimeter
hectare
cubic meter
cubic centimeter
liter
kilogram
gram
milligram
tonne
second
day
year
newton
Symbol
m
km
mm
cm
m'
km2
cm*
mm*
ha
m3
cm3
1
kg
g
mg
t
s
day
yr or
a
N
Comments
Basic S/ unil
The hectare (10,000
m2) is a recognized
multiple unit and
will remain in inter-
national use.
The liter is now
recognized as the
special name for
the cubic decimeter
Basic SI unit
1 tonne = 1,000 kg
Basic S/ unit
Neither the day nor
the year is an SI unit
but both are impor-
tant.
The newton is that
force that produces
an acceleration of
1 m/s2 in a mass
of 1 kg.
English
Equivalents
39.37 in. = 3.28 ft =
1.09yd
0.62 mi
0.03937 in.
0.3937 in.
3.937 X 10'3=103A
10.744 sq ft
= 1.196 sq yd
6.384 sq mi =
247 acres
0.155sqin.
0.00155 sq in.
2.471 acres
35.314 cu ft = .
1 .3079 cu yd
0.061 cu in.
1. 057 qt = 0.264 gal
= 0.81 X 10-4 acre-
ft
2.205 Ib
0.035 oi-tS.43 gr '
0.01 543 gr
0.984 ton (long) •
1.1 023 ton (short)
0.22481 Ib (weight)
» 7.5 poundals
Description
Velocity
linear
angular
Flow (volumetric)
Viscosity
Pressure
Temperature
Work, energy,
quantity of heat
Power
Application of Units
Description
Precipitation^
run-off.
evaporation
River flow
Flow in pipes.
conduits, chan-
nels, over weirs.
pumping
Discharges or
abstractions.
yields
Usage of water
Density
Unit
millimeter
cubic meter
per second
cubic meter per
second
liter per second
cubic meter
per day
cubic meter
per year
liter per person
per day
kilogram per
cubic meter
Symbol
mm
m3/s
m3/s
l/s
m3/day
m3/yr
I/person
day '
kg/m3
Comments
For meteorological
purposes it may be
convenient to meas-
ure precipitation in
terms of mass/unit
area i kg/m 3).
1 mm of rain =
t kg/sq m
Commonly called
the cumec
1 l/s = 86.4 m3/day
The density of
water under stand-
ard conditions is
1,000kg/m30r
1.000g/l
English
Equivalents
35.314 cfs
15.85 gpm
1.83X 10'3gpm
0.264 gcpd
0.0624 Ib/cu ft
Description
Concentration
BOD loading
Hydraulic load
per unit area;
e.g. filtration
rates
Hydraulic load
per unit volume;
e.g. biological
filters, lagoons
Air supply
Pipes
diameter
length
Optical units
Recommended Units
Unit
meter per
second
millimeter
per second
kilometers
per second
radians per
second
cubic meter
per second
liter per second
poise
newton per
square meter
kilonewton per
square meter
kilogram (force)
per square
centimeter
degree Kelvin
degree Celsius
joule
kilojoule
watt
kilowatt
joule per second
Symbol
m/s
mm/s
km/s
rad/s
m3/s
l/s
poise
N/m2
kN/m2
kgf/cm2
K
C
J
kJ
W
kW
J/s
Comments
Commonly called
the cumec
The newton is not
yet well-known as
the unit of force
and kgf cm2 will
clearly be used for
some lime. In this
field the hydraulic
head expressed in
meters is an accept-
able alternative.
Basic SI unit
The Kelvin and
Celsius degrees
are identical.
The usa of the
Celsius scale is
recommended as
it is the former
centigrade scale.
1 joule «1 N-m
1 watt = 1 J/s
English
Equivalents
3.28 fps
0.00328 fps
2.230 mph
15, 8 50 gpm
* 2.120 cfm
15.85 gpm
0.0672/lb/
sec ft
0.00014 psi
0.145 psi
14.223 psi
5F
-- - 17.77
2.778 X ID'7
kwht "
3.725X10-'
hp-hr = 0.73756
10-4 Btu
2.778 kw-hr
Application of Units
Unit
milligram per
liter
kilogram per
cubic meter
per day
cubic meter
per square meter
per day
cubic meter
per cubic metir
per day
cubic meter or
liter of free air
per second
millimeter
meter
tumen per
square meter
Symbol
mg/i
kg/m3 day
m3/m2 day
m3/m3 day
m3/s
l/s
mm
m
lumen/m2
Comments
If this is con-
verted to a
vetocity.it
should be ex-
pressed in mm/s
(1 mm/i o 86.4
m3/m2 day).
English
Equivalents
1 ppm
0.0624 Ib/cu-ft
day
3.28 cu ft/iq ft
SBJS.1!-
3.28ft
0.092 ft
andle/sq ft
B-l
-------
|