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

-------