vvEPA
United States
Environmental Protection
Agency
Office of Resaarch
and Development
Washington DC 20460
Center for Environmental
Research Information
Cincinnati OH 45268
Technology Transfer
Design
Manual
Dewatering Municipal
Wastewater Sludges
-------
-------
EPA/625/1-87/014
September 1987
Design Manual
Dewatering Municipal Wastewater Sludges
U.S. Environmental Protection Agency
Office of Research and Development
Center for Environmental Research Information
Cincinnati, OH 45268
-------
Notice
This document has been reviewed in accordance with the U.S. Environmental Protection
Agency's peer and administrative review policies and approved for publication. Mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
This document is not intended to be a guidance or support document for a specific regulatory
program. Guidance documents are available from EPA and must be consulted to address
specific regulatory issues.
-------
Contents
Chapter Page
1 Introduction 1
1,1 Purpose and Scope 1
1.2 Objectives of Dewatering 2
1.3 Location of the Dewatering Process 2
1.4 Using this Manual 2
2 Preliminary Considerations 5
2.1 Introduction , 5
2.2 Regulatory Concerns 7
2.3 General Performance Capabilities of Mechanical Dewatering Processes 7
2,4 Key Operations Variables Affecting Mechanical Dewatering Performance 11
3 Sludge Characteristics and Preparatory Treatment 13
3.1 Sludge Production and Concentration 13
3.2 Sludge Concentration -- Primary Clarifiers 17
3.3 Characteristics of Waste Sludges 20
3.4 Recirculation from Solids Processing 24
3.5 References 24
4 Process Selection , 25
4.1 Introduction 25
4.2 Sludge Processing Methods/Selection Procedures 26
4.3 Operational Selection Criteria 28
4.4 Sizing of the Dewatering Process 34
5 Conditioning 37
5.1 Introduction 37
5.2 Factors Affecting Conditioning 38
5.3 Inorganic Chemical Conditioning 38
5.4 Organic Polymers 41
5.5 Design of a New Installation 47
5.6 Thermal Conditioning 51
5.7 References 54
in
-------
Contents (continued)
Chapter Page
6 Air Drying Processes 57
6.1 Introduction 57
6.2 Sand Beds . . 57
6.3 Freeze Assisted Sand Bed Dewatering 61
6.4 Vacuum Assisted Drying Beds (VADB) 65
6.5 Wedgewire Beds 70
6.6 Sludge Lagoons , 71
6.7 Paved Beds , 72
6.8 Other Innovative Processes ", 76
6.9 References ; 77
7 Mechanical Dewatering Systems 79
7.1 Introduction 79
7.2 Belt Filter Presses 79
7.3 Centrifuges 88
7.4 Filter Presses 104
7.5 Vacuum Filtration 117
7.6 New Methods of Dewatering 122
7.7 References . 131
8 Case Studies: Air Drying Systems , 135
8.1 Introduction 135
8.2 Upgraded Sand Drying Beds 135
8.3 Paved Drying Beds 139
8.4 Weather Data 142
8.5 References 142
9 Case Studies: Mechanical Dewatering Systems 143
9.1 Introduction 143
9.2 Case Study: Belt Filter Presses, Stamford, CT 143
9.3 Case Study: Centrifuges, San Francisco, CA 145
9.4 Case Study: Centrifuges, Calumet, IL, Calumet STW 150
9.5 Case Study: Centrifuges, Calumet, IL, West-Southwest STW 152
9.6 Case Study: Centrifuges, Denver, CO . . 154
9.7 Case Study: Centrifuges, Ontario, Canada 156
9.8 References 162
Appendix A -- Design Examples/Cost Analyses 163
A.1 Introduction 163
A.2 Determine Sludge Quantities 163
A.3 Upgraded Sand Drying Beds 164
A.4 Vacuum Assisted Drying Beds . 166
A.5 Belt Filter Presses 167
A.6 Solid Bowl Centrifuges 169
A.7 Filter Presses 171
A.8 References 172
IV
-------
Contents (continued)
Chapter , Page
Appendix B - Operation and Maintenance, Mechanical Dewatering Systems 173
B.1 Introduction , 173
B.2 Belt Filter Presses ' 173
B.3 Solid Bowl Centrifuges , 174
B.4 Filter Presses 178
B.5 Vacuum Filters 185
B.6 References 189
Appendix C - Manufacturers and Sources of Equipment 191
C.1 Belt Filter Press Manufacturers 191
C.2 Centrifuge Suppliers 191
C.3 Filter Press Suppliers 192
-------
Figures
Number Page
!
2-1 Sludge Processing Options 6
2-2 Dewatered Sludge Cake, Percent Solids for Raw Primary
and Waste Activated Sludge 8
2-3 Dewatered Sludge Cake, Percent Solids For Mixtures of Digested
Primary (P) and Digested Waste Activated Sludge (WAS) 9
2-4 Dewatered Sludge Cake, Percent Solids For Mixtures of Raw
and Digested Primary and Secondary Sludge and Heat Treated
Primary and Secondary Sludge 10
3-1 Influent BODS Average Monthly Loading, 1985-1986 14
3-2 Process Mass Balance ; : 15
3-3 Cake Solids as a Function of Primary Clarifier Efficiency and PS:WAS Ratio .... 16
3-4 Sludge Yield From Domestic Wastewater With and Without
Primary Clarification 18
3-5 Effect of Floor Configuration on Blanket Location and Sludge Inventory 19
3-6 Combination Primary Clarifier-Thickener 21
3-7 Effect of Feed Solids on Performance of a Rotary Vacuum Filter 22
4-1 General Schematic for Solids Handling, Showing Most Commonly Used
Methods of Treatment and Disposal 26
4-2 Five Stages of Analysis in Selection of a Dewatering Process 27
4-3 Multiple Hearth Furnace and Fluid Bed Reactor Capacities 33
4-4 Conditions for Zero Fuel 33
4-5 Heat Loss Due to Excess Air 33
4-6 Fuel Consumption vs. Stack Temperature and Excess Air 34
5-1 Particle Size Distribution of Common Materials 39
5-2 Polyacrylamide Molecule-Backbone: of the Synthetic Organic Polymer 42
5-3 Typical Configuration of a Cationic Polymer in Solution 42
5-4 Schematic Representation of the Bridging Model for the
Destabilization of Colloids by Polymers 43
5-5 Example of a Dry Polymer Make-Up System 45
5-6 Example of a Liquid Polymer Make-Up System 45
5-7 Example of an Emulsion Polymer Make-Up System 46
5-8 Example of a Filter Leaf Apparatus 48
5-9 Buchner Funnel Apparatus 49
5-10 Time/Filtrate vs. Filtrate Volume 50
5-11 Plot of Specific Resistance vs. Conditioning Chemical Dosage 50
5-12 Capillary Suction Time Apparatus 50
5-13 Heat Treatment Process Flow Diagram 52
5-14 Low Pressure Oxidation Process Flow Diagram 53
VI
-------
Figures (continued)
Number Page
6-1 Sand Bed Details ' 57
6-2 Solids Loading Rates for Sand Beds vs. Solids Content of Applied Sludge 59
6-3 Potential Depth of Sludge that Could Be Frozen if Applied in 8-cm Layers .... 62
6-4 Predicted vs. Measured Sludge Freezing at Duluth, MN 64
6-5 Effect of Freeze Thawing on the Drainage Rate for Anaerobicaliy
Digested Sludge 64
6-6 Plan View of a Vacuum Assisted Drying Bed System 65
6-7 Typical Outdoor Vacuum Assisted Drying Bed 65
6-8 Cross Section of a Typical Vacuum Bed Media Plate 66
6-9 Cracked Bed, Ready for Sludge Removal 68
6-10 Cross Section of a Wedgewire Drying Bed 70
6-11 Cable and Scraper System for Sludge Drying Lagoons 73
6-12 Labor Requirements for Sludge Drying Lagoons 73
6-13 Paved Sludge Drying Bed Designed for Decantation and Evaporation 74
7-1 Simplified Schematic of a Belt Filter Press 81
7-2 Typical Independent High Pressure Section 81
7-3 Typical High Pressure Zone 82
7-4 Typical Clipper Seam for Split Dewatering Belt 84
7-5 Control Panel for Belt Filter Press Dewatering System 85
7-6 Washwater Spray Bar with Cleaning Brushes 86
7-7 Belt Filter Press Installation Using Permanganate for Odor Control 87
7-8 Solid Bowl (Countercurrent) Conveyor Discharge Centrifuge 90
7-9 Solid Bowl Concurrent Centrifuge with Hydraulic Scroll Drive 90
7-10 Comparison Between the Clarifier and the Centrifuge 91
7-11 Sigma Scale-Up Procedure 93
7-12 Four Different Types of Hardfacing Used to Retard Scroll Wear 95
7-13 Effect of Bowl Angle on the Movement of Sludge 96
7-14 Bowl/Scroll Differential Speed Monitoring Box 98
7-15 Effect of SVI on Centrifugal Dewatering of Activated Sludge 100
7-16 Fixed-Volume Recessed Plate Filter Press 106
7-17 Filling and Cake Discharge, Fixed Volume Recessed Plate Filter Press 108
7-18 Filling and Cake Discharge, Diaphragm Press 109
7-19 Specific Resistance vs. Capillary Suction Time ' 112
7-20 Operating Zones of a Rotary Vacuum Filter . 117
7-21 Cross-Sectional View of a Coil Spring, Belt Type Rotary Vacuum Filter ..... 118
7-22 Cross-Sectional View of a Cloth, Belt Type Rotary Vacuum Filter 119
7-23 Rotary Vacuum Filter System 119
7-24 Rotary Vacuum Filter Productivity as a Function of Feed
Sludge Suspended Solids Concentration '. \ . , 121
7-25 Sludge Cake Total Solids Concentration as a Function of the Feed
Sludge Suspended Solids Concentration 121
7-26 Expressor Press . 124
7-27 Dewatered Sludge at the Columbia River WWTP, Portland, OR 125
7-28 Functional Schematic of the Som-A-Press . 126
7-29 Som-A-System 126
7-30 CentriPress and Cake Samples 128
7-31 Effect of Polymer Dose on Solids Recovery 131
7-32 Cake Solids vs. Differential Speed 131
7-33 Schematic of the Screw Press Dewatering System 132
7-34 Sludge Press Function of the Hi-Compact Method 133
VII
-------
Figures (continued)
Number Page
8-1 Schematic of the North Plant, Elgin, IL 136
8-2 Auger Aerator Dewatering Machine 141
9-1 Schematic of Sewage Treatment Plant, Stamford, CT 144
9-2 Sludge Flow Path through Co-Incineration System 144
9-3 Schematic of Southeast WPCP, San Francisco, CA : 146
9-4 Cake Solids vs. Torque 148
9-5 Cake Solids vs. Feed Rate 148
9-6 Effect of Polymer Dose on Cake Solids 149
9-7 Effect of Polymer Dose on Solids Recovery 149
9-8 Automatic Backdrive System 157
9-9 Typical Control Curves for Automatic Backdrive 158
9-10 Centrifuge Dewatering 158
9-11 Polymer Requirements 159
9-12 Centrifuge Power Curve Dewatering Mode •. . 159
9-13 Schematic of Duff in Creek Pollution Control Plant, Ontario, Canada 160
B-1 Sample Log Sheet for Belt Filter Press 175
B-2 Sample Daily Log Sheet for Centrifuge 179
B-3 Sample Monthly Log Sheet for Centrifuge '. 180
B-4 Sample Yearly Performance Trend Log 181
B-5 Material Balance -- Filter Press Test 184
B-6 Sample Test Data Sheet for Filter Press 186
B-7 Material Balance -- Vacuum Filter Test 187
B-8 Sample Test Data Sheet for Vacuum Filter 188
VIII
-------
Tables
Number Page
1-1 EPA Technology Transfer Sludge Management Publications 1
2-1 Estimated Municipal Sludge Production by POTW Size 5
2-2 Distribution of Sludge Disposal/Utilization Practices for 1,011 Surveyed POTWs . . 5
2-3 Operational Variables for Mechanical Dewatering Processes 12
3-1 Typical Production of Primary and Primary-Chemical Sludges 17
3-2 Primary Clarifier Comparison - Sludge Blanket Location and
Sludge Concentration vs. Floor Configuration , 20
3-3 Specific Gravity of Waste Sludges , 21
3-4 Thickening of Waste Sludges 22
3-5 Sludge Dewatering as a Function of Particle Size 23
3-6 Specific Resistance of Various Types of Sludges 23
4-1 Comparative Dewatering Results for Two Test Periods 28
4-2 Operational Selection Criteria for Sludge Dewatering Processes 29
4-3 Compatibility of Dewatering Process with Plant Size 31
4-4 Suitability of Cake Produced by Dewatering Processes for Various
Ultimate Disposal Options 32
4-5 Dewatering-Combustion Costs, Excluding Operating Labor 35
5-1 Effects of Conditioning with Inorganic Chemicals, Organic Polymers, or Heat
on a Mixture of Primary and Waste Activated Sludge 37
5-2 Crystallization Temperatures for Ferric Chloride Solutions 39
5-3 Typical Conditioning Dosages of Ferric Chloride and Lime for Municipal
Wastewater Sludges 41
5-4 Representative Dry Powder Cationic Polymers (Polyacrylamide Copolymers) ... 44
5-5 Representative Cationic Polymers 45
5-6 Typical Dosages of Dry Polymer for Belt Filter Presses 46
5-7 Typical Dosages of Dry Polymer for Conditioning Various Types of Sludges
for Dewatering in Solid Bow! Centrifuges 47
5-8 Typical Dosages of Dry Polymer for Conditioning Various Types of Sludges
on Vacuum Filters 47
6-1 Loading Criteria for Anaerobically Digested, Non-Conditioned Sfudge
on Uncovered Sand Beds 58
6-2 Sludge Filtrate Characteristics After Freeze Thawing 60
6-3 Performance Observed at 13 Vacuum Assisted Drying Beds 67
6-4 Capital Costs for Vacuum Assisted Drying Beds (1984 $) 69
6-5 Cost Comparison of Vacuum Assisted Drying Beds vs. Sand Drying Beds 69
6-6 O&M Cost Comparison of Vacuum Assisted Drying Beds vs. Sand Drying Beds . 69
6-7 Estimated Cost Comparison of Paved Beds vs. Conventional Sand Drying Beds . 76
IX
-------
Tables (continued)
Number
7-1 Comparative Mechanical Dewatering Performance .............. ........ 80
7-2 Summary of Pressures Calculated for Each Roller ...................... 84
7-3 Typical Data for Various Types of Sludges Dewatered on Belt Filter Presses ... 87
7-4 Conventional Gravity Thickening and High Rate Pre-concentration . . ........ 91
7-5 Sludge Dewatering Performance at San Francisco (1982-1983) . . . . ........ 96
7-6 Centrifuge Tests at Littleton/Englewood STP, Colorado ........... ........ 98
7-7 Performance of Fixed vs. Variable Torque Controlled Backdrive •
at Seattle Metro ............................. . ........ • ........ 99
7-8 Centrifuge Performance Characteristics ...................... . ...... 102
7-9 Evaluation Criteria for Centrifugal Dewatering Equipment .......... ....... 1 02
7-1 0 Suggested Capacity and Number of Centrifuges ....................... 1 03
7-11 Performance Data for Solid Bowl Centrifuges .................. '' ....... 105
7-12 Comparison of Iron Conditioners with and without Lime ........... •. ....... 112
7-13 Summary Data -- General Information ............... . ............ 114
7-14 Summary of Operating Problems . . ............................... 116
7-15 Filter Cake Solids -- Average by Conditioning Method .......... ....... 116
7-16 Filter Cake Solids -- Average by Conditioning Method and Sludge Type .... 117
7-17 Typical Dewatering Performance Data for Rotary Vacuum Filters ~ Cloth Media 122
7-18 Typical Dewatering Performance Data for Rotary Vacuum Filters -- Coil Media 122
7-19 Specific Operating Results of Rotary Vacuum Filters ~ Cloth Media ....... 123
7-20 Specific Operating Results of Rotary Vacuum Filters - Coil Media ........ 123
7-21 Advantages and Disadvantages of Vacuum Filtration ............. . ...... 124
7-22 Common Design Shortcomings of Vacuum Filter Installations ............. 124
7-23 Som-A-System Operating Data . : ........................ . ...... 127
7-24 Som-A-System Chemical Conditioning Data ........................ 127
7-25 Results of Chicago WSW CentriPress Study ...... . ........... . ...... 130
7-26 Test Results for HIW Screw Press ................................ 132
8-1 Elgin, IL North Wastewater Treatment Plant Operation and Maintenance
Costs for One Drying Bed Cycle .................................. 138
8-2 Sludge Bed Operation - Belleville, IL, 1985 ............ , ..... . ...... 138
8-3 Costs for Auger Aerator Tractor Used To Improve Sludge Drying Bed
Operations at Village Creek Wastewater Treatment Plant . ..... . ......... 141
8-4 Rainfall and Air Temperature Data for Air Drying Case Studies ..... •. ...... 142
9-1 1984 Centrifuge Performance Test, Sharpies Centrifuge No. 1 ..... . ...... 147
9-2 1 984 Centrifuge Performance Test, Humboldt Centrifuges Nos. 2 and 3 ...... 147
9-3 Operator Comments on Humboldt and Sharpies Centrifuges .............. 150
9-4 Calumet STW: Sample of Operating Data (September 1986} ....... , . ..... 152
9-5 Comparison of Dewatering Systems ... ..................... ; ...... 155
9-6 Centrifuge Bids ...................................... . . ...... 155
9-7 Centrifuge Evaluation for Dewatering and Thickening ....... ..,..:...... 1 56
9-8 Comparison of Bid Specifications and Actual Performance for the ;
Dewatering Mode ............ . ........................ . ...... 158
9-9 Dewatering Equipment Comparison -- $/Day ................ . ...... 1 59
9-10 Ultimate Disposal Costs -- $/Day ............................... 159
9-11 Duffin Creek VVPCP Dewatering System (September 1 986 Operating Data) ... 161
9-12 Duffin Creek WPCP Dewatering System (December 1986 Operating Data) .... 161
9-13 Duffin Creek WPCP Sludge Loading for 1986 ................. . . ...... 162
9-14 Duffin Creek WPCP Sludge Dewatering for 1986 ............... . ...... 162
-------
Tables (continued)
Number Page
B-1 Causes and Prevention of Belt Wear 173
B-2 Number of Operators Required per Belt Filter Press 173
B-3 Causes and Prevention of Operational Problems of Belt Filter Presses 174
B-4 Effect of Process Variables on Recovery and Cake Solids 176
B-5 Effect of Machine Variables on Recovery and Cake Solids 176
C-1 Centrifuge Manufacturer A - Countercurrent Decanter Centrifuges ....... 192
C-2 Centrifuge Manufacturer B - High-G Centrifuges, Concurrent Design 192
C-3 Manufacturer C — Low-G Centrifuges, Concurrent Design 192
C-4 Manufacturer C — High-G Centrifuges, Concurrent Design 192
-------
Acknowledgmen ts
Many Individuals contributed to the preparation and review of this manual. Contract
administration was provided by the U.S. Environmental Protection Agency (ERA), Center for
Environmental Research Information (CERI), Cincinnati, Ohio.
Authors: :
Orris E. Albertson, Enviro Enterprises, Salt Lake City, Utah i
Bruce E. Burris, CWC-HDR, Santa Ana, California
Sherwood C. Reed, U.S. Army Cold Regions Research & Engineering Laboratory, Hanover,
New Hampshire
Jeannette A. Semon, City of Stamford Water Pollution Control Facility, Stamford, Connecticut
James E. Smith, Jr., EPA-CERI, Cincinnati, Ohio
A.T. Wallace, University of Idaho, Moscow, Idaho ;
Technical Direction/Coordination: '
James E. Smith Jr., EPA-CERI, Cincinnati, Ohio '
Contributors and Reviewers: !
Robert K. Bastian, EPA-Office of Municipal Pollution Control, OMPC), Washington, D.C.
Harry E. Bostian, EPA-Water Engineering Research Laboratory (WERL), Cincinnati, Ohio
John A. Drozda, EURAMCA Ecosystems, Inc., Addison, Illinois
Joseph B. Farrell, EPA-WERL, Cincinnati, Ohio
Walter G. Gilbert, EPA-Office of the Inspector General, Washington, D.C.
Ancil Jones, EPA-Region VI, Dallas, Texas
C. Douglas Robinson, U.S. Environmental Products, St. Charles, Illinois
Ludovico Spinosa, CNR-IRSA, Bari, Italy
Reviewers:
J.S. Adamik, Infilco Degremont, Richmond, Virginia
Donald S. Brown, EPA-WERL, Cincinnati, Ohio
S. Brown, Roscoe Brown Corporation, Lenox, Iowa
Carlos Doyle, American Cyanamid Company, Indianapolis, Indiana
Fred Eubanks, USAGE, Washington, D.C.
James A. Louden, Komline-Sanderson, Peapack, New Jersey
Cecil Lue-Hing, Metropolitan Sanitary District of Greater Chicago, Illinois
Wen H. Huang, EPA-OMPC, Washington, D.C.
Orville E. Macomber, EPA-CERI, Cincinnati, Ohio
Stephen W. Maloney, U.S. Army Construction Engineering Research Laboratory, Champaign,
Illinois
Robert W. Okey, CRM Associates, Salt Lake City, Utah
Albert A. Pincince, Camp Dresser & McKee, Boston, Massachusetts
John Resta, U.S. Army Environmental Hygiene Agency, Aberdeen Proving Ground, Maryland
E.D. Smith, U.S. Army Construction Engineering Research Laboratory, Champaign, Illinois
XII
-------
Acknowledgments (continued)
Other Support:
Lisa H. Albertson, Enviro Enterprises, Salt Lake City, Utah
Robert C. Goldberg, JACA Corporation, Fort Washington, Pennsylvania
Virginia R. Hathaway, JACA Corporation, Fort Washington, Pennsylvania
Denis J. Lussier, EPA-CERI, Cincinnati, Ohio
Partial funding support for the preparation of this document was provided by the U.S. Army
Corps of Engineers through the U.S. Army Cold Regions Research and Engineering Laboratory,
Hanover, New Hampshire and the Construction Engineering Research Laboratory, Champaign,
Illinois.
Xlil
-------
-------
Chapter 1
Introduction
1.1 Purpose and Scope
This manual presents up-to-date information on
dewatering processes for municipal wastewater
sludges. The design engineer can use this document
to aid in the selection of an appropriate dewatering
process for a particular application. The manual
revises and updates the information on sludge
conditioning and dewatering previously found in the
U.S. Environmental Protection Agency's Process
Design Manual for Sludge Treatment and Disposal
(October 1979) and Process Design Manual for
Dewatering Municipal Wastewater Sludges (October
1982). Significant advances have been made in
dewatering technology since preparation of these
documents. Also, the regulatory criteria for disposal of
sludges by landfilling, combustion, land application,
and ocean disposal have been tightened.
This manual considers the upgrading of existing
dewatering processes, as well as the designing of
new ones, and pays particular attention to the needs
of small facilities. All currently employed technologies
are considered as well as emerging ones. Proper
sludge management requires the use of several
sludge treatment and disposal processes. Selection of
a dewatering process is not an independent step, nor
is it even the initial step. This selection process is
guided by many factors, including the public's
desires, the final disposal option selected, regulatory
requirements, and the size of the facility. Table 1-1
shows other EPA Technology Transfer publications
which may be used to complement this one when
preparing the feasibility design of a sludge
management system.
Design parameters, performance capabilities, and
design deficiencies for all dewatering processes are
presented. While some cost information (in the form
of a range of values) is included in certain sections of
this manual, the intent is not to give detailed cost
estimating information. Such information can be found
in the Technology Transfer Handbook for Estimating
Sludge Management Costs at Municipal Wastewater
Treatment Facilities (EPA 625/6-85-010). Some
specific cost information is, however, presented with
the case studies and design examples. The manual
Table 1-1. EPA Technology Transfer Sludge Management
Publications
Process Design Manuals
1. Municipal Sludge Landfills (October 1978)
2. Sludge Treatment and Disposal {October 1979)
3. Land Application of Municipal Sludge (October 1983)
Seminar Publications
1. Composting of Municipal Wastewater Sludges (August 1985)
2. Municipal Wastewater Sludge Combustion Technology
(September 1985)
Brochures
i. Environmental Pollution Control Alternatives: Sludge Handling,
Dewatering, and Disposal Alternatives for the Metal Finishing
Industry (October 1982)
Handbooks
1. Identification/Correction of Typical Design Deficiencies at
Municipal Wastewater Treatment Facilities (October 1982)
2, Estimating Sludge Management Costs at Municipal
Wastewater Treatment Facilities (October 1985)
Environmental Regulations and Technology Publications
i. Use and Disposal of Municipal Wastewater Sludge (September
1984)
specifically discusses those processes where the
most extensive and cost-effective dewatering
performance improvements have been made,
including air drying, centrifugation, belt press filtration,
and recessed plate pressure filtration using polymer
conditioning.
This document is current as of the summer of 1987
and includes detailed case history information on the
newer processes and equipment. These include
vacuum assisted dewatering beds; solid bowl
centrifuges with backdrive capability and optimized
bowl design; third generation belt filter presses; and
diaphragm filter presses. This manual describes the
capabilities of these and other dewatering processes
by presenting data from full-scale field testing and
operating installations. In most cases, the information
presented is for sludges produced during primary and
secondary municipal wastewater treatment. Chemical
sludges produced during advanced wastewater
treatment are given minimal coverage.
-------
In general, the manual has been prepared for use by
experienced engineers involved in the design,
selection, and specification of dewatering equipment.
The major types of dewatering processes discussed
in this manual include:
• Air Drying Processes (Chapter 6)
Sand ieds
Freeze Assisted Sand Beds
Vacuum Assisted Beds
Wedgewire Beds
Lagoons
Paved Beds
Other Innovative Processes
» Belt Press Filtration (Chapter 7)
• Centrifugation (Chapter 7)
High-G Machines
Low-G Machines
• Vacuum Filtration (Chapter 7)
• Pressure Filtration (Chapter 7)
Fixed Volume
Variable Volume
All of these processes are in use today, although a
process such as vacuum filtration is rarely seen in a
new installation. The manual does not provide
detailed discussion of mechanical processes which
have been installed at only a few facilities or
processes which do not have a proven background of
performance.
1.2 Objectives of Dewatering
The general objective of dewatering is to remove
water, thereby reducing the sludge volume. This
produces a sludge which behaves as a solid and not
a liquid, and reduces the cost of subsequent
treatment and disposal. In most cases, the percent
solids content of a dewatered sludge is set by the
requirements for subsequent treatment and disposal;
this percent solids content is always significantly
higher than the percent solids content of a thickened
sludge.
1.3 Location of the Dewatering Process
The combination of processes used for solids
treatment prior to dewatering, transport, and disposal
varies widely from plant to plant. Generally, however,
the dewatering process is preceded by one of the
following stabilization processes: anaerobic or aerobic
digestion; thickening by either gravity, centrifugation,
air flotation, or rotating screen type process; and
chemical or heat conditioning. In some cases, raw
sludge, particularly raw primary sludge, may be
dewatered directly, although the handling and the
method of ultimate disposal would have to be
considered. After the dewatering operation, further
stabilization may be provided by composting; volume
and organic reduction may be accomplished by
incineration; or the dewatered sludge may be
ultimately disposed of by transport to either a landfill
or a site for landspreading.
1.4 Using this Manual
This manual has been organized to allow users to
concentrate on areas of interest as easily as possible.
The following brief chapter and appendix descriptions
provide an overview of the manual's organization,
Chapter 2 - Preliminary Considerations
Discusses the size of the treatment facility, regulatory
concerns, and performance Capabilities of various
mechanical dewatering processes. Tables are used to
show the percent total solids achievable with different
mechanical processes.
Chapter 3 - Sludge Characteristics and
Preparatory Treatment
Presents brief discussion of sludge characteristics.
Sludges from primary, biological, and chemical
wastewater treatment are included. Sludge quantity
and quality data are related to how the sludge is
produced. In addition, sludge treatment prior to
dewatering (such as by thickening and stabilization
processes) is discussed in relation to dewatering.
Chapter 4 - Process Selection
Provides guidance for identifying the most cost-
effective system from the alternatives presented in
Chapters 5, 6, and 7. Considered are the method of
disposal, plant size, practicality, and costs. Emphasis
is on upgrading and retrofitting.
Chapter 5 - Conditioning
Presents the purpose and methods of sludge
conditioning, how the methods work, and how one is
selected. Inorganic chemicals, organic chemicals, and
thermal conditioning (alone and in combination) are
discussed. This chapter also includes tests for
selection of chemical dose and determination of
dewaterability by different devices.
Chapter 6 - Air Drying Processes
Proven, cost-effective technologies are emphasized.
Dewatering processes discussed include sand beds,
freeze assisted sand beds, paved beds, vacuum
assisted beds, wedgewire beds, sludge lagoons, and
emerging systems. For each process, the following
topics are discussed: performance data, advantages
and disadvantages, problems experienced, design
criteria, and O&M concerns.
Chapter 7 - Mechanical Dewatering Processes
Systems are discussed with a focus on theory of
operation. This chapter describes belt filter presses,
-------
centrifuges, filter presses, vacuum filters, and
emerging processes. Topics discussed include
performance capabilities, performance experience,
design criteria, and O&M concerns. Wherever
possible this chapter provides comparative
performance information,
Chapter 8 • Case Studies: Air Drying Systems
Detailed case studies for some of the systems
discussed in Chapter 6 are given. Presentations
include actual costs for capital equipment, O&M,
energy, labor, chemicals, and replacement parts.
Chapter 9 - Case Studies: Mechanical
Dewatering Systems
Detailed case studies for some of the systems
discussed in Chapter 7 are given. These include
high-speed centrifuge operation, low-speed
centrifuge operation, recessed plate filter press
operation, and belt filter press operation.
Presentations include actual costs for capital
equipment, O&M, energy, labor, chemicals, and
replacement parts. Included are comparative studies
of mechanical dewatering devices that were made on
either a plant scale or large pilot-plant scale.
Appendix A • Design Examples/Cost Analyses
Examples are presented for upgraded sand drying
beds, vacuum assisted drying beds, belt filter
presses, solid-bowl centrifuges, and recessed plate
filter presses. Step-by-step procedures for design
and cost projections are given. Cost information is
based on the recently issued EPA handbook on
estimating sludge management costs.
Appendix B - Operation and Maintenance:
Mechanical Dewatering Systems
This appendix discusses energy requirements; O&M
record-keeping (data, frequency, etc.); and simple
control tests and observations. The designer can use
this appendix when preparing the dewatering
system's O&M manual.
Appendix C - Manufacturers and Sources of
Equipment
Manufacturers of belt filter presses, centrifuges, and
filter presses appear.
-------
-------
Chapter 2
Preliminary Considerations
2.1 Introduction
The quantity of sludge produced in U.S. municipal
wastewater treatment plants was last estimated from
data obtained in the 1982 EPA Needs Survey. Table
2-1 presents this data for all sizes of treatment
plants. The number of Publicly Owned Treatment
Works (POTWs) in a particular size category is also
given.
It is interesting to note that the smallest plants, <2.5
mgd (0.11 m3/s), represent 91 percent of the POTWs
and produce less than 17 percent of the sludge. In
contrast, the largest plants, >100 mgd (4.38 m3/s)
represent less than 0.3 percent of the facilities and
produce more than 34 percent of the sludge.
A representative survey of U.S. facilities was
performed in 1980 by EPA's Office of Solid Waste to
determine the choice of sludge use/disposal options
by plant size. The results are shown in Table 2-2.
The "Other" category frequently means a lagoon or
temporary storage facility. Note that small to medium
sized facilities more frequently select some form of
land use/disposal option than do the large sized
facilities, which more frequently use incineration.
When either evaluating or selecting a dewatering
process, one must keep in mind the inherent
influence of both the prior wastewater and sludge
treatment processes as well as the subsequent use
or disposal practices. Choice of a use/disposal
Table 2-1.
Estimated Municipal Sludge Production by
POTW Size
POTW Size
mgd
<2.5
2.5 - 5
S - 10
10- 20
20 -50
50 - 100
>10Q
No. of
POTWs
14,168
631
352
187
125
40
41
Sludge
Produced
dry tons/yr
1,189,810
515,504
588,445
622,478
924,896
676,091
2,324,274
Total
percent
17
8
9
9
13
10
34
ton x 0.9072 = Mg.
mgd x 0.0438 = m3/s.
process is in turn strongly influenced by local, state,
and federal regulations.
A dewatering process cannot be evaluated without
considering the other processes involved in the
overall wastewater/solids handling system. This
evaluation or selection can be a complex procedure
because of the large number of possible
combinations of unit processes available for
wastewater treatment and sludge thickening,
stabilization, conditioning, dewatering, and ultimate
use/disposal. Figure 2-1 shows the unit processes
most commonly used to perform most of these
Table 2-2. Distribution of Sludge Disposal/Utilization Practices for 1,011 Surveyed POTWs (by % of dry sludge solids).
Practice
Landfill
Incineration
Land Application
Distribution and Marketing
Ocean Disposal
Other
Small POTWs
(<1 mgd)
31
1
39
11
1
17
100
Medium POTWs
(1 - 10 mgd)
34
1
38
17
-
10
100
Large POTWs
(>10 mgd)
12
32
21
19
4
12
100
Total of
Ail POTWs
15
27
24
18
4
12
100
mgd x 0.0438 = m3/s.
-------
Figure 2-1. Sludge processing options.
PROCESS STAGE
UNTREATED SLUDGE
Primary
Mixed
Chemical
PRELIMINARY
TREATMENT
PRIMARY
THICKENING
LIQUID
STABILIZATION
SECONDARY
THICKENING
CONDITIONING
DEWATERING
FINAL TREATMENT
Degritting
Conditioning
Centrifuge
Gravity
Flotation
Clarifier
Other!
Ill I
Anaerobic
Digestion
Unheated Heated
J 1
Aerobic
Digestion
Unheated I Heated
I L
Chemical
Centrifuge
Gravity
J_ L
Elutriation
Chemical
Thermal
III
Press
Belt
Vacuum
Centrifuge
Bed
I I.I J
Incineration
STORAGE
TRANSPORTATION
FINAL SITE
Drying
Composting
I I
Chemical
Ash
Cake
Compost
I » . T
Landfill
Road
Pipeline
Sea
31 . I
Retail
Agriculture
-------
functions. An evaluation procedure should start at the
bottom of the figure with the use/disposal options and
work back to a decision on the dewatering
technology.
This chapter discusses regulatory concerns and the
capabilities of mechanical dewatering processes.
2.2 Regulatory Concerns
Selection of dewatering equipment is seldom directly
governed by regulations and/or guidelines. Reviewing
authorities are concerned, however, with the
equipment's performance capabilities, its reliability,
and downtime for maintenance. Indirectly, the
selection of dewatering equipment is influenced by
the choice of a use/disposal system for sludge and
the wastewater treatment system.
Before selecting a dewatering process, the design
engineer should consult all regulations for a particular
use/disposal system to see if a minimum cake solids
concentration is required. For example, state
landfilling regulations usually stipulate a required
minimum cake solids concentration. This minimum
level can vary widely from state to state.
2.3 General Performance Capabilities of
Mechanical Dewatering Processes
In recent years there have been great advances in
mechanical dewatering processes. To help the
designer sort these processes out, this section of the
manual compares the performance of several
mechanical dewatering processes. However,
communities with adequate land available should also
study the air drying alternatives presented in Chapter
6, Air Drying Processes.
All of the various methods of mechanical sludge
dewatering have the capability to produce good
recovery (>90 percent) of feed solids and thus the
major differentiation is the cake solids content. The
capital cost and the O&M costs associated with
dewatering may be of secondary concern if there is a
high cost associated with the water content of the
sludge.
The range of cake solids produced by a common
type of dewatering process is the result of several
factors. These factors are discussed in more detail in
the specific sections dealing with each dewatering
unit. However, the key factors are as follows:
a. The ratio of primary to secondary sludge.
Inherently, secondary sludge will retain at least
twice as much water, kg H^O/kg total solids (TS),
as primary sludge.
b. The origin of the secondary sludge. High SRT
(sludge residence time) sludge retains more water
than low SRT sludge. Bulking sludges will retain
more water than non-bulking sludges.
c. The type and quantity of chemical conditioning
can either enhance or reduce the cake solids,
depending on the dewatering process employed.
d. The design and age of the dewatering equipment.
Older equipment cannot compete with modern
(1980 and newer) designs. Further, there may be
several models of the same type of equipment,
some of which will produce a drier cake than
others. Good examples are the various belt press
offerings and the diaphragm recessed plate press
vs. the standard recessed plate press.
e. The design and operation of the dewatering
stations will significantly impact cake solids
content. Dewatering equipment operated at
maximum solids capacity may sacrifice 3-5
percentage points in the final product dryness.
Drier cakes are produced at reduced operating
loadings.
f. Industrial discharges can both enhance or detract
from the capability of a dewatering unit. Fibrous
discharges, for example, can result in belt presses
producing a much drier cake.
Figures 2-2, 2-3, and 2-4 represent typical ranges
of cake solids produced by various means of
mechanical dewatering. This data is representative of
both organic (polymer) and inorganic conditioning;
cake solids are not adjusted for ferric chloride and
lime addition in these figures.
The figures are not recommended or suggested for
design purposes and are provided only to indicate the
relative capabilities of each type of dewatering unit
and the effect of various types of sludges on cake
solids. Depending on specific site factors, better or
worse results can be achieved.
In Figures 2-2, 2-3, and 2-4, ferric chloride and
lime are generally employed for vacuum filters and
filter presses. While analyses of cake solids indicate
that these cakes are higher in percent TS by 2-5
percentage points than a polymer conditioned sludge
from the same device, the water content per unit
weight of sludge solids is about the same. That is, a
ferric chloride and lime vacuum filter cake of 23
percent TS would contain the same weight of
water/kg sludge as would a polymer conditioned cake
of 20 percent TS.
Figure 2-2 shows the characteristic results achieved
when dewatering 100 percent raw primary sludge
(RPS) and 100 percent waste activated sludge (WAS)
using various kinds of dewatering equipment. If the
sludges were dewatered first and mixed after
-------
Figure 2-2. Dowatcred sludge cake percent solids for raw primary and raw waste activated sludges.
Raw
Primary
Sludge
Raw
Waste
Activated
Sludge
Dewatered Sludge Cake, Percent Total Solids
VF
BP
FP
DFP
VF
BP
FP
DFP
GBC
DCG
10 20
C = Solid Bowl Centrifuge
VF = Vacuum Filter
BP = Belt Press
FP = Filter Press
30
40
50
60
DFP = Diaphragm Filter Press
GBC = Gravity Belt Concentrator
DCG = Dual Cell Gravity Unit
-------
Figure 2-3, Dewatered sludge cake percent solids for mixtures of digested primary (P) and digested waste activated sludges
(WAS).
100% P : 0% WAS
70% P : 30% WAS
50% P : 60% WAS
30% P : 70% WAS
0% P : 100% WAS
VF
GBC
DCG
BP
FP
DFP
VF
GBC
DCG
BP
FP
DFP
VF
GBC
DCG
C
^tmmmm
BP
FP
DFP
VF
GBC
DCG
BP
FP
DFP
GBC
DCG
VF
BP
FP
DFP
10 20
C = Solid Bowl Centrifuge
VF = Vacuum Filter
BP = Belt Press
FP = Filter Press
30 40 50
DFP = Diaphragm Filter Press
GBC = Gravity Belt Concentrator
DCG = Dual Cell Gravity Unit
60
-------
FIguro 2-4. Dowatorod sludge cake percent solids for mixtures of raw and digested primary and secondary sludges and heat-
treated primary and secondary sludges.
Raw Primary
and Trickling Filter
Raw Trickling Filter
Digested Primary
and Trickling Rlter
Digested
Trickling Filter
Thermally Conditioned
Primary and Waste
Activated Sludge
Diwatered Sludge Cake, Percent Total Solids
FP
VF
DFP
BP
FP
VF
BP
DFP
FP
VF
DFP
BP
VF
BP
VF
BP
10 20
C = Solid Bowl Centrifuge
VF = Vacuum Filter
BP = Belt Press
30
40
FP
DFP
50
FP = Filter Press
DFP = Diaphragm Filter Press
60
10
-------
dewatering, the final moisture content of the cake
could be calculated from this figure as shown below:
Sludge at 60:40 RPS:RWAS
(P + WAS)
%TS Mixture =
(2-1)
%TS,
%TS
WAS
% TS Mixture =
(60 + 40)
60 40
30 17
% TS Mixture = 23% TS
It is necessary to substitute site-specific numbers
for the sludge proportion of 60:40 and cake solids for
each fraction. These numbers must be consistent
with the sludge at the specific site and representative
of the capabilities of the equipment supply and
process design. This equation can also be used to
estimate the moisture content of dewatered cake if
the mixing is done before dewatering. The estimate is
not conservative - it would be best to carry out
dewatering tests on the mixture.
The filter presses (recessed plate and diaphragm
plate) will always produce the driest cake solids with
the available technology. New dewatering equipment,
now in development or an early stage of application,
may rival the filter presses' ability to produce the
driest cake solids. Belt presses and centrifuges both
achieve approximately the same cake solids, with the
belt press more appropriate for sludges of higher
primary (more structured) content when highest solids
content is required.
Figure 2-3 provides cake solids content for various
methods of dewatering a digested primary and waste
activated sludge. Generally, digesting the primary
sludge before dewatering results in a slightly wetter
cake, probably a result of the finer particles produced
by anaerobic decomposition. However, those plants
with poor grit removal, and hence a higher proportion
of grit in the digested sludge, may produce a drier
digested sludge than would be produced if it were
dewatered in the raw state.
Mechanical thickening devices such as the Gravity
Belt Concentrator and the Dual Cell Gravity Unit
produce a much lower solids content then the other
dewatering devices. However, their simplicity and
ease of operation can be well suited to small plants
that either stockpile and/or land spread dewatered
sludge.
Figure 2-4 provides dewatering information for raw
and digested trickling filter plant sludges. These
sludges behave similarly to an activated sludge
plant's waste solids where the sludge ratio is about
65:35 RPS:WAS.
Heat-treated sludges are often dewatered without
chemical conditioners. However, they will require
polymers to control the suspended solids recycle with
centrifuges and belt presses. However, dosage will be
low - about 20-25 percent of that required for the
sludge before thermal treatment.
Some general conclusions can be drawn from these
figures. They are:
a. Solid bowl centrifuges and belt presses can
produce about the same cake solids. However,
belt presses with high pressure attachments can
generally produce 2-3 percentage points higher
cake solids with sludges of good structural
characteristics.
b. The low energy gravity drainage units that produce
a cake of a low solids content may be attractive
for small plants and where land spreading disposal
is practiced.
c. The recessed plate filter presses will produce 6-
10 percentage points drier cake than the
continuously fed dewatering units; addition of the
diaphragm can increase the cake solids content
3-5 percentage points more.
d. In general, digested sludge cakes will have a
solids content 2-3 percentage points lower than
that of the raw sludge, with the exception of those
with a low volatile content due to grit.
e. The range in cake solids from plant to plant will be
quite wide due to specific site conditions and
design and operating factors.
2,4 Key Operations Variables Affecting
Mechanical Dewatering Performance
Table 2-3 presents the more significant design and
operating variables that impact the performance of the
specific dewatering device. (More detailed information
is provided in the various sections of the manual
reviewing each dewatering process.) These variables
affect the rate of production, solids recovery, and
cake solids, as well as the type and quantity of
conditioning chemicals and dewatering aids
employed.
Table 2-3 lists only those variables that affect
performance and that can be changed in the field and
during design of the dewatering station. Variables
11
-------
Tablo 2-3. Operational Variables
Dewatering Processes
for Mechanical
Vacuum ..Filter
1 Drum Speed/Cycle Time
1 Drum Submergence
1 Sludgo Food Concentration
2 Quantity of Wash Water
Used
2 Filter Madia Used
2 Conditioning Chemicals -
Typo and Dosage
3 Vat Agitation
Bolt Filter Pross
1 Bolt Speed
1 Sludge Food Concentration
1 Polymor Conditioner
* Type & Dosage
• Dosage
• Point of Addiiion, Contact
Time, Mixing
2 Boll Tension
2 Bolt Type
2 Washwater Flow & Pressure
Solid Bowl Centrifuge
1 Bowl/Scroll Differentia!
Speed
1 Pool Depth
2 Potymor Conditioner
• Dosage
• Point of Addiiion
2 Mode of Differential Speed
Control
3 Sludge Feed Concentralion
Recessed Plate
1 Pressure of Feed Sludge
1 Filtration Time
1 Conditioning Chemicals
2 Use of Precoat
2 Frequency of Cloth Washing
2 Filter Cloth Used
Diaphragm Plate Press
1 Diaphragm Pressure
1 Conditioning Chemicals
• Type & Dosage
• Point of Addition,
2 Pressure of Feed Sludge
2 Filtration Time
2 Diaphragm Squeezing Time
2 Filter Cloth Used
2 Frequency of Cloth Washing
2 Sludge Feed Concentration
Gravity/Low Pressure
Dewatering
1 Polymer Dosage
1 Retention Time
2 Sludge Feed Concentration
2 Bell Speed
2 Force Applied By Rollers
2 Depth of Dewatered Sludge
in Cylindrical Devices
Legend: 1 Very important variable
2 Significant variable
3 Less important variable
controlled by the equipment suppliers, such as the
bowl angle of a centrifuge, have not been included.
The table does not include sludge feed rate. If you
set the variables shown for a vacuum filter, feed rate
is fixed. The same is true for pressure filters. In a belt
press, increasing feed rate will flood or starve the
filter. In a centrifuge, you can vary feed rate at will -
you will just get a more or less dry cake and less or
more solids in the centrate. Feed solids concentration
has been included since some dewatering processes
are more sensitive to feed concentration than others.
12
-------
Chapters
Sludge Characteristics and Preparatory Treatment
3.1 Sludge Production and Concentration
3.1.1 Introduction
There are several sources of wastewater sludges;
these sludges can vary widely in characteristics and
quantity. From the standpoint of quantity per unit of
flow, the principle variables are the strength of the
wastewater, whether chemicals are utilized in the
process, and the degree of treatment,
The typical wastewater sludges are classified as
primary, biological, and chemical. The biological
sludges produced are activated sludge and fixed film
sludges from rotating biological contactors and
trickling filters. The activated sludge may have
primary sludge solids incorporated into the biomass
when primary clarifiers are not employed.
Chemical sludges may be produced simultaneously
with primary sludge or biological sludge through the
addition of metal salts for precipitation of phosphorus,
or they can be made in a separate tertiary treatment
stage. Lime is sometimes used in the primary
treatment stage and also in a tertiary stage, when
softening of the effluent is required for reuse. The
reader is also referred to the EPA Technology
Transfer Process Design Manual - Phosphorus
Removal (EPA-625/1 -87/001) for a discussion of
the production and dewatering characteristics of
chemical sludge.
In some cases, well designed sludge handling
systems were actually marginal in operation due to
inaccurate estimates of wastewater treatment
loadings of BODs (5-day biochemical oxygen
demand) and TSS (total suspended solids) and the
subsequent sludge production. These problems
occurred for a variety of reasons as outlined below:
• Low estimate of unit sludge yield/unit of COD
(chemical oxygen demand) or BODs removal
Use of average weekly or monthly
TSS inputs
and
» Inaccurate estimate of primary treatment
efficiency
• Effects of BODs and TSS recycle ignored or
underestimated
• Seasonal discharges of BODs and TSS
overlooked.
3.1.2 Primary Sludge
The raw primary sludge (RPS) production is easily
determined from the total flow and the influent and
effluent TSS (total suspended solids) of the primary
clarifier. Care should be taken to ensure that the
influent sample, which does not contain recycled
solids, is the same as the primary clarifier influent.
Some adjustment of the influent is necessary to
account for recycled solids removed by primary
clarification. Even with good operation, recycled TSS
can amount to 15-20 percent of influent TSS, and
the BODs recycle is usually 8-15 percent of the
influent
• No allowance for the normal peak day/average
discharge characteristics of larger industrial
facilities
The influent loadings and resulting sludge production
should be analyzed and developed into a frequency
plot, which would indicate the frequency of a specific
TSS and BODs influent loading (kg/d) vs. < % time
(frequency). Similar graphs should be plotted for RPS
produced and PE (primary effluent) BOD§ (kg/d vs. %
time). Figure 3-1 is a typical example.
A mass balance of the overall process should be
prepared to ensure accounting of all TSS and BODs.
An example is shown in Figure 3-2, and indicates
the importance of identifying the magnitude of the
recycle BODs and TSS.
Unlike secondary sludge, the volatility of primary
sludge may vary considerably from day to day and
seasonally. This is particularly true of sewage
systems with combined sewers and/or substantial
infiltration and inflow.
The domestic and commercial discharges of volatile
suspended solids (VSS) would not vary widely
throughout the year. Some short-term increases
may be noted due to "first-flush" effects during
sudden wet weather conditions. First-flush effects
13
-------
Figure 3-1.
Influent BOOS average monthly loading, 1985-
1986.
30.000
20.000
J" 10,000
0
o
CD
50th Percentile = 24,000 Ib/day
80th Percentile = 27,500 Ib/day
0.51 2
I 10
5 20
4—'—I ' t—i—I 1—I—i—I—i—H 1
I 30 I 50 I 70 I 90 I 98 |99.5| 99.9 I
40 60 80 95 99 99.8 99.99
Ib/day x 0.4536 = kg/day.
occur with the transport of accumulated solids in the
sewers and street washing where there are combined
sewers. Where there are seasonal or variable
industrial discharges, the VSS may vary widely, and
the primary removals of TSS and BOD5 may ajso
vary, depending on the nature of the solids.
Designers should anticipate that reductions in the
primary sludge volatile content will generally be
accompanied by proportionally more sludge, even
though there may be only a small or no increase in
the VSS loading. An example follows:
TSS. Ib/d
VSS, Ib/d
Volatile %
Dry Weather
Wet Weather
100.000
130.000
78,000
78,000
78
60
Ib/d x 0.454 - kg/d
The sludge production at the lower volatile content is
30 percent higher than the dry weather sludge
quantity. While the lower volatile content sludge is
somewhat easier to handle due to the grit content,
the sludge handling design should anticipate the
higher quantity of sludge solids. In existing plants, the
past operating records should be scrutinized to
determine if there is a significant variation in the
actual sludge volatility and the sludge quantity
projected.
The efficiency of primary clarification is important
because not only is the primary sludge easier to
handle, but the unit yield (kg/kg) of secondary sludge
is partially dependent on the TSS/BODs in the
clarified mixture. Figure 3-3 presents the yield and
the dewatered cake concentration of various
RPS:WAS ratios. When primary clarifiers are not
employed, the total quantity of solids produced is
lower than RPS + WAS, but the water retention
characteristics of the biological solids increase. While
the absolute values shown in Figure 3-3 vary,
depending on sludge characteristics and the
mechanical equipment employed, the general
relationship holds.
In cases where the highest possible cake solids are
required, good primary treatment should be provided.
The primary clarifier requirements can be
experimentally determined using laboratory settling
tests, if the wastewater is available or by evaluating
the performance characteristics of existing units at
various flow rates. The clarifier performance is
strongly influenced by the overflow rate (OFR,
m3/m2/d or gal/sq ft/d) and the clarifier sidewater
depth (SWD). Good performance of circular primary
clarifiers will be achieved when:
ENGLISH OFR, max < IOSWD2 (SWD = 6 < 10ft)
OFR, max < 100 SWD (SWD = 10 < 15 ft)
METRIC OFR, max < 4.5 SWD2 (SWD = 2 < 3 m)
OFR, max < 12.25 SWD (SWD = 3 < 5 m)
For rectangular clarifiers, the length of the flow path is
most important to overcome inlet disturbances. The
depth of the basin is also significant. For basins less
than 30 m long, the length to width ratio is >5:1.
Basins that are 30-65 m long and 4-5 m deep will
provide excellent results, even at rates up to and
exceeding 67.2 m3/m2/d (1,650 gal/sq ft/d).
3.7.3 Chemical Treatment of Raw Wastewater
When chemicals are added to the raw wastewater for
removal of phosphorus or coagulation of non-
settleable solids, larger quantities of sludges are
formed. The quantity of solids produced in the
chemical treatment of wastewater depends upon the
type and amount of chemical(s) added, the chemical
constituents in the wastewater, and the performance
of the coagulation and clarification processes. It is
difficult to predict accurately the quantity of chemical
solids that will be produced. Jar tests are preferred as
a means for estimating chemical sludge quantities;
As discussed in Section 5.5.3.1, Table 3-1 provides
estimated quantities of suspended and chemical
solids removed in a hypothetical primary
sedimentation tank processing wastewater that has
been treated with lime, aluminum sulfate, or ferric
chloride. The use of polyelectrolytes may greatly
enhance the solids capture in the clarifier. The
removal of TSS is usually in the range of. 75-85
percent and BOD5 removal is 55-70 percent
depending on the specific wastewater characteristics.
14
-------
Figure 3-2. Process mass balance.
Inplant
Wastewater
20,66
795,600
595,800
313,000
17
18
19
0,033
7,300
4,500
6,600
0,025
1,100
700
1,900
0.052
24,600
18,BOO
11,100
0.007
6,200
3,800
4,700
FLOW (mgd)
TSS (Ib/dayS
VSS (Ib/day)
BOD ilb/day)
The chemically defined soluble BODs will be
appreciably less than the filtrate soluble BODs since
some colloidal BODs will be agglomerated and
settled.
The solids precipitated are defined as Cax(P04)y,
CaCO3, FePO4, Fe(OH)3> AIPO4 and AI(OH2)3-
However, some of these precipitates will be hydrated
and represent more sludge than shown in Table 3-1.
This is why the metal salt precipitates are so
voluminous. When sludge solids are dried at 103-
105°C (217-221°F) for solids analysis, some of this
water is lost during the test procedure. However, the
hydrate moisture adversely affects the ability of the
sludge to be thickened and dewatered.
3.1.4 Biological Sludge Yield
Sludge yields will vary widely from plant to plant
depending on the overall treatment plant
configuration, wastewater characteristics, and the
biological kinetics/parameters employed for design. A
wastewater with a high COD/BODs ratio will produce
more excess biological sludge solids. Net yields
are determined from Equations 3-1 and 3-2.
YN = a BOD5R - b (M)
1/SRT = a (PR/M) - b
where,
SRT = M/YN
SRT = aerobic sludge age, days
PR = BODsR (removed), kg/d
M = MLSS or MLVSS, kg
a = synthesis, kg/kg BODsR
b = endogenous decay, dayl
YN = net yield, kg TSS or kg VSS
(3-1)
(3-2)
15
-------
Figure 3-3. Cake solids as a function of primary clarifier
efficiency and PS:WAS ratio.
*
1
o
20
cc
V)
p
sr
40
60
30r
CO
P
8
28
18
Domestic Wastewater
TSS/BODj = 1.0
COD/BODS = 2.0
SBOD./BODB - 0.33
j I i i
20
40 60
PS, % of Mixture
80
100
Both the synthesis value "a" and endogenous decay
value "b" are reported in the literature on a TSS and
VSS basis and thus the origin of these values must
be distinguished. As shown in the WPCF Manual of
Practice No. 8, Treatment Plant Design, the
COD/BODs and TSS/BODs values must be specified
in order to project a net yield as a function of SRT
and temperature. Thus, the values of "a" and "b" are
variable and both temperature and SRT dependent.
Since sludge yields are higher at lower temperatures,
it is necessary to use the colder period of the year to
project maximum sludge production, when the organic
loading is uniform throughout the year.
Rgure 3-4 indicates the range of sludge production
to be expected at temperatures between 10 and 30°C
(50-86° F) as a function of the SRT with and without
primary treatment In this case, the wastewater is
specifically identified as typical domestic wastewater
at 400 mg/I COD, 200 mg/I BODs, and TSS, and
primary clarifier effluent is as noted. This curve
should not be used for wastewaters having different
relative COD/BODs/TSS ratios either in the raw
wastewater or in the settled primary effluent.
When the COD/BODs ratio of the primary effluent
exceeds the values shown in Figure 3-4, higher
excess biological solids production will often occur if
the COD is removable by adsorption/oxidation. Los
Angeles-Hyperion, Columbus-Southerly, and
Columbus-Jackson Pike all have primary effluents of
2.3-2.7:1 COD/BODs. Secondary sludge yields of
0.75-0.90 kg EAS/kg BOD5 are produced (EAS =
excess activated sludge (WAS + effluent TSS)j. Not
suprisingly, the COD yields (kg EAS/kg CODR) are
more consistent. The BOD test may be, on occasion,
a poor indicator of the yield.
3.J.5 Biological Phosphorus Yield
When phosphorus is removed, the net sludge yield
will increase measurably. The increase will be similar
to that experienced when Fe
or AI+
addition is employed to precipitate the phosphorus
chemically. That is, the excess phosphorus
precipitated by manipulation of the biological
environment is also an inorganic salt of K, Ca and
Mg. It is recommended that the following procedure
be used to determine this excess sludge.
Influent BOD5:
Effluent SBODs:
.-. BOD5R:
Influent TP:
Effluent STP:
Calculated sludge yield:
Normal sludge P:
Excess bio-P removal
140 mg/I
5 mg/I
135 mg/I
8 mg/I to bio-treatment
2 mg/I (soluble)
65 mg/I VSS
80 mg/I TSS
2.0 percent VSS
= 8 - 2 - 0.02(65)
= 4.7 mg/I P
Additional sludge = 4.5 x 4.7 = 21 mg/I
(4.5 mg/mg AP)
Total sludge production = 80 + 21 = 101 mg/I
The value of 4,5 was based on an average MW of
140 for the inorganic phosphorus crystals in the
biological cells. In this example, the biological sludge
yield would have increased from a calculated value of
0.61 mg EAS/mg BODsR to 0.75 mg/mg, or 21
percent higher.
Typical yield coefficients found in all textbooks and
other reference materials do not allow for this higher
sludge yield. However, recent papers in biological
phosphorus removal have confirmed the MW is about
140.
The phosphorus balance [primary effluent (PE) to
final effluent (FE)] provides an easy and direct
method to determine the net yield, SRT, and system
MCRT (mean cell residence time). MCRT in this
16
-------
Table 3-1, Typical Production of Primary and Primary-Chemical Sludges1
Mode of Operalion Dosage of Chemical Raw TSS Removed
Raw BODs Removed Chemical Sludge Produced
Plain Sedimentation
Polymer Added
CaO Aided2
FeCI3 (as Fe)3
AI2SO4 (as Al)3
mg/l
-
0,5 - 3.0
200
12
12
mg/l
120
150
160
160
160
mg/l
60
90
120
120
120
mg/l
-
-
128
47
46
1 Based on 200 mg/l BODg, 200 mg/l TSS, and 10 mg/l TP in raw sewage; primary effluent < 2 mg/l total phosphorus,
2 Varies due to permanent hardness in the water, used 35 mg/l precipitated as
3 May require polymer addition to enhance clarification.
discussion includes the clarifier sludge inventory.
Equation 3-3 describes the procedure.
SRT =
(MLSS-kg)(%P/lOQ)
(PETP - FETP)(m3/d)(l ,000)
MLSS + Clarifier TSS
- '• -
MLSS
(3-3)
(SRT)
MLSS MLSS + Clarifier TSS
Sludge Yield = or
SRT
MCRT
The method is quite accurate If phosphorus removal
is relatively constant and the ratio of MLSS to the
clarifier sludge is relatively constant. Three- or five-
day running averages are better estimates of YN,
SRT, and MCRT.
3.1.6 Chemical Phosphorus Precipitation in the
Biological System
The use of metal salts for precipitation of phosphorus
in suspended film biological systems is widely
practiced. The most common salts used are ferric
chloride and sulfate and similar salts of aluminum.
Pickle liquor, ferrous sulfate, and ferrous and
aluminum chloride are also employed. When metal
salts are used, it may be necessary to provide
additional alkalinity to the aeration basin.
The metal salts are generally employed in excess
molar ratio, i.e., moles AI:P or Fe:P. The excess
metal salts form hydroxides of the metal and
precipitate. The sludges produced are as follows:
AIPO4 = 121/31 or 3.9 mg/mg P removed
Ai(OH)3 = 77/26 or 3.0 mg/mg excess Al
FePC>4 = 151/31 or 4.9 mg/mg P removed
Fe(OH)3 = 107/56 or 1.9 mg/mg excess Fe
The residual soluble total phosphorus as a function of
the molar ratio is approximately as follows:
Molar Ratio
(metal:TP)
1.0
1.5
2.0
Residual STP
(mg/l)
2
1
0.3
The residual phosphorus at low levels is highly
dependent on pH, and it may be more economical to
increase the pH by adding alkalinity which will not
produce sludge.
3.2 Sludge Concentration - Primary
Ciarifiers
Sludge feed concentration is an important factor in
sludge processing units such as digestion and
dewatering. In larger plants separate sludge pre-
thickening is economically viable, but this is not true
for most plants of less than 0.2 m3/s (5 mgd). [Over
95 percent of the sewage treatment plants in the
United States are less than 0.2 rr>3/s ( 5 mgd)]. While
it is not a specific goal of this manual to provide
information regarding pre-concentration of sludges,
it !:: necessary to consider the impact of sludge
concentration on sludge dewatering and sludge
handling in general.
Most of the smaller plants in the United States will not
have separate thickening and will utilize primary
treatment units to co-settle and thicken raw primary
and waste secondary solids. The problems
encountered in this procedure are well known and
documented. Many of the problems associated with
using the primary clarifier as a thickener are the result
of excessive solids retention time in the clarifier. It is
sludge retention time, not liquid retention time, that is
the primary cause of odors in primary clarifiers. Such
odors result from increasing the sludge blanket
(inventory) over the sludge withdrawal pipe to
maximize the sludge concentration and prevent rat-
holing of the sludge. It is, however, possible to use
17
-------
Figure 3-4, Process mass balance.
Q
o
1.3
1.2
1.1
1.0
{2 0.9
0.7
0.6
f
55 0,5
I
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Sludge Yield Without Primary Clarification
- 10°C
TSS/Total BOD. = 1.0
Inert TSS = 0%
j I
I
j
4 5 6 7 8910 15 20 30
Solids Retention Time, days
Sludge Yield With Primary Clarification
10°C
COD/BOD, =
TSS/BOD, =.
Primary Treatment
@ 60% TSS Removal
30% Inerts In Primary
Effluent TSS
j_
i i i
J I
0.4 0.60.8 1 1.5 2 345 678910 15 20 304050
Solids Retention Time, days
this procedure effectively with minor changes in the
design approach and negligible additional costs.
Where the primary clarifier must serve a dual function
of clarifying the wastewater as well as delivering a
concentrated sludge for digestion or dewatering, the
conventional primary clarifier design configuration is
inappropriate. This is particularly true for primary
clarifiers in smaller plants, where it may be necessary
to have 1.5-2.0 days SRT in the clarifier to create a
1.0 to 1.5 m (3-5 ft) sludge blanket above the
sludge withdrawal pipe. The build-up of sludge to
produce a thicker underflow interferes with
clarification (lower efficiency) and sometimes results
in gasification, odors, and floating sludges.
However, in the new plants there is an easy remedy
for this problem. Construct the smaller clarifiers with
the standard thickener floor slope of 2.75:12. In larger
clarifiers, use a dual slope clarifier where the inner
slope of 2.75:12 is the thickening zone and the outer
zone is 1:12. On a primary clarifier, only 40-50
percent of the diameter would be required for sludge
thickening; thus it is sufficient to modify the floor
slope at mid-radius.
Figure 3-5 shows the types of clarifier floor
configurations of which only three are suitable for
efficient combined clarification and thickening. Type A
is the design most commonly employed, but it is not
suitable for combined clarification and thickening.
Types B, C and D all can provide much better
performance in terms of thickened sludge
concentrations and lowest sludge inventory; hence,
they provide freshest sludge and highest flexibility in
terms of sludge removal and ease of operation.
Pertinent dimensions for a 30 ft diameter and 80 ft
diameter primary clarifier-thickener are also shown
in Table 3-2.
Process data used' to construct the sludge level in
Table 3-2 are provided below. Underflow
concentrations were based on general experiences
with sludge thickening.
RPS @ 120 mg/l = 1,000 Ib TSS/mgd
WAS @ 80 mg/I 667 Ib TSS/mgd
Total = 1,667 Ib TSS/mgd
9.1 m (30 ft) diameter @ 1.36 m3/m2d (800 gal/sq
ft/d):
TSSp (TSS removed) = 943 Ib TSS/d
24 m (80 ft) diameter @ 32.6 m3/m2d (800 gal/sq
ft/d):
TSSR = 6,706 Ib TSS/d
Sludge blanket depth in small clarifiers is a two-fold
problem in units with the 1:12 floor slope. As shown
in Table 3-2, the 1.0 day SRT sludge depth is only
18
-------
Figure 3-5. Effect of floor configuration on blanket location and sludge inventory.
TYPE A d«
dswd
RECOMMENDED
OPERATION
Separate
Thickening
Slope = 1:12
TYPE B
Combined Clarification
and Thickening
d « 15 m
Slope = 2.5:12
TYPEC
d,
Combined Clarification
and Thickening
d > 15-40 m
Slope = 1:12 and 2.5:12
dt = 0.5d
TYPE D
Combined Clarification
and Thickening
d Tf 15-40 m
Slope = 1:12 and 2.5:12
dt = 0.4d
dh = 0.6-1.2m
LEGEND
d = Diameter
dw = Depth of Water
ds = Depth of Sludge
swd = Side Water Depth
dcwd = Center Water Depth
dt = Diameter at Steeper Slope
dh = Height of Vertical Wall
vs
19
-------
Tablo 3-2. Primary Clarifier Comparison - Sludge Blanket
Location, and Sludge Concentration vs. Floor
Configuration (see Figure 3-5 (or
nomenclature)
Type
Slopo
30-lt diamoter
dswd> N
scwd« 't
ds«, ft
dw.lt
dh.'t
VC, H3
Vg, H3/d
Underflow,
%TSS
A
1:12
9.0
10.3
1.9
8,4
-
313
544
2.75
B C
> 2.5:12 1:12
> 2.5:12
9.0 NR
12.1
2.9
9.0 +
-
737
428
3.5
D
1:12
> 2.5:12
9.0
14.1
4.3
9.0 +
3.0
6122
313
4.5
80-it diamelor
dswd< 't
SCwcf, It
ds". It
d«, It
dh,lt
VG. H3
Vg, ft3/d
Undorflow,
%TSS
n.O NR
14.3
3.2
11.0 +
-
5,582
3,546
3.0
4
11.0
16.8
5.0
11.0 +
-
6,623
2,680
4.0
11.0
18.8
6.2
11.0 +
2.6 ;
7,726
2,364
4.5
1 Basod on 1,000 Ib PS/d + 600 !b WAS/mgd sludge inventory @
0.75 % TSS in underflow (1-day SRT),
2 Based on conical plus cylindrical volume below cone.
NR • not recommended.
V(j « volume of conical section
Vs • volume of sludgs at 1 -day inventory.
ft x 0.03048 * m.
cu ll x 0.0283 - m3.
0.58 m (1.9 ft) in clarifier type A. In smaller plants,
normal practice would be to pump a fewer number of
hours/day at a proportionally higher rate. This
procedure generally results in a diluted sludge being
processed with less efficiency.
Type D in Figure 3-5 will provide the highest sludge
concentration since it most closely approximates a
thickener design, has the deepest sludge inventory
and has the sludge inventory closest to the
withdrawal point. This design has been available and
employed in both municipal and industrial facilities for
wastewater treatment. The Type D unit is illustrated in
Figure 3-6 and, as shown in Figure 3-5, the sludge
blanket would still be in the lower zone with a sludge
SRT of over one day in the unit. The slope is 0.5 to
1.0:12 in the outer area and 2.5 to 3.0:12 in the
thickening zone.
Type B is a design that is simpler and quite suitable
for clarifiers up to about 15 m (50 ft) diameter. The
bottom slope would be 2.5 to 3.0:12 with the steeper
slopes oriented to the smaller diameter. As a general
rule of thumb, the depth (dc) of the coned section
should not be less than 1 m (3 ft). Beyond 15-18 m
(50-60 ft) diameter, the depth of reexcavation
becomes a factor and much: of the coned volume is
not needed for sludge storage.
Type C units are most often installed in large
industrial clarification-thickening operations for both
process and wastewater treatment. The configuration
has been employed for both primary and secondary
treatment stages. As in Type D, the slope of the outer
zone is usually decreased as the diameter increases,
since sludge conveyance to maintain inventory and
depth are only critical in the thickening zone. Outer
zone floor slopes of 0.5:12 are common above 30 m
(100 ft) diameter to minimize the overall depth of the
unit. While Type C and D clarifiers larger than 60 m
(200 ft) in diameter have been installed, most
municipal wastewater treatment plants with more than
15,000 sq ft of clarifier capacity will employ separate
thickening.
Those smaller treatment facilities (less than 5 mgd)
employing primary clarifiers may also find it
advantageous to use a floor slope of 2.0 to 2.5:12 on
secondary clarifiers. This slope will reduce the
operational problems associated with maintaining a
high MLSS (mixed liquor suspended solids) since
there will be a higher return sludge concentration and
lower waste sludge volume. The plant will also be
more efficient and easier to operate, both in the wet
and solids ends.
In larger plants without primary clarifiers, pre-
concentration of the WAS is recommended prior to
sludge dewatering for two reasons: (1) to reduce the
volume; and (2) to provide some buffer capacity
between the sludge wasting schedule and the
dewatering operating schedule. The purpose of pre-
concentration should be to increase the waste
activated sludge (WAS) concentration by 50 to 100
percent, i.e., from 0.6-1.0 percent to 1.5-2.0
percent TSS. These high-rate gravity thickening
systems can operate at 3 to 5 times the solids
loading versus those where ultimate compaction is
required. Flotation and centrifugal thickening are also
employed to reduce waste sludge volumes prior to
subsequent processing and dewatering and will
produce 4-7% TS. .
3.3 Characteristics of Waste Sludges
3,3,1 Specific Gravity and Volatility
The specific gravity of sludge will be in part a function
of the amount of grit and fine inert particles in the
sludge. These inorganic particles will have a specific
gravity of 2.5-2.9. Where there is good degritting,
the specific gravity of sludges will have the volatile
and specific gravities shown in Table 3-3(2-4). The
specific gravities of sludge solids are quite low and
will vary depending on the source.
The specific gravity of fixed film biological sludge is
generally higher than that of waste activated sludge.
20
-------
Figure 3-6. Combination primary clanfier-thickener.
Effluent Pipe
Feed Well
Influent Pipe
Sludge Pipe
Table 3-3. Specific Gravity of Waste Sludges
Volatility
Range of
Specific Gravity
Sludge Type
percent
9/cc
RPS
WAS
TF & RBC
RPS + WAS
75-80
80 - 85
75 -80
75 -85
1 + 0.010{%TSS)
to
1 +0.012 (%TSS)
1 + 0.007 (%TSS)
to
1 + 0.012 (%TSS)
1 + 0,015 (%TSS)
to
1 +• 0.025 (%TSS)
1 + 0.004 (%TSS)
to
1 +• 0.006 (%TSS)
RPS = Raw primary sludge.
WAS = Waste activated sludge.
TF = Trickling filter.
RBC = Rotating biological contactor.
This is evidenced by a lower SVI and generally higher
settling rates. The specific gravity of the sludges after
anaerobic digestion will increase due to reduction of
some of the hydrous fractions and the increased inert
content.
3.3.2 Pre-Concentration or Thickening of Waste
Sludges
Raw primary sludges are the easiest to thicken
followed by fixed film sludges. Waste activated sludge
is most difficult to thicken, particularly if the SVI is
high. Chemical sludges produced from Ihe addition of
metal salts thicken similarly to waste activated sludge
at a SVI = 100 ml/g, but they are more stable. Aging
of sludge after removal from the raw wastewater or
the aerobic environment causes deterioration of the
thickening quality.
The general experience in thickening sludges is
shown in Table 3-4. The results achievable in the
primary clarifier are dependent on the clarifier design
as reviewed earlier. Thickening increases the solids
content of sludge slurry by a partial, but substantial,
removal of the liquid phase. The purpose is to reduce
the sludge volume to be stabilized, dewatered, or
hauled away. Figure 3-7 shows the importance of
thickening prior to mechanical dewatering. Thickening
can be accomplished by partial thickening in a
primary or secondary clarifier, a gravity thickener, a
dissolved air flotation thickener, a centrifuge, a gravity
or low pressure belt press, or a rotary drum device.
Gravity thickening of raw or digested primary sludge
is almost always an efficient and economical process.
Anaerobically digested primary sludge is normally
thickened by gravity in the secondary digester. The
use of primary basins to capture and to thicken both
wastewater influent and recirculated WAS solids, may
not always be a cost effective and efficient practice in
larger plants. The WAS solids may not resettle well in
hydraulically overloaded or septic primary tanks.
Hence, this practice results in the production of more
WAS due to an increased solids load on the aeration
system. Poorer thickening results when the primary
basins are employed to concentrate the WAS solids,
particularly if the bottom configuration is not
conducive to thickening.
The use of gravity thickeners for both RPS + WAS
has had mixed results. Most of the poor results can
be traced to one or more of the following causes:
21
-------
Tablo 3-4. Thickening of Waste Sludges
Type of Sludpa
TSS Concentration, percent
Primary Clarltisf
FIotation/DAF
Gravity Thickener
Belt Thickener1
Centrifuge
RPS 5-7
WAS 3 - 5
FFS2 3 - 5
RPS + WAS 2.5-4 4-6
RPS + FFS 3-5 4-6
8-10 9-12 9-12
2-2.5 4-6 4-6
2.5-3 5-7 5-7
4-5 5-7 . 5-7
5-6 5-10' 6-10
1 Poiymors required. >
2 Fixod film sludge.
Figure 3-7. Effect of feed solids on performance of a rotary
vacuum filter.
90r
80
70
3
3
£ 60
C
8
10
12
14
Sludge Solids Concentration, %
a. RPS + WAS feed concentration is >0.5 percent
TSS.
b. RPS is very septic.
c. WAS is > RPS fraction.
d. Secondary dilution water is inadequate.
e. Floor slope is < 2.5:12, causing excessive solids
retention.
f. Sludge is not removed continuously.
Properly designed and operated gravity thickeners
work effectively on mixtures of RPS and WAS
throughout the United States. Misusing them as
sludge storage zones causes operator grief. If storage
is necessary, it must be placed after the gravity
thickeners.
The use of other thickening methods such as
dissolved air flotation, basket or solid bowl
centrifugation, low pressure belt filtration and the
rotary drum system has increased because these
methods can also give reliable and effective results
when thickening WAS.
3.3.3 Particle Surface Charge and Hydration
Sludge particles have a negative surface charge and
try to repel each other as they are brought together.
Additionally, sludge particles weakly attract water
molecules to their surface (hydration) either by weak
chemical bonding or by capillary action. Although the
water is only weakly held at the particle surface, it
does resist thickening and interfere with dewatering.
Chemical conditioning is used to overcome the
effects of surface charge and surface hydration.
Typical chemicals are organic polymers; lime, ferric
chloride and other metallic 'salts. Generally they act
by reducing or eliminating the repulsive force, thus
permitting the particles to come together or flocculate.
Water can be more readily removed at a higher rate
during the subsequent mechanical dewatering. Sludge
conditioning is discussed in detail in Chapter 5.
3.3.4 Particle Size
Particle size is generally recognized as a very
important factor influencing dewaterability. As the
average particle size decreases, the surface area and
surface-to-volume ratio for a given sludge mass
increases. The effects of increasing the surface area
include:
• Greater repulsion between particles due to the
larger area of negatively charged surface
* Greater attraction of water to the particle surface
due to more sites for chemical joining.
Particle size is influenced by both the sludge source
and prior treatment. Primary sludge, in addition to
containing more inorganic and fibrous materials, has a
larger average particle size than secondary sludge.
This is because fine suspendable and colloidal solids
tend to pass through the primary clarifier. Sludge
particles passing the primary ciarifier are then
removed in the secondary clarifier along with the less
dense, flocculated cellular material that is created
during biological treatment. The activated sludge
process, in addition to removing most of dissolved
BOD, functions to capture, remove, and hence
recover most of these residual materials by
biocoagulation and flocculation. As a result, activated
22
-------
sludge is finer than primary sludge. It is normally
comprised of 60 to 90 percent or more cellular
organic material and contains a very large amount of
water.
Individual particles of activated sludge are usually
aggregated to an extent through bio-flocculation.
Table 3-5(5) shows the relative difficulty of removing
water from an unflocculated primary digested sludge
containing various particle size fractions. As can be
seen, the Specific Resistance to filtration of the
unfractionated sludge is dominated by the Specific
Resistance of material under 5 microns in size, even
though this material constitutes only about 14 percent
by weight of the total solids. Specific Resistance is, in
effect, a measure of the relative dewaterability of a
sludge. The lower the Specific Resistance, the
greater the sludge's dewaterability. Specific
Resistance has been defined as the pressure
required to produce a unit rate of flow though a cake
having a unit weight of dry solids per unit area when
the viscosity of the liquids is unity. Specific values are
determined from laboratory filtration experiments.
Table 3-6. Specific Resistance of Various Types of
Sludges
Table 3-5.
Sludge Dewatering as a Function of Particle
Size
Mean Diameter
Specific
Resistance
Percent of
Total Particles
microns sec2/g
Original, unfractionated 10.4 x 109
sample
> 100
5- 100
1 - 5
< 1
2.3 x 109
4.6 x 109
13.8 x 10»
-
10.2
75.5
8.5
5.9
Table 3-6(5) contains typical Specific Resistance
values for different types of sludges, both chemically
treated and untreated. Since the maximum Specific
Resistance for feasible mechanical dewatering is
normally quoted at <10.0 x 107 sec2/g, none of
these sludges would be readily dewaterable. Table
3-6 shows that Specific Resistance values can vary
significantly. Experience indicates that properly
conditioned raw primary sludge is almost always the
most readily dewatered, followed by well-conditioned
digested primary sludge and then activated sludges,
in increasing order of difficulty.
Sludge stabilization by aerobic and anaerobic
processes results in the destruction of a portion of
the organic matter and the production of hydrous
particles, which are more difficult to dewater.
However, a significant portion of the original hydrous
sludge is also destroyed in the stabilization process.
The consequence is that the residual digested sludge
is sometimes more difficult to dewater, sometimes
Type of Sludge
Raw
Raw (coagulated)
Digested
Digested (coagulau.
Activated
Specific Resistance
sec2/g
10 - 30 x 109
3 - 10 x 107
3 - 30 x 109
2 - 20 x 107
4 - 12 x 109
easier. But, in any case, the quantity is reduced 30-
40 percent from the raw state.
3,3.5 Compressibility
If sludge particles were idealized incompressible
solids, the solids would not deform, and the void
space between particles would remain constant
during mechanical dewatering. In such an ideal
situation, resistance to filtration would be proportional
to sludge depth, and there would be no increase in
resistance to filtration as dewatering progresses.
Unfortunately, sludge particles are practically always
hydrophilic and compressible to a degree, which
results in particle deformation and a reduction in the
void area between particles. This reduction in void
volume inhibits the movement of water through the
compressed portion of the sludge cake, and reduces
the rate of dewaterability.
Proper conditioning improves dewaterability primarily
by producing a flocculent matrix of solids in relatively
clear water prior to filtration. When this matrix is
deposited on a filtering medium, the bulk cake retains
a substantial porosity. However, too high a pressure
drop across the sludge floe will trigger the conditioned
sludge cake to collapse, and will result in a decreased
filtration rate. The net result of conditioning is quicker
removal of water, principally due to the higher rate of
water removal at the start of the filtration cycle.
3.3.6 Sludge Temperature
As sludge temperature increases, the viscosity of the
water present in the sludge mass decreases.
Viscosity is particularly important in centrifuge
dewatering since sedimentation is the key component
of the process (1). See Section 7.3 for further
discussion.
3.3.7 Ratio of Volatile Solids to Fixed Solids
Sludges tend to dewater better as the percentage of
fixed solids increases. One high-G centrifuge
manufacturer uses the percentage of fixed solids as a
key parameter in sizing equipment (R.T. Moll,
Sharpies-Stokes Div., Pennwalt Corp., personal
communication, 1982). (See Section 7.3 for a
description of low-G and high-G centrifuges.)
According to this manufacturer, the sludge cake from
centrifugal dewatering of an anaerobically digested
23
-------
mixture of primary and waste activated sludge shows
a positive change of 5 percent in its solids
concentration when the percentage of volatile solids
in it decreases from 70 percent to 50 percent.
However, since digestion also produces smaller
particles, the higher surface area results in more
moisture. The above approximation of volatile content
to cake solids must be cautiously employed and
should be pretested whenever possible.
3.3.8 Sludge pH
Sludge pH affects the surface charge on sludge
particles. Hence pH will influence the type of polymer
to be used for conditioning. Generally anionic
polymers are most useful when the sludge is lime
conditioned and has a high pH, while cationic
polymers are most suitable at a pH slightly above or
below neutral. In some cases, cationic polymers can
be effective up to pH 12 and has been employed for
lime stabilized sludge at New Haven, CT. Polymer
technology is continuing to advance.
3.3.9 Septicity
Septic sludge is more difficult to dewater and requires
higher dosages of chemical conditioners than fresh
sludge. This phenomenon has been experienced at
many locations and is most likely due to a reduction
in the size of sludge particles, to the generation of
gases that remain entrained in the sludge, and to the
change in surface characteristics created by bio-
conversion Wetter cake and lower sludge production
are common results from dewatering septic sludge.
For this reason raw sludge storage should be
minimized as an operating practice.
aeration if it exists. Pre-aeration will enhance the
removal of solids while freshening the wastewater.
i
3.5 References
1. Vesilind, P.A. Treatment and Disposal of
Wastewater Sludges, Revised Edition. Ann Arbor
Science Publishers, Ann Arbor, Ml, 1980.
2. Status of Oxygen/Activated Wastewater
Treatment. EPA-625/4-77-003, U.S.
Environmental Protection Agency, Center for
Environmental Research Information, Cincinnati,
OH, 1977. ;
3. Anderson, M.S. Comparative Analysis of Sewage
Sludges. Sewage and Industrial Waste 28{2),
1965.
4. Dick, R.I. Sludge Treatment. In: Physiochemical
Processes for Water Quality Control, edited by W.
J. Weber, Editor. Wiley Interscience, 1972,
5. Coackley, P. and F. Wilson. Proceedings of the
Filtration Society. Filtration Separation, Jan-
Feb:61, 1971.
3.4 Reclrculation from Solids Processing
The return flows emanating from sludge thickening,
digestion, conditioning, and dewatering will recycle
TSS and BODs. If the primary clarification is not
hydraulically overloaded, the majority of these solids
will resettle in the primary clarifier. Contrary to popular
opinion, there will be no significant increase in
primary clarifier effluent TSS due to recycle loads of
25 percent or more of the influent TSS
concentrations. The treatment plants at York, PA,
Dubuque, IA, and New Haven, CT all experience TSS
recycles up to and exceeding 100 percent of the
influent TSS without impairing primary effluent TSS.
However, if the primary clarifiers were hydraulically
overloaded and/or allowed to go septic, high recycle
levels of BODs an£| TSS would be a serious problem.
Thus, the bulk of TSS in the recycle stream will
resettle and can be contained. This is not true for the
soluble BODs fraction of the recycle. This recycle
loading must be added to the anticipated primary
effluent BOD5.
Return streams should always pass through the
primary clarifiers (if present) and through pre-
24
-------
Chapter 4
Process Selection
4.1 Introduction
Above all, the design engineer must ensure that
capacity limitations in the sludge processing system
are not the direct cause of impaired effluent quality.
That is, the design should provide for sufficient
standby capacity or an alternative mode of sludge
handling, whereby solids can be removed from the
wet-end processing in an orderly manner -- even
if the primary means of sludge disposal is unavailable
or has failed in some manner. This criterion applies
equally well to plants small and large, whether utilizing
mechanical or non-mechanical means of sludge
disposal.
Alternative methods (standby capacity) for non-
mechanical methods of sludge dewatering can
include:
• Multiple units - one standby unit is the best
alternative. For example: two operating sand beds
and one bed unit available as a spare.
» Liquid land spreading - land spread sites must be
approved in advance and liquid storage must be
available and consistent with state regulatory
requirements. Sixty to 120 days storage may be
required in areas where the ground becomes
frozen.
» Storage on-site -
requirements.
per state regulatory
• Liquid haulage to another plant - procedure
should be approved in advance by the alternate
facility through written agreement between the two
authorities.
Alternative methods for mechanical dewatering
equipment can include:
• Duplication of capacity or one standby unit -
recommended as best alternative.
• Maintenance of existing pre-expansion
dewatering equipment in addition to new
equipment - often this appears to be a feasible
approach, but in reality fails since disuse results in
disrepair and, eventually, the old equipment is
inoperable when needed.
» Liquid haulage - satisfactory for small plants,
provided the alternative mode has been guaranteed
by permits and prior agreements and that it is
suitable for year-round disposal.
• Storage • has limitations, except in smaller plants
(the storage capacity must exceed the time
required to repair equipment).
• Contract disposal services - not available in all
areas.
The evaluation of sludge dewatering alternatives must
consider the possibility of upstream and downstream
processes being out-of-service. Consider, for
example, a gravity thickener, which concentrates the
raw primary sludge (RPS) and waste activated sludge
(WAS) to S percent TSS, being out-of-service. If
there is only one thickening unit, what alternative
provisions are available to partially concentrate the
sludge? Can downstream process units handle the
more dilute sludge volume?
If there are two gravity thickeners, the off-line unit
could reduce the underflow concentration to about
3.5-3.8 percent TSS due to the doubling of the unit
loading (kg/m2/d). Or, if there is a single combustion
unit, is there an acceptable alternative method of raw
sludge disposal available? If not, provision for lime
stabilization should be included in the dewatering
plan.
There is seldom a defendable reason for having only
one dewatering unit in a wastewater treatment
system, except in very small systems which have
adequate sludge storage/disposal capacity by an
approved alternative method. In no case should the
lack of sludge dewatering equipment require storing
sludges in the wet processing operations; this
includes the sludge thickening units. Sludge
concentration processes prior to dewatering should
be duplicated unless it can be demonstrated that the
dewatering equipment can process the more dilute,
higher volume sludge in the operating time available.
The unconcentrated sludge may reduce the
25
-------
dewatering unit's operating capacity by 50-70
percent.
4.2 Sludge Processing Methods/
Selection Procedures
When either evaluating or selecting a dewatering
process, the designer must consider both prior
treatment processes and subsequent disposal
practices. A dewatering process cannot be evaluated
without also examining the other processes involved
in the overall solids handling system. Such an
evaluation can be a complex procedure because of
the vast number of combinations of unit processes
available for thickening, stabilization, conditioning,
dewatering, and ultimate disposal. Figure 4-1
presents a general schematic of a typical solids
handling system and the unit processes most
commonly used to perform each of these functions.
The strategy for selection of a dewatering process at
either new or existing plants requires up to five stages
of analysis, as shown in Figure 4-2. These stages
represent a screening procedure in which the
dewatering processes under consideration are given
increasing scrutiny as more detailed cost, operational,
and design data are collected and evaluated. The
components of each of these stages are discussed
below:
Stage 1 - Initial Screening of Dewatering
Processes
Rrst, a large number of factors should be reviewed so
that incompatible processes can be eliminated prior to
the initial cost analysis. Factors to be considered in
the initial screening include:
* Compatibility with plant size and existing facilities
• Type and quantities of sludge produced
Compatibility with the ultimate disposal technique
(Selected dewatering process must be able to
produce required cake solids concentration)
Compatibility with available labor and land
Degree of conditioning required
Environmental considerations
Field experience with processes at other similar
operating installations.
Stage 2 - Initial Cost Evaluation
Based on the best estimates of design and
operational criteria for the feasible dewatering
processes, an initial cost evaluation should be
conducted. In some cases, 10 to 20 complete solids
handling alternatives, which may include 4 to 5
different dewatering processes, are evaluated in this
initial stage. In general, no more than 3 to 5 of the
lowest cost alternatives are selected for more detailed
evaluation,
Stage 3 - Laboratory Testing
Laboratory testing may be conducted on the
dewatering processes selected in Stage 2 to refine
design criteria for the more favorable dewatering
techniques. This laboratory testing may be conducted
at the plant or by equipment manufacturers in their
laboratories. This testing will have limited value unless
conducted on a representative sludge that has not
undergone change in transport. Laboratory tests
should be conducted near the source of sludge in
order for the full range of fresh sludge characteristics
to be tested. A test on one sample of sludge has
negligible value and could be a cause of design
errors. Substantial day-to-day variation in mixed
sludge characteristics can occur. "Too good to be
true" results most often are just that. (See Chapter 5
for more information on laboratory tests.) Reid testing
is preferred except for smaller plants.
Figure 4-1. General schematic for solids handling, showing most cpmmonly used methods of treatment and disposal.
THICKENING
PRIMARY SLUDGE
•Source Thickening In
Primary Clarifier
•Gravity
THICKENING
SECONDARY SLUDGE
•Dissolved Air
Flotation
•Solid-Bowl Centrifuge
•Bolt or Drum
Thickener
STABILIZATION
•Anaerobic Digestion
•Aerobic Digestion
•Wet Air Oxidation
•Aerobic-Anaerobic
Digestion
•Chlorine Oxidation
•Lime Stabilization
•Ferric Chloride
•Lime
•Lime & Ferric
Chloride
•Polymer
•Heat Treatment
•Elutriation
•Freeze-Thaw
•Solid-Bowl Centrifuge
•Belt Rlter Press
•Vacuum Filter
•Filter Press
•Drying Beds
•Sludge Lagoons
•Gravity/ Low Pressure
Devices
•Compost
•Incineration
•Drying
i >
•Land Spread
•Landfill
•Land Injection
26
-------
Figure 4-2. Five stages of analysis in selection of a
dewatering process.
Stage Initial Screening of
1 Dewatering Concepts
i
Stage
,2
i
Stage
3
,
Stage
4
\
Stage j
5 c
Initial Cost
Evaluation
r
Laboratory
Testing
.
Field Level
Testing
r
Final Evaluation
3ased on Detailed
)esign Parameters
Stage 4 - Field Pilot Testing
If the plant is large and/or the cake moisture content
is critical (and more than one dewatering method may
be feasible), pilot studies are often warranted. Since
the sludge dewatering properties of even apparently
similar sludges may vary widely, pilot studies greatly
reduce the risk of improperly selecting and sizing
dewatering equipment. The cost of a thorough
evaluation is small compared to the benefits gained.
If it is necessary to test two types or more of
comparable dewatering equipment, the tests should
be conducted simultaneously to eliminate potential
differences related to the sludge composition. Sludge
variations, due to a number of reasons, can distort
the comparison. An example, shown in Table 4-1, is
the result of two series of tests conducted at the
same plant in Ohio.
Table 4-1 indicates that two dewatering studies at
the same facility produced widely different results.
Moreover, there was not a similar comparative
difference between the two types of equipment for
the winter and summer testing. The differences found
must also be considered in light of the degree of
optimization achieved. Short-term testing may not
have fully evaluated the range of operation or
optimized the critical chemical conditioning step.
The centrifuge data in Table 4-1, which was
produced by full-scale operation, indicate the
magnitude of the problem that could have been
encountered had the centrifuge installation been sized
on winter test performance results. Production rate
and cake solids content were much lower in the
summer tests. The differences were a result of storm
flows adding inert material to the sludge and changes
in industrial discharges. In the winter tests (Series I),
storm flows had added inerts to the sludge; in the
summer tests (Series II), a high TSS discharge from
a brewery had a more adverse impact on the
performance of the centrifuge than on the
performance of the diaphragm plate press.
Ideally, pilot testing should be carried out over an
extended period of time. However, extended testing is
often not practical. Test programs should evaluate a
sufficiently wide range of PS:WAS ratios to ensure
testing of worst-case situations, preferably during
colder weather when the sludge water viscosity and
secondary sludge yields are higher. Further, a full
range in operating capacity should be investigated to
determine the effect on cake solids and capital and
operating costs.
Stage 5 - Final Evaluation Based on Detailed
Design Parameters
After Stage 4 is completed, accurate scale-up and
sizing of equipment is performed by the design
engineer with the aid of the equipment manufacturer.
At this time, estimates of the capital cost, labor,
energy, chemical, and maintenance material
requirements for the dewatering process under
consideration can be refined. This information can be
supplemented with data from other plants using the
same process. The researching of similar equipment
performance and the manufacturers' service record is
highly recommended. Additionally, the operating utility
can make input from performance and operational
problems experienced in Stage 4 field evaluations.
Based on accurate capital and operation and
maintenance cost information, a final cost evaluation
can be made in conjunction with an evaluation of
other parameters. Stage 5 concludes with selection of
the dewatering process and, in many cases, the
preferred manufacturer. All generic equipment is not
created equally.
The equipment and supplier selected should have
widely demonstrated the capability to meet the design
requirements in either similar plants or by adequately
supervised pilot studies at the subject facility. When a
new design of equipment is employed, the
manufacturers' prior practices need to be carefully
scrutinized and adequate safeguards provided to the
utility. Evaluations properly conducted will not stifle
new developments.
Throughout this five-stage process, many trade-
offs will have to be made. In many cases, the total
annual cost of two or more solids treatment systems
are essentially identical (±10%), and the decision
must be made on some basis other than cost.
Frequently, such a decision is based upon capital vs.
27
-------
Tabta 4*1. Comparative Dewatering Results for Two Test Periods
Sorios 1 - Winter
C-1
C-2
C-3
DPP-1>
DPP-2'
Sories II • Summer
C-1
C-2
DPP-11
DPP-21
PS.-WAS Ratio
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
Feed
l/s
3,8
5.0
6.3
65.02
80.02
2.5
3.8
-
Rate
kg TS/rtr
314
382
477
20.93
14.2
282
423
15.13
19,5
CakeTS
percent
24.3
23.6
17.3
21.9
27.1
17.6
13.3
24.3
27.2
SS Recovery
percent
92
86
85
99 +
99 +
97
96
99 +
99 -f-
Polymer Cost
$/Mg
'31.03
23.27
19.64
11.64
13.18
17,09
29.17
5.81
8.64
Tola) Chemical
Cost
$/Mg
31.03
23.27
19.64
28.31
29.85
17.09
29.17
22.48
25.31
C « 0,74 m diameter x 2.34 m long solid bowl centrifuge.
DPP * Lab diaphragm plate press.
Polymer addition w/precoat @ 12% of sludge solids.
FoCI3 © $0.25/kg
Pracoat @ $0.66/kg
1 Used 3*4% FeCIs to improve flocculafon.
2 press time, min.
3 Cako discharge, kg TS/m2 of plate area/hr.
O&M cost considerations, ease of equipment
operation, energy requirements, performance, or
other factors such as prior plant experience with
similar equipment. A point to keep in mind is that the
decision is often not clear-cut.
The overall complexity of analysis will vary depending
on the size of the plant and whether a new solids
handling system is being designed or an old one
upgraded. If the solids handling system is completely
new, there will probably be fewer constraints on the
processes to be evaluated, conditioning method to be
used, and ultimate disposal techniques to be
considered. In other situations, if the entire treatment
facility is new, or if it is being upgraded from primary
to secondary, sludge of the correct composition will
not be available to conduct field tests. Stage 4 is
generally not conducted for most of the small
capacity plants, those less than 0.13 m3/s (3 mgd).
For the small plant, it is usually more economical to
design facilities based on laboratory or bench-scale
testing {often performed by the manufacturer of the
equipment) and after experiences at other plants, and
using generous factors of safety in design, than it is
to conduct the field-scale testing. The field-scale
testing may result in a recommendation for smaller
and thus less expensive equipment, but the reduction
in cost is unlikely to offset the extra time and cost of
the full-scale testing.
While past experiences with similar equipment and
facilities can provide useful input, engineers must
ensure that they are current in the process and
equipment technology under consideration and that
design requirements are similar. A design practice of
using the same process technology for all plants is
highly questionable. It is the design engineer's
responsibility to make independent evaluations of the
equipment's performance and not to accept other
evaluations by parties with a financial interest in the
outcome without first checking the reduced data.
All designs for sludge processing and dewatering
systems should undergo a thorough "what if"
evaluation. That is, evaluation of all probable
situations that could occur in the future plant
operation. The limiting conditions should be defined
and analyzed for any adverse impact on liquid
processing capabilities. Sludge storage in the liquid
treatment process is not an acceptable alternative,
since effluent quality degradation will occur soon
thereafter. Further, the hydraulic peaking capacity of
the clarification operations will be reduced by sludge
storage. Sludge processing equipment should be
selected and sized to process the sludge produced in
a timely manner.
A formal ranking of the alternative dewatering
methods versus key selection criteria is
recommended. The key criteria should have
appropriately weighted values. The formal ranking
results should be internally and sometimes
independently critiqued to ensure all considerations
have been adequately evaluated.
4.3 Operational Selection Criteria
The criteria employed for selecting a sludge
dewatering process are complex and will vary from
site to site and with the size of the plant. Given
28
-------
similar circumstances, engineers will often select
different dewatering processes based on their past
experiences and personal preferences. Often there is
more than one correct process selection, or the
correct selection could only be established by
exhaustive testing and engineering economic
analysis.
The selection criteria for smaller plants are
considerably different than those for larger facilities. In
many cases, other considerations, such as transport
and land availability, may have a more significant
impact than economics. Table 4-2 presents various
selection criteria versus plant size. However, site-
specific conditions may change the relative
importance of some of the criteria. For example, a
drying bed in a highly developed area would be more
objectionable than it would be in a semirural setting or
in a heavily industrialized area.
Table 4-2. Operational Selection
Dewatering Processes
Criteria for Sludge
Key Criteria
Small Minimum Mechanical Complexity
< 0.08 m3/s Local Repairs and Parts
(<2 mgd) Minimum Operator Attendance
Reliable Without Skilled Service
Unaffected by Climatic Factors
Large Excess Capacity
Handleable Cake
Medium Low Operator Attendance
0.08 - 0.44 m% Local Repair and Parts
(2-10 mgd) Transportable Cake Without Nuisance
Mechanical Reliability
Competitive O&M Costs
Drier Cake
Large Lowest O&M Costs/ton Dry Solids
>0.44 m3/s Lowest Capital Costs/ton Dry Solids
{> 10 mgd) Driest Cake
High Output/Unit
Mechanical Reliability
Transportable Cake Without Nuisance
General Considerations
Compatibility with existing equipment with long-term sludge
disposal
Long-term serviceability/utility
Acceptable environmental factors
Good experience at other operating installations
Competence and quality of local operator and service personnel
Compatibility with plant size
Acceptance by user and regulatory agency
Availability and need of manufacturer's services
The classification of plants as small [<2 mgd (0.08
m3/s)], medium [2-10 mgd (0.08-0.44 m3/s)| and
large (> 10 mgd (0.44 m&Vs)] is rather arbitrary and
is used merely as a generalization to segment
applicable technology. The personnel at a 0.11-m3/s
(2.5-mgd) advanced wastewater treatment plant
treating wastewaters to 5 mg/l BODg, 5 mg/l TSS, 1
mg/l NH4N, and 1 mg/l total phosphorus may be
much more qualified to operate and maintain a
mechanical dewatering unit than the personnel at a
0.53-m3/s (12-mgd) plant using trickling filters to
meet 30/30 criteria. The anticipated quality and
quantity of the O&M staffing are important criteria for
the engineer to consider in the selection of process
technology.
In smaller plants, the lowest initial cost may not be
the best selection if it requires continuous operator
attendance, is mechanically complex, and cannot be
repaired locally. The impact of these considerations
can easily offset the advantage of an alternative lower
cost dewatering process. A mechanical dewatering
process that is operating at 200 kg dry solids/hr, but
requires continuous operator attention, could result in
over $50/ton ($55/Mg) operating labor costs plus the
maintenance costs.
4.3.1 Compatibility With Existing Facilities
Existing facilities, which must be considered
evaluating dewatering processes, include:
in
• Type of dewatering equipment presently used, its
useful remaining life, and its compatibility with
future requirements
» Existing conditioning, chemical storage and feed
facilities
» Existing building used for dewatering and ancillary
equipment
• Existing site constraints
• Existing sludge transport facilities.
4.3.1.1 Existing Dewatering Equipment
Existing dewatering equipment customarily plays a
major role in the selection of additional equipment,
particularly if space has been provided for expansion
of the present dewatering facilities. If existing
equipment is providing satisfactory performance (from
both a cost and operational standpoint), and if the
product cake is suitable for the ultimate disposal
technique, in all likelihood the same dewatering
process would be appropriate for the expansion plan.
This would be particularly true if the dewatering
facilities had been designed to accommodate more
equipment of the same type. In perhaps the majority
of dewatering operations, existing equipment is
performing unsatisfactorily and requires more
chemicals or energy than originally anticipated. In
other cases, the sludge characteristics have
adversely changed, and the existing equipment
cannot be operated at the original design capacity. In
some cases, existing equipment cannot perform as
well or as efficiently as some of the newer but similar
equipment available, or the cake produced by existing
29
-------
equipment is not suitable for the future ultimate
disposal technique.
In a large percentage of the expansions of dewatering
facilities, the plant staff is dissatisfied with the
operation of the existing equipment Typical situations
are: {1} vacuum filter installations where lime coating
of the filter media, filter drum, and filtrate piping
presents an expensive and continuing maintenance
problem; (2) filter press installations that often have
chemical requirements substantially higher than
originally expected; (3) older existing solid-bowl
centrifuge installations where a great deal of scroll
maintenance is required due to abrasive wear and/or
where the operating performance is poor; and (4)
drying beds or lagoons where odors, negative visual
impact, intensive labor requirements, or difficulty with
sludge removal make the process an operations
problem for the plant staff. These types of problems
can cause headaches for the operation and
maintenance staff and can decrease effective
dewatering capacity. Also, operating costs increase
when equipment must be taken out of service for
repairs and/or cleaning.
Variation in sludge characteristics after design and
installation of equipment, and therefore variation in
the ability of the sludge to be dewatered, presents a
particularly vexing problem. Very often, variation of
sludge characteristics leads to higher conditioning
and energy requirements than originally projected,
and, in some cases, the inability to produce a
dewatered cake suitable for ultimate disposal.
Sometimes equipment must be operated at less than
design capacity due to changed sludge
characteristics. An evaluation should be conducted to
determine the likelihood and severity of changes in
sludge feed rate and characteristics. If significant
variations are anticipated, equipment, such as the
centrifuge, that is less sensitive to such changes
should be selected.
More often than not, equipment technology will have
advanced since the original equipment was installed.
In this case and particularly where there is a high
degree of owner/operator dissatisfaction, the obsolete
equipment should be removed and replaced by
modern equipment. Where selection evaluation
indicates that there is an acceptable alternative to the
original, unsatisfactory process technology, it is
generally advisable to employ a different process for
dewatering to ensure plant cooperation. Where
possible, extensive test trials to demonstrate the
advanced technology to the plant operators are
recommended.
Obsolete equipment does not provide good standby
capabilitvl There is a natural resistance to upgrading
or maintaining old, obsolete equipment. If there is
significant operator resistance to use of obsolete
equipment, it will not be in a serviceable condition
when needed. When faced with obsolete equipment,
the engineer and the city should "bite the bullet" and
provide a totally new dewatering station.
4.3.1.2 Existing Building Used for Dewatering
Equipment
If the original design allocated space for an
anticipated expansion, there are generally few
problems associated with installing newer equipment
of similar size. Newer equipment, built to handle a
given sludge volume, is usually lighter than similar
older equipment. However, if larger equipment is
being installed, then the suitability of the housing
facility will be a primary concern.
The engineer must consider the present building's
structural capacity for replacement equipment, and
whether the building has sufficient headroom and
working space for the equipment being considered.
Dewatering equipment, such as solid bowl
centrifuges, belt filter presses, and filter presses
frequently discharge dewatered solids downward.
These machines can be incompatible with buildings
with low roofs, because in some cases they require
elevated mounting to provide space for conveyor
belts under the equipment.
i
Further, heavy equipment, such as a filter press, may
not be compatible with a building originally designed
for a centrifuge installation, even though both have
bottom discharge of cake solids. Centrifuges may
require greater structural support than belt presses.
When there is an existing overhead crane, the new
equipment may exceed the allowable crane capacity,
and it would also need to be replaced. Lack of
operator working space is a means of ensuring poor
maintenance and attendance. Crammed facilities
should be avoided at the cost of new or expanded
building space.
4.3.1.3 Existing Site and Environmental
Constraints
Drying beds and sludge lagoons require considerable
land area. Expanded use of these processes may not
be practical if land is unavailable, or if environmental
constraints make continued ;use unacceptable. In
some cases, existing beds or lagoons can be used in
conjunction with a different dewatering process.
For example, drying beds could be used to produce a
dry product in the warm, drier periods of the year and
the mechanical equipment would be used when the
weather is adverse to dewatering sludge on the
drying beds. The drying beds would constitute the
backup in the winter, and vice versa.
Design engineers need to ask the city and
themselves several questions before deciding on a
dual-technology alternative. Typical questions could
be:
30
-------
1. Which method requires the least labor? Is this
important?
2. Is there adequate staff at the plant now? Will there
be?
3. Are the drying beds well maintained now? If not, is
there any reason maintenance would improve if
there were an alternate mechanical unit available?
4, Is the absence of chemicals a sufficient driving
force to result in the use of beds in warmer
weather? Does the additional labor offset the cost
of chemicals and convenience of a mechanical
dewatering station?
5. Is there an advocate or user who wants the drier
sand bed solids as opposed to a wetter cake?
Unless such questions are posed and truthfully
answered, most dual-technology dewatering
technology systems are not fully used and one
system falls into disuse, if the design engineer cannot
justify the money to refurbish the older facilities, they
should be replaced.
The local environmental setting may force the closing
down of a good operating system such as a sand
bed. An alternative may be to enclose the sand bed
and ensure that it is properly ventilated and that odor
control is provided. Enclosure also brings additional
capacity to the same system since weather impacts
are minimized and the sand beds could be heated.
Further, increase in capacity can be achieved with
small dosages of polymer.
4.3.2 Process Compatibility With Size of Plant
Use of uncomplicated sludge handling systems
increases the chances for successful operation in any
size of plant. Complex equipment is especially
unsulted to small plants for several reasons. First, the
amount of operator time available generally decreases
as plant size decreases. Second, small plants may
not have operations and maintenance personnel with
the required skills. Third, less complicated equipment
is generally less expensive to purchase. Since
standby capacity of about 100 percent is often
provided for small plants (usually with duplicate units
rather than employing a single large unit), it is more
economical to choose the less complicated, less
expensive system.
The foregoing discussion presents a basic approach
to selection of the dewatering process. Generalized
guidance based on results at plants across the United
States is summarized in Table 4-3, which presents
compatibility of different dewatering techniques with
various plant sizes. Designers should use the
information presented in Table 4-3 only as a guide.
Every plant must be considered independently, since
site-specific considerations can have a large
Table 4-3,
Compatibility of Dewatering Process with Plant
Size
Belt and Drum Thickeners
Solid Bowl Centrifuge
Belt Filter Press
Vacuum Filter
Filler Press
Drying Beds
Sludge Lagoons
<0,04
XI
X2
X
X
Plant Size, m^/s
0.04-0.44
X
X
X
X
X
X
X
>0.44
X
X
X
X
X
1 Suitable tor land spreading or injection of sludge.
2 Only low pressure press is commonly used in this flow range.
influence on the dewatering process. For example,
drying beds and sludge lagoons may be cost-
effective at a plant larger than 0.44 m3/s (10 mgd), if
weather is favorable and land available. Using a
structured approach to evaluate the alternatives, in
conjunction with the costs of the ultimate solids
disposal method, will sort out the appropriate
technology.
4.3.3 Process Compatibility with the Ultimate
Disposal Technology
The ultimate disposal method of the residual solids
often dictates the process selection. The most
common methods of disposing of the cake are shown
in Table 4-4. The end-product of each dewatering
method is rated in terms of the product acceptability
vis-a-vis the ultimate disposal of the residuals.
However, this table is not meant to establish the
compatibility of the dewatering process and the end-
disposal means. For example, combustion of a sand
bed product of 30-40 percent TS is very
economical, but the use of sand beds to dewater
sludge solids for combustion may be impractical. On
the other hand, a compost operation may be better
equipped to process the periodic and somewhat
variable sand bed product which can be stockpiled
until needed.
While all sludge cakes meeting EPA's stability
regulations (see 40CFR257) can be land spread, drier
solids will be easier and less costly to transport and
apply to the land. The drier the cake, the less the
probability of any nuisances such as odors, insects,
or liquid runoff. Ideally, sludge applied to land should
either be sufficiently dry so that the spreader can
break it into small pieces or should be a liquid that
can be evenly spread. Injection of liquid sludge can
eliminate nuisances, but it also can be limited by
weather conditions. If (he land is to be "worked" after
the sludge application, then the cake solids
concentration is less of a concern, except where the
sludge cake must be stored for part of the year. In
this case, a drier cake is preferred.
31
-------
Table 4-4. Suitability of Cake Produced by Dewatering Processes for Various Ultimate Disposal Options
Land Spread Combustion/
Dewatenng Process
Cake Solids % Land Spread
Inject
Landfill
Drying
Koy. 1. Good
2. Satisfactory
3, Inappropriate
Nola: Tho ratings apply to the product concentration and not the methodology employed to achieve the cake solids.
Compost
Lagoon
Sand Bods
Vacuum Assisted Bods
Screw Pross
Pavod Bods
Bolt Pross
Conlrifugo
Vacuum Filler
Filter Press
Drum/Boll Thickeners
15-40
30 -60
10-
16
12-
20
30-60
18 -
24
18 -
24
5-7
16-20
26-
34
5-8
1
1
1
2
1
2
1
2
2
2
2
1 1
2
1
1
1 1
2
1
3
21
3
2
1
2
2
2
2
3
2
1
1
3
3
, 1
! 3
; 3 .
; 3
! 3
i 1
; 3
'- 2
'• 3
2
, 3
3
2
1
, 3
2
1
3
2
3
2
1
2
2
2
2
3
2
2
1
3
Landfill operations are becoming more selective in the
materials accepted. Wet cakes, which may result in
liquid discharges, are particularly discouraged. A
minimum cake of 18 percent TS is considered
acceptable for a narrow trench landfill. However,
higher percent TS may be required in some areas of
the United States, and in the future higher percent TS
may be required in many areas. West Germany
currently requires 35-40 percent TS for landfill.
Dewatering of sludge for combustion or drying
requires that the solids content be equal to 24
percent TS, preferably 28-30 percent for an
economical operation. On a typical raw primary
sludge (60 percent) and WAS (40 percent) mixture,
only drying beds and filter presses can reliably deliver
24 percent TS or drier cakes. While many multiple
hearth furnaces (MHF) in the United States can burn
24-25 percent TS sludge cakes autogenously, the
stack outlet gases are below 480 °C (900 °F). Some
new state regulations now require a 650-760 °C
(1,200-1,400°F) stack outlet temperature to ensure
deodorization. Fluid bed reactors (FBR) require about
28 percent TS with heat exchangers to preheat the
air and 34 percent TS without heat exchangers to
burn raw primary and waste activated sludge
autogenously. Combustion gases from FBR units
always exceed 800°C (1,470°F) due to process
configuration.
The combustion capacity of multiple hearth furnaces
and fluid bed reactors as a function of sludge cake
solids concentration is shown in Figure 4-3. Typical
excess air for MHF is 75-125 percent and 30-40
percent for FBR, The moisture content also has a
direct impact on the size of the furnace. A 20 percent
TS cake (4 kg H2O:kg TS) will require a 56 percent
larger furnace than a 28 percent TS cake (2.57 kg
H2O:kg TS) for the same dry solids capacity.
Typical heat content of a 60:40 to 50:50 mixture of
PS:WAS is 5,555±278 kcal/kg VSS (10,000
Btu/lb±500 Btu/lb) and the VSS/TSS ratio is usually
72 ±5 percent for raw sludges. A more precise
means of expressing the effect of cake solids content
on an exhaust temperature is to use the wet cake
heat content, kcal/kg (Btu/lb) wet cake, and the
required excess air and other losses. The relationship
of kcal/kg wet cake, percent TS, and exhaust
temperature for autogenous conditions are shown in
Figure 4-4.
An outlet temperature of 540°C (1,000°F) for a FBR
is the equivalent of preheating air to about 540 °C
(1,000°F) with a 815°C (1,500°F) stack outlet
temperature before the heat exchanger. The
autogenous conditions noted in Figure 4-6 are
based on 50 percent excess air. FBRs will employ
30-50 percent excess air.
The effect of excess air on heat losses in a
combustion unit is illustrated in Figure 4-5. For
example, if a furnace is operating at 540°C (1,000°F)
outlet gases, the heat loss at 40 percent excess air is
about 6 percent of the input. At 125 percent excess
air, the loss is 300 percent higher or 18 percent of
the input.
The effect of operating exhaust temperature and
excess air on fuel consumption is quite apparent in
32
-------
Figure 4-3. Multiple hearth furnace and fluid bed reactor
capacities.
36 r
Figure 4-4. Conditions for zero fuel.3
32
'Is
26 percent TS to reduce
fuel requirements to zero. Curves B and C represent
fuel requirements for MHF @ 540°C (1,000 °F) and
680°C (1,250°F) while Curve A is for a FBR @
815°C (1,500°F) using a preheater. Table 4-5
presents the dewatering and combustion costs for an
array of different dewatering systems, which
illustrates the impact of capital and operating costs on
the overall cost of thermal sludge disposal.
Compost operations can sometimes suffer even more
than combustion from wet cakes. This is particularly
true in colder, wetter regions and where the compost
piles are not protected from the elements. Since it is
important to maintain the compost in a specific
moisture and temperature range, it is necessary to
44
40
JS 36
(a
1 32
o
CO
a 28
•££
to
°24
20
Temp Exit
Gases =
1,000"F
(538 °C)
Temp Exit
Gases =
500 CF
(816°C)
2,000 2,500 3,000
Heat Content, Btu / Ib cake
3,500
"Assumed conditions:
1. Excess air = 50%
2. Misc. tosses = 5%
Figure 4-5. Heat loss due to excess air.
150% 125% 100% 80%
40
36
32
o
5
I
28
24
20
16
12
8
400 800 1,200 1,600 2,000 2,400 2,800 3,200
Stack Gas Temperature, °F
recycle the drier product to bulk the sludge cake and
to reduce the average moisture content. The wetter
the cake, the higher the recycle or new bulking agent,
the more the pile is cooled and the less economical
the composting process.
An example of the effect of sludge cake moisture
content on the recycle rate is demonstrated by
Equation 4-1:
Minimum sludge + recycle TSS = 45% TS (4-1)
Sludge cake concentration = 18% TSS & 30% TS
Sludge cake quantity = 10 Mg/d
33
-------
Figure 4-6. Fuel consumption vs. stack temperature and
excess air.
140
120
100
1
•o
g 80
•C
03
CD
8 60
13
£
40
20
Curve
T °F Excess Air, %
1,000 40
1,000 100
1,250 100
20 22 24 26
Total Solids, %
28
30
4.4 Sizing of the Dewatering Process
Where cake solids content is the governing criterion,
operating at a rate below the rated capacity of the
dewatering unit will increase the solids content. In
some cases, a significant increase in solids content
will result from only a 20-25 percent reduction in
capacity. The rated capacity of manufacturers' units
will generally be substantially higher than that which
should be used for producing the driest cake.
When cake solids are not critical, the polymer cost to
maintain 90-95 percent TSS recovery will be the
governing criterion. The ability to readily move from
maximum dry solids content to maximum solids
recovery rate will vary from machine to machine.
While the centrifuge is flexible in this regard, vacuum
filter and belt and filter press operations must lie in
the range of cake dischargeability. Higher polymer
dosages can increase belt press capacity although
this may reduce cake solids. Rlter press operation
may be quite inflexible in this regard due to cake
dischargeability criteria.
Recycle product concentration = 55% TS
Dry Solids Balance:
(Mass of
Mixture/% TS) = (Cake/% TS) + (Recycle/% TS)
@ 10% TS:
(X/0.45) = (10/0.18) + [(x - 10)/0.55]
X = 93 Mg/d dry recycle solids
or 169 Mg/d product recycle @ 55% TS
@ 30% TS:
X = 37.6 Mg/d dry product solids
or 68 Mg/d product recycle @ 55% TS
At some point, further dewatering may have very little
benefit since the bulking requirements for air flow
may govern. The recycle rates are also highly
dependant on climatic conditions and the design of
the compost process. Areas of high temperature and
high evaporation rates will handle a wetter cake more
economically because they can produce a drier
product recycle. Enclosed compost systems in
colder, wetter areas do not have prohibitive recycle
rates because the recycle product moisture
approaches the minimum feed moisture (sludge •*•
recycle).
34
-------
Table 4-5. Dewatering-Cotnbustion Costs - Excluding Operating Labor (7,500 dry tons/yr @ 60:40 PS:WAS)
Average Dewatering Costs/Ton Sewage Solids - 1986
Cost, $/ton
Method
Centrifuge/Belt Press
Centrifuge/Belt Press
Centrifuge/Belt Press
Diaphragm Plate Press
Recessed Plate Filter Press
Diaphragm Plate Press
Chemical
P
P
P+C
P
F + L
P + A
Cake % TS
18
22
30
30
32
45
Maintenance Power Chemical
3,00 5.00
3.00 4.00
4.00 5.00
4.50 2.50
4.50 2.75
5.00 3.00
16.00
16.00
27.00
20.00
30.00
20.00
Total
24.00
23.00
36.00
27.00
37.25
28.00
Average Combustion Costs for FBR - 1986
(1,000°F - 50% Excess Air)
Method
Centrifuge/Beit Press
Centrifuge/Belt Press
Centrifuge/Belt Press
Diaphragm Plate Press
Recessed Plate Filter Press
Diaphragm Plate Press
Chemical
P
P
P + C
P
F + L
P + A
Cake % TS
18
22
30
30
32
45
Cost, I/ton
Maintenance Power
7.50 5.00
7.00 3.90
6.50 3.10
6.50 2.70
7.00 3.30
6.50 2.70
Fuel
74.80
35.00
0.00
0.00
6.30
0.00
Total
87.30
45.90
8.60
9.20
16.60
9.20
Combined Dewatering - Combustion Cost Summary
Cosl, $/ton
Dewatering Combustion
Method
Centrifuge/Belt Press
Centrifuge/Belt Press
Centrifuge/Belt Press
Diaphragm Plate Press
Recessed Plate Filter Press
Diaphragm Plate Press
Chemical
P
P
P + C
P
F + L
P + A
Cake % TS
18
22
30
30
32
45
Capital Operating Capital
6.00 24.00 33.60
6.00 23.00 28.90
7.80 36.00 25.20
9.00 27.00 24.60
11.10 37.25 26.20
11.10 28.00 23.30
Operating
87.30
45.90
8.60
9.20
16.60
9.20
Total
150.90
103.80
77.60
69.80
89.45
71.60
Operating Cost Basis:
Polymer (P) @ $2.00/lb
Ferric Chloride (F) @ $0.i5/lb
Lime (L) @ $0.05/lb
Coal (C) @ $60.00/ton
Fuel @ $7.50/106 Btu
Sludge @ 9,500 Btu/lb VS
Coal @ 12,000 Btu/lb
Ash (A) - no cost
Capital Cost Basis:
Installed equipment, no buildings or foundations.
$/ton x 0.9072 = $/Mg
°C = (5/9) (°F-32)
35
-------
-------
Chapters
Conditioning
5.1 Introduction
Conditioning prior to dewatering involves the chemical
and/or physical treatment of sludge to enhance water
removal and improve solids capture. The three most
common conditioning systems use inorganic
chemicals, organic polymers, or heat. Table 5-1
shows and compares the effects of conditioning
processes on a mixture of primary and waste
activated sludges.
Conditioning always has an effect on the efficiency of
the dewatering process that follows (1). Any
evaluation of the conditioning process must therefore,
take into consideration capital and operating and
maintenance costs for the entire system. These costs
include the impact of sidestreams on other plant
processes, the plant effluent and resultant air quality.
Some treatment plants are required to remove
phosphorus, although less of these plants must do so
today than a few years ago. This is both because of
state bans on the use of detergents containing
phosphorus and the generally decreased use of
phosphorus in household products. Phosphorus is
often removed by the addition of chemicals, including
ferric chloride, aluminum sulfate (alum), sodium
aluminate, and lime with some kind of polyelectrolyte
to facilitate coagulation and settling. Because of the
enormous materials handling difficulties, the interest
in using lime has declined.
Precipitating phosphorus can as much as double the
amount of sludge requiring treatment and disposal.
The amount of chemicals that must be added is a
function of the amount of phosphorus needing
removal (see Chapter 3). Fortunately, with the
quantity of phosphorus in wastewater diminishing, this
quantity is decreasing.
If precipitating phosphorus removal is contemplated,
laboratory/pilot tests should be performed to
determine the mass and volumes of sludge to be
expected, the degree to which the sludge can be
thickened, and how well the sludge can be
conditioned and dewatered if appropriate. While in
every instance there will be more sludge to contend
with (an additional 30 to 100 percent), the sludges will
thicken and dewater differently. Lime sludges readily
thicken and dewater, while hydroxide sludges (ferric
chloride, alum, sodium aluminate) usually thicken and
dewater poorly - requiring considerable conditioning
with polymers. In fact the best analogy is that
difficult-to-handle hydroxide sludges behave like a
poor quality activated sludge. It is not unusual also to
find a hydroxide sludge that will at best dewater to a
Table 5-1. Effects of Conditioning with inorganic Chemicals, Organic Polymers, or Heat on a Mixture of Primary and Waste
Activated Sludge
Inorganic Chemicals
Organic Polymers
Heat
Conditioning mechanism
Effect on allowable solids
loading rates
Effect on supernatant
stream
Effect on manpower
Effect on sludge mass
Coagulation and flooculation
Will increase
Will improve suspended solids
capture
Little effect
Significantly increases
Coagulation and llocculation
Will increase
Will improve suspended solids
capture
Little effect
None
Alters surface properties and
ruptures biomass cells, releases
chemicals, hydrolysis
Will significantly increase
Will cause significant increases in
color, suspended solids, soluble
BOD, NH3-N, and COD
Requires skilled operators and a
strong preventive maintenance
program
Reduces present mass but may
increase mass through recycle
37
-------
fina! cake solids concentration of 10 to 15 percent on
a centrifuge, belt press, and vacuum filter. These
"still wet" sludges may be hard if not impossible to lift
and unacceptable for landfilling.
5.2 Factors Affecting Conditioning
Wastewater solids are comprised of screenings, grit,
scum, and sludges. Wastewater sludges consist of
primary, secondary, and/or chemical solids with
various organic and inorganic particles of mixed sizes.
The sludges each have various internal water
contents, degrees of hydration, and surface
chemistry. Sludge characteristics that affect
dewatering (and for which conditioning is employed)
are parlicle size and distribution, surface charge, and
particle interaction. Furthermore, such things as
biopolymer production, degree of filamentous growth,
primaryrsecondary sludge ratio, and inorganic content
Particle size is considered to be the single most
important factor influencing sludge dewaterability (1).
As the average particle size decreases, primarily from
mixing or shear, the surface/volume ratio increases
exponentially (1). Increased surface area means a
greater hydration, higher chemical demand, and
increased resistance to dewatering. Figure 5-1
shows relative particle sizes of common materials.
Raw municipal wastewater contains a significant
quantity of colloids and fines that because of their
size, 1 to 10 microns, will almost all escape capture
in primary clarifiers if coagulation and flocculation are
not employed. Secondary biological processes, in
addition to removing BOD, also partially remove these
colloids and fines from wastewater. As a result,
biological sludges, especially waste activated sludges,
are difficult to dewater and have a high demand for
conditioning chemicals.
A primary objective of conditioning is to increase
particle size by combining the small particles into
larger aggregates. Since sludge particles are typically
negatively charged and repel rather than attract one
another, conditioning is used to neutralize the effects
of this electrostatic repulsion so that the particles can
collide and increase in size.
Conditioning is a two-step process consisting of
coagulation and flocculation. Coagulation involves
destabilization of the sludge particle by decreasing
the magnitude of the repulsive electrostatic
interactions between particles. This process occurs
through compression of the electrical double layer
surrounding each particle. Flocculation follows
coagulation and is the agglomeration of colloidal and
finely divided suspended matter by gentle mixing.
If the flocculated sludge is subjected to stress, floe
shearing can occur. Therefore, mixing should provide
just enough energy to disperse the conditioner
throughout the sludge and bring the particles and
colloidal suspensions together. Consideration should
be given to providing individual conditioning for each
dewatering unit, since it is neither always economical
nor good practice to provide one common
conditioning unit for several dewatering units.
Problems can arise in balancing the flow rates of the
various streams when starting up or shutting down
individual units. The location of the conditioning unit,
relative to each dewatering device, requires
optimization.
The amount of conditioning required for sludges
depends on the processing conditions to which the
sludge has been subjected and on the mechanics of
the conditioning process available. Both the degree of
hydration and fines content of a sludge stream can be
materially increased by exposure to shear, heat, or
storage. For example, pipeline transport of sludge to
central processing facilities, weekend storage of
sludge prior to mechanical dewatering, and storage of
sludges for long periods of time have been shown to
increase the demand for conditioning chemicals prior
to all types of dewatering. These factors should be
considered in the design of the complete dewatering
facility (1).
5.3 Inorganic Chemical Conditioning
Inorganic chemical conditioning is associated
principally with vacuum and pressure filtration
dewatering processes. The chemicals normally used
in conditioning municipal wastewater sludges are lime
and ferric chloride. Less commonly, ferrous sulfate,
ferrous chloride, and aluminum sulfate have been
used.
5.3.1 Ferric Chloride
Ferric chloride is added to sludge in conjuction with
lime and is added first. It hydrolyzes in water, forming
positively charged soluble iron complexes that
neutralize the negatively charged sludge solids, thus
causing them to aggregate. Ferric chloride also reacts
with the bicarbonate alkalinity in the sludge to form
hydroxides that act as flocculants. The following
equation shows the reaction of ferric chloride with
bicarbonate alkalinity:
2FeCl3 -i- 3Ca(HC03)2 -> 2Fe(OH)3 + 3CaCI2
(5-1)
Ferric choride solutions are generally used at the
concentration received from the supplier (30 to 40
percent) because dilution can lead to hydrolysis
reactions and the precipitation of ferric hydroxide.
An important consideration in the use of ferric
chloride is its corrosive nature. Special materials must
be used in its handling, with recommended materials
being epoxy, rubber, ceramic, PVC, and vinyl.
Contact with skin and eyes must be avoided. Rubber
38
-------
Figure 5-1. Particle size distribution of common materials.
Limit of
Visibility 100
MICRONS 0.001 in. / Mesh
01 0.01 0.1 1.0 10 / /102 / 103 (1 mm) 10" (1 cm)
10
Angstrom
Units
/
'
¥ r
I I
i i
Ii
i
1 1
V
I
I
I
Colloids
Fine
Medium
Coarse
Large
Clay
Silt
Fine
Sand
Coarse
Sand
Gravel
gloves, face shields, goggles, and rubber aprons
should be used at all times.
Ferric chloride can be stored for long periods of time
without deterioration. Usually it is stored in above-
ground tanks constructed of resistant plastic or in
lined steel tanks. At low temperatures, ferric chloride
can crystallize, which generally means that tanks
must be stored indoors or must be heated. Table 5-
2 (1) shows the freezing temperature of various
concentrations of ferric chloride.
Table 5-2.
Crystallization Temperatures for Ferric Chloride
Solutions
Solution Strength
Freezing Temperature o( an
Unagitaied Solution
% FeCI3
20
40
45
"C
-21
-23
-1
°F
-5
-10
+ 30
5.3.2 Lime
Hydrated lime is usually used in conjunction with
ferric iron salts. Although lime has some slight
dehydration effects on colloids, it is chosen for
conditioning principally because it provides pH
control, odor reduction, and disinfection. CaCOa,
formed by the reaction of lime and bicarbonate,
provides a granular structure that increases sludge
porosity and reduces sludge compressibility.
Lime is available in two dry forms, quicklime (CaO)
and hydrated lime (Ca(OH)2). When using quicklime it
is usually first slurried with water, which converts it to
calcium hydroxide prior to adding it to the sludge.
This process, which is called slaking, produces heat
and thus special equipment is required. Quicklime
normally is available in three grades: high - 88 to 96
percent CaO; medium - 75 to 88 percent CaO; and
low - 50 to 75 percent CaO. These grades can
affect the slaking ability of the material and should be
considered when deciding which grade to purchase.
In general, only quicklime that is highly reactive and
quick slaking should be used for conditioning.
Quicklime must be stored in a dry area, since it
reacts with moisture in the air and can become
unusable (2).
Hydrated lime is much easier to use since it does not
require slaking, mixes easily with water (with very little
heat produced), and does not require any special
storage conditions. However, it is more expensive
and less available than quicklime. Thus, the general
rule of thumb is to obtain and slake quicklime for
applications that require more than 1-2 tons per day.
5.3.3 Dosage Requirements
Iron salts, such as ferric chloride, are usually added
at a dose rate of 20 to 62 kg/Mg (40 to 125 Ib/ton) of
39
-------
dry solids in the sludge feed, whether or not lime is
used. Lime dosage usually varies from 75 to 277
kg/Mg (150 to 550 Ib/ton) of dry solids dewatered.
Table 5-3 (3) lists typical ferric chloride and lime
dosages for various sludges.
Inorganic chemical conditioning increases sludge
mass. A designer should expect one pound of
additional sludge for every pound of lime and ferric
chloride added (1). This increases the amount of
sludge for disposal and lowers the fuel value for
incineration. Nevertheless, use of lime can be
beneficial because of its sludge stabilization effects.
5.3.4 Design Example
A designer has calculated that the rotary drum, cloth
belt vacuum filter that will be used at the plant must
be capable of dewatering a maximum of 272 kg/hr
(600 Ib/hr) of sludge. The sludge will be a mixture of
40 percent primary and 60 percent waste activated
sludge, and it will be anaerobically digested. The
vacuum filter is to operate 7 hours per day, 5 days
per week.
To design for a margin of safety in the chemical feed
equipment, the designer has used the higher values
shown in Table 5-3. Chemical feeders should be
capable of adding 60 kg/Mg (120 Ib/ton) of FeCIs and
210 kg/Mg (420 Ib/ton) of CaO.
• Maximum daily amount of sludge to be dewatered
for this example is (refer to Appendix A for
examples of the calculations used to determine this
value):
272 kg sludge/hr x (7 hr/d) =
1904 kg/d
(4,200 Ib/d)
Maximum amount of FeCIa required per day is:
1,904 kg sludge/d x 60 kg FeCla/I.OOO kg sludge
= 114 kg/d (252 Ib/d)
The FeCIa is available as a 40 percent solution.
That is it contains 1.0 kg of active ingredient per
1.77 liters of solution (4.72 Ib/gal of solution).
114 kg/d x 1.77 liters of product/1.0 kg
= 202 liters of FeCIa solution needed/day
(53.4 gal/d)
Maximum amount of CaO required per day is:
1,904 kg sludge/d x 210 kg CaO/1,000 kg sludge
= 400 kg CaO/d (882 Ib/d)
The quicklime is available at 90 percent CaO:
400 kg CaO/d x 1 kg quicklime/0.9 kg CaO
= 445 kg quicklime/d (980 Ib/d)
The amount of extra sludge produced due to
chemical addition is estimated at one kg for every
kg of FeCIs and quicklime added. Therefore, total
maximum daily dry solids to be disposed of are:
1,905 kg sludge + 114 kg FeCIa + 445 kg
quicklime = 2,464 kg (5,432 Ib) of solids
This is the equivalent of 12,320 kg (27,160 Ib) of
wet solids at a minimum of 20 percent solids.
• Cost associated with this amount of chemicals in
1986 dollars:
= $0.26/kg ($0.12/lb)
quicklime = $0.07/kg ($0.03/lb)
114 kg FeCI3/d x $0.26/kg = $29.64/d
445 kg quicklime/d x $0.07/kg = $31.15/d
1,905 kg sludge/d * 1 Mg/1,000 kg
= 1.9 Mg (2.1 tons)/d
[($29.64 + $31.15)/d] * (1.9 tons/d)
= $31.99/dry Mg ($29.02/ton)
5.3.5 Other Types of Inorganic Conditioners
Other types of inorganic materials have been used to
condition sludge. The following is a brief description
of some of these materials and their uses:
• Coal -- Pulverized coal has been used
successfully as a conditioning agent in centrifuge
and vacuum filter studies done by EPA and others
(4). The recent study by Albertson and Koppers (5)
showed that in a concurrent, solid bowl centrifuge,
cake solids were increased from 7 to 1 4 percent
with fine coal addition in the ratio of 0.1 to 0.3 kg
coal/kg dry sludge solids (0.1 to 0.3 Ib/lb). The
main benefit of fine coal ' addition centrifuge feed
seems to be the improvement in the cake solids
concentration. Because of the increased moisture
removal provided by the fine coal feed, fuel costs
for sludge combustion can be reduced as much as
60 to 90 percent. Some concerns, including those
of safety, however, have arisen concerning
materials handling, dust generation, and incinerator
temperature control with the addition of coal.
• Cement Kiln Dust -- Cement kiln dust has been
used to successfully condition sludge prior to
dewatering on vacuum filters and also for before
and after stabilization. Kiln dust is a byproduct of
the cement and lime industries and is high in
calcium and potassium. About twice the amount of
kiln dust is required to achieve the same pH as
from lime. However, the cost is reported to be
about 30 percent that of lime. Some material
handling problems have been reported, but the
40
-------
Table 5-3. Typical Conditioning Dosages of Ferric Chloride and Lime for Municipal Wastewater Sludges'
Type of Sludge Vacuum Filler Pressure Filler
FeCI3
CaO
FeCI3
CaO
Raw:
Primary
WAS
Primary + TF
Primary + WAS
Primary + WAS (septic)
Elutriated Aerabteally Digested:
Primary
Primary + WAS
Anerobically Digested:
Primary
Primary + WAS
Primary + TF
Thermally Conditioned:
40-80
120-200
40-80
50-120
50-80
50-80
60-120
60-100
60-120
80-120
None
160-200
0-320
180-240
180-320
240-300
0-100
0-150
200-260
300-420
250-350
None
80-120 20-280
140-200 400-500
. ^
None None
* All values shown are tor pounds of either FeChj or CaO per ion ol dry solids pumped to the dewatering unit.
to/ton x 0,5 = kg/Mg
advantages appear to warrant further investigation
(6),
• Ash — Flyash, power plant ash, and sludge
incinerator ash can be used as sludge conditioning
agents to increase a sludge's dewatering rate,
improve cake release, increase cake solids, and in
some cases reduce the dosage of other types of
conditioning agents. As early as 1927 a process
was patented for taking sludge incinerator ash from
an incineration unit back to a vacuum filter to assist
dewatering. Ash has been used both as a precoat
and as a body feed in one manufacturer's
dewatering system with high pressure filtration.
However, usually ferric chloride and lime must be
used with flyash. The City of Indianapolis, Indiana
has successfully used ash as a conditioner on their
rotary belt vacuum filters to minimize conditioner
requirements and enhance cake release from the
media.
5.4 Organic Polymers
During the past decade, important advances have
been made in the manufacture of polymers for use in
wastewater sludge treatment. Polymers are now
widely used in sludge conditioning and a large variety
are now available. It is important to understand that
these materials differ greatly in chemical composition,
functional effectiveness, and cost effectiveness.
Polymers were originally used to condition primary
sludges and easy-to-dewater mixtures of primary
and secondary sludges for dewatering by rotary
vacuum filters or solid bowl decanter centrifuges.
Improvements in the effectiveness of polymers has
led to their increasing use with all types of dewatering
processes. Reasons for selecting polymers over
inorganic chemical conditioners are:
» Little additional sludge mass is produced. Inorganic
chemical conditioners typically increase sludge
mass by 15 to 30 percent.
'• If dewatered sludge is to be used as a fuel for
incineration, polymers do not lower the fuel value.
• They allow for cleaner material-handling
operations.
• They reduce operation and maintenance problems.
Selection of the correct polymer requires that the
designer work with the polymer suppliers, equipment
suppliers, and plant operating personnel. Evaluations
should be made on site and, if possible, with the
sludges to be conditioned. Since new types and
grades of polymers are continually being introduced,
the evaluation of polymers must be an ongoing
process.
5.4.1 Composition and Physical Form
Polymers are long chain, water soluble specialty
chemicals. They can be either completely
synthesized from individual monomers, or they can be
made by the chemical addition of functional
monomers or groups to naturally occurring polymers.
A monomer is the subunit or repeating unit from
which polymers are made through various types of
polymerization reactions. The backbone monomer
most widely used in synthetic organic polymers is
acrylamide. Polyacrylamide, created when the
monomers combine to form a long, thread-like
molecule with a molecular weight in the millions, is
41
-------
shown in Figure 5-2. In the form shown
polyacrylamide is essentially non-ionic. That is to
say it carries no net electrical charge in aqueous
solutions. However, under certain conditions and with
some solids, the polyacrylamide can be sufficiently
surface active to perform as a flocculant. In addition,
it is often used as a flocculation aid in conjunction
with lower molecular weight primary coagulants which
are charged molecules.
Figure 5-2, Polyacrylamide molecule-backbone of the
synthetic organic polymer.
CH ^^~ CH 9-~—*~m CH ^ ^ CH
Anionic-type polyacrylamide flocculants carry a
negative electrical charge in aqueous solutions and
are made by either hydrolyzing the amide group
(NH2) or combining the acrylamide monomer with an
anionic monomer. Cationic polyacrylamides carry a
positive charge in aqueous solutions and can be
prepared by chemical modification of essentially
non-ionic polyacrylamide or by combining a cationic
monomer with acrylamide. When cationic monomers
are copolymerized with acrylamide in varying
proportions, a family of cationic polymers with
different degrees of charge and molecular weights are
produced. These polymers are the most widely used
polymers for sludge conditioning, since most sludge
solids carry a negative charge. The characteristics of
the sludge to be processed and the type of
dewatering device used will determine which of the
cationic polymers will work best and still be cost-
effective. For example, an increasing degree of
charge is required when sludge particles become
finer, when hydration increases, and when relative
surface charge increases.
Polymers are available as dry powders or liquids. The
liquids come as water soluble solutions or as water-
in-oil emulsions. The shelf life of dry powders is
usually one or more years, whereas most of the
liquids have shelf lives of about 6-12 months and
must be protected from wide ambient temperature
variation in storage. Polymers can also be purchased
with various molecular weights and charge densities
which can greatly affect the conditioning
characteristics of the polymer and its reaction with the
sludge.
5.4.2 Structure in Solution
Organic polymers dissolve in water to form solutions
of varying viscosity. The resulting viscosity depends
on their molecular weight, degree of ionic charge and
salt content of the dilution water. At infinite dilution,
the molecule tends to assume the form of an
extended rod because of the repulsive effect of the
adjacent, charged sites along the length of the
polymer chain. At normal concentrations the long
thread-like charged anionic polymer assumes the
shape of a random coil, as shown in Figure 5-3.
Figure 5-3.
Typical configuration of an anionic polymer In
solution.
©
©
e
©
This simplified drawing, however, neither shows the
tremendous length of the polymeric chain, nor the
very large number of active polymer chains that are
available in a polymer solution. It has been estimated
that a dosage of 0.2 mg/l of polymer having a
molecular weight of 100,000 would provide 120 trillion
active chains per liter of water treated.
Dewatering is inhibited by the physical and chemical
characteristics of the sludge particles. Polymers in
solution act by adhering to the sludge particle,
causing the following phenomena to occur (see
Figure 5-4):
• Desorption of bound surface water
* Charge neutralization
» Agglomeration of small patliculates by bridging
between particles.
5.4.3 Dry Polymers
Representative dry polymers are described in Table
5-4. This table does not list the myriad of available
types and proprietary chemical differences capable of
yielding performance advantages in different sludge
systems, but it does show some of the gross
distinctions among the major types.
Dry polymers are available in powdered, granular,
bead, or flake form. The form is usually determined
by the manufacturing process. Dry polymers have
42
-------
Figure 5-4. Schematic representation of the bridging model for the destabilization of colloids by polymers.
Reaction 1
Initial Adsorption at the Optimum Polymer Dosage
Polymer
O
Particle
Destabilized Particle
Destabilized Particles
Reaction 2
Fioc Formation
Flocculation
(Perikinetic or
Qrthokinetie)
Floe Particle
Reaction 3
Secondary Adsorption of Polymer
No Contact with Vacant Sites
on Another Particle
Destabilized Particle
Restabilized Particle
Excess Polymers
Reaction 4
Initial Adsorption Excess
Polymer Dosage
O
Particle
Stable Particle
(No Vacant Sites)
Floe Particle
Reaction 5
Rupture of Roc
Intense or
Prolonged
Agitation
Floe
Fragments
Floe Fragment
Reaction 6
Secondary Adsorption of Polymer
Restabilized Floe Fragment
43
-------
T*b!o6-4. Representative Dry Powder Cationic Polymers
(Polyacrylamlde Copolymers)
Rolativa Caiionic
Density1
Low
Medium
High
Molecular
Weight2
Very high
High
Medium high
Approximate
Dosage
Ib/ton dry solids
0.5-10.0
2.0-10.0
2.0-10.0
1 Low. < 10 mote %
Medium: 10-25 mola %
High: >25 mote %
2 Very High; 4,000,000-8,000,000
High: 1,000,000-4,000,000
Medium high: 500,000-1,000,000
(Won x 0.5 - kg/Mo,
very high activities (the amount of polymeric chemical
contained in the product). The active solids
concentration is usually as high as 90 to 95 percent.
Dry polymer should be stored in a cool, dry area and
should not be exposed to moisture, since it will tend
to cake the polymer and make it unusable.
Dissolving dry polymer requires care. A typical dry
polymer make-up system, shown in Figure 5-5 (7),
contains several important components. The system
should include an eductor or other polymer wetting
device that allows for the proper pre-wetting of the
polymer particles before they enter the mix tank. After
the pre-wetted polymer enters the tank, it should be
mixed slowly until it is completely dissolved. Mixing
should then continue for at least 60 minutes to insure
that all of the polymer is completely dissolved.
Undissolved polymer can cause many problems,
including clogging of pumps and piping and fouling of
the filter belts and cloths. Mixing also allows time for
the polymer to age. During the aging process, the
molecule uncoils and takes on a form that enables it
to cause flocculation of the sludge. If the polymer
solution Is not allowed to age, the polymer will not
perform as expected.
5.4.4 Liquid Polymers
The various liquid cationic polymers, which are either
concentrated water solutions or emulsions suspended
in hydrocarbon oils, are described in Table 5-5.
Liquid polymers are available in various activities
(percent active solids). The concentration of the
polymeric material that the manufacturer can dissolve
in water is usually controlled by the viscosity of the
final solution.
Water solutions of polymer are usually purchased in
either 208-liter (55-gal) drums, in liquid bins of
about 1,040 liters (275 gal), or in bulk quantities of
about 19,000-23,000 liters (5,000 to 6,000 gal). In
colder climates, storage areas for drums, liquid bins
and bulk liquid should be located indoors and heated.
If the bulk storage tank must be located outdoors, the
tank should be heated so that the solution's viscosity
will be low enough to allow pumping. Polymers form
true solutions. Thus, mixing of the concentrated
polymer is not required.
The bulk storage tank should be fitted with a sight-
glass and low and high level sensors. The low level
sensor can be wired into a control panel in the
operations area. This sensor can be set to sound an
alarm when there is a certain amount of polymer
remaining in the tank, enabling plant personnel to
re-order polymer before running out of material. For
example, if it takes 10 days to receive a shipment of
polymer from the day of order, the low level sensor
should be set to ring when a 10-day supply is
reached. The high level sensor is important during the
delivery operation to insure that no polymer is spilled.
Polymer is extremely difficult to clean up and
precautions must be taken to prevent spills.
To safely convey the polymer from drums into the mix
tank, the operator should use either a polymer
transfer pump or a drum lifting device to empty the
drum. Polymer transfer pumps are also required to
pump the material from the storage tank to the mix
tank. They should be of the progressive cavity type
so that the polymer molecule is not subjected to high
shear forces.
It is also advisable to include a timer on the pump
control panel. The operator can then set the timer to
a specified interval, and thus the pump will always
transfer the exact amount of polymer to the mix tank.
The pump should be calibrated on a regular basis
(monthly, for example) to insure that the same
quantity of polymer is being pumped in that time
interval. The designer should insure that the timer
setting can be easily changed.
The preparation system, as shown in Figure 5-6 (7),
for this type of polymer should include a mixing tank
and a storage tank for the diluted polymer. Typically,
the operator will prepare a 0.1 percent solution of
polymer. The concentrated polymer and water should
be mixed for about 30 minules to insure a
homogeneous solution. Once the polymer has been
diluted, it is usually stable for about 24 hours.
Therefore, only enough polymer should be made up
to use in that period.
Emulsions are dispersions of polymer particles in a
hydrocarbon oil. Surface active agents are used to
prevent separation of the polymer-oil phase from the
water phase. Activities as high as 25 to 50 percent
are common with emulsions. Emulsions are available
in 208-liter (55-gal) drums, in 1,040-liter (275-
gal) liquid bins, or in bulk quantities. Storage
requirements are the same as for water solutions of
polymer, except care must be taken that no water
comes in contact with the emulsion until it is ready for
44
-------
Figure 5-5. An example of a dry polymer make-up system.
' Dry Feeder
To Process
Feed Tank
Table 5-5. Representative Cationic Polymers
Type
Solution type:
Mannieh product0
Tertiary poiyamine
Quaternary poiyamine
Quaternary poly DADM
Emulsion type:
Polyacrylamide
eopolymer
Relative
Cationic
Density3
High
High
High
High
Low
Medium
High
Molecular
Weight*3
High to very
high
Low
Very low to
medium
Low to medium
High to very
high
Percent
Solids
4-8
20-50
20-50
20-40
25-60
8 Tertiary amines charge affected by solution pH; lose cationically
in alkaline environments.
b Very tow: < 100,000
Low: 100,000-200,000
Medium: 200,000-1,000,000
High: 1,000,000-4,000,000
Very high: 4,000,000-8,000,000
c Product of polymerization reaction of condensation type.
Specifically, the condensation reaction of a primary or secondary
amine with formaldehyde and a ketone to form a beta amino
ketone.
Figure 5-6. An example of a liquid polymer make-up system.
Mixer
Bulk
Storage
1
Storage
Tank
T-T
3
i
Day
Tank
t—r
mixing and that the temperature of the storage area is
fairly constant. Premature exposure to water will
cause the polymer to coagulate. Mixing and aging of
the emulsion polymers also requires care. An
emulsion polymer make-up system is shown in
Figure 5-7 (7). Compact and portable polymer feed
automation equipment is available for in-line use that
requires no batch mixing or aging tanks.
Initially, the emulsion must be broken. Usually, a
disperser uses high pressure water to contact and
break the emulsion. The poiymer should be aged for
at least 30 minutes after it has been diluted before
use. Follow all manufacturer's recommendations for
mixing and aging of the emulsion to insure optimum
performance.
5.4.5 Polymer Feed
A typical polymer feed system should include a day
tank, polymer feed pumps, dilution water system, and
alternate feed points to the dewatering units. It can
also include an in-line static mixer. The day tank
should be sized to hold a 1-day supply of diluted
polymer or less, be made of fiberglass (which
provides the broadest corrosion resistance) and
should be equipped with a slow-speed mixer and a
sight-glass or level gauge.
The feed pumps should be of the progressive cavity
type to insure that the minimum amount of shear
forces are exerted on the polymer. (Diaphragm
pumps are also used sometimes.) These pumps
should be calibrated weekly to insure accurate
dosages of polymer. By using variable speed pumps,
the operating personnel can adjust the polymer
dosage to compensate for changes in sludge
characteristics. It is important to note that over-
conditioning the sludge is just as bad as under-
conditioning and will produce a sludge that is very
difficult to dewater. Polymer dosages should be re-
evaluated periodically (see Section 5.5.3).
In-line dilution water is a very important part of
polymer use. This water further dilutes the polymer
and makes it disperse more readily in the sludge,
thus conditioning the sludge more effectively. Polymer
manufacturers supply dilution information on the type
of water systems required. Typically 4 to 15 l/min (1
to 4 gpm) of dilution water is needed and depends on
the polymer feed rate.
45
-------
Flfluro 5-7. An example of an emulsion polymer make-up system.
• Fosdline
—r—fc-^
Flowmeter
Water Inlet
Product
Suction
Concentrated Emulsion
Polymer Fesdlina
Product Suction
The location of polymer feed points can greatly affect
the performance of the polymer and therefore, the
dewatering unit. For centrifuges, the polymer feed
point is usually inside the dewatering unit itself (see
Section 7.3). However, for a belt filter press, at least
two or three optional locations should be specified,
one adjacent to the dewatering unit, one about 1 to
1.5 m (3 to 5 ft) upstream and tied into the sludge
feed line, and one about 6 to 9 m (20 to 30 ft)
upstream. Usually connecting the feed points into the
sludge piping upstream of the unit works best. In this
case, the sludge has more time to mix with the
polymer and therefore forms a better floe. However,
overmixing should be avoided since excess shear can
degrade fragile floe.
5.4.6 Typical Polymer Dosages
5.4.6.1 Belt Filter Presses
Compared to other mechanical dewatering processes,
belt filter presses appear to have the greatest need
for optimizing the polymer dosage as a function of the
incoming sludge's characteristics (8). Under-
conditioning results in inadequate drainage of the free
water in the gravity dewatering zone, which can
cause sludge overflow in the gravity zone or extrusion
of the sludge in the pressure section. Under-
conditioned biological solids can also blind or clog the
filter belt. Over-conditioning can cause belt blinding
and problems with cake release from the belt by
making the sludge sticky. Futhermore, overflocculated
sludge may drain so rapidly that the solids are not
distributed evenly across the media. Uneven
distribution can cause tracking problems with the belt
and, moreover, can produce poor quality sludge cake.
Table 5-6 (3) lists typical levels of dry polymer
addition to condition sludge for dewatering on a belt
filter press.
Table 5-6.
Typical Dosages of Dry Polymer for Belt Filter
Presses
Type of Sludge
Raw:
Primary
Primary + TF
Primary + WAS
WAS
Aerobioaliy Digested:
Primary + WAS
Anerobically Digested:
Primary
Primary + WAS
Pounds of
Dry Polymer
per Ton of
Dry Solids
2-9
3-15'
2-20
2-20
4-15
2-10
3-15 :
Typical
Values
5
10
7
10
10
3
6
Cost per
Dry Ton"
4.30-19.35
6.45-32.25
4.30-43.00
4.30-43.00
8.60-32.25
4.30-21.50
6.45-32.25
" $2.15/lb (1986 cost for dry polymer)
Ib/ton x 0.5 = kg/Mg
5.4.6.2 Solid Bowl Centrifuges
Solid bowl centrifuges usually require polymer to
obtain good performance on municipal wastewater
sludges. Table 5-7 (3) lists typical levels of dry
polymer dosages to various sludges for conditioning
prior to dewatering in a centrifuge.
46
-------
Table 5-7. Typica! Dosages of Dry Polymer for
Conditioning Various Types of Sludges for
Dewatering In Solid Bowl Centrifuges
Type of Sludge
Raw:
Primary
Primary + WAS
Anerobically Digested:
Primary
Primary + WAS
Thermally Conditioned:
Primary + WAS
Primary + TF
per Ton of
Dry Solids
2-7
4-15
6-10
7-15
3-5
7-15
Typical
Values
4
8
6
8
3
8
Cost per
Dry Ton"
4.30-15.50
8.60-32.25
12.90-21,50
15.50-32.25
6.45-10.75
15.50-32.25
" $2.i5/lb (1986 cost for dry polymer)
Ib/ton x 0.5 = kg/Mg
5.4.6.3 Vacuum Filters
Many of the vacuum filter installations in the United
States have now converted from ferric chloride and
lime as the conditioner to polymer. With polymer
there are many advantages, such as lower costs and
fewer material handling problems. Further, the mass
of solids to be disposed of will not increase as occurs
with inorganic conditioners, and the volatile content of
the sludge cake will be higher. Table 5-8 (3) shows
amounts of dry polymer to condition different types of
sludge for vacuum filtration.
Table 5-8. Typical Dosages of Dry Polymer for
Conditioning Various Types of Sludges on
Vacuum Filters
Type of Sludge
Raw:
Primary
Primary + TF
Primary + WAS
WAS
Anerobically Digested:
Primary
Primary + WAS
Pounds of
Dry Polymer
per Ton of
Dry Solids
0,5-1
2.5-5
4-10
8-15
1.5-4
5-12
Typical
Values
1
4
6
12
1.5
7
Cost per
Dry Ton*
1.08-2.15
5.38-10.75
8.60-21.50
17.20-32.25
3.23-8.60
10.76-25.80
" $2.l5/lb (1986 cost for dry polymer)
1 Ib/ton = 0.5 kg/metric ton
5.4.6.4 Drying Beds
Sludge added to drying beds can be conditioned with
polymer. Indications are that adding 0.25 to 1.0 kg of
dry polymer/Mg of dry solids (0.5 to 2.0 Ib/ton) can
increase dewatering rates. Chapter 8 provides case
studies with examples of quantities of polymer to be
added to drying beds.
5.4.6.5 Pressure Filters
Engineers and operators at pressure filter installations
are experimenting with polymers to replace inorganic
conditoners. Some of the newer polymers on the
market appear to give good performance on the
filters. Advantages of using polymers would be lower
costs, reduced material handling, and no increase in
sludge mass for final disposal.
5.5 Design of a New Installation
The first design step is to determine how much and
what type of sludge will be produced at the treatment
plant. Calculations such as those shown in Section
7.2.6 can be used to predict, in general, the quantity
of sludge that will be produced. Another method of
estimating sludge quantity is to use the following
averages:
• 3 liters of 4-percent solid, primary sludge are
produced per m3 of wastewater treated (2,980
gal/106 gal);
• 18 liters of 1-percent solid, waste activated
sludge are produced per m3 of wastewater treated
(18,025 gal/106 gal).
These numbers were caculated by averaging values
from four different authors (9). The designer must
take care when using either the generalized equations
or these averages, since they do not take into
account unusual conditions in a particular community
that could impact on the sludge quantity. If at all
possible, the designer should try to determine actual
production rates by performing pilot plant studies of
the waste.
The type of sludge produced is determined by the
wastewater's characteristics and the wastewater
treatment process, i.e., activated sludge, trickling
filter, anaerobic digestion, etc. Each type of sludge
has different conditioning requirements.
Once these parameters are established, the engineer
can design the required equipment.
5.5.1 Design Example
For this example, assume a 18,927-m3/d (5-mgd)
activated sludge treatment plant with primary
clarifiers. After conditioning with polymer, the sludge
will be fed to belt filter presses.
Using the averages shown above, the plant will be
producing 56,397 liters (14,900 gal) of 4-percent
primary sludge and 341,123 liters (90,125 gal) of 1-
percent secondary sludge. If the mixture of sludges
were thickened to 5 percent solids, then 113,346
liters (29,946 gal) of mixed sludge or 5.6 dry Mg (6.2
tons) would have to be dewatered per day. Assuming
that dewatering is to take place on a 5-day week,
7.9 dry Mg (8.7 tons) must be processed per day.
47
-------
Referring to Table 5-6, 0.9 to 9 kg (2 to 20 Ib) of
polymer are required per dry ton of solids. Assuming
a dry polyacrylamide polymer and using the highest
value of 10 kg/Mg (20 Ib/ton), the maximum daily
polymer use will be about 78 kg/d (174 Ib/d). Based
on this, the designer must then calculate the volume
of the mixing tank, storage tank, and day-tank as
well as the capacity of the various transfer pumps. He
must also calculate the amount of storage space for
the dry polymer (number of 55-gallon drums or
volume of the bulk storage tank, depending on what
form of polymer is to be used). The bulk polymer tank
should be sized for not more than a 6- to 8-week
supply. This will insure that the polymer is always
fresh. As an alternative, it may be convenient to
specify automated polymer blending and feeding
equipment.
5.5.2 Additives
Potassium permanganate, which is often used for
odor control, has been shown to reduce polymer
doses on mechanical dewatering units (R.
McRoberts, Carus Chemical, LaSalle, IL, personal
communication, 1982). Treatment plant operating
personnel, who have optimized the polymer use in
conjunction with permanganate use, report a 5 to 15
percent reduction of polymer.
5.5.3 Selection of a Conditioning Chemical
Many factors go into the selection of the appropriate
conditioning chemical to be used at a particular plant.
These factors include such considerations as
performance, material handling, storage requirements,
type of dewatering units, final disposal method and
economics. For example, a plant whose final disposal
method is incineration wants the driest cake possible
with the least mass and highest volatile content.
Therefore, polymer conditioning is usually the better
choice when compared to inorganic chemicals.
Polymer conditioning also proves to be the better
choice if either storage space is at a minimum or if
material handling could be a problem.
Today, most municipal wastwater treatment plants are
selecting polymer conditioning over inorganic
chemical conditioning. Manufacturers' representatives
can be of great assistance in evaluating conditioning
agents. However, there are several tests, which the
designer can perform quickly and inexpensively, that
will provide a great deal of information about the
conditioner's performance. Such tests can also
estimate the quantities of agents that will be required.
If sludge is not available from the plant, the designer
could use either pilot-plant sludge or sludge from a
similar plant. However, once the plant is on-line, the
conditioner must then be re-evaluated.
5.5,3.1 Jar Test (10)
The Jar Test is used to screen conditioning agents
especially when the designer faces a wide variety of
potentially effective products. This test is performed
by taking four to six large beakers of about 1-liter
capacity and filling them with about 600 ml of the
sludge. Solutions of different types of conditioning
chemicals are prepared in accordance with the
manufacturer's instructions. Conditioning chemicals
could include ferric chloride, lime, and up to about
four polyelectrolytes. Each of the conditioning
chemicals (ferric chloride, the four polyelectrolytes,
and ferric chloride and lime in tandem) can then be
added to a different sludge sample at the
manufacturer's suggested dosage levels or at levels
noted previously in this chapter. The beakers are then
placed on a gang stirrer with the filter pan in position
beneath the oversized paddle. The paddle should just
clear the bottom of the pan. The stirrer should be set
to 75 rpm. The diluted chemical is then poured into
the filter pan and mixed for 30 seconds. The operator
then stops the gang stirrer, removes the paddle, and
observes the floe formation and settling.
5.5.3.2 Filter Leaf Testing
The Filter Leaf Test (1,9) is usually used for
evaluating dewaterability, primarily by a vacuum filter.
Further, in some cases, this test has been used to
size a vacuum filter. The test is performed by
assembling a filter leaf apparatus as show in Figure
5-8. The filter cloth should be the fabric intended for
use or monofilament filter cloth.
Figure 5-8. Example of a filter leaf apparatus.
Vacuum Gauge
Flexible Hose
Vacuum Regulating
Valve
Vacuum
Source
A jar test apparatus, as described in Section 5.5.3.1,
is used to prepare chemically conditioned sludge in at
least two-liter batches for each filtration cycle. The
48
-------
conditioning chemicals are placed in the jar test
apparatus, allowing for 2 to 4 minutes of mixing and
flocculation time. The mixing should be slow (about
10 rpm). Flash mixing will adversely affect test
results.
Two liters of the chemically conditioned sludge are
transferred to a beaker, its temperature is measured
and the filter leaf is submerged in it about 5 cm (2 in)
below the surface. Then 51 cm (20 in) Hg of vacuum
is applied to the filter leaf and timing begins. After 45
seconds of form time, the leaf is withdrawn and dried
for 90 seconds.
The cake thickness is measured and the cake is
scraped into a previously weighed dish. The dish and
cake are weighed and transferred to a drying oven.
After air drying, the cake should be desiccated,
weighed, volatilized, desiccated again, and weighed
again.
The following determinations should be made for
each run:
Volume of filtrate, ml
Temperature of filtrate, °C
Wet weight of filter cake, g
Dry weight of filter cake, g
Dry weight of ash, g
Total solids concentration of cake, Ts, percent of
wet weight
Volatile solids concentration of cake solids, Vs,
percent of total solids
This test simulates a 3-minute cycle time, divided
into 45 seconds of form time, 90 seconds of drying
time, and 45 seconds of discharge time. It may be
desirable to use a longer cycle time. Typically, cake
thickness with cycle time and cake total solids
content will increase. Cake solids content can be
further increased by decreasing the ratio of form time
to drying time. Subsequently, experiments might be
conducted to determine yield as a function of percent
solids concentration and/or a function of cycle time.
Also for each run the filter yield, in pounds of cake
solids per square foot of filter per hour, can be
calculated with Equation 5-2:
Filter Yield = Dry Weight of Cake (g) x Cycles/hr
-r [453.6 x Filter Area (ft2)] (5-2)
One of the advantages of the filter leaf test is that it
simulates actual behavior on a vacuum filter. Ease of
cake release from the filter cloth can be estimated,
and the percent moisture of the final cake can be
determined.
NOTE; The actual sizing of a vacuum filter would
require a series of leaf tests, employing one or more
controls in which no characteristics are added. A
range of chemical dosages and combinations should
then be studied.
5.5.3.3 Specific Resistances or Buchner Funnel
Test (11)
The Specific Resistance Test is another method of
predicting conditioning agent performance. A detailed
theoretical description is contained in section 7.4.4 of
this manual. The Buchner Funnel test equipment
consists of a graduated cylinder, Buchner Funnel, and
a vacuum pump as shown in Figure 5-9 (12).
Figure 5-9. Buchner funnel apparatus.
No. 2 Buchner
Funnel
Whatman
No. 2 Paper
Wire Screen
Rubber Stopper
Glass Adapter
with Side Arm
Vacuum Gauge
To Vacuum Pump
T
Pinch Clamp Here
at Start of Test
Volumetric Cylinder
A series of conditioned sludge samples are prepared
in large beakers as previously discussed in Sections
5.5.3.1 and 5.5.3.2. First about 200 ml of the
thickened sludge is placed into the beakers. The
sludge tested should be representative of the sludge
to be used on the dewatering units. This sludge can
be from a pilot-plant or a similar full-scale
treatment plant.
A Buchner funnel is mounted on top of a graduated
cylinder as shown in Figure 5-9, and the funnel is
fitted with a piece of filter paper. For each test, a
portion of the conditioned sludge (50-200 ml) is
poured into the funnel. After 2 minutes of gravity
drainage, the vacuum pump is turned on (15 in Hg).
At about 15-second intervals, the filtrate volume is
measured and recorded until the vacuum breaks or
additional water can not be removed. The sludge
cake is then removed from the filter and placed in a
weighed dish. The wet weight of the cake is
measured and then after drying at 180°C, the dry
weight is measured. Total suspended solids is
determined on the filtrate sample. In addition, the
temperature of the filtrate is also measured. A plot is
made of time/filtrate volume versus filtrate volume, as
shown in Figure 5-10. The slope of the straight line
49
-------
Figure 5-10. Time/filtrate vs. filtrate volume plot.
I
S
S
m
Slope = b = T/V
Filtrate Volume, ml
portion of the graph is "b" and is used to calculate
the specific resistance (r) from Equation 5-3:
r = (2 PA2 b)/pw
(5-3)
where,
r = specific resistance, m/kg
P = pressure of filtration, N/m2
A = area of filter, m2
b = slope of time/volume vs. volume curve,
sec/cmS
y = viscosity of filtrate, N (sec)/m2
w = weight of dry solids/volume of filtrate, kg/m3
Specific resistance must be reported in the units of
m/kg.
Figure 5-11 shows a plot of specific resistance
versus conditioning chemical dose. This plot was
constructed with specific resistance data from a
sludge conditioned with different levels of the same
chemical, in this case a polyelectrolyte. From a plot
such as this, the designer can determine optimum
polymer dose. The optimum conditioner chemical
dosage is that which produces the lowest specific
resistance.
A modification of the Buchner Funnel test can be
used to duplicate the gravity drainage results which
can be achieved on a belt filter press (BFP). This test
uses the apparatus shown in Rgure 5-9, exclusive
of the vacuum pump. A piece of the belt material
which will be used in the BFP is placed in the
Buchner Funnel. A sample of conditioned sludge is
placed into the Funnel and the volume of water
released is measured at regular intervals. Both time
and volume are recorded. The polymer and/or dose of
polymer which gives the greatest volume of free
water in the shortest time should give the best results
on the BFP.
5.5.3.4 Capillary Suction Time
The Capillary Suction Time (CST) (13) is a simple
and quick test that measures the time required for the
Figure 5-11. Plot of specific resistance vs. conditioning
chemical dosage.
I
1014r
101
1012
101'
Optimum Conditioning Chemical Dosage
1234
Conditioning Chemical Dosage, % by Weight
liquid portion of the sludge to travel 1 centimeter or
any other fixed distance. The apparatus (Figure 5-
12) consists of a timing device, an upper plate
containing probes that activate and deactivate the
timing device, and a lower plate that holds the filter
paper and a metal sample container.
Figure s-12. Capillary suction time apparatus.
_ Sludge
, Blotter Paper
J
sr
>
Timer
Electrodes
H h
1 cm
A sample of conditioned sludge is placed in the
sample container. As water migrates through the
paper and reaches the first probe, it activates the
timer. When the water reaches the second probe, the
timer deactivates. The time interval between timer
activation and deactivation is the capillary suction
time and is a measure of the dewaterability of the
conditioned sludge. Capillary suction time is plotted
versus chemical dosage. The dosage that gives the
fastest time is the optimum. Conditioner types and
concentrations should be varied until the optimum
chemical and dosage is found for a particular
dewatering system.
5.5.4 Calculations Associated with Polymer Use
Equations 5-4 through 5-7 can help the designer to
evaluate and control a polymer system.
50
-------
Calculation of Dilute Polymer Concentration from A
Water Solution of Polymer
Cn = Wp/(Vw + Vp)
(5-4)
where,
Cn = diluted polymer concentration, grams/liter
(Ib/gal)
Wp = weight of polymer added to mixing tank per
batch, g (Ib)
Vw = volume of water added to mixing tank per
batch, liter (gal)
Vp = volume of polymer added, liter (gal)
or
where,
Cn = (Vp x Dp) * (Vw + Vp) (5-5)
Dp = density of concentrated polymer in
grams/liter (Ib/gal)
Calculation of Polymer Use and Cost per Dry Ton of
Solids Dewatered
Use:
where:
Pt = Pu/Ws
(5-6)
Pt = kg (Ib) polymer used per dry Mg (dry ton) of
dewatered sludge
Pu = weight of polymer used per day, kg/d (Ib/d)
Ws = dry weight of sludge dewatered per day,
Mg/d (tons/d)
Cost:
where,
PC = Cp/Pt
(5-7)
PC = polymer cost per ton of sludge, $/Mg ($/dry
ton)
Cp = cost of polymer, $/kg ($/lb)
5.6 Thermal Conditioning (14)
The thermal conditioning process enhances the
dewatering characteristics of sludge through the
simultaneous application of heat and presssure. It is a
continuous flow process in which sludge is heated to
temperatures of 177°C to 204°C (350°F to 400°F) in
a reactor under pressures of 1,720 to 2,750 kPa (250
to 400 psig) for 15 to 40 minutes. There are two
basic modifications of the thermal conditioning
process employed in wastewater treatment. In one
modification, Low Pressure Oxidation (LPO), air is
added to the process. The other modification, Heat
Treatment (HT), does not include the addition of air to
the process. Both thermal conditioning processes
produce biologically stable sludge with excellent
dewatering characteristics.
Wastewater sludge contains water and cellular and
inert solids that form a gel-like structure. The water
portion consists of bound water, which surrounds
each solids particlej and water of hydration, which is
inside the cellular solids. Thermal conditioning
improves sludge dewaterability by subjecting the
sludge to elevated temperature and pressure in a
confined reactor vessel: thus coagulating solids,
breaking down the gel-like structure of the sludge,
and allowing the bound water to separate from the
solids particles. In addition, hydrolysis of protein
material in the sludge occurs. Cells break down and
water is released, resulting in coalescence of solids
particles. In its conditioned state, the sludge is readily
dewatered on most dewatering devices to 30 to 50
percent solids, in most cases without addition of
chemicals.
A portion of the volatile suspended solids (VSS) in
sludge is solubilized as a result of the breakdown of
the sludge structure. The solubilizaton of VSS
increases its biodegradability. Although this
solubilization does not change the total organic
carbon content of the sludge, it does result in an
increase in the BODs. The BODs produced is of
primary concern in the recycle of sidestreams. The
solubilization of VSS and the resultant BODs
production for HT systems may be estimated with
Equation 5-8:
where,
VSS
PS
WAS
VSS = 0.1 PS + 0.4 WAS (5-8)
BOD5 = 0.07 PS + 0.3 WAS
= Volatile suspended solids solubilized, dry
kg (Ib)
= Primary sludge, dry kg (Ib)
= Waste Activated sludge, dry kg (Ib)
= 5-day biochemical oxygen demand
produced by VSS solubilization, kg (Ib)
Using these rule-of-thumb procedures, 9.9 kg (22
Ib) of VSS solubilization and 7.3 kg (16 Ib) of BOD5
are produced by heat treatment (HT) of 45 kg (100 Ib)
of a typical mixture of 60 percent primary and 40
percent waste activated sludge. In LPO systems, VSS
solubilization and BODs production are expected to
be approximately the same.
Thermally conditioned sludge can be dewatered on
vacuum filters, belt filter presses, recessed plate filter
presses, centrifuges, or sand drying beds. The
dewatered solids can then be incinerated or disposed
of in a landfill or other land application method.
5.6. 1 Heat Treatment
A schematic diagram of a typical HT system is shown
in Figure 5-13. In this continuous process, raw
sludge is ground to reduce particle size to less than
0.64 cm (0.25 in) and is then pumped through a heat
51
-------
Figure 5-13. Heat treatment process flow diagram.
Raw Sludge
Storage or
Blending
Tank
Grinder
Sludge
Feed
Pump
Reactor
Decant
Liquor
Sludge-Water
Sludge Heat
Exchanger
Cake
exchanger and into a reactor. Normal discharge
pressure from the sludge feed pump is approximately
1,720 kPa (250 psi). In the heat exchanger, the
temperature of the sludge is raised from ambient to
between 149°C and 177°C (300°F and 350°F). The
heated sludge exits the heat exchanger and enters a
reactor feed standpipe, where steam is injected
through a nozzle and the sludge is mixed turbulently.
The steam and sludge proceed upward through the
standpipe and enter the reactor at the top. The hot
sludge is retained for a period of time in the reactor
and is subsequently returned through the heat
exchanger to be cooled at approximately 49°C
(120°), From the discharge side of the heat
exchanger, the conditioned sludge flows through a
control valve, which controls reactor sludge level and
pressure, and into a decant tank. The decant tank
permits rapid settling and compaction of the sludge
particles and the release of gas. The settled sludge is
pumped to a dewatering device. Process off-gases
can be treated by various odor control methods.
5.6.2 Low Pressure Oxidation
A schematic diagram of the LPO system is shown in
Figure 5-14. Raw sludge is first passed through a
grinder where particles are reduced to less than 0.64
cm (0.25 in). The ground sludge is then pumped at
approximately 2,750 kPa (400 psi) through a heat
exchanger followed by an LPO reactor. High pressure
air from the system air compressor is introduced into
the sludge flow upstream of the heat exchanger. The
air improves heat transfer and converts sulfur
52
-------
Figure 5-14. Low pressure oxidation process flow diagram.
Raw Sludge
Storage or
Blending
Tank
Standpipe
Grinder
Sludge
Feed
Pump
Compressed Air
Sludge - To -
Sludge Heat
Exchanger
Sludge
Reactor
Steam
Decant
Liquor
Liquor
Cake
products in the sludge to sulfate, slightly reducing
odors from off-gases. The resulting turbulent flow of
sludge and air proceeds through the heat exchanger
where sludge is preheated by processed sludge
returning from the LPO reactor. The sludge and air
mixture enters the reactor at a temperature between
149°C and 160°C (300°F and 320°F). Steam is
injected directly into the reactor to increase the
sludge/air mixture temperature to between 166°C and
177°C (330°F and 350°F). The combined products
rise slowly in the reactor and a slight heat of reaction
or oxidation occurs, producing a small amount of
heat. From the reactor midpoint to the reactor outlet,
the sludge temperature increases approximately 10°
due to the heat of reaction of the sludge, contributing
to an overall temperature increase from the reactor
inlet to reactor outlet of approximately 40°. Detention
time or "cook time" in the reactor is based on the
volume of the reactor and the height of the discharge
pipe (standpipe or downcomer line). The detention
time is controlled by the air, steam, and sludge flow
rates to the reactor.
After leaving the LPO reactor, the partially oxidized
product flows back through the heat exchanger and
releases heat to the incoming sludge/air mixture.
When the partially oxidized product reaches the
control valve, the temperature ranges between 43°C
and 54"C (110°F and 130°F). This valve controls the
pressure in the reactor. From the valve, the thermally
conditioned sludge and exhaust gases are released.
The settled solids are then pumped to a dewatering
device prior to final disposal. Process off-gases from
the LPO system also can be treated by various odor
control methods.
5.6.3 Economic Considerations
The increase in the cost of natural gas and fuel oil
since the early 1970s has significantly changed the
53
-------
economic feasibility of new thermal conditioning
systems for small plants. Larger installations, greater
than 0.44 m3/sec (10 mgd), that use dewatering and
Incineration with energy recovery may determine that
the addition of a thermal conditioning step would be
an economic asset.
Several factors must be considered regarding the
cost effectiveness of a thermal conditioning system
as a function of plant size.
• Present-day energy costs dictate some form of
heat recovery to make the thermal conditioning
process competitive with other conditioning
processes.
• Thermal conditoning systems require well trained
and skilled supervisors and operators to optimize
the operation and maintenance of the systems.
• Both types of systems should be supported with a
complete inventory of spare parts to reduce
excessive downtime. Also, they require a thorough
preventive maintenance program.
• The unit capital cost of thermal conditioning
systems is in the range of $385 to $550/Mg ($350
to $500/ton) of annual sludge production, when
processing over 9,090 dry Mg (10,000 tons) per
year due to use of multiple treatment units and
standby units rather than larger sized individual
units. At lower loading rates, processing costs
increase significantly, and the comparatively high
cost of support systems (such as boilers, air
compressors, and decant tanks) makes HT/LPO
systems more costly to build than other sludge
conditioning facilities,
5.6,4 Advantages and Disadvantages of HT1LPO
Conditioning
Previous literature on HT/LPO provides a summary of
the advantages and disadvantages of using these
processes to condition wastewater sludges.
Advantages cited include:
» Except for straight waste activated sludge, the
process produces a sludge with excellent
dewatering characteristics. Cake solids
concentrations of 30 to 50 percent can be obtained
with conventional dewatering equipment.
• The processed sludge does not normally require
chemical conditioning to dewater well on
mechanical equipment.
• The process stabilizes the sludge and destroys all
living organisms including pathogens.
• The process provides a sludge with a heating value
of 26,000 to 30,000 KJ/kg (11,000-13,000 Btu/lb)
of volatile solids, suitable for incineration or
anaerobic digestion with energy recovery.
• The process is suitable for many types of sludges
that cannot be stabilized biologically because of the
presence of toxic materials.
• The process is effective on feed sludges with a
broad range of characteristics and is relatively
insensitive to changes in sludge characteristics.
• Continuous operation is not required as with
incineration, since the system can easily be placed
on standby.
Disadvantages cited include:
• The process has high capital costs due to
mechanical complexity and the use of corrosion-
resistant materials, such as stainless steel, in the
heat exchangers.
• The process requires careful supervision, skilled
operators, and a good preventive maintenance
program.
« The process produces a malodorous gas stream
that must be collected and treated before release.
• The process produces dark colored sidestreams
with high concentrations of organics and ammonia
nitrogen.
• Scale formation in heat exchangers, pipes, and the
reactor requires cleaning by difficult and/or
hazardous procedures.
• Subsequent centrifugal dewatering may require
continuous or intermittent polymer dosage to
control recycle of fine particles.
• The daily sludge throughput of the process cannot
be adjusted by a significant amount without
incurring high energy and/or labor costs.
5.7 References
1. Dewatering Municipal Wastewater Sludges.
EPA-625/1-82-014. U.S. Environmental
Protection Agency, Center for Environmental
Research Information, Cincinnati, OH, 1982.
2. National Lime Association. Lime, Bulletin 213.
Arlington, VA.
3. Process Design Manual for Sludge Treatment
and Disposal, EPA-625/1-79-01 1, U.S.
Environmental Protection Agency, Center for
Environmental Research Information, Cincinnati,
OH, 1979.
54
-------
4. Hirota, M., H. Okada, Y. Misaka, and K. Kato.
Dewatering of Organic Sludge by Using
Pulverized Coal. Presented at 47th Annual
Conference of the Water Pollution Control
Federation, Denver, CO, 1974.
5. Albertson, O.E. and M. Kopper. Fine Coal-aided
Centrifigual Dewatering of Waste Activated
Sludge. JWPCF 55(2): 145, 1983.
6. Martin Marietta Corp. Sludge Conditioning with
Cement Kiln Dust, Baltimore, MD, 1981.
7. Calgon Corp. Polymer Make-up Systems,
Pittsburgh, PA, 1980.
8. Belt Filter Press Survey Report. American Society
of Civil Engineers, New York, NY, 1985.
9. Vesilind, P.A. Treatment and Disposal of
Wastewater Sludges, Revised Edition. Ann Arbor
Science Publishers, Ann Arbor, Ml, 1980.
10. Calgon Corp. Chemical Application Bulletin 12-
5d-Jar Test Procedure. Pittsburgh, PA.
11. Coakley, P. and B.R.S. Jones. Vacuum Sludge
Filtration, I. Interpretation of Results by the
Concept of Specific Resistance. Sewage and
Industrial Wastes 28:963, 1956.
12. Metcalf and Eddy, Inc. Wastewater Engineering
Treatment and Disposal, 2nd Edition. McGraw-
Hill, New York, NY, 1979.
13. Gales, R.S. and R.C. Baskerville. Capillary
Suction Method for Determination of the Filtration
Properties of a Solid/Liquid Suspension.
Chemistry and Industry, 1967.
14. Heat Treatment/Low Pressure Oxidation Systems:
Design and Operational Considerations. EPA-
430/9-85-001, U.S. Environmental Protection
Agency, Office of Municipal Pollution Control,
Washington, DC, 1985.
55
-------
-------
Chapters
Air Drying Processes
6.1 Introduction
As used in this manual, "air drying" refers to those
dewatering techniques by which the moisture is
removed by natural evaporation and gravity or
induced drainage. There may be some mechanical
assistance, such as turning and mixing the sludge on
paved beds, or some vacuum assistance but the
movement of water is controlled by natural forces.
The air drying processes described in this chapter
range from the oldest dewatering concept (sand
beds) to some recently developed techniques. They
include:
Sand beds
Freeze assisted sand bed dewatering
Vacuum assisted beds
Wedgewire beds
Sludge lagoons
Paved beds
Other innovative processes.
Air drying processes are less complex, easier to
operate, and require less operational energy than
mechanical dewatering systems. They do, however,
require a larger land area and more labor, primarily for
sludge cake removal. The combination of all of these
factors suggests that air drying processes are
especially well suited for small to moderate sized
communities with design wastewater flows less than
7,500 m3/d (2 mgd). Air drying processes are
technically feasible at greater flows but the need for
the larger land area restricts the economic feasibility
in some locations. Air drying processes should be
given strong consideration for all small to moderate
sized communities, and for larger facilities in arid and
semi-arid climates when land is available. Sand
beds, freeze dewatering, and reed beds can easily
produce a sludge cake with 25 to 40 percent solids
and can exceed 60 percent solids with additional
drying time. These three processes can produce a
drier sludge than any of the mechanical dewatering
devices discussed in Chapter 7.
6.2 Sand Beds
Sand beds have been used successfully for sludge
dewatering since wastewater treatment became a
recognized technology early in this century. In 1987,
they are still the most frequently used technique for
sludge dewatering in the United States. Oewatering
on the sand bed occurs through gravity drainage of
free water followed by evaporation to the desired
solids concentration level. Figure 6-1 illustrates
details of a typical sand bed. (In areas of high
precipitation, covered sand beds have been used.)
Figure 6-1. Sand bed details.
Gate
Sludge
Sidewalls can be constructed of reinforced concrete
as shown in Figure 6-1, or treated timber planks or
concrete planks. The plank type construction has the
advantage of allowing adjustment of the total depth of
the bed, an important feature for the freeze
dewatering process discussed in Section 6.3.
6.2,1 Design Considerations
The critical design parameter is the surface area of
the sand bed required to attain the necessary
drainage and evaporation in the specified time. Most
57
-------
of the sand beds in current use were designed with
per capita loading criteria that were developed
empirically in the early 1900s (1,2). Some State
agencies still specify drying bed criteria on the basis
of per capita loading; so it is necessary to obtain
these values prior to any specific project design.
These per capital loading criteria are still valid if all of
the original conditions are incorporated, but this is not
always the case in modern systems. The preferred
approach is to base design on the mass loading of
solids. Currently accepted loading criteria are
presented in Table 6-1 (1). The values for digested
primary plus waste activated sludge (WAS) for
uncovered beds range from 60 to 100 kg/m2/yr (12 to
20 Ib/sq ft/yr). These values can be increased to 85
to 140 kg/m%r (17 to 28 Ib/sq ft/yr) for covered
beds, and a significant further increase is possible if
coagulants are used to condition the sludge prior to
application on the bed (1). The upper end of these
ranges applies for warm dry climates and to sludges
which drain readily.
Tat>lo6-1. Loading Criteria for Anaerobically Digested,
Non-Conditioned Sludge on Uncovered Sand
Beds
SludaoTypo Mass Loading
Primary
Primary
Primary plus low-rate TF
Primary plus WAS
kg/m2/yr
120-200
100-160
100-160
60-100
x 0,2048 - Ib/sq ft/yr.
A sand bed's performance depends on:
• The required solids concentration in the dewatered
sludge
• The solids concentration in the applied sludge
• The type of sludge to be applied (e.g., stabilized,
thickened, conditioned)
* The drainage and evaporation rates.
The required solids concentration depends on the
technical or regulatory requirements for final sludge
disposal or utilization (see Chapter 2 for discussion).
If no special requirements apply, the sludge cake is
typically "liftable" at about 25 to 30 percent solids
and can be removed from the bed without excessive
sand loss.
Water is removed from the sludge through gravity
drainage and evaporation. The amount of water that
can be removed by drainage is strongly influenced by
the type of sludge applied. Drainage might account
for 25 percent of water removal for some
anaerobically digested primary plus waste activated
sludges, and 75 percent or more for well conditioned
sludges. Significant drainage is typically complete
within 3 to 5 days. However, the unit drainage rate is
less critical than the total percentage of water
removed (percent total solids).
The rate of evaporation is a function of local climatic
conditions and the sludge surface characteristics.
Seasonal evaporation rates can be obtained from
local pan or lake evaporation values. Since the crust
which forms on the sludge surface inhibits
evaporation, the pan evaporation values must be
adjusted when designing the sand bed. An adjustment
factor of 0.6 was experimentally derived (3) (see
Section 6.7 for further discussion). Once "cracking"
occurs (see Figure 6-9), the evaporation rate should
approach the pan value due to the additional sludge
surfaces exposed.
Design equations which relate the initial and final
solids concentration and the amount of water lost to
drainage and evaporation are presented below. These
design equations provide a rational method of
determining the design mass loading on a sand bed
and other critical operational parameters.
The drying time for a single application is given by:
(6-1)
where,
t(j = dewatering time for a single application, mo
yo = initial depth of applied sludge layer, cm (in)
SQ = initial dry solids concentration required for
dewatered sludge, percent
Sf = final dry solids concentration required for
dewatered sludge, percent
D = fraction of water removed by drainage,
percent as decimal
Ev = average pan evaporation during time td,
cm/mo (in/mo)
ke = reduction factor for sludge evaporation vs. a
free water surface, percent as decimal
= 0.6 (a pilot-test is recommended to
determine this value)
The number of applications during the operating
season is given by:
58
-------
N= — =
-D)
(6-2)
where,
N = number of sludge applications
nv = length of operating season, or an increment
if evaporation is significantly different, mo
Evn = average pan evaporation during period nv,
cm/mo (in/mo)
The design solids loading is given by:
(C)(s0)
(6-3)
where,
L
C
= solids loading during period nv, kg/m2 (Ib/ft2)
= Conversion factor (assumes specific gravity
of sludge = 1.04)
= 10.4 (metric units)
= 5.41 (US units)
The annual solids loading is determined by
summation of the results of equation 6-3 for the
different operational periods selected, over a full 12-
month annual cycle. To obtain a first approximation,
assume nv = 12 and use the average annual pan
evaporation for EVn, if the beds are to be operated on
a year-round basis.
The final depth of the dewatered sludge cake is given
by:
in) layer, as recommended in many design texts, at 3
percent solids would produce a dewatered cake at 30
percent solids only 2 cm (0.8 in) deep. Such a depth
is too thin for most mechanical removal techniques
and could result in excessive sand loss.
To keep operation and maintenance costs as low as
possible, the design goal is to achieve the maximum
possible solids loading with the minimum number of
application and removal cycles. Repeated calculations
with Equations 6-1 through 6-4 will converge on
the most effective combination of initial solids
concentration and layer depth for a particular project.
Final optimization of the layer depth is only possible
with operational experience.
The annual solids loading depends on the solids
concentration in the applied sludge, as shown by
Equation 6-3. This relationship is demonstrated by
Figure 6-2 using actual data from 13 operational
systems in Pennsylvania, Ohio, New York, California,
Texas, Illinois, and North Carolina (4,5).
Figure 6-2
2001-
€. 160
Solids loading rates for sand beds vs. solids
content of applied sludge.
120
ra
c.
"•5
2
"6
80
40
2468
Solids Content of Applied Sludge, %
10
(6-4)
where,
yi = final depth of dewatered sludge cake, cm
(in)
Thin layers of sludge will dry faster than a thick layer,
but (as defined by Equation 6-3) the annual solids
loading is independent of the depth of the individual
layers applied. Using too thin a layer has several
disadvantages, including more frequent operation and
maintenance (some dried sludge must be removed
before the next application), greater sand loss from
the bed, and increased costs in general. A 20 cm (8
An increase in solids content from 2 to 4 percent, for
example, would approximately double the loading rate
and could reduce the required bed area by one half.
This relationship demonstrates the potential
advantage of thickening or preconditioning the sludge
prior to application on the bed. However, increasing
the solids content beyond 8 percent is not
recommended since the sludge will not flow and
distribute uniformly on the bed beyond this level. At
existing facilities where expansion may not be
possible, the use of prior sludge thickening might
allow more effective use of the present bed area.
Typically, the total bed area is subdivided into multiple
cells. It is convenient to size the cells so that one or
two can contain the total volume of sludge from a
59
-------
scheduled digester withdrawal. The width of the bed
depends on the removal method. Small to moderate
sized facilities with hand or semi-mechanical
removal are about 6 m (20 ft) wide. Greater widths
are used with mechanical removal methods; each cell
must contain an entry ramp and possibly paved
runway slabs for equipment. Sand beds as long as 30
to 60 m (100 to 200 ft) have been successfully used
with the more dilute sludges. Uniform sludge
distribution on the bed can be difficult, particularly
when polymers are used for conditioning. In these
cases, the bed length should not exceed 15 to 25 m
(50 to 75 ft) and/or multiple distribution points should
be incorporated into the design.
6.2,2 Structural Elements
Sludge can be applied to each cell with a valved pipe
(plug valves) or from an open channel with gate
controls along the perimeter of the bed. The open
channel is easier to clean, but is more difficult to
operate in cold weather. The valves in a pipe network
should be protected from freezing in cold climates,
since the adjacent pipe will not always drain
completely. A splash block on the bed at every
sludge entry point minimizes erosion of the sand.
A minimum sand depth of 30 cm (12 in) is
recommended. In some cases, depths of up to 46 cm
(18 in) can be used to extend the life of the bed.
Since sand is unavoidably removed every time sludge
is taken from the bed, new sand must eventually be
installed. Preferred characteristics for the sand are:
• Clean, hard particles - no clay, silt, or organic
matter
• Effective size 0.3 to 0.75 mm (0.01 to 0.03 in)
• Uniformity coefficient < 3.5.
The gravel layer is usually 20 to 46 cm deep (8 to 18
in), with gravel sizes ranging from 3 to 25 mm (0.1 to
1.0 in). With mechanical sludge removal, greater
depth of gravel is needed to structurally protect the
underdrain network. A thinner layer of coarser stone,
overlain by a suitable permeable geotextile
membrane, can be used with hand removal or
mechanical removal with very light equipment.
Underdrains are usually plastic pipe or clay tile laid
with open joints. The main underdrain pipes should
be at least 10 cm (4 in) in diameter and should be
laid with a slope of at least 1 percent to insure
drainage. Spacing of these main underdrains ranges
from 2.5 to 6 m (8 to 20 ft) depending on the type of
sludge removal planned. Lateral drain pipe branches
connected to the main drains should be on about 2.5
m (8 ft) centers.
Covered beds have often been used in northern
areas to extend the otherwise seasonal dewatering
operations. Standard designs for glass or plastic
enclosures are available. The freeze dewatering
technique described in Section 6.3 allows the use of
uncovered beds throughout the winter in cold
climates. The covered beds might still be desirable in
the marginal zones shown on Figure 6-3 where the
potential for freeze dewatering is limited.
6.2.3 Performance Expectations
The drainage rate during the initial dewatering phase
depends on the type of wastewater and sludge
treatments used, on the concentration of solids, and
the depth of sludge applied. Well stabilized aerobically
digested sludges will tend to drain more completely
than anaerobically digested sludges, for example, and
most raw sludges will drain more readily than
digested sludges. Drying of raw sludges, however, is
likely to result in malodorous conditions in all but dry
warm climates.
The drainage phase is usually measured in terms of
hours or a few days. The evaporative stage lasts as
long as is necessary for the sludge to reach the
desired solids concentration as defined by Equation
6-1. The most significant volume reduction occurs
up to about 30 to 40 percent solids. Further drying
beyond that point achieves little additional volume
reduction but may still be required by regulatory
authorities. Sludge at about 30 percent solids can be
removed from the bed with minimal loss of sand.
The water collected in the underdrainage network is
returned to the wastewater treatment facility.
Characteristics of this liquid will depend on the type of
treatment process used, and may be similar to that
reported in Table 6-2.
Table 6-2.
Parameter
Sludge Filtrate Characteristics After Freeze
Thawing
Digested Sludge
Raw WAS
Aerobic
Anaerobic
BOD, mg/l (dissolved)
COD, mg/l
SS, mg/l
Total P, mg/l
706
1,585
14
28
722
1,815
17
46
1,012
3,325
18
80
6.2.4 Operation and Maintenance
The optimum depth of sludge to apply will be
determined with experience; in general this depth may
range from 20 to 45 cm (8 to 18 in). Any chemical
conditioners should be added continuously during the
pumping operation, at points in the system that will
insure proper mixing. Multiple dosage points for
polymers should be constructed into the system.
These dosage points, at a minimum, should be
located ahead of the pump suction, at the pump
discharge, and ahead of the discharge point to the
bed. It may not be necessary to use all dosage points
60
-------
but the multiple array will allow optimization after
operation commences.
Bed maintenance involves the periodic replacement
of sand lost during sludge removal, leveling and
scarification of the sand surface prior to dosing, and
removal of vegetation. Odors should not be a problem
with well stabilized sludges. To control odors, calcium
hypochlorite, potassium permanganate or ferrous
chloride can be added to the sludge during
application to the bed.
The time required for O & M activities is a direct
function of the size of the system and the number of
operational cycles used in a year. Very small systems
(<100 m2, at 100 kg/m2/yr) might require about 4
hr/yr m2 (0.4 hr/yr/ft2), while larger systems (> 4,000
m2, at 100 kg/m2/yr) need less than 0.5 hr/yr/m2 due
to the economies of scale and increased
mechanization.
6.2.5 System Upgrading
At many locations the original sand bed capacity may
become inadequate. Causes include increased
system flow and sludge production or the use of
chemicals in wastewater treatment that can
significantly increase the mass and volume of sludge
to be handled, as well as adversely change the
drainage characteristics of the sludge. Alum, for
example is often used for phosphorus removal, but
the resulting sludge does not readily drain on sand
beds. A similar situation often occurs with iron salts.
Some of these concerns are reflected in Table 6-1
with the larger sand bed area allowance for "Primary
plus chemicals." Similar allowance must be made for
"Primary plus waste activated sludge" when
aluminum or iron salts are present. Typically, a plant
operating at capacity that decides to add aluminum or
iron salts for phosphorus removal may need to take
action in one or more of several ways. These include
constructing additional drying beds, adding
polyelectrolytes to facilitate dewatering, modifying
sludge application procedures, lengthening sludge
drying times, and replacing sand beds with some type
of mechanical dewatering device. Alternatives that
must be considered include the use of polymer
conditioning alone or combined with sludge
thickening. Either approach may allow the more
effective use of existing sand beds, or in new designs
allow the use of sand beds where land limitations
exist.
In the typical case with anaerobically digested sludge,
about 25 to 60 percent of the total water applied is
removed by drainage. The use of an appropriate
polymer accelerates the particle agglomeration and
increases the amount of water which can be drained.
Polymer conditioning reduces the amount of water to
be evaporated and thereby allows a significant
increase in solids loading.
The use of polymers for conditioning is discussed in
detail in Chapter 5. The selection of polymer type and
optimum dose is based upon the results of tests with
the sludge in question. These tests are usually series
of Buchner funnel or capillary suction time (CST)
tests. Typical costs range from $3.00 to $11.00 per
ton of dry sludge solids (see Chapter 8 for case
studies). Sludge characteristics can change and it is
often necessary to vary the dose or the polymer type
occasionally. Sludges that drain poorly are the best
candidates for polymer conditioning. Polymers are
also a benefit with thin sludges having high
concentrations of fine particles. Sludges of this type
tend to penetrate beyond the sand surface and
eventually plug the bed. Polymers increase drainage
rate and reduce penetration of sludge particles.
Even if polymers have been used, sludges that drain
poorly will usually still require additional drying time in
the evaporative stage to reach the desired solids
concentrations. The use of polymers can increase
bed capacity by inducing greater drainage and
thereby reducing the time required for evaporative
drying, but the evaporation rate is unaffected (6,7).
Problems can occur if either mixing or distribution are
inadequate, which can cause localized deposits and
clumps of sludge which are then slow to drain. To
insure the effective use of the entire bed area when
using polymers, multiple inlet points are
recommended.
6.2.6 Costs
The capital cost for sand beds is strongly influenced
by the cost of land at the project site. Other major
factors include the containing walls and bed bottom,
the application and drainage piping, and any sludge
removal equipment. The major O & M costs are labor,
fuel for equipment, periodic sand replacement, and
conditioning chemicals (if used). Section 5-8 in
Reference 8 can be used to estimate capital and O &
M costs for sand beds. Section 6-5 in the same
reference should also be used if polymer conditioning
is planned. The design example in Appendix A
demonstrates the use of these procedures.
6,3 Freeze Assisted Sand Bed
Dewatering
Freezing and then thawing a sludge will convert a
material with a jelly-like consistency to a granular
type material that drains readily. Solids concentrations
exceeding 20 percent will be realized as soon as the.
material is thawed and 50 to 70 percent can be
achieved with minimal additional drying time (9). The
effects of sludge freezing have been recognized for
over 50 years but until recently a generally applicable
design procedure was not available (9).
Freezing will work with any type of sludge, at any
solids concentration, but is particularly effective with
chemical and biochemical sludges that do not drain
readily. Energy costs for artificial freeze-thawing are
61
-------
prohibitive so the concept must depend on natural
freezing to be cost effective. As shown in Figure 6-3
the feasible area can include most of the northern half
of the United States. The recently developed design
procedure will allow the year-round use of new and
existing uncovered sand beds in colder climates.
6.3.7 Design Considerations
The freeze-dewatering effects will be realized
regardless of the initial sludge concentration or the
degree of stabilization, but the cost effectiveness of
the operation will be influenced by both factors. A
very dilute sludge will increase costs by requiring
more area for freezing beds; thickened sludge in the
range of 3 to 7 percent solids works well. The use of
stabilization for wastewater sludges is recommended
to avoid odor complaints during thawing and drying
and to meet regulatory requirements for final disposal.
The design of a freeze dewatering system must be
based on worst case conditions to insure successful
performance at all times. If sludge freezing is to be a
reliable expectation every year, the design must be
based on the warmest winter during the period of
concern (usually 20 years) and on a layer thickness
which will freeze within a reasonable time if freeze-
thaw cycles occur during the winter. It is essential for
the layer to freeze completely to achieve the
dewatering benefits. In many locations a large single
layer may never freeze completely to the bottom, with
only the upper portion going through alternating
freezing and thawing cycles. Recent research (9) has
indicated that an 8-cm (3-in) deep layer of sludge
is practical for most locations in moderately cold
climates. A thicker layer is feasible in colder climates;
facilities in Duluth, MM, for example, successfully
freeze water treatment sludges in 23 cm (9 in) layers
and a 46-cm (18-in) layer has been used in
Fairbanks, AK. An 8-cm (3-in) layer should be
assumed for feasibility assessment and preliminary
design. A larger increment may then be justified by a
detailed evaluation during final design. The freezing or
thawing of a sludge layer can be described with
Equation 6-5:
Y =
1/2
(6-5)
where,
Y = depth of freezing or thawing, cm (in)
m = proportionality coefficient, dependent on
thermal conductivity, latent heat of fusion
and density of the material being frozen,
cm(°Od)-1/2 [in
Figure 6-3. Potential depth of sludge (cm) that could be frozen, if applied in 8-cm layers (13).
Freezing Not Practical
on a Routine Basis
62
-------
AT»t = freezing or thawing index, °Od (°F»d)
AT = difference between average ambient air
temperature and freezing temperature, °C
t = time period of concern, days
The proportionality coefficient was experimentally
determined with wastewater sludges and should be
valid over the range 0 to 7 percent solids. The
freezing or thawing index is an environmental
characteristic for a particular location. The values are
sometimes published but can also be determined
from weather records. For example:
Average daily air temperatures:
0°C, -3°C,
-7°C, -4°C
Time period t = 4 d
Average temperature during period = -3.5 °C
Freezing index for the period
= [0°C -(-3.
= 14°Od.
Since a layer thickness of 8 cm was suggested for
preliminary design, it is possible to rearrange
Equation 6-1 and solve for the time to freeze the
design layer using Equation 6-6:
A7V = (Y/m)
(6-6)
with Y = 8 cm and m = 2.04 cm
equation 6-6 becomes
This form is used with the local weather records to
determine how many 8-cm (3-in) layers can be
frozen during each winter of the study period. The
year with the smallest number of layers is then the
control year for design.
It can be assumed, for example, that the first layer is
applied in late fall, and Equation 6-6 is then used to
determine the number of days required to freeze the
layer under the average temperature conditions
indicated in the records. Either the intensity or the
duration of the low temperature must be sufficient to
freeze the layer in a continuous period. The
calculations are repeated for the entire winter season
with a 1-day allowance for each sludge application
and cooling, and due account taken of any thaw
periods during the winter. The next layer is not
applied until calculations show that the previous layer
has frozen completely. This procedure can be easily
programmed for rapid calculations with a small
computer or desk-top calculator.
6.3.1.1 Preliminary Feasibility Assessment
A rapid method for preliminary assessment and
design relates the potential depth of sludge which
may be frozen in the "design" year to the maximum
depth of frost penetration at a particular location.
There is a high correlation between the two factors
since they both depend on the same environmental
conditions. The maximum depth of frost penetration
for an area can be found in local records or other
published sources (10). The relationship between the
two factors is defined by Equation 6-7:
]>>=1.76F - 101 (6-7)
where,
EY = total depth of sludge which could be
frozen in 8-cm layers during the
"design" year, cm
Fp = maximum depth of frost penetration into
the soil for the location, cm
In US units, with Y and Fp in inches, the equation
is:
Y=l.76F -38
p
Equation 6-7 is the basis for the map shown in
Figure 6-3. The map and Equation 6-7 are only
valid for preliminary estimates. Detailed weather
records and Equation 6-6 should be used for final
design. Although very effective, freeze dewatering is
a seasonal process. Except in very cold climates it is
not economical to store sludge in the warm months
and depend only on winter freezing for dewatering. In
most parts of the United States it will be more cost
effective to combine winter freezing with polymer
assisted summer dewatering on the same beds. This
combination of techniques would eliminate the need
for large scale sludge storage and reduce the total
number of beds required. Figure 6-4 demonstrates
the application of Equation 6-6. The arrows and
circles on the diagram represent the predicted times
for sludge application and layer freezing at Duluth,
MN. The predicted total depth is the same as was
actually observed at the treatment facility (11). This
design approach has also been independently verified
with full-scale tests in Sweden (12).
6.3.2 Structural Aspects
The basic facility is essentially the same as for
conventional sand drying beds. The major design
difference is increased freeboard to contain the
design depth of frozen sludge. The maximum
potential for freezing will occur when the sludge is
exposed to the extreme weather conditions, so
covering the bed, using windbreaks, or applying the
63
-------
Figure 6-4. Predicted vs. measured sludge freezing at
Duluth, MM.
10
-to
120
tn
I 40
5
*
Msen Weekly Air Temperature
Actual Total Depth, Layer
Freezing (1980-81 Winter! ~~|
. T
Predicted Total Depth if
Frozen in 20-om Layers
I
Dec 1380 Jan 1981 Feb 1981 Mar 1981
™ 20 cm of Sludge Applied
— Layer Frozen
sludge in a deep trench will only reduce the freezing
rates. Trenches are used at the system in Duluth, MN
(11), for sludge storage during the warm months.
Supernatant is decanted prior to the onset of winter.
As soon as a surface layer of sludge is frozen, a hole
is drilled in the ice and sludge is pumped up to freeze
in layers on top of the ice.
An effective way to provide the necessary freeboard,
and still allow exposure of the sludge to winter
conditions, is the use of concrete or timber planks to
increase the sidewall depth of the bed as the winter
progresses. The sludge feed system must be
designed to apply each layer on top of the previously
frozen material. A hydrant and hose combination is
one possibility.
6.3.3 Performance Expectations
Freezing a sludge changes both the structure of the
sludge water mixture and the characteristics of the
solid particles. In effect, the solid matter tends to be
compressed into large discrete conglomerates
surrounded by frozen water. When thawing
commences, drainage occurs instantaneously through
the large pores and channels created by the frozen
water. Cracks in the frozen mass also act as conduits
to carry off the melt water.
Experience in a number of locations (9,12,13) has
shown that the solids concentration will approach 25
percent as soon as the frozen mass is completely
thawed, due to the very rapid drainage. Figure 6-5
shows the drainage rate of a frozen sludge after
thawing as compared to the drainage rate of the
same sludge without freezing.
Figure 6-5. Effect of freeze thawing on the drainage rate
for anaerobically digested sEudge.
9% Solids
JL Unfrozen
i» Frozen and Thawed
200
400 600
Time, hr
800
1,000
The time required for thawing can be estimated with
Equation 6-6, using an "m" value of 3.78 cm
(°C»d)l/2 (1.11 in(°F«d)1>2) (13). In this case the
coolest expected spring/summer temperatures should
be used for design purposes, and the depth to be
used in the equation is the total depth of frozen
material, not the individual layers. In extremely cold
climates it would be possible to freeze more sludge,
in thin layers, than could be thawed in the very short
summer; this is unlikely to occur anywhere in the
continental U.S. However, the time to thaw should be
calculated for all locations above the 150 cm line on
Figure 6-3 to insure that the frozen sludge will thaw
in time for the bed to be used in the conventional
manner during the spring and summer.
The maximum potential response during both the
freezing and thawing portions of the cycle can be
obtained by exposing the sludge on open uncovered
beds. Section 6.3.4 provides operational guidance for
rain or snow conditions during the freezing, thawing,
or drying phases.
The drainage of water during thawing may occur at a
faster rate, and will produce a greater volume when
compared to applying the same unconditioned sludge
to a conventional sand drying bed. Typical
characteristics of this drained liquid are given in Table
6-2 (14). Also, the freezing and thawing process will
not improve the pathogen kill in the sludge. Freezing
conditions preserve rather than destroy most
pathogens.
6.3.4 Operation and Maintenance
The critical operational requirement is to ensure
complete freezing of the sludge layer before the next
is applied. Hand probing with a small pick or axe is
the easiest way to make this determination at small
facilities. Remote temperature sensing devices, such
as thermocouples, can be used but will not directly
indicate when freezing commences or exactly when
the mass is completely frozen since the temperature
64
-------
will remain at 0°C until all of the latent heat is drawn
off. The operator should maintain records of average
daily temperature and other weather conditions during
the freezing, thawing, and drying phases. With this-
data and some experience, the operator can quickly
develop site specific criteria for sludge applications
and removal.
Rainfall during the final drying phase seems to have
few lasting effects. Rainfall during the freezing stages
may thaw some of the previously frozen material. If
time permits, the thawed and drained material can be
removed; otherwise the next sludge application
should not be made until any melt water has drained
or refrozen. A light snowfall, <5 cm (2 in) just prior to
or during the freezing stage is not a concern since
the mass of water involved is small and the snow will
help in the initial sludge cooling. A heavy snowfall will
act as an insulating barrier and retard the freezing
rate. Snow layers greater than 5-8 cm (2-3 in)
should be removed from the bed with a snow blower
or front-end loader prior to the next sludge
application to ensure maximum exposure of the
sludge.
In most cases it will be more cost effective to
combine freeze thaw dewatering with polymer
assisted dewatering and/or thickening in the warm
months on the same drying beds. In order to optimize
bed use, the operator should only apply that quantity
of sludge for freezing that can be removed from the
bed as dried cake by mid-May of each year. The
system design will provide a conservative procedure
based on worst case conditions. The operator can
make the necessary adjustments depending on the
weather conditions during each winter.
6.3.5 Costs
The capital costs for freeze thaw dewatering beds are
essentially the same as described in Section 6.2.6 for
uncovered, conventional sand drying beds. Additional
capital costs may be required for the extra timber or
concrete planks and for thermal protection of any
exposed sludge piping, since the system must
operate throughout the winter. Operation and
maintenance costs should be about the same as for
conventional sand drying beds (see Section 6.2.6).
6.4 Vacuum Assisted Drying Beds
(VADB)
This dewatering technology applies a vacuum to the
underside of rigid, porous media plates on which
chemically conditioned sludge has been placed. The
vacuum draws free water through the plate and
essentially all of the sludge solids are retained on top,
forming a cake of fairly uniform thickness. Figure 6-
6 is a plan view of a typical single bed system
showing the necessary or desirable support
components, Figure 6-7 is a typical outdoor facility,
and Figure 6-8 is a cross section view of a typical
epoxy-bonded media plate.
Figure 6-6. Plan view of a vacuum assisted drying bed
system.
H
N
M
D-
A. Entrance Ramp
B. Off-Bed Level Area
Area Drain
Curbing
Sludge Distribution Piping
Bed Closure System
Media Plates
Corner Drain
Bed Containment Wall
J. Truck Loading Area
K. Area Drain
C.
D.
E.
F.
G.
H.
I.
L. Wash Water Supply
M. Sludge Feed Inventory Tank
(below grade, seldom
required)
N. Control Building with
— Sludge Feed Pumps
— Polymer System
—Vacuum Pumps
— Control Panel
— Filtrate Receiver/Pumps
(below grade)
Figure 6-7. Typical outdoor vacuum assisted drying bed.
6.4.1 Design Considerations
The basis for design is the average annual sludge
(dry solids) production rate and the number of cycles
which can be conveniently carried out in a typical
65
-------
Figure G-8 Cross section of a typical vacuum bed media
plate.
Surface Material
Diameters: .
1 mm to 3 mm
Thickness:
1 /4 in to 1 /2 In
Gravel Support
Material
Diameters:
1/8 in to 1 in
Thickness:
1-3/4 in to 3 in
work week. To insure reliability, the design may
require an increase in the size or number of beds, A
two-bed system should be the minimum standard. If
the sludge production rate exceeds one dry ton of
solids per day, a three-bed system is recommended.
A properly sized three-bed system, using a 24-
hour total cycle time, would utilize two of the beds for
dewatering each operating day with the third bed idle.
Each bed in such a system should be sized to
dewater, at a minimum, 70 percent of the average
daily sludge mass drawn from the treatment system.
This design will allow the dewatering system to be
operated a maximum of 5 days per week and still
provide for the dewatering of 7 days' accumulation of
sludge. In the event the applied sludge did not
properly dewater within the allowed 24 hours, the
third bed could be used during the succeeding days
to dewater one-half of the daily sludge production,
or the other two beds could be used over the
weekend to get the system back on schedule.
A solids loading of about 10 kg/m2/cycle (2
Ib/ft2/cycle) has been found acceptable. Adjustments,
based on the expected efficiency and effectiveness of
the operation, may be considered by the designer.
6.4.2 Structural Elements
The vacuum assisted beds are proprietary devices
and there are minor differences in the components
offered by the various manufacturers. The listing
bolow is a "generic" description of the common
elements (15).
• A support structure, either a level concrete slab or
level graded stone overlaying a sloped concrete
slab upon which the media plates are placed.
• A concrete wall surrounding three sides of the bed
and a bed closure system on the fourth wall, to
allow for the containment of conditioned sludge on
the plates and subsequent removal of dewatered
sludge cake.
A filtrate collection/drainage system between
media plates and the underlying concrete slab.
the
Media plates, sealed around the edges to adjacent
plates and to the walls of the concrete containment
structure.
Polymer feed and mixing systems to introduce
dilute polymer into the sludge feed stream, and to
provide proper flocculation.
Sludge distribution piping, usually located on the
walls of the containment structure.
An air-tight filtrate sump adjacent to the bed and
connected to the filtrate collection/drainage system.
Float operated filtrate pumps located in the sump
to convey collected filtrate to the treatment plant
headworks.
A vacuum system connected to the filtrate sump to
induce a partial vacuum between the
underdrainage system and the layer of conditioned
sludge on top of the media plates,
A source of high pressure, 480 to 830 kPa (70 to
120 psi), clean (no particulates) wash water for
cleaning the surface of the media plates after
removal of the dewatered sludge cake.
Drains for the collection of the media plate
washwater. If the,drains are inside the beds they
must have caps or seals to insure no leakage of
applied sludge.
A conveyance system to pump this washwater
back to the treatment plant. With proper valving the
washwater can flow to the vacuum filtrate sump. A
separate washwater sump is sometimes used, and
on occasion direct gravity flow to the headworks is
possible.
A control panel to operate all mechanical and
electrical components of the system. These are
typically designed to allow either manual or
automatic sequencing of the operational cycle.
In most cases an enclosure covering all
mechanical and electrical components in the
system, including the filtrate pump, the vacuum
system, the polymer system, and the control panel.
If year-round operation in cold climates, or
operation during rainfall is necessary, an enclosure
for the whole facility may be required, with heat
addition during freezing weather.
66
-------
• A front-end loader to allow for mechanical
removal of the sludge cake from the bed.
6.4,3 Performance Expectations
There are reliable operating records for over 20
installations in the United States. These systems
have dewatered a variety of sludges, including:
Waste Extended Aeration --
Thickened and
Unthickened
Aerobically Digested Activated -- Thickened and
Unthickened
Anaerobicaliy Digested --
Thickened and
Unthickened
• Lime Conditioned Primary and Waste Activated
• Imhoff Tank.
Table 6-3 summarizes data from a recent U.S. EPA
Technology Assessment (16) which examined
performance at 13 operating systems. The cycle time
in all cases was 24 hours, including cake removal and
plate washing. The median polymer (liquid emulsion
type) dose was 9 kg/Mg (20 Ib/ton) dry solids.
Table 6-3. Performance Observed at 13 Vacuum Assisted
Drying Beds
Relative Loading
Low8
Median"
High"
Solids
Loading
kg/m2/cycle
3.18
9.18
37.84
Total Solids
Sludge Feed Sludge Cake
percent percent
0.8 9-12
1.5-3.0 14-18
8-10 30-35
kg/m2/eyc!e x 0.2048 = !b«t2/cycle.
a Unthickened oxidation ditch sludge.
b Typical of aerobically digested WAS, thickened by decanting.
c Lime conditioned mixture of primary and thickened WAS.
Vacuum assisted drying beds are a proven
technology. Section 6.4.5 compares the costs for
vacuum assisted beds to conventional sand drying
beds. It must be recognized that the end product from
the two concepts will not usually be the same. The
sand bed will typically produce a sludge cake
exceeding 30 percent solids while the vacuum bed is
typically operated to produce a liftable (12 to 15
percent solids) cake. If regulatory agencies require a
higher solids concentration for disposal, an additional
supplemental drying area may be needed for the
vacuum assisted systems.
Practically all of the problems experienced with early
systems (prior to 1984) have been solved by the
manufacturers. Ongoing research and development
has focused on prolonging the service life of the
media plates and reducing their cost. The major
market to date has been treatment systems with flows
less than 7,500 m3/d (2 mgd), however, the size of
facilities giving serious consideration to this concept
is gradually increasing.
In many cases it is possible to add multiple layers of
polymer treated sludge, with decantation of
supernatant between each. The bed can also be filled
with water to the top of the media plates prior to
sludge application, as described in Section 6-5. This
will increase the gravity drainage rate and, in this
case, the vacuum is only used in the final stage.
Combining these techniques can very significantly
increase bed capacity.
6.4.4 Operation and Maintenance
The typical operational cycle includes the following
activities:
• At the start of the cycle the vacuum pumps are off,
the bed closure system is in place, the filtrate
pumps are on automatic, sufficient polymer is
mixed or otherwise available, and drains in the bed
(if present) are covered and sealed. The filtrate
valve may be open or closed depending on the
mode of operation. The media surface may be dry,
wet, or covered with a thin film of standing water
(< 1.5 cm), depending on the particular system.
* Valves on the sludge feed line are opened, with the
polymer feed pump also operational.
• If the cycle was started with the filtrate valve
closed, it may be opened whenever the media
plates become entirely covered with well
flocculated sludge. Opening this valve allows
gravity drainage ,to begin. If a good separation of
sludge solids and supernatant occurs, it is also
possible to decant the supernatant prior to opening
the filtrate valve, as described in the previous
section.
• When the desired volume of sludge has been
applied to the bed, the polymer and sludge feed
pumps are stopped.
• Gravity drainage is allowed to continue until the
operator decides that the rate of filtrate collection is
too slow. The time for this gravity drainage may
range from 30 minutes to several hours after
sludge application is completed.
• The operator starts the vacuum cycle at the end of
gravity drainage. The vacuum sequence usually
proceeds in discrete steps, beginning at 5 to 8 cm
(2-3 in) Hg for about 1 hour, then increasing to 13
to 15 cm (5-6 in) Hg for another hour with a final
step at 25 to 30 cm (10-12 in) Hg. This highest
67
-------
vacuum level normally continues until the sludge
cake has dried sufficiently to crack and at this point
the system vacuum level is lost. Figure 6-9
illustrates a "cracked" bed with the sludge ready
for removal. Exact values and times vary from
system to system.
Figure 6-9. "Cracked" bed, ready for sludge removal.
Tr~5~'
• An optional evaporative phase may be necessary
to produce a liftable sludge cake. The minimum
solids concentration to achieve this condition is
about 10 to 12 percent for most sludges, but may
be lower for certain others. The time required for
evaporation is variable and can only be determined
on a site specific basis.
• The bed closure system, if used, is removed to
allow access for sludge removal. Typically a small
tractor with a front loading bucket is used at all but
the very smallest installations.
• The front-end loader cannot completely remove
all of the sludge cake. The small amounts left on
the bed must be removed manually with a shovel
or a scoop. Diligent removal of this material is
necessary to permit an optimum final cleaning of
the media plates.
• The manual rinsing with hose and nozzle
commences at the end of the bed furthest from the
drains and progresses toward the drains. This
media plate cleaning shares equally with the
polymer conditioning as the most critical aspect of
system operation. The plates must be scrupulously
cleaned between each cycle of operation if
progressive loss of plate permeability is to be
avoided. This cleaning completes the operational
cycle and the bed is ready for another charge of
sludge.
Selection of polymer, the aging time, and the
effectiveness of mixing and dosage control are
variables subject to various degrees of operator
control, and all of these factors strongly affect
performance. The high cost of polymers makes
overdosing a very expensive activity. In addition,
overdosing may lead to progressive plate clogging
and the need for special cleaning procedures to
regain plate permeability.
Plate cleaning is critically important. If not performed
regularly and properly, the media plates are certain to
clog. The design should incorporate proper sizing and
location of drains, sufficient water pressure, and
selection of a satisfactory hose nozzle. The media
plates will in time show some sign of decreased
permeability, even with good maintenance, due to
accumulation of oils and greases or other substances.
Special cleaning measures are then required. Some
of those used successfully include:
• High pressure, hot water cleaning
* Commercial grade hydrochloric acid at about 1
percent concentration
• Tri-sodium phosphate at about 0.25 to 0.5
percent
• Calcium or sodium hypochlorite at about 1 percent
available chlorine
• Enzyme based cleaners.
6.4.5 Costs
Since vacuum assisted drying is a proprietary
concept, the capital cost will vary with the
manufacturer. The high rate operational cycle allows
a compact system; land costs are not a significant
factor. The manufacturer provides the media plates,
pumps, polymer makeup/feed system, and system
control panel at a typical cost of $640-$860/m2
($60-$80/ft2) of bed surface area. The remaining
capital costs are for concrete work, pipes and valves,
electrical wiring, a control building for uncovered beds
or a complete enclosure for the entire system if
required, and a front-end loader for sludge cake
removal. The range of total capital costs derived from
a review of 29 systems are summarized in Table 6-4
(16).
Vacuum assisted drying beds (VADB) are normally
compared to conventional sand drying beds.
However, the sludge cake removed from the VADB is
typically never as dry as that removed from sand
beds and this difference must be recognized in any
cost comparison. In some cases additional drying
68
-------
Table 6-4. Capital Costs for Vacuum Assisted Drying Beds
(1984 $)
Capital Cosl, $/m2
Range
Low
Median
Average
High
Uncovered
Beds
484
1,485
1,323
1,991
Covered Beds
753
1,614
1,679
2,292
$/m2 x 0,0939 = $/ft2.
may be required prior to final disposal of the VADB
sludge cake and that is not included in the
comparisons in this section. As described in Section
6.2, the solids loading on sand beds is dependent on
sludge characteristics, geographic location, and
whether the beds are covered or not. Tables 6-5
and 6-6 (16) compare the relative capital and 0 & M
costs for these two concepts.
Table 6-5. Cost Comparison of Vacuum Assisted Drying
Beds vs. Sand Drying Beds
System Type Average Capital Cost
$/m2
Vacuum Assisted Beds
Uncovered
Covered (roo( only)
Enclosed in a building
Sand Drying Beds
Uncovered
Covered (roof only)
Enclosed in a building
1,291
1,475
1,678
121
625
1,000
$/mz x 0.0929 = $/sq ft.
Table 6-6.
Hem
O & M Cost Comparison of Vacuum Assisted
Drying Beds vs. Sand Drying Beds
O & M Cost, $/Mg
Sand Beds
Vacuum Beds
Labor
Polymer
Electricity
Front End Loader
Sand Replacement
Media Plate Cleaning
Media Plate Replacement
Total
72
-
-
1
7
-
-
80
39
29
1
1
-
1
6
77
$/Mg x 0.896 = $/ton.
The operational life of the systems compared in Table
6-5 is 20 years, with an assumed replacement of
media plates after 10 years (possibly a very
conservative assumption). The cost values in Table
6-5 do not include land costs which can be a
significant factor for conventional sand beds. The land
costs for possible additional drying of VADB sludge
cake are also not shown.
The 0 & M costs in Table 6-6 are based on fuel and
maintenance for the front-end loader, and assume
an annual replacement of about 8 cm of sand on the
sand beds and a 6-month chemical cleaning cycle
for the vacuum bed media plates. The calculations
assumed a 907 kg/d (2,000 Ib/d) production of
aerobically digested sludge, using 196 m2 (2,112 ft2)
for the vacuum assisted bed and 1,859 m2 (20,000
ft2) for the sand bed.
The cost effectiveness of these two technologies
depends strongly on local climatic conditions, a
critical factor in determining the loading rates on sand
beds. The following relationships were derived from
the information in Tables 6-5 and 6-6 and the
related assumptions.
• An uncovered vacuum assisted bed will be more
cost effective if the solids loading on an uncovered
sand bed is less than 146 to 171 kg/m2/yr (30-35
Ib/ft2/yr).
» A vacuum system with a roof will be more cost
effective if the solids loading on a roofed sand bed
is less than 317 to 342 kg/m2/yr (65-70 Ib/ft2/yr).
• A vacuum system in a building will be more cost
effective if the solids loading on a completely
enclosed sand bed is limited to 440 to 464
kg/m2/yr (90-95 Ib/ft2/yr).
» If the allowable solids loading on the sand bed
exceeds the values given in the three categories
above, then the conventional sand bed will be the
more cost effective alternative.
• The solids content of the final sludge cake is also
an important consideration. The vacuum assisted
beds typically produce a liftable sludge (10-12
percent solids) while sand beds can achieve 50
percent solids or more. The hauling distance for
final disposal and/or the necessary final solids
concentration may require additional drying for the
vacuum assisted product, thus increasing the total
costs for this process.
• The use of polymers with sand beds, or the
combined use of freeze thaw dewatering and
polymer dewatering (as described in Section 6.3),
should make uncovered sand beds competitive,
even in colder climates.
69
-------
6.5 Wedgewire Beds
The Wedgewire, or wedgewater, process is physically
similar to the vacuum assisted systems described in
the previous section. The media in this case consists
of a septum with wedge shaped slots about 0.25 mm
(0.01 in wide). This septum serves to support the
sludge cake and allow drainage through the slots.
Rgure 6-10 illustrates the process.
Flguro 6-10 Cross section of a wedge wire drying bed.
Controlled Differential Head in Vent
by Restricting Rate of Drainage,
Vent
. Partition to Form Vent
Wedgewire Septum
Outlet Valve to Control
Rate of Drainage.
Initially, water enters the bed from beneath, and fills
the bed to a depth of about 1 cm (0.4 in) above the
media surface. Polymer conditioned sludge is then
applied to the bed. During the initial phase the drain
valve is closed so that the water and sludge stand on
the bed. The valve is then partially opened to control
the drainage rate for up to 2 hours. Following this
controlled phase, the valve is opened fully and the
sludge cake allowed to drain naturally. The initial
static period with sludge on the flooded bed allows
the sludge to settle to the media surface and form a
filter zone. In addition, this establishes the potential
for saturated flow conditions through the sludge and
media. Drainage will proceed at a much higher rate
under "saturated" conditions (devoid of air, so a small
hydrostatic suction is exerted on the bed) compared
to wet sludge resting on a dry surface.
6.5.1 Design Considerations
Since wedgewire systems are proprietary devices,
loading criteria are developed in conjunction with the
manufacturer. Bench scale or pilot units can be used
to develop loading criteria and polymer dose for a
particular sludge.
Typical sludge solids loadings range from 2 to 5
kg/m2 (0.4 to 1 Ib/ft2) per operational cycle. The
number of operational cycles per year will vary,
depending on the type of system and other local
conditions. On a routine, 24-hour operational cycle
the annual loading could exceed 1,600 kg/m2 (328
Ib/ft2), including allowances for maintenance
downtime. This annual loading exceeds the loading
on conventional sand beds by an order of magnitude.
An enclosed and possibly heated facility would be
needed to maintain such production in cold climates
with extended periods of freezing weather. The
wedgewire process appears best suited for smaller
treatment systems, in locations with moderate
climates, and where land area may be limited.
In many operational systems the sludge cake is
removed soon after completion of the drainage phase
to maintain high production rates. Typically, the
sludge cake at this point will be 8 to 12 percent solids
(after 24 hours) and is handlable, but still wet.
Production of a drier sludge would require more time
on the bed, or removal to a stockpile area for
evaporative drying.
6.5.2 Structural Elements
The media was originally constructed of stainless
steel but is now predominantly made of preformed
polyurethane modules. The stainless steel requires
additional support in the bed. The interlocking
polyurethane modules are self-supporting and also
create a shallow drainage plenum beneath the media
surface. Either type of media can support small
front-end loaders for removal of the sludge cake.
Small systems have also used large tilting metal
trays. In this case, when the sludge is ready for
removal the whole bed is tilted to a steep angle and
the sludge cake slides out.
Existing sand drying beds can be retrofitted for the
wedgewire process, or a new concrete basin can be
constructed.
6,5.3 Performance Expectations
According to manufacturers of wedgewire systems,
polymer treated aerobically digested sludges can be
dewatered to 8 to 12 percent solids within 24 hours
and treated anaerobically digested sludges can be
dewatered to 16 to 20 percent in the same time
period. Polymer conditioning is necessary for most
sludges and desirable for all. Without conditioning, the
fines, particularly in aerobically digested sludges, may
penetrate the media and either be lost with the filtrate
or accumulate in the drainage plenum. To avoid
solids build-up in the plenum, the floor of new
concrete beds should be sloped to insure positive
drainage.
6.5.4 Operation and Maintenance
The basic operation and maintenance requirements
for wedgewire systems are similar to those described
in Section 6.4 for vacuum assisted beds. Surface
clogging is less likely with the wedgewire process but
is still possible if routine cleaning is not performed
properly.
Polymer conditioning is critical for successful
performance. The polymer dosages required are
70
-------
similar to those used with the vacuum assisted beds
described in Section 6-4. The typical sludge depth
for a single application ranges from 10 to 25 cm (4-
10 in). The optimum for a particular system will be
determined with operational experience. In some
cases, as also described in Section 6.4, it may be
possible to apply multiple sequential layers with
decantation of the supernatant prior to starting the
drainage phase.
There have been reports of damage to the plastic
surfaces when front-end loaders have been
improperly used to remove the sludge cake. The
proper procedure requires driving straight in and
backing straight out. Sharp, skidding turns can cause
structural damage to the molded polyurethane
surfaces.
It is important for the operator to carefully manage the
initial controlled drainage rate to insure maximum
water flow during this phase. If the rate is too slow
the total cycle time will have to be increased, and if
the rate is too high, complete drainage may not
occur. The manufacturer's drainage
recommendations can be used initially and then
modified as necessary with operational experience.
6.5.5 Costs
Land is not usually a significant factor in the capital
costs of wedgewire systems, unless additional land
area is needed for further drying of the sludge cake.
The other construction costs will depend on whether
existing sand beds can be retrofitted instead of
constructing entirely new basins. Construction costs
in the latter case might range from $1,000 to
$1,900/m2 ($93 to $177/ft2). Operating costs may
be slightly less than the vacuum assisted systems
described in Section 6.4, but will still fall in the same
range. The comparisons in Section 6.4.5 to sand
beds should also be approximately applicable to
wedgewire systems.
6.6 Sludge Lagoons
A distinction must be made between sludge drying
lagoons and sludge lagoons primarily intended for
storage. Some drying occurs in storage lagoons but
the primary intent is to provide temporary or semi-
permanent storage.
Drying lagoons are operated on a regular cycle to
dewater sludges. A typical operational cycle includes
the following activities:
• Well stabilized liquid sludge is pumped into the
lagoon, over a period of several months or more.
• Supernatant is decanted, either continuously or
intermittently, from the lagoon surface and returned
to the treatment plant.
» Filling and decanting operations are continued until
the design depth of sludge is reached.
• The surface crust is repeatedly broken up and/or
removed during the drying period.
• Dewatered sludge is removed with some type of
mechanical removal equipment.
• Maintenance and repair is performed while the
lagoon is empty and then the filling cycle is
repeated.
The complete cycle for a single lagoon typically takes
from less than 1 to 3 years, depending on the final
solids concentration required, local climate, the depth
of sludge applied, and management practices (17). Ail
sludge should be stabilized prior to addition to the
lagoon to minimize odor problems. Occasional odors,
flies and mosquitos may still be a problem, so a
remote site is essential.
6.6.1 Design Considerations
Until recently, sludge lagoons were often located in
soils with at least moderate permeability to take
advantage of subsurface drainage and percolation.
That practice is now the exception rather than the
rule in most of the United States due to more
stringent environmental and groundwater protection
regulations. If a groundwater aquifer with drinking
water potential exists beneath the site, it may be
necessary to line the lagoon or otherwise restrict
significant percolation. Unless a sand bottom and
underdrains are then installed, the only sludge
dewatering mechanisms left are decanting
supernatant and evaporation.
In effect, the sludge drying lagoon is similar in
concept to a deep sand drying bed with restricted
drainage. The depth of sludge in the lagoon might be
0.7 to 1.4 m (24 to 48 in) as compared to 0.3 m (12
in) for the sand bed. The recommended solids
loading for the drying lagoons is 36 to 39 kg/yr/m3 of
lagoon capacity (2.2 to 2.4 Ib/ft3/yr). A minimum of
two cells is essential, even at very small systems, to
insure availability of storage space during cleaning,
maintenance or emergency conditions.
Evaporation and decantation are usually the dominant
pathways for water even if an underdrainage network
exists. The required lagoon surface area depends on
the temperature, precipitation, and evaporation rates
for the local area. Equations 6-8 to 6-12 in Section
6.7 can be used to estimate surface area
requirements, or assuming that standing water is
routinely decanted, the design calculations for
evaporation are similar to Equations 6-1 to 6-4.
The evaporation procedures in Reference 18 to
complete retention ponds can also be used. The
water to be removed from the sludge lagoons is the
required portion of the sludge moisture content plus
71
-------
that portion of precipitation that will infiltrate the
sludge mass rather than be removed as supernatant.
The dependence on evaporation tends to favor arid
and somi-arid climates for this dewatering process.
However, the Metropolitan Sanitary District of Greater
Chicago, the Milwaukee Metro Sewerage Authority,
and the City of Philadelphia have all successfully
operated large scale sludge drying lagoons in cool
humid climates (19).
It is possible to facilitate drying with a device that
consists of a tractor with a helical screw in front to
push sludge aside and mix it. This helps to open up
the dried top layer and expose the wet material
below.
6.6.2 Structural Elements
The retaining walls for drying lagoons are typically
earthen dikes 0.7 to 1.4 m (2 to 4 ft) high with a side
slope of 1:3. The lagoon is typically rectangular in
shape to facilitate sludge removal. Required
equipment includes: sludge feed lines and pumps,
supernatant decant lines, and sludge removal
equipment. The last can include trucks, front-end
loaders, bulldozers, or draglines, depending on the
size of the operation.
6.6.3 Performance Expectations
Solids concentrations in the range of 15 to 40 percent
are expected in the sludge removed from the lagoon;
concentrations can be higher in arid climates. These
lagoons share a common problem with other air
drying processes in that a surface crust forms early in
the evaporative stage, which then restricts further
evaporative water losses. This problem is minimized
with the paved drying beds described in Section 6.8
that use mechanical equipment to move around the
bed to turn and mix the sludge. Similar equipment
and procedures can be used in drying lagoons if the
depth of sludge permits. Floating devices can also be
used. Larger scale facilities have used a cable and
scraper system as shown in Figure 6-11.
6.6.4 Operation and Maintenance
The routine operational activities consist of sequential
sludge applications and decantations until the lagoon
contains the design volume of sludge. The periodic
break-up or removal of the surface crust then
insures continued evaporation. Sludge removal is
labor intensive but occurs infrequently. Maintenance
activities include care of equipment and dikes and
control of dike vegetation. Some sludge drying
lagoons may require insect and odor control. The
labor requirements for sludge drying lagoons are
shown in Figure 6-12.
6.6.5 Costs
The capital cost for drying lagoons is significantly
influenced by the cost of land at the project site.
Other major factors include construction of the dikes,
sealing the bottom (if required), underdrainage (if
used), and the other structural elements described in
Section 6.6.2. The construction costs for the lagoon
(with earthen dikes) are similar to the costs for sludge
storage lagoons, or wastewater treatment ponds.
Appendix A-32 in Reference 8 can be used to
estimate these costs. The other capital costs depend
on the intended methods for sludge loading and
removal, and should be determined on a case-by-
case basis. The major 0 & M costs are for labor, fuel,
and maintenance of sludge removal equipment.
Figure 6-12 (20) can be used with prevailing wage
rates to estimate labor costs. The remaining O & M
costs will depend on the equipment and procedures
used and must also be determined on a case-by-
case basis.
6.7 Paved Beds
Until recently, paved beds used an asphalt or
concrete pavement on top of a porous gravel
subbase. Unpaved areas, constructed as sand drains,
were placed around the perimeter or along the center
of the bed to collect and convey drainage water. The
main advantage of this approach was the ability to
use relatively heavy equipment for sludge removal.
Experience showed that the pavement inhibited
drainage, so the total bed area had to be greater than
that of conventional sand beds to achieve the same
results in the same time period.
Recent improvements to the paved bed process
utilize a tractor-mounted horizontal auger, or other
device, to regularly mix and aerate the sludge (21).
This mixing and aeration breaks up the surface crust
that inhibits evaporation, allowing more rapid
dewatering than conventional sand beds. Some of the
equipment was originally developed for composting
operations but serves equally well for paved bed
dewatering. Underdrained beds are still used in some
locations, but the most cost effective approach in
suitable climates is to construct a low cost
impermeable paved bed and depend on decantation
of supernatant and auger/aeration mixing for
evaporation to reach the necessary dewatering level.
Figure 6-13 shows a bed of this type.
6.7.1 Design Considerations
The critical design parameter for paved beds, as with
sand beds and drying lagoons, is the surface area
required to dewater the sludge to the specified solids
level in the specified time. Since drainage is not a
factor in many modern paved bed designs, the only
ways water can be removed is through decantation
and evaporation. These water losses will depend on
the same factors described in Section 6.2, but with
paved beds the use of the mechanical auger/aerator
sustains evaporation near the maximum potential for
sludge. Paved beds can be used in any location, but
since evaporation provides the major pathway for
water loss, they work best in warm, arid and semi-
72
-------
Figure 6-11 Cable and scraper system for sludge drying lagoons.
• Track Cable
Inhaul (Drag) Cable
Crane
Tail
Anchorage
(Bulldozer)
Lagoon Bottom
Figure 6-12. Labor requirements for sludge drying lagoons.
1,200
C 800
o
X!
400
10
20
30
40
50
Solids Loading, tons/yr x 1,000
arid climates. Assuming the same degree of effort
with the auger/ aerator, the design solids loading on a
bed, or the bed area will be directly related to the
potential evaporation, and precipitation in the local
area. The design loading rate for the system in
Roswell, NM, is 244 kg/m2/yr (50 Ib/ft2/yr); the
loading during a pilot test in Wichita, KS, was 127
kg/m2/yr (26 Ib/ft^/yr). In more humid climates the
allowable loading might be even lower (3).
As shown in Figure 6-13, these completely paved
beds incorporate devices to draw off the supernatant,
and with some sludges it may be possible to draw off
20 to 30 percent of the water in this manner. If the
sludge has particularly good settling characteristics, it
may be possible to use several fill and decant cycles
prior to the evaporative stage. The rate of evaporation
for a particular site can be determined with small
scale pilot studies or assumed as a fraction of the
pan evaporation rate for water in the local area
(usually routinely available). A study in New Mexico
(3) indicated that the evaporation rate from mixed and
aerated sludge was about 58.7 percent of the free
water pan evaporation for the site. That relationship
should be generally valid for other locations also. At
large scale projects, where land costs can be very
significant, a pilot test to determine this ratio should
be used to optimize the design.
The water losses and bed area required for a paved
bed system can be determined with Equations 6-8
to 6-12.
1 -
(6-8)
where,
W0 = total water content in applied sludge,
kg/yr (Ib/yr)
1.04 = assumed specific gravity of sludge solids
S = annual sludge production, dry solids, kg
Ob)
so = dry solids in applied sludge, percent as
decimal
The water content after decantation is given by:
r 1 -s
WD=
I s,
(6-9)
where,
WQ =, total water remaining after decantation,
kg/yr (Ib/yr)
73
-------
Figure 6-13 Paved sludge drying bed designed for decantation and evaporation.
Sludge Layer, 30 cm typ.
3-4 m typ.
15
76 cm Minimum
Total Bed Depth
Slope 0.2 - 0.3%
Sludge Level
-Soil Cement 20 cm
Cross Section View
sd = dry solids in sludge after decantation,
percent as decimal
The water content to be removed by evaporation is
given by:
In US units, the final term becomes: (P)(A)(62.4). If
the system does not allow decantation, use WQ in
equation 6-10 instead of WQ.
The evaporation rate for a given location is given by:
1-s
(P)(A)(1,000)
(6-10)
where,
WE = water to be evaporated after decantation,
kg/yr (ib/yr)
Se = dry solids required after evaporation,
percent as decimal
P = annual precipitation, m (ft)
A = bed area, m2 (sq ft)
(6-11)
where,
Re = evaporation potential for sludge on a
mixed and aerated paved bed, kg/m2/yr
(Ib/sq ft/yr)
ke = reduction factor for sludge evaporation vs.
a free water surface, percent as decimal
= 0.6 (pilot test to determine this value is
recommended for large projects)
En = free water pan evaporation rate, cm/yr
(ft/yr)
In US units, the 10 becomes 62.4.
74
-------
The area required for a paved drying bed system can
be estimated by combining Equations 6-10 and 6-
11.
(PXAK1000)
A =
R
where,
(6-12)
A = paved bed area, m2 (sq ft, use 62.4
instead of 1,000)
The example below demonstrates the application of
this procedure.
Assume:
S = 365,000 kg/yr total dry solids in sludge
produced
Scj = 15 percent
se = 35%
P = 0.5 m/yr
Ep = 127cm
Use equation 6-11 to determine evaporation rate:
Re = (10)(0.587)(127) = 745 kg/m%r
Use equation 6-12 to determine the total area
required:
A = [(1.04) (365,000) (5.67 - 1.86)
+ (0.5) (A) (1,000)] -r 745.5
A = 1,940 + 0.67 (A)
A (1 - 0.67) = 1,940
A = 5,891 m2 (0.5891 ha, 1.46 ac)
The solids loading for this case would be 62 kg/m2yr.
The total design area should be divided into at least
three beds for all but the smallest operation to provide
operational flexibility. A detailed month-by-month
analysis of weather records and expected sludge
production rates will determine the optimum number
of beds required. It may not be necessary to use all
of the beds in the hot dry months. For example, the
system designed for Roswell, NM has a total of seven
beds, six of which need to be used in December; only
three are required in June due to increased
evaporation and decreased sludge production. To
insure a conservative design, the equations and the
example presented above assume that all
precipitation that falls on the bed must be removed by
evaporation. For small systems, it may be
advantageous to plan the orientation of the bed for
reception of maximum solar radiation.
6.7.2 Structural Elements
Paved beds have been constructed with concrete and
asphalt pavement, with and without drains. However,
the most economical approach may be to use soil
cement as the paved surface as shown in Figure 6-
13. Other structural features are also shown in the
same figure. Information on construction of soil
cement pavements can be obtained from the Portland
Cement Association.
A long rectangular configuraton improves efficiency
by reducing the time required for turning the
auger/aerator vehicle. A variety of inlet and
decantation structures are also possible. The
minimum total depth of the bed is about 0.8 m (2.6 ft)
to provide some freeboard above the typical 30-cm
(12-in) sludge layer. In some systems up to 1 m (3
ft) of liquid sludge is applied in the initial layer and the
freeboard must be correspondingly increased.
Other major system components include the sludge
and decantation piping, and the auger/aerator
vehicles. A variety of vehicle sizes and configurations
are available and the designer should seek the
assistance of the manufacturers in determining the
optimum sizes and number for a particular operation.
6.7.3 Performance Expectations
The use of digested, or otherwise stabilized, sludge is
necessary to avoid odor complaints and to satisfy
regulatory requirements for final sludge disposal. The
decantation phase might require 2 to 3 days for
sludge settling and 1 to 2 days to decant each
increment of sludge added, depending on the sludge
characteristics. If drainage is allowed by the design, it
should also be essentially complete during the time
allowed for sludge settling and decantation.
The final evaporative drying period will depend on the
climatic conditions occurring after the sludge is
applied and on the regular use of the auger/aerator
equipment. Solids in the range of 40 to 50 percent
can be achieved in 30 to 40 days in an arid climate,
for a 30-cm (12-in) sludge layer, depending on the
time of year and the effectiveness of decantation (3).
A 1-m (3-ft) sludge layer in the same climate might
require 100 to 250 days to reach 50 percent solids,
depending on when the sludge was applied.
6.7.4 Operation and Maintenance
The major operational tasks are sludge application,
decantation, mixing and aeration, and sludge removal.
Depending on the size of the operation and the time
of year, the sludge on the bed should be mixed
several times a week to maintain optimum
evaporation conditions. Labor requirements at the
75
-------
Roswell, MM system are estimated to be about 0.3
hr/yr per Mg of dry solids processed (3). Maintenance
requirements include routine care of the
auger/aeration equipment, the sludge pumping and
piping network, the decantation piping, and the bed
and dikes. If the site experiences freezing weather in
the winter months, the valves and pumps in the
system need to be protected and checked periodically
during the critical freezing periods.
6.7.5 Costs
Capital costs are strongly influenced by the cost of
land at the project site. Other major capital costs
include the containing walls and pavement,
application and decantation piping (and drainage
piping if used), the auger/aerator, and the sludge
removal equipment. In many cases, the same vehicle
can be used for both tasks. The major O & M costs
are labor and fuel for the equipment. Table 6-7 (3)
compares the cost of a paved bed operation to
conventional sand beds in the same location.
Tablo 6-7. Estimated Cost Comparison of Paved Beds vs.
Conventional Sand Drying Beds
Paved Bed
Horn Sand Drying Bed vy/Augec/Aerator
Number of Beds
Tolal Aroa, nfl
Solids Loading, kg/mz/yr
Labor, hr/yr
Capital Costs, $
O & M Costs, $/yr
Total Present Worth, $
16
60,600
108
8,580
1,465,000
100,000
2,500,000
7
26,200
243
1,700
520,000
25,000
780,000
6.8 Other Innovative Processes
A number of other processes, which do not fit directly
into the categories previously discussed, have been
proposed and tested at both the pilot and full-scale
level. In some cases the distinctions are minor. The
"Solar" sludge drying beds used in the arid
southwestern United States differ little in concept
from drying lagoons discussed in Section 6.6 or the
paved beds described in Section 6.7. In all cases the
key to success is the mixing and turning of the sludge
and break-up of the surface crust. The evaporative
losses will be very rapid in arid climates. There are
two other processes that are unique; one incorporates
additives of various types and the other utilizes an
underdrained sand bed and growing reeds or
bulrushes to dewater the sludge.
6.8.1 Additives
In some cases additives such as sawdust, wood
chips, etc., are mixed with the sludge as a
preparatory step for composting or vermistabilization.
These processes are usually described as
stabilization processes but a significant degree of
dewatering does occur with both.
In a composting process, the initial sludge solids
concentration might be about 20 percent. Wood
chips, sawdust or some other bulking agent is added
so the solids content of the mixture is between about
40 and 45 percent. The solids content of the compost
product following composting might approach 50 to
60 percent (22). There is considerable moisture loss
to the atmpshere. Most of the energy for evaporating
water comes from the aerobic oxidation of the sludge,
but contact with the air by forced aeration or turning
of windrows is needed to carry the moisture away.
Sawdust is also used as a bulking agent in a process
that uses earthworms for sludge stabilization and
dewatering (vermistabilization). In this case, thickened
sludge (3-4 percent solids) is sprayed onto beds
containing sawdust and earthworms. Typical sludge
loading rates are 85-90 kg/m2/yr (17-18 Ib/ft2/yr),
which is near the low end of the range for
conventional sand beds.
After 6 to 12 months the mixture of earthworms,
castings, and sawdust is removed and the
earthworms separated (by screening) for use in the
next cycle. The solids concentration of the stabilized
material ranges from 15 to 25 percent. This process
has been demonstrated at the pilot scale level at
Lufkin, TX, and elsewhere (23). A heated enclosure
over the beds would be required in all but the
warmest climates to sustain activity during the winter
months. Laboratory scale experiments at Cornell
University successfully stabilized and dewatered raw
sludges with this process (24). A large scale version
of this process might be a cost-effective alternative
since thickening, digestion, conditioning, and
dewatering are all eliminated.
In the Pacific Northwest, sawdust has also been used
as a bulking agent for sludge drying in a process
similar to a paved bed operation. The use of sawdust
or other agents will only be economical when the
materials are locally available at little or no cost.
The reed bed process combines the elements of an
underdrained sand bed with a dense stand of
vegetation to obtain sludge dewatering. Most of the
operational reed beds have been planted with the
common reed Phragmites but rushes or cattails could
also be used.
The bed is actually constructed as a deep trench and
lined to prevent exfiltration. A 25-cm (10-in) layer
of washed gravel encloses the underdrain pipe, and is
overlain by a 10-cm (4-in) layer of sand. The root
stock of the reeds is planted on 30-cm (12-in)
centers on the gravel layer, at a depth of about 10 to
15 cm (4-6 in). The bed is flooded with water to a
depth of about 10 cm (4 in) for several weeks to
encourage plant development. The freeboard above
the sand layer is at least 1 meter to provide for long-
term sludge storage. Sludge is not applied until the
plants are well established.
76
-------
The vegetation plays an essential role in the
dewatering process. The root system absorbs water
which is then lost to the atmosphere via
evapotranspiration. More importantly, the penetration
of the plant stems and the root system maintains a
permanent pathway for continuous drainage of water
from the sludge layer. Reeds and similar plants have
the capacity to transmit oxygen from the leaves to the
roots so there are aerobic microsites (adjacent to the
roots) in an otherwise anaerobic environment that
assist in stabilization and mineralization of the sludge.
Stabilized, thickened sludge, at about 3 to 4 percent
solids is applied in 10-cm (4-in) layers. Solids
concentrations less than 3 percent are not cost
effective for optimum bed use and concentrations
above 4 percent wiil not flow properly and will not be
uniformly distributed within the dense vegetation. A
layer can be applied about every 20 days with warm,
dry weather conditions.
An operating system in Washington Township, NJ
(25) was designed for an annual loading of 3,5 m
(11.5 ft) of aerobically digested sludge at 3 percent
solids, solids loading of 100 kg/m2/yr (20 Ib/sq ft/yr).
The average loading on 16 operational systems in
New Jersey, New York, and North Carolina is about
81 kg/m2/yr (17 Ib/sq ft/yr) which is at the low end of
the range for conventional sand beds. The final
annual layer of dried and further stabilized sludge (at
about 90 percent solids) will be about 10 cm (4 in)
deep and this residue can be left in place. A 10-year
operational cycle has been planned for several
systems in New Jersey (26). At the end of this period
the accumulated sludge and the sand layer are
removed. A new layer of sand is installed and new
vegetation planted if necessary. An annual harvest of
the vegetation is recommended, when the plant is
dormant but before the leaves have been shed. The
harvested material can be burned, composted, or
otherwise disposed of.
These systems have been successfully operated in
New Jersey on a year-round basis, with only 20 to
30 days downtime for adverse weather conditions.
Since the dewatering benefits will be minimal during
the dormant season for the plants and during
prolonged freezing weather, it is likely that a longer
downtime will be required for locations with more
severe winters than New Jersey.
Multiple beds are required for every installation. With
a 10-year cycle, a minimum of 12 beds would be
necessary to allow for one out of service each year
and one for emergencies. When a bed is to be
cleaned, sludge applications are stopped for that bed
in early spring, the vegetation is harvested in early
fall, and the sludge residue and sand are removed by
early winter. The dried sludge removed from the bed
is similar in character to composted sludge with
respect to pathogen content and stabilization of
organics, due to the long detention times combined
with the final 6-month rest period.
The major advantage of the reed bed process is the
infrequent need for sludge removal and bed cleaning.
The major disadvantage is the need for the annual
vegetation harvest. However, the total volume of
harvested vegetation and sludge residue on a 10-
year operational cycle is still less than the sludge
cake volume requiring disposal if the same amount of
sludge were dried on a conventional sand bed.
6.9 References
When an NTIS number is cited in a reference, that
reference is available from:
National Technical Information Service
5825 Port Royal Road
Springfield, VA 22161
(703) 487-4650
1. Manual of Practice 20 - Sludge Dewatering.
Water Pollution Control Federation, Washington,
DC, pp. 19-44, 1983.
2. Imhoff, K. and G.M. Fair. Sewage Treatment.
John Wiley & Sons Inc., New York, NY, 1956.
3. Innovative Sludge Drying Study, City of Roswell,
New Mexico. EPA Project Report C-35-1052-
01, U.S. Environmental Protection Agency,
Region VI, Dallas, TX, 1985.
4. Haseltine, T.R. Measurement of Sludge Drying
Bed Performance. Sewage and Industrial Wastes
23(9): 1065, 1951.
5. Rolan, A,T. Determination of Design Loading for
Sand Drying Beds. Jour. North Carolina Section,
AWWA and WPCA L5(1):25-40, 1980.
6. Novak, J.T. and M. Langford. The Use of
Polymers for Improving Chemical Sludge
Dewatering On Sand Beds. In: Proceedings of
the 30th Purdue Industrial Waste Conference,
May 1975, Ann Arbor Science Publishers, Ann
Arbor, Ml, pp. 94-106, 1976.
7. Novak, J.T. and G.E. Montgomery. Chemical
Sludge Dewatering on Sand Beds. ASCE EED
Journal 101(1):1-14, 1975.
8. Handbook: Estimating Sludge Management
Costs. EPA 625/6-85-010, U.S. Environmental
Protection Agency, Center for Environmental
Research Information, Cincinnati, OH, 1985.
9. Reed, S.C., J. Bouzoun, and W. Medding. A
Rational Method for Sludge Dewatering Via
Freezing. U.S. Army Corps of Engineers Cold
77
-------
Regions Research and Engineering Laboratory,
Hanover, NH, 1985.
10, Whiting, D.M. Use of Climatic Data in Design of
Soil Treatment Systems. EPA 660/2-75-018,
U.S. Environmental Protection Agency, Center for
Environmental Research Information, Cincinnati,
OH, 1975,
11. Schleppenbaeh, F.X. Water Filtration at Duluth.
EPA-600/2-84-003, U.S. Environmental
Protection Agency, Center for Environmental
Research Information, Cincinnati, OH, 1983.
12. Farrell, J.B., J.E. Smith Jr., R.B. Dean, E.
Grossman, and O.L. Grant. Natural Freezing for
Dewatering of Aluminum Hydroxide Sludges.
Journal AWWA 62(12):787, 1970.
13. Hernebring, C. and E. Lageson. Conditioning of
Sludge by Natural Freezing, Report No. 27. Dept.
Water & Sewage Technology, University of Lulea,
Sweden, 1986.
14. Rush, R.J. and A.R. Stickney. Natural Freeze-
Thaw Sludge Conditioning and Dewatering.
Report EPS 4-WP-79-1, Environment
Canada, Ottawa, ONT, 1979.
15. Design Information Report, Design, Operational,
and Cost Considerations for Vacuum Assisted
Sludge Dewatering Bed Systems. JWPCF
59(4):228-234, 1987.
16. Condren, A.J., A.T. Wallace, LA. Cooper, and J.F.
Kreissl. Design, Operational and Cost
Considerations for Vacuum Assisted Sludge
Dewatering Bed Systems. Contract Report 68-
03-1821, U.S. Environmental Protection Agency,
Center for Environmental Research Information,
Cincinnati, OH, 1985.
17. Process Design Manual for Sludge Treatment
and Disposal. EPA 625/1-79-001, U.S.
Environmental Protection Agency, Center for
Environmental Research Information, Cincinnati,
OH, pp. 9-3 to 9-16, 1979.
18. Design Manual: Municipal Wastewater
Stabilization Ponds. EPA 625/1-83-015, U.S.
Environmental Protection Agency, Center for
Environmental Research Information, Cincinnati,
OH, pp. 135-143, 1983.
19. Tchobanoglous, 6. Wastewater Engineering
Treatment Disposal Reuse. McGraw-Hill, New
York, NY, pp. 653-656, 1979.
20. Performance Evaluation and Troubleshooting at
Municipal Wastewater Treatment Facilities. EPA
430/9-78-002, U.S. Environmental Protection
Agency, Office of Municipal Pollution Control,
Washington, DC, 1978.
21. Technology Evaluation of Brown Bear Tractor for
Sludge Dewatering, J.M. Montgomery Inc., EPA
Contract Report 68-03-1821, U.S.
Environmental Protection Agency, Waste
Engineering Research Laboratory, Cincinnati, OH,
1984.
22. Sludge Treatment and Disposal; Sludge
Treatment, Volume I. EPA 625/4-78-012, U.S.
Environmental Protection Agency, Center for
Environmental Research Information, Cincinnati,
OH, 1978.
23. Donovan, J. Engineering Assessment of
Vermicomposting Municipal Wastewater Sludges.
EPA-600/2-81-075, NTIS No. PB 81-196-
933, U.S. Environmental Protection Agency,
Center for Environmental Research Information,
Cincinnati, OH, 1981.
24. Loehr, R.C., J.H. Martin, E.F. Neuhauser, and
M.R. Malecki. Waste Management Using
Earthworms. NSF Report ISP 3016764, Cornell
University, Ithaca, NY, 1984.
25. Banks, L. and S.F. Davis. Wastewater and
Sludge Treatment by Rooted Aquatic Plants in
Sand and Gravel Basins. In: Proceedings of the
Workshop on Low Cost Wastewater Treatment,
Clemson University, Clemson, SC, pp. 205-218,
1983.
26. Engineer's Report—Washington Township
Utilities Authority Sludge Treatment Facility.
Costic & Associates, Long Valley, NJ, 1983.
78
-------
Chapter?
Mechanical Dewatering Processes
7.1 Introduction
Some conditions favoring the use of mechanical
dewatering are as follows;
• Aesthetics: Developed areas, where land is a
premium and the use of open air drying might be
offensive, are prime candidates for the use of
mechanical dewatering.
• Climate: Adverse weather conditions are not
conducive to non-mechanical drying methods.
* Costs: Hauling liquid material a significant distance
is not as cost-effective as hauling dewatered
material.
• Site limitations: Lack of available land within an
economical liquid haulage distance (or sludge that
is not suitable for land applications) will favor the
use of mechanical dewatering.
While vacuum filtration remained the most popular
method of sludge dewatering into the early 1970s,
this practice was sharply curtailed by the end of the
decade. No significant improvements in vacuum
filtration have occured recently, and none are
currently foreseen. Improvements in belt presses and
centrifuges resulted in a shift to these devices; they
were more cost-effective and generally produced
better results than vacuum filtration. The oil shortage
in the mid-seventies produced a renewed interest ,in
filter presses, especially in combustion operations
where fuel costs had increased four- to eight-fold
in a matter of two to three years. The fact that both
belt presses and centrifuges were more efficient in
the use of polymers weighed heavily in their favor.
Recently, filter press operations have had success
with polymers. This development should lead to
further interest in filter presses, particularly for
composting, combustion, and restricted landfills
(moisture limited).
Lime and ferric chloride conditioning have become a
less favored option than polymer conditioning for
several reasons. With these conditioning chemicals,
operating and maintenance (O&M) costs are higher
and the operation is less clean and more labor
intensive. The seventies saw an increase in the cost
of lime, as much as 200-300 percent. Further, lime
and ferric chloride conditioning increases the gross
solids for disposal by 20-25 percent, whereas
polymer produces a negligible increase in solids
quantity.
The comparative results obtained with the various
methods of mechanical dewatering can vary from
sludge to sludge. It is realistic to compare data of
different devices only when they have been tested on
the same sludge, during the same time period, and at
the same relative level of capacity. That is, one
should not compare results of one mechanical
dewatering unit operating at 50 percent of its capacity
with one at 100 percent of its capacity.
Still, it is possible to compare mechanical dewatering
devices operating at maximum practical capacity in
relative terms to each other in terms of cake solids,
chemical costs, and solids recovery. This comparison
is shown in Table 7-1. There are not detectable
factors without testing that can result in substantial
differences in cake and polymer costs between
mechanical dewatering devices. For example, a belt
press may produce a range of cake solids between
19 and 30 percent TS on a mixture of 50 PS:50 WAS
when a large volume of data on sludges from several
sources is compared. Because of the great variability
between sludges, larger projects and those requiring
a drier cake should be field tested whenever possible.
The implication is not that it is all right to blunder on
small projects. Rather, the cost of testing is a high
fraction of the total capital cost for small projects and
correction of errors is frequently relatively easy.
7.2 Belt Filter Presses
7.2.1 Introduction
Belt filter presses (BFP) have been used in Europe
since the 1960s and in the United States since the
early 1970s. They were initially designed to dewater
paper pulp and were then modified to dewater
sewage sludge. The European models were the first
used in the United States. However, the difference
between U.S. sludge and European sludge led to
performance problems (low cake solids and poor
solids capture). American manufacturers began
79
-------
Tablo 7-1.
Comparative Mechanical Dewatering
Performance
Dowatoring Unit
Belt press
Conlnfugo
Vacuum filler
Filter pross - Lo P
Filler pross - Hi P
Filter press - Oia P
Screw pross
Low pross drum/bells
Cake
Solids1
% TSS
X
X ± 2
X-4
X + 8
X + 10
X + 12
X-2
X- 10 +
Recovery
%TSS
90,-952
90-952
85-90
98 +
98 +
98 +
90 +
90 +
Polymer
Cost3
%/ton DS
Y
0.8 Y
0,9 Y
1,1 Y
1.1 Y
1.1 Y
1.2 Y
0.8 Y
' Relative to bolt pross, X denotes base level.
2 Controlled by polymer dosage.
3 Relative to belt press, Y denotes base level.
building and selling presses in the United States.
However, the early American-made models were
also plagued with many problems, mainly mechanical
failures of rollers and bearings. Since the American
units were designed with the same principles used for
the design of belt conveyors, they were much lighter
than their European counterparts. The bearings and
roller shafts were unable to withstand the forces
generated during the dewatering operation.
By the late 1970s, American manufacturers made
significant improvements in their units. For example,
they followed the European practice of using bearings
rated at an LIQ life of 100,000 hours. Further, they
increased the diameter of the roller shafts and
improved the materials of construction, thus reducing
the number of failures considerably. Since these
Improvements have been made, the use of belt filter
presses has increased. With the newest models on
the market, mechanical failures are limited and
process performance has improved. Although there
are now about 20 belt filter press suppliers in the
United States, only about 5 are considered to be
major manufacturers.
7.2,2 General Description
Belt filler presses are designed on the basis of a very
simple concept. Sludge sandwiched between two
tensfoned porous belts is passed over and under
various diameter rollers. For a given belt tension, as
roller diameter decreases, an increased pressure is
exerted on the sludge, thus squeezing out water.
Although many different designs of belt filter presses
are available, they all incorporate the following basic
features:
• Polymer conditioning zone
• Gravity drainage zone
• Low pressure zone
» High pressure zones.
Figure 7-1(1) shows these zones identified on a
simplified schematic of a belt filter press.
The polymer conditoning zone can be a small tank,
approximately 265-379 I (70-100 gal) located 0.6-
1.8 m (2-6 ft) from the press; a rotating drum
attached to the top of the press; or an in-line
injector. Each press manufacturer usually supplies the
polymer conditioning unit with the belt filter press.
The gravity drainage zone is a flat or slightly inclined
belt which is unique to each press model. In this
section sludge is dewatered by the gravity drainage of
the free water (interstitial water in the sludge slurry).
The engineer should expect a 5 to 10 percent
increase in solids concentration in the gravity
drainage zone from the original feed sludge (2,3). For
example, a primary/waste activated sludge mixture fed
to the press at about 3.0 percent solids will be about
10-12 percent solids at the end of the gravity zone.
Problems such as sludge squeezing out from
between the belts and blinding of the belt mesh can
occur if the sludge does not drain well in this zone.
This free water drainage is a function of sludge type,
quality, conditioning, screen mesh, and the design of
the drainage zone.
The low pressure zone, also called the wedge zone
by some manufacturers is the area where the upper
and lower belts come together with the sludge in
between, thus forming the sludge "sandwich." The
low pressure zone is very important since it prepares
the sludge by forming a firm sludge cake which is
able to withstand the shear forces within the high
pressure zone.
In the high pressure zone, forces are exerted on the
sludge by the movement of the upper and lower belts,
relative to each other, as they go over and under a
series of rollers with decreasing diameters. Some
manufacturers have an independent high pressure
zone which uses belts or hydraulic cylinders to further
increase the pressure on the sludge (see Rg. 7-2),
thus producing a drier cake. A dry cake is especially
important for plants that use incineration as the final
disposal method and need the driest cake possible.
7.2.3 Theory of Operation
The high pressure zone is critical to good press
performance (high cake solids and recoveries). A
design manual prepared by Rubel and Hager, Inc. (4)
contains models to describe the various effects in a
typical high pressure zone (see Figure 7-3). This
design manual provides these equations to familiarize
the engineer with the belt filter press design process.
The equations can be used to calculate the following
parameters:
• Pressure on the sludge cake due to drive torque
(force required to pull the belt through the press):
80
-------
Figure 7-1. Simplified schematic of a belt filter press.
Independent High
Pressure Section
High Pressure
Shear Zone
Free Drainage
Zone
Figure 7-2. Typical independent high pressure section.
Pw^^jij^W0 55, p ' 'It
* fj«-*'i4',J
v;;f !,.•>"• s$-,**'tA'^|
psii = 2F1/D= 1,700 HPV(D) (fpm)
where,
(7-1)
psi-) = maximum pressure on the sludge cake due
to FI
FI = Ib of force due to drive torque per inch of
belt width
D = roller diameter, in
HP' = drive horsepower per inch of belt width per
belt
fpm = belt speed, ft/min
• Pressure on the sludge cake due to belt tensioning
(for presses that use pneumatic or hydraulic
cylinders to tension the belts)
psi2 = 2F2/D + 2P Cos [a/(DWY/2)]
where,
(7-2)
psi2 ~ average pressure on sludge cake due to Fa
F2 = Ib of force due to take-up tension per inch
of belt width - required to prevent slack
belts and to provide traction for the drive
rolls
P = resultant force from tensioning roller
actuator. It is the pressure (force) you set
and can easily measure.
a = angle between belt force resultant and
actuating cylinder axis
D = diameter of roller, in
W = active belt width, in
Y = belt wrap angle at take-up roller
• Pressure on the sludge cake due to belt elasticity
psia = 2F3/D
where,
(7-3)
81
-------
Flguro 7*3. Typical high pressure zone.
W = 80 in
HP' = 0.0127
Roll
Diameter, in
LI, In
L*,!n
0
TlD©
IST
35
34
16
244°
76,6
30
16
15
130°
34.0
24
15
16
114°
23.9
21
16
11
160°
29.3
13
11
17
185°
21.0
10
17
10
180°
15.7
8
10
24
180°
12.6
D
E
e
A
LI
LO
= average pressure on the sludge cake due
to Fa
= Ib of force due to belt elasticity per inch of
belt width, where Fa = 2eE/D
= roller diameter, in
= modulus of elasticity of the bell {i.e.,
stress/strain before yield point)
= belt strain, A/LI
= belt stretch (l_o-L|), cm or in
= tangent length of belt entering roller
= length of outer belt around roller between
tangent points on adjacent rollers
= length of inner belt around roller between
tangent points on adjacent rollers
Total pressure on the cake at any roller:
+ psi3 (7-4)
+ F3]/D
psi = psii +
or
psi = 2[Fi +
With these equations the engineer can calculate the
total pressure on the sludge cake at each roller to
ensure that there is a gradually increasing pressure
on each successive roller. These equations also allow
determination of roller and shaft diameters and
bearing size requirements, which can then be
compared to the belt press manufacturer's
specifications.
An example using these equations to evaluate a belt
filter press design is shown below and is reproduced
with permission of the author (4). The parameters
needed to evaluate a design follow:
D = diameter of each roller, in
LI = tangent length of belt entering each roller
L2 = tangent length of belt leaving each roller
0 = angle of wrap of the belt around the roller.
The above parameters are available from the
manufacturer's specifications. Other parameters
follow:
Q = sludge throughput rate, Ib of solids/minute
IA = cake thickness at the entrance to the high
pressure zone, in
IB = cake thickness at exit
CA = cake solids concentration at the entrance to
the high pressure zone as a decimal
CB = cake solids concentration at the exit of the high
pressure zone as a decimal
82
-------
E = modulus of elasticity from belt manufacturer's
specifications, Ib/sq in
HP' = drive horsepower per inch of belt width per belt
W = belt width, in
t-| = cake thickness at entrance of roller, in
t£ = cake thickness at exit of roller, in
ta = average cake thickness at roller, (t^ + t2)/2
L = effective length of belt (portion over each
drum), D8/360.
For this example, the magnitudes of the values for D,
LI, L2 and 0 are shown in Figure 7-3. The
magnitudes of the other parameters follow:
W = 80 in
Q =125 Ib/min
IA = 4.0 in
CA = 0.1
Cb = 0,38 (desired value)
HP' = 0.0127 HP/in
fpm = 2.3Q/(W x IA x CA) = 9-0 fpm
Procedure:
1. Calculation of cake thickness at exit of high
pressure zone.
IB = (0.95 x IA x CA)/CB = 1.0 in
(7-5)
This equation assumes 95 percent solids capture
across the unit.
2. Plot: FI + Fa, in Ib/in vs. D8/360
3. Calculate:
e + A/Li = (2ta)/{l360(L1 + L2)/n6] + D> (7 - 6)
4. Calculate:
Fa = 939e-[(100/2,OOOe2)+1]-M08 (7-7)
5. SF = F! + F2 + F3
6. psi = 2/(dSF)
Comments and Interpretation of Results (Table 7-2):
1. The value of IA will have some maximum allowable
value according to the press design.
2. ta is the value at the center of the roller, halfway
between tangent points.
3. The total pressure (psi) calculated for each roller is
compared to the pressures specified by the
manufacturer. In this example, the sharp increase
in pressure between rollers 4 and 5 suggests that
sludge would either extrude into the belt mesh,
clogging the belt or squeeze out from between the
two belts. It is important to remember that the
pressure should increase gradually from roller to
roller and from zone to zone. Sharp increases of
pressure can indicate operating problems.
4. In this example, the belt tension (SF) is greater
than 250 Ib/in in many cases. A force of this
magnitude tends to deform most dewatering belts,
which suggests problems with belt tracking and
alignment as the belt wears.
5. The angle of wrap (6) should be as close to 180°
as possible.
6. Use FI, FZ, and F3 for analyzing the size of the
rollers and bearings in the other zones. Failure to
include F$ when analyzing the high pressure
section rollers and bearings could result in bearing
and/or shaft failure.
7. The design engineer should require a submittal of
the calculations from the manufacturer to confirm
bearing and shaft design.
8. To calculate the reaction at each bearing for rollers
in the high pressure zone with a bearing on each
end of the roller, use Equation 7-8:
R = W[2(Fj
sin(6/2)
(7-8;
7.2.4 Mechanical Description
The mechanical components of a belt filter press
generally include the following:
» Dewatering belts
» Rollers and bearings
» Belt tracking and tensioning system
• Controls and drives
• Belt washing system.
The dewatering belts are usually woven from
monofilament polyester fibers. There are various
weave combinations, air permeabilities and particle
retension capabilities available from belt
manufacturers. These parameters greatly influence
how well the press will perform. For example, a
sludge with a high concentration of activated sludge
(80 - 100 percent) may require a belt with a high air
permeability and high solids retention capability, while
a primary sludge might require just the opposite.
Therefore, since sludge types and qualities vary
considerably from plant to plant, it can be important
for the press manufacturers and engineers to try to
evaluate different weaves, permeabilities, and solids
retention capabilities for each installation to ensure
optimum performance. The initial belt type is usually
provided by the supplier based on the sludge type
and his past experience.
Such an evaluation is very simple for a plant that is
already in operation and is generating sludge that can
be treated on a belt filter press. For a newly designed
83
-------
Tnblo 7-2. Summary of Pressures Calculated for Each Roller
Roll
1-2
psi
1
2
3
4
5
6
7
36
30
24
21
13
10
8
94
16
15
16
11
17
10
16
15
16
11
17
10
24
244
130
114
160
185
180
180
43
53
58
63
67
70
73
4.0
2.9
2.4
2.1
1.7
1.4
1,2
2.9
2.4
2.1
1.7
1.4
1.2
1,0
3.45
2.65
2.25
1.90
1.55
1.30
1.10
0.079
0.092
0.082
0.094
0.102
0.096
0.074
174
188
178
190
197
192
169
217
241
236
253
264
262
242
12
16
20
24
41
52
61
plant, the belts must be evaluated differently. Sludge
from a similarly designed plant can be tested or
surveys can be made of similar plants to determine
what type of belt they are using. Further, belt filter
press manufacturers can supply information on how
to relate belt characteristics to type of sludge
expected. Usually, the belt filter press manufacturer
can recommend the belt best suited to the type of
sludge expected at the plant. Once the plant is
operating, belts with different characteristics can be
tried to obtain optimum performance. The belts
should be designed for ease of replacement with a
minimum of downtime to insure continuous
dewatering. There are two different types of belts,
split and continuous. The split belts are joined
together with a splicing device called a clipper seam
(see Rgure 7-4). Split belts are the most common
type on the market and can be used on all models of
belt filter presses. The continuous or seamless belt
can only be used on certain presses. The designer
should consult the press manufacturer for a specific
belt recommendation. The manufacturers of
continuous belts claim longer life than with the split
type. However, there is no available data to
substantiate this claim. Continuous belts are more
difficult to install.
The rollers and bearings are the main mechanical
components of the belt filter press. As stated earlier,
the rollers provide the pressure and forces that allow
dewatering to occur, but they also insure proper belt
support and tension. The tensioning device is the key
control once the roller sizes have been fixed. Roller
diameter and shaft sizes are key design parameters
and should be carefully evaluated (Section 7.2.3).
The bearings are extremely important components
since they support and guide the rollers.
Press controls should be centralized either on the
press itself or on a remote control panel (Figure 7-5)
and should include automatic sequential start-up and
shutdown systems, instrumentation for tracking and
tensioning of the belts, pressure gauges, and safety
interlocks. In addition, many engineers include
running time meters, sludge and polymer pump
controls, and other auxiliary devices that allow for a
more efficient operation. The panels should be NEMA
Figure 7-4. Typical clipper seam for split dewatering belt.
12 and should be well designed with centralized
controls and adequate safety interlocks. If possible,
they should be located in a separate control area
away from noise, odors, and moisture that might
affect the controls. Each piece of equipment which
makes up the sludge dewatering system should be
interconnected so that each unit is started in the
proper sequence. For example, the press and
conveyors should be started before the sludge and
polymer feed pumps. In addition, automatic shutdown
of equipment must be provided, also in the proper
sequence. If a piece of equipment downstream fails,
everything upstream of that unit must shut-down,
i.e., if the press fails for any reason, the polymer and
sludge feed pumps must shut-down automatically.
Automatic shutdown of dewatering equipment should
occur for any of the following fault conditions (3):
Belt drive failure
Sludge conditioning tank failure
Belt misalignment
Insufficient belt tension
Loss of pneumatic or hydraulic system pressure
Low belt wash water pressure
Emergency stop
84
-------
Figure 7-5. Control panel for belt filter press dowatering
system.
. Upper and Lower
-Screen Tension Controls
Automatic
Start
ss Drive
Controls
Automatic
Shutdown
Bleed
Valve
Hydraulic
Pump
Controls
Conveyor
Drive
Controls
Main Air
Gauge
Hydraulic
Pressure
Emergency
Stop
• High sludge level on gravity drainage section
• Polymer feed pump shut-down (should stop feed
pumps).
The belt washing system should include a high
pressure water pump, a set of spray bars for cleaning
both the upper and lower belts, and a spray cleaning
device. Belt washing occurs after the cake has been
removed from each belt. It is washed from the side
opposite that which is in contact with the sludge cake.
Either potable water or high quality plant effluent
could be used as wash water for the presses. Since
some belt presses require as much as 3.16 l/s (50
gpm) per meter belt width of wash water, it is more
economical to use plant effluent. Therefore, when
designing a system ensure that the pump, piping, and
nozzles are capable of handling high quality plant
effluent. However, when plant effluent is used, a
highly efficient filtration system must be installed
upstream of the press to ensure that the effluent is
free of solids that can clog the spray nozzles. The
spray nozzles should be designed for easy access to
enable efficient and thorough cleaning which ensures
complete cleaning of the filter belts. Many of the new
models of presses are equipped with stainless steel
brushes within the spray header to automatically
clean the nozzles without removing them from the
press. Figure 7-6 (5) illustrates this type of spray
header. This type of cleaning system should be
specified since they are easy to use and require
much less operator time than the manually cleaned
systems.
7,2.5 Performance Characteristics
Belt filter presses can be used to dewater most
sludges generated at municipal wastewater treatment
plants. However, the sludge must be conditioned with
polymer to ensure optimum performance. Polymer
produces a phenomenon known as superflocculation
(2). Superflocculation is the formation of large, strong
floe which causes free water to drain easily from the
sludge in the gravity drainage zone of the belt filter
press. Superflocculation also produces a sludge that
can withstand the pressures generated during the
dewatering process and prevents the sludge from
squeezing out from between the dewatering belts.
Only polymer can produce this phenomenon. Some
plants have tried to dewater lime conditioned primary
sludge on belt filter presses. (H, Johnson,
Ashbrook-Simon-Hartley, personal communication,
1987; J. Labunski. Parkson Corp., personal
communication, 1987.) The results have been poor,
with low cake solids, low solids capture, and blinding
of the belt mesh. If the designer wants to use lime
stabilization and landfilling as the final disposal
method, a post-lime-stabilization system should be
used. First, the sludge is conditioned with polymer
and dewatered, and then lime is added to the sludge
cake (6). New Haven, CT is successfully dewatering
a high pH (lime added) raw primary and waste
activated sludge using a cationic polymer. However,
this can be site specific and results vary widely.
Usually cationic polymers are used for sludge
conditioning. Sometimes a two polymer system, such
as a cationic polymer following either an anionic or a
nonionic polymer, will be used on a belt filter press to
improve cake release from the upper dewatering belt.
The polymer must be selected carefully to ensure
optimum performance. Always contact the polymer
manufacturer's representative for help screening and
testing polymers; this service is normally provided
free.
Typically, polymer and sludge are piped to the
conditioning unit. When designing the polymer
conditioning system, it is important for the designer to
locate polymer feed points at several locations: one at
the conditioning unit itself, one about 0.6 to 0.9 m (2
to 3 ft) upstream of the unit directly into the sludge
feed piping, and one about 7.6 m (25 ft) upstream.
With two polymer feeds, the anionic or non-ionic
polymer may have to be added before the sludge
pump. Feedpoint location is especially important for a
new installation where sludge characteristics are not
known, but is also important for any plant, since
sludge characteristics can change periodically.
Sometimes the sludge will condition better when
there is a longer contact time with the polymer, but at
other times it requires a shorter period. Therefore, by
designing polymer feed points at several locations,
85
-------
Figure 7-6. Washwater spray bar with cleaning brushes.
Handwhool
^—Seal Wash Tube-
r/Hi *.
. Wire Brush —
=4 \ ..-.„.. ,.^ Gasket
• Orifice Plate
Threaded Bushing
Seal
Wash Tube
Nozzles
Washbox
Bypass Drain
the designer can ensure flexibility and optimum
performance. For more information on sludge
conditioning, refer to Chapter 5. Typical dewatering
data for various types of sludges is shown in Table
7-3 (3).
Typical polymer conditioning costs for belt filter press
dewatering range from a low of $2.65/Mg ($2.41/dry
ton) to a high of $91.15 Mg ($82.86/dry ton) with an
average of $24.38/Mg ($22.l7/dry ton).
Odors can be a problem during belt filter press
operation. Odors can be controlled with good
ventilation systems, ensuring that the sludge is kept
fresh, and using chemicals such as potassium
permanganate to neutralize the odor-causing
chemical.
Potassium permanganate, KMNO4, is a strong
oxidizing agent which rapidly ozidizes the hydrogen
sulfide in sludge to odorless sulfate. It is packaged as
a dry, crystalline, water-soluble powder. The
permanganate solution can be fed directly into the
suction side of the sludge transfer pump so that the
pump itself can act as a mixer for the permanganate
and sludge. The permanganate feed point should be
upstream of the polymer feed point by a distance
which yields a travel time of about 1 minute (7) to
ensure the most effective use of the KMN04. A
typical installation is shown in Figure 7-7. It should
also be noted that some plants use hydrogen
peroxide for the same purpose as potassium
permanganate.
A dosage of about 0.5-2.0 kg/Mg (1-4 Ib/dry ton) is
typical, but the dosage depends on the concentration
of hydrogen sulfide in the sludge. Permanganate adds
about $1.00/Mg to the cost of dewatering the sludge
but it produces several benefits besides the
destruction of the sulfides. These include (7):
• Improved sludge dewaterability - slight increase in
cake dryness using permanganate
• Reduction of polymer - increased sludge
production with no additional polymer
• Sulfide-free recycle stream from the belt presses.
7.2.6 Design of a New Installation
This design example (8-10) is for a proposed 0.22-
m.3/s (5-mgd) secondary, activated sludge, treatment
plant with primary clarifiers. The plant will be
dewatering raw sludge during an eight hour day, five
days per week. The sludge will be thickened before
dewatering to 5.0 percent solids. How many and what
size presses will be required?
Belt presses are not sized on the basis of wastewater
flow to the plant, but on the basis of the weight or
volume of sludge to be dewatered. The following
calculations show how the required number of
presses can be determined.
• Determination of the amount of primary sludge:
Influent total suspended solids concentration is
determined by laboratory analysis (11) of the
86
-------
Type of Sludge
Raw:
P
WAS
P + WAS
P + TF
Anaerobically Digested:
P
WAS
P + WAS
Aerobically Digested:
P + WAS
P+TF
Oxygen Activated:
WAS
Thermally Conditioned:
P + WAS
Feed Solids
percent
3-10
0.5-4
3-6
3-6
3-10
3-4
3-9
1-3
4-8
1-3
4-8
Solids Loading Rate
kg/hr/m belt width
360-680
45-230
180-590
180-590
360-590
40-135
180-680
90-230
135-230
90-180
290-910
Polymer Dose
9/kg
1-5
1-10
1-10
2-8
1-5
2-10
2-8
2-8
2-8
4-10
0
Cake Solids
percent
28-44
20-35
20-35
20-40
25-36
12-22
18-44
12-20
12-30
15-23
25-50
Figure 7-7. Belt filter press installation using permanganate for odor control.
Sludge Dewatering Room
Belt Press Po|ymer Feec| Point Belt Press
[
;*>li>c
Chemical Feed Room
Potassium
Permanganate
Feeder ^_
'WW^f .rw ww ww w w ,,,,,„„),_.,
Sludqe Transfer Pump
V
.^. j
1
irn i i i — i — i — *-H*
Sludge Tank
Permanganate Feed Point
wastewater that will be flowing through the plant.
For this example, assume 220 mg/l (8).
Total suspended solids
= 220 mg/l x 5 mgd x 8.34 Ib/Mgal/mg/i (7-9)
= 9,180 Ib/d (4,168 kg/d)
A primary clarifier will remove an average of 60
percent of the suspended solids. Therefore:
Total primary sludge = 0.6 x 9,180 Ib/d
= 5,508 Ib/d (2,501 kg/d)
Assume the primary sludge is 3.5 percent solids.
Total volume of primary sludge
= (5,508 Ib/d) * [{0,035)(8.345 lb/gal)J
= 18,858 gal/d (71.38 m3/d)
• Determination of the amount of waste activated
sludge: Influent BODs is determined by analyzing
the wastewater (11). For this example, assume 220
mg/l (8), a 30 percent removal of BOOs in the
primary clarifier, and a total plant removal of 90
percent of the BODs.
BOD Removal
= 0.7 (220 mg/l) - (0.1 )(0.7)(220 mg/l) (7-10)
= 139 mg/l
BOD Removed
= 139 mg/l x 5mgd x 8.34 Ib/Mgal/mg/l
= 5,780 Ib/d (2,622 kg/d)
Assume a sludge yield coefficient of 0.5 @ SRT of
16 days:
Solids production = 0.5 (5,780 Ib/d)
= 2.892 Ib/d (1,313 kg/d)
87
-------
Total waste activated sludge will then be equal to the
amount of suspended solids in the primary effluent
plus the solids production in the aeration system
minus the effluent suspended solids.
WAS = 0.4 (9,180 Ib/d) + 2,890 Ib/d - 918 Ib/d
= 5,646 !b/d (2,563 kg/d)
Assume that the waste activated sludge is 1
percent solids:
Total volume of WAS
= (5,646 lb/d)/[(0.01)(8.34 Ib/ga!)]
= 67,657 gal/d (256 m3/d)
» Determination of the volume of the 5 percent
thickened sludge to be dewatered per day.
The total volume of the 5-percent mixed sludge
would be about 27,000 gpd.
The plant wants to dewater only 5 days per week,
therefore:
(27,000 gpd x 7)/5 = 37,800 gpd (143 m3/d)
(7-11)
A typical belt filter press has a hydraulic loading of
40 gpm per meter of belt width:
37,800 galions/40 gpm = 945 minutes = 16 hr
Therefore, a 1-m belt press would dewater the
sludge produced at this plant in a 16-hr day.
However, sludge is to be processed on an 8-hr day
and it is also important to design for excess capacity.
Therefore, two 1.5-m units should be used. With
both units operating, the sludge would easily be
dewatered in 8 hr. Further, there is enough excess
capacity so that if one unit were out of service, the
other unit could process all of the sludge in a 10.5-
hr day, thus preventing a build-up of sludge in the
plant.
Following is a summary of equipment and operating
recommendations for the design of belt filter press
installations:
A. Equipment
• Use durable materials for equipment construction.
• Provide sturdily constructed, properly coated
frames.
• Use long-life bearings (l_io life of at least 100,000
hr)
• Use high strength rollers.
• Provide continuously acting tension/tracking
systems.
* Use high quality, durable, and properly woven
materials for belts.
• Ensure that both the press and belt manufacturers
provide high levels of quality control.
B, Performance
* Consult manufacturers for design and performance
data early in planning stage,
» Confirm performance data with other operating
installations and/or through pilot testing.
• Specify high quality equipment and require a
performance bond.
• Assure system integration by specifying that the
dewatering system be the responsibility of a single
supplier.
C. Auxiliary Equipment
* Provide sludge blending prior to dewatering to
enhance continuity of feed sludge.
• Use a macerator upstream of the belt filter press to
ensure a homogenous feed.
* Provide flexibility in points of polymer application
and type of polymer used.
» Use continuously acting high pressure sludge
pumps such as progressive cavity or rotary lobe.
• Provide positive ventilation for odor control in the
dewatering area. Provide carbon or chemical odor
control system.
D. Controls
• Provide instrumentation to monitor such operating
parameters as sludge, filtrate, and washwater flow
(including provisions for sampling).
• Integrate ancillary system controls with those for
the dewatering equipment and interconnect key
control functions.
• Protect controls from the rnoist, corrosive operating
environment.
£ Safety
• Provide non-skid walkways and floors.
• Provide adequate access to equipment.
• Assure installation and maintenance of emergency
stop systems, drive guards, and other protective
equipment.
• Educate operators to follow safety precautions;
assure adherence to rules.
F. Operations
• Monitor system performance to assure optimum
operation.
• Assure that sludge is properly conditioned.
* Assure good gravity drainage of sludge.
G. Operator Training
» Provide operation and maintenance training upon
completion of installation.
• Provide ongoing training to maintain skills.
7.3 Centrifuges
7.3.1 Introduction
Centrifugal dewatering of sludge is a process that
uses the force developed by fast rotation of a
cylindrical bowl to separate the sludge solids and
liquid. In this basic process, when a sludge-water
88
-------
mixture enters the centrifuge, it is forced against the
bowl's interior walls, forming a pool of liquid and
sludge solids. Density differences cause the sludge
solids and the liquid to separate into two distinct
layers. The sludge solids, "cake," and the liquid,
"centrate," are then separately discharged from the
unit. The two types of centrifuges used for municipal
sludge dewatering, basket and solid bowl, both
operate on these basic principles. They are
differentiated by the method of sludge feed,
magnitude of applied centrifugal force, method of
solids and liquid discharge, cost, and performance. A
third centrifuge type, the disc-nozzle centrifuge, has
been used for thickening waste activated sludge
(WAS), but does not produce a dewatered material. It
will not be discussed in this manual.
Engineers have long recognized the potential of
centrifugal dewatering devices for handling both
domestic and industrial waste slurries and sludges.
Recent improvements in materials, design, and
process technology have not only produced sturdier,
more sophisticated centrifuges, but also have given
rise to new operational procedures and design
practices. These developments have alleviated or
eliminated past problems, and have increased
process performance levels.
A five-decade evolution preceded the centrifuge's
acceptance in municipal wastewater treatment.
Though simple in concept, the practical aspects of
centrifuge design are quite involved and are still
controversial. Following the 1902 development and
testing of the first perforated basket centrifuge in
Cologne, Germany, Herman Schaefer joined forces
with Dr. Gustav Ter Mer in 1907 to produce improved
basket units. These batch units were installed in
Germany and the USSR. A modified Schaefer-Ter
Mer unit was installed during 1920 in Milwaukee,
Wisconsin to thicken activated sludge. It could not
produce clear centrate at an economical rate and this
finding was further verified during the twenties and
thirties.
A solid-bowl conveyor centrifuge with continuous
feed and discharge of solids was developed and
tested during the thirties for dewatering of raw and
digested sludges. Two decades later an improved
version was successfully installed and operated in
California. It marked the start of a gradual acceptance
of the solid bowl centrifuge by the wastewater
treatment industry. This unit's bowl was 1.5 times as
long as its diameter; produced a centrifugal force in
the range of 900 to 1,500 Gs; and produced cake
solid concentrations of 20-35 percent with a solids
recovery of 50 to 70 percent. An improved solid bowl
centrifuge, producing a force of over 3,000 Gs and
with a bowl to length diameter ratio of 2.5:1, was
introduced in the early 1960s. It had better recovery
of sludge solids and higher cake solids. In 1965 a
German unit was produced that had a bowl length
diameter ratio of 2.9:1 and operated at forces in the
range of 2,500 to 4,000 Gs. While it gave even better
performance, wear was a significant factor.
In the 1970s, significant improvements were made in
centrifugal designs and sludge conditioning. In the
early 1970s German manufacturers developed a slow
speed concurrent centrifuge, which could produce
high recoveries and acceptable cake solids; they
introduced the use of the hydraulic backdrive
arrangement. This backdrive device provided
automatic control of "the centrifuge bowl and conveyor
differential speed, using the change in the conveyor
torque output to adjust differential speed to maintain
constant torque. In 1974, a major advancement in
protection of wearing surfaces was the development
of special hardened, easily replaceable tiles on the
wearing surfaces. Four- to eight-fold increases in
the operating life of the conveyors resulted.
From 1974 to 1986, there were other refinements in
construction of centrifuges to enhance performance
and reduce O & M costs. Machine sizes and capacity
grew with sludge centrifuges available up to 183 cm
diameter (72 in) x 427 cm (168 in) bowl length and
with capacities exceeding 37.9 l/s (600 gpm). These
machines are used in both thickening and dewatering
of sludges. Both high-speed and low-speed
centrifuges are employed for all types of sludge.
As the advancements in solid-bowl centrifuges have
primarily been occurring in the conveyor type, there
has been a declining interest in the basket centrifuge
for municipal sludge dewatering.
7.3.2 General Centrifuge Type and Description
The solid-bowl centrifuge, also called decanter,
conveyor, or scroll centrifuge, is characterized by a
rotating cylindrical-conical bowl. A helical screw
conveyor fits inside the bowl with a small clearance
between its outer edge and the inner surface of the
bowl. The conveyor rotates, but at a slightly lower or
higher speed than the bowl. This difference in
revolutions per minute (rpm) between the bowl and
the scroll is known as the differential speed, which
allows the solids to be conveyed from the zone of the
stationary feed pipe, where the sludge enters, to the
dewatering beach, where the sludge cake is
discharged. As shown in Figure 7-8, the scroll
pushes the collected solids along the bowl wall and
up the dewatering beach, located at the tapered end
of the bowl, for final dewatering and discharge.
The differential speed between the bowl and
conveyor is maintained by several methods. Earlier
designs used a double output gearbox that imparted
different speeds as a function of the gear ratio. It was
possible to vary the output ratio by driving two
separate input shafts. Eddy current brakes are also
used to control the differential. Latest designs provide
automatic speed control as a function of conveyor
torque that can maximize solids concentration.
89
-------
Figure7*8. Solid-bowl (countercurrent) conveyor
dlschargo centrifuge.
Georbox
Drive
Pulley
Feed
Liquid
Discharge
T
Scroll
(Conveyor) Solid
Discharge
The solid-bowl centrifuge operates in one of the two
modes: countercurrent or continuous concurrent. The
major differences in design pertain to the location of
the sludge feed ports, the removal of centrate, and
the internal flow patterns of the liquid/solids phases.
In the countercurrent centrifuge, influent sludge is
introduced through the feed pipe at or near the
junction of the cylindrical and conical sections of the
bowl; the solids move to the conical end of the
machine while the centrate flows in the opposite
direction. This design is shown in Figure 7-8. Under
the influence of centrifugal force, the sludge solids
are pushed against the bowl wall. The solids are then
moved gradually by the rotating conveyor along the
bowl wall, up the dewatering beach. From there, they
drop into a sludge cake discharge hopper. Centrate,
the partially clarified liquid containing smaller and
some finer unflocculated solids, flows around (and
through) the conveyor toward the liquid discharge
end. Depending on the sludge particle characteristics,
gravitational forces and residence time, a portion of
these solids settle to the outer wall and are also
conveyed to the solids discharge end while the liquid
flows over the adjustable weir. The length of the
conical section (drying beach) above the pool level
may vary considerably depending on the specific
sludge characteristics and the centrifuge
design/operation. In general, the lower the structural
strength of the cake, the smaller the desirable drying
beach length.
In the concurrent model, shown in Figure 7-9, feed
slurry is introduced at the opening opposite the
dewatering beach. The settling zone then begins near
or at the feed point, and the solids travel the full
length of the bowl. While general construction is
similar to the countercurrent design, the centrate
flows in the same direction as the sludge solids
(concurrent flow) and is withdrawn by a skimming
device or return tube located near the junction of the
cylindrical bowl and the conical section. Clarified
centrate then flows into channels inside the scroll hub
and returns to the feed end of the machine, where it
is discharged over adjustable weir plates through
outlet ports built into the bowl head.
Figure 7-9. Solid-bowl (concurrent) centrifuge with
hydraulic scroll drive.
Polymer
Feed
Line
Lube
Reservoir
Conveyor/
Scroll
Housing
Conveyor/
Bowl
Differential
Gearing
Drive
Pulleys
The concurrent design of conveyor centrifuges is
most often operated at speeds lower than the
countercurrent design and in the range of 700 to
1,500 gravities depending on machine size and
sludge properties. It has been categorized as a
"low-speed" or low-G centrifuge, which operates
at 50-75 percent of the gravitational force of a
high-G centrifuge.
7.3.3 Applications
A solid-bowl centrifuge's ability to be used either for
thickening or dewatering provides flexibility and is a
major advantage. For example, a centrifuge can be
used to thicken ahead of a filter press, reducing
chemical use and increasing solids throughput.
During periods of downtime of the filter press, the
solid-bowl centrifuge can serve as an alternate
dewatering device. Another advantage for larger
plants is the solid bowl centrifuge's sludge throughput
capability, which allows the largest single units of any
type of dewatering equipment. The larger centrifuges
are capable of handling 19 to 38 l/s (300 to 600 gpm)
per unit depending on the sludge's characteristics.
The centrifuge also has the ability to handle higher
than design loadings, such as a temporary increase in
hydraulic loading or solids concentration, and the
percent solids recovery can usually be maintained
with the addition of a higher polymer dosage. The
cake solids concentration will likely decrease, but the
centrifuge will handle the higher solids loading.
As discussed in Chapter 3 for all sludge dewatering
processes, it is helpful to at least partially thicken the
feed to the centrifuge, so that capacity is not limited
due to the excessive water content of the sludge.
Ideally, the feed solids are sufficiently pre-
concentrated such that the centrifuge liquid (Sigma)
and solids (Beta) capacity are reached at the same
time. Diluted feeds can result in reduced solids
capacity as well as increased polymer requirements.
90
-------
High rate preconcentration of dilute sludges in a
gravity thickener using loadings that are 200-400
percent higher than normal gravity thickener loadings
can improve performance, reduce costs due to partial
thickening of the sludge, and provide a more stable
operation using the reservoir of sludge in the
thickener. Recommendations for gravity thickener
loadings and underflow concentrations for both
conventional and high rate are provided in Table 7-
4.
Table 7-4.
Conventional Gravity Thickening and High Rate
Pre-Concentration
Conventional
High Rale
Sludge
RawP
Raw WAS
(PS + WAS)
(60:40)
Loading
ka/m2/d
98
20
40
Underflow
percent
8-10
1.75 + SDI
4.5-5.5
Loading
kg/m2/d
196
74
98
Underflow
percenl
5-6
0.5 + SDI
3.5-4.0
SDI = Sludge Density Index.
7.3.4 General Design Theory and Considerations
The solid-bowl centrifuge is essentially a high
energy (g) settling unit. Particles entering the liquid
pool settle toward the outer wall aided by gravity and
resisted by the same factors that slow settling and
resist compaction in clarifiers and thickeners. The
capacity of the centrifuge is also affected by the type
of solids, rate of solids removal, solids concentration,
etc. Figure 7-10 compares the clarifier and
centrifuge. Whereas the clarifier is large, operating at
1 G, the centrifuge clarifier-thickener operates at
high multiples of gravity to separate liquid and solids
in a shallow, small volume unit. The high gravitational
force permits a greater removal of water, resulting in
a dewatering function.
Although there are some major deficiencies, the
settling velocity of a particle in a fluid under the
influence of gravity can be defined by Stokes Law as
follows:
V = g (Psolid - Pliquid)d2-M,800n (7-12)
where,
V
9
Psolid
Pliquid
d
= settling velocity of solid particle in the
fluid, m/s
= acceleration due to gravity, m/s2
= particle density, kg/m3
= liquid density, kg/m3
= mean particle diameter, m
= viscosity of liquid, kg/(m x s)
For settling in a centrifuge, the same relationship can
be used but with g, the acceleration of gravity,
Figure 7-10. Comparison between the clarifier and the
centrifuge (courtesy Pennwalt Corp., Sharpies-
Stokes Div.)
A centrifuge has the same basic characteristics as a clarifier. It Is a
clarifier that has been wrapped around a center line so that it can
be rotated to generate g's.
By looking at the design of a centrifuge as a clarifier, several design
improvements can be incorporated.
Weirs
For example, redesigned overflow weirs reduce material turbulence
at the liquid overflow.
Conveyor Speed
Control
The installation of an eddy current brake controls conveyor speed
(i.e., cake removal rate) and makes full use of the solids
compaction volume.
Surface Area
Feed System
Cake Removal
Weirs
Cake Compaction
„. , ,, _ , Volume
Sidewall Depth
In sum, the centrifuge's basic design elements, which are like a
clarifier's, can be refined to take full advantage of surface area,
detention time, weir design, and other factors.
91
-------
replaced with G, the centrifugal force produced on
the particle by the rotation of the bowl.
The centrifugal acceleration force (G) defined as
multiples of gravity is a function of the rotational
speed of the bowl and the distance of the particle
from the axis of rotation. In the centrifuge, the
acceleration force, G, is calculated as follows:
G = (2nN)2R-r- 60
(7-13)
where,
N = rotational speed of centrifuge, rev/s
R = radius of rotating body of liquid, m
The value of G can be substituted for g in the earlier
equation to determine the rate of settling velocity as a
function of R. This theory presents the basis to argue
for a very shallow liquid film for settling. In practice,
this must be modified by sludge compaction
requirements, turbulence zones, clearance
requirements between solids and clarified liquid
zones, etc., such that medium to deep pools are now
used for sewage sludges. Heavier solids, like CaCOa,
can be efficiently dewatered using shallower pool
depths since they readily compact and are easily
conveyed (scrolled) out of the machine.
The centrifuge has three functions that are not
entirely compatible. The first is clarification or removal
of solids from the liquid suspension, and the second
is consolidation of the settled particles against the
bowl wall. Lastly, it is necessary to convey and further
dewater these solids during their transport out of trie
bowl. At a specific gravitational force, the
effectiveness of clarification and solids concentration
will each be a function of the pool volume, a large
volume favoring good clarification and a small volume
favoring high solids concentration. That is, detention
volume available for clarification will affect the
hydraulic rate, separation efficiency and, perhaps, the
chemical dosage. However, maximizing the
clarification volume (minimize sludge volume, hence
sludge depth) reduces the time for the solids to
compact and a lower sludge concentration must be
the result. Generally, the operating mode is a
compromise between objectives. If scrolling (moving)
of the solids is difficult because the movement needs
to resuspend the particles, it may be necessary to
compromise both the clarification rate (reduce feed
rate) and solids concentration (increase pool depth).
The clarification capacity of a solid-bowl centrifuge
has historically been measured by its Sigma (£) value
as defined by Ambler with certain assumptions (12).
The Sigma value is essentially the averaged surface
area of a settling tank equivalent to the sedimentation
capacity of the centrifuge, and it is given by the
following formula:
S = 2n L (w2/g) (0.75 r^ + 0.25 r22) (7-14)
where,
£ = theoretical hydraulic capacity, m2
L = effective clarifying length of centrifuge bowl,
m (inlet to liquid outlet)
w = angular velocity of centrifuge bowl, rad/s
g = acceleration due to gravity, 9.8 m/s2
(2 = radius from centrifuge centerline to the liquid
surface in the centrifuge bowl, m
H = radius from centrifuge centerline to the inside
wall of the centrifuge bowl, m
A simplified method of calculating Sigma that is
applicable only to a solid-bowl centrifuge is given by
the equation:
S = Pvw2/g [In (r2/ri)]
(7-15)
where,
Pv = the pool volume, m3
The Sigma value can be used to estimate relative
performance characteristics between centrifuges but
has limitations (12). It can be and is used for scaling
up of results from machines of comparable operating
conditions and physical configuration. However, the
use of Sigma to compare a high-G centrifuge
capability to a low-G capability (or ones where
relative pool depths are markedly different) can be
and often is invalid. The scale-up also may be
modified by the manufacturers' experiences with the
specific machines involved.
Scale-up of a smaller test machine to a larger,
similar solid-bowl centrifuge using Sigma is shown in
Figure 7-11. The use of Sigma is based on an
application where the centrifuge is only clarification
limited and solids capacity is unquestioned.
Clarification capacity can generally be enhanced
using chemicals to help flocculate and settle the
particles. The effectiveness of polymers and the
greatly increased clarification area (longer bowls)
have generally resulted in the sludge consolidation -
- transport being the limiting process rate factor.
This is particularly true when high solids
concentrations are desired or required for a
cost/effective sludge handling operation. Sigma can
be the limiting factor when handling dilute feeds.
The solids limiting capacity of a solid bowl centrifuge
has been designated as the Beta value, and it is used
like Sigma. Solids capacity of a centrifuge is reached
at the point where the compacted sludge volume in
the pool interferes with the solids-liquid separation
and solids recovery declines. The total pool volume
(V) in a centrifuge is equal to (see f:igure 7-11):
V = n
- r22) L
(7-16)
92
-------
Figure 7-11. Sigma scale-up procedure.
Centrate
r/f.'/l/t,/ MTTT
i/li/.MLJLig
•?
Feed
Q
Solids
= 2nL
Test Centrifuge A
L = 0.861 m
r, = 0.178 m
ra = 0.143m
co = 366 rad/s
IA = 2,134 m2
QA = 9 mVhr
QB =
' + 0,25^)
Test Centrifuge B
L = 1.625 m
r, = 0.318 m
r2 = 0.254m
co =262 rad/s
SB = 6,574m2
QB = ?
= 27.7 m'/hr
The volume of sludge, Vs, in the pool with a surface
at r3 is:
Vs = n(n2-r32)L
(7-17)
The solids flow into a centrifuge is defined as Qs kg
TSS/hr, and the volumetric flow of the solids by
QS/YC, where Yc is the specific weight of the
compacted solids, kg TSS/m3 in the cylindrical
portion of the bowl. If there is no slippage of solids in
the bowl, then the particle travel time (T) can be
calculated from a formula for the conveyor or scroll
as:
T = UAw SN
(7-18)
where,
L = length of cylinder, m
Aw = differential speed, rad/s (speed at which
solids are conveyed out of centrifuge)
S = spacing between conveyor blades, m
N = number of leads
The volume occupied by solids during steady state
operation is:
V =
or
Vs = (QS/YC) ( UAw SN)
(7-19)
Since the surface area of the inside wall is A =
2nr\L, then the depth of the cake, y, is:
y = (VS/A) - [Qs/Yc 2Awnri SN]
VS/V =
-------
the sludge with time, can provide misleading results.
It has been reported (13) that the use of
formaldehyde at a dosage of 5 mg/l prevented
changes in the dewatering characteristics of the
sludge.
Where small, continuous field-scale tests are
practical, there are some limitations in terms of
magnitude of the scale-up. It is recommended that
the scale-up factor using Beta and Sigma not
exceed 3 whenever possible.
Since there are substantial differences in the design
approach to the load chambers of high-G machines,
it is suggested that comparative field tests provide the
most useful information for design. This is particularly
true when the cake solids concentration could have a
significant impact on the economics of downstream
operations. If tests involving two machines are
contemplated, they should be run concurrently.
Side-by-side operation will alleviate any concern
that the sludge characteristics changed during the
tests between the two machines. It is further
recommended that the test machines be similar to the
full-scale units planned. The use of torque controlled
back-drives, for example, should be included if that
is the plan for the full-sized machines. The testing
range of flow and solids rates tested should be
adequate to provide a full description of the operating
characteristics of the machine. Ideally, these tests
should also be conducted during the colder months of
the year, since that will be the most difficult time to
dewater the sludge. If it is done during warmer
periods, then the cake solids should be discounted to
account for cold weather operation.
Since the feed solids are split between the centrate
and the cake, it is necessary to use a recovery
formula to determine solids capture. Recovery is the
mass of solids in the cake divided by the mass of
solids in the feed. If solids contents of the feed,
centrato and cake are measured, it is possible to
calculate percent recovery without determining total
mass of any of the streams. The equation for percent
recovery Is given below:
R = 100 (CS/F) [(F - CC)/(CS - Cc)] (7-24)
where,
R = recovery, % TSS
Cs = cake solids, % TSS (or TS)
F = feed solids, % TSS
Cc = centrate or overflow solids, % TSS
7.3.5 Centrifuge Components, Operation and
Control
While not specifically part of the centrifuge, the
foundation upon which it rests is an important design
consideration. The base provides a solid foundation
on which to mount and support the centrifugal unit.
Vibration isolators, normally mounted between the
base and foundation, help reduce the vibration
created by the centrifuge. The base is normally
fabricated steel or cast steel of sufficient mass to
sustain vibration and reduce harmonic effects caused
by minor imbalance.
The centrifuge's case serves as a guard, protecting
the rotating assembly and reducing the noise level. It
also contains and directs the cake solids and centrate
as they are discharged from the rotating assembly.
The case may be fabricated from carbon steel and
coated, but the cake and centrate discharge housings
should be of stainless steel of SS316 quality or
better.
The variables listed below will be discussed in this
section. Many are preset by the manufacturer; some
can be controlled by the operator.
Bowl diameter
Bowl length
Bowl rotational speed
Beach angle
Beach length
Pool depth
Scroll rotational speed
Scroll pitch
Feed point of sludge
Feed point of chemicals.
Condition of scroll blades
7.3.5.1 Bowl and Conveyor
The bowl configuration is cylindrical and conical in
shape, though the proportions of each section will
vary depending on manufacturer and application. The
angle and length of the conical section, which acts as
the dewatering beach, have an important affect on the
performance for individual applications and will be
specific to each manufacturer's machine. The bowl
diameters of dewatering centrifuges range up to 183
cm (72 in) with bowl lengths of up to 427 cm (168 in)
in 1986. In 1975, the largest units in the United States
were 91 cm (36 in) diameter x 244 cm (96 in) long.
The scroll, a helical screw conveyor, is fitted
concentrically into the bowl. The central core of the
scroll contains feed tubes and ports for the discharge
of the centrate. Design of the scroll may vary
considerably in terms of the pitch diameter, number
of conveyor leads, and in openings for the centrate to
pass through to the discharge weirs. The scroll may
contain baffles to prevent the incoming feed from
disturbing the previously consolidated cake in a
countercurrent centrifuge. Improved conveyor designs
are often jealously guarded secrets and the final
decision regarding conveyor details must be left to
the manufacturer.
The material of construction for the centrifuge bowl
and conveyor ranges from high strength carbon steel
94
-------
to common stainless steels such as SS316 and
SS317. Special alloys are used for some applications,
but normally this is not required for municipal sludges.
Where there is a concentration of chlorides above
200 mg/l, there is concern for chloride stress
corrosion of stainless steels. Plants receiving chloride
wastes or infiltration of sea water may have serious
mechanical problems if this is not considered.
Abrasive wear on scroll conveyor blades or flights has
traditionally been the item of greatest maintenance. It
is influenced by sludge abrasiveness, the centrifugal
force at the bowl wall, the differential speed, and the
abrasion resistance of the material used to form scroll
blade tips. Figure 7-12 shows the various types of
hardfacing that have been used to reduce wear on
scroll tips. These include many different welder
applied metallic hardfacings (such as Colmonoy #6,
Eutalloy, and Stellite) as well as tungsten carbide and
ceramic tiles. Field replaceable ceramic tiles have
recently been recommended by low-G centrifuge
manufacturers because of their long life, relatively low
replacement cost, and ease of replacement.
However, they are more fragile than metallic
hardfacings, tending to chip easily. They also may
occupy more space in the bowl and do not form as
smooth a surface on the conveyor blades as do
metallic hard facings. Ceramic tiles can be glued onto
the flights although in some cases they are both
glued and bolted to the flights. One manufacturer of
low-G centrifuges using ceramic tile hardsurfacing
material will routinely guarantee scroll conveyor life
for 15,000 to 20,000 hours between rebuilds.
Figure 7-12. Four different types of hardfacing used to
retard scroll wear.
Bird Ceramic Tile
Impco Hardfaced Insert
Sharpies Stellite Tile
Sintered Tungsten
Carbide Tile
The Severn Trent Water Authority in Birmingham,
England reported in 1981 that the original ceramic tile
conveyor lasted 25,000 hours and the replacement
units had operated 27,000 hours and were still
satisfactory (T. Wood, Severn Trent Water Authority,
Birmingham, England, personal communication,
1981). The machine averaged 165 hours per week
operation and was a low-G centrifuge. Similar
experiences with ceramic and sintered tungsten
carbide tiles have been reported in the United States
at San Francisco, Oakland, CA, Port Huron, Ml, and
Lorain, OH.
Sintered tungsten carbide tiles have demonstrated
useful lives greater than 30,000 hours, but they are
generally more expensive than ceramic tiles.
However, the cost may decrease as additional
suppliers and refurbishing plants employ these
materials. Sintered tungsten carbide tiles are
generally welded to the flights and are usually
required for only the portion of the conveyor blade
near the dewatering beach. One high-G centrifuge
manufacturer warrants scroll conveyor life for 30,000
hours using highly abrasion resistant sintered
tungsten carbide tiles. Experience with low-G
concurrent flow centrifuges at the Los Angeles
County Sanitation District's Carson Plant has
indicated that conventional welder applied hardfacing
has an operating life of only 5,000 hours.
Bowl and scroll geometry varies considerably from
one manufacturer to another. In general, increasing
the bowl diameter will increase both the capacity of
solids conveying and clarification. On the other hand,
an increase in the bowl length improves only
clarification capacity.
The beach angle is usually kept at 8 to 10 degrees to
help prevent slippage of the conveyed solids. As the
solids emerge from the pool, the buoyancy effect is
lost and it becomes more difficult to convey fine,
hydrous, and soft solids, such as waste activated
sludge against the high G forces of the centrifuge.
Shallow beach angles, deep pools, and conveyor
design configurations also work together with a
hydraulic effect, which essentially helps push the
settled solids up the beach and eliminates slippage
problems. Although a shallow beach angle increases
conveyor capacity and improves centrate quality, a
wetter cake is produced due to the loss of beach
drainage area. Conversely, a steep beach angle
produces a drier cake but at the expense of centrate
quality and conveyor capacity. With a steep beach
angle, there are higher resistive forces to conveyance
as shown in Figure 7-13. Albertson and Guidi (14)
reported that the force of slippage can increase 10-
fold at the pool-beach interface. As a result, a
portion of the solids are generally resuspended or
leaked through the beach and are ultimately lost over
the centrate weir. The ultimate decision of the beach
angle depends on the relative importance of greater
cake solids or centrate quality.
95
-------
Figure 7-13. Effect of bowt angle on the movement of
sludge.
Center of Rotation
G =» Centrifugal Force
g = G sin a = Slippage Force
Bowl speed is normally not varied on most centrifuge
models once the unit is installed. The solid-bowl
centrifuge operates at speeds equivalent to 600-
3,000 times the force of gravity and are categorized
into low- and high-G centrifuges. Low-G units
have operating speeds equivalent to 600-1,800 Gs,
and high-G units operate at 2,000-3,000 Gs. The
gravitational force is directly proportional to the bowl
diameter and the square of the bowl speed. Thus,
since the G force takes into account both bowl speed
and bowl diameter, it is a better method of describing
solid-bowl centrifuges than bowl speed alone.
The question of which type of centrifuge, high-speed
or low-speed, performs best cannot be answered
generally. The results have been site and sludge
specific; higher gravitational forces have produced
slightly poorer to significantly better cake solids and
recovery. Sludges with higher structural
characteristics, as with a paper fiber content, will
generally respond well to a higher gravitational force,
while weak structured cakes may even respond
negatively to increased gravitational forces.
Arguments for the low-speed, concurrent centrifuges
have been as follows:
» Feed introduction at far end of bowl reduces
turbulence, improves clarification.
« Feed introduction does not disturb partially
consolidated sludge.
• Chemical consumption is lower due to less
turbulence.
» High solids and clear centrate can be produced at
lower speeds without loss of machine capacity.
• Lower speed requires less power.
» Lower speed reduces wear and other O & M costs.
• Lower speed produces less vibration.
Arguments for high-speed, countercurrent
centrifuges have been as follows:
* Speed can always be reduced if not required.
• Most municipal sludges are not tested prior to
installation of the centrifuge to determine the
performance characteristics as a function of speed
or G force.
• Concurrent centrifuges will wear the full length of
the bowl and scroll, whereas countercurrent units
wear only a portion of the bowl, resulting in less
repair costs per repair interval.
In general, the high-speed, countercurrent
centrifuges will be smaller in size than the concurrent,
low-speed unit of similar hydraulic and solids
capacity. Some operating/design principles also differ,
and further, it is not always possible to reduce the
gravitational force since the higher speed would be
required to maintain clarification capacity. Without
testing, some of these considerations will not be
readily apparent.
There are specific benefits to both design concepts
that often can be determined only by side-by-side
evaluations. A basic rule to follow is: operate at the
minimum speed possible that still meets the capacity
and other performance characteristics necessary for a
cost-effective dewatering system.
The City of San Francisco has both high-G (HSC)
and low-G (LSC) centrifuges dewatering
anaerobically digested primary and pure oxygen
activated sludge (G. Davies, City of San Francisco,
personal communication, 1983; J. Loiacono, City of
San Francisco, personal communication, 1986). Also,
there is an existing vacuum filter (VF) station.
Average results reported by the City are provided in
Table 7-5. Additional results and comments
regarding operation and performance are found in
Chapter 9.
Table 7-s.
Sludge Dewatering Performance at San
Francisco (1982-1983)
HSC
LSC
VF
Feed, l/s
TSS, kg/hr
Cake, % TS
Recovery, % TS
Polymer, kg/Mg
Power, kWh/m3
Gravities, G
7,32
634
23,6
96.9
4.1
1.9
1,880
8.64
740
22.7
96.7
3.6
1.1
1,000
-
-
16.2
88
-
-
-
HSC - 74 crn diameter x 208 cm L (29 in x 62 in)
LSC - 90 cm diameter x 225 cm L (35 in x 96 in)
The difference in cake solids was negligible
particularly when one considers that the LSC was
handling 17 percent more solids and operating at a
lower polymer dosage. At San Francisco, the solids
content of the cake is not strongly affected by the
gravitational force.
96
-------
Extensive field trials (15) were conducted at the
Littleton/Englewood STP, Colorado with a high-
speed and a low-speed centrifuge for thickening
WAS (TWAS) and dewatering digested primary and
waste activated sludge [D(P + WAS)]. The results of
the tests are summarized in Table 7-6.
The HSC was able to thicken the WAS to 7 percent
TSS without polymer addition while the LSC required
polymer at a cost of $4.40/Mg DS to achieve 7.0
percent TSS at a higher recovery. The results on the
digested sludge were very similar. However, it was
necessary to operate the LSC at higher than the
customary G force of 1,000-1,400 to achieve
equivalent performance. This sludge was amenable to
the use of higher G forces to maximize performance.
Centrifuges require a stationary feed pipe, which is
inserted through the center of the scroll housing for a
distance necessary to reach the discharge port. The
feed pipe may enter the centrate or the cake
discharge end of the centrifuge depending on the
machine size and design.
Polymer conditioning of sludge could still be termed
an art. Thus, it is important to have maximum
flexibility for polymer addition. Provisions for adding
polymer before and after the feed pumps, centrifuge
inlet and into the centrifuge feed discharge ports
through an independent polymer feed line should be
provided as part of the design. See Chapter 5 and
Appendix B for greater discussion of Chemical
Conditioning.
7.3.5.2 Differential Bowl and Conveyor Speed
Assembly
The gear unit is generally of a planetary or cyclo-
gear type and works together with the backdrive to
control the differential speed between the bowl and
the conveyor. By controlling the differential speed,
optimum solids residence time in the centrifuge and
the cake solids content can be provided. A backdrive
of some type is considered essential when
dewatering raw or digested primary - secondary or
secondary sludges due to the presence of fine
particles. The backdrive function can be
accomplished with a hydraulic pump system, an eddy
current brake, DC variable speed motor, or a Reeves
type variable speed motor. The two most common
backdrive systems are the hydraulic backdrive and
the eddy current brake.
The control of the bowl-conveyor speed differential
has had a series of evolutionary changes over the
past 25 years. The early centrifuges were provided
with a fixed gear ratio, which, in turn, fixed the ratio of
the bowl speed to the conveyor speed. An
improvement was the use of an auxiliary motor, which
controlled the speed of the previously fixed output
shaft of the gearbox. The eddy current brake
backdrive is now provided on high-G centrifuges.
The eddy current brake is attached to the pinion shaft
of the gearbox and consists of a stationary field coil
and a brake rotor on the shaft. When a DC voltage is
applied to the stationary field coil, magnetic flux lines
are created in the brake rotor. The amount of flux in
the rotor is a function of the speed differential
between the rotor and the field coil as well as the DC
current applied to the field coil. This flux produces
eddy currents, which create a resistance to turning,
or a braking action. Thus, varying the DC voltage
applied to the stationary field coil will change the
speed differential between the bowl and the scroll.
While the eddy current backdrive differential can be
easily set by the operator, it is still possible to
overload the conveyor and cause blockage of the
centrifuge due to overtorque.
The most versatile backdrive arrangement is a
hydraulic pump design. This arrangement is widely
used since it completely eliminates the need for a
gearbox and a mechanical or electrical backdrive. The
hydraulic unit is now employed by many centrifuge
suppliers since it can assure that the centrifuge
conveyor is not overloaded and can maintain the
bowl-conveyor differential at the optimum level. A
unit manufactured in Switzerland is used by many
centrifuge manufacturers to provide automatic
differential control. It has a low-speed hydraulic
motor that drives the centrifuge scroll independently
of the bowl. A pump unit powers the hydraulic motor
and its control system senses the scroll torque and
regulates the differential speed to prevent blockage.
The hydraulic scroll speed control can operate at
lower differential speeds than the fixed gearbox
differential. As a result, an increase in cake dryness
and recovery will be possible. Lower differential
speeds also reduce the scroll tip speed, which, in
turn, mitigates the wear on both the conveyor and the
bowl shell or facing strips. Since any change in solids
loading, hence torque, will be automatically
compensated for by a change in differential speed,
the operation of the centrifuge can be tuned for
maximum retention of solids within the bowl and
scroll, without the risk of choking the machine. That
is, the differential speed is increased (or decreased)
in proportion to an equivalent change in torque.
Furthermore, increased throughput is possible, since
the automatic torque-related scroll speed controller
can allow the feed rate to be increased without
danger of plugging.
An all-hydraulic system eliminates the problem of
gearbox failure and another problem, torsional
vibration or chattering, is minimized. Reliability and
improved performance are the major reasons that the
all-hydraulic system is widely employed. Since the
scroll is directly driven by the adjustable volume
hydraulic system, it also eliminates the need for a
separate backdrive/eddy current brake system, a
common maintenance problem. As the feed solids to
the machine change, the scroll backpressure is
97
-------
Tabfa 7-6.
Centrifuge Tests at Littleton/Englewood STP, Colorado
HSC
TWAS
D (P + WAS)
HSC - 42.5 cm diameter x 125.7 cm L (16.75 in x 49.5 in)
LSC - 45.7 cm diameter x 134.6 cm L (18 in x 53 in)
LSC
TWAS
D(P + WAS)
Food, l/s
Cako, % TS
Rocovory, % TSS
Polymor, S/Mg DS
Power, kWh/Mg
Gravities, Q
Operating Costs, $/Mg
3.15
7
75
0
364
2,500
15.45
3.79
18
90
14.55
472
2,500
20.00
3.15
7
85
4.40
273
1,185
13.64
3.79
17
95
14.55
364
1,800
18.18
continuously monitored and the analog hydraulic unit
automatically increases or decreases the scroll speed
according to pressure. Because the scroll differential
speed is independent of the bowl speed, there is no
loss of torque in the event that the main drive motor
shuts down. In contrast, the gearbox-driven
centrifuge must contend with torque- related
problems. Once the torque limit is reached in a
gear-driven centrifuge, either a shear pin breaks, a
clutch disengages, or a current relay trips, thus
causing loss of torque and differential speed, and
increasing the risk of the machine plugging.
The hydraulic oil pressure is a direct measure of the
torque; as such, the torque loading on the scroll can
be read at any time and can also be used for feed
rate control. The control unit can also be used to
control polymer dosage based on the clarity of the
centrate. A representation of the one unit's electronic
monitoring box is pictured in Figure 7-14.
Controlling Bowl/Scroll Differential Speed to Optimize
Performance
The differential speed controls the solids residence
time within the centrifuge and thus, it can greatly
influence cake concentration and machine capacity.
Increasing the differential between the bowl speed
and the scroll speed normally results in a wetter
sludge cake and higher machine throughput.
Conversely, a decrease in the differential speed
produces a drier cake and decreased machine
throughput and may result in poorer recovery or
higher polymer dosage to maintain recovery. One of
the problems of operating at too low a differential
speed is the creation of a pile of solids in front of the
scroll conveyor blades, allowing some of the fine
solids to be skimmed from the top of the cake into
the centrate. Another danger is plugging the
centrifuge if solids are removed at a slower rate than
they are fed to the machine.
A backdrive unit can generally provide an increase in
cake solids content of 4 percent or more relative to a
comparable machine without a backdrive. The
Figure 7-14. Bowl/scroll differential speed monitoring box.
u u u u
HydrnultkDnicK ' HYdRPfiESSUflE t
, f , ^ ,it,» ,'tfy, i ^
VISCOTHERM AGjrvi»E704 'lt\l^l
backdrive increases the overall stability of the
centrifuge performance when the feed solids
characteristics vary. Retrofitting a centrifuge with an
analog hydraulic backdrive is cost-effective. At
Seattle Metro wastewater treatment plant, the existing
centrifuges were modified to provide torque controlled
conveyor differential speed. The cake solids
increased 3-5 percentage points and the estimated
savings in haulage were $39,000 per month. The test
results of a study conducted by Seattle Metro (16)
are shown in Table 7-7. As shown in the Seattle
studies, the polymer cost can increase when
operating at the minimum differential speed. The
lower differential speed produced auto torque control,
the sludge inventory was higher, and more polymer
was used to control effluent solids. The cost of
98
-------
Table 7-7. Performance of Fixed vs. Variable Torque Controlled Backdrive at Seattle Metro
Differential Control Feed Rale Cake TS Recovery
Haulage Savings @ 50.8 m3/hr = $39,000/month.
Liquid Polymer
Fixed Differential
Automatic Backdrive
Fixed Differential
Automatic Backdrive
Fixed Differential
Automatic Backdrive
m3/hr
50.8
50.8
45.2
45.2
37.5
37.5
percent
16
20
16
21
16
19
percent TSS
80
81
85
86
85
86
kg/Mg
103
106
62
77
45
72
haulage, fuel, etc., must be balanced against
chemical cost.
Tests were also conducted on a 74 cm (29 in) dia. x
208 cm (82 in) long centrifuge in Columbus, OH.
Providing torque directed conveyor speed control
manually produced a cake that was 4-7 percentage
points drier than other units on-line which were
handling the same feed solids. Similar results have
been experienced elsewhere.
7.3.5.3 Miscellaneous
Overflow Weirs
Although the pool depth is variable on solid-bowl
units, several hours of labor may be required to adjust
the overflow weirs. As such, it is not a popular
method of operational control. The pool depth
regulates both the quality of clarification and the
dryness of solids. Thus, while increasing the pool
depth will normally result in better solids recovery at a
specific feed rate, the cake produced will be wetter.
However, it may be necessary to adjust the weirs if a
major change in feed rate is required. When the pool
depth is translated to residence time in the centrifuge,
there may be little or no difference in the recovery
and cake solids. However, this is not necessarily true
for all types of waste sludges, nor for pool depths that
leave little or no dry beach. Over the years, pool
volumes have increased to accommodate the greater
space occupied by solids under longer retention
times. Additional depth and greater cross-sectional
area reduce turbulence and permit solids to become
compacted without interfering with the clarification
process. The use of deep pool operation in thickening
and dewatering waste activated sludge minimizes the
slippage force on the beach, resulting in improved
conveying efficiency.
Sludge Feed Pumps and Piping
Control of sludge feed rate demands a sludge
pumping system that can handle varying sludge
consistencies and centrifuge loadings. For this
reason, progressive cavity pumps are the
overwhelming preference of centrifuge designers.
Lobe pumps also provide a steady flow, are positive
displacement, and thus suitable for centrifuge feed
pumps. Centrifugal pumps, on the other hand, are
less adaptable to changes in sludge consistencies.
Varying sludges affect the pumping rate and, as such,
appropriate flowmeters and controllers - and
possibly variable speed drives - are recommended
for positive control of centrifuge loading.
7.3.6 Process Variables
In addition to machine variables, there are a number
of process variables that affect the performance of a
centrifuge. These variables are listed below.
Feed rate
Sludge characteristics
Particle size
Particle shape
Rheology
Solids concentration
Liquid viscosity
Liquid density
Temperature
Type of chemicals added.
Amount of chemicals added
7.3.6.1 Feed Rate
One of the most important control variables during
centrifuge operation, as already noted, is the feed
rate of the centrifuge, both from a hydraulic and
solids loading standpoint. The hydraulic load to the
centrifuge affects the clarification ability, while the
solids loading is a function of the conveying
capabilities. Increasing the hydraulic load will
decrease the centrate clarity and may increase the
chemical consumption. A corresponding change in
the differential speed is required when changes in
solids loading occur if the centrifuge is initially
operating at maximum solids residence time. The
most concentrated cake is achieved at minimum
differential speed and at a feed rate to match the
reduced volumetric conveying capacity.
7.3.6.2 Sludge Characteristics
The identifiable characteristics of the various sources
of waste solids have an impact on the dewatering
efficiency as measured by unit capacity, product
dryness, and solids recovery. Thus, the design of the
wastewater treatment plant is an important
consideration in the sizing of the centrifuge and the
expected performance characteristics.
99
-------
Larger and heavier particles are most easily captured
by the centrifuge. Finer particles that cannot be
settled separately must be agglomerated by
chemicals to a size that will settle in the pool to the
outer wall, where they can then be conveyed to the
discharge point. As the proportion of finer particles
increases, the sludge becomes more difficult to
flocculate and requires increasingly higher dosages of
chemicals to maintain a high capture of the feed TSS.
Also, as the proportion of finer particles increases, the
cake moisture content will also increase. If the finer
particles are also hydrous, as is the case with
activated sludge or alum sludge, the moisture content
can increase significantly. The sludge cake produced
will also change characteristics; it will have much
more of a thixotropic nature as the proportion of fine
and hydrous particles increases. Also, the sludge will
be more plastic when the WAS fraction increases.
The sludges with a high proportion of fine and
hydrous particles will also have poor structural
characteristics. That is, the solids will have a
tendency to flow. This characteristic will affect the
conveying of solids from the centrifuge. As the sludge
becomes more fluid, it will resist being conveyed up
the slope of the conical portion of the bowl to the
discharge point. If the upward frictional force of the
conveyor is less than the centrifugal force on the
solids, the solids may slip back into the pool.
The ash content of a sludge affects the final cake
solids. Generally, about the first 10-15 percent of
the inerts in the sludge are associated with the
organics and thus have little impact. However, as the
ash content exceeds 25 percent, a definite
improvement in cake solids is noted. The added
inerts will usually be fine silt, which dewaters readily
and thus produces a higher cake solids. When
comparing operating data, the designer must evaluate
the possible effects of inert content. The Sludge
Volume Index (SVI) of the secondary sludge can also
have a profound impact on both the feed rate as well
as the dewatered cake concentration. Vesilind and
Loehr (17) found that the centrifuge capacity was
impacted by SVI and their results are shown in Rgure
7-15. The higher values of SVI will also produce a
wetter cake.
7.3.6.3 Chemical Conditioning.
As is described in detail in Chapter 5 and Appendix B,
both inorganic and organic chemicals are used for
dewatering applications. For the most part, solid-
bowl centrifuges use organic polyelectrolytes for
flocculating purposes. Polymer use improves centrate
clarity, increases capacity, often improves the
conveying characteristics of the solids being
discharged and often increases cake dryness. Anionic
polymers may yield a better operation if aluminum or
ferric salts are present. A number of centrifuge
installations are using FeCIa in conjunction with
cationic and anionic polymers. It may be necessary to
Figure 7-15. Effect of SVI on centrifugal dewatering of
activated sludge.
400 r
8
>
tn
300
200
100
50 100
Sludge Volume Index, ml/g
150
200
use a dual-polymer system if there is polymer
treatment upstream of the centrifuge.
7.3.6.4 Sludge Temperature
Warm sludges will dewater better than cold sludge.
Winter to summer sludge cake concentrations may
vary by as much as 2-4 percentage points. The
probable reason for the improvement in the summer
cake is the decrease in liquid viscosity, which
improves the liquid drainage. Another reason for the
drier cake, though perhaps to a lesser degree, is the
decrease in liquid density during warmer
temperatures.
Secondary sludge quantities and proportions increase
in winter, which, in turn, increase the moisture.
Heating of the sludge will significantly improve the
cake solids. However, this practice is rarely found to
be feasible, unless there is an available source of
usable waste heat, generally steam. One pound of
steam is needed to heat 10 pounds of sludge to
about 66*C(150°F).
7.3.6.5 Performance Characteristics
Primary to secondary ratio will have a profound effect
not only on the capacity of the centrifuge, but also on
the cake concentration and the polymer dosage.
Further, as mentioned elsewhere, the SVI of the
secondary fraction will also have an impact on these
same parameters. Even if the sludge is digested,
some of the effects of the primary to secondary
sludge ratio and the SVI appear to carry through,
affecting the centrifuge performance as well as other
mechanical dewatering equipment. Wherever
possible, the plant design should be directed toward
100
-------
producing a minimal amount of secondary sludge.
This strategy will enhance the performance of the
centrifuge as well as reduce the operating costs. It is
difficult to compare data from different locations
unless the primary and secondary sludge ratio and
the SVI characteristics of the secondary fraction are
known. Further, addition of chemicals for phosphate
removal complicates this comparison and may make
any conclusion invalid. The performance
characteristics of a solid-bowl conveyor centrifuge
on various sludges are provided in Table 7-8.
Mixtures of sludges can be estimated on a weighted
mass basis; this is assuming that each fraction
dewaters proportionately to its weight in the mixture.
The procedure is as follows:
RPS 100 kg @ 30% TSS =333 kg
RWAS 80 kg @ 16% TSS = 500 kg
AI(OH)3+AIPO4 30 kg @ 14% TSS * 214kg
210kg 1,047kg
Cake TS = 210/1,047 = 20.1% TS
Polymer requirements can be determined in the same
manner. The higher cake concentrations are generally
achieved with the more favorable ratio of primary
sludge. With the chemical sludges resulting from
phosphorus removal, the range of solids can be due
to a number of factors not well understood. The
higher ratio of hydroxide precipitates will tend to
reduce the solids content and part of the cause is
generally an increasing fraction of secondary solids.
Centrifuge cake solids can be increased (2-5
percentage points) by using excessive polymer
dosages, i.e., dosages above that necessary for 90-
95 percent TSS recovery. Polymer costs are the
lowest when the machine is running at a reduced
capacity. Maximizing the capacity of the centrifuge
will not only increase the polymer cost but also
produce a wetter cake.
7.3.7 Pre-treatment
The sludge, prior to being pumped to the centrifuge,
should be ground into a particle size in the range of
0.64 cm (1/4 in) or smaller. This is particularly true for
the units in smaller treatment plants, which may have
relatively small openings for feed inlets and
discharge. In the very large plants, grinding of the
sludge may not be necessary.
While materials of construction for abrasion
resistance have greatly improved, good grit removal
should be incorporated into the plant design. At a
minimum, plus 65 mesh grit should be removed at
peak flows entering the plant. This means that at
normal flows, removal of plus 100 mesh should be
readily achieved.
The best operation of the centrifuge will be achieved
if the feed rate of solids loadings is relatively steady.
In section 7.3.3, the benefits of preconcentration of
the sludge using high rate thickening was discussed.
The result of minimizing fluctuations will generally be
higher solids concentration, better recovery, and
lower polymer dosages. Operator attendance will also
be minimized.
7.3.8 Cake Solids and Centrate Handling
Depending upon the application of the sludge
characteristics and composition, the cake structural
characteristics will vary widely. The cake may vary
from a wet, sloppy mass to a relatively dry, firm solid
mass, and in some cases, the solids will be a loose
bulked product. Due to the high energy prior to when
the sludge solids leave the machine, the solids will
generally be massive, even if drier than a vacuum
filter cake, for example. This is due to the thixotropic
nature of the sludge and the conditions under which it
exited the centrifuge. Thixotropic solids lose their
structural integrity when energy, such as vibration, Is
applied. These solids are normally conveyed from the
centrifuge by belt conveyors, screw conveyors, or
specially designed pumps. The pump method has
gained popularity since it is a very clean means of
moving the sludge solids a considerable distance.
Specially designed progressive cavity pumps,
employing separate feeder mechanisms, have been
used to pump centrifuge cake up to 61 m (200 ft),
depending on the sludge concentration. More
recently, ram type pumps, which are capable of
pumping the cake greater distances, are being
employed. These methods are not only effective; they
are also less costly and much easier to maintain than
the conventional conveying mechanisms.
Under normal operating conditions utilizing polymers,
the centrate solids will constitute 5-8 percent of the
feed solids. These solids are normally recycled back
to the head of the plant or to a concentration unit
prior to the centrifuge. These solids should not
constitute any significant additional load on the
clarification devices in the main plant stream. There
will be a BOD associated with the centrate, both from
the solids fraction as well as the soluble BOD
contained in the liquid. It is recommended that the
calculations of the feed sludge to the centrifuge
include a factor of at least 10 percent by volume
increase to account for a recycled load.
The centrate piping, in general, must be of adequate
size and slope to allow for air venting the windage
generated by the centrifuge. The manufacturer should
be consulted for recommendations and the
representative should review and approve the final
drawings before bidding or installing the centrifuge.
The centrate piping must also be sufficiently large to
handle a mixture of foam and water. Often when
polymers are employed, a stable foam can be
produced. If the piping is too small, liquid will back up
into the centrifuge. Liquid backup can be a problem
particularly if there are multiple centrifuges
101
-------
Table 7-8. Centrifuge Performance Characteristics
Sludge Cake TS
Raw Primary
Anaerobically Digested Primary
Raw WAS
Anaerobically Digested WAS
Raw (Primary + WAS)
Anaorobically Digested (P +• WAS)
AHOH)3 + Aff>O4
Fo(OH)3 + FeP04
Exl, AoraUon or Aor. Digested Sludge
Cam(P04)x
CaCO3
percent
28-34
26-32
14-18
14-18
18-25
17-24
12-16
12-16
12-16
12-18
40-50
Solids Recovery
percent
90-95
90-93
90-95
90-95
90-95
90-95
90-95
90-95
90-95
90-95
90-95
Polymer
kg/Mg
1-2
2-3
6-10
6-10
3-7
3-8
1-3
1-3
6-10
1.5-3
0
$/Mg*
5-10
10-15
30-50
30-50
15-35
15-40
5-15
5-15
30-50
7-15
0
* Based on $5/Mg. 40:60 to 60:40 P:WAS mixtures.
discharging into one centrate line. Here again, the
manufacturer should be consulted during review for
approval of the engineering design.
7.3.9 General Equipment Selection Criteria
The criteria used for selecting a centrifuge will be
different depending on the plant size. That is, the
criteria for smaller plants will emphasize reliability
without complex servicing and maintenance while
performance would be secondary. The generalized
major (M) and secondary (S) criteria have been set
forth in Table 7-9.
Toblo 7-9. Evaluation Criteria for Centrifugal Dcwaterlng
Equipment
Plant Size (mgd)
Small Med,
(<2) (2-10)
Quick Startup
Reliability w/o Skilled Service
Minimal Operator Attendance
Parts Available Overnight
Local Ropairability
Cleanliness
Low Noiso Level
Low Initial Cost
Maximum Cake Solids
Recovery Solids >90%
Polymer Cost
Power Cost
Labor Cost
M » Major; MS = Significant;
1 » Excellent; 2 « Good; 3 -
M
M
M
M
M
M
M
M
S
M
S
S
M
S =
Fair
MS
MS
MS
M
MS
M
M
MS
MS
M
SM
SM
M
Secondary
Large
(>10)
S
S
S
S
S
M
MS
MS
M
M
M
M
M
(Minor)
Cent,
Rating
1
2
1
2
2/3
1
2
2
2
1
1
3
1
Small plants generally have low staffing levels, which
are best served by unit operations that require
minimal operator attendance. Large plants can use
one operator to attend several machines and thus the
labor cost per ton can be relatively small. On the
other hand, driest possible cake could be very
necessary for an economical combustion or compost
operation. In this case, a higher capital and O&M cost
for power, polymer, and labor could be offset by fuel
savings and increased capability of the combustion or
compost operations.
7.3.9.1 Capacity
It is not easy to compare capacities and performance
of centrifuges of different manufacturers. As already
noted, when low-speed centrifuges are compared to
high-speed ones of similar capacity, they will be
larger in diameter. While a centrifuge supplier will
have a database that can be utilized to determine the
capacity of these machines, independent analysis of
operating units is recommended.
Appendix C includes tables that represent the bowl
diameters and lengths of various suppliers. With
some units, it is necessary that the engineer be
assured that the selected design unit is still being
manufactured. Further, new units of different sizes
and configurations would normally be added to a
manufacturer's line.
The capacity of sludge dewatering to be installed at a
given plant is a function of the size of the plant,
capability to repair malfunctioning machines on-site
or locally, and the availability of an alternative disposal
means. Proportionally, a smaller plant will have a
higher percentage of standby capacity than a larger
plant. On the other hand, the larger plant may have
two or three spare machines while the small plant has
one spare machine and an acceptable and available
alternative disposal method. Some general guidelines
102
-------
relating the minimal capacity requirements are
incorporated into Table 7-10. This table is based on
the assumption that there is no alternative mode of
sludge disposal and that the capacity to store solids is
limited. These considerations may vary from case to
case and must be considered individually by the
design engineer.
Table 7-10. Suggested Capacity and Number of Centrifuges
Plant Size
mgd
2
5
20
50
100
250
Sludge
Flow"
m3/d
40
50
320
800
1,600
4,000
Operation
hr/d
7
7,5
15
22
22
22
Centrifuges
Operating -t-
Spare
@ m3/hr
1 +1 @6
1 + 1 @ 12
2 + 1 @ 12
2+1 @ 18
3 + 2 @25
4 + 2 @ 45
* Sludge production is about 0.1 kg/m3 (1,600 Ib TSS/mgd).
mgd x 0.0438 = m3/s.
7.3.9.2 High-G vs. Low-G Centrifuge
For the medium to larger size plants, use of higher
gravity rotational forces may increase the solids
content as well as the capacity of a centrifuge.
However, this is not true for all applications, and it is
desirable to determine the ability of the dewatering
operation to load to the high G force range. In any
case, the operation of the centrifuge should be
evaluated at different gravitational forces to ensure
that the minimum G forces are employed for the
specific application. This will extend the operating life
of the machine. The question of whether high G
forces are better than low G forces for dewatering
cannot be answered easily. Solids with high structural
strength respond more favorably to high G forces.
Conversely, if the sludge contains substantial
quantities of fines or hydrous materials, such as alum
sludges and waste activated sludge, then low G
forces can be equally effective. Only side-by-side
tests can determine the comparative results.
7.3.9.3 Differential Speed Control
All centrifuges should be specified with an easily
adjustable, differential speed control device. One
option is the eddy current brake, which provides a
readily adjustable fixed speed between the bowl and
the conveyor. A better device is the hydraulic torque
controlled device, which eliminates the maintenance
associated with a gearbox and provides automatic,
optimum control of the differential speed.
It is generally recommended that, for even the small
machines, an automatic torque-controlled backdrive
be employed. This device will optimize the machine's
performance and produce the best results possible in
terms of cake solids and recovery. Also, this type of
drive is efficient, compensating for the normal
variations found in small plants. The drive's ability to
eliminate centrifuge overloading is sufficient reason in
itself to install one. The machine will shut down if the
bowl and scroll lock together and, at that point, the
feed pump is automatically turned off.
7.3.9.4 Chemical Feed Control
Electronic monitoring of the torque and control of the
speed differential comprise the basic equipment
required to translate centrate clarity into chemical
feed pump control. For example, as the clarity
decreases, as measured by light transmittance, the
polymer dosage will automatically be increased to
maintain the preset clarity. If the chemical dosage is
excessive, the torque setting is manually reduced to
lower the sludge level in the pool. Since centrate feed
control will provide optimal polymer dosing, save
chemicals, and reduce operator attendance, this
device should be incorporated into both large and
small dewatering stations.
While there are devices available that can give a
readout of solids concentration in mg/l, a light
scattering absorption unit connected to the centrate
line will probably provide more than adequate
information regarding the changes in the recovery
rate of the centrifuge. This continuous readout of the
turbidity as a function of polymer doses will help to
improve the dewatering operation as well as provide a
record of the chemical consumption.
7.3.9.5 Abrasion Resistance
While there are applications where normal hardfacing
has provided good life, the average wastewater
treatment sludge is abrasive and, in some cases,
highly abrasive. For this reason, it is recommended
that the ceramic or sintered carbide tiles be used for
sewage sludge dewatering units. These tiles will give
a normal life of 15,000-30,000 hr and are suitable
for the level of maintenance normally found in
wastewater treatment plants.
The manufacturer will also recommend that other
portions of the centrifuge be tiled in order to extend
the life of the entire machine.
7.3.9.6 Materials of Construction
Experience has shown that both carbon steel and
stainless steel will provide satisfactory service.
Stainless steel machines can be more easily
disassembled than carbon steel units after several
years of service. However, tight tolerances on
stainless steel may also be difficult to separate since
the metal tends to grab. Chloride resistance is a
consideration where sea water or higher than normal
chlorides are present in the wastewater. Stainless
steels are usually unsatisfactory for chloride
exposure.
The use of stainless steel or carbon, to a large
extent, depends on the engineer's and user's
103
-------
preferences. There is no strong technical reason to
support either material since both materials provide
satisfactory long-term service,
7.3.JO Performance Data for Solid Bowl
Centrifuges
Table 7-11 includes operating data for a number of
centrifuge installations that were installed since 1980,
The results vary widely, which may be due to the
machine design, solids loading, sludge feed
characteristics, planned operating results, or a
combination of these factors.
Suppliers have advised that there is limited data on
newer machines that have been installed in the period
from 1983 to the present. Since there has been
continued advancement in the area of centrifuge
design, the consulting engineer should carefully
evaluate the new designs and the performance of
these units. Any questions should be resolved by
on-site testing.
7.4 Filter Presses
7.4,1 Introduction
RIter presses for dewatering were first developed for
industrial applications and, until the development of
diaphragm presses, were only slightly modified for
municipal applications.
The original or early models of the press were
sometimes called plate and frame filters, because
they consisted of alternative frames and plates on
which filter media rests or are secured. The frames
provide both structural integrity and spacing between
the plates. The frames could be changed to provide
different cake thicknesses. The unit had a fixed and a
movable end, which promoted pressure maintenance
during the filtration cycle. There are few, if any, plate
and frame units in service for municipal applications
today, because this configuration is not particularly
suitable for the filtration of hydrous pseudoplastic
materials like municipal sludges. However, test filters
sometimes are plate and frame, and they are used to
determine the optimum cake thickness.
The equipment commonly in use for the dewatering
of municipal sludges falls into one of two categories.
The fixed-volume recessed plate filter and the
diaphragm filter press; the latter was introduced within
the last ten years.
A typical fixed-volume recessed plate filter press is
shown in Figure 7-16.
Pressure filtration whether it be in a recessed plate
filter press or in the diaphragm filter press is defined
as confined expression. Precoating the filter media or
substantial chemical conditioning of the sludge is
normally required. This is particularly true for such
difficu!t-to-dewater materials as waste activated
sludge or aerobically digested sludge. A chemical
conditioning station is therefore almost invariably a
part of a facility that uses filter presses to dewater
municipal sludges.
There is usually a higher degree of operator activity
associated with filter presses than with most other
types of dewatering. As a result, filter presses, for the
most part, have been employed in wastewater
treatment facilities > 1.1-2.2 m3/s (25-50 mgd). On
the other hand, this equipment can produce a very
dry cake, probably the driest cake produced by
conventional dewatering equipment. Hence, it has
substantial attractiveness.
7.4.2 Equipment Description
The recessed plate filter press shown in Figure 7-16
consists of a series of plates, each with a recessed
section that forms the volume into which the sludge is
pumped for dewatering. Filter media are placed
against each wall and retain the sludge solids while
permitting passage of the filtrate. The surface under
the filter media is specifically designed to facilitate the
passage of the filtrate while holding the filter cloth.
Sludge is pumped with high pressure pumps into the
volume between the two plates and individual pieces
of filter media. The filtrate passes through the cake
and the filter media and out of the press through
special ports on the filtrate side of the media.
The pumping of sludge into the press continues up to
pressures sometimes in excess of 1,380 kPa (200
psi). When solids and water fill she void volume
between the filter cloths and ultimately no further
filtrate flow occurs, pumping is stopped. Shortly
thereafter, the press is opened mechanically, and the
cake is removed. Practice has separated the
operation of recessed plate filters into two principal
categories: low pressure units and high pressure
units. Low pressure units operate between 350-864
kPa (50-125 psi) as the terminal pressure; high
pressure units operate between 1,040-1,730 kPa
(150-250 psi). Typically, the low pressure units will
terminate at about 691 kPa (100 psi) and the high
pressure units at about 1,380 kPa (200 psi). There
are several ways of maximizing the filtrate removal,
including good conditioning and stepping the
pressure. Stepping is particularly effective for the high
pressure units, using increments of 350-520 kPa
(50-75 psi).
The diaphragm press is a comparatively new device,
having been commercialized in the U.S. in the 1980s.
It operates during the initial filling period, if the sludge
is properly conditioned, very much like a gravity
drainage deck and is able to drain considerable
amounts of water at substantially zero headloss
across the medium. Overall the diaphragm press
operates like the recessed plate press, typically up to
pressures between 690-1,040 kPa (100-150 psi).
104
-------
Table 7-11. Performance Data for Solid Bowl Centrifuges
o
01
Plant
Victor, TX
Pinhoie, CA
SSF, CA
Port Huron, Ml
Oakland, CA
San Rafael, CA
Valdese, NO
San Francisco, CA
Detroit, Ml
Blue Plains, DC
Petaluma, CA
Denver, CO
Sludge Type
Oxid. Ditch
D (P + WAS)
D(P + WAS)
R(P + WAS)
D (P + WAS)
D (P + WAS)
R (P + WAS)
D(P + WAS)
WAS
D(WAS)
D (P + WAS)
D (P + WAS)
Sludge Mixture
P:S:C
percent
100:0
50:50
50:50
40:58.5:1.52
40:60
50:502
45:55
66:342
0:100
40:60
(DP:RWAS)
80:20 (TF)
40:60
Sludge Feed
Rate
gpm/unit
40-60
45-65
140-160
35-61
150
60-70
40-75
125-250
15
150-200
40
900
Sludge Feed
Rate
Ib TS/d
500
3,600
24,000
11,122
140,000
10,000
1,495
145,000
.3
142,809
7,000
248,000
Avg. Feed
Solids Cone.
%TSS
1 .0-4.0
1 .5-2.0
2.2-3.2
2.2-6.9
2.0
2.0-3.5
3.0-5.0
1.0-3.0
1.5-3.5
5.3-6.8
3.5-4.0
2.3-2.5
Avg. Cake
Solids Cone.
%TSS.
17
14
134
20
20
22
18
18
15
17
21
19
Solids
Recovery
percent
98.9
95.0
90.0
94.0
85.0
93.0
90.0
92.0
89.0
98.4
96.0
90.0
Dewatering
Ib/ton DS
26.7
10.0
13-15
9.0
18.0
10.0
5.0
16.0
8.9
50.0
80.0
10,1
9.0
20.0 .
Chemicals
Type
Percol 767
Polymer
Cat. Polymer
Polymer
Cyanamid
E1125
Allied Coll.
FeCi3
Polymer
Percol 757
FeCI3
Galloway
4450 Emul.
Percol 757
Percol 757
Percol 752
Chemical Cost
$/lb
1.65
2.45
2.30
2.65
0.98
1.11
0.12
2.15
2.25
0.12
0.85
1.71
2.25
1.10
$/ton DS
44.06
24.50
32.20
23.77
17.64
11.00
0.60
34.40
20.03
6.00
68.00
17.27
20.25
22.00
R = Raw; D = Digested; WAS = Waste Activated Sludge; P = Primary; S = Secondary; C = Chemical; TF = Trickling Filters.
11986/1987 data.
2 Unknown quantity of chemical P removal sludge in secondary sludge.
3 Centrifuge only used when wasting rate is excessive and continued blending of primary and secondary sludges would overload belt filter presses.
4 Sludge composted with rice hulls requiring wetter cake.
-------
Figure 7-16. Fixed-volume recessed plate filter press (courtesy Eimco Process Equipment Co.).
Filtrate
Discharge
Stationary
Head
Slurry
Feed
Movable Head
Light
Curtain
Control
Panel
Shifter
Carriage
Plate Centering
Guide
Hydraulic
Cylinder
Light Curtain
Cylinder Bracket
The release of water at low pressures helps maintain
the integrity of the floe. After water release appears
complete following the initial filling period, pumping is
stopped and the diaphragm cycle is initiated. The
diaphragm pressure is applied, using either air or
water on the reverse side of the diaphragm, and
pressures up to 1,380-1,730 kPa (200-250 psi) are
applied to the sludge for additional dewatering. In
addition, the confined expression operation, which
follows when the diaphragm pressure is applied
effectively, releases substantial additional quantities of
water (Cake solids will increase 5-8 percent).
A most significant aspect of the diaphragm press is
that its construction and mode of operation allow the
use of organic polymers as an alternative to ferric
salts and lirne conditioning techniques. Although there
still is the same tendency to squeeze sludge into the
media itself, the tendency is reduced by the
elimination of substantial quantities of water prior to
the start of the squeezing operation.
It should be noted that there has been an evolution in
the diaphragm press to a simpler design sometimes
known as a diaphragm plate press. In this design,
both the cloth and the diaphragm are built into the
plate. There fewer moving parts, longer cloth life, and
much lower O&M costs.
Based on typical Filtration operations, it can be
expected that 70-85 percent of the water will be
removed during the low pressure portion of the cycle
of the recessed plate and diaphragm press. Similar
performance can be obtained from a fixed-volume
recessed plate press by stepping the pressure at two
or three intermediate levels. The diaphragm press,
however, usually produces a drier cake than that
obtained from the fixed-volume recessed plate. Also,
there is a substantially greater uniformity of solids
concentration in the cake produced with a diaphragm
press. With the low solids feed material continually
being supplied to the recessed device, a very low
solids cake fraction is produced near the feed point.
This problem, of course, is not present in the
diaphragm press because the pumping cycle is only
the first part of the overall cycle and the diaphragm
tends to remove water uniformly. Also, the cycle time
for a given cake solids concentration is generally less
in the diaphragm press.
There appears to be a less frequent need for
precoating the diaphragm press than is usually
106
-------
encountered with the fixed-volume device. The
implication is that the diaphragm press improves the
dischargeabiiity of the cake due to the higher cake
solids content. Another advantage of the diaphragm
press is that the sludge only needs to be pumped in
at pressures up to, but rarely exceeding, 865-900
kPa (125-130 psi). The higher pressures during the
diaphragm cycle may be supplied by clean water
pumps or air pumps, thereby reducing the overall
maintenance cost associated with high pressure
delivery devices.
While polymers are uniquely successful in
conditioning pure waste activated sludge and mixed
sludges for dewatering in the diaphragm press, it
would be misleading to say that most of the polymer
conditioning success has occurred with these units.
Actually, with mixed primary and secondary sludges,
pure polymer conditioning has been most successful
in low pressure presses operating at 520-1,040 kPa
(75-150 psi), typically using one or two steps to
achieve the ultimate pressure. However the low
pressure recessed plate unit does not provide for the
final high pressure water removal that the diaphragm
press does and this could be the key to better cake
discharge from the cloth. The operating sequences
for fixed-volume recessed filter and diaphragm
presses supplied by different manufacturers are
shown in Figures 7-17 and 7-18, respectively.
7.4.3 Basis for System Design
This section provides an understanding of the
following important properties:
» Cake solids concentration
• Throughput rate
» The recovery fraction, or the fraction of those
solids delivered to the machine that exit the
machine as cake and are not recycled to some
other portion of the facility.
The cake solids concentration achievable with a
particular sludge will regulate the cost of downstream
operations and often determine the need for additional
upstream operations such as thickening. There is a
relationship between the cake solids concentration
and the throughput in that, with filter presses, higher
solids are almost always achievable. This is true for
any given operating circumstance, if one is willing to
increase the cycle time and, therefore, decrease the
rate of throughput. The designer's challenge in this
regard is to maximize throughput and solids
concentration consistent with specific operating
conditions.
The third critical design parameter is solids recovery.
Systems that do not recover a substantial quantity of
solids can experience an increased need for media
washing and cause a buildup of fine solids in some
process loop, especially one that goes to a thickener
or perhaps to the wet end of the plant. Solids losses
above 2-3 percent of feed solids are usually traced
to torn media or sludge adhering to the media on
discharge and washed off to be recycled. Sometimes
this buildup of fines can lead to higher effluent
suspended solid concentrations. In any event,
recovery in excess of 95 percent is an important
design objective of a system and is necessary to
prevent both the excessive recycle of solids and the
possible impact on some aspects of the wastewater
treatment plant's operation. In this regard, filter
presses generally are superior, with solids recovery
typically greater than 98 percent.
7.4.4 Design Procedures
This section contains a review of the methods
employed for predicting solids concentration,
throughput, and recovery based on rather simple and
straightforward laboratory tests. Pilot operations, if
feasible, offer the best way of obtaining data on all
three of the important design aspects. However, pilot
operations are often not possible, in which case it
becomes necessary to design from bench-scale
information.
These procedures are based on one or both of two
properties of sludge slurry systems and reflect the
ease or difficulty in separating the water phase and
the solid phase from each other. These properties are
the Specific Resistance and the Capillary Suction
Time (CST). These two properties have been defined
and methods for their measurement are described in
Sections 5.5.3.3 and 5.5.3.4 respectively. Specific
Resistance and CST are used to develop design
information on throughput and final solids
concentration. Most of the relationships discussed
below utilize the Specific Resistance test as the basic
guide in estimating yield and cake solids. Yet, the
CST is also a useful test, and this section of the
manual contains several references to its use.
There are some significant dimensional
considerations which must be discussed for a full
understanding of Specific Resistance. Christensen
(18) has summarized typical values of Specific
Resistance for water and wastewater sludges and
commented on the disparity in the use of units to
describe Specific Resistance. He points out that
sec2/g probably is an incorrect unit assignment
because of the manner in which these units were first
employed. Gale (19) has pointed out that sec2/g as a
unit describing Specific Resistance was a result of
improperly using g/cm2 for pressure difference across
the filter cake. It has been suggested that meters per
kilogram (m/kg) is a more satisfactory unit.
Christensen suggests the use of terameters per
kilogram (Tm/kg) as the best possible unit.
Christensen has also noted that proper conditioning,
generally speaking, changes the Specific Resistance
by a factor of 102 to 103. Raw wastewater sludges
107
-------
Figure 7-17. Filling and cake discharge, fixed volume recessed plate filter press,
Filter Cloth
Cake Forms in
This Volume
Filtrate
Sludgo Feed
Filter Plata Assembly
Holds Filter Cloth
Filter Cloth
Air Supply
; Filter Plate
Filter Medium
Under Air
Pressure
Accumulated
Cake
Gasket
Filtrate
have Specific Resistance values of 10-100 Tm/kg.
Adequately conditioned sludges have Specific
Resistance values of about 1,0 Tm/Kg, and well
conditioned sludges have Specific Resistance values
on the order of 0.1 Tm/kg. The conversion factors to
go from sec2/g to cm/g and m/kg are 9.81 x 102 and
9,81 x 103, respectively.
The CST test provides a substantial amount of
information about the ease in separating the water
portion from the organic solids portion of sludge. For
example, unconditioned waste activated sludge has a
Capillary Suction Time of 100-200 seconds. For a
filter press to function, dewater, and release the
waste activated sludge cake, a Capillary Suction Time
of 10 seconds or less is required.
The following series of relationships show the
development of the significant equations that govern
flow through porous media and, hence, filtration
phenomana. In the 1800s, Poiseuelle described the
velocity in a circular capillary tube as:
U = (d2g/32u) (Pg/L)
All in compatible units, where,
U = linear velocity
d = diameter of capillary
P = pressure differential
p. = viscosity of the liquid
(7-25)
L = length of capillary
g = gravitational constant
D'Arcy also in mid-1800s, showed that (d2g/32u) is
a constant by noting that the flow through sand beds
may be described by U = (Kj) x (P/u) x (L), the
other symbols are defined as before.
These equations were modified by Kozeny (20), who
introduced porosity and specific surface in the
equation:
(7-26)
Again, in consistent values, where,
c = the porosity
S0 = the Specific Surface
K = a constant equal to 5
All other symbols are as indicated earlier.
Hence, with filtration:
U = (1/A) (dv/dO)
and
(7-27)
108
-------
Figure 7-18. Filling and cake discharge, diaphragm press.
System Ready
Membrane Filter Plates
Sludge
Feed
Inlet
Filter Cloth
Filtration
Recessed Filter Plates
Sludge
Feed
Inlet
Filtrate Outlet
«— Filtrate Outlet
Membrane Squeeze
Air Inlet Ports
Filter Cake Complete
Filter Cake
- Filtrate Outlet
where,
U = K, (P/pL)
K! = (g e3)/[5 (So)2 (1 - e)2]
dv/d6 = rate of flow of liquid across cake
This form leads to the conventional expression:
dv/d0 = PA/uRL
where,
L = cake thickness
R = specific resistance, sec2/g
A = area of cake
but,
LA = cake volume also
v V = cake volume
v = volume of solids deposited per unit of filtrate
hence,
LA = v V
L = v V/A
and then by substitution,
dv/dO = PA%RvV
(7-29)
Carman (21), noting that R must include all
resistance, developed an equation with two terms -
one for the cake and one for the media:
dv/d6 = PA2%(rvV + Rm A)] (7-30)
where,
r = Specific Resistance of cake
v = cc of cake deposited by 1 cc of filtrat3
Rm = initial resistance of 1 cm2 of filtering aurface
109
-------
For compressible cake, vV becomes Vc where c is
the weight of dry solids per unit volume in the
unfiltered slurry. The general equation then becomes:
dv/dO = PA2/fp(rcV + Rm A)] (7-31)
Integrating the expressions and neglecting the
resistance of the media, the time for filtration
becomes:
6 = prcV2/2PA2
(7-32)
If OA7 is plotted against V, a straight line is obtained
whose V slope (b) is:
b = prc/2PA2
(7-33)
Therefore, the Specific Resistance may be calculated
from Buchner funnel test data as described in Section
5.5.3.3, where,
r = 2bPA2/pe
(7-34)
Specific Resistance has been used in calculating
yields from pressure filters. Coackley (22) reported a
procedure in 1957. Mininni, Spinosa, and Misiti (23)
have presented a procedure for predicting the filtrate
flow rate and cake concentrations for fixed-volume
pressure filter filtration. These workers observed that
49, the filtrate flow rate or flux, after the initial period of
drainage or while the cake is being formed, is
described by the expression:
4> = atb
(7-35)
where t is time, and a and b are coefficients which
can be determined if the Specific Resistance, initial
solids concentration, filtrate viscosity, and maximum
operating pressures are known. The final cake
concentration then can be calculated by making a
material balance, assuming that the dry solids
density, the filtering time, the conditioner dosage, the
slurry concentration, filter press chamber volume, and
filtration surface areas are all known. The reported
agreement between predicted and actual values is
excellent.
Wilhelm (24) obtained the following expression from
classical filtration theory:
Log 9 = log [KA2(Sc/c)2] + log [(fpf)2]
where,
0
A
Sc
c
£
= filtration time, minutes
= filtration area, cm2
= cake solids concentration by weight fraction
= feed solids concentration, gm/cm3
= cake thickness, cm
= feed density, gm/cm3
Wilhelm's procedure provides excellent correlation
when the cycle time is plotted against his correlating
factor:
KA2 ( Sc)2/c2
To obtain good replication on different runs with the
same sludge requires the solution of similar ranges of
pressure to obtain "K" values.
The role of the filter media and the relationship
between the character of the material being filtered
and the media has been described in a study by
Christensen and Sipe (25). They developed the
following equation which, like the Carman equations,
separates the resistance associated with the cake
itself from that associated with the medium. The
equation is:
t_
-
V +
PtA
(7-36)
where,
t
V
P
R
PI
A
= time
= filtrate volume
= absolute viscosity
= Specific Resistance (in the case of m/kg)
= mass of cake deposited per unit volume of
filtrate
= total pressure drop across cake and medium
= filtration area
Rm = resistance of the medium
If the medium resistance is negligible, the preceding
equation can be rearranged to a form similar to
Carman's equations:
where,
log t = (pRc/2PcA2) + 2 log v (7-37)
Pressure drop across cake
The authors suggest that the rearrangement of the
equation offers the second way to plot filtration data,
i.e., log t versus log v. When this is done, according
to the equation, the data should plot as a straight line
with the slope of two. The intercept at a convenient
point, such as volume = 1, can be used to calculate
the Specific Resistance, since all of the other factors
in the first term of the equation are known. The
authors also point to many significant deviations when
the slope is equal to two; one is that the equation can
be written as t = KV", where n varies. In the t/V vs.
V plot (which will be linear only if n is equal to two
when the data is approximated by a straight line), the
intercept of that line will be negative when n is greater
110
-------
than two and positive when n is less than two. The
t/V intercept is proportional to the medium resistance.
Constructing a log t versus log V plot of the data from
a filtration experiment with significant medium
resistance is equivalent to satisfying the equation: log
t = log (KiV2 -f «2V) where Kj represents the
constant items in the initial term of the first of
equation of Christensen and Sipe and where r<2
represents the constants in the second term.
Noting a consistent deviation from theoretical
practice, Notebaert et al. (26) have proposed a
modification of the standard cake filtration model to
account for the deviations. The significant findings of
Notebaert et al. were summarized by Christensen and
Sipe (25) as follows:
• The assumption that the medium does not become
fouled or clogged is not realistic.
• If the particles in the sludge are of the same order
as the pores in the filter medium, the medium will
clog. If the particles are a great deal larger than the
pores in the medium, clogging will still occur but it
will occur over a much longer period.
• If the medium clogs, resistance will be high at the
start of the filtration cycle while the medium is
clogged, but will increase slowly afterwards.
Therefore, the average Specific Resistance will be
decreasing throughout filtration. If the cake
becomes clogged, the clogging will continue
throughout filtration with a continuous increase in
the average Specific Resistance.
• The slope of the log t versus log V plot is indicative
of the physical processes described. If the medium
is clogging, the slope will be less than two. If the
cake is clogging, the slope will be greater than two.
Selection of the optimum filter media, based on the
manufacturer's specification characteristics of the
media (which will include data such as air flow, the
weave, the fabric, etc.) is not yet possible. However,
the Metropolitan Waste Control Commission of the
Twin Cities (27) has carried out a detailed and
comprehensive study on pressure filtration. When
wastewater, without solids, comes in contact with the
filter medium, the media resistance will increase. This
is probably due to bacterial growth, since the
presence of chlorine decreases the rate at which
resistance increases. Pressure and the impact of
pressure on the fibers themselves increases the
extent and rate of blinding. The workers observed that
polypropylene and nylon are the two most commonly
used materials for filter cloth. The authors, in their
literature review, pointed to Purchas' work. He tried to
relate filtrate clarity, resistance to flow, cake solids,
ease of discharge, cloth life, and tendency to blind to
media characteristics. They also noted that criteria
set forth by Warring might be the best and most
reliable guide for establishing a good model. These
criteria are:
» How small a particle can the media retain?
» What is its resistance to flow?
• What is the relationship between buildup of
particulates in the medium to the rate of flow?
Finally, these workers concluded from their own
studies that the resistance of the filter cloth increased
markedly with use. Periodic washing with water or
with acid reduced media resistance. The effect of
media resistance on press operation was found to be
very significant. Cake solids decreased from 60
percent to 33 percent on a full-scale press due to
increased media resistance. Filtration rates, filtration
yield, mass of dry solids deposited, and cake percent
solids all decreased because of increases in media
resistance. Any model of a pressure filtration process
must include a term for media resistance.
7.4.5 Support Equipment and Processes
Sludge must maintain some structural integrity during
the pressing period, since a massive structure will
prevent the movement of water through the filter cake
to the discharge or filtrate side. To this end, one of
the essential parts of a filter press system is the
sludge conditioning subsystem. For recessed volume
filter presses, the most common conditioning
technique for digested and waste activated sludges -
and probably the most common for primary sludge -
is the addition of iron salts and lime. The average
quantity required is on the order of 5 percent ferric
chloride and 20 percent lime, though values as low as
3 percent ferric chloride and 10 percent lime and as
high as 10 percent ferric chloride and 40 percent lime
have been reported and are sometimes required.
In general, the ferric chloride requirement is a
function, at least in part, of the sludge alkalinity. The
role of lime in sludge conditioning has been
discussed extensively by Webb (28) and Sontheimer
(29), but is still somewhat unclear. Lime's solubility
above pH 11 or 12 is only on the order of one gram
per liter. As a result, much of the lime must exist as
partially hydrated calcium hydroxide, which probably
acts structurally to provide channels for water to
move through to the filtrate side.
In studies with four iron salt conditioners and lime,
Christensen and Stule (30) obtained both CST data
and Specific Resistance data. The results are shown
in Table 7-12 together with a correlation between
these two for the particular sludge under study. This
data is presented in Figure 7-19 and shows a
remarkably good correlation. Briefly, the ferric
conditioners performed the best and the chloride form
was more effective than sulfate ore which appeared
111
-------
Table 7-12. Comparison of Iron Conditioners With and Without Lime (14)
Total Sludge
Solids1
percent
5.5
5.5
5.5
5.5
5.5
5.5
5.5
5.5
7,0
7.0
7.0
7.0
Iron Conditioner
FeSO4»7H2O
FeCI2»4H2O
Fe2(SO4)3«6H20
FeCi3»6H2O
FeSO4»7H2O
FeCl2«4H2O
Fe2(SO4)3»6H20
FeCI3»6H2O
FeSO4*7H2O
FeCI2»4H2O
Fe2(S04)3«6H2O
FeCI3*6H2O
Iron Dose
percent
1.72
1.72
1.72
1.71
3.44
3.44
3.44
3.44
3.44
3.44
3.44
3.44
CST After Iron
Addition
sec
208
157
41
26
180
139
27
19
480
2
117
58
Lime Dose
percent CaO
15
15
15
15
30
30
30
30
20
20
20
20
Specific Resistance After
Iron and Lime Addition
10" rn/kg
14.0,
7.9
5.0
2.6
6.0
2.9
2.3
1.2
11.0
5.6
5.3
1.8
* Tho ncuvated sludge ornployed was a raw mixed sludge approximately 50% primary and 50% waste activated on a dry solids basis.
2 No result because of a lab accident.
Flguro 7-19. Specific Resistance vs. Capillary Suction Time.
14r
12
2-
20 40 60
CST, sec
80
100
to produce a slightly poorer floe and a more poorly
conditioned cake than did the chloride.
The role of calcium was also studied extensively, and
it was concluded that calcium is involved in a
chemical link with the iron floe. However, calcium
chloride was used and the pH was raised with sodium
hydroxide to obtain a somewhat synthetic situation.
Extensive reports exist which relate the necessity to
clean calcium hydroxide scale off both media and
plates, indicating that considerable quantities of
calcium hydroxide exist when the sludge is
conditioned with lime. Two other important
observations are that aging the sludge can as much
as double the Specific Resistance in one hour. The
actual impact of aging depends on the flocculants
used. Also, it is clear that the pH has a substantial
effect. The Specific Resistance dropped from 1.6 x
1012 m/kg at a pH value of around 11.3 to about 0.2
x 1012 m/kg at a pH of 12.45. This clearly is a pH
related phenomenon and not tied into any physical
property of calcium because no precipitation of
calcium hydroxide was observed.
The use of polymers for sludge conditioning is
expanding. Polymers can produce very nearly the
same cake solids and do not result in a 15-30
percent increase in cake weight and volume. Dosing
procedure, flocculation requirements, and filter press
pressure-time relationships necessary to optimize
polymer dosage and cake release are site specific.
About 75-80 percent of the conversion to polymer
trials appear successful. Polymer costs are 30-70
percent of ferric and lime costs.
7.4.6 Operational Factors and Performance
Characteristics
This section deals with the machine and process
variables that affect the efficiency of filtration when
dewatering with a conventional recessed plate filter
112
-------
press or a diaphragm filter press. The first part is
devoted to machine variables, which are developed or
derived from the unique and special characteristics of
the machine in use. The second part discusses those
variables that arise from the process or the unique
characteristics of the sludge to be filtered.
In a conventional filter press, the operator controls the
following variables:
• Pressure of the feed sludge
• The rate at which the pressure is applied and the
pacing of flow to the filter press
• The overall filtration time, including such variables
as the time at each pressure level in multiple
pressure level operations
• The use of precoat or body feed and the amount of
material used
• Conditioning chemicals
- Type
- Dosage
- Location
- Mixing efficiency
• Floccutation efficiency
• Cloth washing frequency
• The nature of the filter media used,
A similar set of machine variables exists for the
diaphragm filler press. They are:
• Pressure of the feed sludge and the rate at which
feed sludge is added to the machine.
• Filtration time
» Diaphragm pressure
• Diaphragm squeezing time
• Rate at which the diaphragm pressure is increased
• Conditioning chemicals
- Type
- Dosage
- Point of addition
- Mixing efficiency
- Flocculation efficiency
• Filter media used
• Cloth washing frequency.
Changes in these parameters are predictable up to a
point, and mechanisms exist to evaluate the effect of
varying each one for optimizing the system.
Precoat generally does not need to be used when
inorganic conditioning chemicals, particularly ferric
chloride and lime, are used. Heavy doses of organic
polyelectrolyte may also preclude the use of body
feed or precoat. Precoat is normally used in cases
where the particle size is extremely small or
considerable variability in filterabilty and substantial
loss of fine solids to and through the filter media are
anticipated. A final decision about the need for
precoating may require lab or field experimentation
with the specific sludge.
When substantial quantities of lime are used, cloth
washing may require both an acid and a water wash.
Therefore, a medium is needed that is resistant to
both acid and alkaline environments. In those
instances where polyelectrolytes are used, the
washing operation normally is accomplished with only
clean water since the sludge imbedded in the media
is backflushed to the waste.
Few process variables, as opposed to machine
variables, are likely to be controllable by the operator.
Process variables include:
• The type of sludge to be dewatered. Raw or
digested primary sludge, waste activated sludge,
trickling filter sludge, RBC sludge, or mixtures
thereof have varying effects on the dewatering
process.
• The age or the freshness of the sludge.
Conditioning, particularly conditioning with
polyelectrolytes, is much more dependable and
reproducible when the sludge is fresh. The Specific
Resistance increases with time. Therefore, it is
desirable to dewater the sludge in as fresh a
condition as possible.
• Prior chemical conditioning. Prior chemical
conditioning tends to confound the use of chemical
conditioning at the dewatering device. This is
particularly true when polymers are used and if
polymers have already been used somewhere
upstream from the dewatering system. If this
condition exists, the best remedy is to use a small
quantity of the polymer of the opposite or neutral
charge, followed by the normal dose of the
polymer usually employed. Establishing charge
reversal with the polymer of the opposite charge
eliminates the confounding effects of the old,
partially degraded polymer on the sludge surface.
» The solids concentration achievable in the final
clarifier or in subsequent thickening operations.
Generally speaking, it is desirable to send to the
dewatering device a feed sludge with the highest
possible solids content.
• Solids capture. If the cloth is unbroken and cake
cleanly discharged, suspended solids recovery is
113
-------
about 99 percent. When the cloth is washed, the
effluent solids are somewhat higher.
» Cake concentration. The cake concentration must
be sufficiently high to readily discharge from the
cloth. Variables affecting cake concentration have
been reviewed earlier.
• Throughput rate. The throughput will be dependent
on the water release characteristics of sludge, type
and amount of chemicals, and the desired
minimum cake solids.
* Conditions under which the sludge was produced.
The filterability of sludge, particularly waste
activated sludge, is strongly dependent on the
conditions under which the sludge was produced.
This consideration probably applies to those
municipal wastewater treatment plants receiving
substantial quantities of high carbohydrate
industrial wastes that may produce, on occasion, a
nitrogen deficient situation in the activated sludge
portion of the plant. However, nitrogen deficient
activated sludge has a considerably higher Specific
Resistance when untreated than activated sludge
grown under nitrogen enriched conditions. In
addition, the final Specific Resistance after
chemical conditioning is not as good as that
achieved with activated sludge grown under excess
nitrogen conditions. The conditioned Specific
Resistance of the nitrogen- poor sludge generally
runs two to three times that of the activated sludge
grown under high nitrogen conditions when
properly chemically treated (31).
7.4.7 Survey of Finer Presses
In order to provide detailed information on the
operating experience of filter presses, a survey was
made of SO municipal wastewaler treatment plants by
Terraqua Corp. of Hunt Valley, MD (32) for the City of
Baltimore in 1984. An effort was made to update the
information to April 1987. The plants ranged in size
from 0.66 to 265 m3/min (0.25 mgd to 100 mgd).
Table 7-13 summarizes general information on the
filter press installations. Of the 50 plants, 42 were
dewatering anaerobically digested or raw sludges: 21
of each sludge type. One plant was elutriating
anaerobically digested sludge, five were processing
aerobically digested sludge, one was thermally
conditioning anaerobically digested sludge, and one
was dewatering an alum sludge from a tertiary
treatment process.
The majority of plants, 41 of them, had their presses
in operation at the time of the study. Of the remaining
plants, four had taken their presses out of service
(one temporarily, three permanently), two planned to
take them out of service, two had presses under
start-up, and one had not yet completed installation.
Most of the plants were landfilling or incinerating their
filter cake.
Table 7-13. Summary Data - General Information
Number of Plants
Type of Sludge Processed
Anaerobically digested
Anaerobically digesied/elutriated
Raw
Aerobically digested
Other (thermally conditioned and chemical)
Total
Operating Status of Presses
In service
Abandoned use of press
Out of service temporarily
Planned to be taken out of service
Under startup
Under construction
Total
Filter Cake Disposal Method1
Landfill
Incinerate
Land Apply
Incinerate/Landfill
Landfill/Land Apply
Compost/Land Apply
Total
Press Manufacturers
Passavant
Edwards & Jones
Eimco(Shriver)
Sperry
Netzsch
Hoesch
Envirex (NGK)
Clow
Ingersoll-Rand (Lasla)
Total
Longest in continuous operation:
Statesville, NC - since 1974 - Passavant
Plate Material2
Cast Iron
Polypropylene
Not reported
Total
Operating Experiences3
Positive
Negative
Mixed
Total
21
1
21
5
2
50
41
3
1
2
2
1
50
24
11
5
3
2
1
46
14
12
11
4
3
3
1
1
1
50
33
14
4
51
29
12
8
49
1 Excluding plants under construction or which have abandoned
use ol press.
2 One plant has both plate types.
3 Excluding one plant under construction.
Three manufacturers of recessed chamber presses
dominated the installations: Passavant, Edwards and
Jones, and Eimco (Shriver press). Other recessed-
chamber press manufacturers with municipal
installations included Sperry, Netzsch, Hoesch, and
Clow. Two diaphragm press manufacturers were
surveyed - Envirex (NGK press) and Ingersoll-
Rand (Lasta press). The manufacturer with the
longest continuous operating history was Passavant,
which has an installation at the Statessville, NC plant
operating since 1974. The majority of installations
were relatively recent, since about 1980. Most of the
114
-------
presses used cast iron plates (33 plants), but
polypropylene plates were used by some
manufacturers for new presses; polypropylene had
been used as a replacement for cast iron plates by
some plants.
Based on the telephone conversations and on the
written comments received from a mai! survey, a
rating of positive, negative, or mixed was given to the
attitude of operations personnel toward operating and
maintaining their filter press system. In general, a
negative rating was assigned to plants which have
taken their press system out of service due to
excessive operating and/or maintenance costs, which
reported serious maintenance problems, or which had
a very negative reaction toward the installation. A
mixed rating was given to plants which had less
serious operating problems and where the operator's
attitude was more positive than negative. A positive
rating was given where the personnel were generally
satisfied or, in some cases, enthusiastic about the
press, although some problems may have been
reported. The experience of the majority of the plants
was rated as positive (29 of 49 plants, excluding one
plant under construction). Twelve had negative
ratings; of these, four plants have ceased using the
press. Eight plants had mixed operating experience.
Table 7-14 presents a summary of operating
comments itemized by type of problem and grouped
by the overall rating of operating experience. A large
number of plants (28) reported no significant
problems with the press installation. It was commonly
reported even by plants with both positive and
negative overall attitudes toward the press system
that costs were high for operating labor and for
chemical conditioning, and that well trained, motivated
operators and mechanics were a necessity. Plants
with positive attitudes seemed to be able to overcome
this difficulty by good training and supervision, while
plants where the attitude was negative seemed, in
contrast, to be overcome by it.
Table 7-15 presents a summary of data on
conditioning chemical dosages and filter cake quality.
Most of the plants, 29 of them, were conditioning with
ferric chloride and lime only, six were using a precoat
of ash or diatomaceous earth with ferric and lime
conditioning, and six were conditioning with polymer
alone. The less frequent conditioning methods
included ferric chloride/lime with ash and
polymer/precoat with and without ash. One plant was
using lime alone, but was dewatering an alum sludge
from a tertiary phosphorus removal system.
Reported filter cake solids content averaged 37
percent for the plants using ferric/lime conditioning
and only slightly less at 34 percent for plants using
polymer alone. Plants using a precoat reported one of
the highest solids contents, 42 percent with ferric and
lime, but this method had the disadvantage of
maximizing performance at the expense of additional
inert material in the cake. The two plants using ash
as a conditioning material reported 45 percent with
polymer and a precoat and 32 percent with ferric and
lime. However, the amount of ash added, from 63
percent to 100 percent, adds significantly to the
amount of filter cake to be disposed. Two plants were
using polymer with a precoat and reported an average
cake solids of 33 percent. Table 7-16 presents a
summary of filter press performance on anaerobically
digested and raw sludges for the three most common
conditioning methods (ferric/lime, polymer, and
ferric/lime/precoat). Surprisingly, in each case, the
best results were obtained on digested rather than
raw sludge: 38 percent vs. 36 percent for ferric/lime,
35 percent vs. 31 percent for polymer, and 44
percent vs. 32 percent for ferric/lime/precoat. This
data should be interpreted carefully, however because
numerous other factors can influence press
performance. Such factors include operating
pressure, cloth condition, feed solids percent,
primary/secondary sludge ratio, chemical dosage, and
press cycle time. An example is Watertown, NY,
which reported a 37 to 44 percent solids cake with
polymer-only conditioning, but required 22.5 kg/Mg
(45 Ib/ton) of polymer and a very long (4-hr) cycle
time.
Chemical dosages and cycle times are also listed in
Table 7-16. As shown, chemical dosages for
anaerobically digested sludges averaged (in order by
conditioning method): (1) 7 percent ferric, 26 percent
lime without precoat; (2) 18.5 kg/Mg (37 Ib/ton)
polymer; and (3) 9 percent ferric, 32 percent lime with
precoat. Chemical dosages for raw sludges averaged:
(1) 7 percent ferric, 23 percent lime without precoat;
(2) 6 kg/Mg (12 Ib/ton) polymer; and (3) 8 percent
ferric, 20 percent lime with precoat. Average cycle
times varied from about 1.5 to 2.5 hr.
7,4.8 Genera/ Equipment Selection Criteria
Ohara et al., (33) in writing about the Hyperion
system, developed the following set of criteria, in
order of priority, for selecting the most cost effective,
functional, safe and environmentally sound system:
• Meets all environmental and legal requirements
• Has minimum energy, resource and economic
requirements
• Minium suspended solids remain on the liquid side
stream, whether il be concentrate, filtrate, or
supernatant
• Provides capture of sludge solids
• Provides maximum cake solids (minimum percent
moisture in the cake)
• Has maximum operational reliability, flexibility, and
ease of use
115
-------
Tabla 7-14. Summary of Operating Problems
Horn No,
Positive
No significant problem 1 1
High maintenance costs or unspecified mechanical 1
problems
High operating or conditioning costs 4
Well trained and/or motivated operators and mechanics 4
needed
Sludge feed problems (line clogs, feed pumps) 2
Pressoloctrieal or instrumentation programs 4
Excessive cloth wear or tears 1
Conditioning system problems (corrosion, line clogs, poor 1
uniformity or conditioning)
Stay boss waar or failure 1
Plato suspension pin breakage 1
Difficult to obtain spare parts 2
Ammonia release problem 1
Plate shilling mechanism problem 2
Cake discharge system problems (conveyors, drip trays) 1
Plate cracking or breakage
Hydraulic power unit leakage
Poor cake solids
Press frame twisting
Plate coating wear 1
Rapid cloth blinding
Filtrate drain lime accumulation
Tabla 7-15. Filter Cake Solids - Average by Conditioning «
Method
•
No. o! Plants1 Percent Solids
Mean2 Std. Dev. *
Ferric lime only 29 37 s,3
Forric/Iime/pracoat 6 42 7.8
Limo only (alum sludge) 1 38 -
Ferric/Iime/ash 1 32
Polymer only3 8 34 4.2
Potymer/procoat 2 33 4.2
Pofymer/ash/preeoat i 45
of Plants (by overall rating)
Negative Mixed
-
5 2
4 3
4
1 3
1
1 4
2 2
3
3
2
1 2
1
1 1
1 1
1
1
1
-
1
1
Requires minimum maintenance
Total No. of Plants
Reporting
11
8
11
8
6
5
6
5
4
4
4
4
3
3
2
1
1
1
1
1
1
and downtime
Has maximum flexibility to meet changing needs
Can meet established construction schedules.
1 Excukling plants thai have abandoned use o( press or are under
construction.
2 Using midpoint data tor plants reporting a range of values.
3 Excuding thermally conditioned sludge.
116
-------
Table 7-16. Filter Cake Solids - Average by Conditioning Method and Sludge Type1
No. Plants Average
Conditioning Method
and Sludge Type
Reporting Cake Solids Ferric Only (%)
Lime Only (%)
Polymer Only
(Ib/ton)
Cycle Time2
(minutes)
percent
Min Max Avg Min Max Avg Min Max Avg Min Max Avg
Ferric/lime
Anaerobic
Raw
Polymer
Anaerobic
Raw
Ferric/lime/precoat
Anaerobic
Raw
11
14
2
3
4
1
38
36
35
31
44
32
3 15 7 12
1.5 12 7 15
7 10 9 16
8
43
30
40
-
26
23
32
20
52
70
20 45 37 60
7 18 12 50
90
168
330
240
105
105
-
100
155
1503
82
99
120
1 Using midpoint of data tor plants reporting a range of values.
2 Excuding diaphragm press installations.
3 Only two plants reporting, wide variation.
7.5 Vacuum Filtration
7.5.1 Introduction
The most common means of mechanically dewatering
municipal wastewater sludge until the mid-1970s
was vacuum filtration. Vacuum filters were patented in
England in 1872 by William and James Hart. The first
United States application of a vacuum filter in
dewatering municipal wastewater treatment plant
sludge was in the mid-1920s. Until the late 1950s,
the drum or scraper-type rotary vacuum filter was
the most common design of vacuum filter employed.
The belt-type filter using stainless steel (SS) coils
was introduced by Komline Sanderson in 1951. Since
then and until the mid 1970s, the belt-type filter with
natural or synthetic fiber cloth, woven SS rnesh, or
coil springs media has been the dominant means of
mechanically dewatering sewage sludge.
A vacuum filter consists of a horizontal cylindrical
drum which rotates while partially submerged in a vat
of sludge. The filter drum is partitioned into several
compartments or sections. Each compartment is
connected to a rotary valve by a pipe. Bridge blocks
in the valve divide the drum compartments into three
zones, which are referred to as the cake formation
zone, the cake drying zone, and the cake discharge
zone.
The filter drum is submerged to about 20-35 percent
of its depth in a vat of previously conditioned sludge;
this submerged zone is the cake formation zone.
Vacuum applied to this submerged zone causes
filtrate to pass through the media and sludge particles
to be retained on the media. As the drum rotates,
each section is successively carried through the cake
formation zone to the cake drying zone. This zone
begins when the filter drum emerges from the sludge
vat. The cake drying zone represents from 40 to 60
percent of the drum surface and ends at the point
where the internal vacuum is shut off. At this point,
the sludge cake and drum section enter the cake
discharge zone, where sludge cake is removed from
the media. Figure 7-20 illustrates the various
operating zones encountered during a complete
revolution of the drum.
Figure 7-20. Operating zones in a rotary vacuum filter.
There are essentially two variations currently on the
market: the drum filter and the belt filter. In the case
of the drum filter, the media covers the drum and the
sludge is removed by a roll discharge or a doctor
blade. The belt filter may employ conventional media
or coil media, but when conventional media are
employed, the belt leaves the drum for discharge and
is washed before recontacting the drum.
Figure 7-21 shows a cross sectional view of a coil
spring, belt-type vacuum filter. This filter uses two
117
-------
Figure 7-21. Cross-sectional view of a coil spring, belt type rotary vacuum filter.
Wash Water
Spray Piping Internal Piping
Vacuum Gauges
Vacuum and
Filtrate Outlets
Coke Discharge
^
Sludge Level
Agitator Drive
Agitator
Vat
layers of stainless steel coils arranged around the
drum. After the cake drying or dewatering cycle, the
two layers of springs leave the drum and are
separated from each other. In this way, the cake is
lifted off the lower layer of springs and can be
discharged from the upper layer. Cake release from
the coils is usually not a problem if the sludge is
properly conditioned. After cake discharge, the coils
are spray washed and returned to the drum just
before the drum reenters the sludge vat.
The coil springs, which have 7 to 14 percent open
area, act to suppprt the initial solids deposit which in
turn serves as the filtration medium. Because of the
open area of the springs, it is important that the feed
solids concentration be high; that is, it should contain
sufficient fibrous material to prevent the loss of fine
solids. Sludges with particles that are both extremely
fine and resistant to flocculation dewater poorly on
coil filters, and solids capture is low. A cloth medium
is required when filtering unthickened sludge that is
predominantly secondary solids.
Rgura 7-22 shows a schematic cross section of a
fiber cloth, belt-type rotary vacuum filter.
In this type of unit, the media leave the drum surface
at the end of the drying zone and pass over a small-
diameter discharge roll to facilitate cake discharge.
Washing of the media occurs after discharge and
before return to the drum for another cycle. This type
of filter normally has a small-diameter curved bar
between the point where the belt leaves the drum and
the discharge roll. This bar aids in maintaining belt
dimensional stability and ensures adequate cake
discharge. Scraper blades, additional chemical
conditioner, or the addition of fly ash are sometimes
required to obtain cake release from the cloth media.
This is particularly true at wastewater treatment plants
that produce sludges that are greasy, sticky, and/or
contain a large quantity of waste activated sludge. In
general, cloth media made from staple fiber produces
cleaner filtrate but has lower throughput than cloth
media made from monofilament fiber.
7.5.3 Design Procedures
The best way of carrying out bench-scale studies
involves the use of Capillary Suction Time, Specific
Resistance, and Filter Leaf Tests. These test
procedures are all described in detail in Chapter 5.
7.5.4 Support Equipment
Vacuum filters are normally supplied with auxiliary
equipment including vacuum pump, vacuum filtrate
receiver and pump, and sludge conditioning
apparatus. Figure 7-23 shows a typical complete
rotary vacuum filter system.
118
-------
Figure 7-22. Cross-sectional view of a cloth, belt type rotary vacuum filter,
Filter Drum
Cloth Filter Media
Discharge Roll
Cake
Discharge
Wash
Trough
Internal Piping
Drum Drive
Sludge
s Level
Filter Vat
Filter
Agitator
Figure 7-23. Rotary vacuum filter system.
Metering
Ferric Chloride
• Mixing Tank
Air to
Atmosphere
Filtrate
Pump - w-ter
Washings Return^- Conveyor
to Primary
Silencer
Water to
Primary
Vacuum
Pump
Sludge Inlet
119
-------
Usually, one vacuum pump is provided for each
vacuum filter, although some larger plants use fewer
than one pump per filter with the pumps connecting
to a common header. Until the 1960s, reciprocating-
type dry vacuum pumps were generally specified, but
since the early 1970s wet-type vacuum pumps have
been universally used. The wet-type pumps are
more easily maintained and provide sufficient
vacuum. Wet-type pumps use seal water, and it is
prudent to use potable water. If the water is hard and
unstable, it may be necessary to prevent carbonate
buildup on the seals through the use of a
sequestering agent. The vacuum pump requirement is
normally 0.7-1.0 m3/min of air /m2 of drum surface
area at 33 kN/m2 absolute pressure (1.4-2.0 cfm/ft2
@ 5 psi). If the expected yield is greater than 20-40
kg/m2 hr (5-10 Ib/ft2 hr) and extensive sludge cake
cracking Is expected, an air flow 2.0-2.5 times
higher should be used.
Each vacuum filter must be supplied with a vacuum
receiver located between the filter valve and the
vacuum pump. The principal purpose of the receiver
is to separate the air from the liquid. Each receiver
can be equipped with a vacuum-limiting device to
admit air flow if the design vacuum is exceeded (a
condition that could cause the vacuum pump to
overload). The receiver also functions as a reservoir
for the filtrate pump section. The filtrate pump must
be sized to carry away the water separated in the
vacuum receiver, and it is normally sized to provide a
capacity two to four times the design sludge feed rate
to the filter.
The filtrate pump should be able to pump against a
minimum total dynamic head of between 12-15 m
(40-50 ft), which includes a minimum suction head
of 7.5 m (25 ft). Centrifugal pumps are commonly
used but can become air-bound unless they have a
balanced or equalizing line connecting the high point
of the receiver to the pump. Typically, non-clogging
centrifugal pumps are used with coil filters because
they permit a somewhat higher solids concentration in
the filtrate. Self-priming centrifugal pumps are used
most frequently, since they are relatively maintenance
free. Check valves on the discharge side of the
pumps are usually provided to minimize air leakage
through the filtrate pump and receiver to the vacuum
pump.
7.5.5 Operating Factors and Performance
Characteristics
7.5.5.1 Machine Variables
The principle machine variables that impact on
vacuum filter operation are as follows:
• Filter media used
• Quantity of wash water used
» Drum speed
• Vacuum level
* Conditioning chemicals - type and dosage
• Drum submergence
* Vat agitation.
Establishment of the drum speed, optimum vacuum
level, conditioning methods, drum submergence and
optimum media selection can all be accomplished on
bench scale. The drum speed establishes the cycle
time and the submergence sets the form time and
drying time. The media selection is normally made at
the time of equipment start-up by the equipment
supplier. The trend over the past few years has been
to select a monofilament fabric, since they seem the
most resistant to blinding and have a reasonably long
life.
A change in conditioning procedures, sludge mixture,
or sludge holding time (time held before conditioning
and dewatering) impact on the efficiency of a given
medium.
7.5.5.2 Other Process Variables
With belt filters, wash water at a pressure of at least
480 kN/m2 (70 psi) must be available. Throughput is
usually estimated from data gathered with clean
media. It is generally observed that where there is
insufficient cloth washing, increasing the amount of
wash water will increase the machine throughput and
will help to increase cake dryness. Vat agitation is
necessary for proper cake formation, but over-
agitation will result in breaking up the sludge floe,
poor solids capture, and lower feed rates. The
addition of scraper blades, use of excess chemical
conditioner, or addition of fly ash are sometimes
required to obtain cake release from cloth media
vacuum filters.
The feed solids concentration has a critical effect on
the filter's production rate and the final solids
concentration. The higher the suspended solids
concentration of the feed sludge, the greater will be
the production rate of the rotary vacuum filter (Figure
7-24), and the suspended solids concentration of
the cake (Figure 7-25).
Generally, municipal wastewater treatment plant
sludges are not concentrated beyond about 5-6
percent total solids (TS) [primary sludge + waste
activated sludge (PS + WAS)], since above this
concentration the sludge becomes difficult to pump,
mix with the chemicals, and distribute after
conditioning to the filter. Increasing production rates
without higher sludge feed concentrations requires
higher chemical dosages and results in higher cake
moisture. Both of these consequences affect the cost
of sludge dewatering and ultimate disposal.
The lowest feed sludge suspended solids
concentration for successful vacuum filtration is
generally considered to be 3.0 percent (PS + WAS).
Below this concentration it becomes difficult to
120
-------
Figure 7-24. Rotary vacuum filter productivity as a function
of feed sludge suspended solids concentration.
65
50
.c
£
25
10
5
O Digested
• Primary
• Blended
A Activated
i
i
i
I
i
4 5 6 7 8 9 10 11 12
Feed Solids, %
Figure 7-2S. Sludge cake total solids concentration as a
function of the feed sludge suspended solids
concentration.
35
30
25
20
o
V)
15
10
Primary
Activated
Chemicals Used:
11 g/I CaO
5
o
-
I l l
0123
i
4
1
5
Feed
i i
6 7
Solids, %
3.7
j
8
g/i
i
9
FeCI3
|
10
|
11
produce sludge filter cakes thick enough or dry
enough for adequate discharge. For this reason, it is
extremely important that the design and operation of
the preceding sludge processes take into
consideration the need for an optimal solids
concentration when dewatering on vacuum filters.
Existing operation can often be improved by
increasing feed solids.
7.5.5.3 Performance Data
As with all types of mechanical dewatering
equipment, optimum performance depends upon the
type of sludge and its solids concentration, type and
quality of conditioning, and how the filter is operated.
Selection of vacuum level, degree of drum
submergence, type of media, and cycle time are all
critical to optimum performance. Tables 7-17 and
7-18 contain expected performance data for cloth
and coil media rotary vacuum filters, respectively, for
the sludge types indicated. Higher solids
concentration (+3 to 4 percentage points) are
produced by operating at about 50 percent of
maximum rate.
Tables 7-19 and 7-20 contain operating data of
several wastewater treatment plants using cloth
media and coil media, respectively.
The efficiency of solids removal, or percent solids
recovery, is the actual percentage of feed solids
recovered in the filter cake. Solids removals on
vacuum filters with adequate chemical conditioning
range from about 85 percent for coarse mesh media
to 98 percent with close weave, long nap media. The
recycled filtrate solids impose a load on the treatment
plant and should normally be kept to a practical
minimum. However, it may be necessary to reduce
the percent recovery in order to deliver more filter
output and thus keep up with sludge production.
7.5.6 Equipment Selection Criteria
Due to high power costs and the heavy use of
inorganic conditioners, vacuum filters are not often
selected for use in new facilities. In some cases,
refurbishing old equipment may be indicated to
minimize capital costs. However, older units should
generally not be used for standby capacity unless
they are refurbished. The factors discussed in this
section are included to provide guidance primarily for
such applications.
Table 7-21 lists some of the advantages and
disadvantages of vacuum filtration relative to other
dewatering processes, and Table 7-22 lists design
shortcomings that have been noted at a number of
vacuum filter installations.
The significant points to be examined, if refurbishing
or reusing vacuum filters, are:
Media selection
Feed solids
Conditioning requirements and the design of the
conditioning subsystem
Sludge holding time before and after conditioning
Filtration rate.
Other environmental considerations are shown in
Chapter 4.
121
-------
Tabto 7-17. Typical Dawaterlng Performance Data for Rotary Vacuum Filters - Cloth Media
Sludge Type
RawP
WAS
P+WAS
P + TF
AnaorobicaUv Digested:
P
P+TF
P + WAS
Elulrialod Anaorobically Digested:
P
P + WAS
Thofmallv Conditioned:
P + WAS
Feed Solids Cone.
percent
4.5-9,0
2.5-4.5
3-7
4-8
4-8
3-7
5-10
5-10
4.5-8
6-15
Chemical Dosage1,
FeCI3
20-40
60-100
25-40
20-40
30-50
40-60
40-60
25-40
30-60
03
kg/Mg dry solids
CaO
80-100
120-360
90-120
90-120
100-130
150-200
125-175
0-50
0-75
0
Yield*
kg dry solids/m!i/hr
17-40
5-15
12-30
15-35
15-35
17-40
20-40
15-35
20-40
Cake Solids
percent
27-35
13-20
18-25
23-30
25-32
18-25
20-27
27-35
18-25
35-45
' All values shown are for pure FeC^ and CaO. Dosage must be adjusted for anything else.
2 Filter yield depends to some extent on feed solids concentration. Increasing the solids concentration normally gives a higher yield
3 Somes heat treated sludge requires some conditioning to maintain recovery at a high level.
1 Ib/ton - 0.5 hfl/Mfl
11b/ft2/hr = 4.9 kg/m2/hr
Sludoe Type
RawP
TF
P + WAS
Anaorobically Digested:
P+TF
P + WAS
Elutriated Anacrobicallv Digested:
P
Feed Solids Cone.
percent
8-10
4-6
3-5
5-8
4-6
8-10
Chemical Dosage1,
FeCI3
20-40
20-30
10-30
25-40
25-40
10-25
kg/Mg dry solids
CaO
80-120
50-70
90-110
120-160
100-150
15-60
Yield2
kg dry solids/m2/hr
30-40
30-40
12-20
20-30
17-22
20-40
Cake Solids
percent
28-32
20-32
23-27
27-33
20-25
28-32
' AH values shown are tor pure FeCIs and CaO. Dosage must be adjusted for anything else,
2 Filler yield depends to some extent on feed solids concentration. Increasing the solids concentration normally gives a higher yield.
1 Ib/lon - 0.5 kg/Mg
1 Ihm2/hr - 4.9 kg/m2/hr
Equipment sizing may be accomplished by using
information from Tables 7-18 through 7-20. This
data may be augmented through the use of the filter
leaf test as discussed in Chapter 5, A series of leaf
tests will provide a range of values for solids loading
and cake solids. The design conditions may be
selected from these findings. A scale-up value of 0.8
Is usually applied to obtain the final sizing. These
procedures have been developed over many years
and have excellent reproducibility and a high degree
of confirmation.
7.6 New Methods of Dewatering
An extensive review of developments in municipal
wastewater treatment plant sludge technology was
undertaken for this section. This review included a
literature search and contact with both commercial
and non-commercial resources. Significant findings
are identified and discussed below, including the
Expressor Press, Som-A-System, CentriPress,
Screw Press, and Sun Sludge System.
7.6.1 Expressor Press
A major manufacturer of dewatering equipment
recently developed a modified twin belt press for use
primarily in the industrial market. Substantial tests
have been conducted with municipal sludges and
various kinds of fibrous industrial waste sludges. The
device, named the Expressor (R) or Expressor Press,
consists of, in its basic form, two or three S rolls
(wraparound) and a series of five P rolls (direct) on
which the pressure can be individually varied. An
Expressor Press with this configuration is shown in
Figure 7-26.
122
-------
Table 7-19, Specific Operating Results of Rotary Vacuum Filters - Cloth Media
Location
Willoughby, Eastlake, OH
Tamaqua, PA
Grand Rapids, Ml
Grand Atkinson, Wl
Frankemuth, Ml
Oconomowoc, Wl
Genessee City, Ml
Sludge Type
P + WAS •*• septic
Anaer. Dig. (P + WAS)
Therm. Cond. (P + WAS)
WAS
WAS
Anaer. Dig. (P + WAS)
P + WAS
Feed Solids
Cone.
percent
4-6
6
10-15
3-4
3.7
2.3
8
Conditioner
Used'
% by weight
FeCI3 - 3
CaO - 14
FeCI3 - 3
CaO - 23
None
FeCI3 - 6
CaO - 16
FeCI3 - 8
CaO - 14
FeCI3 - 6
CaO - 20
FeCI3 -
CaO- 16
Cake Solids
percent
20
18
50
19
15
18
27
Yield
kg dry
solids/m2/hr
14-24
15
30
15-17
15
12-15
27
Filtrate
mg/l
SS 20-30
SS 5,000
BOD 10,000
SS 500-1, 100
BOD 10
1 All values shown are for pure FeCI3 and CaO. Dosage must be adjusted for anything else.
Table 7-20. Specific Operating Results of Rotary Vacuum Filters - Coil Media
Location Sludge Type
Conditioner
Used1 Cake Solids
Yield
Blytheville, AR
York, PA
Wyomissing Valley, PA
Bayonne, NJ
Woodbridge, NJ
Shadyside, OH
Arlington, TX
TF
Anaer. Dig. (P + WAS)
Anaer. Dig. TF
Anaer. Dig, P
P
Anaer. Dig.
TF
% by weight
FeCI3 - 18
CaO - 47
FeCI3 - 40
CaO - 125
FeCI3 - 31
CaO - 136
FeCI3 - 14
CaO- 120
FeO3 - 20
CaO - 160
FeCI3 - 32
CaO - 1 65
FeCI3 - 32
CaO - 174
percent
33.1
21.1
18.2
30.9
29.7
29
25.2
kg dry
solids/m2/hr
50
23
30
38
40
20
43
1 All values shown are for pure FeCIs and CaO. Dosage must be adjusted for anything else.
In a second configuration, a unit called the Hybrid
Expressor Press contains a gravity drainage section,
four or five S rolls, and the five variable pressure P
rolls. Depending on the model being considered, the
P roll pressure can be varied from zero above the belt
tension up to 200 kg/cm (1,000 Ib/lineal inch). This
new unit is capable of producing a very dry cake from
the most difficult sludges with the use of press aids.
A variety of press aids have been employed, but the
most widely investigated material has been sawdust.
The unit can produce an autogenous cake from waste
activated sludge using between 50 and 125 percent
sawdust by dry weight, based on the content of
sludge solids. The water displacement by the press
aid varies from slightly over one to as much as three
kg H2O/kg press aid added. The water displacement
is based on the kg H2O/kg sludge solids with and
without press aid. The cake produced varies from 30
to 40 percent solids and, in some instances, runs
somewhat higher than 40 percent.
Other press aids have been tested, including sand,
soil, finely divided paper, fly ash, and coal fines. All
work to some degree to increase the cake resistance
to shear in the P rolls and hence permit higher
123
-------
Table 7-21.
Advantages and Disadvantages of Vacuum
Filtration
Figure 7-26. Expressor press.
Advantages
Disadvantages
Operation is easy to understand
bocauso formation and
discharge of sludge cake are
easily visible.
Does not require highly-skilled
operator.
WiH continue to operate even if
tho chemical conditioning
dosago is not optimized,
although (his may cause
discharge problems.
Coil spring medium has very
long Ido compared to any cloth
medium.
Has low maintenance
requirements for a continuously
operating piece of equipment,
except in certain cases with
lime conditioning.
Consumes a large amount of
energy per unit of sludge
dewatered.
Vacuum pumps are noisy.
Lime and Ferric chloride
conditioning can cause
considerable maintenance
cleaning problems.
The use of lime for conditioning
can produce strong ammonia
odors with digested sludge.
Best performance is usually
achieved at feed solids of 3-
4%. However, some well
conditioned sludges are filtered
successfully at concentrations
of <2%.
Ferric chloride and lime
conditioning costs are higher
than polymer conditioning costs.
Polymer conditioning is not
always effective on vacuum
filters.
Tabto 7-22.
Common Design Shortcomings of Vacuum Filter
Installations
Shortcomings
Resultant Problem
Solution
Improper Filler
Media
Filler blinds,
provides inadequate
solids capture
and/or poor cake
release.
Replace media after
testing for optimum.
Improper chemical
conditioning used
Inadequate water
pressure for spray
nozzles
Poor solids capture,
low solids loading
rate, and low cake
solids concentration.
Improperly cleaned
media.
Change to correct
chemical
conditioners.
Provide booster
pumping to maintain
484 kPa (70 psi}
minimum pressure.
pressure and, in turn, higher solids content. Press
aids in the 30 to 80 mesh region seem to be the most
effective. With materials not particularly resistant to
shear, such as paper and fiber, the particle size
seems to have little impact on final sludge solids.
The press has also been tested on primary sludges
and on mixtures of primary and waste-activated
sludge from a pulp and paper manufacturing facility.
Typical cake concentrations varied from 40 to 47
percent solids without a press aid. Wastes from the
manufacture of pulp and/or paper would seem to work
particularly well with this equipment because of the
fibrous nature of the primary sludge.
Also of interest is the ability of the press to produce
an alum sludge cake of 40 to 60 percent solids using
soil as a press aid in one test and sawdust in another.
In each case, the press aid used was approximately
100 percent of the weight of dry solids of the alum
sludge.
Determination of the pressure profile is a function of
the sludge, the sludge blend, and the quantity and
nature of the press aid used. On primary and waste
activated sludges in the normal proportions (i.e.,
approximately 50-50) and on pure waste activated
sludge, the P roll pressures are usually tapered and
will vary from 10 kg/cm (56 Ib/in) on the first roll to
60-250 kg/cm (336-1,401 Ib/in) on the last roll.
The dewatered sludge shown in Figure 7-27 is from
the Portland Columbia River Wastewaler Treatment
Plant, and was dewatered during a demonstration
study at that facility. The activated sludge feed varied
from 2.5 to 3.5 percent solids, and each test was run
at approximately 100 percent of the press aid by
weight. Sawdust additive yielded a cake in the range
of 30 lo 40 percent solids, while the paper press aid
produced a cake from 35 percent to somewhat over
40 percent solids.
Solids capacity of the press varies from 225 to 600
kg/m hr (102 to 272 Ib/m hr) and an acceptable
hydraulic feed rate ranges from 1.6 to 3.2 l/s (25-51
gpm) on a 1-m (39-in) wide machine. The basic
press has been investigated for further dewatering of
cake derived from other dewatering equipment.
The press is of interest because it produces an
autogenous cake with a modest amount of press aid.
However, as of March 1987, there are no commercial
installations on wastewater sludges. Initial units have
been used in industrial and food processing
applications. This device is similar to the high
124
-------
Figure 7-27. Dewatered sludge at the Columbia River WWTP, Portland, OR.
in*" » i sff*
y> ' * . i
e .<•,. '•f "*,t
vl«"^.1
125
-------
pressure attachment available on the Parkson
Magnum 3000 and 3500 belt presses.
7.6.2 Som-A-System
The Som-A-System Screw Press consists of a
vertical, rotating screw enclosed by dual stainless
steel screens.'The screens and screw are encased in
a stainless steel housing with a removable cover on
each side. Tiny perforations in the inner screen allow
only water to escape. The outer screen has larger
holes and easily collects the pressate, which sprays
inside the housing and drains into a receptacle.
Brushes are located along the edge of the screw to
sweep the cake that builds up on the screen, allowing
a clear opening for the pressate to escape.
The feed enters at the bottom of the screw press. A
buildup of sludge cake on the screw is recommended
to get good sludge dewatering. As the pressate
drains, the cake becomes progressively drier and is
pushed to the top, where it is discharged into a
waiting dump truck or hopper. A back pressure
system Is located below the discharge chute and
gives the cake a final squeeze before discharge. One
plant, however, removed this cone, which collected
hairballs, with no adverse affect to its operation or
dewatering results (35). The Som-A-Press and
Som-A-System are shown in Figures 7-28 and
7-29, respectively.
Flguro 7-28. Functional schematic of Som-A-Press
Figure 7-29. Som-A-System (courtesy of Siomat Corp.)
Cake Discharge Chute
A. Feed enters at inlet.
B. Variable speed auger carries liquid/
solids stream vertically along the
dewatering screen barrel.
C. Most of the free liquid in the feed
discharges through the screen
openings in the initial 1 /3 to 112 of
the barrel length.
D, Solids build up, forming a plug.
E. Adjustable cone mounted on the
auger shaft restricts passage of the
cake and gives a final squeeze to the
mass.
F. Revolving plug cutter pushes solids
from the barrel into the discharge
chute.
Sludges that floe easily and are fibrous are the most
conducive to a screw press operation. Feed
concentration is critical to achieving high cake solids.
The higher the feed concentration, the higher the
cake solids and the unit capacity (kg/hr). Table 7-23
reports feed solids, cake solids, and solids recovery
from several different plants using the Som-A-
System (35). Key to the action of the unit is bridging
of the holes in the screen, because the bulk of the
particles in the sludge will be finer than the holes in
the screen (36). Consequently, proper sludge
conditioning is essential. Table 7-24 presents the
polymer usage of several plants using the Som-A-
System (35).
A slow screw speed will yield a better cake, although
it will also decrease throughput. High flow rates and
screw speeds generally result in a discharge of wet
sludge. A variable speed pump regulates the feed
rate to the screw press. At Pinetop, AZ the plant
generally keeps the feed rate at 2,5 l/s (40 gpm), near
the maximum. Sludge that has been aerobically
digested at a 20 day detention will easily yield an
acceptable 12-15 percent cake at a feed rate of 2.5
l/s (40 gpm). However, if the sludge had a lower
detention time, a feed rate of 2.5 l/s (40 gpm) would
produce a wetter cake.
Operation and maintenance of the Som-A-System
is simple. Depending on the sludge, the press can
normally be operated with only periodic checks. Many
plants simply turn the machine on in the morning and
periodically check the feed solids, the cake solids,'
and the level of the waiting dump truck. Some
operations require more attention to the feed sludge
126
-------
Table 7-23. Som-A-Systcm Operating Data
Plant Sludge Average Plant Flow Feed Solids TSS
Provo, UT
Feed Rale
Cake Solids Solids Recovery
Camden, NY
Churchville, NY
New Canaan, CT
Danville, VA
Pinetop, AZ
Sunriver, OR
Frisco, CO
Aerob. Digested
Aerob. Digested
Aerob. Digested
WAS/Stab. scum
(ram DAF
Aerob, Digested
Asrob. Digested
Asrob. Digested
mgd
0.6
0.11
0.25
16.2
0.4
0.5
1.0
percent
1-2
2.5-3.5
1.0-1.5
5-6
8
2
0.5-0.75
2
gprn
10-24
10.5
30-40
15
30-40
40
35-40
15-182
403
percent
10
12.3
12-17
21-23
28-30
12-15
7-12
11
percent
85
-
84-94
86
90
88-90
85 1
-
Anaer. Digested
1.5
30
7-15
87-94
1 Normally the solids recovery runs 90-94%.
2 Undersized polymer pump limits feed rate to 15-18 gpm - new pump ordered.
3 With larger pump, expect to run presses at 40 gpm.
Table 7-24. Som-A-System Chemical Conditioning Data
Plant Polymer
Camden, NY
Churchville, NY
New Canaan, CT
Danville, VA
Pinetop, AZ
Sunriver, OR
Frisco, CO
Provo, UT
Name
Percol 767
Percol 757
Percol 757
Cationic
Percol 757
Allied CC4450
Percol 757
Percoi 763
Dosage
5-20
1.57
6-9
21
8-10
21
12
21
$/lb
-
2.80
3.25
0.92
-
1.65
2.70
-
$/ton DS
-
4.39
24.38
19.32
27.00
34.65
32.40
-
to ensure that the proper concentration - and not
water - is being fed to the press. The unit is also
relatively easy to disassemble. General maintenance
involves routine lubrication and washing the screens
to prevent buildup of sludge, which can prematurely
wear the brushes. Repairs reported by plants have
been limited to replacement of inner and outer
screens and brushes. One plant replaced the brushes
after approximately 1,500 hours (35).
The low capital cost of this screw press is a primary
attraction and comparative economic evaluations
point favorably to it. For one plant, the Som-A-
System was approximately $55,000 less than a belt
filter press bid for the same job. It is ideal for
operations with limited space requirements, since the
system occupies, at a maximum, approximately 3 m2
(32 sq ft) of floor space.
Potential drawbacks include low unit capacity, higher
polymer dosage, and lower cake solids. Capacity of
the presses can be a deterrent because the small
throughput demands a multiplicity of units, which can
be more difficult to control. A few plants (35)
expressed disappointment about the amount of
polymer required, and were experimenting in an effort
to reduce the quantity.
7.6.3 CentriPress
Based on observed field demonstrations at the 1987
IFAT Conference held in Munich, Germany, there
have been significant improvements in the capabilities
of a newly designed solid bowl continuous flow
centrifuge. The improvements were in the area of
cake solids concentration. In testing, the centrifuge
was operated in parallel with a filter press system.
The new centrifuge design, called the CentriPress,
produced as a high a cake solids as the filter press
system. Figure 7-30 shows the CentriPress and
examples of cake that it produced.
A Model S2-1, 45-cm dia. x 135-cm long (18-in
x 53-in), centrifuge at the Marienfelde STP was
operating on digested primary and waste activated
sludge. This same sludge was fed to 91.5-cm
diameter x 274-cm long (36-in x 108-in)
centrifuges which were dewatering the plant sludge to
a cake product of approximately 22 percent TS. The
CentriPress was producing a granular cake of 30-32
percent TS. The "standard" centrifuges produced a
cake with a 60 percent higher moisture content. Both
centrifuge installations were recovering in excess of
90 percent of the feed solids and the products are
shown in Figure 7-30.
A larger unit, 91.5 cm dia. x 274 cm long (36 in x 108
in), is operating at Vienna, Austria WWTP. This unit is
dewatering a heated primary and waste activated
sludge to a cake solids content of 40-42 percent TS.
Results from the centrifuge are comparable to those
produced by a recessed plate filter press.
The manufacturer has advised that orders have been
taken in Europe for the new machine, and that
demonstration of the centrifuge capabilities were
initiated in the spring of 1987 in the United States.
Details of the centrifuge construction were not made
available prior to printing.
127
-------
Flguro 7-30, CentriPress and cake samples.
Humboldt-Wedag Centri Press
From Left to Right, Pressate, Press Cake, Ash Recycle, Centrifuge
Cake, and Centrate
Granular Centrifuge D (P + WAS) Cake @ 30-32% TS
128
-------
Test trials using a Humboldt-Wedag CP2-1 were
performed by the Metropolitan Sanitary District of
Greater Chicago (MSDGC) at the West-Southwest
STP. This plant currently employs high-speed
centrifuges for dewatering a digested primary and
waste activated sludge, which has an original solids
ratio of 0.21 PS: 0.79 WAS. The existing centrifuges
produce a cake of 14-16 percent TS.
The tests were conducted using two types of cationic
polymers as shown in Table 7-25. One of the
polymers was not cost effective for the plant's
digested sludge. The tests used different feed rates
and differential speeds, with the polymer adjusted to
maintain the TSS recovery in the range of 85-95
percent. The key results of Table 7-25, using
American Cyanamid 2540C polymer are as follows:
Cake Solids, %
Solids Recovery, %
Polymer Dosage, kg/Mg
$/Mg
26.2-33.9
78.4-97.9
3.23-15.93
5.86-29.22
Figure 7-31 shows the effects of polymer dosage on
the solids recovery of the CentriPress. About 5 kg/Mg
(10 Ib/ton) of cationic polymer was required to
maintain the solids recovery in excess of 90 percent
TS. Table 7-25 does not indicate that higher
dosages of polymer were beneficial to improve cake
solids, although recoveries above 95 percent were
achieved.
The use of low differential speeds appears to be the
key to achieving good cake solids. As shown in
Figure 7-32, there was a good correlation between
cake solids and centrifugal force at about 2,600 g's.
7.6.4 H1W Screw Press
This Korean screw press is being evaluated for
dewatering sludge from liquid to cake, Second stage
(cake to drier cake) operations have also been
evaluated. The HIW screw press, shown in Figure 7-
33, is continuously fed a polymer conditioned sludge.
Once inside the unit, the sludge receives a gradually
increasing pressure as it progresses through the
screw press. The maximum pressure before
discharge may exceed 10 kg/cm2 (147 !b/sq in). In
some instances the dewatering may be enhanced by
heating (a normal experience with screw presses)
prior to the dewatering screw. This screw press is
said to be relatively simple and easy to maintain.
Also, the low operating speed helps keep repair costs
to a minimum. HIW reports that there are over 100
units in operation (or installed) for various types of
wastewater treatment and are providing satisfactory
service. An adequate U.S. database is not yet
available. Some results reported are shown in Table
7-26.
More data and a better definition of the feed sludge is
required to fully evaluate the possibility of the screw
press replacing conventional dewatering equipment.
Past excessive secondary solids losses must be
evaluated as a function of the cake solids content
produced.
During May of 1986, the Municipal Sanitary District of
Greater Chicago (MSDGC) tested a pilot HIW screw
press. The unit was tested on primary and
anaerobically digested sludge at the West-
Southwest STP. The test was performed over a
period of two days and approximately twelve separate
runs were undertaken. Sludge flow rate, dilute
polymer concentration, and polymer flow were varied.
With an average sludge feed concentration of 4.5
percent, the test unit attained the following average
results.
Cake concentration: 17.5%
Solids recovery : 94.5%
Pressate concentration: 3,720 mg/l (0.43 Ib/gal)
Polymer usage: 8 dry kg polymer/dry Mg
solids (16 Ib/dry ton)
Based on these pilot-test results, the MSDGC
decided to purchase a full-size screw press, to be
delivered in mid-1987. MSDGC anticipates that a
full-size screw press may be a cost-effective
alternative to centrifugation due to the following
considerations:
Low initial cost
Lower electric power consumption
Equal to or higher cake concentrations
Slow operating speed (low G force)
Lower maintenance cost
Comparable polymer cost.
7.6.5 Sun Sludge System
The Sun System (Hi-Compact) of pressing sludge
was developed in Japan, and has been licensed for
marketing and manufacture for Europe and the United
States. The principle of the process is to develop a
structured material from a cake of poor dewatering
characteristics, and to form liquid channels. The cake
is then subjected to high pressures. To that end,
dewatered sewage sludge is reduced to pellets which
are subsequently coated by a powdery layer of a
drainage substance such as ash, pulverized coal, etc.
Compressing a stack of these pellets results in a
compact block interwoven with a network of drainage
layers; the water being removed by pressing flows
through a line of least resistance to the nearest
drainage layer as shown in Figure 7-34.
In the system, sludge is first dewatered by
conventional dewatering equipment such as vacuum
filters, centrifuges, or continuous belt filters to a 20-
25 percent solids concentration. This material is then
conditioned in a unit called a disintegrating pelletizer,
129
-------
Tables 7-25. Results of Chicago WSW CentrlPress Study
Run Machine Data Sludge Data
It
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
172
18
19
20
21
22
23
24
25
G-
Forco,
Q's
2,300
2,300
2.300
2,300
2,600
2,600
2,600
2,600
2,600
2,600
2,600
2,600
2,600
2,600
2,600
2,600
2,600
2,600
2,600
2.600
2,600
2,600
2,600
2,600
2,600
Diff.
Speed
3
2
1.8
2
2
2
3.5
5.5
5.5
2.2
2.8
7.5
5.8
2.5
2.5
2,7
2.5
2.8
2,9
2.2
2
2
5
1.5
1.2
Feed
Rate,
gpm
27
27
25
31
26.5
26.5
32
32
32
16
16
50
50
22
22
22
22
22
22
22
22
22
26.5
18
18
Feed
Cone.,
%
4.18
4.19
4.00
4.09
4.16
4.10
3.96
4,07
4.10
4.12
3.80
4.13
4.11
4.23
4.25
4.20
4.25
4.25
4.00
4.10
4.08
4.16
4.09
4.05
4.08
Feed
Solids,
ton/d
6,78
6.79
6.01
7.61
6.62
6.52
7.61
7.82
7.89
3.96
3.65
12.43
12.34
5.59
5.62
5.55
5.62
5.62
5.28
5.42
5.39
5.50
6.51
4.38
4.41
%
Volatiles
48.3
48.0
47.0
46.7
48.4
48.9
48.7
49.0
49.2
48.6
50.1
48.8
48.1
50.7
50.5
50.1
49.4
49.4
49.8
49.2
50.6
50.5
49.3
49.1
49.4
Polymer Dala
Flow
Rate,
gpm
3.45
3.09
3.28
4.10
3.9
3.9
3.73
2.63
4.58
3.27
2.67
5.57
5.57
3.91
4.4
3.2
2.73
2.73
2.4
2.4
3.57
3.50
3.50
2.5
2.9
Polymer
(Dry),
Ib/lon
14.67
13.12
15.73
8.41
10.61
10.28
9.42
6.46
11.15
19.83
17.57
10.76
11.93
21.84
24.45
31.85
33.25
33.25
31.12
30.31
45.34
42.8
36.16
38.39
44.23 -
Cake
Solids,
%
29.1
26.3
29.6
29.2
33.2
33.9
29.7
26.2
27.4
30.2
31.1
29.8
28.4
28.5
28.4
28.9
29.2
32.0
29.4
31.3
32.8
34.8
30.0
30.8
29.7
Performance
Centrate
Solids,
%TSS
4,200
1,700
5,000
1,000
3,200
2,700
7,200
10,000
4,200
1,900
2,400
1,400
3,700
2,900
1,500
1,300
2,900
1,700
1,100
2,200
1,300
2,200
2,400
1,000
1,200
Capture,
%
91.27
96.57
89.00
97.89
93.21
94.17
83.85
78.42
91.15
95.99
94.41
97.07
92.20
94.10
96.98
97.34
94.11
96.51
97.61
95.30
99.20
95.31
94.84
97.85
97.45
Cost,
$/ton
11.24
10.06
12.06
6.45
8.74
8.88
7.76
5.33
9.19
16.34
14.48
8.87
9.83
18.20
20.37
26.56
76.48
76.48
71.58
69.71
104.28
98.44
83.17
88.30
101.73
1 Tests 1 through 16 used American Cyanamid 2540C polymer.
2 Tests 17 through 25 used Allied Chemical Percol 778F525 polymer.
which first breaks and forms the sludge cake into
small particles and then coats the particles with a dry
powder, forming sludge-like pellets. The dry additive
used should be mostly water insoluble and should not
break up at the high pressures used. Materials such
as diotomaceous earth, gypsum, calcium carbonate,
incinerator ash, coal powder, bone meal, dried pulp,
sawdust, and soil have been used, either alone or in
combination with each other. The conditioners should
be added in the ratio of 40-60 percent by weight per
unit dry weight of the original sludge cake. The
effective sludge particle or pellet's diameter should
not be greater than 20 mm {0.8 in). Best performance
occurs when the effective diameter of the pellets is
between 3 and 5 mm (0.1-0.2 in). Also, the
conditioning agents should coat only the surface and
should not be kneaded into the sludge pellets for
maximum effectiveness.
The conditioned sludge cake particles are conveyed
to a hydraulic press where additional water is
removed, and a cake of greater than 40 percent
solids is produced. The pelfetized sludge is pressed
between two sheets of filter cloth that cover thick
plates that have a number of perforations 2-10 mm
(0.1-0.4 in) in diameter. Compression is carried out
in two steps. The initial compression step is usually at
15-25 kg/cm2 (210 Ib/sq in) for 45 seconds followed
by a pressure of 30 kg/cm2 (430 Ib/sq in) for 5
minutes. In practice, the compression has occurred at
15 kg/crn2 (210 Ib/sq in) for 45 seconds, followed by
a pressure of 30 kg/cm2 (430 Ib/sq in) for one minute.
The pelletized sludge cake is compressed by an oil
hydraulic cylinder to form a disc-shaped solid with a
40-55 percent solids concentration. As an example,
a mixture of primary and waste activated sludge
having a 2 percent solids concentration could first be
dewatered by a belt filter press to a solids
concentration of 25 percent and then, with the Sun
Sludge System, could be further dewatered to a
solids concentration of 55 percent.
130
-------
Figure 7-31. Effect of polymer dose on solids recovery.
Humboldt CentriPress CP2-1
Feed Rate = 25-32 gpm
Feed Solids = 3.0-4.1% TSS
At Chicago WSW WWTP:
Digested (0.21P:0.79WAS)
Polymer Used: American Cyanamid 2540C
100
90
I" 80
70
10
11
12
Polymer Dose, Ib/ton
Figure 7-32. Cake solids vs. differential speed.
Humboldt Centri Press CP2-1
• Feed Rate = 25-32 gpm
Feed Solids = 3.0-4.1% TSS
At Chicago WSW WWTP:
Digested (0.21 P:0.79WAS)
Polymer Used: American Cyanamid 2540C
35
34
33
32
H 31
£
rf 30
T3
29
28
27
o
C/J
o
28
25
• G = 2,300 Gs
VG = 2,600 Gs
Differential Speed, rpm
The Ashigara Works of Japan has successfully been
using this process for waste activated sludge
treatment since mid-August 1982. The excess
sludge is dewatered by a belt press to a water
content of 80 percent, then pelletized and conditioned
with incinerator ash and further dewatered to a water
content of 50 percent or less. Ash is added in the
ratio of 10 to 15 percent by weight of the amount of
belt press cake (or 50-75 percent of the dry solids).
Sludge cake is incinerated and heat recovery
equivalent to 30 l/hr (7.9 gal/min) of fuel additive is
practiced.
At the 1987 IFAT show in Munich a field
demonstration of the process produced a cake of 55
percent from a 32 percent sewage cake mixed with
sludge ash (50 percent by dry solids weight) every 3
minutes. The unit was pilot scale producing in excess
of 1,000 kg cake/hr (2,204 Ib/hr). This would be
equivalent to 370 kg/hr (816 Ib/hr) of sewage sludge
solids. In this demonstration, pressures up to 60
kg/cm2 (853 Ib/sq in) were employed and the press
time was shortened to about three minutes.
While the product from the press is very hard, it is
also quite friable. It can be easily fragmented into
particles which are dry to the touch and can easily be
transported pneumatically. The pelletizing/pressing
operation at Munich reduced the moisture content
from 3.5 kg H20/kg TS to 1.3 kg H2O/kg TS (sludge
only basis). The feed would have been suitable for
boiler feed and would produce an equilibrium
temperature of about 1,090°C (2,000°F).
While the process mechanics look favorable, the
machine design capable of long-term operation at
50-60 kg/cm2 (710-850 Ib/sq in) will need to be
further evaluated.
7.7 References
1. Parkson Corp.
Manual.
Operation and Maintenance
2.
3.
Belt Filter Press Survey Report. American Society
of Civil Engineers, New York, NY, 1985.
Design Information Report on Belt Filter Presses.
U.S. Environmental Protection Agency, Center for
Environmental Research Information, Cincinnati,
OH, 1985.
4. Peterson, R.L., Rubel and Hager Inc. Belt Filter
Press Design. Presented at the 54th Annual
_j Water Pollution Control Federation Conference,
7 Detroit, Ml, 1981.
5. Ashbrook-Simon-Hartley,
Maintenance Manual.
Operation and
6. Semon, J. Post Lime Stabilization, JNEWPCA
16(1), 1982.
7. Boll, J.E. Potassium Permanganate for Sludge
Odor Control. Carus Chemical Co., LaSalle, IL,
1986.
131
-------
Figure 7-33. Schematic of the screw press dewatering system (courtesy of Hoilin Iron Works JHIWJ).
Water Polymer Powder
1 n r*^
Water
"1 nap
From Sludge
Thickener
, Polymer Feed
Pump
Polymer Dissolving
Tank
n
Water Shower or
Air Spray
Polymer and Sludgo
Mixing Tank
Sludge Feed
Pump
Tablo 7-36. Test Results for HIW Screw Press
Feed Cake Solids
Sludgo P/S Ratio Solids Solids Recovery
Digested A
Digested B
Primary A
Primary B
Paper Mill 1
Paper Mill 2
Paper Mill 3
Paper Mill 4
10/90
10/90
NR
NR
0/100
60/40
50/50
0/100
percent
4.65
4.93
2,85
2.37
3.45
4.08
2.95
2.4
percent
20.9
25.3
20.5
21.2
48.6
44.6
42.3
23.0
percent
93.0
97.8
95.0
95.9
99.0
98.9
98.9
95.4
Polymer
Dosage
kg/Mg
9.1
7.6
13.4
16.0
1.0
NR
NR
NR
' Not reported.
8. Metcalf and Eddy, Inc. Wastewater Engineering.
McGraw-Hill, New York, NY, 1982.
8. Vesilind, P.A. Treatment and Disposal of
Wastewater Sludges. Ann Arbor Science
Publishers, Ann Arbor, Ml, 1979.
10. Manual of Practice No, 8 - Wastewater
Treatment Plant Design. Water Pollution Control
Federation, Washington, DC, 1982.
11. Standard Methods for the Examination of Water
and Wastewater. American Public Health
Association, New York, NY, 1985.
12. Ambler, C. The Evaluation of Centrifuge
Performance. Chemical Engineering Progress
48(3): 150-58, 1975.
13. Nissen, J.A. and P.A. Vesilind. Preserving
Activated Sludge. Water and Sewage Works,
August, 1974.
14. Albertson, O.E. and E. Guidi, Jr. Advances in the
Centrifugal Dewatering of Sludges. Water
Sewage Works 114RN:R-133, 1967.
15. Owen, W.F. Comparative Testing of High and
Low Speed Solid Bowl Centrifuges, Littleton
Engtewood Wastewater Treatment Plant. Owen
132
-------
Figure 7-34. Sludge press function of the Hi-Compact
method (courtesy of Humboldt-Wedag),
iwinnn n
^n
Stack of
CoatedPellets
Rltrate
n n ft n pi n n
Filter Cloth
Compact Cake Interwoven •
with a Network of
Drainage Layers
Engineering & Management Consultants, Inc.,
Engiewood, CO, 1983.
16. Uohida, B. Acceptance Tests for the Viscotherm
Backdrive System, Internal Memo. Seattle Metro,
WN, October 15, 1982.
17, Vesilind, P. Aarne and J. Loehr. Unpublished
Report, 1970.
18. Christensen, G.L. Communication. JWPCF
55(4):417-19, 1983.
19. Gale, R.S. Filtration Theory with Special
Reference to Sewage Sludge. Water Pollution
Control (GB) 66:662, 1967.
20. Kozeny, J. S-Ber. Weiner Akad., ABTA Ha
136:271, 1927.
21. Carmen, P.C. Journal Society Chemical Industry
57:225, 1938.
22. Coackley, P. Laboratory Scale Filtration
Experiments and Their Application to Sewage
Sludge Dewatering. In: Biological Treatment of
Sewage and Industrial Wastes, Vol. II, edited by
McCabe and Eckenfelder, Reinhold Publishing,
New York, NY, 1957.
23. Mininni, G., L. Spinosa, and A. Misiti. Evaluation
of Filter Press Performance for Sludge
Dewatering. JWPCF 56(4): 331-36, 1984.
24. Wilhelm, J.H. The Use of Specific Resistance
Data in Sizing Batch-Type Pressure Filters,
JWPCF 50(3):471-83, 1978.
25. Christensen, G.L and J.R. Sipe. The Application
of Sludge Filtration Models to Media Selection. In:
Proceedings of the 14th Mid-Atlantic Industrial
Waste Conference, Ann Arbor Science, Ann
Arbor, Ml, 1982.
26. Notebaert, F.F., D.A. Wilms, and A.A. Van Haute.
A New Deduction with a Larger Application of the
Specific Resistance to Filtration of Sludges,
Water Research 9:667, 1975.
27. Greenwood, S.J. and W. Maier. Computer
Simulations and Process Studies of Pressure
Filtration for Sludge Dewatering. Metropolitan
Waste Control Commission of the Twin Cities
Area, conducted by the Department of Civil and
Mineral Engineering, University of Minnesota,
1982.
28. Webb, W.J. A Study of Conditioning Sewage
Sludges with Lime. Water Pollution Control (GB)
73:192, 1974.
29. Sontheimer, H. Effects of Sludge Conditioning
with Lime on Dewatering. In: Proceedings of the
3rd International Conference on Advances in
Water Pollution Research, Munich, 1967.
30. Christensen, G.L. and D.A. Stule. Chemical
Reactions Affecting Filterability in Iron-Lime
Sludge Conditioning. JWPCF 51(10):2499-512,
1979.
31. Wu, Y.C., E.D. Smith, and R. Novak. Filterability
of Activated Sludge in Response to Growth
Conditions. JWPCF 54(5):444-56, 1982.
32. Terraqua Resources Corp., Filter Press
Operations Survey for the City of Baltimore, MD,
1984.
33. Ohara, G.T., S.K. Raksit, and D.R. Olson. Sludge
Dewatering Studies at Hyperion Treatment Plant.
JWPCF 50(5):912-25., 1978.
34. Bennett, E.R., D.A. Rein, and K.D. Lisntedt.
Economic Aspects of Sludge Dewatering and
Disposal. JEED ASCE 99:55, 1973.
133
-------
35. Survey of Eight Wastewater Treatment Plants,
Enviro Enterprises Inc., Salt Lake City, UT, April
1987.
36. Lash, L., P. Steele, and J.H. Petersen. Sludge
Dewatering Innovation - The Screw Press.
Proceedings of the Utah Water Pollution Control
Association, Provo, UT, 1985.
134
-------
Chapters
Case Studies: Air Drying Systems
8.1 Introduction
Two air drying technologies are described in this
chapter: upgraded sand drying beds that use polymer
to improve dewatering and paved drying beds that
use a composter/agitator to increase the rate of
dewatering.
8.2 Upgraded Sand Drying Beds
A site visit was conducted to write the case study for
an upgraded sand drying bed at Elgin, Illinois.
Information on three other upgraded sand drying beds
was obtained through personal communication.
8.2.1 Detailed Case Study - Elgin, Illinois
The Sanitary District of Elgin, Illinois operates three
wastewater treatment plants. The North Plant is a
0.15-m3/s (3.5-mgd) activated sludge plant.
Following is a description of the plant as it operated in
1986.
The activated sludge process is a conventional,
complete-mix design with surface mechanical
aerators. The plant has an average daily flow rate of
0.14 m3/s (3.15 mgd) and is undergoing a
construction expansion to 0.25 m3/s (5.75 mgd).
The waste activated sludge (WAS) at the North Plant
is wasted to the primary clarifiers. The combined
primary sludge (P) plus WAS is digested in two-
stage anaerobic digesters. Following digestion, the
sludge is dewatered on sand drying beds and the
dewatered sludge is hauled away to various land
application sites. A plant schematic for the North
Plant is shown in Figure 8-1.
Polymer is added in-line to the sludge as it flows
from the primary digester to the secondary digester.
The polymer serves as a settling aid during
supernatant removal operations. Using polymer
enables the District to remove more supernatant and
produce a thicker digested sludge. The polymer
dosage applied to the digester is about 114 liters (30
gal) of liquid cationic polymer to 454 m3 (120,000 gal)
of sludge, four times per month. The approximate
polymer dosage is 19.5 g/kg (39 Ib/ton) of solids, or
about $7.16 per Mg ($6.50/ton) of solids. The first
stage digester is both heated and mixed, while the
second stage digester is neither heated nor mixed.
The digested P + WAS has an average solids
concentration of 5 to 6 percent.
Polymer is also added to the sludge as it flows onto
the drying beds. The approximate polymer dosage is
32 g/kg (65 Ib/ton) of solids, or about $12.04/Mg
($10.92/ton) of solids. (This polymer dosage is higher
than average due to inefficient mixing, as described
below.) Polymer addition to the sand drying beds and
to the anaerobic digesters has been practiced for at
least 8 years. The Sanitary District of Elgin first
constructed the North plant in 1962 to a design
capacity of 0.035 m3/s (0.8 mgd). In 1972 the plant
was expanded to a design capacity of 0.11 m3/s (2.4
mgd). Later the plant's average design capacity was
upgraded on paper to 0.15 m3/s (3.5 mgd). By the
mid-1970s the plant, during the spring of each year,
produced more sludge than could be dewatered. This
was the time when polymers were first tested to aid
dewatering on the sand drying beds. Soon thereafter,
the use of polymer became a common practice. The
main reason for its use is that the sludge dewaters
easier and faster with polymer. The total dewatering
time required is usually 3 to 4 weeks during the non-
winter months. If polymer is not added to the sludge,
the dewatering time is typically 5 to 6 weeks. Sludge
is applied to the beds year-round even though the
sludge freezes fairly often during winter months.
As the digested sludge flows toward the drying bed
and discharges onto a concrete splash pad, liquid
polymer is added to the sludge through a separate
hose that discharges onto the same splash pad. (This
is an unusual and inefficient method for polymer
addition. It is more typical and more efficient for
polymer to be added in-line, upstream of the drying
bed.) The concrete splash pad and sludge feed line
are located between two sludge beds. Gates on
either side of the splash pad can be raised or lowered
to allow sludge to flow onto either the right or left
sludge bed. Liquid sludge is added to the drying bed
to an initial depth of about 23 cm (9 in). During warm
summer months, a dewatered sludge with a solids
concentration of between 28 to 32 percent solids can
be removed in 3 to 4 weeks. During spring and fall
months, the dewatered sludge solids concentration
135
-------
FIguro 8-1. Schematic of the North Plant, Elgin, IL.
Parshall Flume —^ ^_ Sludge Division Box
Flow Division
Chamber
Influent
Sewer
Sewage Grinder
& Parshall Flume
HI
Preaeration
& Grit Removal
Tank
Control Building
^
Digester
phlo'rlneh I ^
lation Pumping Chamber
Return Sludge Pumping Station
Effluant Sewer
to Fox River
Sk
dge Dr
ying B<
ids
-•*+! -—i-l—».•
I I I, I I
Sludge Drying Beds
I-
LEGEND
Sowage
Sludge
________ chlorine
__.__.— . Supernatant
Influent Sewers
can be as low as 20 percent. During winter months,
sludge has occasionally been removed from the beds
in a frozen state.
Dewatered sludge is removed from the drying beds
with a front-end loader. The loader can easily
remove the sludge because the sand bed Has
concrete strips that run lengthwise down the entire
length of the bed. The concrete strips are 0.7 m (2 ft,
3 in) wide, 20 cm (8 in) thick, and are separated by
open sand strips 0.8 m (2 ft, 9 in) wide. Vitrified clay
tile underdrains are located beneath the sand strips.
Sand bed filtrate is returned by gravity to the plant
headworks.
Following mechanical removal of most of the
dewatered sludge, small quantities of sludge
remaining in the sand, alongside the concrete strips,
or at the edge of the beds are removed manually with
a rake. This step prevents clogging of the sand
filtering media. The sand is then loosened as needed
with a rake to allow better percolation through the
sand. If the sand is not loosened, it becomes packed
down, reducing percolation. Following raking of the
sand strips, the entire bed surface is covered with
about a 2.5-cm (1-in) layer of sand. The sand is
scraped off the concrete strips and is used to fill the
void areas created during the last cleaning of the bed.
Because of the construction of the current plant
expansion, the total number of beds available for
sludge application was reduced from 14 to 7 starting
in August of 1984. The total sludge bed area before
the reduction of bed area in 1984 was about 2,750
m2 (29,600 sq ft), or an average of 116 m2 (2,114 sq
ft) per bed. In the period of August 1983 through July
1984, the average daily wastewater flow rate was
0.13 m3/s (3.1 mgd). The sludge production was
3,230 m3 (854,000 gal) at an average digested sludge
solids concentration of 5.6 percent. Thus, the
digested sludge production is about 1,350 dry Mg of
solids/yr/m3/s (65.4 dry tons/yr/mgd) of average plant
flowrate. This amounts to only 3,710 kg solids/d/m3/s
(358 Ib/d/mgd). One reason for such a low amount is
136
-------
that the influent sewage strength is quite weak; during
the period of August 1983 through July 1984, the
average influent BODs and suspended solids
concentrations were 98 mg/L and 126 mg/L,
respectively.
The District Engineer cited industrial waste
contributions as one of the problems at the North
Plant. For the activated sludge system, the average
F/M ratio was 0.6 and the average aeration basin
hydraulic detention time was 3.9 hours during 1983-
84. The annual average solids loading rate applied to
the North plant sludge beds during the year before
the reduction to 7 beds, based upon an average
wastewater flow of 0.13 m3/s (3.1 mgd), is 66.4 kg
solids/m2/yr (13.6 Ib/sq ft/yr).
The polymer system was installed by plant operators
at a minimal construction cost. The operation and
maintenance cost for the drying beds is broken down
in Table 8-1, which describes the cost to fill and
clean one average size bed.
During the period of August 1983 through July 1984,
the total weight of dry sludge solids remaining
following anaerobic digestion was 183 Mg (201 tons).
At an average digested sludge solids concentration of
5.6 percent, the total volume of sludge dewatered
was 3,230 m3/yr (854,000 gal/yr). At a depth of 23
cm (9 in) of sludge and an average sludge bed area
of 116 m2 (2,114 sq ft), the average volume of sludge
applied per bed is 45 m3 (11,900 gal). This volume of
sludge results in a total number of sludge applications
to a drying bed of 72 per year, or 5.1 applications per
year per bed. This total is low because of the low-
strength influent wastewater and low sludge
production; the sludge beds were not used at their full
capacity in 1983-84. Based upon an average cost of
$135.95 per sludge bed application (from Table 8-
1), the total annual cost during 1983-84 was
$135,95 x 72, or $9,800/yr.
Currently, with about two-thirds of the North plant
sludge beds removed, only a portion of the sludge
can be dewatered on the sand drying beds. Dry
sludge is hauled away by a contract hauler to various
land application sites for a 1986 price of $7.52/m3
($5.57/cu yd) of dry sludge. Assuming the dewatered
sludge cake has a solids concentration of 30 percent
and a specific gravity of about 1.15, the cost of
hauling dry sludge is $21/Mg ($19/ton) of dry solids.
Undewatered liquid sludge is hauled away to land
application sites for 0.77 cents/I (2.9 cents/gal) of
liquid sludge.
8.2.2 Case Study - Belleville, Illinois
The following case study describes sludge dewatering
at the wastewater treatment plant for Belleville, Illinois
as it operated in 1986. The plant, located about 32
km (20 mi) southeast of St. Louis, is a 0.4-m3/s (8-
mgd) activated sludge wastewater treatment plant.
Current average daily flows are about 0.28 m3/s (6.5
mgd). Waste activated sludge is returned to the
primary clarifiers and the combined sludge is pumped
to anaerobic digestion. Sludge treatment at the plant
consists of two-stage anaerobic digestion and
dewatering on sand drying beds or storage in
lagoons.
There are 11 sand drying beds and a total sludge bed
area of 11,700 m2 (125,900 sq ft). A liquid cationic
polymer is added to the digested sludge before it
flows onto the drying beds. The typical polymer
dosage is 11 g of liquid polymer/kg (22 Ib/ton) of
solids, or about $H.62/Mg ($10.54/ton) of solids.
The polymer is pumped into the sludge discharge line
from the anaerobic digesters at a point about 244 m
(800 ft) before the sludge beds. Digested sludge is
normally applied to the beds at a solids concentration
of 3 to 3.5 percent and at a depth of 36 to 38 cm (14
to 15 in).
The sludge beds are used regularly from March
through October. During the winter months, the
sludge is only occasionally applied to the beds since
dewatering has nearly ceased. The sludge takes
normally about 6 to 8 weeks to dry with the polymer
and about 12 weeks without polymer. Application of
polymer to the sand drying beds has been practiced
at Belleville since 1965. The Wastewater Treatment
Division Superintendent considers it more cost-
effective to dewater sludge on sand drying beds with
polymer due to the reduced drying time provided by
the polymer.
The dewatered sludge is removed from the drying
beds at about 30 to 35 percent solids. Sludge is
removed from the drying beds manually with shovels.
Two men shovel sludge from the bed into a truck for
transfer to a stockpile site. Stockpiled sludge is
removed from the plant site for use by farmers,
gardeners, nurseries, etc. Sludge removal from one
bed takes about 24 man-hours. Cost estimates for
the sludge bed operations for 1985, prepared by the
Wastewater Treatment Division Superintendent, are
presented in Table. 8-2. The total operation and
maintenance cost estimate is $21,236/yr, or
$56.55/dry Mg ($51.30/dry ton) of solids. The solids
loading rate during 1985 is 54 kg/m2/yr (11 Ib/sq
ft/yr).
8.2.3 Gwmnett County Water Pollution Control
Department, GA
The Gwinnett County Water Pollution Control
Department, located about 40 km (25 mi) east of
downtown Atlanta, operates 11 wastewater treatment
plants. Polymer addition to improve dewatering on
sand drying beds is practiced at three of the plants.
The design capacities of the three plants are 0.20,
0.13, and 0.09 m3/s (4.5, 3, and 2 mgd). The sludge
bed area at these plants is 2,900 m2 (31,000 sq ft)
for 9 beds, 2,000 m2 (21,000 sq ft) for 6 beds, and
930 m2 (10,000 sq ft) for 3 beds, respectively.
137
-------
Table 8-1. Elgin, IL North Wastewater Treatment Plant Operation and Maintenance Costs for One Drying Bed Cycle (average
bed area = 196 m2 (2,114 sq ft)
Operation and Maintenance Costs
Horn
Labor - Add sand
Add sludge
Add polymer
Labor - Clean bed
Haul dewatered sludge
Electricity - 2,2-kW (3-hp) sludge feed pump
0,25-kW (0.33-hp) polymer pump
Fuol, diesol - for loader, haul truck, and sand
replacement
Polymer - 78 1/45,000 1 sludge, 1.01 kg/I {8,4? ib/gal)
Sand, replacement
Total Cost por Bod
Units
2hr
2 hr
0.12 MJ
(3.3 kWn)
571
{15 gal)
81 kg
(178 Ib)
8,845 kg
{1 9,500 Ib)
Unit Cost
$15/hr
$l5/hr
S1.67/MJ
($Q.06/kWh)
$0.22/1
($0.85/gal)
$0.37/kg)
($0.168/lb)
$3.73/Mg
($3.38/ton)
Cost
$30.00
$30.00
$0.020
$12.75
$30.00
$33.00
$135.95
Table 8-2. Sludge Bed Operation - Belleville, IL, 1985
Eleven (11) Drying Beds -125,900 total sq ft
Eslimaio Costs
Bod filling: 28 beds x 8 hr = 224 x $14 =
Stripping: 28 beds x 24 hr = 672 x $13 =
Sand - Power-Vehicle, elc.
Polymer: $l0.54/dry ton x 414 tons =
Total
$3,136.00
$8,736.00
$5,000.00
$4,364.00
$21.236.00
$21,236 * 414 dry tons = $51.30/dry ton
Tho sludge bods wore constructed in 1950, and the polymer
systom in use at this time was purchased in 1967. The vehicle
used for stripping the beds is a 1971 Chevrolet pickup truck that
was modified for sludge bed use at a cost of $2,500 in 1984.
AH three wastewater treatment plants are activated
sludge plants with single-stage nitrification and no
primary clarifiers. The two largest plants also remove
phosphorus with the addition of alum to the
secondary clarifier. All sludge is aerobically digested.
Sludge is typically removed from the aerobic digester
at about 4 percent solids for the two largest plants
and at 2 percent solids at the smallest plant. At two of
the three plants, a liquid cationic polymer is pumped
into the digested sludge line as sludge flows to the
drying bed. At the third plant, a crude mixing box is
used to allow the sludge to mix with the polymer. The
polymer dosage is 5 to 6.5 g liquid polymer per kg
(10 to 13 Ib/ton) of solids, at a cost of $8-10 per Mg
($7-9/ton) of solids.
Sludge is applied to the sand drying beds year-
round. The approximate depth of sludge applied is 10
to 13 cm (4 to 5 in) per application. In the summer
the typical drying time is 2 to 3 weeks. The sludge
surface on the beds normally cracks overnight when
using the polymer. Without polymer, the typical drying
time is 4 to 5 weeks, and the sludge surface takes 2
to 3 days before cracking. The solids loading rate
achieved at each of the three plants using polymer on
the sand drying beds ranges from 200 to 220
kg/m2/yr (41 to 44 Ib/sq ft/yr).
Polymer has been applied to the sand drying beds for
2 1/2 years at the largest plant, 1 1/2 years at the
0.13 m3/s (3 mgd) plant, and nearly 1 year (by
February 1987) at the smallest plant. Sludge is
removed manually with pitchforks at two of the plants.
The total labor time required is 9 to 10 man-hours
per bed. At the third plant, concrete strips in the bed
allow the use of a front-end loader to remove
sludge. Use of the front-end loader cuts the labor
time to only 3 man-hours per bed.
Plant administrators have plans to eliminate the use
of drying beds; rainy weather severely hampers
dewatering on the sand beds and there are neighbors
within about 180 m (200 yd) of the plants who
occasionally complain about odors. A belt press in a
trailer has recently been purchased. The belt press
has dewatered the sludge to greater than 20 percent
solids. It is uncertain if a belt press or centrifuge will
be used for future mechanical dewatering. Dewatered
sludge is currently hauled about 16 to 19 km (10 to
12 mi) to a landfill for disposal.
8.2.4 Chicago, Illinois
The Metropolitan Sanitary District of Chicago
operates the West-Southwest Wastewater
Treatment Plant (WSW Plant). Following is a
description of the plant as it operated in 1986. The
plant has an average daily design capacity of 53 m3/s
(1,200 mgd) of activated sludge secondary treatment
and a design capacity of 26 trfl/s (600 mgd) of Imhoff
tanks available for primary treatment. In addition,
138
-------
there are conventional primary clarifiers with a
maximum hydraulic design capacity of 48 m3/s (1,100
mgd). The average daily flow rate is about 35 m3/s
(800 mgd), divided nearly evenly between the Imhoff
tanks and the remainder of the WSW Plant. There is
an annual average Imhoff tank sludge solids
production of about 64 Mg/d (70 tons/d); of this, about
36 to 41 Mg/d (40 to 45 tons/d) of solids are applied
to sand drying beds, and about 23 to 27 Mg/d (25 to
30 tons/d) of solids are applied directly to lagoons.
Polymer has been used nearly continuously to aid
dewatering on the sand drying beds at the WSW
Plant since about 1969-1970. Sludge is applied to
the drying beds regularly from April until November.
For the winter period, the last sludge application is
usually in December, with a top-off application in
January/February, Dewatered sludge is removed with
a diesel-powered digging machine that operates on
a system of rails. This digging machine is also used
to break up the surface crust that forms on the drying
sludge. The original rail sludge removal system was
installed in 1931.
A liquid cationic polymer is used to aid dewatering.
The polymer is diluted with 9 I of water/I of polymer
and is stored in a 360,000-1 (94,000-gal) tank. A
metering pump discharges polymer solution into the
sludge lines as the sludge flows to the drying beds.
The digested Imhoff tank sludge is removed at an
average solids content of about 5.5 to 6 percent.
Sludge is applied on the sludge beds to an average
depth of about 28 cm (11 in). There are 12 sand
drying beds and a total bed area of 11.0 ha (27.3
acres). The average polymer dosage applied to the
sludge beds is 15 to 20 g of liquid cationic polymer/kg
(20 to 40 Ib/ton) of solids. The polymer cost is
$0.17/kg ($0.075/lb) as delivered, which results in a
polymer cost of $2.50 to $3.30/Mg ($2.25 to
$3.00/ton) of solids.
Dewatered sludge is removed from the sludge beds
after an average non-winter drying time of 35 days
during 1985. This drying time is probably several days
longer than normal since 1985 was considered to be
a wet year. The average dewatered sludge solids
content during 1985 was 33 percent. Dewatered
sludge is removed from the plant site by rail car and
is stored temporarily in piles in a storage area off-
site. The sludge is used by various non-agricultural
users, by the Parks District, and for reclamation of
completed sanitary landfills as a topsoil material.
During 1985 a total of 12,468 Mg (13,744 tons) of dry
sludge solids were applied to the sand drying beds.
This total results in an annual solids loading of 113
kg/m2/yr (23.1 Ib/sq ft/yr). The total operation and
maintenance costs for operation of the sand drying
beds during 1985 are as follows:
Item Cost
Polymer $3/Mg ($3/ton) dry solids
Maintenance $36/Mg ($33/ton) dry solids
Labor $21/Mg ($!9/ton) dry solids
Total O&M Cost $60/Mg ($55/ton) dry solids
One reason the maintenance cost was so high in
1985 is that the digging machine used to remove the
sludge is a very old piece of equipment that requires
frequent repairs. A new digging machine, expected to
be operational in September 1986, has an estimated
cost of $1,700,000.
8.3 Paved Drying Beds
A post construction evaluation of a sludge handling
system presently used by the City of Fort Worth,
Texas, was prepared for the U.S. Environmental
Protection Agency (1). This section is based on this
evaluation. The system to be evaluated was a
tractor-mounted auger aerator dewatering device.
Although the drying beds in Fort Worth are unpaved
sand beds, the paved bed concept is still applicable
to this case study.
8.3.1 Post Construction Evaluation - Fort Worth,
Texas
The City of Fort Worth purchased an auger aerator
machine in March of 1984. The City has been using it
since that time in their sludge handling facilities to
increase the drying rate of sludge on the sludge beds,
reduce insect problems, and to windrow dried sludge
to facilitate removal from the sludge beds and loading
in trucks.
The Village Creek Wastewater Treatment Plant is a
4.38-m3/s (100-mgd) design flow activated sludge
wastewater treatment plant, owned and operated by
the City of Fort Worth, Texas. This wastewater
treatment facility includes primary sedimentation
tanks, diffused air activated sludge treatment,
secondary clarification, effluent filtration, and
chlorination prior to discharge to the Trinity River.
Existing plant flow has averaged 3.87 m3/s (88.4
mgd) and has varied between 3.36 m3/s (76.8 mgd)
and 4.75 m3/s (108.4 mgd) over the period of
October 1982 to July 1984. AH sludge production and
performance figures are based on monthly averages
of plant performance for this 22-month period. The
average BOD and suspended solids concentrations
for influent sewage are 271 and 240 mg/l,
respectively. Primary and secondary sludges are
collected and handled separately. Primary sludge is
collected and thickened to about 5 percent solids
prior to digestion in six anaerobic digesters. Waste
activated sludge is collected and thickened by
centrifuges to an average of 3.9 percent solids prior
to digestion in four anaerobic digesters.
139
-------
Raw sludge production and feed to the digesters has
averaged 116,000 kg/d (256,000 Ib/d). After digestion
and volatile solids reduction, the average dry solids to
the sludge drying beds are 82,100 kg/d (181,000
Ib/d). These solids are sent to the drying beds as
liquid sludge at an average rate of 0.029 m3/s (0.667
mgd), which reflects an average digested sludge
solids concentration of approximately 3.25 percent
solids. The drying beds are located approximately 1.3
km (0.8 mi) from the treatment plant.
The sludge drying beds are located on a 120-ha
(296-ac) site and the total bed area including all 52
beds is approximately 93.5 ha (231 ac). At the
present time, eight drying beds have been removed
from service to provide a buffer zone between the
sludge drying area and adjacent residential housing.
This effectively removes 15.5 ha (38.3 ac), thereby
leaving 78.0 ha (192.7 ac) of sludge beds available
for use.
There are no detailed records of the sludge drying
bed performance over the past years. The
engineering staff felt that the available sludge drying
bed area was adequate and probably matched the
4.38 m3/s (100 mgd) design capacity of the treatment
plant during wet years. In dry years the operators felt
that they could complete two full drying cycles on
each bed. There appears to be little or no excess
capacity in the sludge drying beds.
During normal operation of the sludge drying beds,
the City of Fort Worth had several concerns. There
has been a constant and chronic odor problem
associated with the sludge drying beds. The odor
nuisance was compounded by recent construction of
new houses adjacent to the sludge drying bed site.
The land, which was previously zoned industrial or
commercial, had been re-zoned to residential. As
new homes are built and families move in, the impact
of fly and odor problems at the drying bed site will
become more critical. The Village Creek drying beds
have experienced a severe psychoda fly problem
especially in the spring months. A secondary, also-
less troublesome, fly problem has existed in the fall.
In the summer and winter months, psychoda flies
have not been a severe problem.
The dense fly population, the odor, from the beds, and
the close proximity of new homes combine to create
a serious public relations problem. The operators
report that at one time the psychoda fly population
consisted of very small flies that could pass through a
typical house screen. One can imagine the impact
this had on neighboring residents.
The flies breed in the dry crust that forms on the top
of the sludge as it sits in the drying bed. The auger
aerator is used to break up this crust and mix it with
the wet sludge, which has two beneficial effects. First
by breaking up the upper layer of crust, wet sludge is
exposed to the atmosphere to accelerate drying. It
also mixes fly eggs and larva into the sludge and
removes the dried crust, on which the flies land and
deposit their eggs. The effect of the tractor on insect
nuisance is discussed further in the qualitative
analysis of the machine later in this section.
The device is a four-wheel tractor with a hydraulic
motor-driven horizontal auger mounted in front. It
was originally designed as a high-production earth-
handling machine that would backfill cross-country
pipeline trenches with loose excavated material that
had been stockpiled alongside the pipeline trench.
The engine is a 168-kW (225-hp) turbocharged
after-cooled diesel. The tractor is a four-wheel
drive unit and can be steered with the rear axle, the
front axle, or with both axles. At the driver's option,
both axles can be operated in tandem to "crab" the
machine. The all-wheel drive permits the tractor to
travel in wet sludge beds without getting stuck. A
photograph of the tractor auger machine is shown in
Figure 8-2.
The auger is protected from excessive wear by
replaceable inserts. In practice, the auger is driven
through wet sludge beds to break up the crust, and
through drier sludge beds to turn the sludge and
expose new surfaces for drying. According to one
tractor operator, it takes the machine about 1 hour to
completely mix a 2-ha (5-ac) wet sludge bed filled
to its normal depth of 0.46 m (1.5 ft). It takes 2 to 3
hours to turn and windrow a drier sludge bed.
The auger aerator machine is currently handling a
digested sludge drying bed input of over 82 Mg/d (90
tons/d) operating 8 hours/d and 5 days/week. The
treatment plant staff indicated that they felt they could
more effectively control insects if they had a second
unit. However, the one unit brought great
improvements in operations.
The costs of utilizing the auger aerator machine are
calculated and shown below and in Table 8-3.
Annual costs are estimated based upon 2,000 hours
of machine operation annually. Fuel costs are based
upon a fuel consumption of 2.6 liters (7 gal)/hr at a
cost of $0.27/liter ($1.00/gal). Present worth is based
upon a 10 percent interest rate and an equipment life
of 10,000 hours, which is equivalent to 5 years. Labor
cost is assumed to include a single operator at a cost
of $12.00/hr.
The cost per metric ton of dry solids is determined by
dividing the total annual cost by the mass of dry
metric tons influent to the beds, annually at 29,900
Mg dry solids/yr (33,000 tons/yr). Thus, the cost per
metric ton was computed to be $3.54/Mg ($3.20 per
ton) over the life of the tractor.
The Village Creek Wastewater Treatment Plant has
received several benefits from the use of the tractor
140
-------
Figure 8-2. Auger aerator dewatermg machine.
auger equipment. Insect abatement has been an
important benefit, according to both engineering and
operating staff at the plant. The fact that fly eggs and
larva are mixed continually into and below the surface
of the sludge has dramatically cut the fly population
and its impact on the surrounding housing. However,
the staff feels that their single machine is barely
adequate to handle the 81 + ha (200+ ac) of sludge
beds and mix them with sufficient frequency to break
the fly cycle. The importance of this problem will
increase in the near future as additional nearby
residences are constructed and families move in.
Odor control is another important consideration. The
drying beds cover nearly 1.3 km2 (1/2 square mile). A
piped deodorant system has been installed
surrounding the sludge bed area. The system is
operated on a continuous basis, but, if there are
severe odor problems, effective odor masking cannot
be accomplished with a deodorant system. The
tractor auger is used in several ways to control odors.
By mixing and turning the sludge frequently, the
Table 8-3. Costs tor Auger Aerator Tractor Used To
Improve Sludge Drying Bed Operations at
Village Creek Wastewater Treatment Plant
Estimated Cost
Item ($)
Capital Costs
Auger Aerator Tractor 200,000
Annual Cost Equivalent3 52,800
Annual O&M Costs
Operation
Fuelb 14,000
Labor= 24,000
Maintenance01 15,000
Total O&M, $/yr 53,000
Annual Capital Recovery 52,800
Total Annual Cost6 105,800
Cost per dry ton of solidsf 3.20
Total Present Worth of Auger Aerator Systems 401,000
a Using Capital Recovery Factor of 0.2638, based on i = 10%,
5-year life (10,000 hr @ 2,000/yr).
b? gal/hr x $l.00/gal x 2,000 hr/yr.
c Man-hour rate of $i2/hr x 2,000 hr/yr.
d 7.5% of capital cost.
e Annual cost equivalent •*• total annual O&M.
' Total annual cost •»• 33,000 dry tons/yr of sludge.
g Present worth factor = 3,791, based on i = 10%, 5-year life.
drying time is accelerated, thereby cutting the total
acreage of wet sludge at the site, resulting in less
surface area for odor release. Other options are
possible with the machinery on the site. Wet sludge
and dry sludge could be mixed and windrowed for
composting operations. Composting helps keep
sludge aerobic and thereby reduces odor potential.
Reduction of drying time is an important
consideration. Such a reduction increases the
capacity of the sludge beds, and can so delay the
future installation of alternative dewatering processes
or enlargement of the sludge beds. Thus, this
machinery has lengthened the projected life of the
sludge bed drying system for Fort Worth. It has made
the sludge beds more effective in drying sludge, in
that less area is required to dry the existing sludge (or
the sludge drying capability has increased for the
existing area). It has also cut down the nuisance
associated with the sludge beds, which may make
them more acceptable to nearby residents.
The accelerated drying times have permitted the
establishment of a buffer zone between the sludge
beds in use and the new housing. The City has
dedicated eight sludge beds for use as a buffer zone.
This zone creates a sludge-free area about 150 m
(500 ft) wide between sludge beds and nearby
nouses. Prior to the purchase of the auger aerator
machine, all sludge beds were needed during wet
years to dewater the plant's digested sludge output.
Another benefit derived by the purchase and use of
the tractor auger is that it is an active, visible show of
141
-------
Tablo 8-4, Rainfall and Air Temperature Data for Air Drying Case Studies
Mean Annual Rainfall, in/yr
Caso Study Location
Elgin, II.
Boltovilto, IL
Gwinnet Co., 6A
Chicago, IL
Ft Worth. TX
Station Location
Aurora, IL
Belleville, IL
Atlanta, GA
Chicago, IL
Dallas/Ft Worth, TX
Maximum
n/a
n/a
71.4
49.4
50.6
Minimum
n/a
n/a
31.8
21.8
18.6
Mean
3S.61
36.81
48.62
35.23
32. 12
Air Temperature
"C
9.3
12.9
16.42
9.4
10.7
op
48.71
55.3 1
61.52
48.93
8S.62
1 Yearly average for the period 1951-1980.
2Yoaily average for the period 1944-1983.
3Ycai1y average for the period 1958-1983.
n/a - not available.
in/yr x 2.54 - cm/yr.
good faith to nearby residents. Such equipment
serves as evidence that the treatment plant is taking
action to resolve the complaints of neighbors.
An additional benefit that the engineering staff
foresees is that the tractor auger will still be useable if
they convert to a sludge composting operation.
Already, woodchips have been stockpiled in one
section of the drying bed site for use as a bulking
agent in pilot studies with sludge composting. The
auger aerator is directly applicable for composting
operations because of its mixing and windrowing
capabilities.
Another benefit of the auger aerator is that operator
acceptance has been very good. The machine has an
air-conditioned cab with a stereo and is pleasant to
operate. Because of the comfort, the operators are
eager to use the unit. The "user-friendly" cab has
two major benefits. First, the operators take pride in
the condition and appearance of the tractor auger. At
the time of this site visit in 1984 the machine was
being thoroughly cleaned at the end of each day.
Second, since operators enjoy using the machine, it
is more likely to be used as intended. Therefore, it is
more likely that the sludge beds will be maintained in
a way that accelerates the drying time and reduces
insect and odor problems.
It should be noted that the tractor auger also has one
disadvantage. The machine may be creating a long-
term problem since the auger mechanism has a
tendency to scrape and mix sand from the sand
drying beds with the sludge. Since the sand is well
mixed with the sludge, it is removed when the sludge
is trucked from the site. In the long term this removal
could cause a lowering of the sand beds and could
ultimately require the replacement of sand materials
to avoid exposing or damaging the underdrain
system. However, since the Village Creek Plant's
underdrain system is sad to be installed nearly 1.8 m
(6 ft) below the surface, the sand removal is not an
Immediate concern.
8.4 Weather Data
Table 8-4 provides annual average rainfall and air
temperature data for weather station locations near
the locations described in this chapter. This weather
data can be used to see if differences in dewatering
rates between locations are caused by differing
rainfall amounts and air temperatures.
8.5 Reference
1. Technology Evaluation of Brown Bear Tractor for
Sludge Dewatering. J.M. Montgomery Inc., EPA
Contract Report 68-03-1821, U.S.
Environmental Protection Agency, Water
Engineering Research Laboratory, Cincinnati, OH,
1984.
142
-------
Chapters
Case Studies: Mechanical Dewatering Systems
9.1 Introduction
Six case studies on mechanical dewatering processes
are presented in this chapter. They were selected to
illustrate the use of these processes as well as to
relate performance and operation and maintenance
experiences. Where possible, case studies were
selected on the basis of availability of information,
data, and side-by-side comparisons of different
processes. This manual does not aim to endorse any
particular manufacturer's dewatering or conditioning
process. It is recognized that there are, as indicated
in Appendix C, other manufacturers of the same
types of equipment and conditioners. Further some of
the unmentioned manufacturers may have had both
better and more economical experiences than those
illustrated by the case studies. The case studies
include:
» Stamford Water Pollution Control Facility: Stamford,
CT - Parkson Belt Filter Press
• Southeast Water Pollution Control Plant: San
Francisco, CA - Comparison of Sharpies and
Humboldt-Wedag Centrifuges
• Calumet Sewage Treatment Works, Metropolitan
Sanitary District of Greater Chicago (MSDGC):
Calumet, IL - Harima Centrifuge
* West-Southwest Sewage Treatment Works,
Metropolitan Sanitary District of Greater Chicago
(MSDGC): Calumet, IL - Sharpies Centrifuge
• Metropolitan Denver Central Plant: Denver, CO -
Comparison of Various Centrifugal and Belt Press
Dewatering Processes with the Original Vacuum
Filtration Equipment
» Duffin Creek Water Pollution Control Plant:
Toronto, Canada - Jones and Edwards Membrane
Filter Press using Polymer Conditioning
9.2 Case Study: Belt Filter Presses,
Stamford, CT
The City of Stamford's Water Pollution Control Facility
(shown in Figure 9-1) is a 0.88-m3/s (20-mgd)
conventional, activated sludge treatment plant. Its
wastewater is about 15 percent industrial and 85
percent commercial/domestic. The average influent
BOD§ and suspended solids are 160 mg/l and 120
mg/l respectively. Effluent BODs and suspended
solids average 7 mg/l and 20 mg/l respectively.
The plant has primary sedimentation followed by
mechanical aeration. The average MLSS is 2,500
mg/l, SVI is 100 ml/g, and SRT is 6 days. Sludges
from the underflow of the primary and secondary
clarifiers are combined, degritted using
hydrocyclones, and gravity thickened. The typical
underflow concentration from the thickeners is 2.5-
3.0 percent solids.
Stamford is unique in that it has the only operational
co-incineration system for the disposal of sludge
and municipal refuse in the United States (see Figure
9-2). Using progressive cavity pumps, the sludge is
pumped to flocculation tanks. Calgon WT-2136
polymer, which is used to condition the sludge, is
added approximately 7,6 m (25 ft) upstream of these
tanks. The conditioned sludge is dewatered and the
cake is then discharged to the pug mill. There the
dewatered sludge is combined with previously dried
sludge to produce a mixture which is approximately
60-65 percent solids. This mixture is discharged
from the pug mill and conveyed to the rotary dryer. A
portion of the hot gas which would normally be
wasted through the incinerator stack enters the dryer
at the same area as the sludge mixture. As the dryer
rotates (5 rpm), the sludge is cascaded through these
hot gases, thus evaporating the moisture in the
sludge. The dried sludge (90 percent solids) is
discharged through a diverter gate and divided into
two streams - one that goes back to the system as
dry recycle and the other that goes to the incinerator
and is burned. The heat value of the sludge is about
12,800 kJ/kg (5,500 Btu/lb).
When the plant was built, centrifuges were installed
for dewatering. However, they were unable to achieve
the requisite 22 percent solids needed to allow
optimization of the sludge drying and incineration
system. Also, they were extremely high maintenance
items and the solids capture with polymer addition
was less than 60 percent. In 1979, these units were
143
-------
Figure 9*1. Schematic of sewage treatment plant, Stamford, CT.
SCREENING
Influent
(Raw
Sowaga)
Bar
COMMINUTION
PRIMARY
SETTLING
AERATION ZONE
SECONDARY
SETTLING
DISINFECTION
Aeration Basins
ScreenI
A
Screening to
Incinerator Pit
Grit to
Landfill
THICKENING
Primary Sludge
Waste
Activated Sludge
Return Activated
Sludge
Activated Sludge
DEWATERING
DRYING
INCINERATION
Thickened
Sludge
Belt
Pre
/
Filter
sses
Rotary
Dryer
^ Waste Heat
Lime
™ Addition
Municipal
Incinerator
I
» Landfill
East Branch
Stamford Harbor
Alternate Disposal System
When Incinerator Is Down
Figure 9-2. Sludge flow path through co-incineration
system.
Feed
SolWs
2.5-3%
60-65%
Belt Presses
Solids
— >
\
Uva-
bottom Bir
Pug Mill
Mixer
" Solids
Rotary Dryer
i
Incinerator
Solu
replaced with two 2-meter Parkson MP-80 belt
filter presses. These units were equipped with an
additional high pressure section to ensure the driest
possible sludge cake. The City specified a minimum
cake solids concentration of 24 percent with a 95
percent capture.
During the first two months of operation, cake solids
concentrations of less than 20 percent were produced
but the solids capture was an excellent 98 percent. In
the third month of operation, the cake solids
concentration began to increase and rose by the end
of the month to 25-26 percent. Solids capture
remained at 98 percent. It was theorized that,
because the centrifuges had such an extremely poor
solids capture, fine solids were continuously
circulating through the plant. With the high solids
capture efficiency of the belt filter presses, these
fines were gradually being removed. Once that
occurred, the sludge dewatered more easily and a
drier cake was produced.
Initially, belt or screen life was poor (less than 500
hours). The short screen life was attributed to
operator inexperience and a sludge that was
somewhat difficult to dewater during that period.
Experimentation with different mesh screens led to
the selection of the Scandiafelt FE-3366 screen.
This screen has a high air permeability and allows for
good drainage of water from the sludge. This feature
prevented uneven buildup of the sludge prior to the
extra-high pressure section, thus preventing
creasing and wearing of the screen cloth.
Typical maintenance problems for these units were
failure of ballbearings and occasionally broken roller
shafts. Bearing life increased substantially when the
press operators were given the responsibility of
greasing the bearings.
In 1985, the MP-80 units were traded in on the
improved Parkson Series 3000 presses. They have
been operated for 24 hours per day on an av/erage of
6 days per week for over 2 years (15,000 hours) with
no mechanical failures at all (bearings, shafts, drive
units). Screen life has been outstanding with upper
screen lives of as much as 10,000 hours and lower
144
-------
screen lives of about 5,000 hours. At one time, a
series of lower screens failed at the clipper seams
after only 200-300 hours of operation. However, this
failure was due to a problem with the belt
manufacture and not with either the presses or their
operation.
Currently, with a raw feed sludge solids of 2.5-3.0
percent (50:50, P:WAS), a cake solid of 27 percent is
achieved with 98 percent capture. Approximately 40
kg of Calgon WT-2136 cationic polymer per dry Mg
(80lb/dry ton) of sludge is used at a cost of
$l3.22/dry Mg ($12.00/dry ton). In addition,
potassium permanganate is added to the feed sludge
at the rate of 0.5 kg/Mg (1 Ib/dry ton) for odor control.
This adds about $l.10/dry Mg ($1.00/dry ton) of
sludge, but the cost is justified by giving the operators
a better working environment.
9.3 Case Study: Centrifuges, San
Francisco, CA
9.3.7 Sludge Characteristics and Processing
The San Francisco Southeast Water Pollution Control
Plant (WPCP) is a 3.5-m3/s (81-mgd) pure oxygen
activated sludge wastewater treatment plant. Pertinent
design features include:
Primary clarinets:
7 @ 12 m x 64 m x 3.7 m SWD
(38ft x 210 ft x 12 ft SWD)
4 @ 11 m x 78 m x 3.4 m SWD
(37 ft x 256 ft x 11 ft SWD)
Aeration basins:
8 @ 1.8-13 m x 13 m x 4.3 m SWD
(6-42.5 ft x 42.5 ft x 14 ft SWD)
Secondary clarifiers:
16 @ 36.6 m dia. x 4.6 m SWD
(120 ft dia. x 15 ft SWD)
The 1986 operational characteristics of the plant are
as follows:
Primary Secondary
Influent Effluent Effluent
Flow, mgd
BOD5, mg/I
TSS, mg/I
TP, mg/I
NH4N, mg/I
Temp.
81
201
246
21 °C
127
106
11
18
3
21
The operational characteristics of the activated sludge
include a HRT of 1.8 hours and an estimated 1.9 day
mean cell residence time (MCRT) and 0.7 day1
F/M. The pure oxygen activated sludge system is
operating in a non-nitrifying mode of operation. The
aeration system consists of six stages of complete
mix reactors in each of the eight trains. A schematic
for the San Francisco Southeast WPCP is shown in
Figure 9-3.
The excess primary and biological sludge produced
from the plant is estimated to be as follows:
PS - 43,090 kg/d (95,000 Ib/d) 60%
WAS - 29,485 kg/d (65,000 Ib/d) 40%
TOTAL- 72,575 kg/d (160,000 ib/d) 100%
The primary clarification efficiency is 55 percent
removal of TSS and 35 percent removal of BODg.
The high efficiency of primary treatment results in a
60:40 ratio of PS:WAS in the sludge mixture. The
sludge yield is about 0.70 Ib EAS/lb BODs removed
based on 5 mg/I SBODs in the final effluent. The SVI
averages about 150 ml/g and normally ranges from
100 to 200 ml/g. Sludge is anaerobically digested
prior to centrifugation.
9.3.2 Centrifuges
Two Sharpies PM75000 centrifuges were installed in
1982 and two more were added in 1985. These
countercurrent flow centrifuges are driven by 112-
kW (150-hp) motors and are equipped with manually
controlled eddy current backdrives. The PM75000
has a 0.7-m (29-in) bowl diameter and a 2.3-m
(92-in) bowl length and operates at 2,300 rpm and
2,500 rpm or a peripheral force of 2,180 to 2,500 g's.
The differential speed is normally in the range of
2.5-4.0 rpm. The bowl and scroll are fabricated from
ACI CFB SS and the housing is 316 SS with a cast
iron base. The scroll tips and other wear points are
protected by replaceable tungsten carbide tiles.
Sludge is fed to the centrifuges by 0-15.8 l/s (0-
250 gpm) variable flow progressive cavity pumps. The
feed rate is normally 9.46 l/s (150 gpm) and is
adjusted to account for varying sludge concentrations
and characteristics. Polymer is added to the
centrifuge in the bowl just downstream of the feed
port. Cake is removed from the centrifuge discharge
hopper with progressive cavity sludge cake pumps
and a cake conveyor.
The Humboldt-Wedag centrifuge installation consists
of two Model S4-1 concurrent flow units which
started up in December 1982. The S4-1 centrifuges
have a bowl diameter of 0.9 m (36 in) and a bowl
length of 2.4 m (96 in). The bowl speed is 1,400 rpm
(1,000 g's) and the units are equipped with automatic
torque controlled hydraulic backdrives. The differential
speed is normally in the range of 2 to 4 rpm. The
bowl and scroll are fabricated from carbon steel. The
housing and base are made of carbon steel. The
scroll and other wear points have Udalite and/or
ceramic hardfacing.
Sludge feed to the centrifuges is controlled by two
0-15.8 l/s (0-250 gpm) sludge pumps. Normal feed
145
-------
Figure 9-3. Schematic of the Southeast WPCP, San Francisco, CA.
Cake Storage (2)
TUF Pumps (4)
,. Oxygon
I fl H fl ' r Saturation
! U U U f Building
rcr^—
f... 1
r 2
f. 3
S 4
«,... 5
* 6
f, 7
| z.
780 -DAF
8
7
6
TUP.
-Oxygen
Aerators
PPR «• Pumped Plant
Recycle
RAS - Return
Activated
Sludge
TAS « Thickened
Activated
Sludge
PS-
-PPR
PE Pump Station
IvflJUU
\S_fi_fi_y
PS Pumps (41 —
rt
H J
/
Parshal! Flymas
LEGEND
TUF = Thickener
Underflow
WAS = Waste
Activated
Sludge
OS = Digested Sludge
ML = Mixed Liquor
PE = Primary Effluent
PI = Primary Influent
HS = Heavy Solids
SM/DEW = Scum/
Dewatering
PS = Primary Sludge
SB = Secondary
Bypass
SE = Secondary
Effluent
SI = Secondary
Influent
rate is 9.46 l/s (150 gpm). Polymer is added to the
feed zone area of the centrifuge. Centrifuge cake is
transferred from the centrifuge hopper with
progressive cavity pumps. Ferric chloride is added to
the sludge upstream of all centrifuges for struvite
(magnesium ammonium phosphate scale) control.
The centrifuges were evaluated by City personnel and
technicians and engineers from the centrifuge
suppliers. Centrifuge setup was optimized by the
suppliers prior to the test conducted in mid-1984.
The performance test results are presented in Tables
9-1 and 9-2. The tests employed Centrifuge No. 1
(Sharpies) and Centrifuges 2 & 3 (Humboldt-Wedag)
for tests conducted on the same feed of PS and
WAS. The ratio of PS:WAS was about 1:1 at the time
of the tests. (Note that these tests are not necessarily
representative of current performance and that one of
the requirements of the test was a minimum 95
percent recovery).
In Figure 9-4, the respective cake solids as a
function of the scroll shaft torque is presented. As
indicated, the driest cake is produced at the highest
torque, which generally represents the maximum
residence time of solids in the bowl. This conclusion
is somewhat modified by other factors such as feed
rate, solids discharge rate, solids recovery, and
polymer dosage.
Both types of centrifuge demonstrated a decline in
cake solids with increasing feed rate. Since the feed
TSS was relatively constant, the probable effect was
increased solids discharge rate as a function of feed
rate. Both effects on cake solids are shown in Figure
9-5. In this case, if the design objective was to
maximize the cake solids, it would be necessary to
operate the centrifuge at less than the maximum
capacity, possibly to 50-70 percent of maximum
capacity.
146
-------
Table 9-1. 1984 Centrifuge Performance Test, Sharpies Centrifuge No. 1 - San Francisco Southeast WPCP*
6/4
Feed Flow, gpm
Feed, %TS
Feed, %VS
Polymer Cone., %TS
Polymer feed, gpm
Polymer dose, Ib/ton
Cake %TS
Centrate TSS, mg/l
Recovery, %
Energy, kWh
Energy, kWh/gpm
Torque, kg/m2
150
1.9
67.3
0.30
3.8
7.9
18.4
540
97,5
60.8
0.41
333
150
1.9
67.7
0.39
4.6
12.4
19.7
533
97.5
58.1
0.39
369
148
1.9
67.9
0.32
6.4
14.4
22.2
860
95,8
-
486
6/5
194
1.9
69,4
0.34
5.0
8.9
16.5
1,100 1
95.0
68.9
0.36
360
199
1.9
69.7
0.34
6.5
11.4
18.9
,270
94.1
72.6
0.36
440
6/6
250
1.9
67.7
0.22
8.0
7.4
13.9
1,060
95.2
48,2
0.3
312
6/7
125
1.8
68.4
0.21
2.4
4.5
14.4
1,570
92.3
52.4
0.42
280
126
1.9
68.0
0.21
4.0
7.0
17.2
1,120
94.7
53,8
0.43
330
126
1.8
70.0
0.17
5.0
7.5
19.5
920
95.3
54.1
0.43
350
6/8
173
1.8
70.4
0.23
6.0
8.8
19.2
660 1
96.7
67.0
0.39
390
173
1.8
68.6
0.24
7.5
11.3
21.7
,140
94.3
68.4
0.4
450
" Bowl speed: 2,300 rpm
Polymer: Percol 757
Pond depth: not reported.
Table 9-2. 1984 Centrifuge Performance Test, Humboldt Centrifuges Nos. 2 and 3 - San Francisco Southeast WPCP*
Feed Flow, gpm
Feed, %TS
Feed, %VS
Polymer Cone., %TS
Polymer feed, gpm
Polymer dose, Ib/ton
Cake %TS
Centrate TSS, mg/l
Recovery, %
Energy, kWh
Energy, kWWgpm
Torque, ib-in
6/5
188
1.9
69.4
0.34
5.5
10.1
17.8
2,620
87.8
41.3
0.22
50
»2
193
1.9
69.7
0.34
6.3
11.4
17.3
2,650
87.6
41.9
0.22
60
6/6 #2
247
1.9
67.7
0,22
8.0
7.5
14.7
2,220
89.7
48.2
0.19
30
6/7 #3
125
1.8
68.4
0.21
2.4
4.5
11.9
1,850
91.1
31.9
0,26
20
126
1.9
68,0
0.21
4.0
7.0
19.0
1,320
93.7
32.2
0.26
50
125
1.8
70.0
0.17
5.0
7.6
22.8
980
95,0
32,1
0,26
80
6/8
171
1.8
70.4
0,23
6.0
8.9
21.7
580
97.1
37.0
0,22
80
#3 '
172
1.8
68.6
0.24
7.5
11.4
20.9
1,700
91.5
37.4
0.22
90
" Bowl speed: 1,400 rpm
Polymer: Percoi 757
Pond depth: 520 mm.
Both types of centrifuge exhibited sensitivity of cake
solids concentration to the polymer dosage. White the
two lowest cake solids results were due to
underdosing, there was a significant improvement in
cake solids with an increased dosage. As shown in
Figure 9-6, the increased polymer dosage of 50
percent beyond the amount necessary to achieve
85-90 percent recovery can result in a 3-5
percentage points increase in cake soiids. This may
be only true for a specific polymer, in this case Percol
757. The lower cake solids again generally resulted
from the highest solids rate.
At a polymer dosage of 14-18 kg/dry Mg (7-9 Ib/dry
ton) of solids, the Humboldt centrifuge produced 1-3
percentage points drier cake solids at equivalent feed
rates. While the Sharpies cake concentration
increased with dosages up to 22-24 kg/Mg (11-12
Ib/ton), the Humboldt cake became slightly wetter
with dosages exceeding 18 kg/Mg (9 Ib/ton). For
these tests Humboldt modified the polymer injection
with a nozzle that created a 5.6 kg/cm2 (80 psi) pump
discharge pressure. San Francisco no longer runs the
system in this manner.
In general the Sharpies unit had a slightly better
recovery than the Humboldt centrifuge at the
equivalent polymer dosage. The lowest recovery
measured was 87.6 percent for the Humboldt
centrifuge and 92.3 percent for the Sharpies unit.
Average recovery was 91.7 percent and 95.3 percent
for Humboldt and Sharpies, respectively. While the
147
-------
Flfluro 9-4. Cake solids vs. torque.
Figure 9-5. Cake solids vs. feed rate.
25
20
#
in
a
Q 10
6
0
'
26
20
#
i is
d 10
S
0
25
A A
/ * * * * , »
. A ^ i" is
A A o
J
a 10
A Sharpies
i i i 1 I 0
"
A A
" * A /
'A A
A Sharpies
Feed Rate = 900 Ib TS/hr per 100 gal/min
i i i i
» 300 350 400 450 BOO ^fV^H, 16° I^TS/hr "°
Torque, Ib/in Feed Rg{e> ga|/min
25
8 •
* * ^
m 2 15
1
3 10
• Humboldt
5
l 1 l i J
*' »
* .
*
• Humboldt
111!
20 40 60 80 100 120 160 200 240 280
Torque, .Bar Feed Rate, gal/min
148
-------
Figure 9-6. Effect of polymer dose on cake solids.
100
$
95
90
85
80
* Humboldt
A Sharpies
8 10 12 14
Polymer Dose, Ib/ton dry solids
16
recovery data for Sharpies indicated that the polymer
dosage could be reduced, the reduction in recovery
would be at the expense of a wetter cake as shown in
Table 9-1. The recovery versus polymer dosage
data is presented in Figure 9-7.
Figure 9-7. Effect of polymer dose on solids recovery.
tft
1
1
o
25
20
15
10
5
n
_
• A
J* ? A
* ^ A « •
~* A«
*
-
• Humboldt
A Sharpies
l i i i l I
8 10 12 14 16
Polymer Dose, Ib/ton dry solids
9.3.3 Operation and Maintenance
The operating hours on the centrifuges as of
September 1986 are as follows:
Centrifuge
No. Manufacturer
1
2
3
4
5
6
Sharpies
Humboldt
Humboldt
Sharpies
Sharpies
Sharpies
Date
Installed
1982
1982
1982
1982
1985
'1985
Operating
Hours
13,368
17,415
20,71 1
14,903
3,912
1,638
The plant reports that there has been no appreciable
wear on the scrolls of either machine. Most of ihe
wear has occurred in the area of feed inlet and
sludge cake discharge. A detailed breakdown of costs
is not available. The total amount spent up to
September 1986 is about $450,000 for labor and
materials - $300,000 for Sharpies and $150,000 for
Humboldt. However, the $300,000 for Sharpies
includes the cost to refurbish machines purchased
from New York City. The gross maintenance cost of
the centrifuges (including refurbishing of New York
City machines) is $6.25/hr of operation. At 11.0 l/s
(175 gpm) and 2 percent TSS, the cost would be
$7.86/Mg dry solids ($7.13/ton).
The operating costs for power for the centrifuge are
as follows:
Flow
Sharpies
Humboldt
(gpm)
125
175
250
kWh/gpm
0.43
0.40
0.31
$/hr
3.76
4.90
5.43
$/ton
6.68
6.22
4.82
kWh/gpm
0.26
0.26
0.20
$/hr
2.28
2.70
3.50
$/lon
4.05
4.33
3.11
Based on $0,075/kWh, sludge @ 2% TSS, and 85% recovery.
The Sharpies recovery is estimated to be 90% and the Humboldt
recovery is estimated to be 80%.
The chemical dosage and costs for the centrifuge
operation are as follows:
Polymer (Perec/ 757 - Allied Colloids):
8.9 Ib/ton @ $2.25/lb = $20-03/ton
60 Ib/ton @ $0.12/lb = $6.00/ton
Total = $26.03/ton ($28.69/Mg)
There are two operators per shift responsible for the
centrifuge operation. Their duties include operation
and adjustment of the centrifuges and auxiliary
systems, batch polymer makeup, monitoring the
sludge hopper level, and operating the cake
conveyors, as well as some housekeeping.
The operating staff's comments regarding their
experiences with the two types of centrifuges, high
speed countercurrent and low speed concurrent, are
summarized in Table 9-3.
Currently, the percent excitation (Sharpies) needs to
be monitored as it relates to the inventory of cake in
the bowl and requires a manual adjustment of the
backdrive. However, San Francisco intends to start
up a programmable controller supplied by Sharpies to
monitor the percent excitation and adjust the
backdrive automatically.
149
-------
Tabfo 9-3. Operator Comments on Humboldt and Sharpies Centrifuges - San Francisco Southeast WPCP
ConlnkiQO Type Advantages Disadvantages
Humbddl
Shafptos
(automatic PC backdrive
control is being installed)
• Lower initial cost
• Less power required
* Simpler design, easy to operate, fewer hours for
operators
• Automatic torque control backdrive accomodates
varying feed solids
• Larger tolerances
» More operating hours between servicing
• Wear items can be replaced on site
• Will use less polymer when operating properly
• Service oriented, local representatives
• More gauges and monitoring devices available
• Recovery is not as high with identical polymer
dosage
• Hydraulic system subject to hydraulic fluid leakage
• More sensitive to operational upsets, e.g., plugged
cake pumps backing into.machine
• Needs more operator attention
At San Francisco, the higher speed machine has
shown greater wear due to the speed and softer
stainless steel, and also due to San Francisco's fine
grit (100-200 mesh) in the sludge slurry. However,
this problem is partly offset by the higher speed
machine's ability to capture more of a high volatile
sludge Introduced by the activated sludge.
9.4 Case Study, Centrifuges, Calumet, IL,
Calumet STW
The Metropolitan Sanitary District of Greater Chicago
(MSDGC) operates seven sewage treatment works,
of which four have complete solids processing
systems. The two largest of these are the Calumet
Sewage Treatment Works (STW) and the West-
Southwest STW. The Northside plant purges sludge
to the West-Southwest plant. The dewatering
oporations at the two largest of the MSDGC's sludge
treatment works are described below.
9.4. f Sludge Characteristics and Processing
The Calumet STW is a 9.6-m3/s (220-mgd)
single-stage conventional activated sludge plant.
Pertinent design features include
Primary clarifiers:
32, each @ 1,020 m3 (270,000 gal)
total surface area - 9,510 m2 (102,400 sq ft)
Aeration basins:
31, total volume - 192,800 m3 (51.1 Mgal)
Final clariliers:
32, total volume - 179,600 m3 (47.6 Mgal)
total surface area - 41,090 m2 (442,300 sq ft)
The operational characteristics of the plant, based on
1985 yearly averages, are as follows:
Flow, mgd
BODs, mg/l
TSS, mg/l
NH4N, mg/l
Temp., °C
Influent
235
151
303
14.7
14
Primary Secondary
Effluent Effluent
131
215
19
21
12.4
Based on 1985 figures, the plant averages an SRT of
8.6 days and an F/M ratio of 0.14 Ib i3OD«j/lb MLSS/d.
Since 1982, the plant has installed additional primary
clarifiers, aerated grit chambers, new fine screens,
and a new secondary system.
The plant anaerobically digests the primary and waste
activated sludge in the following percentages:
PS - 51,710 kg/d (114,000 Ib/d) 33.7%
WAS- 101,600 kg/d (224,000 Ib/d) 66.3%
TOTAL- 153,310 kg/d (338,000 ib/d) 100.0%
The primary clarification efficiency is 13 percent
removal of BOD5 and 29 percent of the TSS. The
SVI averages approximately 110 ml/g.
9.4.2 Centrifuges
A total of five Ishikawajima-Harima Heavy Industries
(IHI) centrifuges, model number HS-805M, were
installed in 1981 at the Calumet plant. The units are
driven by 112-kW (150-hp) motors and are
equipped with manually-controlled backdrives. The
bowl speed is 1,560 rpm (1,750 g's) and the normal
differential speed is 12 rpm. The inside bowl
dimensions measure 800 mm (31.5 in) in diameter
and 2,650 mm (104 in) in length.
The units have a rated capacity of 33.2 I/s (525 gpm).
The digested sludge is fed to the centrifuges by three
7.5-kW (10-hp) feed sludge pumps at a rate of
11.4 I/s (180 gpm) and an average feed concentration
of 2.7 percent of total solids. Polymer is injected into
the sludge at a rate of 0.2 I/s (2 gpm) just before the
feed enters the centrifuge. The average polymer
dosage is 5 kg dry solids/Mg (9.9 Ib/ton). However,
the plant notes that the centrifuges require a
significantly higher polymer dosage when the
concentration of the feed sludge drops below 2.4
percent of total solids.
At a 2.7 percent feed concentration and a polymer
dosage of 5 kg dry solids/Mg (9.9 Ib/ton), the units
produced the following results:
150
-------
Cake Solids, percent Solids Recovery, percent
Summer
Winter
min.
14.5
15,2
max.
16.7
16.2
avg.
15.6
15.7
min.
91.0
.
max.
95.4
.
avg.
93.8
-
September 1986 operating data for each of the five
centrifuges are presented in Table 9-4.
Cake is transferred from the centrifuge discharge
hopper by conveyor belts and removed from the plant
by trucks. Baffles were added to the chutes on the
centrifuge discharge hopper to prevent the cake from
splashing off the conveyor belts. The trucks have also
been modified to reduce leakage of wet cake, but
trucks with tailgates prove less efficient. Pumping or
screw conveyor systems are currently being
considered for centrifuge cake transfer. Pumping
appears to be the cleanest operation and several
systems were successfully tested in 1986. No
decision regarding the installation of a pump system
has been reached at this time. The plant also installed
secondary belt scrapers at the discharge end of each
conveyor to improve the capture of solids. Previously,
the solids ended up in the building drain system.
9.4.3 Operation and Maintenance
Each of the five centrifuges at the Calumet complex
are operated between 3,500 and 4,200 hours per
year. The IHI scrolls had to be retrofitted with sintered
tungsten carbide tiles to extend the life to the
10,000-12,000 hr range. Currently, the scrolls on at
least one unit are repaired each year at a cost of
$75,000 for sintered carbide tiles. While the plant
expects future repair costs of the scrolls to decrease,
an additional $15,000 is generally estimated for in-
house support costs. Over 60 percent of these costs
were associated with retrofitting the tiles.
At least one set of bearings is replaced each year,
normally after 10,000 hours of operation or whenever
the scroll is removed for maintenance. The cost of
five main bearing sets is $37,500 and in-house
support costs can be as high as $8,000.
The 1985 sludge dewatering costs were estimated to
be as follows:
Sludge Dewaterinq
Polymer (Percol 763):
@ 9.9 Ib/ton dry solids =
$/ton Dry Solids
15.84
Power @ 123 kWh/ton dry solids
+ Water @ 3,370 gpm/hr + Maintenance = 35.00
Labor =
Total =
22.00
72.84
The total cost, including transport and placing of the
cake in the landfill, ranges from $83 to $110/Mg dry
solids ($75-100/ton).
The Calumet plant has made several modifications to
improve the operation and maintenance of the
dewatering complex. Some of these changes include:
Drainage System: Since it was not feasible to replace
internal building drains, which later proved to be too
small, the clean water sump pumps were replaced
with solids-handling trash pumps. Also, a new line
replaced the main drain pipeline, which recycled
drainage to the treatment plant, to prevent further
settling and clogging. Areas in and around the
centrifuge complex were paved to prevent clogging of
street drains. The pavement also makes it easier to
clean up sludge spills.
Polymer Transfer and Mixing Systems: To reduce
losses and allow quicker transfer of material, the 5-
cm (2-in) liquid polymer transfer piping was replaced
with 15-cn (6-in) piping. The "star feeder" for
metering the dry polymer into the transfer system
frequently jammed and was modified. The plant also
modified the polymer wetting and mixing system to
ensure proper initial wetting of the material. Sufficient
water quantity and pressure, coupled with adequate
clearance at all points, prevent partially wetted
polymer from gumming up the small openings into the
feed system. The plant also converted the polymer
mixing liquid from treatment plant effluent to city
water. This change resulted in reduced polymer
usage and less frequent maintenance of the sight
glasses.
Digested Sludge Feed System: The feed surge
channel was raised 1.2 m (4 ft) and, as a result, the
enlarged channel now provides better suction to the
feed pumps. The plant also installed larger pumps,
equipped with shear impellers to minimize clogging,
to handle the maximum sludge feed rate.
Several other areas still require additional
modifications to optimize the operations at the
Calumet plant. Two of these concerns are noted
below.
Struvite Formation: Although the conversion from
plant effluent to city water significantly reduced the
rate of struvite formation, this crystalline product
(magnesium ammonium phosphate) continues to
cause problems at the centrifuge complex. The plant
has made other modifications in its operations, but
their effect on reducing the rate of struvite formation
has been negligible. These changes included:
• Using a polymer in solution with a pH < 7.0.
» Injecting a chelating agent into the feed sludge.
» Flushing water (for dilution purposes) in the
centrate lines.
151
-------
Tablo 9-4. Calumet STW; Sample of Operating Data (September 1986)
Centrifuge Average Sludge Feed Polymer Dosage
No. 1
No, 2
No. 3
No. 4
No. 5
mgd
0.255
0.260
0,220
0.230
0.256
percent solids
2.54
2.54
2.54
2.54
2.54
Ib/dry ton
10.2
8.4
9.4
12.3
9.4
Solids Recovery
percent
95.2
95.9
94.3
92.9
95.6
Cake Solids
percent
14.9
16.1
14.6
16.4
15.1
Planl personnel remarked that the addition of anti-
sealants to the polymer proved unsuccessful because
the anti-sealants deactivated the basic polymer, to
the point where it was ineffective as a dewatering aid.
The ultimate solution appears to entail the removal of
soluble phosphorus by iron (or other heavy metal)
precipitation. This additional step would decrease the
pH of the centrate, which, in turn, would reduce the
rate of struvite formation.
Hydrogen Sulfide Gas Formation: The Calumet
centrifuge complex experienced two major hydrogen
sulfide gas incidents in 1986. This toxic gas was
stripped from the sludge during the centrifugation
process in excessive concentrations. Although the
ultimate solution lies within the treatment plant (or at
the generation point of the dissolved sulfides into the
sewer system), the control of the gas at both the
digester and centrifuge complex is very important.
Preliminary testing indicated that the levels of
hydrogen sulfide gas can best be controlled by the
addition of zinc chloride or ferric chloride into the
sludge stream at either the digester or centrifuge
complex. However, some have suggested that power
venting of the centrate and cake zones would be the
best solution.
Once the majority of the deficiencies were corrected,
the Calumet centrifuges have operated fairly reliably.
They are capable of producing a good sludge cake
and an excellent centrate with reasonable polymer
requirements, provided that the feed sludge solids
concentration remains greater than 2.5 percent. The
centrifuges themselves have proven to be good
machines.
However, the problem at the Calumet complex is that
overall cost of centrifugation is high. Contributing to
this high cost are machine malfunctions, the large
number of electrically- and mechanically-driven
support systems, and high seasonal labor rates.
9.5 Case Study, Centrifuges, Calumet, IL,
West-Southwest STP
9.5,1 Sludge Characteristics and Processing
The West-Southwest Sewage Treatment Works
(STP) in Calumet, IL is a 52.6-m3/s (1,200-mgd)
single-stage activated sludge plant. Peak flow
averages 63.1 m3/d (1,440 mgd). Pertinent design
features include:
Primary clarifiers:
29 @ 30.5 m x 308 m x 3.4 m SWD
(100ft x 1,011 ft x 11 ft)
108 Imhoff tanks @ 24 m x 24 m x 10-11 m SWD
(80 ft x 80 ft x 33-36 ft)
Aeration basins:
32 @ 132 m x 10 m x 4.6 m SWD
(434 ft x 34 ft x 15 ft) each pass (4 passes/tank)
Secondary clarifiers:
96 @ 38.4 m dia. x 4.3 m
(126 ft dia. x 14 ft) SWD
The operational characteristics of the plant, based on
1986 data, are as follows:
Influent
Primary
Effluent
Sec.
Effluent
West Southwest West Southwest
Side Side Side Side
Flow, mgd
BOD5< rng/l
TSS, mg/l
NH4N, mg/I
Temp. °C
327
153
362
6.2
486
273
708
10.1
-
-
105
222
8.0
-
-
401
234
11.0
-
788
9.3
12
3.5
16
Based on 1986 figures, the plant averages an SRT of
5.6 days and an F/M ratio of 0.272 Ib BOD5/lb
MLSS/d. The source, quantity and disposition of the
raw sludges produced by the West-Southwest STP
are as follows:
Type Raw Sludge Disposition Location
dry tons/d
PS
WAS
635
165
500
135
Anaerobic
Digestion
Anaerobic
Digestion
Anaerobic
Digestion
Recycle
West Side Imhoff Tanks
(Unheated)
Wast Southwest Imhoff
Tanks {Unheated)
West Southwest
Digesters (Healed)
Return to Plant
152
-------
The primary clarification efficiency is 66.9 percent and
38.7 percent respectively for the Southwest and West
sides of the plant. The SVI averages 67 ml/g.
The estimated sludge mixture for dewatering is as
follows:
PS - 32,66 kg/d (72,000 Ib/d) 21%
WAS - 123,380 kg/d (272,000 ib/d) 79%
TOTAL - 156,040 kg/d (334,000 Ib/d) 100%
9.5.2 Centrifuges
Eleven Pennwalt Sharpies Super D-Canter PC
81000 centrifuges were first used in 1981 and an
additional unit was installed in 1984. They are
countercurrent flow centrifuges driven by 93.4 kW
(125 hp) motors. The knits are equipped with
changeable backdrive pulleys which allow backdrive
speed differentials of 22, 16, 9, and 6 rpm. The
backdrives have 11-kW (15-hp) motors, and the
gearbox ratio is 98:1. One DC variable speed
backdrive has been purchased and will be installed in
1987. The PC 81000 has a 0.6 m (25 in) by 2.92 m
(115 in) bowl which operates at 2,160 rpm.
The bowl and scroll are fabricated from stainless
steel. The base is cast iron and the cover is plexiglas.
The scroll flighting (feed zone area) is protected with
replaceable tungsten carbide tiles. The liquid area
scroll flighting is protected with Stellite hard surfacing.
Other areas susceptible to accelerated wear are
protected by either replaceable tungsten carbide
inserts or hard surfacing.
Feed sludge is provided by three constant feed
centrifugal pumps. These pumps feed the distribution
system with individually controlled feed valves at each
centrifuge. These valves allow infinite flow
adjustment. The feed rate is normally in the range of
9.5-11.4 l/s (150-180 gpm). Polymer is added to
the centrifuge feed inside the feed zone. Each
centrifuge has a 0-0.6 l/s (0-10 gpm) progressive
cavity polymer pump. However, the capacity of these
pumps is being increased to 0-1.3 l/s (0-20 gpm).
At a 3.51 percent TSS feed concentration and a
polymer dosage of 115 kg/Mg (230 Ib/ton), the units
produced the following 1986 results:
Cake Solids, percent Solids Recovery, percent
Summer
Winter
mm, max,
11.9 17.9
10.9 14.0
avg,
15.4
12.6
mm.
70.6
60.5
max. avg.
94.6 88.4
96.5 84.3
Cake is removed from the discharge hopper by a
system of conveyor belts and is loaded on railroad
dump cars for disposal.
9.5.3 Operation and Maintenance
The operating hours of the centrifuges as of
September 1986 are as follows:
Unit
No.
1
2
3
4
5
6
7
8
9
10
11
12
Operating
Hours
32,000
31,800
32,200
29,400
30,900
29,000
30,200
30,600
30,300
31,200
30,300
15,200
In general, the scrolls are repaired after 30,000 hours
of operation at a cost of $12,000-$17,000. In 1986,
four units had the scrolls repaired. As of 1986, the
West-Southwest units have not had the bearings
replaced. The conveyor bearings were replaced after
10,000-15,000 hours at a cost of $300.
The 1986 maintenance costs, shown below, reflect an
average feed rate of 9.46-11.8 l/s (150-189 gpm)
@ 3.51 percent TSS and 88.7 percent recovery.
Machines
Only
Support
Equipment
Total
$/hr
$/dry ton
4.86
3.81
3.21
2.46
8.17
6.27
Operating costs, based on $0.053/kWh, sludge feed
concentration of 3.51 percent TSS, and 88.7 percent
recovery, were as follows:
Flow, gpm
kwh/gpm
$/hr
$/ton DS
180
0.412
3.50
2.80
The 1986 chemical dosage and cost is shown below:
Polymer Ib/dry ion $/lb $/lon DS
2540 C
230
0.0528
12.14
153
-------
The 1986 sludge dewatering costs were estimated to
be:
Sludpo Dowaloring
Cost/Unit
$/ton DS
Polymer @ 230 to/ton DS $0.053/lb 12.14
Powor
@ 52.8 kWh/ton DS (centrifuge)
@ 23.0 kWh/ton DS (support eq)
@ 75.0 kWh/lon DS (total) $0.053/kWh 4.02
Water
@ 0.22 gph/hr (centrifuge)
@ 0.10 gph/hr (support eq)
@ 0.32 gph/hr (total) $0.05/1.000 gal 0.04
Labor
@ 0.084 MH/lir (centrifuge)
@ 0.092 MH/hr (support eq)
@ 0.176 MH/hr (total) S16.45/MH 2.04
Maintonanco 6.27
Total Dowaloring Cost 24.51
During May of 1986, the MSDGC tested a pilot screw
press (manufactured by Hoilim Iron Works Company,
Ltd.) on anaerobically digested sludge at the West-
Southwest STP. The test results and screeen press
are described in detail in Section 7.6 of this Manual.
Recently MSDGC has conducted dewatering studies
on a new type of centrifuge, the Humboldt
CentriPress CP2-1. The tests produced the following
average results:
Cake-
Recovery -
Polymer Cost -
29.4%
92.7%
$12.15/ton
The complete test data for the Chicago tests are
included in Section 7.6 of this manual.
9.6 Case Study, Centrifuges, Denver, CO
The following report was excerpted from a 1986
paper entitled "Digested Sludge Thickening and
Dewatering at the Metropolitan Denver Central Plant"
(1). In 1981, a study conducted for the Metropolitan
Denver Sewage Disposal District No. 1 (MDSDD No.
1) recommended a solids handling system that
included centrifugal dewatering and on-site
composting of the digested sludge, or alternatively,
thickening the waste sludge for agricultural reuse. A
solid-bowl centrifuge was installed in 1982 and
continuously operated reliably, leading the District to
purchase a second, identical unit in 1985. Installation
of the centrifuge substantially reduced labor
requirements at the plant and caused a reduction in
the unit cost of processing the waste sludge.
Wastewater flow to the Central Plant averages 6.8
rrA's (155 mgd), generating approximately 6.2 Mg (70
dry tons)/d of digested waste solids. The solids are a
combination of primary and waste-activated sludges,
which are anaerobically digested before disposal. The
digested sludge is either thickened for agricultural
reuse or dewatered for composting, using two high-
capacity centrifuges.
The decision to purchase centrifuges for dewatering
was made after on-site testing of two belt filter
presses and two centrifuge units during late 1978 and
early 1979. From these tests, District personnel
concluded that centrifugation had the lowest overall
costs. Test results and costs are summarized in
Table 9-5 and are compared to the vacuum filter
units, which were in operation at that time.
The recommendation for centrifugation was based on
the following:
• Centrifuges have lower capital, operation, and
maintenance costs.
• Centrifuges had proved higher capacity for peak
loads of sludge.
• Centrifuges required only polymer for conditioning,
compared to belt presses, which needed both
polymer and ferric chloride.
• Centrifuges have higher throughput capacities than
belt presses.
• Centrifuges can both thicken and dewater.
The centrifuge pre-purchase specification was
directed to both high and low speed units and
included a bid evaluation procedure. The bidders
were required to include a guaranteed loading (in
terms of solids and throughput), power consumption,
and polymer dosages. The equipment comparison for
dewatering is summarized in Table 9-6.
Based on the data presented in each of the bids, an
economic evaluation was completed using the
following parameters:
• Electrical energy at current local rates over a
period of five years.
• Polymer use at current costs to MDSDD No. 1 in
quantity lots over a period of five years.
• Throughput per unit and resultant cake solids.
• Any significant differences in installations and
support system costs.
• Any exceptions taken to the design requirements
or specifications that might in any way affect the
use of this equipment.
On the basis of the evaluation procedure and the
equipment costs, a Humboldt-Wedag centrifuge was
154
-------
Table 9-5. Comparison of Dewatering Systems - Metropolitan Denver Central Plant1
Item Bell Press Centrifuge
Existing Vacuum Filter
B
Capital Recovery
4-yr(1983), $/Mg
10-yr(1989), $/Mg
Chemicals, $/Mg
Operations Labor, $/Mg
Power, $/Mg
Water, $/Mg
Total2, $/Mg
Cake Solids, percent
Solids Recovery, percent
7.56
3.03
48.03
6.47
0.22
2.07
67.38
17.5
91
13.61
5.45
44.25
6.47
0.33
1.93
72.04
17.0
91
13.72
5.49
24.16
4.31
2.28
-
49.96
14.03
90-95
12.36
4.95
20.13
4.31
3.88
-
45.63
12.03
94-95
-
54
6.47
1.58
0.36
62.41
9.5
75-80
1 Based on processing 86.9 Mg/d.
2 Based on 4-yr capital recovery value
3 Based on these results, plant personnel predicted a cake solids of 16 percent.
Table 9-6.
Centrifuge Bids - Metropolitan Denver Central Plant
Bid Specification Humboldt-Wedag
Unit A
UnitB
$530,693
$366,000
$630,500
Design Requirements
Volatile Matter, %
Guaranteed Loading
dry tons/d"
dry Ib/hr*
gpm"
Min. Feed Rate, gpm
Feed Cone., %TS
Min. TS Recovery, %
Min. Cake Solids, %
Dewatering Thickening Dewatering Thickening Dewatering Thickening Dewatering Thickening
61 61 -
96
8,000
500
144
12,000
750
57.72
4,810
300
115.4
9,620
600
60
5,000
300
120
10,000
600
300
2.5-3.2
go
16
600
2.5-3.2
go
6
Max. Polymer, Ib/ton"
Max. HP Draw, hp/gpm*
10.0
0.27
2.5
13
120
4-5
171
15
220
8
220
Values supplied by bidders.
selected. The cost comparison is summarized in
Table 9-7 and shows that the evaluated cost of the
Humboldt machine is 19 percent less than Unit A and
49 percent less than Unit B for solids dewatering.
The centrifuge installed at the Central Plant is a
Humboldt-Wedag S6-1, and it is believed to be the
largest centrifuge in daily use in a municipal facility in
the U.S. This concurrent unit has a bowl dimension of
1.4 m x 4.3 m (56 in x 168 in) long. The bowl is
driven by a 149-kW (200-hp) main drive motor and
the scroll by a 56-kW (75-hp) hydraulic backdrive
system. The backdrive is controlled by a
microprocessor (see Figure 9-8). During automatic
operation, a user-programmable control curve is
keyed into the microprocessor. The control curve
defines the relationship between the scroll conveyor
torque (measured as hydraulic pressure) and the
differential speed. Figure 9-9 shows a control curve
in which the torque is measured as hydraulic
pressure at the inlet to the backdrive hydraulic motor.
The operating point on the curve is a function of the
slurry feed rate and the feed solids concentration.
The control point is set up with a "warning" torque
set point, which, when reached, will cause the
backdrive to increase the differential speed to its
maximum. This increase clears solids from the
machine and thereby avoids an automatic shutdown.
The centrifuges installed at the Central Plant have
been modified, based on research completed by
Humboldt-Wedag. The angle and spacing of the
scroll conveyors have been changed to optimize
detention time in the centrifuge. Another change was
to raise the maximum torque delivered to the
backdrive hydraulic motor. The higher torque results
in larger throughput and drier cakes.
The centrifuge was started in May 1982 and test
results are summarized in Table 9-8. The Denver
155
-------
Table 9-7. Centrifuge Evaluation for Dewatering and
Thickening - Metropolitan Denver Central Plant
Horn
Unit A
Capital
Polymer
Energy
Tola!
UnitB
Capital
Polymer
Energy
Total
Humboldt
Capital
Polymer
Energy
Total
Five-Year Cost
Dewalering
366,000
2.807,142
178,809
3,352,951
630,500
3,239,010
327,817
4,197,327
530,693
2,159,340
120,696
2.810,729
(1982$)
Thickening
366.000
1.926.653
254,803
2,547,456
630,500
3,425,160
476,824
4,532,484
530,693
1.070,363
241,392
1,842,448
Assumptions: Costs straight-lined for 5 years
Power - $0.04536/kWh
Polymer - $0.93/kg ($2.04/lb)
Thickening - 136 m3/hr, 105 Mg/d (600 gpm, 115
dry tons/d)
Dewatoring - 78m3/hr, 53 Mg/d (300 gpm, 58
dry tons/d)
sludge is difficult to dewater; part of the problem is
believed to be the preponderance of waste activated
sludge in the final product. Ferric chloride is used to
adjust the pH of the sludge and avoid scaling (struvite
deposition) of piping and equipment.
The centrifuge has performed reliably since its
installation and has simplified management and
handling of waste sludges at the Central Plant. The
success of the centrifuge installation prompted the
District to purchase a second, identical machine as a
backup.
Higher loading rates became achievable after larger
feed pumps were installed. The results of the cake
production tests then are presented in Figures 9-10
and 9-11. This also was after old sludge was
cleaned out and equilibrium was established. This
data shows a substantial improvement over the
results of acceptance tests and establishes a broader
range of flow rates than were originally maintained
under the acceptance testing. The operational
performance results in Figure 9-10 show that the
centrifuge is capable of achieving greater than 19
percent total solids (TS) with the range of 31-57 l/s
(500-900 gpm) flow rate. The cake dryness can be
controlled within the narrow range of 19.0-19.4
percent TS, while varying the feed rate through the
use of the microprocessor. At differentials of 1 to 9
rpm for the corresponding flow rates of 31-57 l/s
(500-900 gpm), the cake dryness and percent
capture are 19 and 90 percent, respectively. The
polymer feed rate varied from 11 kg/Mg (22 Ib/ton) to
16 kg/Mg (32 Ib/ton) of sludge, as shown in Figure
9-11. The variation in polymer feed rate was
attributed to the inverse relationship with the feed
solids concentration (Figure 9-11). The horsepower
used per unit of flow decreased with increasing flow
(Figure 9-12).
The main feature of this installation is the positive
impact on the District's annual operating costs. Table
9-9 compares the 1986 centrifuge operation and
maintenance costs (O&M) costs to the last year
(1981) of operating the vacuum filters. Table 9-9
shows that, on a unit cost basis, the centrifuge cost is
61 percent of the vacuum filter cost. The cost
difference is attributed to the higher centrifuge
throughput. Because the centrifuge was able to
process all of the sludge (more than twice the
quantity processed by the vacuum filters), chemical
and electrical costs are higher but the cost per ton
processed is lower. The personnel costs reflect one
full-time operator per shift (total of three) and a
maintenance staff consisting of one mechanic, two
utility repairmen, and one electrician. This staff
services not only the centrifuge but all support
equipment such as conveyors, process controllers,
and chemical/sludge feed equipment. Materials costs
reflect spare parts used for routine preventive and
corrective maintenance. Chemical requirements for
polymer are 218 Mg (240 tons)/yr and for ferric
chloride are 481 Mg (530 tons)/yr. Outside services
are primarily for rebuilding the centrifuge bowl and
scroll ($45,000).
The digested sludge is dewatered (19 percent TS) for
composting on-site, or, alternatively, thickened
sludge (8 percent TS) is transported to agricultural
lands where it is injected. The ultimate disposal costs
for 1981 and 1983 are summarized in Table 9-10.
The centrifuge installation has resulted in a 1986
annual savings of $627,259.
The use of a centrifuge has shown benefits for larger
treatment plants, when compared to continuous belt
filter presses. These larger units require less space,
less operator attention, and have a lower initial cost.
The Central Plant would have required at least six 2-
m belt presses to achieve the same capacity as one
centrifuge. Based on the earlier on-site testing, the
cake solids would not have exceeded that produced
by this centrifuge.
9.7 Case Study, Centrifuges, Ontario,
Canada
9.7.7 Sludge Characteristics and Processing
The Duffin Creek Water Pollution Control Plant
(municipalities of York and Durham) is a 1.8-m3/s
(40-mgd) conventional activated sludge facility. It is
designed for the removal of phosphorus in
accordance with the Canadian requirements for the
156
-------
Figure 9-8. Automatic backdrive system.
RPM Meters
Hydraulic Pressure
Transmitter
Backdrive
Motor
Hydraulic
Pump
A = Automatic
M <= Manual
Electronic
Displacement
Controller
Manual
Backdrive
Speed
Control
protection of the Great Lakes. Ferrous sulphate is
added to the aeration system for the removal of
phosphorus. Design features of the plant include:
Primary clarifiers:
4 @ 24 m x 23 m x 3.7 m SWD
(80 ft x 77ft x 12ft)
Aeration basins:
16 @ 23 m x 23 m x 5.8 m SWD
(75 ft x 75 ft x 19ft)
Secondary clariGers:
8 @ 41.2 m dia. x 3.7 m
(135 ft dia. x 12 ft)
The 1986 operational characteristics of the plant were
as follows:
Flow, mgd
BODs, mg/l
TSS, mg/l
TP, mg/l
Influent
35.6
147
266
6.0
Secondary
Effluent
35.6
20.8
17.8
0.9
The plant was designed on the basis of an F/M of 0.3
and a Mean Cell Retention Time (MCRT) of 8 days.
The actual operating conditions for 1986 were as
follows: 0.2 F/M and 15 days MCRT. A schematic of
the Duffin Creek Pollution Control Center is shown in
Figure 9-13.
At Duffin Creek, the excess waste activated sludge
and the primary sludge are digested in conventional
digesters with the designed retention time of 30 days.
The blend sent to the digesters is approximately 70
percent WAS and 30 percent PS. The material, at
present, is digested for 20 days and then dewatered
at a concentration typically between 4 and 6 percent
solids. The daily quantity varied between 15 Mg/day
and approximately 19 Mg/d in 1986.
The waste activated sludge system is presently
producing 45,450 kg WAS/d and the primary clarifier
is capturing 38,290 kg suspended solids/d. The
waste, RPS + WAS, is sent to digestion after blending.
The SVI averaged approximately 60 ml/g in 1986, and
varied between a low of 35 and a maximum value of
78 ml/g.
157
-------
Hguro 9-9. Typical control curves for automatic backdrive.
Typical Control Curve
for Dewatering
10
9
P DMX8
. 7
1 6
« 5
1 4
I 3
£
1
0
—
—
-
•
-
.
f ! 1
0
i <
! y
I S
4
)
I
y
'
IS
r
1
**
I
1
i i i
100 200 30
Pressure, bar
bar x 100 = kPa.
O " First Control Point
• » Second Control Point,
Reduce to 0.8 x
Control Point
A = Maximum, Reduce to
120 Bar
• a Automatic Shutdown
DMX - Maximum Differential Speed
DMN •> Minimum Differential Speed
Typical Control Curve
for Thickening
DMX10
9
P 8
* DMN7
1 6
CO g
1 4
1 3
° 2
1
0
1
^r
- (
-
_
_
-
I i
r
, ,
1 I 1
1
1
I 1 1
) 100 200 300
bar x 100 - kPa.
Pressure, bar
O • First Control Point
A ** Maximum, Reduce to
120 Bar
• m Socond Control Point
I =- Automatic Shutdown
DMX >* Maximum Differential Speed
DMN - Minimum Differential Speed
Table 9-8. Comparison of Bid Specifications and Actual
Performance for the Dewatering Mode
Item Bid Specification Actual Performance
Feed Rale, gpm
Feed Cone. %TS
Solids Recovery, %
Polymer Dose, Ib/ion
Ferric Dose, Ib/ton
HP Draw, hp/gpm
500.0
2.5-3.2
90.0
10.0
-
0.27
Average
424.0
2.6
89.0
14.2
47.0
0.37"
Range
200-600
2.3-3.5
82-95
10.4-19.8
25-70
TS Recovery
" 0.32 hp/gpm @ 800 gpm
Figure 9-10. Centrifuge dewatering.
100
^o
*I 98
85
35
30
25
20
i- 22
^ 20
.g 18
1 16 -
^ 14-
•i 12 -
0 10 -
Polymer Dose
Cake Solids
200 300 400 500 600 700 800 900 1,000
Feed Rate, gal/min
158
-------
Figure 9-11. Polymer requirements.
34 r
32
o
Q.
o
Q-
28
26
24
22
20
I
I
J_
1.5 2.0 2.5 3.0 3.5
Feed Solids Concentration, %
Figure 9-12. Centrifuge power curve (dewatering mode).
800 r
Q.
O>
oc
•a
700
600
500
400
0.10
0.15
0.20 0.25
kwh/gpm
0.30
0.35
9.7.2 Filter Presses
Four membrane filter press systems are used for
dewatering sludges. Each filter is currently equipped
with 66-1,200 mm x 1,200 mm plates. The system,
however, has been expanded to 83 plates. Sludge is
fed to the units with positive displacement pumps
manufactured by Thomas Willett & Co., Ltd. A
measured amount of polymer is injected into the
Table 9-9. Dewatering Equipment Comparison
Metropolitan Denver Central Plant
Activity Vacuum Filter (1981$) Centrifuge (1986 $)
Personnel
Materials
Chemicals
Electricity
Outside Services
Total
Inflation (5%/yr)
Sludge, tons/d
Unit Cost
$/day
766.50
166.61
1,199.76
154.57
19.31
2,306.75
2,944.75
29
101.52
$/day
1,173.04
390.80
2,659.36
299.45
130.23
4,652.88
75
62.04
Table 9-10. Ultimate Disposal Costs -
Central Plant
Activity Vacuum Filter (1981$)
Personnel
Materials
Chemicals
Electricity
Outside Services
Total
Inflation (5%/yr)
Sludge, tons/d
Unit Cost
$/day
4,566.33
2,167.02
1,199.76
174.36
600.44
8,707.91
11,113.91
73
152.25
Metropolitan Denver
Centrifuge (1986 $)
$/day
4,944.56
1,297.64
2,659.36
363.60
130.23
9,395.39
75
125.27
Willett pump on suction stroke and returns to the
storage tank on the discharge stroke. Polymer is
taken from the storage tank by a Moyno pump and
sent to the Willett pump. Electric solenoids determine
the end destination, that is, in pump or in tank. This
system is very reliable for delivering a constant
volume. It is called a "Polymeter System" and is
supplied by Allied Colloids.
The sludge is conditioned with polymer, but without
precoat. A polymer dissolving system is used at
Duffin Creek. The humidity is sufficiently low in the
Toronto area so the hydroscopic properties of the
polymer have not adversely affected the dewatering
operation.
Augers and screws are used to convey all sludges at
the Duffin Creek facility. The augers, manufactured
by Asdor, Ltd., are installed in 4.6 m sections and are
coupled with flexible connectors and grease fittings.
The filter presses at Duffin Creek have been in
operation since 1985. The filter presses replaced belt
presses, installed in 1981, which produced only 18-
21 percent TS. The plant encountered the normal
159
-------
Figure 9-13. Schematic of Duffln Creek Pollution Control Plant, Ontario, Canada.
(1
U
V
Storage
Building
Return and Waste
Sludge Pumping Station
Secondary
Clarifiers
Chlorination
Station
160
-------
start-up difficulties; however, they were
compounded by the discharge of unusually large
quantities of difficult-to-filter waste activated
sludge. Although digested waste activated sludge is
currently fed to the presses, in the beginning the
presses successfully dewatered raw sludge to 28-30
percent cake solids. Actual daily data for the months
of September and December 1986 is compiled in
Tables 9-11 and 9-12, respectively. The data
indicates that the solids content of the cake exceeded
the 30 percent requirement during most of the period.
In 1986, this trend continued during September
through December as shown in the monthly
summaries in Table 9-13. Table 9-14 presents
1986 operating data for the filter press operation.
Table 8-11. Duffin Creek WPCP Dewatering System
(September 1986 Operating Data)
Cake Dry Total Dry Dry
Date Cycles/day Solids Solids Solids/Cycle
September percent Mg/d Mg
1 36 31.2 35.06 0.97
2 36 32.5 36.74 1.02
3 33 33.1 33.95 1.03
4 34 31.5 33.28 0.98
5 40 32.4 39.36 0.98
6 28 30.8 26.33 0.94
7 18 29.8 16.23 0.90
8 22 32.3 21.50 0.98
9 40 34.4 43.28 1.08
10 31 31.6 30.69 0.99
11 37 32.9 38.09 1.03
12 35 33.1 35.54 1.02
13 21 34.8 21.90 1.04
14 23 34.1 24.55 1.07
15 37 34.0 37.19 1.01
16 36 31.8 34.16 0.95
17 12 32.0 11.94 1.00
18 38 33.0 38.36 1.01
19 40 33.8 41.64 1.04
20 29 33.3 29.56 1.02
21 19 34.0 20.35 1.07
22 8 32.0 7.76 0.97
23 32 32.6 31.69 0.99
24 34 34.0 36.16 1.06
25 39 35.1 42.83 1.10
26 31 35.6 33.12 1.07
29 20 33.0 20.86 1.04
30 21 34.3 23.32 1.11
Average 30 33.0 30.19 1.02
Table 9-12. Duffin Creek WPCP Dewatering System
(December 1986 Operating Data)
Cake Dry Total Dry Dry
Date Cycles/day Solids Solids Solids/Cycle
December percent Mg/d Mg
1 30 34.6 32.49 1.08
2 32 35.3 34.70 1.08
3 37 34,0 39.26 1.06
4 36 33.5 37.99 1.06
5 36 34.0 38.56 1 .07
6 22 34.0 23.61 1 .07
7 21 34.0 21.97 1.05
8 34 34.6 36.62 ; 1.08
9 33 36.5 37.42 1.13
10 34 38.8 42.39 1.25
11 35 37.5 41.34 * 1.18
12 26 36.0 29.07 1.12
13 28 37.3 32.15 1.15
14 21 35.8 23.46 1.12
1R OK ' O"7 "I .It CO 1 1Q
13 O3 i3/.l *» 1 .Oo 1. 1 tS
16 35 33.4 36.24 1.04
17 35 34.3 37.39 1.07
18 33 34.0 35.31 1.07
19 33 33.9 34.94 1.06
20 16 34.0 17.01 1.06
21 16 37:3 20.32 1.20
22 32 30.3 29.88 0.93
23 33 31.2 31.61 0.96
24 7 30.9 6.75 0.96
25 '8 34.8 9.11 1.14
26 25 34.3 26.80 1.07
27 28 36.0 31.81 1.14
28 17 31.3 ' 16.64 0.98
29 35 32.5 35.12 1.00
30 30 37.3 35.41 1.18
31 27 31.9 26.52 0.98
Average 28 34.5 30.43 1.08
squeeze time. The average core blow time is 25
seconds. The core blow, however, is incomplete, and
consequently core solids restrict the cake discharge.
The poor core discharge is partially finked to the
extremely high solids content of the core. In addition,
the pipe design at the core blow line restricts the
solids. To improve cake discharge, the core blow
receiving pipes must be larger than the core diameter
and must not make any abrupt changes in direction.
The typical polymer dose is 5 to 6 kg/Mg (10-12
Ib/ton) of 100 percent active material, with a low dose
The system, as indicated, uses only polymer. Precoat
is not employed. In general, the operators indicate the
cake release is fair to excellent. However, they have
to observe the cake discharge to ensure it is
complete.
The operating cycle time is typically 90-95 minutes,
which includes 45 minutes of fill and 50 minutes of
of 3 kg/Mg (6 Ib/ton) and a high of 8 kg/Mg (16
Ib/ton). Allied Colloids' Percol 757 is used at a cost of
$2.80/kg dry powder. In recent months, the ability of
the diaphragm plate presses to achieve 33-36
percent TS on digested sludge has resulted in an
autogenous operation of the fluid bed reactors. At
18-21 percent TS, the furnace capacity was reduced
40 percent and fuel usage was over 400 l/Mg dry
solids. The furnace was designed to receive a
minimum of 30% TS from the belt presses.
161
-------
Table 9-13.
January
February
March
AprH
May
Juno
Jury
August
September
October
November
December
Average
Total
Tobto 9-14.
January
February
March
April
May
Juno
July
August
September
October
November
December
Average
T.^»«*l
Total
Duff In Creek
% solids
4.25
4.63
4.77
5.15
4.90
5.40
5.23
7.09
7.40
6.04
5.97
5.25
5.51
WPCP Sludge Loading for 1986
Raw Sludge
% volatile
56.33
58.78
57.18
61.80
63.44
60.97
55.80
50.69
42.60
51.43
52.54
53.00
55.38
Duffin Creek WPCP Sludge
Volume
571.66
503.62
787.60
759.13
809.92
682.77
800.62
835.64
878.47
678.09
683.91
943.42
744.57
8,934.85
% Solids
Mg
4.2
4.0
4.1
4.2
3.8
3.9
4.3
5.3
7.3
7.4
8.1
7.0
5.3
Digested Sludge Presses
volume, m3
32,045.2
23.462.1
24,971.2
24,443.0
37,544.3
24,608.0
34,322.0
33,120.0
27,530.0
38,858.0
22,448.0
25,096.0
348,447.8
Dewatering
% solids
4.61
3.63
4.11
4.60
4.10
5.38
5.86
6.73
8.25
10.00
8.14
7.04
6.04
for 1906
CST Cake Solids
sec
465.2
417.7
384.6
449.4
522.6
575.9
554.0
506.8
587.4
511.9
537.3
481.4
499.52
%
26.7
26.6
26.4
27.6
26.6
26.1
27.8
28.1
32.9
31.9
34.9
34.5
29.2
% volatile volume, m3 % solids % volatile volume, m3
47.85 10,187.9 26.47 43.84 2,135.5
49.49 9,120.4 26.47 . 44.12 1,898.9
46.84 14,829.9 26.00 49.95 2,986.8
47.28 14,706.9 27.64 45.83 2,755.1
55.75 18,627.2 26.17 47.10 3,025.3
49.00 13,611.9 26.20 45.83 2,677.1
45.60 12,970.7 27.84 44.21 2,883.3
42.50 14,868.3 29.40 40.06 2,851.0
36.18 21,143.0 32.90 35.70 2,659.6
34.94 6,982.0 32.11 32.69 2,129.0
36.15 8,542.0 35.06 32.54 1,968.2
35.77 14,223.0 34.59 34.94 2,725.5
43.95 29.24 41.40
159,813.3 30,695.5
replaced the cloths with new ones. Recently, they
have changed the type of cloth from RILSAN to
polypropylene and now plan to change the cloths
every six months.
In general, the operators like the filter press operation
at Duffin Creek and generally feel it is a cleaner
operation than the original belt press installation. One
exception would be when blowout occurs and the
cover screens are not in place. The system, however,
is now equipped with "fail-safe" curtains (presently
provided by ISB in Montreal), which provide a warning
signal and automatically stop the press cycle if the
safety screens are not in place.
The initial operation used plant air for both
instruments and the compression cycle. As a result,
the instrument part of the operation was starved for
air during the compression periods. To correct the
problem, a second compressor was installed
specifically for the instruments.
9.7.3 Operation and Maintenance
After 10,880 operating hours, the cloths were
changed for the first time. Prior to the cloth change,
filtration rates and cake solids had remained at a
satisfactory level.
The gritty character of the digested sludge has
created operating and maintenance problems at
Duffin Creek. Fine grit passes through the filter
medium and lodges in the plates; these plugged
points hinder cake release. As a result, the plant
removed the cloths after five months and cleaned the
entrapped fines. After another four months, the plant
The operators feel the polymer system is
fundamentally a good one. However, as the
proportion of primary to secondary sludge varies
(even after digestion), the polymer dosage will also
change, thus affecting the cake release. An ideal
situation would be to supply consistent quality of
sludge to the unit.
9.8 Reference
1. Williams, R.B. and J.K. Nelson. Digested Sludge
Thickening and Dewatering at the Metropolitan
Denver Central Plant. Presented at the 59th
Annual Conference of the Water Pollution Control
Federation, Los Angeles, CA, October 5-9,
1986.
162
-------
Appendix A
Design ExampleslCost Analyses
A.1 introduction
This appendix presents design examples for several
dewatering processes. The examples are prepared for
two different sizes of treatment plants: 0.088 m3/s (2
mgd) and 0.88 m3/s (20 mgd). Information from the
design chapters is used as the basis for the
calculations. Cost analyses are also presented to
show how to prepare cost estimates by using the
EPA publication Handbook: Estimating Sludge
Management Costs (\).
Note: Construction cost estimates presented in this
Appendix are total base capital costs obtained from
the EPA Handbook (1). Total base capital costs
(TBCC) for sludge dewatering processes presented in
this Appendix include structural, mechanical,
equipment, electrical, and instrumentation costs. They
do not include costs for engineering design,
construction supervision, legal costs, administration,
interest during construction, and contingencies. In
order to estimate the total project construction cost,
these non-construction costs must be estimated and
added to the process TBCC costs derived from the
cost curves in the EPA Handbook. In addition, the
costs presented in this Appendix are based on last
quarter 1984 costs, and must be adjusted for inflation
for use in later years. An example of how to estimate
the total project construction cost given the TBCC is
shown for upgraded sand drying beds under Section
A.3.2 for the 0.088-m3/s (2-mgd) plant. More detail
on these cost update procedures can be found in
Section 2.6 of the EPA Handbook (1).
A.2 Determine Sludge Quantities
Assume Activated Sludge Plant with Primary
Clarifiers
Sludge Type:
Primary Sludge (P) + Waste Activated Sludge
(WAS)
Sludge Quantities:
P - 150 kg solids/Ml treated (1,250 Ib/Mgal)
WAS - 90 kg solids/Ml treated (750 Ib/Mgal)
Sludge Solids Concentrations:
P - 5.0% from primary clarifier
WAS - 0.5% from secondary clarifier
Thickened WAS - 4.0% from dissolved air
flotation thickener
Determine Sludge Volumes Before Anaerobic
Digestion
Assume Sludge Solids Specific Gravity
= 1.4 for primary sludge
= 1.25 for WAS
Sludge Specific Gravity [Equation 2-3 from (1)]:
SSG =
SS
(100KSPG)
(100 - SS)
100
where,
SSG = sludge specific gravity (dimensionless).
SS = sludge suspended solids concentration,
weight percent.
SPG = sludge solids specific gravity
(dimensionfess).
Primary Sludge SSG
= 1/[[(5)/(100)(1.4)l + {(100 - 5)/100]] = 1.01
Thickened WAS SSG
= 1/J[(4)/(100)(1.25)] + 1(100 - 4)/100]] = 1.01
Therefore, combined P + WAS has SSG = 1.01.
Sludge Volume:
SV =
(DSSKWQ)
(SS)(l.Okg/D(SSG)
where,
SV =
DSS =
SS =
SSG =
1.0 =
sludge volume, I/Ml treated
dry sludge solids produced, kg/Ml
sludge suspended solids concentration,
percent
sludge specific gravity (dimensionless)
density of water, kg/I
163
-------
SV = ((150 kg/Ml treated)/(0.05)(1,0)(1.01)]
•*• [(90 kg/Ml treated)/(0.04)(l.0)(1.01)]
= 5,200 I/Ml treated (5,200 gal/Mgal)
Solids Content before Digestion
= (100)(150 + 90)/[(5,20Q)(1.0)(1.01)] percent
= 4.6 percent solids
Assume anaerobic digestion destroys 50% of volatile
solids. Assume 70% of sludge solids are volatile.
Mass of solids before digestion:
150 + 90 = 240 kg/Ml
Solids Destroyed = (240 kg/MI)(0.70)(0.50)
= 84 kg/Ml (700 Ib/Mgal)
Solids Remaining = 240 kg/Ml - 84 kg/Ml
= 156 kg/Ml (1,300 Ib/Mgal)
Solids Content after Digestion
= (100)(156 kg/MI)/(5,200)(1.0)(1.01) percent
= 3.0 percent solids
A.3 Upgraded Sand Drying Beds
A.3.1 Design Examples
When designing sand drying beds for a wastewater
treatment plant, the required size or bed area
depends on the geographic location of the plant. In an
area with a high evaporation rate, the bed area can
be smaller than in an area with a low evaporation
rate.
From Chapter 6 of this manual, typical loading criteria
for anaerobically digested primary sludge plus waste
activated sludge are 60 to 100 kg/m2/yr (12 to 20
Ib/sq ft/yr). For polymer conditioned sludges these
criteria can reasonably be increased by 50 to 100
percent to 120 to 200 kg/m2/yr (25 to 41 Ib/sq ft/yr).
For this example, use 140 kg/m2/yr (29 Ib/sq ft/yr).
This loading rate would be appropriate for
southwestern and southern regions of the United
States. For northern regions of the United States, a
loading rate of 80 to 100 kg/m%r (17 to 21 Ib/sq ft/yr)
is more appropriate.
Design Example for a 0.088-m3/s (2-mgd) Plant
Sludge Quantity
Digested Sludge Solids
= (0.088 m3/s)(156 kg/MI)(86,400 s/d)/( 1,000 m3/MI)
= 1,190 kg solids/d (2,600 Ib/d)
Digested Sludge Volume
* (1,190 kg/d)/(0.03)(1.0 kg/l)(1.01)
= 39,300 l/d (10,400 gpd)
Determine Required Sand Bed Area (based on solids
loading from above)
Bed Area = (1,190 kg/d)(365 d/yr)/140 kg/m2/yr
= 3,100 m* (33,000 sq ft)
Check Number of Sludge Applications per Year
[assume normal application depth of 23 cm (9 in)]
Total Volume Available per Application
= (3,100 m2)(23 cm)(1,000 l/m3)/100 cm/m
= 713,000 I (188,000 gal)
Number of Applications Per Year
= (39,300 l/d)(365 d/yr)/(713,000 I/application)
= 20 applications/yr
If the plant is located in the northern part of the
United States, 20 applications per year is too high for
design purposes. In the southern and southwestern
United States, this loading rate is reasonable.
Determine Number of Beds
If beds are 7.6 m wide by 30.5 m long (25 ft by 100
ft), the area of one bed is 232 m2 (2,500 sq ft).
Approximate No. Beds
= 3,100m2/(7.6m)(30.5m)
= 13,4 beds, use 14 beds
Fix Width @ 7.6 m (25 ft)
Length = 3,100 m2/(i4)(7.6 m) = 29.1 m (95.6 ft)
Design Example for a 0.88-m3/j; (20-mgd) Plant
The design approach is the same as in the previous
example with the exception that larger sand beds,
such as 30 m x 60 m (100 ft x 200 ft), would be
used. The overall bed area will be determined.
Sludge Quantity
Digested Sludge Solids
= (0.88 m3/s)(156 kg/MI)(86,400 s/d)/(1,000 m3/MI)
= 11,900 kg/d (26,000 Ib/d)
Digested Sludge Volume
= (11,900 kg/d)/{0.03)(1.0 kg/l)(1,01)
= 393,000 l/d (104,000 gpd)
Determine Sand Bed Area Required
Bed Area = (11,900 kg/d)(36S d/yr)/140 kg/m2/yr
= 31,000rn2 (330,000 sq ft)
A.3.2 Cost Analyses
Cost Analysis for a 0.088-m3/s (2-mgd) Plant
Determine Capital Cost
Refer to Figure 5-13 on page 85 of the EPA
Handbook: Estimating Sludge Management Costs (1).
Note: English units only are used in the EPA
Handbook.
Need Sludge Volume per Year (from Design
Example):
10,400 gpd x 365 d/yr = 3.80 x 108 gal/yr
Figure 5-13 is based on sludge loading rates of 15
Ib/sq ft/yr at 2 percent solids and 22 Ib/sq ft/yr at 4
164
-------
percent solids, or about 18.5 Ib/sq ft/yr at 3 percent
solids. The design example size was based on 29
Ib/sq ft/yr. Using a curve based on a low loading rate
to size the sand bed would yield a bed (and cost)
which is too large.
Adjust sludge volume by ratio of 18.5/29 = 0.64:
0.64 x 3.8 Mgal/yr = 2.4 Mgal/yr
From Figure 5-13 @ 3% solids:
Capital Cost = $140,000
Note: This includes a land cost valued at $3,120 per
acre. If a significantly different cost for land is used,
the capital cost needs to be adjusted, as shown in
Section 5.8.1 of the EPA Handbook (1).
Note: The drying beds in the EPA Handbook (1) have
no plastic or asphalt lining beneath the beds for
groundwater protection and no concrete tracks for
easy equipment access. If the user elects to include a
PVC liner, add about $2.50/sq ft. If the user elects to
include concrete tracks, add about $2.00/sq ft.
Example - Convert capital cost to total project
construction cost
Update cost to May 1987.
Engineering News Record (ENR) construction cost
index in May 1987 is 4367. ENR index for last quarter
1984 is 4171:
4367/4171 = 1.047
Updated capital cost = $140,000 x 1.047
= $146,600
Add non-construction costs to updated cost:
Engineering Design @ 10% of $146,600
= $14,700
Construction Supervision @ 5% of $146,600
= $7,300
Legal and Administrative Costs @ 20% of $146,600
= $29,300
Contingencies @ 15% of $146,600
= $22,000
Subtotal Cost = $219,900
Interest During Construction (assume 10% Interest
and 1-year Construction Period):
0.10 x 1 yr x 1/2 x $219,900 = $11,000
Total Project Construction Cost = $230,900
More detail on these cost update procedures can be
found in Section 2.6 of the EPA Handbook (1).
Determine O&M Cost
Use Figure 5-14 of the EPA Handbook. Use actual
sludge volume to determine O&M cost.
From Figure 5-14 @ 3.8 Mgal/yr and 3% Solids:
Base Annual O&M Cost = $16,000/yr
Use Figure 5-16 of the EPA Handbook to determine
O&M cost components. From Figure 5-16:
Labor: 750 hr/yr x $13.50/hr = $10,100/yr
Diesel Fuel: 2,500 gal/yr x $l.35/gal = $3,400/yr
Materials: $2,000/yr
Total =
$15,500/yr
$l5,500/yr is close enough to total O&M of
$l6,000/yr - within accuracy of curves.
Determine Costs for Polymer Feed System
A typical polymer cost in $/ton dry solids is $11/ton,
from the case studies on air drying systems in
Chapter 8. From O&M Cost Figure 6-22 from the
EPA Handbook (1), the assumed cost of polymer is
$2.80/lb.
Based upon this dry polymer cost, the approximate
dry polymer dosage is:
($11/ton)($2.80/lb) = 3.9 Ib/ton dry solids
Therefore, use a polymer dosage of 4 Ib/ton when
determining the capital costs.
Figures 6-19 and 6-20 from the EPA Handbook (1)
are used to estimate the capital costs for polymer
feed systems treating a sludge with 2 percent and 4
percent solids, respectively. Since the design
example has 3 percent sludge solids, an average of
the capital costs from each figure will give a useable
cost.
Use Figure 6-19, based on 3.8 Mgal/yr and 2%
solids:
Capital Cost = $31,000
Use Figure 6-20, based on 3.8 Mgal/yr and 4%
solids:
Capital Cost = $33,000
For 3% solids sludge, use an average of $32,000 for
the capital cost.
For the O&M cost, similarly:
Figure 6-22 gives an O&M cost of $14,000/yr.
Figure 6-23 gives an O&M cost of $20,000/yr.
For 3% solids sludge, use an average of $17,000/yr
for the O&M cost.
165
-------
Cost Analysis for a 0.88-m3/s (20-mgd) Plant
Determine Capital Cost
Use same approach as for 0.088-m3/s (2-mgd)
Plant.
Need Sludge Volume per Year:
104,000 gal/d x 365 days/yr = 38.0 x 106 gal/yr
Adjust Sludge Volume due to Loading Rate:
0.64 x 38,0 Mgal/yr = 24 Mgal/yr
From Figure 5-13 of the EPA Handbook (1) @ 3%
solids:
Capital Cost = $1,100,000
Determine O&M Cost
Use Figure 5-14 and actual sludge volume of 38
Mgal/yr.
Base Annual O&M Cost = $140,000/yr
Determine Costs for Polymer Feed System
See approach used for 0.088-m3/s (2-mgd) plant.
A.4 Vacuum Assisted Drying Beds
A.4.1 Design Examples
Design Example for a 0.088 m3/s (2 mgd) Plant
Note: Since Vacuum Assisted Drying Beds (VADBs)
are not typically used in 0.88-m3/s (20-mgd)
plants, a Design Example has been prepared only for
the smaller 0.088-m3/s (2-mgd) plant.
Sludge Quantity
Sludge Solids = 1,190 kg/d (2,600 Ib/d)
Sludge Volume = 39,300 l/d (10,400 gpd)
Select a Solids Loading Rate
A typical upper limit solids loading rate for a VADB,
without decanting any supernatant, is about 10
kg/m2/cycle (2 Ib/sq ft/cycle). However, one can
assume that some decanting of supernatant can be
accomplished and that a solids loading of 15
kg/m2/cycle (3 Ib/sq ft/cycle) can be used. A typical
cycle can be completed in 24 hours or less.
It is desirable to have a minimum of two VADBs at a
plant, and three are preferable for flexibility. Each bed
should be sized to handle, at a minimum, 70 percent
of an average daily sludge quantity.
Minimum Bed Size = (1,190 kg/d)(0.70)/l5 kg/m2/d
= 55 m2 (590 sq ft)
Standard media plate sizes are 0.6 m x 0.6 m (2 ft x
2 ft) or 0.6 m x 1.2 m (2 ft x 4 ft).
Standard VADB sizes are 6. 1 m x 6. 1 m (20 ft x 20 ft)
or 6.1 m x 12.2 m (20 ft x 40 ft). The area calculated
above is in between the standard VADB sizes of 37
and 74 m2 (400 and 800 sq ft).
Check bed solids loadings at 37 and 74 m2 (400 and
800 sq ft):
(1,190 kg/d)/37
= 32 kg/m2/cycie
(6.6 Ib/sq ft/cycle)
(1,190kg/d)/74m2 = 16 kg/m2/cycle
(3.3 Ib/sq ffcycle)
The solids loading for a 37-m2 (400-sq ft) bed
would be too high to be acceptable as a design basis.
The solids loading for a-74 m2 (800-sq ft) bed
would be very conservative if three beds are chosen.
More than likely, if the 74-m2 (800-sq ft) bed was
chosen, only two beds would be designed.
A better design is to select three beds of about 56 m2
(600 sq ft) each. This design allows more flexibility
than with two overly large beds. With three beds, the
dewatering system can operate an average of 5 days
per week, yet it can handle a full 7 days'
accumulation of sludge,
Size Polymer Feed System
From Section 6.4.3, the median polymer dosage at 1 3
VADB systems was approximately 10 g/kg (20 Ib/ton)
of solids. It is necessary to determine the volume of
liquid emulsion type polymer to use. For example, a
liquid emulsion type polymer may be 12 percent
polymer by weight.
Thus, total weight of polymer is:
(10g/kg)/(0.12) = 83 g polymer/kg solids
(166!b/ton)~
If sludge is flowing to the VADB at a velocity of 0.6
m/s (2 ft/sec) through a 15 cm (6 in) pipe, the sludge
feed rate is:
Sludge Flow
= Velocity x Cross-sectional Area
= (0.6 m/s)(n/4)(0.15 m)2(i,000 l/m3)(60 s/min)
= 636 l/min (168gpm)
Each sludge flow volume of 636 l/min (168 gpm) has
the following quantity of sludge solids:
(636l/min)(1.0kg/I)(1.01)(0.03) = 19 kg/min
(42 Ib/min)
Polymer Requirement
= (83 g polymer/kg solids)(19 kg/inin)/1,000 g/kg
= 1.58 kg/min (3.48 Ib/min)
166
-------
If polymer is diluted
polymer:
5 volumes water per 1 volume
Feed Volume = (6)(1.58 kg/min)/(1.0 kg/I)
= 9.5 l/min (2.5 gpm)
These calculations indicate that the polymer system
should be capable of feeding a minimum of 7.5 to 75
l/min (2 to 20 gpm) of polymer feed solution and 1.25
to 12.5 kg/min (2.75 to 27.5 Ib/min) of polymer.
A.4.2 Cost Analysis
The EPA Handbook (1) does not contain costs for a
vacuum assisted drying bed. See Section 6.4.5 of this
Design Manual for cost information.
A.5 Belt Filter Presses
A.5.7 Design Examples
Design Example for a 0.088-m3/s (2-mgd) Plant
Sludge Quantity
Sludge Solids = 1,190 kg/d (2,600 Ib/day)
Sludge Volume = 39,300 l/d (10,400 gpd)
Determine Size and Number of Belt Presses
At a plant of this size, one can assume that sludge
dewatering operations would be restricted to 5 days
per week and 8 hrs per day.
(1,190 kg/d)(2,600 lb/d)(7 d/wk)
= 8,330 kg/wk (18,200 Ib/wk)
(39,300 l/d)( 10,400 gpd)(7 d/wk)
= 275,100 i/wk (72,800 gal)/wk
From Section 7.2.6, a typical sludge throughput is 2.5
l/s (40 gpm) per meter belt width.
Hours Required Per Week
= (275,100 l/wk)/(2.5 l/s)(3,600 s/hr)
= 30.2 hr/wk/m of belt width
With 5 days per week and 8 hrs per day, there are 40
hrs per week available. 30.2 hrs is less than 40 hrs,
therefore one 1-m belt press would provide enough
capacity @ 6.0 hr/d.
Operation would be for 5 days/wk. One belt press
would typically be sufficient for a plant of this size,
provided there is some other backup means of
dewatering or storing sludge for several weeks at
most.
Size Polymer Feed System
Typical polymer requirements (from Section 7.2) are
2-8 g/kg (3-15 Ib/ton).
Sludge Solids = (8,330 kg/wk)/(30.2 hr/wk)
= 276 kg/hr (608 Ib/hr)
Design to provide sufficient capacity lo feed up to 10
g/kg (20 Ib/ton) polymer.
Polymer Requirements
= (276 kg/hr)(10 g polymer/kg)
= 2,760 g polymer/hr (6.2 Ib/hr)
If polymer solution is @ 0.1 percent:
Feed Volume
= (2,760 g/hr) (0.001 )(1 kg/l)(1000 g/kg)
= 2,760gl/hr (729 gal/hr)
Thus, the Polymer Feed system should be capable of
feeding 3,000 g/hr (6.6 Ib/hr) of dry polymer and
3,000 l/hr (790 gal/hr) of polymer solution.
Volume of Sludge Cake Produced
Determine Sludge Cake Specific Gravity (SSG).
Assume the sludge cake is 20 percent solids and that
the Digested Sludge Solids have a specific gravity of
1.4.
SSG [Equation 2-3 from (1))
= 1/[[(20)/(1QQ)(1.4)1 + [(100 - 20)/100H
= 1.06
Sludge Volume [Equation 2-1 from (1)|
= (276 kg/hr)(6.0 hr/d)/(1.06)(1.0)(0.20)
= 7,800 l/d (275 cu ft/d)
= 7.8m3/d (10cuyd/d)
Belt Press Filtrate
Washwater requirements (vary with belt press
manufacturer):
Assume 3.16 l/s (50 gpm)/m belt width (per
Section 7,2).
Solids Capture: Assume 85% (worst case).
Solids in Filtrate: 276 kg/hr x (1 - 0.85) = 41 kg/hr
(90 Ib/hr)
Filtrate Flow Rate = Washwater Flow Rate
+ Digested Sludge Volume
- Dewatered Sludge Volume
= (3,16l/s)(3,600s/hr)
+ (275,100 l/wk)/(30.2 hr/wk)
- (7,800 l/d)/(6 hr/d)
= (11,400 + 9,100 - 1,300) l/hr
= 19,200 l/hr
Filtrate Volume Per Day
= (19,200 l/hr) (6 hr/d)
= 115,200 l/d (30,432 gpd)
Filtrate Solids Concentration
= (100)(41 kg/hr)/(19,200 l/hr)(1.0 kg/I)
= 0.21 percent solids
= 2,100 mg/l
167
-------
Design Example for a 0.88-m3/s (20-mgd) Plant
Sludge Quantity
Sludge Solids = 11,900 kg/d (26,000 Ib/d)
Sludge Volume = 393,000 l/d (104,000 gpd)
Determine Size and Number of Belt Presses
Assume sludge dewatering operations are 16 hr/d and
7 days/wk. A typical sludge throughput is 2.5 l/s (40
gpm)/m belt width,
(2.5 l/s/m)(16 hr/d)(3600 s/hr)
= 146,000 l/d/m belt width required operating at
16 hr/d
(393,000 l/d)/( 146,000 I/d/m)
= 2.7 m of belt required
At first glance, it appears that two 1.5-m belt
presses could handle the sludge volume. However, if
one belt press is out of service, could the other unit
handle the total plant sludge if operated 24 hours a
day?
Check: 146,000 l/dy/m x 1.5m x (24 hr/16 hr)
« 329,000 l/d, < 393,000 l/d sludge flow
Thus, two 1.5-m belt presses are not sufficient.
Try two 2-m belt presses, Again, if one unit is out of
service, the other unit must process the total sludge
volume operating 24 hr/d.
Sludge Throughput in 24 hr
= (2.531 l/s/m)(24 hr/d)(3,600 s/hr)
= 219,000 l/d/m
(393,000 l/d)/( 219,000 l/d/m)
= 1.8 m of belt required
One machine could temporarily handle the total
sludge flow. Check the operating time for two 2-m
belt presses:
Hours Required Per Day
= (393,000 l/d)(4 m belt)/(146,000 l/d/m)
= 10.8 hr/d
Polymer Feed System
See approach for (0.88 m3/s) 2 mgd plant.
Volume of Sludge Cake Produced
Sludge Solids Feed Rate
= (11,900kg/d)/(1Q.8hr/d)
= 1.100 kg/hr (2,430 Ib/hr)
Sludge Volume
= (1,100 kg/hr)(10.8 hr/d)/(1.06)(1.0 kg/l)(0.20)
= 56,000 l/d (14,800 gpd)
= 56 m3/d (73 cu yd/d)
Belt Press Filtrate
Assume solids capture 85% (worst case).
Solids in Filtrate: 1,100 kg/hr x (1 - 0.85)
= 165 kg/hr (364 Ib/hr)
Filtrate Flow Rate = Washwater FJow Rate
+ Digested Sludge Volume
- Dewatered Sludge Volume
= (3.16 l/s)(3,600 s/hr)
+ (393,000 l/wk)/(10.8 hr/wk)
- (56,000 l/d)/(10.8 hr/d)
= (45,000 + 36,400 - 5,200) l/hr
= 76,600 l/hr
Filtrate Volume per Day
= (76,700 l/hr)(10.8 hr/d)
= 828,000 l/d (219,000 gpd)
Filtrate Solids Concentration
= (100)(165 kg/hr)(76,700 l/hr)(1.0 kg/I)
- 0.22 percent solids
= 2,200 mg/t
A.5.2 Cost Analysis
Cost Analysis for a 0.088-m3/s (2-mgd) Plant
Determine Capital Cost
Use Figure 5-4, Base Capital Cost of Belt Filter
Press, page 75 of EPA Handbook (1).
Need Sludge Volume per year (from Design
Example):
(10,400 gpd)(365 days/yr) = 3.80 x 106 gal/yr
Figure 5-4 is based upon a solids loading rate of
500 Ib/hr per meter for 2 percent solids and 650 Ib/hr
per meter for 4 percent solids, or about 575 Ib/hr/per
meter for 3 percent solids.
Check Solids Loading in Design Example:
(40 gal/min-m)(60 min/hr)(8.34 lb/gal)(1.01)(3/100)
= 606 Ib/hr/ m
606 - 575 Ib/hr. No need to adjust curve for this
difference.
Figure 5-4 is based upon 8 hr/day, 7 days/wk
operation.
8 hr/d x 7 d/wk = 56 hr/wk operation
From Figure 5-4 @ 3% solids and 3.8 Mgal/yr:
Capital Cost = $270,000
From Design Example:
With one 1-m belt press, a 40 hr/wk operation
(including startup, shutdown, and cleanup) is
168
-------
common. With fewer hours of operation, the belt
press would need to be larger. Therefore, the capital
cost obtained from the curve is too low.
Adjust sludge volume by ratio of 56/40 = 1.40
1.40 x 3.8 Mgal/yr = 5.3 Mgal/yr
From Figure 5-4 @ 3% solids and 5.3 Mgal/yr:
Capital Cost = $290,000.
Note: The adjustment procedure for different hours of
operation is also illustrated on page 27 of the EPA
Handbook (1).
Determine O&M Costs
Use Figure 5-5 of the EPA Handbook, which is for
Base Annual O&M.
For the actual sludge flow of 3.8 Mgal/yr and 3%
solids:
Base Annual O&M Cost = $0.012 million/yr
= $12,000/yr
This is based upon a labor cost of $13.50/hr and an
electricity cost of $0.094/kWh. Using the same
procedure and Figure 5-6, the labor man hours,
material costs, and electrical energy costs in kWh/yr
can be obtained.
Determine Costs for Polymer Feed System
Capital and O&M costs for the polymer feed system
can be computed using the EPA Handbook (1), as
shown in Section A.3.2 of this manual.
Cost Analysis for a 0.88 m3/s (20 mgd) Plant
Determine Capital Cost
Use Figure 5-4 of the EPA Handbook (1).
Sludge Volume per year = 38 x 106 gal/yr
Figure 5-4 is based on 8 hr/day, 7 d/wk operation.
From the design example, operation is 10.8 hr/d, 7
d/wk. Thus, the belt presses could be smaller (the
capital cost would be too high).
Adjust Sludge Volume by ratio of 8/10.8 = 0.74
0.74 x 38 Mgal/yr = 28 Mgal/yr
From Figure 5-4 @ 3% solids and 28 Mgal/yr,
Capital Cost = $650,000.
Determine O&M Costs
From Figure 5-5 @ 3% solids and 38 Mgal/yr,
O&M Cost = $75,000/yr.
Determine Costs for Polymer Feed System
Use procedure shown in Section A.3.2 of this manual
to obtain costs from the EPA Handbook (1).
A.6 Solid Bowl Centrifuges
A.6.1 Design Examples
Design Example for a 0.088-m3/s (2-mgd) plant
Sludge Quantity
Sludge Solids = 1,190 kg/d (2,600 Ib/d)
Sludge Volume = 39,300 l/d (10,400 gpd)
Sludge Solids Concentration = 3%
Determine Size and Number of Centrifuges
Refer to section 7.3 of this manual. Table 7-10
shows suggested capacities and numbers of
centrifuges for various plant sizes. For the 0.088-
m3/s (2-mgd) plant, the sludge flow is 40 rr»3/d,
which is essentially equal to the 39,300 l/d used in
this example. The number of hours of operation,
seven, is reasonable as is the number of centrifuges
(one duty, one standby). It is assumed that centrifuge
operations are 7 hr/d, 7 d/wk.
Write Performance Specification
As described in Section 7.3, for a specific sludge flow
rate, the actual size (both diameter and length) of the
required centrifuge varies from supplier to supplier.
The low speed centrifuge compared to the high-
speed centrifuge of similar capacity will be larger in
diameter.
The typical engineering design of a centrifuge
dewatering facility will select the number of units and
will specify the performance which must be achieved.
Typically, the design engineer does not select the
actual model and size of centrifuge required, unless
on-site, side-by-side testing of several different
centrifuge models has been conducted.
A typical performance specification includes both
design sludge characteristics and performance
requirements. Two such sections from a complete
specification on a centrifuge are presented for the
0.088-m3/s (2-mgd) plant as follows:
Job Conditions
A. Design Sludge Characteristics
1. Mixed anaerobically digested sludge to
centrifuge
a. Total solids
1) Average: 3%
2) Range: 2.5-3.5%
b. Volatile solids
1) Average: 54% of total solids
2) Range: 50-70% of total solids
c. Ratio WAS to primary sludge before digestion
1) Average: 38:62
2) Range: 35:65 to 50:50
d. Average temperature: 21 °C
169
-------
Design and Performance
A. The following performance requirements shall be
met for the sludge characteristics specified herein.
1. Digested sludge dewatering
a. Cake solids: 18% minimum
b. Recovery of suspended solids: 90% minimum
c. Polyelectrolyte allowed to achieve required
cake solids and recovery: 4 g polymer/kg dry
sludge solids max. (8 Ib polymer/ton max.)
d. Design flow rate: 5.7 m3/hr (25 gpm)
A complete engineering specification for solid-bowl
centrifuges contains many more requirements than
merely the performance specification. Other items
that would be included are descriptions of required
submittals of shop drawings and product data,
equipment, O&M manuals, product delivery
requirements, and guarantees. Specifications would
also include material requirements for bowls,
conveyors, and for the type of abrasion resistant
materials to be used. Also included are requirements
for the fabrication and manufacture of the many
components of the centrifuge including the motor and
drive, the type of backdrive system, and the electrical
controls.
Polymer Feed System
Although the performance specification requires the
centrifuge to use a maximum polymer dosage of 4
g/kg (8 Ib/ton), the polymer system should be
designed to be able to deliver a minimum of 10 g/kg
(20 Ib/ton).
Volume of Sludge Cake Produced (Assume 18
Percent Solids):
Determine Sludge Gate Specific Gravity, SSG
SSQ {Equation 2-3 from (1)]
= 1/[[(18)/(100)(1.4)] -i- [(100 -18)/100]]
= 1,05
Sludge Volume [Equation 2-1 from (1)]
= (1,190kg/d}/(1.05)(1.0)(0.18)
= 6,300 l/d (1,700 gpd)
= 6.3 m3/d (8.2 cu yd/d)
Centrate
Centrate Volume = Digested Sludge Volume
- Dewatered Sludge Volume
= 39,300 l/d - 6,300 l/d
= 33,000 l/d (8,700 gpd)
Assume 90% solids capture (worst case):
Solids in Centrate: 1,190 kg/d (1 - 0.90) = 119 kg/d
(262 Ibid)
Centrate Solids Concentration
= (100)(119 kg/d)/(33,000 l/d)(1.0 kg/I)
= 0.36 percent solids
= 3,600 mg/l
Design Example for a 0.88-m3/s (20-mgd) Plant
Sludge Quantity
Sludge Solids = 11,900 kg/d (26,000 Ib/d)
Sludge Volume = 393,000 l/d (104,000 gpd)
Determine Size and Number of Centrifuges
Refer to Section 7.3.
For a 0.88-m3/s (20-mgd) plant size, the sludge
flow is 320 m3/d, which is 19 percent lower than the
sludge volume assumed in these design examples.
The number of hours of operation (15) is too low for
12-mS/hr (53-gpm) centrifuges. This design should
select three 13-m3/hr (58-gpm) centrifuges, based
on 15 hr/day operation. The 15 hr/day operation is
desirable because it allows 2-shift operation and one
hour for required cleanup.
Write Performance Specification
The performance specification would be identical to
the one for the 0.088-m3/s (2-mgd plant) with the
exception that the design sludge flow rate is 13 m3/hr
(58 gpm).
A.6.2 Cost Analysis
Cost Analysis for a 0.088 m3/s (2 mgd) Plant
Determine Capital Cost
Use Figure 5-1, Base Capital Cost of Centrifuge, of
the EPA Handbook (1).
Sludge Volume per Year (from Section A.3,2)
= 3.80 x lQ6gal/yr
Figure 5-1 is based upon 8 hr/d, 7 d/wk operation. In
this example, only 7 hr/d operation is assumed. Thus,
the centrifuges in the example need to be slightly
larger than assumed in Figure 5-1.
Adjust annual sludge volume by ratio of 8/7 = 1,14
1.14 x 3.8 Mgal/yr = 4.3 Mgal/yr
From Figure 5-1 @ 3% solids:
Capital Cost = $260,000.
Determine O&M Cost
From Figure 5-2 and actual sludge volume of 3.8
Mgal/yr,
O&M Cost = $30,000/yr.
Cost Analysis for a 0.88 mS/s (20 rngd) Plant
Determine Capital Cost
Use Figure 5-1, page 72 of EPA Handbook (1).
170
-------
Sludge Volume per Year (from Section A.3.2)
= 38 Mgal/yr
Adjust annual sludge volume by ratio of 8/7 = 1.14
1.14 x 38 Mgal/yr = 43 Mgal/yr
From Figure 5-1, Capital Cost = $620,000
Determine O&M Cost
From Figure 5-2 and actual sludge volume of 38
Mgal/yr,
O&M Cost = $82,000/yr
Polymer Feed Systems
Figures 6-19 and 6-20 of the Handbook would be
used to obtain polymer feed system capital costs as
was done in section A.3.2. The design polymer
dosage should be 10 Ib/ton.
For O&M costs, Figures 6-22 and 6-23 would be
used. The operating polymer dosage should be 6
Ib/ton.
A.7 Filter Presses
A.7.1 Design Examples
Design Example for a 0.088 m3/s (2 mgd) Plant
Sludge Quantity
Sludge Solids = 1,190 kg/d (2,600 Ib/d)
Sludge Volume = 39,300 l/d (10,400 gpd)
Sludge Solids Concentration = 3%
At this size plant, assume 8 hr/d, 5 d/wk operations.
(1,190kg/d)(7d/wk) = 8,330 kg/wk
(18,200 lb)/wk
(39,300 l/d)(7 d/wk)
275,100 l/wk
(72,800 gal)/week
Determine Dewatered Sludge Cake Volume
Assume 40 percent solids cake and that this is
equivalent to a cake density of 1,140 kg/m3 (71 Ib/cu
ft)
Cake Volume per Day of Operation
= (8,330 kg/wk)/(5 d/wk)(0.40)(1,140 kg/m3)
= 3.65 m3/d (12i cu ft/d)
= 3,650 l/d
Determine Size of Filter Press Required
Assume one press capable of dewatering all the
sludge volume per day. Assume in an 8-hour day
that, conservatively, 3 filter cycles could be
completed.
Required Filter Volume = (3,650 l/d)(3)
= 1,2201 (43cuft)
Number of Filter Presses Required
For a wastewater treatment plant this small, only
0.088 m3/s (2 mgd), a filter press would be selected
for dewatering only if (1) a very dry sludge cake is
required, and (2) very little land area is available. In
this case, dewatering reliability is probably critical,
and two filter presses should be provided. One filter
press would serve as a standby unit.
Polymer Dosage
A draft EPA report on recessed plate filter press
design and operation (2) describes conditioning with a
highly charged cationic polymer to produce sludge
cake of 35% solids. One facility reported reducing its
chemical conditioning cost from $76/dry ton of
dewatered sludge to $24/dry ton by using polymer
instead of lime and ferric chloride. Determination of
the size and costs for a polymer feed system can be
conducted as done in Section A.3.2 of this manual.
Before a filter press installation is designed with only
polymer conditioning, numerous on-site tests of the
particular sludge and polymer must be conducted.
Not all sludges can be dewatered on a filter press
that uses only polymer conditioning.
Filtrate
A filter press can normally achieve a solids capture of
95 to 99% when conditioned with lime and ferric
chloride. For polymer conditioning, a worst case
capture of 90% can be assumed. Calculation of the
filtrate quantity would be similar to the filtrate
calculation for belt filter presses in Section A.5.1.
Design Example for a 0.88 m3/s (20 mgd) Plant
Sludge Quantity
Sludge Solids = 11,900 kg/d (26,000 Ib/d)
Sludge Volume = 393,000 l/d (104,000 gpd)
Sludge Solids Concentration = 3%
At this plant, assume 16 hr/d, 7 d/wk operation.
(11,900kg/d)(7d/wk) = 83,300 kg/wk
(182,000 Ib/wk)
(393,000 I/d)(7 d/wk) = 2,751,000 l/wk
(728,000 gal/wk)
= 2,751 m3/wk
(3,600 yd3/wk)
Determine Dewatered Sludge Cake Volume
Assume 40% Solids and 1,140 kg/m3 (71 Ib/cu ft)
Cake Volume per Day of Operation
= (8,330 kg/wk)/(7 d/wk) (0.40)(1,140 kg/m3)
= 26.1 m3/d (922 cu ft/d)
Determine Size of Filter Press Required
If a typical cycle time is 2.5 hours, then during a 16
hour work day,
(16 hr/d)/(2.5 hr/cycle) = 6.4 cycles/d
(use 6 cycles/d)
Required Filter Volume
= (26.1 m3/d)/( 6 cycles/d)
= 4.3 nr>3(152cu ft)
171
-------
Number of Filter Presses Required
Two filter presses should be available for service for
reliability. Two filter presses of 4.3-m3 volume each
should be installed, although the individual designer
can vary the actual size selected.
One reason for choosing two presses is that it allows
a reduction to 8 hr/d, 7 d/wk operation. Then, if one
filter press is out of service, the 16 hr/d and 7 dy/wk
operation could be used. Also, if the sludge proves
more difficult to dewater than anticipated, there will be
sufficient capacity available. '
A.7.2 Cost Analyses
Cost Analysis for a 0.088 m3/s (2 mgd) Plant
Determine Capital Cost
Use Figure 5-7, Base Capital Cost of Recessed
Plate Filter Press, page 78 of the EPA Handbook (1).
Sludge Volume per Year = 3.80 x 106 gal/yr
Figure 5-7 is based on 8 hr/d, 7 d/wk operation. In
this example, operation will be only 5 days per week.
Adjust annual sludge volume by ratio of 56/40 = 1.4
1.4 x 3.8 Mgalfyr = 5.3 Mgal/yr
From Rgure 5-7 @ 3% solids:
Capital Cost = $300,000 for 1 unit
Appendix A-10 from the EPA Handbook (1) lists the
background assumptions for the cost figures on the
filter press. For total press chamber volumes below
450 ft3, only one filter press is included. Therefore, to
get a cost for the two filter presses (one standby), the
cost for one must be multiplied by 2. Some reduction
in cost is possible, for example 15 percent, in
construction of two identical units.
(1-0.15) x $600,000 = $510,000
Determine O&M Cost
From Rgure 5-8 and actual sludge volume of 3.8
Mgal/yr,
O&M Cost = $15,000/yr
Cost Analysis for a 0.88 m3/s (20 mgd) Plant
Determine Capital Cost
Use Figure 5-7 of the EPA Handbook (1).
Sludge Volume per Year = 38 x I06 gal/yr
Figure 5-7 is based on 8 hr/day, 7 days per week
operation. In this example, operation will be 16 hr/day,
7 days per week.
Adjust annual sludge volume by 8/16 = 0.50
0.50 x 38 Mgal/yr = 19 Mgal/yr
From Figure 5-7 @ 3% solids,
Capital Cost = $520,000
For this design, two 4.3-m3 (152-cu ft) presses,
the cost would be calculated as for the 0.088-m3/s
plant:
(1-0.15) x $1,040,000 = $880,000
Determine O&M Cost
From Figure 5-8 and actual sludge volume of 38
Mgal/yr,
O&M Cost = $58,000/yr
A.8 References
1. Handbook: Estimating Sludge Management Costs,
EPA-625/6-85/010, United States Environmental
Protection Agency, Center for Environmental
Research Information, Cincinnati, OH, p. 540,
1985.
2. Recessed Plate Filter Presses (Design Information
Report). EPA-600/M-86/017, United States
Environmental Protection Agency, Center for
Environmental Research Information, Cincinnati,
OH, 1986.
172
-------
Appendix B
Operation and Maintenance, Mechanical Dewatering Systems
B.I Introduction
This appendix presents operation and maintenance
information for several mechanical dewatering
systems. Sections have been included for belt filter
presses, centrifuges, filter presses, and vacuum
filters. Sample log sheets can be used in the
development of a record-keeping program.
B.2 Belt Filter Presses
Replacement of the filter belts is one of the most
common maintenance items. The main reasons for
failure of the belts are tearing at the clipper seam,
inferior quality belt material, ineffective tracking
systems and/or poor operation and maintenance.
Clipper seam failure usually occurs because of
inferior construction of the seam or sharp edges on
the doctor blades. The seams should be epoxy
coated on each side and doctor blades should be
made of polyethylene instead of metal. Furthermore,
ineffective tracking systems can cause the edges of
the belt to wear and fray. Poorly designed and
maintained high pressure washwater pumps and
spray nozzles can also cause belt wear. The pump
may not deliver sufficient pressure and flow to the
spray nozzles, which in turn may clog. Without an
even flow of water across the belt, it will blind.
Blinding leads to belt wrinkling and creasing, thus
reducing belt life. Because of improvements in belt
and press design, especially in the area of clipper
seam construction and belt tracking and tensioning
systems, belt life has increased tremendously.
Average belt life is about 2,700 running hours with a
range of 400-12,000 running hours (1). Table B-1
(2) shows typical causes of belt wear problems and
possible solutions.
Belt filter presses are probably the most energy
conservative and therefore, the most economical
mechanical dewatering units to operate, since they
have very low power requirements. The average
requirement is about 5.7 kW (8 hp) per meter belt
width, which is considerably lower than other types of
mechanical dewatering equipment. Some models are
as low as 0.8 kW (1 hp) per meter belt width (3).
Staffing requirements are also low. Table B-2 (3)
summarizes typical operator requirements.
Table B-1. Causes and Prevention of Belt Wear
Cause Preventive Measure
1. Inferior belt material
and inaccurate
dimensions
2. Wear at clipper
seam
3. Misalignment of
rollers
4. Belt shifting or
creasing
Purchase only high-quality belts with a
guaranteed life from well-known
manufacturers.
Purchase belts with highly durable seams,
especially those epoxy coated on both
sides. Check doctor blades to ensure that
there are no sharp edges on which the
clipper can get caught and tear.
Check and adjust roller alignment. Ensure
well-functioning tensioning/tracking
systems.
Ensure that sludge is distributed evenly
across the belt. Ensure that belt washing
system is adequate.
Table B-2.
Number of Operators Required Per Belt Filter
Press
Number of Units
1
2
3
4 or more
Number of Operators Per Unit
1.09
0.66
0.36
0.33
In order for a belt filter press or for that matter, any
other type of mechanical dewatering unit to function
effectively, it must be compatible with both upstream
and downstream processes. Often, operational
problems can be reduced simply by specifying the
appropriate equipment and by properly training the
operators of the equipment. Table B-3 (2) lists
typical causes and preventive measures to reduce
operational problems associated with belt filter
presses.
Process control is extremely important to ensure
optimum performance of the dewatering system. It is
important for the operator to keep records of all press
performance parameters. A typical operational log
sheet is shown in Figure B-1. The operator can
determine how well the press is performing from the
information provided by this log. Sample points must
be located at various places throughout the system.
At least once per shift, a sample should be taken of
173
-------
Tablo B-3. Causes and Prevention of Operational Problems of Beit Filter Presses
Cause
Preventive Measure
1. Improperly conditioned sludge due lo:
a Varying characteristics of sludge feed
b. Improper chemical selection or dosage rate
a Improper point of application.
2. Insufficient gravity drainage of sludge
3. Loss of sludge from between belts
4. Poor housekeeping
5. Poor safoly practices, including:
a. Removal of spray and other equipment guards to facilitate
operation
b. Doacttvation of trip-wire and limit switches to facilitate
access to unit.
1. Improve conditioning practices by:
a. Assuring continuity of sludge through blending of sludge prior to
dewalering
b. Polymers should be carefully selected and tested. Selection and
dosage rate should be checked frequently, particularly when
changes in sludge characteristics are expected
c. The point al which polymer is applied should be reviewed and
revised.
2. Evaluate:
a. Press speed/drainage time
b. Polymer selection and sludge conditonihg system.
3. Reduce belt tension
4. Train operators to properly maintain the press area. Provide steam
cleaning equipment lo assist cleanup.
5. Provide safety training and stringently enforce rules to keep safety
equipment in place.
a Design guards for ease of replacement after removal
b. Design safety equipment to minimize interference with operation.
the feed sludge to the press, cake discharge, and
filtrate. However, composite sampling at these three
locations is the best choice. A composite sample will
give a better picture of press performance. Total
solids should be determined on the feed and cake
samples and total suspended solids should be
determined for the filtrate. Standard Methods for the
Examination of Water and Wastewater (4) should be
used as the source for all laboratory protocol. The
press operator should be required to make the
following calculations:
Throughput, kg/hr = Q C x 1,000
where,
Q = flow to press in m3/hr
C = concentration of solids in feed sludge in mg/l
Throughput, Ib/hr = 8.345 Q C
where,
Q = flow to press in gal/hr
C = chemical fraction of total solids in feed
sludge
Capture, %
= 100 x f(%Cake Solids)(%Feed-%Filtrate)]
* [(%Feed Solids)(%Cake-%Filtrate)]
B.3 Solid Bowl Centrifuges
In the operation of centrifuges, the performance
criteria of high TSS recovery, driest cake solids and
high sludge capacity are not mutually compatible. To
achieve the driest cake, the centrifuge solids capacity
may be limited and higher polymer dosages may be
required. While a lower solids t ,;overy will increase
cake solids, this is not an acceptable long-term
operating mode. The fines lost to the centrate will
eventually be recycled until removed from the system
or lost, thus impairing the final effluent.
The general effect of process variables on the cake
solids and recovery is shown in Table B-4. While
temperature is not a common variable available to the
operator, heating of the feed sludge can result in a
significant increase in cake concentration. In some
cases, polymer addition result sin increased moisture
content because of the capture of fines; in other
cases, moisture content is decreased due to higher
structural strength and better liquid expression from
the cake.
Table B-5 shows the machine variables that can
affect solids recovery and cake solids. While higher
bowl speeds can produce a drier cake, at some point
the g forces will be so high that the cake will not
convey out of the pool. Deepening the pool under
these conditions may alleviate the conveying
problems, but the cake will be wetter. However,
reducing the differential speed can effect a solids
"dam" in the centrifuge and thereby still maintain high
solids, with the ability to convey the solids. Generally,
weak structured solids, like alum and activated
sludge, often respond best to lower gravitational
forces. Reducing the differential speed will produce a
drier cake, but at some point it will impair solids
recovery due to the depth of sludge in the pool.
Often, it is possible to produce a drier sludge cake by
increasing the polymer dosage. This may be cost
effective for a combustion operation. For example, if
the cake solids were increased from 20 to 24 percent
TS with $11.00/Mg dry solids ($10.00/ton) additional
174
-------
TYPE OF SLUDGE:
PRESS NUMBER:
IB
m
DATE
TIME
TEST
NO.
FEED
GPM
TOTAL
SOLIDS, %
LBS/HR
POLYMER
TYPE
ML/MIN
PPM
LBS/D.T.
FILTRATE
GPM
PPM
RECOVERY,
%
CAKE
DRY
SOLIDS, %
HP/
PSI
SCREEN
TENSION
WASH
WATER,
PSI
01
•D
"
o
1C
ffl
S>
o
®
13
i
cn
REMARKS:
-------
Tabto B-4. Effoct of Process Variables on Recovery and
Cake Solids*
Process Feed Feed
Variables Rale Consistency Temperature Flocculants
To improve + + +•
recovery
To improve - - +• -, +
cako solids
* The process variables are highly interactive and any evaluation of
a single parameter requires maintaining the other process
variables constant. Machine variables should not by varied until
final optimization.
- sign; reduction in the process variable produces the desired
effect.
+ sign: increase in the process variable produces the desired
effect.
Tablo B-S.
Machine
Variables
Effect of Machine Variables on Recovery and
Cake Solids'
Bowl Rale
Pool Volume
Conveyor
Speed
To improve
recovery
To improve
cake solids
• In each case the TSS recovery is considered to be about 90
percent for comparative evaluation and other factors are held
constant, he,, food rate, feed solids, etc. If cake structural
characteristics are weak, cake solids and recovery may decrease
due to solids slipping back down the beach slope. In that case it
is necessary lo increase polymer dosage to maintain recovery.
- sign: reduction in the machine variable produces the desired
pifeci.
+ sign: increase in the machine variable produces the desired
effect.
polymer cost, the fuel used by the furnace could be
reduced 165-210 i/Mg dry solids (40-50 gal/ton),
saving about $30.00/Mg dry solids ($27.00/ton) in
fuel costs. Further, the furnace solids capacity would
increase 30-40 percent. The operator should
periodically test this condition.
High polymer usage can be due to a number of
factors. Full-scale and laboratory testing can be
conducted to help identify the reasons for the high
polymer dosage. Operators should periodically
evaluate new polymers as sludge properties change
and new products become available. The procedure
suggested for evaluating high polymer dosage and
the resulting cake solids concentration is outlined
below.
There are a number of reasons for excessive polymer
usage. These causes may be evaluated step-by-
step:
1. The polymer is not effective for the specific
sludge.
Only a series of polymer trials can determine the best
polymer. Lab tests can generally be used to sort out
the best two or three. The sludge should appear
granular when it is poured two or three times from
beaker to beaker.
2. The polymer is not fully into solution and only
partially effective.
The polymer should be tested at the normal age
employed, then allow 6-8 hours more before
retesting. The higher the aging concentration, the
slower the polymer molecules go into solution and
unravels. Use lab tests at the normal age solution
concentration and also at half concentration. Test
sludge flocculation at 4 hours and 12 hours.
3. There is a charge on the sludge that must be
neutralized first.
Mix 1 kg/Mg dry solids (2 Ib/ton) of anionic polymer
(dry basis) with 2-3 liters of sludge, so there is
enough treated sludge for several cationic polymer
tests. Wait 10 minutes after mixing in the anionic
polymer, then add increasing dosages (e.g., 6-12-
18 Ib/ton or 3-6-9 kg/Mg) of cationic polymer to
each sludge mixture in turn. Repeat the process with
another batch of sludge, using 2 kg/Mg dry solids (4
Ib/ton) of anionic polymer. Visually compare the floe
structure of these mixtures to a chemically dosed
mixture without anionic polymer. The matrix of these
mixtures is shown below.
Cationic
Polymer
Dosage
(Ib/ton)
Anionic
6
12
18
Polymer
0
x,
X
X
Dosage
2
X
X2
X
(Ib/ton)
4
X
X
*3
Record visual observations using 200 ml sludge in
300 ml beakers and pouring it back and forth three
times. Good polymer treatment will result in solids
granulation and free water. Run tests marked Xj, Xg,
and Xa first to determine where to concentrate
testing.
It is possible that 10-20 kg/Mg dry solids (20-40
Ib/ton) FeCIa will be effective in place of anionic
polymer. It may be a lower cost alternative to anionic
polymer. Corrosive properties of FeCIa must be kept
in mind.
4. The polymer and sludge are not effectively
mixed at the introduction point.
If the full-scale results are the same as the lab tests,
mixing is generally not the problem. When lab tests
indicate lower dosage, inadequate mixing is a
possible cause. Breakup when mixing can be due to
a weak floe or shear, in which case another polymer
should be considered.
5. The centrifuge is overloaded on a solids basis,
which is impairing the clarification efficiency.
Solids overloading can result in high polymer dosage.
If the scroll differential speed is increased and the
176
-------
centrate clears, this indicates that the centrifuge was
exceeding its Beta capacity (solids discharge
capacity) and thus affecting clarification. This is
especially true when the objective is to produce the
maximum cake solids. If the cake becomes too wet at
the higher differential speed, the feed rate must be
reduced and then the differential speed to maximize
cake solids.
The evaluation procedures 1 through 4 above will
isolate problems associated with polymer selection,
sludge characteristics, and polymer handling. Once
the polymer and feeding have been optimized, the
best polymer can be evaluated, using Step 5 to
determine whether a limiting solids capacity is
affecting polymer cost and performance.
The following tests to determine solids capacity are
recommended:
A. Set up centrifuge to operate at a specific
maximum torque setting, auto or manually
controlled, which will be considered 100 percent
capacity,
B. Feed centrifuge 25, 50, 75 and 100 percent of
solids capacity.
C. Adjust polymer dosage for each flow to about the
same effluent clarity-slightly dirty (90-95
percent recovery). Use lab centrifuge spin tests
for accuracy and quick analysis of the probable
recovery.
D. Allow 15 minutes for stabilization and then take
three samples of feed, cake, and centrate at 15,
20 and 25 minutes (after machine conditions have
been established) for composite.
E. Measure:
Feed rate
Feed TSS-keep the same, if possible
Cake TS
Centrate TSS-keep the same, as much as
possible
Polymer rate
Differential speed
F. Plot:
1. Cake solids rate (Ib/hr or kg/hr) vs. cake solids
(% TS)
2. Cake solids (Ib/hr or kg/hr) vs. polymer dosage
(Ib/hr or kg/hr)
3. Recovery (% TSS) vs. feed rate (Ib TSS/hr
and gpm or kg TSS and l/s) - should be
constant if polymer is adjusted to same clarity.
The cake solids (percent TS) will normally decrease
slightly with increasing solids rate (Ib/hr or kg/hr). If
the differential speed is increasing significantly during
the test in order to maintain torque within limits and
centrate quality, then the cake may be significantly
wetter at higher feed solids rate (Ib/hr or kg/hr).
Polymer dosage may also increase to help maintain
solids volume in the centrifuge. If the polymer dosage
is increasing and the cake becoming wetter, then the
solids capacity of the centrifuge has probably been
exceeded.
If there is a significant difference in the TSS recovery,
this can cause the results to be misleading. Recovery
should be maintained in the range of 90-95 percent
of the feed solids.
To fully evaluate the effect of auto torque control,
conduct the following tests. Based on the earlier test,
use a feed rate that indicates that the unit is within
+ 10 percent of its full capacity. Operate the
differential speed manually at 25, 50, 75 and 100
percent of full torque.
A. Adjust chemical dosage as necessary.
B. Adjust eddy current or auto torque backdrive as
necessary.
C. Operate for 15 minutes for stabilization and
sample at 15, 20 and 25 minutes for composite.
D. Determine:
1. Differential speed, rpm
2. Feed rate, gpm and Ib TSS/hr or l/s and kg
TSS/hr
3. Polymer rate, gpm and Ib/ton dry solids or l/s
and kg/Mg dry solids
E. Plot:
1. Differential speed (rpm) vs. cake solids (%
TSS)
2. Differential speed (rpm) vs. polymer (Ib/ton or
kg/Mg)
3. Polymer dosage vs. cake solids (%) and cake
rate (Ib TS/hr or kg TS/hr)
This test series should be run on the same feed
concentration and recovery as much as possible so
that solids discharge rate does not affect results.
These results will define the optimum operation and
determine the benefits available from the addition of
an automatic torque-controlled backdrive. These
optimization tests should be run periodically, even
when the machine has an auto backdrive.
177
-------
6.3.7 Record-keeping
Good record-keeping helps to detect long-term
changes in the performance of the centrifuge.
Performance differences could be due to the wear of
the scroll, changes in sludge characteristics that
require changes in bowl speed, improper pond
setting, inappropriate bowl/scroll differential speed,
etc. Performance fluctuations should lead to an
optimization study of the centrifuge. As indicated
earlier, polymer should also be continually evaluated.
Communication with other centrifuge operators will
also help to keep abreast of new developments.
Data log sheets should be maintained at the
centrifuge and periodically checked to determine
changes in the performance characteristics of cake
solids, recovery, solids capacity, and polymer dosage.
Suggested data log sheets for daily and monthly
operation of the centrifuge are shown in Figures B-2
and B-3. Representative samples should be
collected and analyzed daily. This data, upon review
could provide the basic insight necessary to optimize
the operation.
On a regular basis, long-term trend lines should be
plotted for the following parameters:
Feed Rate, Ib TSS/hr or kg TSS/hr
Solids Concentration, % TS
Polymer Dosage, Ib/ton TSS or kg/Mg TSS
Solids Recovery, %
An example is shown in Rgure B-4. This trend log
should be posted and updated weekly in an area
visible to the operators.
As with any piece of machinery, the maintenance
program is an essential part of a successful
operation. Observations, such as an increase in
vibration, could lead to a planned overhaul, thus
avoiding a damaging breakdown. Reviewing and
understanding the manufacturer's O&M manual will
help lower O&M costs, reducing downtime and
improving performance.
B.4 Filter Presses
Rlter presses operate in a discontinuous or fill-and-
dump mode. Since this mode of operation varies
substantially with time, it imposes some special
requirements when evaluating performance and
developing effective operating patterns.
For example, it is not possible to take grab samples
at any time during the operation on the dewatered or
cake side of the operation and obtain any genuinely
valuable information about performance. As a
consequence, to obtain a full description of machine
operation, a specific, directed sampling program must
be created. Ideally, one would want to determine the
volume of filtrate, which could be used to confirm the
cake solids for any particular solids and total volume
input to the machine.
B.4.1 The Sludge Conditioning System
At the onset of any performance evaluation or prior to
startup, one must recognize the process key that
controls all performance evaluations dealing with
domestic wastewater sludges is proper sludge
conditioning.
Many sludge conditioning problems do not derive
directly from planned conditioning techniques but
from failure of the ancillary or support equipment.
This is particularly true when using ferric chloride (or
sulfate) and lime to condition sludges, although
polymer conditioning systems can and do have major
maintenance problems. The purpose of the sludge
conditioning system is to promote water release,
increase cake solids, to permit a high solids recovery
(generally greater than 98 percent) and to enhance
cake removal. The conditioning system may involve
the feeding of iron or aluminum salts, lime, and/or
polymer; the delivery of a precoat material such as
diatomaceous earth or incinerator ash; and/or or the
addition of a body feed. All or any one or more of
these materials may be employed on a particular
sludge.
The management of the lime feed storage and
preparation system is tedious, difficult, and
demanding. Yet, when lime is employed, it is
absolutely essential that it be available at the mixing
tank. The operator should determine the availability of
lime and the functioning capability of the lime feed
system on a routine basis. Designers are urged to
assure the easy maintenance and operation of lime
receiving, storage, slaking, dissolution, and delivery
systems. The very nature of hydraled lime slurries
requires special attention. The engineer should design
a system that the operator can rely on to function
properly.
Too little lime means the sludge usually is improperly
conditioned. Too much lime increases the rate at
which filter cloths and plates become fouled with
calcium hydroxide and calcium carbonate, that is,
scale. Thus, excessive lime conditioning increases
the frequency of acid washing, leading to increased
downtime.
Sludge aging should be minimized, as should be the
storage of sludge for any substantial period between
the wet-end and the solids handling train or in the
sludge handling train. Varying the holding time, for
example, from one day to three days, will adversely
effect the ability to condition the sludge and will
increase the chemical requirements, particularly of
polymers but also of iron salts and lime. Ferric
chloride and lime are usually selected as the
conditioning chemicals when complete control over
178
-------
DAILY CENTRIFUGE OPERATOR LOG
-4
CO
DATE;
TIME
2:00
4:00
6:00
8:00
10:00
12:00
14:00
16:00
18:00
20:00
22:00
24:00
MACK. NO.
— .
RATE
GPM
% TSS
% TSS
ml/50 ml
CAKE
% TSS
POLY FEED
SE1TING spm
H^O
?tpin
SAMPLES
TAKEN
TORQUE
FT-LB
DIFF SPEED
rpra
COMMENTS
a>
•a
5
a.
to
o
03
-I
r*
"31
ID
Si
FEEDTOTALIZER
END
START
POLYMER TOTALIZER
END
START
POLYMER TYPE
OPERATORS
2.
TOTAL
TOTAL
-------
CO
o
MONTH:
YEAR:
MONTHLY CEHTRIFOGE PBRFORHAHCE LOG
PLANT: "
DATE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
TOTAL
MIN.
AVG
MAX
HACII.
NUMBER
FEED SOLIDS
GFH
1 TS8
LIJ/HR
O'FLOW
I TSS
CAKE
I TSS
L8/HR
RECOVERY
% TSS
POLYMER
,8/TON DS
OPERATION
MRS/DAY
COMMENTS
•n
S
i
n
u
CO
la
•o
o
i
o
10
«>
I
I
i
I
to
a
CENTRIFUGE TYPE:_
MODEL:
BOWL SPEED:
POLYMER:
RPM
SLUDGE TYPE
-------
YEARLY PERFORMANCE TREND LOG (SAMPLE)
FEED TSS
Ib/hr
RECOVERY
00
CAKE SOLIDS
% TS
POLYMER
DOSAGE
ID
i
Ui
a
•a
&
•o
-------
the aging and condition of the sludge prior to
conditioning and dewatering is not possible. While
ferric chloride and lime probably remain the primary
choice in this situation, polymers are also being used
quite effectively and at usually a reduced chemical
cost. When using polymers, it is essential that the
amount and kind of polymer required be evaluated on
a routine basis.
The pour test, described elsewhere in this document,
is particularly useful in providing a quick and quite
effective evaluation of the polymer dose requirement.
Generally one test early in the day is sufficient for
those facilities where the sludge feed is relatively
constant. However, in other cases more tests are
needed to ensure adequate conditioning. Generally,
mixed primary and waste activated sludge, or pure
waste activated sludge without other treatment or
conditioning history, will require a cationic polymer in
varying doses. Most often the doses are in the range
of 2.2 to 7.7 kg of active poIymer/Mg of solids
processed (2-7 Ib/ton). In many cases polymer
conditioning will not work unless the polymer is added
directly to the feed sludge entering the press. That is,
if not paced with the varying sludge feed, polymer
conditioning will not work.
Since operating conditions (i.e., the SRT, F/M for the
influent wastes, holding time, temperature, etc.) vary
and may cause conditioning changes, it is necessary
for the operator to keep a log of the polymer dosage
required to reach a given degree of cake solids with a
particular blend or a particular kind of sludge. If the
dosage requirements change dramatically, it is the
operator's responsibility to determine the origin of
these differences. Possible explanations for these
differences are listed below.
• When other sludge conditioning is practiced in the
flowsheet, for example conditioning prior to flotation
thickening, it may alter the effectiveness of
downstream polymer additions, possibly to such a
degree that additional treatment is required. In this
case, often two polymers are required downstream
from the first conditioning step to eliminate the
residual effects of the first polymer.
» When lime, alum, or iron salts are used in the
wet-end of the process (e.g., alum or iron salts is
added in the primary tanks for phosphorus removal
and the primary sludge is separately thickened and
separately pumped to the dewatering operation),
the makeup or blend of primary and waste
activated sludge will change with time. Also, the
quantity of alum or iron will vary in the same
fashion. Sludges that have been treated with large
quantities of either iron or aluminum in the wet-
end of the treatment plant may require two
polymers or, at a minimum, an anionic polymer for
effective conditioning. Often the presence of these
carryover materials will simply alter the quantity of
cationic polymer required. Such wet-end
conditions demand a routine, careful appraisal of
polymer needs.
• The criteria that should be used to determine the
correctness of the conditioning chemicals, whether
inorganic or organic, are: 1) cake solids, 2) solids
recovery, and 3) the rather subjective property of
cake release from the cloth. Good cake release
simply implies that the cake falls away cleanly from
the filtration medium and does not penetrate into
the medium or foul the medium so that continuous
washing is required. Secondary considerations,
such as how fast the filtration rate through a given
cloth deteriorates, may also assume primary
importance, particularly with iron and lime. If
overconditioning is practiced on a long- term
basis, especially with lime, an excessive rate of
cloth blinding can be expected. The cake solids
objective will be a function of the conditioning
method and the blend of primary or waste activated
sludge that is being dewatered. Recovery is a
function of the same variables, but generally is not
a significant factor unless cloths are torn.
If the sludge being dewatered contains a high
fraction of primary sludge (greater than 40
percent), the cake will discharge cleanly even with
marginal or less than optimum conditioning
practices. This is particularly true when using iron
and iron salts and lime, but is also true when using
polymer conditioned sludge. The greatest difficulty
with cake release or cake discharge is usually
encountered in operations dewatering 100 percent
waste activated sludge, which is conditioned with
polymer but without precoat or body feed.
Where the cake tends to be overly sticky or wet, the
following factors should be considered prior to
additional dewatering operations.
• Determine the optimum dose of the polymer being
employed, and determine if the needs of the
conditioning system are being met. For example, is
polymer dosed at the actual feed rate?
• If body feed or precoat is used, check the timing of
the application, quantity, and sequence of
operation. Is there enough feed volume to apply
the precoat evenly? Consult the manufacturer for
further input.
» Routinely check (particularly those plants that have
a history of possibly requiring two polymers) to see
if the condition of the sludge has changed
sufficiently to merit a change in polymer addition,
types of polymer used, or sequence of addition.
* Be sure to follow the manufacturer's
recommendations for mixing and aging of the
polymer.
182
-------
• If a polymer change has been recently introduced,
make sure the active fraction of the new polymer is
the same as that of the previous polymer. For
example, some emulsion polymers are
approximately 50 percent active; others are as low
as 20-25 percent active. Failing to take into
account the activity of the material, by modifying
the calculation for the quantity of material, could
reduce the actual dose to approximately 50 percent
of the desired value, thus causing conditioning
failure. The unit cost in $/ton should be the basis
for selecting a polymer.
B.4.2 Determining Mass Balances
It is possible to make a complete mass balance by
determining all of the parameters that may impact the
quality of the cake or filtrate. In Figure B-5, a
material balance for a filter press test is shown.
Often, it is not practical to try to measure all of the
filtrate volume, particularly with full-sized units,
because of the quantity and relative inaccessibility of
the filtrate system. Indeed, total quantity involved may
make it impractical to capture all of the filtrate
material.
It is possible to make a reasonable material balance
without knowing the filtrate volume if a good
composite of samples has been taken during the
dewatering cycle to determine the solids
concentration in the filtrate. If the filtrate is clear
enough, one may assume 100 precent recovery.
Generally, if the recovery is in excess of 98 or 99
percent, the mass balance calculations employing 100
percent recovery will not be seriously or adversely
affected. In this case, one assumes that all of the
cake and precoat solids are discharged with the cake
and the loss to the filtrate is zero.
The following equations allow the designer to
determine the mass balance quantities without
determining the actual quantity or volume of filtrate.
Feed solids rate = 5 Qp F
where,
Qp = feed rate, gal/min
F = feed solids, % TSS
where,
R
Cs
Cc
R = 100 [(Cs/F)l [(F-CC)/(CS-CC)]
= recovery, % TSS
= cake solids, % TSS (or TS)
= feed solids, % TSS
= centrate or filtrate solids, % TSS
B.4.3 Startup
At startup, sufficient jar tests should have been done
to determine the optimum or near optimum chemical
conditioning mode and the chemical requirements. It
is essential that the feed of the required chemicals be
the same as used in the calculations and the mixing
be adequate to assure proper distribution of the
polymer or other conditioning chemicals. It is much
easier to achieve good mixing in a 1.0-1 beaker than
in a line carrying 12.6-126.2 l/s (200-2,000 gpm) of
4 percent sludge. If two chemicals are employed,
such as ferric chloride and lime, or ferric chloride and
a polymer, the ferric chloride should be applied first
and at least 10 seconds of mixing time should be
permitted prior to the addition of the second
conditioning chemical. During startup, particular
attention should be paid to: 1) flocculation and water
release, 2) recovery, 3) cake solids, and 4) cake
release, the primary objectives of the conditioning.
Cake release is one of those properties that is difficult
to measure in a bench-scale test. It is possible in
bench-scale work to determine whether cake
release is quick, easy, and clean, or difficult.
However, in the ranges between difficult and easy, it
is hard to make anything other than a subjective kind
of judgment. As the operator gains experience, his
judgment will become more accurate. Usually cake
release optimization must await full-scale operation,
although broad range estimates of chemical
requirements can be made on the basis of bench-
scale tests.
Poor cake release results from the development of
inadequate resistance to shear in the cake and, thus
penetration of the cake into the pores of the filter
medium. This usually occurs with a sloppy and
inadequately dewatered cake. However, in some
instances, it may occur with a cake that has been
dewatered as well as possible, indicating the need to
change conditioning procedures. If a sticky cake is
encountered and chemical conditioning with ferric
chloride and lime has been employed, the cake
release may be enhanced or improved through the
increase in the quantity of either iron, lime or both. It
should be noted that iron is the coagulant and primary
flocculant when this pair of chemicals is employed.
Lime is used for flocculation, pH correction, and, at
high lime doses, for structural stability.
When poor cake release is experienced with polymer
conditioning, there are a variety of possible solutions
that need to be examined, including:
« Alter polymer dose.
• Change the polymer to a similar but different type
- that is, from a high molecular weight cationic to
a medium molecular weight cationic.
• Examine the use of a multi-polymer system.
» Examine the use of ferric chloride and lime.
» Examine the use of body feeds or precoats.
• Extend the press time to reduce capacity.
183
-------
CO
-u
V
TS
TSS
v;8;
V
TS
TSS
V
Fe (OH),
Cl
V
TS
TDS
TSS
TS
TVS
TSS
TVSS
NVSS
10 %
Ib
Ib
Ib
llg) %TS
Ib
Ih
1 © %
Ib ITSS)
IK rrn«;i
I
% (S % Vol
% @ % Vnl
% © % Vol
Ib
Ib
Ib
Ib
Ib
* Umo as Ca (OHlj
v la %TS „ ...
_c /- solids
i a ID Balances
TS Ih
TVS Ib
TSS to
TVSS Ih
NVSS Ib
V 1 ffl % TS
TS Ib
TSS Ib
'
V 1
Filter
^ Press
V 1 to)
TDS % <5> % Vol
TSS % m % voi
,, TS Ib
' TVS Ib
Ciik» • TSS Ib
TVSS Ib
Wet Wt Ib @ % TS NVSS Ib
TS Ib
TVS Ib
TSS Ib
TVSS _ Ib
NVSS Ib
•n
lET
i
o
Q
19
Ifl
Test No. .
Date
-------
Actually, precoats and body feeds may be employed
with ferric chloride and lime to increase their
resistance to shear and to limit the penetration into
the filtration medium. If a high volatiles solids contenl
is desired, polymers should be employed because of
the negligible alteration in the volatile fraction that
they present. If sticking problems still persist after
optimization of the polymer choice, it is likely precoat
or a limited quantity of body feed or both may be the
next best choice to mainlain a high volatile fraction
and improve cake release. If these fail, then it may be
necessary to use inorganics to bring about the
desired cake properties, p
As indicated earlier, record-keeping is absolutely
essential for all operations where conditioning is
required, and the rate of dewatering is an important
consideration. A sample data sheet is provided in
Figure B-6.
B.5 Vacuum Filters
This section will consider the operation of both drum
filters and belt filters. These are the two common
types of vacuum filters employed in the dewatering of
domestic sludges. The differences between these two
types of filters are outlined below.
• Drum filters have the cloth on the surface of the
drum 360°.
• Drum filters usually employ doctor blades or a
similar device to scrape the sludge from the
surface of the filter medium.
» Belt filters have equipment for removing the filter
medium from the surface of the drum immediately
prior to discharge. Doctor blades, discharge rolls,
or similar discharge devices can be employed. The
removal of the belt from the filter drum also
facilitates washing. The belt filters can be
continuously washed in the area after cake
discharge and prior to the cake form part of the
cycle.
B.5.1 Sludge Conditioning System
For years, primary sludge and then primary plus
trickling filter sludges were dewatered on vacuum
filters using ferric chloride and lime. These operations
were generally successful and produced an excellent
cake that was manageable as a dry solid. However,
since these operations used substantial
concentrations of ferric chloride and lime, the mass of
sludge was increased approximately 15 to 25 percent.
Most new plants, existing plants, and plant
modifications contemplate the use of polymer
conditioning systems. Polymers can be successfully
employed for the conditioning of 100 percent waste
activated sludge on either drum- or belt-type filters.
However, certain changes are required in the filter
operation. Generally, operators try to achieve a final
cake thickness in the range of 0.6-1.3 cm (1/4 to 1/2
in) when dewatering on a vacuum filter. This
thickness is often associated with heat-treated
sludges or sludges conditioned with ferric chloride
and lime, or other agents for bulking in addition to
lime.
There is no solids matrix to support the compressible
waste activated sludge and to provide a channel for
movement of water to the filter medium when
conditioning with lime. Therefore, sludges high in
biological material, typically 75-100 percent waste
activated sludge, tend to collapse and lose their
structure early in the filtering operation, leaving the
outer portion of the cake extremely moist and
resulting in a typical non-dischargeable filter cake.
To obtain a drier cake and a better cake release,
several modifications in the filter operation must
accompany the use of a polymer conditioning system:
• Generally, the cycle time must be shortened to
effectively decrease the form time, hence cake
thickness.
• If the cycle time cannot be shortened, the form
time should be decreased by altering the extent of
immersion of the vacuum filter in the sludge slurry.
In this case, it may be necessary to change the
bridging.
• It is almost always essential to increase the rate of
vat agitation to avoid the problem of separation of
solid and liquid phases in the vat. Further, it is
often necessary to shear the larger floes to restrict
cake thickness.
These changes will generally result in a final cake
thickness in the. range of 0.3-0.6 crn (1/8 to 1/4 in).
However, this cake will crack at the discharge roll of a
belt filter and will normally be removed easily by a
doctor blade on a drum filter. The cake is no longer
too thick to collapse its own structure and it retains
the water in the surface of the cake during the drying
portion of the cycle. This moisture causes resistance
to discharge by virtue of the physical features, that is,
the sticky character of the cake and the penetration
into the filter medium.
fi.5.2 Determining Mass Balances
Mass balances may be accomplished by using the
same kind of mass balance sheet and information
provided in Section B.4 on filter presses. The
operator should develop a mass balance sheet
specifically for the type of vacuum filter used. Figure
B-7 provides an example of a mass balance sheet.
Operating data sheets should contain full information
on conditioning techniques and operational history of
both the wet-end and dewatering system. An
example of a data sheet is shown in Figure B-8.
185
-------
Figure B-6, Sample test data sheet tor filter press.
PRESSURE FILTRATION DATA
TEST
Chamber Area, ft
Cake Thickness, mm
Final:
Initial:
Dry Solids Density, g/cc:
Liquid Density, g/cc:
Time
minutes
1.
3.
5.
7.
9.
11.
13.
15.
17,
19.
Filtrate pH:
Pressure
Volume
Test No:
Test Date:
Slurry Solids, wt%:
Data:
g-ml
Cake Weight: Total
Wet:
Dry:
Part
Tare
Line // of Constant Pressure:
Output: Metric
English
Time
minutes
Pressure
Volume
2.
4.
6.
8.
10.
12,
14.
16.
IS.
20.
Filtrate Solids, mg/1:
CHEMICAL ADDITION:
FeCl3
CaOH2
Polymer
Other
186
-------
Figure B-7. Material balance - vacuum filter test.
Polymer
V.
TS.
TSS.
VSS.
@
.Ib
Jb
. Ib
Body Feed
V.
TS.
TSS.
TS
Jb
. Ib
Fed,
— y.
Fe (OH)3.
Cl- .
. Ib (TSS)
. Ib !TDS)
Feed Sludge
V.
TS.
TSS.
Lime as Ca (OH)2
. I @
. Ib
. Ib
Solids
Balances
Input-Outlet
_% TS
TS.
TVS.
TSS.
TVSS.
NVSS.
. Ib
. Ib
. Ib
. ib
. Ib
Precoat
V
TS.
TSS.
@
TS
. Ib
. Ib
V.
TS.
TDS.
TSS.
TS.
TVS.
TSS.
TVSS.
NVSS.
_% Vol
_% Vol
_% Vol
. ib
. Ib
Ib
Jb
. Ib
•*- Precoat H20 •
Cake
Wet Wt
TS.
TVS
TSS.
TVSS.
NVSS.
Jb @
Ib
Jb
Ib
Jb
Jb
i TS
V.
TS.
TDS.
TSS.
TS.
TVS.
TSS
TVSS.
NVSS.
Filtrate
-I @ -
_% © -
-% ® -
_% © .
_lb
_lb
_lb
_lb
_lb
_% Vol
_% Vol
_% Vol
Test No. .
Date
187
-------
Company
Address
TABLE NO.,
MATERIAL TO BE FILTERED:
...„„„,„% Suspended Solids, Consisting of
% Liquid. Consisting of
VACUUM FILTRATION TEST DATA SHEET
Dale Tested
By
Location
Slurry Temperature °C/'
Slurry pH .,
Specific Gravity of Liquid
% Dissolved Solids in Liquid
'F.
Filter Area Sq. Ft.
Filter Cloth
Slurry Feeding Technique: Bottom Feed rj
Top Feed Q
Air Flow Meter: Gas Meter p
Rotameter rj
Avg. Filtrate Suspended Solids, Mg/L
o
z
i-
:
oc < 5
0. J O
I
Ul
<
>
c
Q
w
IUS£
^U
«
occ
o
AIR
METER
HEADING
FILTRATE
VOLUME.
ML.
WASH
VOLUME.
ML.
CAKE
THICKNESS,
INCHES
CAK6 WEIGHT
GRAMS
WET
DRY
CAKE
MOISTURE,
%
&
3d
U.OT
ffis
<&
-
o
o
X
H
>-
E
O
WET WT. I
+
FILTRATE WT.
W,
DRY
LBS.
PEH
SQ.
FT,
•n
S
i
CO
§
~Z
to
Q.
ti
S
DJ
O
c
c
3
00
00
REMARKS:
-------
The quantity of lime that will be in the solid rather
than the liquid phase can be estimated by subtracting
the estimated dissolved Ca(OH)2 from the total
amount added::
Dissolved Ca(OH)2
FeCIa Neutralization, 70% of Fe dose
Dissolved phase from pH 7.0-10.0, 1.2 g/l
Solid phase - All remaining calcium will be
present as (Ca{OH)2 or CaCOa in the cake
B.6 References
1. Baskerville, R.C. et al. Laboratory Techniques for
Predicting and Evaluating the Performance of a
Filterbelt Press. Filtration and Separation 15{5):445,
1978.
2. Design Information Report on Belt Filter Presses.
U.S. Environmental Protection Agency, Center for
Environmental Research Information, Cincinnati,
OH, 1985.
3. Belt Filter Press Survey Report. American Society
of Civil Engineers, New York, NY, 1985.
4. Standard Methods for the Examination of Water
and Wastewater. American Public Health
Association, New York, NY, 1985.
189
-------
-------
Appendix C
Manufacturers and Sources of Equipment
C.I Belt Filter Press Manufacturers
The list which follows is a compilation of many of the
current (1987) belt filter press manufacturers. This list
does not attempt to include all manufacturers of belt
presses, consequently, some suppliers may not
appear.
Arus Andritz Inc.
Arlington South Industrial Park
1010 Commercial Boulevard
South Arlington, TX 76017
817/465-5611
Ashbrook-Simon-Hartley Co.
11600 East Hardy
Houston, TX 77093
713/449-0322
Ralph B. Carter Co.
192 Atlantic Street
Hackensack, NJ 07602
201/342-3030
EIMCO
Process Equipment Co.
P.O. Box 300
Salt Lake City, UT84110
801/526-2000
Envirex Inc.
A Rexnord Company
1901 South Prairie Avenue
Waukesha. WI53186
414/547-0141
Komline-Sanderson Engineering Corp.
100 Holland Avenue
Peapack, NJ 07977
201/234-1000
Parkson Corp.
P.O. Box 408399
Fort Lauderdale, FL 33340
305/974-6610
Roediger Pittsburgh
R.J. Casey Industrial Park
Columbus & Preble Avenues
Pittsburgh, PA 15233
412/231-7979
C.2 Centrifuge Suppliers
The listing below includes many of the current
suppliers of centrifuges in the United States. Since
there are substantial differences in the bowl sizes and
the gravitational forces for the various machines,
Tables C-1 through C-4, inclusive provide this
background information for three major
manufacturers. The technical information provided in
Section 7.3, used in conjunction with these tables, will
assist in evaluating available and comparable
machines. Capacity cannot generally be assigned to
each machine since it is a function of the feed
characteristics, product requirements, and other
variables set forth in the technical discussions.
Alfa Laval, Inc.
2115 Linwood Avenue
Fort Lee, NJ 07024
201/592-7800
Bird Machine Company, Inc.
100 Neponset Street
South Walpole, MA 02071
615/668-0400
Broadbent Inc.
2684 Gravel Drive
P.O. Box 185249
Fort Worth, TX76118
817/595-2411
Centrico, Inc. (Westfalia)
100 Fairway Court
Northvale, NJ 07647
201/767-3900
Clinton Centrifuge
P.O. Box 217, Dept. B
Hatboro, PA 19040
215/674-2424
Dorr Oliver, Inc.
77 Havemeyer Lane
Stamford, CT 06904
203/358-3200
191
-------
GCl, Inc.
P.O. Box 217
220 Jacksonville Road
Hatboro, PA 19040
215/443-7878
Humboldt-Wedag
3260 Pointe Parkway
Atlanta, GA 30092
404/448-4748
Ingersoll-Rand, Inc. (Kruger)
150 Burke Street
Nashua, NH 03061
603/882-2711
Pennwalt Corporation
Sharpies-Stokes Division
955 Mearns Road
Warminster, PA 18974
215/443-4000
Table C-3.
Tablo C-1. Manufacturer A
Centrifuges
Bowl Siza, mm
460x1,370
610x1,830
610x2,440
760 x 2,440
760 x 3,050
915 X 2.7SO
915 x 3,660
1,120x3,350
1,120x4,470
- Countercurrent
Decanter
Maximum Gravities
3,500
3,000
3,000
3,000
2,100
2,100
2,100
1,980
1,980
Tablo C-2.
Manufacturer B -
Concurrent Design
High-G Centrifuges,
Bowl Sizo, mm
152x353
184x129
356 x 787
356 X 1,257
425x1,257
508 x 1,270
508x1,930
610x1,930
635 x 1,651
635 x 2,286
737 x 2,336
889 x 3,302
1,016x3,556
Maximum Bowl
Speed, rpm
6.000
6,000
4,000
4,000
3,250
3,300
3,300
2,850
3,000
2.700
2.600
2,400
2,000
Maximum Gravities
3,060
3,700
3,180
3,180
2,510
3,090
3,090
2,770
3,190
2,590
2,780
2,860
2,270
Noto: Normal operation on sludge is at less than maximum bowl
speeds shown.
Manufacturer C -
Concurrent Design
Low-G Centrifuges,
Bowl Size, mm
250 x 750
350 x 950
450x1,350
530 x 1,400
530 x 2,200
600 x 1,800
600 x 2,500
900 x 2,500
1,100x3,300
1,400x3,300
1,800x4,400
Maximum Bowl
Speed, rpm
4,400
2,250
2,325
1,965
1,450
2,050
2,050
1,600
1,265
1,100
860 .
Maximum Gravities
2,680
990
1,360
1,440
620
1,410
1,410
1,290
980
950
740
Note: While maximum speeds are shown, best operation may be
at speeds considerably lower than those shown.
Table C-4.
Manufacturer C
Concurrent Design
- Higli-G Centrifuges,
Bowl Size, mm
Maximum Bowl
Speed, rpm
Maximum Gravities
600 x 1,800
600 x 2,500
750 x 2,500
900 x 2,500
1,100x3,300
3,000
2,700
2,500
2,100
1,600
3,020
2,445
2,620
2,630
1,570
Note: While maximum speeds are shown, best operation may be
at speeds considerably lower than those shown.
C.3 Filter Press Suppliers
The following list includes the major manufacturers of
filter presses in the United States. This list does not
attempt to include all suppliers of filter presses,
consequently, some manufacturers may not appear.
Ametek, Inc.
Valley Foundry and Machine Division
251 OS. East Avenue
Fresno, CA 93706
209/233-6135
Clow Corporation,
Box 68
Florence, KY 41042
606/283-2121
Dorr Oliver, Inc.
77 Havemeyer Lane
Stamford, CT 06904
203/358-3200
Duriron Company, Inc.
Box 1145
Dayton, OH 45401
513/226-4000
192
-------
Edwards and Jones
17 Leslie Court
Whippany, NJ 07981
201/428-2828
Eimco RED
Box 300
Salt Lake City, UT84110
801/526-2000
Envirex, Inc.
1901 S. Prairie Avenue
Waukesha, Wl 53186
414/547-0141
Ertel Engineering Company
P.O. Box 3245
Kingston, NY 12401
212/226-6023
Hoesch Industries
Box 461
Wharton, NJ 07885
201/361-4700
Ingersoll-Rand
150 Burke Street
Nashua, NH 03061
603/882-2711
JWI Inc.
2155 112th Avenue
Holland, Mi 49423
616/399-9130
Koppers Environmental
1900 Koppers Bldg.
Pittsburgh, PA 15219
412/227-2000
Kubota, America Environmental
405 Lexington Avenue
New York, NY 10174
212/490-8050
Netzsch, Inc.
119 Pickering Way
Exton, PA 19341
215/363-8010
Passavant Corporation
P.O. Box 2503
Birmingham, AL
205/853-6290
Perrin William, Inc.
432 Monarch Avenue
Dept. B, Ajax
Ontario, CANADA L15 2G7
416/683-9400
R&B Filtration Systems
Division of Buderus Corp.
2211 Newmarket Parkway, Suite 150
Marietta, GA 30067
404/955-9335
Star Systems, Inc.
101 Kershaw Street
P.O. Box 518
Timmonsville, SC 29161
803/346-3101
Sperry, DR, Company
112 North Grant Street
North Aurora, IL 60542
312/892-4361
Treatment Technologies, Inc.
10 Poplar Road
Honey Brook, PA 19344
215/273-2977
*U.S COVWMEKTPRINTINGOFF1CE1 99 3 .750- 00* "so 165
193
-------
-------
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No, G-35
Official Business
Penalty for Private Use, $300
Please make all necessary changes on the above label,
detach or copy, and return to the address in Ihe upper
left-hand corner.
If you do not wish to receive these reports CHECK HERE o;
detach, or copy this cover, and return to the address in the
upper left-hand corner.
EPA/625/1-87/014
------- |