United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
EPA-600/2-79-1 23
August 1979
Research and Development
Evaluation of
Dewatering
Devices for
Producing
High-Solids
Sludge  Cake

-------
                RESEARCH  REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was  consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental Health Effects Research
      2.  Environmental Protection Technology
      3.  Ecological Research
      4.  Environmental Monitoring
      5.  Socioeconomic Environmental Studies
      6.  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports  .
      9.  Miscellaneous Reports

This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment,  and  methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology  required for the control and treatment
of pollution sources to meet environmental quality standards.
 This document is available to the public through the National Technical Informa-
 tion Service, Springfield, Virginia 22161.

-------
                                       EPA-600/2-79-123
                                       August  1979
     EVALUATION OF DEWATERING DEVICES
                    FOR
     PRODUCING HIGH-SOLIDS  SLUDGE CAKE
                    by

              Alan F. Cassel
                    and
            Berinda P. Johnson
      District of Columbia Government
   Department of Environmental Services
 Water Resources Management Administration
          Washington, D.C.  20032
          Contract No. 68-03-2455
                                  D-I  Agency

                                  n 1.6.7Q
              Project Officer
            Roland V. Villiers
       Wastewater Research Division
Municipal Environmental Research Laboratory
          Cincinnati, Ohio  A5268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
    OFFICE OF RESEARCH AND DEVELOPMENT
   U.S. ENVIRONMENTAL PROTECTION AGENCY
          CINCINNATI, OHIO  45268

-------
                                 DISCLAIMER

     This report has been reviewed by the Municipal Environmental Research
Laboratory, U. S. Environmental Protection Agency,  and approved for publica-
tion.  Approval does not signify that the contents  necessarily reflect the
views and policies of the U. S. Environmental Protection Agency,  nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
                                     11

-------
                                  FOREWORD

     The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people.  Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.

     Research and development is that necessary first step in problem solu-
tion and it involves defining the problem, measuring its impact, and search-
ing for solutions.  The Municipal Environmental Research Laboratory develops
new and improved technology and systems for the prevention, treatment, and
management of wastewater and solid and hazardous waste pollutant discharges
from municipal and community sources, for the preservation and treatment of
public drinking water supplies, and to minimize the adverse economic, social,
health, and aesthetic effects of pollution.  This publication is one of the
products of that research; a most vital communications link between the
researcher and the user community.

     This report presents the results of a year and a half study of the
capability of various mechanical dewatering devices to produce high solids
sludge cake.  Results show that a number of alternative methods are presently
available that are capable of dewatering municipal sludge solids to 30-40
percent cake dryness.  This is quite significant since it impacts in sub-
sequent cost savings in disposal of sludge solids.  At 30 percent solids,
sludge burns autogenously.  This eliminates the need of costly auxiliary fuel
to incinerate sludge.  Also, high cake solids means less sludge to haul and
land dispose.  This decreases land disposal costs.
                                         Francis T.  Mayo
                                         Director
                                         Municipal Environmental Research
                                           Laboratory,  Cincinnati
                                     111

-------
                                  ABSTRACT

     Pilot-scale dewatering tests were made to establish design and operating
parameters for dewatering municipal wastewater sludges on recessed plate
filter presses (both diaphragm and fixed volume types), continuous belt
presses, and retrofit units for a vacuum filter.  Results from the 1.5-year
study showed that when dewatering lime and ferric chloride-conditioned
sludges, the recessed plate presses consistently produced a 30-40% solids
filter cake.  Feed solids to the units averaged 5% total solids with a range
from 2.4 to 10%.  Various ratios of waste-activated to primary sludge solids,
with emphasis on the 2/1 ratio, were tested.  Successful operation of the
belt presses on the Blue Plains sludge was largely a function of the
percentage of waste activated sludge in the feed mixture.  Cake solids from
25-30% were attained when the polymer conditioning dosage was optimized.
When used as a retrofit device to a vacuum filter, the belt press gave cake
solids in the 30-40% range during laboratory-scale tests.  Full-scale
demonstration, however, was not achieved because an adequate system for
delivering filter cake to a belt filter has not yet been developed.

     Design parameters are developed to dewater a mixture of 67% secondary
and 33% primary sludge in a full-scale plant installation.  The estimated
costs for dewatering plus final disposal by either incineration or
composting are also presented.

     This report was submitted in fulfillment of Contract No.  68-03-2455 by
the Water Resources Management Administration, Department of Environmental
Services, District of Columbia, under the sponsorship of the U.  S. Environ-
mental Protection Agency.
                                     iv

-------
                                  CONTENTS

Foreword	iii
Abstract	iv
Figures	vi
Tables 	 ....... 	   ix
Abbreviations and Symbols	,	xi
Acknowledgements	xiii

     1.  Background and Introduction 	    1
     2.  Summary and Conclusions 	    6
     3.  Recommendations	  .   10
     4.  Type of Sludge Processed	11
     5.  Sludge Conditioning 	   14
     6.  Test Results - Dewatering Units	20
            Diaphragm filter press (variable volume press)  	   20
            Fixed volume filter press	59
            Continuous belt filter press	75
            Vacuum filter retrofit - Envirotech hi-solids filter ....   84
            Vacuum filter	88
     7.  Special Tests	 .	90
            Correlation with specific resistance 	 ....   90
            Dewatering of variable sludge concentrations  	   94
            Material balance 	   96
            Conditioning with polymer	96
            Tests on press cake processing	96
     8.  Process Design	102
            Continuous belt press	102
            Filter press 	  103
            Chemical conditioning	103
            Filter press design	105
            Multiple-hearth incinerator design 	 ....  109
     9.  Dewatering and Disposal Costs	112

Appendices
     A.  Laboratory analyses 	  119
     B.  Data sheets	121
     C.  Determination of specific resistance	138
     D.  Material balance	152
     E.  Derivation of costs	157
     F.  Full-scale unit specifications	174

Glossary	176
                                      v

-------
                                  FIGURES


Number
                                                                  Page
   1          District of Columbia Wastewater Treatment Plant.         2
              Present Facilities

   2          District of Columbia Wastewater Treatment Plant.         3
              Future Facilities

   3          Lime Requirements vs.  Percent Secondary Sludge         19

   4          NGK Diaphragm Press                                    21

   5          NGK Sludge Mix Tank                                    21

   6          NGK Pump Assembly                                      2?

   7          NGK Control Panel                                      22

   8          Process Schematic for  NGK Diaphragm Press              24

   9          Schematic of Filtration  and  Squeezing  in                25
              Diaphragm Press

  10          Schematic of Discharge and Washing  in                   27
              Diaphragm Press

  11          Feed Volume vs. Time!  NGK Runs on  3/4/77               32

  12          Feed Pressure vs.  Time:   NGK Runs on 3/4/77             35

  13          Filtrate  Volume vs. Time:  NGK Runs on  3/4/77           37

  14          Effect  of  Increasing Squeeze  Times  - NGK Press          38

  15          Filtrate  Volume vs. Time:  NGK Runs on  11/1/77          39

  16          Process Flowsheet  for  Continuous Run                    50

  17           Lasta Diaphragm Press                                   52

  18           Schematic of  Filtration, Discharge, and Washing         56
              in the Lasta Press
                                  vi

-------
Numbers
                                                                  Page
  19          Passavant Filter Press                                 61

  20          Sample Sections from Passavant Cake                    61

  21          Nichols Filter Press                                   67

  22          Sample Sections from Nichols Cake                      67

  23          Comparative Yield Data                                 74

  24          Schematic of Parkson Belt Press                        76

  25          Parkson Laboratory Belt Press                          77

  26          Magnum Press Test Results                              78

  27          Magnum Press Test Results                              78

  28          Unimat Belt Press                                      81

  29          Schematic of Envirotech Hi-Solids Filter               84

  30          Chemical Dosages vs.  Percent Secondary Sludge.          86
              Envirotech Tests

  31          Process Yield vs.  Rv                                   91

  32          Process Yield vs.  Rp                                    92

  33          Process Yield vs.  CST/(Percent  Solids  of               93
              Conditioned Feed)

  34          CST/(Percent Solids  of  Conditioned  Feed)                93
              vs. Rv

  35          CST/(Percent Solids  of  Conditioned  Feed)                95
              vs. Rp

  36          Rp vs.  Rv                                              95

  37          Cake from NGK Diaphragm Press                           98

  38          Cake Breaker                                           98

  39          Incinerator  Outlet Temperature vs.                     110
              Percent  Conditioners

C-l          Buchner  Funnel Apparatus                               142
                                   vii

-------
Numbers
                                                                Page
 C-2          Passavant Series 275 Resistance Meter               142




 C-3          CST Instrument                                      145
                                  viii

-------
                                 TABLES


Number                                                          Page

  1          Blue Plains Wastewater Treatment Plant               12
             Operating Parameters

  2          Chemical Specifications                              15

  3          Material Balance Tests for Ca and Fe                 16

  4          Material Balance Tests for Total Solids              16

  5          NGK Filter Performance vs. Chemical                  30
             Conditioning

  6          Runs to Optimize Pumping Time                        33

  7          Extended Runs - 3/8/77                               41

  8          Filtrate Quality vs.  Chemical Conditioning           41

  9          NGK Filter Cloths                                    42

 10          NGK Runs on 2/1 Sludge                               45

 11          Typical Results on Diaphragm Press  -                 48
             August  Runs

 12          Lasta Runs on 2/1 Sludge                             54

 13          Comparison Runs on 2/1 Sludge                        58

 14          Typical Results on Model 2400 High-Pressure Press     62
             (38 mm  Plate)  - August Runs

 15          Runs on Model  2400 High-Pressure  Press  with 2/1       64
             Secondary/Primary Sludge

 16          Runs on Model  600 High-Pressure Press                 65

 17          Typical Results on Low-Pressure Press -              68
             August  Runs
                                   ix

-------
Number                                                           Page

  18          Runs on Low-Pressure Press with 2/1                  70
              Secondary/Primary Sludge

  19          Comparison Runs                                      72

  20          Parkson Press as a Retrofit to Vacuum Filters        79

  21          Unimat Belt Press Results on 2/1 Sludge              82

  22          Unimat Press as a Retrofit to Vacuum Filters         83

  23          Hir-Solids Filter Results                             87

  24          Comparison Runs - Vacuum Filter/Filter Press         89

  25          Dewatering Costs                                    113

  26          Belt Press Costs

  27          Incineration Costs

  28          Land Disposal Costs

  29          Total Disposal Costs                                116

 F-l          Filter Press Specifications                         174

 F-2          Filter Media Specifications                         175
                                   x

-------
                          ABBREVIATIONS AND  SYMBOLS
ABBREVIATIONS

Btu/lb
Cal/gm
cm
cfm
d
ft
g (gm)
gal
hr
in
kg
1
Ib
m
mg/1
ml
mm
MG
MGD
min
min/rev  (MPR)
mo
N/m2
Ns/m2
RPM
sec (s)
wt
SYMBOLS

BOD5
CaC03
Ca(OH)2
CaO
COD
C.ST
F:M
 Btu per pound
 Calorie per  gram
 Centimeter
 Cubic feet per minute
 Day
 Foot  (feet)
 gram
 Gallon
 Gallons per  day
 Gallons per  minute
 Hour
 Inch
 Kilogram
 Liter
 Pound
 Meter
 Milligrams per liter
 Milliliter
 Millimeter
 Million Gallons
 Million gallons per day
 Minute
 Minutes per  revolution
 Month
 Newtons per  square meter
 Newtons-seconds per square meter
 Revolutions  per minute
 Second
 Weight
Five Day Biochemical Oxygen Demand
Calcium Carbonate
Calcium Hydroxide
Calcium Oxide
Chemical Oxygen Demand
Capillary Suction Time, sec
Food to Mass Ratio in Secondary
  Aeration System
                                    xi

-------
 FeCls
 Fe(OH)3
 Hg (Inches Hg)

 MLSS
 MLVSS
 mm H20
 NH3
 NC>3
 P
 PH
 P04
 psig
AP
 Rp

 Rv

 SRT
 TKN
 TS
 USDA

 VS
Ferric Chloride
Ferric Hydroxide
Mercury (Inches of Mercury Pressure-
  Gage)
Mixed Liquor Suspended Solids
Mixed Liquor Volatile Suspended Solids
Millimeters of Water Pressure-Gage
Ammonia
Nitrate
Total Phosphrous
Hydrogen Ion Concentration
Phosphate
Pounds per square Inch-Gage Pressure
Pressure Drop
Specific Resistance to Filtration,
  dimensionless (pressure)
Specific Resistance to Filtration,
  cm/g (vacuum)
Sludge Retention Time
Total Kjeldahl Nitrogen
Total Solids
United States Department of Agricul-
  ture
Volatile Solids
Percent
Pound
Viscosity of water
                                     xii

-------
                              ACKNOWLEDGEMENTS

     We would like to thank all the people whose dedication and cooperation
contributed to the success of the project.  We would like to extend a
special thank you to the staff of operators - Roger Benfield, leader, Jerry
Ballengee, Dave Willhite, and Mark George - for the quality of data they
collected; to Felix Costanzo for assembling and maintaining the various
test units; to Bill Ruby for laboratory analyses; and to Marco Garcia and
Pete Repak for the special tests run throughout the course of the study.

     Particular appreciation is extended to all the manufacturers who
provided their time and equipment to the study.  We are deeply indebted to
Dr. James E. Smith, Jr., of the EPA in Cincinnati, Ohio, who was instru-
mental in bringing this project into being and providing technical assist-
ance in getting the study underway.  Thanks are also due to Mr. R.V.  Villiers,
project officer, and Dr. J. B. Farrell, with the Ultimate Disposal Section
of EPA's Municipal Environmental Research Laboratory in Cincinnati, Ohio.
                                     xiii

-------
                                  SECTION 1

                        BACKGROUND AND INTRODUCTION
     The District of Columbia's Wastewater Treatment Plant at Blue Plains
receives flow from the District of Columbia and from suburban Maryland and
Virginia jurisdictions.  Approximately two million residents produce an
average daily flow of 1.06 Mm3/day (280 MGD).  Wastewater is treated by
primary sedimentation and a secondary waste activated sludge process
with chemical addition for solids capture and phosphorus removal.  Sludge
treatment is accomplished by two methods:  gravity thickening and raw sludge
dewatering with subsequent composting or trenching; or gravity thickening,
anaerobic digestion, elutriation and dewatering with subsequent land
spreading.  (See Figure 1.)

     An on-going expansion and upgrading of the plant, with completion
scheduled for mid 1980, will add nitrification and multi-media filtration
to the wastewater train.  Sludge production will increase from its current
level of approximately 150,000 kg/day dry solids (165 dry tons per day)
to 340,000 kg/day solids (374 dry tons per day).  To handle the additional
sludge quantities, original plans had called for gravity thickening of
primary sludge, air flotation thickening of all waste activated sludges,
blending, vacuum filtration, and incineration.  (See Figure 2.)  All units
except the incinerators have been installed.  Because of the large amount
of fuel oil which would be required to incinerate the vacuum filtered
sludge cake, approval of the incinerators has been deferred by EPA pending
further study.  An initial study conducted by Camp, Dresser and McKee,
Inc.  recommended a dual disposal system of composting and incineration
and pointed out that if a high-solids sludge cake were produced, incineration
could be accomplished with minimum quantities of auxiliary fuel.  The study
estimated that to incinerate 374 dry tons per day of an 18% vacuum filter
cake, annual fuel oil usage would be approximately 16,000,000 gallons.
Whereas, to incinerate a 35% solids cake, fuel usage would decrease to only
5,500,000 gallons per year.  This 35% solids cake would be autocombustible
and fuel would be required primarily for the afterburner to control toxic
organics in the off-gas.  Fuel usage in the multiple hearth furnace itself
    Camp, Dresser & McKee, Inc. Alternative Sludge Disposal Systems for the
    District of Columbia Water Pollution Control Plant at Blue Plains.
    December, 1975.

-------
Aerated
* Grit Chambei

Primary
r
Sedimentation


Sludge
Thickening






^^ uigestic


>n



Secondary Secondary



^ Vacuum
Filtration
Aeration Clorificat o
f

Vacuum
Elutriation — *-
Filtration

Composting
" or
Trenching
i To
River
land
» •
Disposal
Figure 1.  District of Columbia Wastewater Treatment Plant.   Present facilities.

-------
Raw
Wastewater



Prir
^


D
•—
Gravity Flotation
Thickeners Thickeners




..4-
Vacuum
Filtration
"iltration To
isinfection River

Ash
— ^ Incineration • fr to
Disposal
Figure 2.  District of Columbia Wastewater Treatment Plant.   Future facilities.

-------
would be needed only for control of the position of the burning hearth.
Camp, Dresser and McKee, therefore, recommended that the vacuum filters
be replaced by filter presses so that an autocombustible sludge cake would
be produced.

     In mid-1976 the District of Columbia received approval from EPA to
conduct a one year study of alternative dewatering devices for producing a
high-solids sludge cake.  The study was funded from two EPA sources.
EPA Region III allowed expenses up to $186,992 as an addendum to an
existing capital outlay project.  EPA's Municipal Environmental Research
Laboratory in Cincinnati, Ohio provided $49,693 and technical direction
for the study.

     The study was performed in the EPA-DC Pilot Plant using the existing
equipment available in that facility as well as equipment provided by
various manufacturers.  A project engineer, a part-time chemical engineer,
four sewage disposal plant operators, one chemist, and one mechanic conducted
the entire study.  The study period officially commenced on September 1, 1976
and ended on October 31, 1977; some preliminary test work actually began
in April, 1976.  The study was completed with expenses well below the
budgeted amount.

     The purpose of the study was to compare the operation of the various
dewatering devices on selected ratios of waste activated to primary sludges.
Because the plant had already purchased and installed 30 vacuum filter units,
the District was interested only in evaluating devices that would produce
significantly higher solids than the vacuum filters.   Specifically, the
District was interested in the results with a 2:1 ratio of waste activated-
to-primary sludge solids.  All units were operated in an attempt to produce
an auto-combustible sludge cake.  For sludges conditioned with inert
chemicals,  this requires a solids content of approximately 35%.  The type
of units tested and their suppliers included:

     1.   Vacuum Filter - Pilot model owned by EPA.
     2.   Vacuum Filter add-on devices - supplied by

         a)  Envirotech Corporation, Salt Lake City,  Utah.
         fa)  Parkson Corporation, Fort Lauderdale, Florida.
         c)  Komline-Sanderson Corp., Peapack, New Jersey.

     3.   Belt Press - supplied by Parkson and Komline-Sanderson.
     4.   Filter Press - fixed volume @100 psig pressure supplied by
         Neptune-Nichols, Belle Mead, New Jersey.
     5.   Filter Press - fixed volume @225 psig pressure supplied by
         Passavant Corp., Birmingham, Alabama.
     6.   Filter Press - diaphragm type supplied by

         a)  NGK Insulators, Ltd.,  Nagoya, Japan.   Envirex Corporation has
             since purchased the rights to manufacture and market this press
             in the United States.

-------
         b)  Dart Industries, Paramus, New Jersey.
         c)  Ingersoll-Rand Corp., Nashua, New Hampshire.

This report presents the results of testing the above dewatering units.
Chemical conditioning with lime and ferric chloride was examined in detail
to establish the requirements as a function of the ratio of waste activated-
to-primary sludge solids.  A correlation between bench-scale filterability
tests and filter press performance was developed to monitor the conditioning
step.  Polymer conditioning was evaluated as an alternative to lime and
ferric chloride conditioning.

     Detailed design parameters for each of the filter presses were developed
for a 2/1 secondary/primary sludge ratio.  Comparison runs on these presses
with the same batch of sludge provided valuable information on the advan-
tages/disadvantages of each.  Filter press cake was used in a variety of
experiments to test cake shredding, incineration  (solid waste furnace,
multiple hearth furnace, and coal-fired boiler), and composting (static pile
method).

     The belt presses were, used to provide design criteria for the thickened
sludges.  These presses were also modified to function as add-on units to
further dewater vacuum filter cake.

     Capital and operating costs and utility consumption are detailed for
the dewatering units.  Total disposal costs for dewatering plus both
incineration and composting are also presented.

-------
                                SECTION 2

                         SUMMARY AND CONCLUSIONS
Chemical Conditioning

1.  The lime and ferric chloride dosages required to produce a filterable
    sludge varied with the percentage of waste activated sludge.  Fibrous
    primary sludge filtered quite readily; waste activated sludge required
    greater quantities of conditioners and was more difficult to dewater.
    Generally, a 3/1 ratio of lime-to-ferric chloride was optimum for
    conditioning the Blue Plains sludge.

2.  Laboratory tests showed that over-agitation of the conditioned sludge
    was detrimental to the filtration process.  Floe deterioration with
    both time and high shear was a major factor in determining chemical
    requirements.

3.  The addition of lime and ferric chloride to the sludge mixture
    increased the final dry weight of the filter cake by a corresponding
    amount.  All of the iron and 80% of the .calcium exited with the cake
    solids during filtration operations.

4.  Bench-scale filterability tests were found to be useful when optimizing
    and controlling the lime and ferric chloride dosages.

5.  Polymer conditioning of the 2/1 mixture of secondary-to-primary sludge
    was generally ineffectual.   No single polymer was found which could
    adjust to the daily variations in the quality of sludge received from
    the primary and secondary treatment processes.

Filter Press-General

1.  Each of the filter presses  was capable of dewatering all sludge ratios
    in the range of 2.4-10% total feed solids to at least a 30% solids
    cake.   The diaphragm press, however, was the only unit capable of
    dewatering the marginally conditioned sludges to the 35% solids
    required for an autocombustible cake.

2.  Once a minimum chemical conditioning requirement of lime and ferric
    chloride for adequate dewatering was established, increases in
    filtration yields (up to 20%) were obtained by slight increases in
    chemical dosages.

-------
 3.   In all the presses,  suspended solids  recovery in the filter cake
     was greater than 99%.   The quantity of  suspended solids  in the filtrate
     was affected primarily by the type of filter cloth used  and the degree
     of chemical conditioning.

 4.   The filter presses did not satisfactorily dewater polymer  conditioned
     sludges.

 5.   The average specific resistance-to-filtration parameter  was correlated
     directly  with filter press yield.

 Filter Press-Diaphragm Unit

 1.   On the average,  the  NGK press,  using  conditioning of  19.6%  lime and
     6.5% FeCl-j,  dewatered  the  2/1 secondary-to-primary sludge  to a
     38.7% solids cake with a yield  of  2.39  kg/hr/m2  (0.49 Ib/hr/ft2).
     The pumping  pressure required to feed the press  was  always  less than
     7  kg/cm2  (100 psig).   The  pumping  cycle time  averaged 17 minutes and
     was controlled by monitoring  the total  solids feed rate.  A squeezing
     pressure  of  15.0 kg/cm2 (213  psig) was  generally used.   The squeezing
     cycle time  (18 minutes) was controlled  by filtering  to a specified
     filtrate  flow rate.

 2.   Equivalent results were obtained on the Ingersoll-Rand Lasta press.
     The full-scale yield for the  unit, however, was  somewhat higher at
     2.93  kg/hr/m2  (0.60  lb/hr/ft2)  for the  2/1 secondary-to-primary sludge.

 3.   Different filter cloths were  tested on both the  NGK and Lasta  units.
     All  gave  acceptable  filtrate  quality  but cloth  life, resistance to
     abrasion, etc.,  could not be  effectively  evaluated in our study.

 4.   The  cloth washing system in each of the presses  also could not be
     adequately evaluated during the study.  Maintenance of satisfactory
     cloth permeability by high-pressure sprays or acid washing  is  an
     area  that generally requires more study.

 Filter Press - Fixed Volume Unit

 1.  The high-pressure press (225 psig)  had an average filtration yield
    of 1.51 kg/hr/m2 (.31 Ib/hr/ft2) and required 62.3% more filtration
    area than the NGK diaphragm unit to produce equivalent results.  The
    low-pressure press (100 psig)  had an average full-scale yield  of
    1.07 kg/hr/m2  (.22 lb/hr/ftz)  and needed 126.8% more filter area than
    the NGK diaphragm press to produce equivalent results.

2.  Cycle time on the presses  averaged  2-3 hours and was determined by
    filtering to a specified filtrate flow rate.

3.  The cakes from the fixed volume presses  always contained  a dry outer
    section and a wetter  inner core.  This resulted in a substantial
    variation in the solids content across the cake.

-------
 Continuous Belt Press

 1.  Because of the highly variable sludge at Blue Plains, no polymer was
     found that could adjust to these variations and adequately condition
     the sludge at all times.  The operation of the belt press, therefore,
     was not consistent.

 2.  With thickened sludge feeds, the press capacity, final cake solids,
     and polymer consumption were all affected by the percentage of waste-
     activated sludge.  The unit performed best when dewatering high
     percentages of fibrous, primary sludge.

 3.  Suspended solids recovery in the filter cake averaged only 95%.
     Because of the stringent advanced waste treatment standards at Blue
     Plains, this level of recovery would be insufficient for continuous
     operation at this plant.

Vacuum Filter Retrofit Unit

 1.  The Envirotech Hi-Solids filter was discounted as an option for Blue
     Plains.  It was capable of increasing the cake from a rotary vacuum
     filter to only 25% solids.

 2.  The use of the high-pressure section of the continuous belt press to
     further dewater the vacuum filter cake showed great promise.  Cake
     solids of 35% were achieved in bench-scale work; however, demonstration
     of the system' in a full-scale test was not successful because of
     problems with feeding the vacuum filtered cake to the press.

Filter Cake Processing

 1.  The filter press cake was composted with wood chips by the static-pile
     method.  A good final product was produced with projected costs less
     than those for composting with vacuum filter cake.

 2.  Filter press cake with a solids content of at least 35% is considered
     a low-value fuel.  It will burn in a multiple hearth incinerator
     without auxiliary fuel to produce an exit temperature of 800° F.  It
     can also be co-burned with municipal refuse in a rocking-grate furnace.
     Because of the high ash content of the cake (up to 50%), however, it
     has been rejected as a fuel for a coal-fired boiler.

Economics

 1.  The belt press ($32.39 per ton) and the vacuum filter ($39.10 per ton)
     provide the lowest cost for dewatering.

 2.  Dewatering costs for each of the filter presses are nearly equal
     with unit costs of approximately $55.00 per ton.

-------
Total disposal costs for filter pressing and incineration are
approximately $88 per ton.  This compares to the total cost for
vacuum filtering and incineration at $130 per ton.  Therefore,
savings of nearly $4,000,000 per year for a 250 ton-per-day plant
are possible by selecting filter presses for dewatering.

Total disposal costs for filter pressing and composting (including
the cost of hauling the press cake 25 miles) are approximately $102
per ton.  This compares to the total cost of vacuum filtering and
composting (including hauling) of $155 per ton.  Choosing a filter
press rather than a vacuum filter, therefore, will result in annual
savings of nearly $5,000,000 for a 250 ton-per-day plant.

-------
                                 SECTION 3

                             RECOMMENDATIONS
!•  Filter presses should be installed at Blue Plains, designed to dewater
    the total quantity of sludge (average 374 dry tons per day) to be
    processed by either incineration or composting.  This study showed
    that the diaphragm-type press offers the most flexibility and provides
    the best product.  However, a final decision on the type of press to be
    utilized should be deferred until full-scale facilities of each type
    are inspected.  Regardless of the type of unit chosen, a single
    large-scale unit should be purchased, installed and operated for
    several months to provide valuable design information prior to a
    large-scale committment of funds.

2.  Additional test work should also be conducted to determine whether the
    specific resistance parameter can be used to successfully monitor
    and control the chemical conditioning process.  As outlined in Section 8
    (Process Design) a pilot-scale horizontal vacuum filter, adjusted to
    simulate the Buchner funnel filtration test, would be used for this
    purpose.  When a pilot-unit becomes available, this work can be carried
    out in conjunction with existing vacuum filter operations.
                                    10

-------
                                  SECTION 4

                          TYPE OF SLUDGE PROCESSED
     During the study period, the wastewater treatment system included degrit-
ting, primary sedimentation, and a high-rate waste activated secondary pro-
cess, with chemical addition for phosphorus removal.  Table 1 shows average
operating parameters of the system for FY 1977  (October, 1976 through Septem-
ber, 1977).  The primary and secondary sludges are blended in a thickening
operation and a portion of this undigested sludge is dewatered on vacuum
filters.

     Throughout the study period, the plant experienced continual operating
problems because of an overloaded sludge processing system.  Recycle loads,
especially from gravity thickening, created operating difficulties in secon-
dary.  The recycle flows (only 5% of total flow) contributed 22% of the BOD
loading and 25% of the suspended solids loading to the wastewater treatment
train.  This recycle problem, together with normal operating problems expe-
rienced with chemical addition outages and the rapidly changing biology in
secondary, caused a highly variable sludge product.  The combined sewer system
in the District of Columbia also contributed to the problem; heavy rains
washed large quantities of solids into the primary sludge thus changing the
character of that product.   As a result, the sludge dewaterability varied
daily.

     In setting up the study, the engineers attempted to simulate the condi-
tions that would exist in the future full-scale plant.  When the systems
are completed, all primary sludge will be gravity thickened separately.  The
sludges from secondary and nitrification will be combined and air float
thickened.  All chemical precipitate will be included with the waste activated
sludges.  Backwash water and solids from the multi-media filters will be
returned to secondary.   Calculations of future sludge production show that
the plant will produce an average ratio of 33% primary solids/67% waste acti-
vated solids.

     The pilot plant had the capability to receive either blended thickened
sludge from the plant or primary and secondary sludges separately.   Initial
test work with the plant thickened sludge gave good filtration results; how-
ever, the overloaded plant  thickeners tended to wash out the fine solids.   In
order to get a more realistic product for dewatering, separate pilot-scale
gravity thickeners for the  primary and secondary sludges were placed in opera-
tion.  A major variable for study was the dewaterability of various ratios of
secondary to primary sludges.   Consequently,  the separate gravity thickening
systems were used and the two sludges blended as necessary on a dry solids
                                      11

-------
TABLE 1.  BLUE PLAINS WASTEWATER TREATMENT PLANT OPERATING PARAMETERS
Primary Treatment
Flow, Mm -Vday (MGD) including plant recycle
Influent suspended solids, mg/1
Influent BOD5, mg/1
Detention time, hours
Surface Loading Rate, m3/day/m2 (gpd/ft2)
Suspended solids removal, %
BOD5 removal, %
Sludge Production, kg/Mm3(lb/MG)
Sludge wasted, % total solids
Sludge wasted, % volatile solids
Mean
1.1 (292)
183
186
2.6
42.3 (1039)
45.0
37.0
78,960 (657.5)
0.525
72.4
Secondary Treatment
Aeration Tank
    Flow, Mm3/day (MGD)
    Influent suspended solids, mg/1
    Influent BODs, mg/1
    Influent phosphorus, mg/1 as P
    MLSS/MLVSS concentrations, mg/1
    Detention time, hours
    SRT, days
    F:M, days"1
Sedimentation Tank
    Detention time-, hours
    Surface loading rate, m3/day/m2 (gpd/ft2)
Chemical Addition
    Ferric chloride, mg/1
    Polymer (anionic), mg/1
Process Performance
    Suspended solids removal, %
    BOD5 removal, %
    P removal, %
    Sludge wasted, kg/Mm3  (lb/MG)
    Sludge wasted, % total solids
    Sludge wasted, % volatile solids
    Percent, biological solids/chemical solids
    Fe content of waste sludge, %
                   ate, m3/dav/m2  (gpd/ft2)
                   , kg/day/m*,  (lbs/day/ft2)
                   ciency, /•>
Gravity Thickening
Hydraulic loading rate,
Solids loading rate, kg.
Solids capture efficiency

Vacuum Filtration
Feed, % total solids
Lime addition, % of feed solids
FeCl3 addition, % of feed solids
Cake solids content, %
Filter yield, kg/hr/m2 (lb/hr/ft2}
                                                        1.1 (288)
                                                        102
                                                        121
                                                        6.2
                                                        1297/852
                                                        1.58
                                                        0.64
                                                        1.51

                                                        2.63
                                                        33.1 (813)

                                                        23
                                                        0.20

                                                        72
                                                        77
                                                        63
                                                        118,410  (986)
                                                        1.4
                                                        66.1
                                                        80/20
                                                        10.0
29.9 (734)
118.8 (24.3)
77
                                                         7.0
                                                         21.8
                                                         7.5
                                                         23.2
                                                         15.3 (3.12)
                                      12

-------
basis.  Air  flotation thickening of the secondary sludge would have been the
preferred method.  Unfortunately, the  logistics of running an air  float thick-
ener proved  too cumbersome.  Gravity thickening of the secondary sludge did
produce the  required 4 to 6% solids.

     The primary sludge was delivered  to the pilot plant and gravity thickened
to 6 to 10%  solids for use in the study.  Normally, sludge was delivered
continually  from Monday morning to Friday and wasted as necessary  to keep a
sludge blanket in the thickener.  The  capacity of the thickener far exceeded
the requirements for dewatering.  The  thickener was drained each Friday and
fresh sludge started each Monday.  Because the primary sludge at Blue Plains
is very high in fiber content, a shredder was installed in the primary sludge
delivery line  (0.5% solids stream).  It was used intermittently to keep the
rags and trash from plugging the transfer pumps.

     The secondary waste sludge was pumped directly from the secondary clari-
fiers to a gravity thickener in the pilot plant.  This thickener was used
primarily as a holding tank.  Each evening, sludge was pumped in at a low flow
rate.  The flow was then cut off early in the morning and the contents allowed
to thicken to 4-6% solids content.  Sludge was used from this source during
the day and  the remaining contents drained each afternoon.  Such an operation
kept the sludge as fresh as possible.  By severely limiting the overflow rate
practically all the fines in the secondary sludge thickeners were  captured.

     A typical 2/1 secondary/primary sludge had the following characteristics

          Percent solids                        4 to 6
          pH                                    6.2 to 6.8
          Density, gm/cc (Ib/gal)               1.006 (8.4)
          Temperature, winter °C                8-15
          Temperature, summer °C                25-30
          % iron as Fe                          7
          % volatile solids                     60-65

     For  all test runs, the sludges were blended in the following manner:
Thickened secondary sludge (at 4-6% solids) was pumped with a Moyno pump to
a calibrated mixing tank;   the volume was measured, the sludge density
measured, and a sample analyzed for percent solids on an O'Haus Moisture
Balance.  The quantity of dry solids in the tank was then calculated.
The primary sludge was also analyzed for density and % solids, and the pounds
of primary sludge calculated for a given volume.  Based on the ratio of
secondary to primary solids required, the volume of primary sludge was then
Moyno pumped to the mixing tank.   This method did give some experimental
error;  however, only approximate ratios were required for the type of tests
run.

     When plant thickened sludge was used for testing, it was pumped from the
gravity thickeners to a tank truck (700 gallons) and  then transported  one-
quarter mile to the pilot plant.   A Moyno pump was used to pump the truck
contents to the filter feed tank.
                                      13

-------
                                 SECTION 5

                             SLUDGE CONDITIONING
     Prior to either pressure or vacuum filtration, wastewater sludges must
be chemically conditioned.  A filter press will generally use ferric chloride
and lime for conditioning.  A vacuum filter will use either ferric chloride
and lime or ferric chloride and polymer; belt presses will generally use
polymer conditioning alone.  For the purpose of this study, lime, ferric
chloride and various polymers, either singly or in combination with one
another, were examined for their suitability in conditioning the sludge.  The
study attempted to optimize each of these chemicals for each of the dewatering
units.  Other chemicals, such as aluminum chlorohydrate or ferrous sulfate,
were not tested because they were either too costly or in short supply.

CONDITIONING WITH LIME AND FERRIC CHLORIDE

     The lime used for the study was a bagged, pulverized, high calcium (94%
CaO) quicklime.  Lime dosages are reported as the weight of the lime as pur-
chased.  The ferric chloride used was purchased as a 30% by weight solution
and was diluted as necessary.  Results are reported on a 100% FeCl3 basis.
Table 2 gives the specifications for the lime and  ferric chloride.  Percent
chemicals (either lime or FeCl3) are calculated as:

                Ibs dry weight of chemical	 x 100 = % chemical
                Ibs dry incoming sludge solids

Material Balance Tests

     The addition of lime to the thickened sludge stream is expected to form
calcium carbonate (insoluble) and calcium hydroxide (soluble).  The
quantities normally required will raise the pH of the solution to 11.0 or
above.  Ferric chloride reacts at this high pH to form the insoluble ferric
hydroxide (Fe(OH)3).   The disposition of these metal ions was determined
with a material balance test on a Buchner funnel.   At three different chemical
dosages, approximately 225 ml of conditioned sludge was filtered; the feed,
cake and filtrate were all analyzed for calcium (Ca) and iron (Fe) content.
The Ca and Fe determinations were made with an Atomic Absorption Spectropho-
tometer.  Table 3 shows the test results.   Note that the weight of Ca and Fe
in the feed do not balance exactly with the Ca and Fe in the filtrate and
cake.   However, the tests are useful in that they show that approximately
80% of the calcium and 100% of the iron will exit with the cake. .The
remainder of the calcium stays with the filtrate,  probably as calcium hydrox-
ide.


                                     14

-------
                      TABLE 2.  CHEMICAL SPECIFICATIONS
Lime

     Type

     CaO
     Available CaO
     Size

     Concentration  (as used)
     Specific gravity

Ferric Chloride

     Type

     Concentration  (as used)
     Specific gravity
Pulverized CaO from Warner Co.,
  Beliefonte, Pa.
94.84%
92.1%
100% will pass a 100-mesh sieve
1.95 micron average particle diameter
1 Ib/gallon (11.3% by weight)
1.06
Liquid @ 30% by weight from DuPont
titanium dioxide manufacture
1 lb/gallon (10.9% by weight)
1.103
     Additional tests were run on the Buchner funnel for the purpose of
performing a total solids balance (determined by evaporating the samples to
dryness).  These tests, at different chemical dosages, were carefully run
to determine the increase in dry solids due to the addition of lime and ferric
chloride.  Thickened sludge (primary and waste activated) was conditioned and
approximately 200 ml were filtered.   Weights and total solids of the
unconditioned sludge, the conditioned sludge, the cake and the filtrate were
measured.  The results are shown in Table 4.  The tests show that as chemicals
are added, the weight of solids actually increases above the initial weight
of sludge plus chemicals added.  This is most likely due to CaCOg formation.
The total solids in the final cake,  however, match very closely with the
initial weight of sludge plus the weight of lime and ferric chloride added.
The conclusion we reached from these tests was that both the lime and ferric
chloride weights must be accounted for in the filter cakes off either a
vacuum filter or filter press.  In all calculations we therefore assumed that
for every pound of lime and ferric chloride added for conditioning the final
dry weight of the filter cake also increased by an identical amount.

Important Considerations In Conditioning

     A considerable amount of trial-and-error work on the filter presses
and the bench studies showed the following:

     1.  For Blue Plains sludge, the minimum amount of FeCl3 needed for
         conditioning was approximately 5% by weight of sludge solids.
                                     15

-------
TABLE 3.  MATERIAL BALANCE TESTS FOR Ca AND Fe
FEED SLUDGE
% CHEMICAL
lime/FeCl3
15.6/5.4
20.8/6.9
26.1/8.7

Ca
1.107
1.450
1.604
gms
Fe
.770
.824
.837
FILTRATE
gms
Solids
15.7
16.5
16.3
Ca
.232
.277
.389
Fe
.00012
.00014
.00014
Ca
.865
1.106
1.294
CAKE
gms
Fe
.806
.788
.892


Solids
15.1
15.6
15.7
 TABLE 4.  MATERIAL BALANCE TESTS FOR TOTAL SOLIDS

UNCONDITIONED
SLUDGE
gms +
25.68
23.93
23.51
20.36
19.74


LIME
gms +
4.08
5.09
6.30
3.07
4.06


FeCl3
gms
1.36
1.70
2.08
1.02
1.36

TOTAL SOLIDS
COMPUTED IN FEED
gms
31.12
30.72
31.89
24.45
25.16 -
TOTAL SOLIDS
MEASURED IN
CONDITIONED SLUDGE
gms
32.67
32.67
33.77
25.67
26.69
TOTAL SOLIDS
MEASURED IN
CAKE
gms
31.32
30.80
32.00
23.95
24.71
TOTAL SOLIDS
MEASURED IN
FILTRATE
gms
1.96
2.20
2.50
1.56
1.88

-------
2.  Three parts of lime per part of FeCl3 worked all the time.  How-
    ever, the optimum lime:FeCl3 ratio could be in the range from 2:1
    to 4:1.

3.  The FeCl3 was always added first and allowed to mix thoroughly
    before adding the lime.  The FeCl3, however, forms a weak floe
    which can be easily broken up by too vigorous mixing; consequently
    care has to be exercised during the mixing.

4.  After the lime has been thoroughly mixed in, the sludge should be
    filtered as soon as possible.  The following tests were made to
    assess floe deterioration.  Specific resistance tests were run
    with a Buchner funnel and the results reported as Rv.  For good
    filtration on the pilot-scale NGK filter press the Rv value should
    be less than 27 x 1010 cm/gm.  (See Appendix C for the Rv procedure),
    These laboratory tests were conducted in a 1500 ml beaker with a
    single paddle stirrer  (1" high x 3" wide).  Maximum speed of the
    stirrer was 100 RPM.  Sludge was added to the beaker with the mixer
    at 100 RPM.  FeCl3 (6.2%) was added and mixed in thoroughly
    (approximately 5 to 6 minutes).  Lime (18.6%) was added and mixed
    in (approximately 5 to 6 minutes).  When a visual check showed
    that the chemicals were well dispersed, this was called time=0.
    At various time intervals samples were grabbed and the Buchner
    funnel test made to determine Rv.

         In Run #1, the mixer was allowed to operate at 100 RPM for
    the duration of the test.  This run showed that the specific
    resistance increased rather quickly:

         Time  (min)                      Rv (cm/g)

            0                            17.5 x 1010
           10                            94.8 x lOJ-jj
           20                            125  x 1010

         In Run #2, the mixer also ran at 100 RPM for the entire test.
    By visual examination, the operator picked the time at which the
    sludge appeared to change.  This test showed that the breakdown
    occurred in approximately five minutes:

         Time  (min)                      Rv (cm/g)

            0                            23.1 x 1010
            5                            76.1 x 1010 sludge appeared
                                                       milky
           11                            98.9 x 1010

         In Run #3, the mixer was slowed to 20 RPM after the lime and
    FeClo had been mixed in thoroughly.  This slow speed was barely
    adequate to keep the sludge mixed.  A longer time was needed before
    noticeable floe deterioration:
                                17

-------
          Time (min)                            Rv (cm/g)

              0                                 11.8 x 10^
             10                                 27.3 x 10 "
             20                                 43.5 x 10   floe began to
                                                         , „   deteriorate
             30                                 93o2 x 10

          The above tests confirmed the visual observations made throughout
     the entire test period; the same deterioration of sludge floe was
     observed many times in the NGK mixing tank.  Marginally conditioned
     sludges were especially susceptible to over mixing or too-long storage
     times.  If, however, the sludges were conditioned well above the
     marginal level, this rapid deterioration was less pronounced.
     Consequently, it should be noted that lime and Fed  usage can be
     minimized by proper design and operation of the conditioning system.

Chemical Requirement vs. Secondary/Primary Sludge Ratio

     A test in a Buchner funnel was used to show the effect of the ratio
of secondary/primary sludge on chemical dosage requirements.  This test was
made for seven different sludge ratios.  For each ratio, the sludges were
blended in the proper proportions and then conditioned, with lime and ferric
chloride.  In all cases, a 3:1 weight ratio of lime:  FeCl_ was used.  The
dosage was considered to be optimum if the sludge could be filtered down to
a good cake in less than 3-4 minutes.  The results of this test are shown in
Figure 3.  It should be noted that this graph shows only a trend, rather
than absolute chemical requirements.  The Blue Plains sludge varies to the
extent that these results would not be duplicated if the test was repeated
on a subsequent day.  This trend, however, is exactly what was found with all
the filter press runs.  Generally, primary sludge, because of its high fiber
content, filters quite readily with only low chemical requirements.
Secondary sludge which is composed of small biological solids is more
difficult to condition and filter.  As the percentage of secondary sludge
increases, the chemical requirements also increase.  If enough lime and
FeCl_ are added, though, the sludge can always be made to dewater.

CONDITIONING WITH POLYMER

     This topic will be discussed under each of the dewatering unit
sectionso
                                    18

-------
30--
             SO
                                                  BO
"OO
                   40         8O
                  °lo  SECONDARY
Figure 3.  Lime requirements vs percent secondary sludge.
                                 19

-------
                                  SECTION 6

                       TEST RESULTS - DEWATERING UNITS


DIAPHRAGM FILTER PRESS (VARIABLE VOLUME PRESS)

     The diaphragm-type filter press, a relatively new innovation in the
wastewater treatment industry in the United States, was tested most exten-
sively during the study.

NGK Diaphragm Press

     NGK Insulators, Nagoya, Japan, has been manufacturing and marketing a
diaphragm-type filter press in Japan for several years.  A pilot-scale model
of their NR-PF-II filter press was provided to the District for the duration
of the study.  This unit was used not only to provide design parameters for
a diaphragm press, but also to study several other factors associated with
any sludge dewatering operation.  Envirex Corporation, Waukesha, Wisconsin,
has since purchased the rights to manufacture and market this press in the
United States.

Facilities—
     The filter press system included the following equipment:.

     1.   Press - 5.8 m2 (62.4 ft2) filtration area.  Contained six chambers,
         with twelve 800 mm (31.5 inches) square plates.  Spacing between
         plates was 25 mm (1.0 ineh).  Every other plate was equipped with
         rubber diaphragms.   As is typical of filter presses, the surface of
         the plate behind the filter cloth resembles the surface of a waffle
         iron, to allow removal of filtrate that passes through the filter
         cloth.  The surface of the rubber diaphragm in contact with the
         filter cloth also has a raised grid pattern for this purpose.
         The press was equipped with a hydraulic closing mechanism and an
         overhead cloth vibrating and washing unit.  See Figure 4.

     2.   Sludge mix tank - 1.0 m3 (264 gallon) tank, with variable-speed
         mixer, equipped with three turbine-wing type agitator blades.  See
         Figure 5.

     3.   Pump assembly - sludge feed pump, squeezing water pump, and cloth
         washing pump.  See Figure 6.

              a.  Feed pump - a diaphragm-type piston pump rated at 100 1/min
                  (26 gpm) and pressures up to 7 kg/cm2 (100 psig).
                                      20

-------
OP
 a
 o
 CO
 e
 m
 )-••
 x
 rt

-------
Figure 6.  NGK pump assembly.
Figure 7.  NGK control panel.
              22

-------
              b.  Squeezing water pump - a multi-stage turbine pump rated
                  at 60 1/min  (15.8 gpm) and pressures up to 17.9 kg/cm2
                  (255 psig).

              c.  Cloth washing pump - plunger-type pump rated at 92
                  1/min (24 gpm) and pressures up to 70 kg/cm2 (1000 psig).

     4.  Water storage tank - 500 1 (132 gallons).

     5.  Air compressors - with receivers (two), each rated at 7 kg/cm2  (100
         psig) for operating ball valves and core blowing.

     6.  Control panel - with relays, timers, etc.  Allowed either a totally
         automatic or manual mode of operation.  See Figure 7.

     The District provided the following equipment to complete the system.

     1.  Primary sludge thickener - total volume of 28.4 m3 (7500 gal) and
         overflow surface area of 8.9 m  (96 ft^).

     2.  Secondary sludge thickener - total volume of 20.8 m3 (5500 gal) and
         overflow surface area of 6.0 m2 (65 ft2).

     3.  Moyno transfer pumps - (two) each rated at 37.8 1/min (10 gpm).

     4.  Lime slurry makeup and storage tank - 757 1 (200 gal) with agitator.

     5.  Ferric chloride makeup and storage tank - 378 1 (100 gal).

     6.  Batch feed tanks - ferric chloride tank, 37.8 1 (10 gal); lime tank,
         56.8 1 (15 gal).

     7.  Calibrated filtrate collection tanks - 378 1 (100 gal) and 56.8 1
         (15 gal).

Operation—
     A complete cycle for the NGK filter press included pumping, squeezing,
and cake discharge operations.  A typical cycle was as follows.  See Figure 8.

     Primary and secondary sludges were pumped into the sludge mix tank at
the desired test ratio.  Solids content of the mix was measured and the
chemical dosage computed as a percentage of dry sludge solids.  FeCls (usually
5-10% by weight of dry sludge solids) was added by gravity and mixed in at
an agitator speed of approximately 95 RPM.   Lime (usually 15-30% by weight
of dry sludge solids) was also added by gravity and mixed in at 95 RPM.  After
visual examination showed the chemicals to be well mixed (about 10-15 min-
utes), the mixer was slowed to a speed just sufficient to prevent stratifica-
tion ( approximatly 28 RPM).  The filter press was closed by actuating the
hydraulic unit, which held a constant pressure of 200 kg/cm2 (2844 psig) on
the plates during the entire press cycle.  The filtration cycle began when
the sludge feed pump was started.   Pumping time was normally 10 to 20 min-
utes, allowing a sludge feed of 227 to 303 liters (50 to 80 gal).  Figure 9


                                      23

-------
         SECONDARY
          SLUDGE
         THfCKENER
IS5
          PRIMARY
          SLUDGE
         THICKENER
                                     LIME
                                    SLURRY
                                     TANK
 FeCI3
STORAGE
 TANK
                                          BATCH
                                          TANKS
                                                                                    WATER
                                                                                    TANK
                                                                       WASHING
                                                                         PUMP
                             SLUDGE
                               MIX
                              TANK
                                      SLUDGE
                                       PUMP
                                    SQUEEZING
                                      PUMP
                                                                        FILTER PRESS
                                                       FILTRATE
                                                                        CAKE
                          Figure 8.  Process  schematic for NGK diaphragm press.

-------
                                                Cloth Suspension
Isi
Ul
                       Sludge

                                                              Squeezing Water
                                                    ••Filtrate
                                                                                              Diaphragm
                                                                                               Cake
                  Filtrate
                                  FILTRATION
SQUEEZING
                      Figure 9.   Schematic of  filtraticm and  squeezing in diaphragm press.

-------
shows a detailed view of the filtration operation.  Sludge is fed through a
bottom feed port into the empty chamber.  Filtered water passes through both
cloths to collection ports on the ends of the plates.  At the end of the
pumping cycle, the squeezing pump was started immediately to pressurize the
diaphragm.  Squeezing time was usually 10-25 minutes at a pressure of 15
kg/cm2 (213 psig).  Figure 9 also shows how the pressurized water expands
the rubber diaphragm behind one of the cloths in each chamber.  The cake is
compressed to approximately half its original thickness as the filtrate
passes through both cloths.

     At the end of the squeezing cycle, the sludge feed lines and filtrate
lines were blown out with pressurized air.  The press ram was opened and the
plate shifter carriage moved two plates into position for cake discharging.
The overhead vibrating unit subsequently positioned itself over these two
plates, lowered its two vibrating shoes onto the cloth support bars, and shook
the four cloths with an eccentric cam action, thereby discharging two cakes.
The shifter carriage then moved another two plates and the process repeated
automatically.  The pilot unit discharged six cakes, each measuring 686 mm
(27 inches) square and approximately 13 mm (0.5 inch) thick.   At the end of
the discharge cycle, the shifter carriage, the vibrating unit and the plates
all moved back into position, ready for another run.  At this point, the
cloth washing cycle was initiated when required.  The cycle was also com-
pletely automatic and similar in operation to that of the cake discharge.
The overhead vibrating and wash unit was equipped with two spray bars, which
washed four cloths at one time.   The cloths were washed each morning and
evening,  and depending on the type of tests,  after each run.   Figure 10 shows
detailed views of both the cake discharge and washing operations.   The cloths
are attached to the plates at the bottom but  are suspended from springs at
the top.   The cloth moves away from the top of the plate to facilitate both
the discharging and washing operations.

     Data sheets 1 through 4 in Appendix B were used in recording data for
the test  runs.  Raw data was recorded on sheets 1 through 3  and results
summarized on data sheet 4.   The data sheets  are filled out for a typical
set of runs,  with calculations detailed in an accompanying explanation.

Test Data to Establish Design Parameters—
     In order to develop design parameters for the dewatering of a 5% solids
sludge to produce a 35% solids cake on a diaphragm filter press,  the follow-
ing parameters should be optimized:

     1.   Chemical requirements
     2.   Feed pump pressure
     3.   Pumping time
     4.   Squeezing pressure
     5.   Squeezing time
     6.   Filtrate quality
     7.   Filter cloth selection
     8.   Filter yield—as a function of all of the above.

     Because of the constantly changing filtration characteristics of the
sludge,  it was extremely difficult to compare test results from one day

                                      26

-------
                                Washing Cylinder
                    Vibrating shoe
                               Washing Nozzle
                              •Plate
CAKE DISCHARGE                                    CLOTH WASHING
     Figure 10.   Schematic of discharge and washing  in diaphragm press.

-------
with those of another.  Consequently, it became necessary to try to collect
enough data for a good comparison study of these parameters from the same
batch of sludge, preferably in one days' time.  The following sections discuss
the methods used to optimize each of the parameters.  The section on filter
yield discusses average overall results for the 2/1 secondary/primary sludge.

     Chemical requirements—Generally, there are three possible ranges of
chemical conditioning:

     1.  The level of lime/FeCl3 dosages below which no dewatering will occur.
         It is obvious that this level must be defined and appropriate
         measures taken to insure that all sludges are conditioned above it.
         Adding underconditioned sludge to a press will cause many wasted
         manhours in cloth rejuvenation, either by high pressure spray wash-
         ing or acid washing.

     2.  The level of chemical dosages in the optimum range where good
         filtration will occur.  Higher chemical dosages within this range
         give slightly higher cake solids and filter yields.  At a particular
         installation, the operator can choose to operate at the upper or
         lower end of this range depending on sludge quantities to be fil-
         tered.  Obviously, cost savings in chemical will inspire the op-
         erator to stay in the lower end of the range as much as possible.

     3.  The level of chemical dosages at very high lime/FeCl3 addition where
         the chemicals are overdosed but dewatering readily occurs.  This is
         a very safe  level for operation, with very little chance of press
         failure but  chemical costs are extremely high.  The increased
         quantities of inert solids, in the final cake also result in decreased
         yields and could cause problems in further processing.

     Table 5 shows the effect of varying chemical dosages on the major filter
press  parameters, final cake solids content and filter yield.  Full-scale
yield  is defined as the weight of sludge solids per square meter of filtration
area per hour.  The total cycle time used in  calculating this full-scale yield
includes the pumping  and squeezing times plus 19 minutes mechanical turn-
around time  (based on the manufacturer's recommendation for their largest
press).  See Appendix B, Explanation of Data  Sheet  4.  The runs on 2/18, 2/23,
3/4, 7/12 and 6/23 clearly show levels  at which the sludge will not dewater.
With all of these poorly conditioned runs, the filter cloths required
considerable cleaning.  In cases where  the conditioning was adequate for
dewatering, higher chemical  dosages  generally gave  either higher cake solids
and/or higher yields.  Examination of the data shows that on the average,
an increase of  5 percentage  points of lime and 2 of FeCl3 can give a 20%
increase in full-scale yield.  But,  if  the objective is to obtain a specific
cake solids content,  for example  35%, at minmum chemical addition, a point
is reached beyond which chemical addition is wasteful.  The runs on 3/10
show that no benefit  in yield  is  gained by increasing the chemical dosage
above  22.8% lime/6.7% FeCl3.   In  fact,  the very high dosages  (45.5%
lime/13.3% FeCl3) actually showed a  decreased yield because of the quantity
of inert chemicals in the  final sludge  cake.
                                      28

-------
      Chemical  dosage was,  in  summary,  largely  a  function of  the  sludge
 characteristics  existing at the  time.   Establishing  the proper dosage is a
 trial-and-error  procedure  that must be performed on  each batch of  sludge to
 be  filtered.   A  simple  Buchner funnel  test and the use of the Capillary
 Suction  Time meter  can  aid in defining this  dosage.   These methods are
 discussed  further in the section on specific resistance tests.

      Feed  pump pressure—The  sludge feed pump  supplied with  the  NGK press
 was  capable of delivering  pressures from 3 to  7  kg/cm^ (43 to 100  psig).
 Numerous tests to optimize the terminal pump pressure were inconclusive.
 The  more difficult  sludges generally would give  higher filter yields if
 the  pump pressures  were above 5  kg/cm^  (71 psig).  The easier to filter
 sludges, such  as those  with high primary ratios, could be handled  with lower
 pressures  of 3 kg/cm2  (43  psig).  In large installations, optimization of
 pump pressure  should be done  under continuous  operating conditions, while
 considering filter  yield,  chemical conditioning  requirements, and  especially,
 filter cloth life.

      Pumping time—With a  diaphragm press, the pumping as well as  the squeez-
 ing  cycle  time must be  optimized so that the filter  yield will be  maximized.
 The  automatic  control system  supplied with the press provided the  option of
 operating  with a preset pumping  time for each  cycle.  This mode of operation
 is best, however, only  if  the sludge filterability does not  change, i.e., if
 a constant lime/FeCl3 dosage  gives consistent  results on the filter press.
 With the Blue  Plains sludge,  this was not the  case.   Sludge  filterability and,
 hence, required  chemical dosages varied almost daily.  With  some sludges a
 pumping  time of  5 minutes  was sufficient; with others, 25 minutes was best.
 Figure 11  shows  the variation of total  feed  volume with time in several
 press runs with  different levels of conditioning.  Data for these runs of
 3/4/77 are shown in Table  5.  For runs  1 and 2 in which the sludge was well
 conditioned, the feed rate remained quite high (over 5 gpm) until  the ninth
 minute.  After that time,  the slopes began to  flatten out as the resistance
 to filtration  started to increase.  In  run 3,  a poorly conditioned sludge,
 the  feed rate  dropped off  and the resistance to filtration began to increase
 after only the fourth minute.

      In a  diaphragm-type press,  the pumping  cycle is used primarily for adding
 filterable solids to the press,  and the pumping cycle time should be optimized
 to this end.   For example, in Table 6,  several runs  are shown for  4/1/77 and
 4/6/77 in  which  successively  longer pump times were  used.  In each case, as
 the pump cycle was  extended, a greater quantity of solids was added to the
press (as  evidenced by  cake dry weights).   Notice,  though, that a correspond-
 ing  increase in yield was not obtained.  The key to  optimizing the pumping
cycle, i.e. to obtain the maximum yield for maximum  solids input to the press,
lies in knowing the solids addition rate for each successive minute of
pumping.   Once this rate drops below the expected average solids yield on
the press  (kg/hr),  then the pumping cycle should be terminated.   Determination
of this rate was made for a generalized sludge feed to the NGK press and was
correlated to a terminal sludge volume rate.

     From early test work,  we established an average rate of 2.4 kg total
solids/hr/m2 (0.5 Ib/hr/ft^)  as a reasonable production rate for a full-scale


                                      29

-------
TABLE 5.  NGK FILTER PERFORMANCE VS.  CHEMICAL CONDITIONING

DATE
2-18-77
2-18-77
2-23-77
2-23-77
3-4-77
3-4-77
3-4-77
3-10-77
3-10-77
3-10-77
3-10-77
7-8-77
7-8-77
7-12-77
7-12-77
7-12-77
4-29-77
4-29-77
4-29-77
5-19-77
5-19-77

RATIO
SEC/PRIM
1/1
1/1
1/1
1/1
1/1
1/1
1/1
1/1
1/1
1/1
1/1
1/1
1/1
1/1
1/1
1/1
2/1
2/1
2/1
2/1
2/1

% CHEMICALS
LIME/FeCl3
26.4/7.8
17.8/5.0
32.0/9.4
18.3/5.4
29.4/8.6
27.5/8.1
14.2/4.1
45.5/13.3
34.3/10.0
22.8/6.7
17.1/5.0
18.9/6.4
14.0/4.6
16.8/5.6
14.4/4.9
12.2/4.1
25.6/8.4
19.5/6.5
15.5/5.2
18.6/6.2
13.8/4.6
CYCLE TIME
(min)
PUMP /SQUEEZE
20/10
20/10
20/10
20/10
16/8
16/8
16/8
16/10
16/10
16/10
16/10
16/14
16/16
14/19
12/23
11/23
21/19
20/19
18/25
15/15
13/26

% CAKE
SOLIDS
35.5
23.6
34.6
23.2
38.4
31.6
24.1
39.6
38.5
36.0
31.6
44.0
41.2
41.8
39.9
29.0
39.8
35.7
35.1
38.0
35.5

FULL-SCALE YIELD
kg/m2/hr
3.14
2.06
2.69
1.76
2.94
3.57
2.87
2.85
3.25
3.27
2.51
3.80
3.46
2.76
2.22
1.44
2.09
1.84
1.49
2.78
1.71

CAKE
DISCHARGE
excellent
poor (wet)
excellent
poor (wet)
excellent
excellent
poor (wet)
excellent
excellent
excellent
good
excellent
excellent
excellent
excellent
poor
excellent
excellent
excellent
excellent
excellent

-------
TABLE 5.
DATE
5-3-77
5-3-77
5-3-77
5-26-77
5-26-77
6-23-77
6-23-77
6-23-77
8-19-77
8-19-77
10-18-77
10-18-77
6-30-77
6-30-77
6-30-77
9-15-77
9-15-77
11-1-77
11-1-77
11-1-77
RATIO
SEC/PRIM
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
3/1
3/1
3/1
3/1
3/1
1/0
1/0
1/0
% CHEMICALS
LIME/FeCl3
18.6/6.2
17.4/5.8
13.7/4.5
20.3/6.6
15.7/4.8
27.0/9.0
24.7/8.3
17.2/5.9
21.5/7.2
14.2/4.8
20.1/6.6
14.8/5.0
24.6/8.3
19.9/6.8
15.7/5.4
26.8/8.9
21.3/7.1
25.2/8.3
19.7/6.6
14.8/5.0
CYCLE TIME
(min)
PUMP/SQUEEZE
18/19
9/24
6/20
18/19
16/19
16/18
15/18
11/18
20/16
18/21
14/15
17/18
19/13
20/18
18/21
24/21
23/21
17/16
17/17
13/21
% CAKE
SOLIDS
38.0
38.1
37.5
36.1
29.3
38.4
39.2
29.4
42.4
36.0
41.0
35.5
40.0
39.6
40.4
36.0
31.2
38.5
36.8
34.5
FULL-SCALE
YIELD
kg/m^/hr
1.97
1.30
1.05
2.00
1.71
2.07
1.95
1.47
3.19
2.12
3.18
2.47
2.97
2.47
2.23
1.96
1.63
2.67
2.73
2.01
CAKE
DISCHARGE
excellent
good (thin)
good (thin)
excellent
excellent
excellent
excellent
poor (wet)
excellent
excellent
excellent
excellent
excellent
excellent
excellent
excellent
good
excellent
excellent
good

-------
11O —
1OO _
B     B    10   IS   14
 TIME  C MINUTES?
                                                    IB   IB
        Figure 11.  Feed Volume vs. time.  NGK runs on 3/4/77.
                                32

-------
                                      TABLE  6.   RUNS TO OPTIMIZE PUMPING TIME
UJ


DATE
4-1-77



4-6-77


5-19-77





% CHEMICALS
LIME/FeCl3
15.4/4.5
15.1/4.3
15.1/4.6
14.1/4.3
18.0/5.4
18.0/5.4
18.0/5.4
20.4/6.8

18.6/6.2
13.8/4.6

CYCLE TIME
(MIN)
PUMP /SQUEEZE
5/15
10/15
15/15
20/15
5/15
10/15
15/15
19/12

15/15
13/26


% CAKE
SOLIDS
44.9
42.7
33.8
34.5
43.0
39.1
37.5
33.6

38.0
35.5


CAKE DRY
WEIGHT (kg)
8.0
10.9
12.1
13.8
8.0
9.4
10.7
16.8

16.4
11.3

FULL-SCALE
YIELD
kg/m2/hr
1.7
2.2
2.1
2.2
1.7
1.8
1.8
2.7

2.8
1.7



REMARKS
.thin cake
thin cake


thin cake
thin cake

squeeze time
too short

marginal
conditioning

-------
installation size NGK press.   For the 5.8 m2 pilot press this rate is

     2.4 kg solids /hr/m2 x 5.8 m2= .23 kg solids/min (.51 Ib/min)
          60 min/hr

For a 7.5% total solids feed  and a specific gravity of 1.02, the volumetric
feed rate is
    .23 kg solids/min x ^°9 ^ to^a]  feed
                        7.5 kg solids
                                           =3.0 1/min feed (.8 gpm)
This terminal pumping rate, through calibration of the NGK mix tank, was
found to be equivalent to 1/8 inch per minute.  During later test work (i.e.
after 4/28/77)  when the sludge flow rate dropped to 1/8 inch per minute
for three consecutive minutes, the pumping cycle was terminated.  The variable
output of the sludge feed pump and the difficulties in measuring 1/8 inch
necessitated a three minute time to ensure that a good measurement was taken.

     The runs on 5/19/77 (Table 6) show the effects of this procedure on three
different levels of sludge conditioning.  The first run, with good condi-
tioning, gave a high yield with a rather long pump time of 19 minutes (low
cake solids resulted from an error in the determination of the squeezing
time).  The second run was still with good conditioning and a high yield
resulted.  The third run, with marginal conditioning, achieved the required
pumping rate in only 13 minutes.  This indicated to the operators that the
sludge was not well conditioned.  It was then necessary to squeeze for a
slightly longer time, so that a good cake release would result.

     A secondary advantage of using this procedure involved the response of
the press to poorly conditioned sludges.  In general, with a poorly condi-
tioned sludge, this rate was usually achieved in 5 to 10 minutes and a thin
cake was produced.  This thin cake, however, would further dewater under
extended squeezing times and thus give a good discharge from the filter cloth.
Earlier runs showed that poorly conditioned sludge, when allowed to form
a thick cake, did not dewater well under extended squeezing  and cake sticking
and resultant cloth blinding occurred.  This new operational procedure thus
gave a way for the filter press to compensate for errors that had occurred
in the conditioning step.  This same method, applied to a well conditioned
sludge produced a thick cake with maximized yield.  In effect, this method
gave the best filter performance for the sludge and conditioning available.

     Recognizing the problems with instrumentation that could occur in obtain-
ing a sludge flow rate on a larger filter press, we examined two other
methods for optimizing the pump time:  1) rate of pump pressure buildup and
2) filtrate flow rate.  The rate of feed pump pressure buildup gives an
indication of the resistance to filtration that exists during the dewatering
process.  It can be used to indicate a poorly conditioned sludge and alert
the operator to take corrective action.  It cannot be used, however, to
define the cycle endpoint for a well conditioned sludge.  In Figure 12, the
feed pressure curves for the runs of 3/4/77 are shown.  The best .conditioned
sludge  (Run #1) built up pressure slowly, indicating little resistance to
filtration.  At the end of the pumping cycle  (16 minute mark) , the pressure

                                     34

-------
                                                 PRESSURE   CPSIO1
UJ
Ln
                OP
                C
                OJ

                4^
                                             ID
                                             0
U
0
n
o
o
0
M
0
o
0
ID
0
0
0
                           It
                           ft  -
fD
fD
CU
fD
CO
CO
C
l-j
fD
CO

ft
H-
§
C
3
CO
O
3

H
m

0^
2
Z
C
H
m
01
M







0)


0

J
0

J
10
g


_a
m
                                                                               ni
ra
0
a
C
z
0
•
1 u



^^^~






2

9
C
z
0
u
'„»
ft
w
o
0
r
i
m
ft
1
J
n
u
jj
C
Z
0
io
n
M
en
1
r
i
m
«••
p
La
o
o
J
n
u
n
c
z
p
la
10
(0
ft
o
o
• r
i
m
«••
o
•
Q
o
0
•n
n
u

-------
was still rising at a steady rate.  The under-conditioned sludge (Run #3)
built up pressure very rapidly.  Resistance to filtration was nigh  and
the pressure reached a limiting value well before the end of the cycle
(also 16 minute mark).

     The filtrate flow-rate is an easily measured parameter  and a correlation
which would determine the end of the pumping cycle can be readily developed
from the filtrate flow rate data collected.  Figure 13, for example, shows
the filtrate collected vs time for the 3/4/77 runs.  At the 16 minute mark
(the end of the pump cycle), Runs #1 and #2 still had a fairly steep slope,
whereas Run #3 had already begun to flatten out.    These curves show that
since filtrate for run #1 and #2 was still being discharged at a high rate
at the end of the cycle, had the pumping times for these runs been extended
higher yields would have resulted.  Run #3 pumping, though, should have been
terminated near the 8 minute mark, where the slope began to flatten out.

     Because of the need to test different ratios of blended sludges and
variable terminal pump pressures, we decided to use the more direct method
of measuring the sludge feed rate to determine the end of the pumping cycle.
On a full-scale installation, however, either sludge feed rate or filtrate
flow rate could be used.

     Squeezing pressure—The squeezing pressure was developed by applying
pressurized water to the diaphragms.  During startup, the NGK engineers
recommended that the press be operated at 15 kg/cm2 (213 psig).  However,
the pump could deliver any pressure up to 17.6 kg/cnr (250 psig).  Numerous
tests were run throughout the study period to try to optimize the squeezing
pressure level and the rate at which it was applied.  None of the tests
showed a marked difference in final cake solids or squeezing time over the
range of 7 to 17.6 kg/cm^ (100 to 250 psig).  The Blue Plains sludge appeared
to dewater independently of pressure within this range.  Under normal
operation, the full pressure was applied to the diaphragm immediately after
the sludge pump stopped.  Tests were also run applying the pressure in step
increments up to the final squeezing pressure.  Again, no difference in
results could be determined.  Therefore, in nearly all the press runs a
squeezing pressure of 15 kg/cm^ (213 psig), applied as per the manufacturer's
design, was used.

     Squeezing time—In an optimum pumping cycle for a well conditioned
sludge in the NGK press, approximately 75-85% of the filtrate will be
collected during pumping.  The squeezing cycle is then really a cake
consolidation step removing relatively small quantities of filtrate.
Generally, the squeezing cycle increases the cake solids from approximately
20% at the end of pumping to 35-40% solids before discharging.  Figure 14,
for example, shows the results of tests run on 4/13/77.  For these five runs
the lime/Feds dosage was 20.7%/6.2%; the pump time was 18 minutes.  As the
squeeze time was increased from 5 to 25 minutes, the cake solids increased
from 25% to over 40%, with a corresponding decrease in process yield.

     During the initial part of the study  (prior to 4/28/77), the squeezing
time was preset by the operator.  As with the pumping time, however, this
method was good only if the sludge and conditioning remained constant.  With

                                      36

-------
u>
                  90
                  ac
               0)
               ui
                  ao
                  so
                  10
                                            RUN NO. 1
                                           1O          IB
                                         TIME  CMIN8)
                         Figure 13.   Filtrate volume vs.  time.  NGK runs on 3/4/77.

-------
Co
oo
               if 5.0
               I

               ° 4.B
               w

               Q
               a 3.s
               ID
               HI
               U
               0 3.4
               B
               a

                 3.O
                  B.B
                                              I
I
                                                                                                40
                                                                                                38
                                                                                                3B
                                          0)
                                          Q
                                          n
                                                                                                 3O
                                          U

                                          o
                                          0
                                      28
                                                                                                 24
                                             1O          IB

                                              TIME CMIIMUTES)
           20
                                                                                              3O
                                  Figure 14.   Effect of increasing  squeeze times - NGK press.

-------
  7O

HI

< 80
E
b
IL
  ao
vo    • 3O

     3

     0 BO
  10
                                        RUN MO.1  3B.B»la  SOLIDS


                               RUN  NO.B   3B.BOIo  SOLIDS-
                                                            RUN  NO.3   34.B°lo  SOLIDS
                            1O         IB         20

                                      TIME  CMINS1
                                                                                 30
3B
4O
                        Figure 15.  Filtrate volume vs. time.  NGK runs on 11/1/77,

-------
a well conditioned sludge, squeezing times of 8 to 10 minutes were normally
sufficient to produce the required 35% solids cake; with the marginally
conditioned sludges squeezing times of over 20 minutes were required to
reach this level.  Examination of many press runs showed a correlation between
the terminal filtrate flow rate and final cake solids.  The data from the
runs on 4/13/77 is typical:


Squeeze Time (min)        % cake solids          Terminal filtrate rate
                                                      1/min (gpm)

        5                      25.5               3.78        (1.00)
       10                      31.3               1.89        (0.50)
       15                      35.0               0.87        (0.23)
       20                      37.3               0.57        (0.15)
       25                      40.6               0.00        (0.00)

Based on results such as these, a procedure was developed to ensure final
cake solids of 35% for each run.  Beginning on 4/28/77, after pumping the
sludge feed to a rate of 3.0 1/min, the squeezing cycle was extended until
a terminal filtrate rate of 0.57 1/min (0.15 gpm) was obtained.   When this
rate held steady for three consecutive minutes (to obtain good measurement),
the squeezing cycle was terminated.  Subsequent tests showed that this
terminal rate method worked well with all levels of conditioned sludge.
Figure 15 shows the filtrate curves on three levels of conditioning 100%
secondary sludge for this procedure.  Data for these 11/1/77 runs are included
in Table 5.  Run #l-was slightly over-conditioned; Run #2 was conditioned
at an average level, and Run #3 was marginally conditioned.  In each case
the press cycle was the optimum for the type of sludge dewatered; cake solids
of about 35% resulted for each run.

        Filtrate quality—-The overall quality of the filtrate is affected
not only by the selection of cloth, but also by the chemical conditioning.
Normally, only total solids and suspended solids were analyzed on filtrate
samples.  On 3/8/77, tests were run to characterize the filtrate for other
parameters.  The filter cloth was not washed between these eight runs, so
that a normal plant operation would be simulated.  Feed solids were 6.0% on
a 1/1 ratio of secondary/primary sludge.  Cycle time for each of the runs
was 16 minutes pumping and 8 minutes squeezing.  Table 7 shows the filtrate
parameters analyzed.  Because of filtration difficulties at that time, all
runs were made with high chemical dosages.  Each of the pollutant levels is
consistent with that for a water in contact with undigested sludge.  The
average percent total volatile solids in the filtrate was 24.8%, indicating a
high soluble chemical content  (mostly lime).  The average percent volatile
suspended solids was 60%, indicating that filtrate suspended solids were
mostly organic.

        Further tests were conducted to determine the effect that chemical
conditioning had on filtrate quality.  See Table 8.  The first three runs
on 3/10/77 were well conditioned and gave good cake results.  Note that the
filtrate total solids (mostly lime) decreased as the chemical addition rate
decreased.  The last run was underconditioned and gave poor cake results and

                                      40

-------

% CHEMICALS
LIME/FeCl
26.7/7.8
26.8/7.8
26.8/7.8
29.3/8.5
29.3/8.5
29.3/8.5
30.0/8.8
30.0/8.8
28.5/8.3

DATE
3-10-77
3-10-77
3-10-77
3-10-77
7-12-77
7-12-77
7-12-77
11-1-77
11-1-77
11-1-77


pH BOD
11.9 554
11.9 916
11.9 694
11.7 684
11.7 682
11.8 684
11.8 420
11.9 374
11.8 626

% CHEMICALS
LIME/FeCl3
45.5/13.3
34.4/10.0
22.8/6.7
17.1/5.0
16.8/5.6
14.4/4.9
12.2/4.1
25.2/8.3
19.7/6.6
14.8/5.0


COD
1560
2067
2061
1640
1668
1785
1730
1867
1923
TABLE
% CAKE
SOLIDS
39.6
38.5
36.0
31.6
41.8
39.9
29.0
38.5
36.8
34.5
11 '
F I L T R A T

P04 TKN NH3
35.8 173 85.3
110 253 107
63.6 214 105
44.7 244 106
56.7 240 96.9
69.2 296 99.2
47.5 226 96.9
64.0 258 110
61.4 238 101
E mg/1
TOTAL

TOTAL SUSPENDED
N03 ALKALINITY SOLIDS SOLIDS
.70 1801
.74 2367
.80 1987
.75 2021
.77 1961
.79 1959
.74 2063
.72 2008
.75 2020
8. FILTRATE QUALITY VS. CHEMICAL
CAKE
DISCHARGE
excellent
excellent
excellent
good
excellent
excellent
poor
excellent
excellent
good
TOTAL SOLIDS
mg/1
10120
8910
8189
9404
9815/10130*
9736/10024*
8963/8057*
6773
5879
5546
7286 144
9431 1540
8575 464
8307 151
8203 62
8140 55
8471 127
8474 47
8361 S23 (Averaee')
CONDITIONING
SUSPENDED SOLIDS
mg/1 pH
87 11.5
70 11.6
80 11.5
2604 11.5
120/18* 11.5
216/45* 11.5
438/198* 11.5
28
22
1 8,9 -

-------
high filtrate suspended solids.  High filtrate suspended solids were common
for a poorly conditioned sludge.  The fine particles were not well flocculated
and passed through the filter cloth.  The runs on 11/1/77 show identical
results.

        The following typical run shows the filtrate solids as sampled at
various time intervals during the pumping cycle.  This was taken from Run
#1 on 5/3/77 shown in Table 5.
                                 Filtrate                   Filtrate
                               Total solids             Suspended solids
Time (min)                        mg/1                        mg/1

   1.5                            8242                        716
   3.0                            8169                        550
   5.0                            7687                         55
  10.0                            7655                         45
  18.0                            7604                         19
The  suspended solids drop off rapidly after three minutes.  This is due to
the initial filtration being through the cloth.  Once a"cake is formed in
the chambers, the accumulated solids then act as the primary filter media.
The composite sample of filtrate during the pumping cycle averaged 7826 mg/1
total solids and 308 mg/1 suspended solids.  The squeezing cycle for the
same run showed 7713 mg/1 total solids and  50 mg/1 suspended solids.  Note
that the suspended solids were lower for the squeezing cycle than for the
pumping cycle.  This was true for all the press runs.  The runs on 7/12/77
(Table 8) show several cases in which the filtrate collected from the pumping
and squeezing cycles were analyzed separately.

     Filter cloth evaluation—Filter cloth selection depends on resistance
to wear and abrasion, the ease of cake release, and filtrate quality desired.
Three different filter cloths were supplied with the NGK filter press, each
of which was tested during the study.  Because of the limited number of runs
made and the method of operation (day time only), resistance to wear and
abrasion could not be determined by this study.  Cloth media specifications
are given in Table 9.


                      TABLE 9.  NGK FILTER CLOTHS

                                                         AIR PERMEABILITY
                                  MATERIAL            at A P = 12.7 mm
 TYPE      CONSTRUCTION         WARP/FILLING               Cm3/Sec/Cm/
NY 516
TR 520
NY 51-4
Plain
Herringbone
twill
Twill
polyamide/polypropylene
polyester /polyester
polyamide/polyester
4.-0
11.0
93.0
42

-------
      The NY 516 cloth, the tightest weave, was the first cloth tested.
 A total of 151 runs (5/26/76 thru 11/1/76) were made on this cloth, mostly
 with plant thickened sludge.  Filtrate analyses for 71 of these runs showed
 excellent results with an average of 16.7 mg/1 suspended solids.  Cake
 discharge, however, was not always the best.   For nearly all the runs, except
 those that were over conditioned, the cloth shaker was needed for cake
 discharge.  The cloth surface was rough and particles of cake tended to stick.
 Underconditioned sludges blinded the cloth very readily and excessive
 scrubbing was required.  The cloth washing system, operated at 350 psig, did
 little to rejuvenate the cloth after such a run.   Resistance to abrasion
 and wear seemed to be very high.

      The second type cloth tested was the NY  51-4 media.   This cloth had the
 most open weave  and a very smooth surface.   With the exception of very
 poorly conditioned sludges, cake discharge was almost always good.   The
 first set of NY 51-4 cloths, operated from 11/2/75 through 4/7/77,  were
 worn badly after 182 runs due to  over-zealous brushing while cleaning.   The
 brushing was required during early 1977 when  difficulties were encountered
 with dewatering the sludge.   The  second set of NY 51-4 cloths were operated
 for 213 runs from 4/8/77 to 10/7/77.   Filtrate analyses on 271 samples  with
 both NY 51-4 cloths averaged 525  mg/1 suspended solids on all types of  sludge.
 This cloth gave the best overall  cake solids,  since it provided little
 resistance to filtration.   Filtrate quality was,  however,  a drawback.   The
 cloth also seemed to show little  resistance to wear.   The second set were
 beginning to tear in places after only 130 runs,  but  this  may have  been due
 to  the alternate wetting and drying (which is  known to stretch fibers)
 caused by our operational schedule.   Further  evaluation on a continuous basis
 is  required before any  definite conclusions can be made.

      The TR 520 cloths  were tested  from 10/13/77  through  11/30/77  for a total
 of  80 runs.   Only 22 of these runs  were analyzed  for  filtrate quality giving
 52.4  mg/1 suspended  solids.   Even though this was  a textured,  heavy cloth,
 cake  discharge  was  excellent—equal to  the NY  51-4  cloth.   Sufficient runs
 to  determine  abrasion resistance  were not made, but the cloth seemed to be
 more  sturdy than  the NY 51-4  cloth.   The TR 520 cloth  appears  to provide the
 best  compromise for  both  good discharge  and acceptable  filtrate  quality.

      Cloth  washing requirements are more a function of chemical  conditioning
 than  cloth  selection.  When the conditioning was optimum, up  to  15 runs
were made on  each of the cloths before they required washing.  The cloth
washing  system, while designed to operate at 70 kg/cm2  (1000 psig), only
produced a  maximum pressure of 24.6 kg/cm2 (350 psig).  A defective pressure
gage and regulator valve caused this problem but was not discovered until
all tests were  completed.

     Filter yield—A major purpose of the study was to develop design parame-
ters to dewater a 2/1 ratio of secondary/primary sludge from an initial 5%
solids mixture to a 35% solids cake.  A total of 142 runs were made on this
sludge ratio.  Recognizing that sludge variability was an important factor
during the study and that many types of experiments were made, only the runs
that gave at least a 35% solids cake were used to produce representative


                                      43

-------
design conditions for dewatering the Blue Plains sludge on a year-round basis.
According to the manufacturer, scale-up from the pilot press data to a full-
scale unit can be made directly.  Filtration (pumping) and squeezing cycle
times will be identical.  A mechanical turn-around time, though, must be
included in the total cycle time in order to obtain the full-scale yield.
(See Appendix B, Explanation of Data Sheet 4 for details).

     These runs, tabulated in Table 10, are summarized below:

          Chemical dosage                  19.6% lime/6.5% FeCl3
          Cycle time                       16.9 min pumping/
                                           18.1 min squeezing
          Final cake solids                38.7%                       9
          Full-scale yield                 2.39 kg/hr/m2  (0.49 lb/hr/ftz)

This data was used to develop average design parameters for a full-scale press
installation at Blue Plains.  Cost- estimates for the NGK press, given in
Section 9, are based on these average values.

Special Tests—

     Dewatering of variable sludge ratios—A secondary purpose of the study
was to observe the effect of dewatering various ratios of secondary to
primary sludge solids.  Tests on the Buchner funnel, Figure 3, had shown that
more chemicals were required as the percentage of secondary sludge increased.
Tests on the NGK press confirmed these results and also showed the effects
that the sludge ratio had on filter yield.  .During the month of August, 1977
the sludges were fairly consistent in their filterability.  During that month,
seven different sludge ratios were .tested.  With each sludge ratio at
least three runs at three different chemical dosages were made; one over-
conditioned, one average conditioned, and one marginally  conditioned.  The
results are averaged for .each sludge ratio in Table 11.   Note the general
trend that sludges high in primary solids give high cake  solids and high
yields with relatively low chemical dosages.  Once the ratio of solids in-
creases above 1/1 secondary to  primary, the secondary sludge is the control-
ling factor and the sludges become more difficult to dewater.

     Three day continuous run—From 10/4/77 - 10/7/77 the NGK pilot unit was
operated continuously for a period of 72 hours.  The primary objectives of
this test were:

     a.  To simulate a  full-scale installation and thereby obtain represen-
         tative operating parameters,

     b.  to test the effectiveness of a.continuous chemical  conditioning
         scheme;

     c.  to establish diagnostic and monitoring procedures for  a full-scale
         system,

     d.  to operate the unit  under stress conditions  in order to evaluate
         the mechanical design, and

                                       44

-------
TABLE 10.  NGK RUNS ON 2/1 SLUDGE


DATE
3-15-77

3-16-77

3-30-77

3-31-77

4-1-77

4-5-77

4-7-77




4-12-77
4-13-77


4-26-77
4-28-77

4-29-77


5-3-77


5-4-77

5-11-77
5-12-77

5-17-77
5-18-77
5-19-77

5-20-77
5-23-77

5-24-77
5-26-77

% CHEMICALS
LIME/FeCl3
26.0/8.1
16.0/4.6
15.9/4.6
19.6/5.7
21.6/6.3
21.6/6.3
14.9/4.2
15.0/4.2
15.4/4.5
15.1/4.3
21.9/6.4
21.9/6.4
20.8/6.0
20.8/6.0
20.3/6.0
20.3/6.0
20.3/6.0
17.2/5.0
20.7/6.2
20.7/6.3
21.6/6.6
21.5/6.4
19.3/5.9
24.0/6.9
19.5/6.5
25.5/8.4
15.5/5.2
18.6/6.2
17.4/5.8
13.7/4.5
24.4/8.4
20.2/6.8
19.4/6.2
19.6/6.4
20.2/6.6
17.6/5.9
20.3/6.8
18.6/6.2
13.8/4.6
19.9/6.8
17.1/5.8
19.5/6.5
14.6/4.9
20.3/6.6
CYCLE TIME
(MIN)
PUMP/ SQUEEZE
30/15
30/15
30/15
20/15
20/15
20/15
20/15
20/15
5/15
10/16
5/15
10/15
5/15
10/15
15/15
20/15
25/15
18/25
18/15
18/20
18/25
20/20
17/20
22/20
20/19
21/19
18/25
18/19
9/24
6/20
19/14
16/16
15/22
17/22
15/21
11/23
18/19
15/15
13/26
11/22
16/18
19/20
18/16
18/19

% CAKE
SOLIDS
41.7
44.4
36.4
39.9
41.4
40.6
38.8
35.8
44.9
42.7
43.8
38.3
41.3
40.0
38.8
36.7
37.7
38.3
35.0
37.3
40.6
36.6
37.8
39.7
35.7
39.8
35.1
38.0
38.1
37.5
37.1
37.8
36.6
38.0
37.0
37.7
35.8
38.0
35.5
35.8
36.1
38.1
37.4
36.1
FULL-SCALE
Yield
kg/hr/m2
2.78
3.30
2.54
2.94
3.23
3.13
2.72
2.69
1.75
2.17
1.93
2.27
1.78
2.47
2.75
2.28
2.52
1.89
2.32
2.05
1,93
2.09
1.78
2.18
2.07
2.35
1.66
1.97
1.30
1.05
2.57
2.48
1.65
1.87
1.82
1.63
1.87
2.78
1.71
1.46
1.91
2.17
2.48
2.00
         45

-------
TABLE 10.


DATE
6-15-77
6-22-77

6-23-77

7-8-77
7-11-77

7-12-77
7-14-77
7-27-77
7-28-77
8-2-77

8-4-77
8-17-77

8-18-77
8-19-77

8-23-77
8-24-77
8-25-77

9-1-77
9-2-77

9-9-77

9-14-77

9-21-77
9-22-77
10-13-77
10-18-77


10-19-77






% CHEMICALS
LIME/FeCl3
19.9/6.6
20.4/6.9
25.4/8.4
24.7/8.3
27.0/9.0
20.6/6.8
25.9/8.7
13.9/4.6
19.0/6.4
20.9/7.0
26.8/8.9
16.8/5.6
22.1/7..4
17.9/5.9
14.7/4.9
18.0/6.0
17.5/5.8
19.6/6.6
21.5/7.2
14.2/4.8
16.6/5.5
18.3/6.2
20.1/6.7
19.3/6.5
17.9/6.0
18.9/6.3
17.8/6.0
19.1/6.4
19.1/6.4
24.3/8.1
29.2/9.7
27.7/9.2
24.1/8.0
15.0/5.0
23.4/7.8
20.1/6.6
14.8/5.0
20.0/6.7
20.0/6.7
20.0/6.7
20.2/6.8
20.2/6.8
20.9/7.0
CYCLE TIME
(MIN)
PUMP/ SQUEEZE
16/15
11/21
17/18
15/18
16/18
16/16
17/16
18/20
14/17
20/22
16/14
17/19
21/16
20/18
14/23
16/15
16/16
18/16
20/16
18/21
15/20
21/23
18/18
17/17
20/20
15/18
16/21
16/22
12/22
13/22
19/17
16/16
18/20
20/15
18/17
14/15
17/18
18/17
17/19
15/17
18/17
19/17
17/18

% CAKE
SOLIDS
36.5
41.7
35.5
39.2
38.4
41.0
39.6
37.7
39.2
42.5
44.9
40.7
41.4
40.8
40.2
39.9
39.1
40.2
42.2
36.0
36.4
41.4
43.0
40.9
38.8
38.7
38.5
37.2
36.7
36.3
35.5
41.0
36.1
35.7
37.3
41.0
35.5
39.2
38.0
45.3
37.1
38.3
34.7
FULL-SCALE
Yield
kg/m2/hr
2.21
1.96
1.93
1.95
2.07
3.22
2.60
2.64
2.54
2.22
3.70
2.97
3.07
2.83
2.16
2.91
2.70
2.76
3.19
2.10
2.12
2.19
3.07
2.76
2.35
2.45
2.39
2.05
1.88
1.90
2.30
2.67
1.79
2.69
2.44
3.18
2.47
2.53
2.33
3.03
2.66
2.59
2.25
   46

-------
                            TABLE 10.
DATE

10-25-77
10-26-77
10-27-77

10-28-77
11-2-77
% CHEMICALS
LIME/FeCl3

19.3/6.4
25.7/8.5
15.8/5.3
15.8/5.3
12.0/4.0
12.1/4.0
19.7/6.6
20.4/6.8
CYCLE TIME
   (MIN)
PUMP/SQUEEZE

    19/15
    19/14
    18/18
    16/18
    18/20
    15/21
    20/16
    20/17
% CAKE
SOLIDS
37,
41,
36,
37,
35.
36.
40.
39.8
FULL-SCALE
   Yield
 kg/m2/hr

   2.82
   2.42
   2.62
   2.63
   2.31
   2.05
   2.94
   2.62
Averages   19.6/6.5
                  16.9/18.1
                38.7
               2.39
                                47

-------
                             TABLE  11.   TYPICAL  RESULTS  ON DIAPHRAGM PRESS - AUGUST RUNS
00

SEC/PRIM
RATIO
0/1
1/2
1/1
2/1
3/1
4/1
1/0

NO. OF
RUNS
2
2
3
13
5
3
4

% CHEMICALS
LIME/FeCl3
13.4/4.5
13.2/4.4
15.6/5.2
18.3/6.1
22.7/7.6
21.4/7.2
23.5/7.8
PROCESS
CYCLE/TIME
(MIN)
23
32
35
36
39
34
34

% CAKE
SOLIDS
54.0
48.1
47.4
40.0
40.7
43.1
38.9

FULL-SCALE
YIELD (kg/hr/m2)
4.95
3.62
3.46
2.63
2.41
3.06
2.28
FILTRATE SOLIDS
mg/1
TOTAL/ SUSPENDED
8366/1926
8156/602
8558/308
8238/386
10358/380
7927/197
7971/297

-------
      e.   to acquaint  plant engineers,  maintenance,  and operating personnel
          with the design and operation of  a filter  press.

      The  flowsheet for  the process  is  shown in  Figure  16.   Primary  sludge
 (1% solids) and  secondary sludge  (0.75% solids) were gravity  thickened  to
 8% and 4% solids,  respectively, and mixed  in the  blending  tank  in a
 secondary/primary solids ratio of 2/1.   A  recycle rate within this  tank of
 10-20 gpm provided the  necessary agitation for mixing.  From  here,  the
 blended sludge (5.3%  solids)  was pumped continuously at 1.5 gpm to  the
 chemical  conditioning system.  A Komline-Sanderson  Rotary  Drum  Conditioner,
 with internal baffles for mixing, was  used as the conditioning  tank.  Ferric
 chloride  (13% weight  solution) was  added by a positive displacement pump to
 the sludge feed  line; lime slurry (6.5% -  13% weight solution)  was  added
 to  the middle of the  conditioning drum.  In order to minimize floe  deterio-
 ration during the  conditioning step, drum  speed was maintained  at one RPM,
 and the sludge was  detained only a  few minutes before  overflowing to the press
 feed tank.   This tank held enough sludge for 2-3  press runs and had an
 average sludge retention time of 1.5 hours.

      Automatic filtration and squeezing cycles of about  20 minutes each were
 used during press  runs.   These cycles were  checked and adjusted every fifth
 run by measuring the  sludge and filtrate flow rates.   The cloth washing
 cycle was  initiated only  when required;  hence, the turn-around  time between
 successive  runs  averaged  only 10 minutes.

      Sludge filterability was monitored each run by Buchner funnel and CST
 tests on the  sludge leaving the conditioning tank.  A  CST of  15  seconds and
 a Buchner  funnel filtrate  rate of 80 ml/2 min was used as an  indicator of
 acceptable  filterability.   Samples of sludge, cake,  and  filtrate were taken
 every fifth run  for laboratory analysis.

     A total of  76 runs were made, 72 on the 2/1 secondary/primary mixture
 and 4 on the 100% secondary sludge.   Approximately 5600 gallons of sludge
were  filtered during the operation.   Press down-time was minimal and 59.3
operating hours were logged.  Results of the laboratory analyses are summa-
rized in the following table:

     Sludge feed solids/volatile  solids    5.32%/67.5%
     Conditioned sludge volatile  solids    46.9%
     Lime dosage (average)                 22.9% (of sludge solids)
     FeCla dosage (average)                7.3% (of  sludge solids)
     Cycle time  (avg)  pump/squeeze/
      mechanical                          20/20/5 minutes
     Cake  wet weight (total)               2682  kg
     Cake solids/volatile solids            36.3/48.2%
     Cake  dry weight                       973 kg
     Cake  sludge  solids                     748 kg
     Yield  (average)                       2.17  kg/hr/m2
     Filtrate suspended  solids             83.2  mg/1
     Filtrate total solids                 9465  mg/1
     Cloth washed                          16 times  (every 4.75  runs)
     Cloth used                            NY 51-4

                                      49

-------
         PRIMARY
         SLUDGE
           1%
SECONDARY
 SLUDGE
  0.75%
FeCl3
STORAGE
 TANK
 LIME
STORAGE
 TANK
                            4%
Ul
o
     10-20
      gpm
                                                                                            FILTER PRESS
                                                                                            PUMP TIME 20 mln.
                                                                                            SQUEEZE TIME 20 mln.
                                                        CONDITIONING
                                                            TANK
                                                                       FEED
                                                                       TANK
                                                                                      60-80
                                                                                      gal/batch
                                                                                                                  FILTRATE
                                                                    6 Cakes
                                                                    100 ///batch
                                                                    35 - 40%  SOLIDS
     1.5 gpm
                                  Figure 16.  Process flowsheet  for continuous run.

-------
     The chemical addition system proved to be the major bottleneck of the
operation.  Frequent plugging and clogging of the lime slurry feed line
caused the sludge feed to be under-conditioned for several of the runs.
Consequently, poor cake discharge and excessive cloth blinding occurred.
Because of these problems in the lime system, the bench-scale filterability
tests were invaluable.  Interruptions in the lime delivery were immediately
evident by high CST values (e.g. 257 sec) and low Buchner funnel filtrate
rates (e.g. 14 ml/2 min).  Thus, the operators were able to slug dose the
feed tank in order to avoid disastrous press runs.

     Problems with the electrical functions of the cake discharge and cloth
wash mechanisms caused some minor delays in the automatic operation of the
press.  But otherwise, the press performed extremely well during the extended
operation.

     Plant personnel, who initially were unfamiliar with the press, were
generally pleased with its operation.

     The entire project, therefore, was considered highly successful.  Several
important design suggestions evolved from these continuous runs and will be
discussed later in the design section of the report.

Lasta Diaphragm Press

     From 10/25/77 to 11/3/77 Ingersoll-Rand, Nashua, New Hampshire, provided
a trailer mounted demonstration unit of their Lasta press for testing.  This
unit was also a Japanese-made press and is manufactured under license from
Ishigaki Mechanical Industry Co., Ltd.   Tests were run for comparison with
the NGK diaphragm press.

Facilities—

     1.   Press - 1.64 m^ (17.65 ft^) filtration area; contained four chambers
         with eight 600 mm (23.6 inches) square plates; every other plate was
         equipped with concave rubber diaphragms.   See Figure 17.

     2.   Tank assembly

         a.   sludge conditioning tank - 0.7 m3 (184 gal)  tank with variable
             speed mixer

         b.   sludge storage tank -  1.3  m3 (350 gal) tank

         c.   lime slurry  tank - 1.3 m3  (350 gal)  tank,  with constant speed
             mixer

         d.   ferric chloride  tank - 0.6 m3 (150 gal)  tank,  with constant speed
             mixer

         e.   water storage  tank - 0.6 m3 (150 gal)  tank
                                      51

-------
      4.

      5.
Pump assembly

a.  three (3) constant speed pumps - cake wash, cloth wash, dia-
    phragm pressurization

b.  three (3) variable speed pumps - sludge, lime, and ferric
    chloride delivery

c.  one (1) vacuum pump - diaphragm deflation

d.  hydraulic pump - opening and closing press

Air compressor with receiver - core blow and instrumentation control

Filter cake conveyor
Operation—
     Conditioned sludge was prepared in the NGK mix tank and pumped to the
Lasta conditioning tank for use during the tests.  As with the NGK press, the
Lasta press cycle included pumping, squeezing, cake discharging, and cloth
washing operations.  The pumping cycle, during which sludge was fed to the
press, averaged 10 minutes at pressures of 4.6 - 7.0 kg/cm2 (65-100 psig).
Sludge feed volume ranged from 4.5-60.9 liters (1.2-16.1 gallons) and entered
the filtering chambers via special dispersion nozzles located at the top
                      Figure 17.  Lasta Diaphragm Press.
                                      52

-------
 center of the filter plates.  Filtrate was discharged through side nozzles
 at the bottom of the filter plates.

      At the termination of the pumping cycle, the squeezing cycle began
 immediately.  An average cycle length was 20 minutes at pressures of 14.8
 kg/cm2 (210 psig) provided by stored, recirculated water.  At the end of
 the cycle, the sludge and filtrate lines were blown out by compressed air
 and the diaphragms were gravity drained and returned to their original shape
 by vacuum suction.

      Cake discharging and cloth washing operations began at the conclusion
 of the squeezing cycle.   As shown in Figure 18,  these operations were com-
 pletely different from those of the NGK press.   The Lasta unit released all
 four cakes simultaneously by a "traveling" motion of the filter cloths.   The
 cloths moved downward around the bottom of the plates in a U-turn  fashion
 which caused the cakes to release.   After the discharging was completed,  the
 cloths then returned to  their original positions.   (The press also had doctor
 blades located at the bottom of the plates to assist in difficult  cake
 releases.)   The filter cloths moved downward  a second time for washing on
 both sides by low pressure,  7 kg/cm2 (100 psig),  spray showers located near
 the bottom of the press.   Drip pans were closed  over the discharge port  in
 order to  catch spent wash water and prevent rewetting of the  filter cake.

      Standard laboratory  analyses were performed on  samples of the sludge
 cake,  and filtrate.

 Test  data—
      Over the two week test period,  35 runs were made on this  press;  the
 results are  presented  in  Table 12.   For each  batch of sludge,  two  to  three
 runs were usually made at  varying cycle times to optimize  the  yields  and
 cake  solids.   For the 2/1  sludge mixture  tested,  these runs are clustered
 together  in  Table 12; for  the  100%  secondary  sludge  tested, typical runs at
 different conditioning levels  are shown.  As  shown by  this  data, the  press
 performed quite well and produced cake  solids of at  least  35%  in most cases.
 At  this time,  though, the  feed solids  content was high,  and the sludge was
 easily filtered,  even at low chemical  dosages. (Yields were calculated by
 adding a  full-scale mechanical time of  10.5 minutes to the process cycle
 time.)

     Three different filter cloths, with the following specifications, were
 tested on the press:
                                                                AIR
                                              THICKNESS     PERMEABILITY
TYPE          CONSTRUCTION      FILLING           ^        Cm3/min/cm2

 891           2x2 twill   polypropylene       1.46           1500
 920           2x2 twill   polypropylene       1.17            800
 940           2x2 twill   polypropylene       1.02           2400
                                      53

-------
TABLE 12.  LASTA RUNS
DATE
10-25-77
10-25-77
10-26-77
10-26-77
10-26-77
10-26-77
10-26-77
10-26-77
10-27-77
10-27-77
10-27-77
10-27-77
10-27-77
10-27-77
10-27-77
10-28-77
10-28-77
10-28-77
10-28-77
10-28-77
% CHEMICALS
LIME/FeCl3
19.3/6.4
19.3/6.4
25.7/8.5
25.7/8.5
25.7/8.5
15.8/5.3
15.8/5.3
15.8/5.3
12.0/4.0
12.0/4.0
12.1/4.0
12.1/4.0
12.1/4.0
11.9/4.0
11.9/4.0
19.7/6.6
19.7/6.6
15.0/5.0
15.0/5.0
15.0/5.0
CYCLE TIME
(MIN)
PUMP/ SQUEEZE
8/8
7/10
15/16.5
10/17
5/15
15/17
10/17
5/15
20/30
15/30
10/25
15/30
5/25
5/18
5/22
5/20
10/25
10/25
15/30
5/20
% CAKE
SOLIDS
35.1
33.2
34.8
34.6
36.1
35.4
34.8
38.0
38.0
39.7
40.7
38.5
39.9
35.3
33.6
41.9
42.5
40.9
42.5
43.6
CAKE
THICKNESS
(mm)
12
11
20
18
15
20
15
12
18
12
10
8
6
9
7
12
15
14
13
9
FULL SCALE
YIELD
(kg/hr/m2)
4.05
3.37
3.37
3.56
3.56
3.61
3.03
2.64
2.69
2.15
1.32
1.42
1.42
1.54
1.23
3.12
3.03
2.82
2.61
2.99
                                (continued)

-------
                                 TABLE 12.   CONTINUED


DATE
11-2-77
11-2-77
11-2-77
*10-31-77
*ll-l-77
*ll-l-77

% CHEMICALS
LIME/FeCl3
20.4/6.8
20.4/6.8
20.4/6.8
15.8/5.1
19.7/6.6
25.2/8.3
CYCLE TIME
(MIN)
PUMP/ SQUEEZE
5/15
5/20
8/20
5/20
3/18
5/15

% CAKE
SOLIDS
39.2
40.5
41.6
42.6
37.3
35.2
CAKE
THICKNESS
(mm)
11
11

10
11
FULL SCALE
YIELD
(kg/hr/m2)
3.37
3.03

o f.-\
2.99
3.92
*100 % Secondary Sludge

-------
Ul
Jfe
                                  Filter cloths
                          ix/itz^	7
                     /	1> jytBylir..|../N'   /
                Filtering
                chamber
                Pressate
                                    Feed slurry
         Diaphragm
                          o    o
                         FILTERING I
                                                   CAKE DISCHARGE III
                                                                                      •Showers •
                                                 WASHING OF FILTER CLOTHS
                           Figure 18.  Schematic of  filtration, discharge, and washing in the
                                        Lasta Press.   (Source:  Ingersoll-Rand)

-------
 The best  filtrate  quality was  obtained  using  the  891  cloth;  suspended  solids
 averaged  87.5 mg/1 for  runs made on  the 2/1 sludge.   Cake discharge was  good
 for all three cloths, but sufficient runs were not made to evaluate resistance
 to wear and abrasion.

 Comparison with NGK press—
     During the test period, simultaneous runs were made on  the NGK press.
 Sludge was prepared in  the NGK mix tank; a portion was pumped to the Lasta
 conditioning tank   and  the remainder was fed  to the NGK unit.

     In Table 13,  comparable runs for the 2/1 sludge  mixture are shown.  The
 full-scale NGK yield assumes a 19 minute mechanical cycle  with cloth washing
 every 20  runs.  The full-scale Lasta yield assumes a  10.5 minute mechanical
 time  with cloth washing every four runs.  "Equivalent full-scale yields"
 were calculated for cycle times at which the  lowest cake solids were achieved
 for either press.   The  performance of both presses, as shown by the average
 cake solids achieved, was essentially equal.  The main advantage of the Lasta
 press was its shorter mechanical time (10.5 min vs. 19 min), positive cake
 discharge, and ease and speed  of cloth washing.  Additionally, the optimum
 cycle on  the Lasta  unit usually had a shorter pump time than the NGK press,
 which resulted in  a thinner cake for discharge.

     The main disadvantage of  the Lasta unit  is the quantity of total filtra-
 tion area which is  currently available on the full-scale Lasta unit.  When
 comparing equivalent yields, the Lasta unit was much  higher, averaging 3.31
 kg/hr/m2 as compared to 2.70 kg/hr/m2 for the NGK press.  This difference
 represents an additional 22.6% filtration area that the NGK unit would require
 in order to dewater the same quantity of sludge to the same cake solids as the
Lasta press.   However,  the largest NGK press has 145% more filtration area
available than the largest Lasta unit (NGK-500 m2; Lasta-204 m2);  therefore,
fewer NGK units would be required.
                                      57

-------
                                   TABLE 13.  COMPARISON RUNS ON 2/1 SLUDGE
m
oo
% CHEMICALS
DATE
10-25
10-26
10-26
10-27
10-27
10-28
11-2
LIME/FeCl
19.
25.
15.
12.
12.
19.
20.
3/6.4
7/8.5
8/5.3
0/4.0
1/4.0
7/6.6
4/6.8
EQUIVALENT
FULL-SCALE
% CAKE EQUIVALENT YIELD
SOLIDS % CAKE (kg/hr/m2)
NGK
37.1
41.7
36.9
35.5
36.2
40.5
39.8
LASTA
35.1
34.6
35.4
38.0
40.7
42.5
39.2
SOLIDS
35.0
34.6
35.4
35.5
36.2
40.5
39.2
NGK
3.15
2.93
2.82
2.32
2.05
2.96
2.68
LASTA
4.05
3.56
3.61
3.08
2.15
3.37
3.37
EQUIVALENT
CYCLE TIMES
(min)
PUMP /SQUEEZE
NGK
19/15
19/5
18/14
18/20
15/21
20/16
20/16
LASTA
8/8
10/17
15/17
20/23
10/21
5/19
5/15
FILTRATE
SUSPENDED
SOLIDS
(mg/1)
NGK
26
37
69
13
100
_
46
LASTA
67
26
43
21
71
44
400
       Avg.
17.9/5.9
38.2
37.9
36.6
                          2.70
3.31   18.4/15.3 10.4/17.1  485
                                                                                        96

-------
 FIXED VOLUME FILTER PRESS

      The fixed volume press is a standard recessed plate filter press  which
 produces cakes of a constant thickness.   These presses are designed to operate
 at terminal pressures ranging from 7 kg/cm2 (100 psig) to 15.8 kg/cm2 (225
 psig).   For the study, a 225 psig unit was supplied by Passavant Corp., and
 a 100 psig unit was supplied by Neptune  - Nichols Inc.  This report refers
 to the 100 psig press as a "low-pressure unit"  and the 225 psig press as a
 "high-pressure unit".   Both presses were operated to develop design data for
 dewatering a variety of sludge ratios, but most work was centered on the 2/1
 secondary to primary ratio.   The units were also used for comparison tests
 with each other and with the diaphragm type press.

 High-Pressure Press

 Facilities—
      The Passavant  system included the following equipment.

      1.   Press - Passavant Model 2400  -  Produced up to 6 circular cakes,  each
          597 mm (23.5 inches)  in diameter.   Each chamber had a filtration
          area of 0.56 m2  (6.0  ft2).  Several stainless steel plates  were
          supplied to provide chamber thicknesses of 30 mm (1.18  inch),  34 mm
          (1.34 inch),  and 38 mm (1.50  inch).   A hydraulic mechanism was
          provided for press  closing.   (See  Figure 19.)

      2.   Feed tank  - 1135 1  (300 gal)  cylindrical closed tank; capable  of
          withstanding air pressures  up to  21 kg/cm2  (300 psig).

      3.   Air compressor - operated at  pressures  up  to  21 kg/cm2  (300 psig)  for
          feeding the press and  core  blowing.

      4.   Filtrate collection tank -  378  1  (100  gal)  calibrated plastic vat.

 Operation—
      For  all  runs,  the sludge was blended and conditioned  in  the  NGK mix  tank
 prior to  pumping to  the feed tank.   Before  each  run  the  cloths were  wetted
 with  tap  water and  scrubbed  with a stiff-bristle nylon brush.  The press was
 then  closed hydraulically.   The sludge feed valve was opened  and  the high-
 pressure  compressor  started.  Within 15 to  20 minutes the  full pressure of
 15.8  kg/cm2  (225 psig) was attained and was held for the remainder of the run,
 thus  providing the  sole driving force for dewatering the sludge.  The run
was ended when either the filtrate rate reached 0.1 gal/hr/ft2 or when three
hours filtration time had elapsed.  Usually 3 cakes were made, but at times
when  greater quantities of feed sludge were available, up to 5 cakes could be
produced.  At the end of the run all cakes were weighed and analyzed for
percent solids.  Generally, the availability of laboratory oven space per-
mitted no more than one or two cake samples for analysis for percent solids.
Data sheets 5 and 6 in Appendix B summarize a typical run on the high-pressure
press.  Explanations are provided with each data sheet.
                                      59

-------
Test Data-
     All test runs with the high pressure-press were conducted in August and
September, 1977.  During this time, sludge temperatures ranged from 24°C to
30 °C.  In August, the feed before conditioning averaged 5.7% solids  and
the sludge dewaterability was good.  In September, the feed averaged from
3.4 to 4.0% solids before conditioning and the dewaterability was poor.  Thus,
higher chemical dosages were required to filter the sludges than during
August.

     The data in Table 14 are typical results with this press for a variety
of secondary to primary sludge ratios.  These tests were all made in August
using a 38 mm (1.5 inch) cake.  The filter cloth used was a nylon monofilament
of twill weave, with an air permeability of 76.7 cm3/s/cm2 @AP= 12.7mm
The full-scale yields were computed by adding 20 minutes mechanical turn-
around time to the process cycle time (based on the manufacturer's recommen-
dation for their largest press).  As the ratio of secondary sludge increased,
the chemical requirements increased, and the cake solids and yields decreased.
Because of the open weave cloth on this press, the filtrate suspended solids
were sometimes high, particulary when the sludge was marginally conditioned.
The average cake density for the runs was 1123.0 kg/m3  (70.1 lb/ft3).  Cake
discharge from this press was not the best; a thin mat of fibrous sludge
remained on the cloth after each run, especially around the center feed hole.
No precoat was used for any of these runs.

     The cake always showed a very dry outside portion and a much wetter inner
section.  Because of the limited size of the air compressor receiver, the
core blow at the end of each run was generally ineffective in removing all
the solids from the center feed hole.  During cake sampling, pie shaped pieces
were taken which included the proper proportions of this wet inner core and
the dry outside section.  To insure that these slices were indeed represen-
tative, the variation in dryness across the cake was periodically checked.
A wedge was divided into four quarters as shown in Figure 20, and each quarter
was analyzed for  percent solids.  These results were:
 Date

 8/15
 8/16
 8/17
 8/18
 8/19
 8/23
17.4
41.6
19.7
21.6
26.1
16.1
           solids  (section)
             II     III        IV
21.7
47.4
33.0
28.7
35.9
23.4
37.7
52.1
43.9
41.2
41.7
36.5
39.3
49.0
42.5
41.9
42.3
40.8
                              Theor.
                              % solids
32.
48.
38.
36.
39,
32.8
  % solids of
adjacent wedge

     31.2
     47.5
     37.0
     36.6
     37.4
     33.6
 Section I was 10.6% of the total volume;  section II was 22.8%;  section III was
 35.1%; and section IV was 31.5%.  The theoretical percent solids was
 calculated by multiplying these percentages by the percent solids in each
 section.   The correlation of the last two columns is quite good, indicating
 that our method was correct.  These results also show that a standard recessed
 plate press will always have a variation in percent solids across the cake.
                                       60

-------
        Figure 19.  Passavant Filter Press.
Figure 20.  Sample sections from Passavant cake.
                      61

-------
                  TABLE 14.  TYPICAL RESULTS ON MODEL 2400 HIGH-PRESSURE  (38 mm PLATE)  -  AUGUST RUNS
ro

SEC /PRIM
RATIO
0/1
1/2
1/1
2/1
3/1
4/1
1/0

NO. OF
RUNS AVERAGED
2
2
3
13
5
3
4

% CHEMICALS
LIME/FeCl3
13.4/4.5
13.2/4.4
15.6/5.2
18.3/6.1
22.7/7.6
21.4/7.2
23.5/7.8
PROCESS
CYCLE
TIME (MIN)
160
125
140
169
146
137
148

% CAKE
SOLIDS
47.6
35.9
38.6
34.3
33.2
36.9
29.5

FULL-SCALE
YIELD (kg/hr/m2)
3.37
2.54
2.69
1.94
2.04
2.39
1.70
FILTRATE SOLIDS
mg/1
TOTAL/ SUSPENDED
6305/36
7818/36
7930/37
7610/260
-10399/384
7724/43
7770/324

-------
 With the more difficult filtering sludges this variation was even greater as
 shown by the data on 8-15 (100% secondary), and 8-17 through 8-23 (all 2/1
 sludge).  The 8-16 run on 100% primary showed only slight variation because
 this sludge was easily filtered.

      Table 15 presents the results of runs on the 2/1 secondary to primary
 sludge.   When the sludge was properly conditioned, the high-pressure press
 gave acceptable cake solids (over 35%),  using the 38 mm cake,  with cycle
 times approaching three hours.  Marginally conditioned sludges as shown on
 8/4, 8/19,  and 9/1 gave very poor results for this thick cake.  In September,
 two new  plates were installed to  provide cakes measuring 30 mm (1.18 inches).
 The end  plates could not be changed,  however.   So on subsequent runs,  the two
 inside cakes measured 30 mm (1.18 inches),  while the two outside cakes
 measured 34 mm (1.34 inches).   When the  cakes  could be weighed and analyzed
 separately, a yield was computed  for  each thickness (runs  on 9/21 and  9/22).
 The runs after 9/14 show that  the thinner cakes always contained slightly
 higher cake solids,  but with some measurable  sacrifice in  overall yield.
 No  comparisons could be made between  the runs  in August and those in September
 because  the sludge filterability  had  changed so drastically.

      During August,  1977,  additional  high pressure runs were made on a
 Passavant Model  600  bench-scale press.   This press,  152 mm (6  inches)  in
 diameter, could  produce cakes of  various  thicknesses  from  25 mm to  38  mm.
 The press was  fed  from a small  tank pressurized with  nitrogen  to  15.8  kg/cm2
 (225 psig).  For most  runs,  the conditioned sludge was  sampled from  the NGK
 mix tank.   Results are  presented  in Table 16.   In general,  the tests showed
 that higher cake solids  were produced by  the thinner  cakes  (see the  compari-
 son tests on 8/17, 8/18,  and 9/1).  Any  change in full-scale yield because
 of  the different cake thicknesses was not readily apparent  from this data.

      Some comparison runs between the Model 600  and Model 2400  presses were
 conducted for  the  38 mm (1.5 inches)  cake:
  DATE

8-17-77
8-18-77
8-30-77
8-30-77
9-1-77
   CYCLE TIME
     (rain)
M-600  M-2400
   90
  120
  115
  110
  130
200
190
180
180
180
              YIELD
            (kg/hr/m2)
         M-600  M-2400
3.12
2.54
3.37
2.83
1.81
1.47
1.54
1.54
                       % CAKE
                       SOLIDS
                  M-600   M-2400
37.8
35.8
26.7
39.2
35.9
37.0
36.6
28.4
31.9
29.3
     The above table shows that with the Model 600 press,  cycle times were much
 shorter and the resultant yields were much higher.  In some cases higher
 cake solids were also achieved with this unit.  No satisfactory explanation
 has yet been given for these apparently inconsistent results.

     The scale-up factor (filtration area per plate) for the Model 2400 to
 full-scale is only 12.9 to 1.0,  while for the Model 600,  it is 198 to 1.

                                      63

-------
TABLE 15.  RUNS ON MODEL 2400 HIGH-PRESSURE PRESS WITH 2/1 SECONDARY/PRIMARY SLUDGE.

DATE
8-2-77
8-2-77
8-4-77
8-17-77
8-18-77
8-19-77
8-19-77
8-23-77
8-24-77
8-25-77
8-25-77
9-1-77
Average
9-13-77
9-14-77

9-14-77

9-21-77

9-22-77


% CHEMICALS
LIME/FeCl3
21.9/7.4
17.9/5.9
14.7/4.9
18.0/6.0
19.6/6.6
21.5/7.2
14.2/4.8
16.6/5.5
18.3/6.2
20.1/6.7
19.3/6.5
17.9/6.0
18.3/6.1
20.1/6.7
24.3/8.1

29.2/9.7

27.7/9.2

30.0/10.0


CYCLE
TIME (MIN)
120
150
120
200
190
180
180
180
190
160
180
180
169
180
180

180

140

160


% CAKE
SOLIDS
37.4
36.1
28.4
37.0
36.6
37.4
29.6
33.6
34.1
36.2
35.4
29.3
34.3
28.7
30.1
33.6
36.2
37.3
34.4
36.4
35.9
36.2
CAKE
THICKNESS
(mm)
38
38
38
38
38
38
38
38
38
38
38
38
38
30-34
34
30
34
30
34
30
34
30

CAKE
DISCHARGE
excellent
excellent
not noted
good
good
good
fair- good
fair- good
fair
not noted
good
poor

not noted

good

good

good

good

FULL SCALE
YIELD (kg/hr/m2)
2.73
2.25
2.06
1.81
1.81
1.94
1.60
1.73
1.63
2.26
1.88
1.54
1.94
1.10
1.27

1.48

1.85
1.79
1.67
1.45

-------
                           TABLE 16.   RUNS ON MODEL 600 HIGH-PRESSURE PRESS

DATE
8-17-77
8-17-77
8-17-77
8-17-77
8-18-77
8-18-77
8-23-77
8-24-77
8-24-77
8-25-77
8-30-77
8-30-77
8-30-77
8-30-77
9-1-77
9-1-77

RATIO
SEC/PRIM
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
3/1
3/1
3/1
3/1
2/1
2/1

% CHEMICALS
LIME/FeCl3
18.0/6.0
18.0/6.0
17.5/5.8
17.5/5.8
19.6/6.6
19.6/6.6
16.6/5.5
18.3/6.2
30.0/10.0
20.1/6.7
14.5/4.9
14.5/4.9
28.7/9.7
28.7/9.7
20.0/6.7
20.0/6.7

CYCLE
TIME (MIN)
75
90
120
120
120
105
155
130
100
90
115
110
110
90
130
130

% CAKE
SOLIDS
43.1
37.8
41.7
35.9
35.8
40.4
40.4
39.6
32.6
24.1
26.7
27.0
39.2
40.3
39.0
33.9
CAKE
THICKNESS
(mm)
25
38
32
38
38
32
32
32
32
32
38
32
38
32
32
38

FULL SCALE
YIELD (kg/hr/m )
N.A.*
N.A.
N.A.
N.A.
3.12
3.17
2.39
2.64
2.20
2.05
2.54
2.00
3.37
3.38
2.57
2.83
*Not available.

-------
 Therefore, Model 2400 data was used for comparisons with the other presses
 and for full-scale design.  As with the other presses,  scale-up  from pilot
 data to the full-scale press can be made directly.   The process  cycle time
 (filtration time) is assumed to be identical for both size units.   The
 total cycle time, however, must be adjusted to include the mechanical turn-
 around time in order to obtain the full-scale yield.

 Low-Pressure Press

 Facilities—
     The Nichols system included the following equipment:

     1.  Press - 0.37 m2 (4 ft2) filtration area.  Produced two octagon shaped
         cakes each measuring 330 mm (13 inches) across.  Plates  were rubber
         coated steel, with a chamber thickness of 25 mm (1.0 inch).  Spacers
         were available to produce a cake thickness of 32 mm (1.25 inches).
         The press was closed by a manually operated screw.  (See Figure 21.)

     2.  Feed tank - 113.5 1 (30 gal.) cylindrical closed tank; could be
         pressurized with air up to 10.5 kg/cm2 (150 psig).

Operation—
     For all the runs on this press, the sludge was blended and conditioned
in the NGK mix tank  prior to pumping to the feed vessel.  Before each run,
the cloths were wetted with tap water and scrubbed with a stiff-bristle nylon
brush.  The press was closed and sealed as tightly as possible by manually
turning the screw.  The inlet valve to the press was opened and the feed tank
pressurized slowly with air to reach a pressure of 7 kg/cm2 (100 psig) within
5 to 10 minutes.  This pressure was maintained throughout the entire run and
provided the sole driving force for dewatering.  Filtrate was collected from
a drain pipe and a drip pan under the plates during the run.  Because there
were no gaskets between the filter plates, and the filter cloths provided the
only seals, up to 50% of the filtrate was collected from the drip pan.  Early
test work established that the run was complete when the filtrate rate reached
25 ml/min or less.  At the end of the run both cakes were weighed and sampled.
A triangular shaped section, as shown in Figure 22, was taken and analyzed for
percent solids.  Data Sheets 7 and 8 in Appendix B summarize a typical run on
the low-pressure press.  Explanations are provided with each data sheet.

Test Data—
     Test work with the low-pressure press throughout the year showed that
the press could dewater sludges over a range of sludge temperatures from
11 °C to 30 °C.  The data in Table 17 are typical results with this press
on a variety of secondary to primary sludge ratios with an average
unconditioned feed solids concentration of 5.7%.  These tests were all
conducted in August during comparison studies with the high-pressure unit
and the diaphragm press.  The sludges tested at that time contained a high
proportion of septic solids; however, their dewaterability was quite good
even at low chemical dosages.  The full-scale yield was computed h-y adding
20 minutes mechanical turn-around time to the cycle  time (based on- the
manufacturer's recommendation for the largest press  available).  Attempts  to
run 100% primary  sludge on this small pilot press failed because solids

                                      66

-------
       Figure 21.   Nichols Filter Press.
Figure 22.  Sample sections from Nichols cake.
                     67

-------
00
                          TABLE 17.  TYPICAL RESULTS ON LOW PRESSURE PRESS - AUGUST RUNS

SEC/PRIM
RATIO
1/2
1/1
2/1
3/1
4/1
1/0

NO. OF
RUNS
2
3
13
5
3
3

% CHEMICALS
LIME/ Fed 3
13.2/4.4
15.6/5.2
18.3/6.1
22.7/7.6
21.4/7.1
23.1/7.7
PROCESS
CYCLE
TIME '(MIN)
128
127
140
140
113
143

% CAKE
SOLIDS
38.5
39.0
35.0
34.1
35.6
32.3

FULL-SCALE
YIELD (kg/hr/m2)
1.61
1.59
1.39
1.32
1.46
1.08
FILTRATE SOLIDS
mg/1
TOTAL/ SUSPEND ED
7382/44
8532/57
8082/50
10267/69
7786/36
7125/49

-------
 plugged the small (1 inch diameter) feed line.

      The general trend, as shown in Table 17, was that as the ratio of second-
 ary sludge increased, the chemical requirements increased, and the cake
 solids and yields decreased.  The exception to this was the 4/1 sludges which
 seemed to filter extremely well.  The filtrate quality was unaffected by
 changes in the sludge ratios.  The cloth used for all these runs was the
 Nichols 4709/40 cloth (a monofilament fabric with a 2 x 2 twill weave, and
 an air permeability of 20.3 cm3/s/cm2 @Ap = 12.7 mm H20) .  The average cake
 density was 1141 kg/m3 (71.2 Ib/ft3).  Except for the marginally conditioned
 sludges, cake release from the cloth was generally very good.   But, as with
 the high-pressure units,  the cake from this press was drier on the outer
 sections than at the inner core.

      Table 18 presents the results of the 13 individual runs on the 2/1
 secondary/primary sludge.   The first three runs in the table,  8/2/77 and
 8/4/77,  show the effect of chemical conditioning.   Low chemical dosages
 (marginal conditioning)  as in the run on 8/4/77 gave poor results  on this
 press.   The first eight runs in the table were the results with the 25 mm
 (1 inch) thick cake;  the last 5 were with the 32 mm (1.25 inches)  thick cake.
 Increasing the cake  thickness to 32 mm (1.25 inches) provided  some tradeoffs.
 The averages of the  two  sets of runs showed that resultant cake solids were
 slightly lower with  the  thicker cake,  but the overall full-scale yields were
 nearly  identical at  approximately 1.4 kg/hr m2 (0.29 Ib/hr ft2).   The in-
 creased  cake thickness  also  required increased cycle times.

 Comparison Runs

      The comparison  runs  in  August,  1977, were designed to establish  the
 operating conditions  for  the three  types  of presses— low-pressure fixed
 volume,  high-pressure fixed  volume,  and  diaphragm.   These  tests were  run
 on seven different secondary to  primary  sludge ratios.   With each  ratio,  at
 least three  runs  were made:   one with  the sludge over-conditioned;  one with
 the sludge conditioned in  a  good, safe range;  and  one with the  sludge
 marginally conditioned.  The lime/FeCl3 dosages  required to produce the above
 conditions were determined from  experience;  filterability  was checked by
 specific resistance and CST  tests prior to  each  run.

 Facilities—

     1.  Diaphragm press - NGK unit 5.8 m2  (62.4 ft2) filtration area.

     2.  High-pressure fixed volume -  Passavant Model 2400 with 1.67  m2
          (18 ft2) filtration area (3 cakes).  All  tests used the 38 mm plates.

     3.  Low-Pressure fixed volume - Nichols unit  -  .37 m2  (4 ft2)  filtration
         area.  Cake thickness was either 25 mm or 32 mm.

Operation—
     Thickened primary and secondary sludges were  independently pumped to the
NGK mix tank and blended in the proper proportions for the test.  Percent
solids of each sludge were determined prior to blending to be certain  that


                                      69

-------
TABLE 18.  RUNS ON LOW-PRESSURE PRESS  WITH 2/1  SECONDARY/PRIMARY  SLUDGE


DATE
8-2-77
8-2-77
8-4-77
8-17-77
8-19-77
8-19-77
8-25-77
9-1-77
Average
8-17-77
8-18-77
8-23-77
8-24-77
8-25-77
Average

% CHEMICALS
LIME/FeCl3
22.1/7.4
17.9/5.9
14.7/4.9
18.0/6.0
21.5/7.2
14.2/4.8
20.1/6.7
17.9/6.0
18.3/6.1
17.5/5.8
19.6/6.6
16.6/5.5
18.3/6.2
19.3/6.5
18.3/6.1
PROCESS
CYCLE
TIME (MIN)
100
120
140
110
140
110
130
150
125
180
150
180
170
140
164

% CAKE
SOLIDS
37.5
37.5
31.0
35.3
35.1
34.1
36.2
35.8
35.3
34.4
34.5
32.5
34.0
36.4
.34.4
CAKE
THICKNESS
(mm)
25
25
25
25
25
25
25
25
25
32
32
32
32
32
32

CAKE
DISCHARGE
ecxellent
excellent
fair - good
excellent
excellent
excellent
excellent
good

good
excellent
good
excellent
good


FULL SCALE
YIELD (kg/hr/m2)
1.71
1.51
1.12
1.51
1.17
1.56
1.32
1.22
1.39
1.27
1.51
1.22
1.32
1.66
1.40

-------
the blend was accurate.  FeCla solution at one Ib/gal was added and allowed to
mix in.  Lime solution at one Ib/gal was then added and mixed in.  Within 15
minutes after adding the lime, portions of the conditioned sludge were pumped
to the Passavant and Nichols feed tanks.  The remainder was fed to the NGK
press.  The three presses were all started at the same time.  The operational
procedures, (i.e. cycle times, pressures, etc.) were those described
previously.

Test Data—
     Individual test runs on the three presses are presented in Table 19.  All
runs on the 2/1 sludges are included, while only the best runs with average
conditioning are presented for the other sludge ratios.  The full-scale yields
were calculated with these turn-around times:  NGK - 19 minutes, Passavant -
20 minutes, and Nichols - 20 minutes.  Several important conclusions were
derived from this table:

     1.  With a properly  conditioned sludge,  all three types of presses pro-
         produced  the required 35% cake solids.

     2.  On the average,  the diaphragm press  gave both higher cake solids
         (40.0% vs 34.3%  and 34.9%) and higher yields than the fixed volume
         presses.  Cake solids were approximately the same on both the high
         and low pressure presses,  but the high-pressure press gave signi-
         ficantly greater yields  than the low-pressure unit.

     3.  The diaphragm press was  the only unit capable of satisfactorily
         dewatering the marginally  conditioned sludges.   For example,  on
         8/4 and 8/30 both the Passavant and  Nichols presses  had poor runs,
         with wet,  sloppy cakes and extremely low yields.   With the same
         sludge the diaphragm press gave a good cake discharge and high cake
         solids,  but at reduced yields.   The  diaphragm press,  because  of its
         separate squeezing cycle,  provided a much more flexible operation.
         These poorly conditioned sludges  were pumped for shorter cycles
         and the squeezing time was increased slightly to give thin,  dry
         cakes.   The fixed-volume presses  did not  have this option,  so once
         the sludge was fed to  these presses   no  corrective measures  could
         be taken.

    4.   The marginally conditioned runs on 8/19  and 9/1  gave  poor cake solids
         on the  Passavant  press,  but  acceptable results on the Nichols press.
         This  indicated that  cake thickness had more of an effect  than
         pressure in determining  cake solids  content.

    5.  As  the percentage  of primary sludge  increased, the cake  solids  and
        yields  improved for all  presses.  Cake solids approached  the  50%
         solids level for high primary ratios.

    6.  For the properly conditioned 2/1 sludge runs  the average  filtrate
        suspended  solids were:   NGK  - 197 mg/1; Passavant - 26.5 mg/1;
        Nichols - 49.1 mg/1.  The calculated percent  recovery of  inlet
        suspended solids in the  filter cake is:  NGK - 99.74%; Passavant -
        99.97%; and Nichols - 99.93%.

                                     71

-------
TABLE 19.  COMPARISON RUNS
FULL-SCALE YIELD

DATE
8-2-77
8-2-77
8-4-77
8-17-77
8-17-77
8-18-77
8-19-77
8-19-77
8-23-77
8-24-77
8-25-77
8-25-77
9-1-77
Average
8-5-77
8-30-77
8-8-77
8-10-77
8-11-77
8-16-77
8-15-77
RATIO
SEC/PRIM
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
2/1
3/1
3/1
4/1
1/1
1/2
0/1
1/0
% CHEMICALS
LIME/FeCl3
22.1/7.4
17.9/5.9
14.7/4.9
18.0/6.0
17.5/5.8
19.6/6.6
21.5/7.2
14.2/4.8
16.6/5.5
18.3/6.2
20.1/6.7
19.3/6.5
17.9/6.0
18.3/6.1
25.5/8.5
14.5/4.9
19.5/6.5
15.2/5.0
15.5/5.2
15.7/5.3
28.1/9.3
% CAKE SOLIDS
NGK
41.4
40.8
40.2
39.9
39.1
40.2
42.2
36.0
36.4
41.4
43.0
40.9
38.8
40.0
42.4
37.2
43.4
47.0
48.2
55.2
41.4
PASS
37.4
36.1
28.4
37.0
—
36.6
37.4
29.6
33.6
34.1
36.2
35.4
29.3
34.3
37.0
28.4
38.8
38.2
40.7
47.5
31.2
NICHOLS
37.5
37.5
31.0
35.3
34.4
34.5
35.1
34.1
32.5
34.0
36.2
36.4
35.8
34.9
35.4
29.3
35.6
39.6
39.0
-
32.6
NGK
3.06
2.83
2.17
2.90
2.69
2.77
3.19
2.12
2.13
2.19
3.07
2.77
2.34
2.63
2.52
1.80
3.25
3.64
4.17
4.92
2.72
(kg/hr/m2)
PASS
2.73
2.25
2.06
1.81
-
1.81
1.94
1.60
1.73
1.63
2.26
1.88
1.54
1.94
2.33
1.47
2.38
2.53
2.74
4.42
1.66
NICHOLS
1.71
1.51
1.12
1.51
1.27
1.51
1.17
1.56
1.22
1.32
1.32
1.66
1.22
1.39
1.27
1.07
1.46
1.95
1.90
-
1.32

-------
      Overall, the diaphragm press performed better than either of the fixed
 volume presses.  But since these presses also produced the required cake
 solids with the 2/1 sludge, the comparison for cost and design purposes
 was based on the respective yields achieved in reaching the 35% solids cake.
 The runs that reached the required cake solids for the low and high-pressure
 presses are shown in Figure 23.  Full-scale yields are plotted vs the date
 of the run  and the eight points are averaged for each unit.   The NGK cycle
 time was then adjusted to a level which would give the same cake solids as
 obtained on the other presses.   This was done by recalculating the squeezing
 time to the point where the desired percent solids were reached.   For example,
 the first NGK run on 8/2 had a  pump/squeeze time of 21/16 minutes and a final
 solids of 41.4%.   The Passavant and Nichols cake solids were  37.4%.   Filtrate
 collection data for the NGK run showed that this 37.4% solids level  was
 reached after 9 minutes of squeezing.   Therefore,  a recalculated NGK cycle
 time of 21 minutes pumping, 9 minutes  squeezing, and 19 minutes mechanical
 time was used,  and a full-scale yield  of 2.94 kg/hr/m2 obtained.   These
 adjusted NGK yields are also plotted in Figure 23  and the eight runs averaged.
 To produce a 36.3% solids cake  the NGK average yield was 3 31 kg/hr/m2;
 the Passavant yield was 2.04 kg/hr/m2;  and the Nichols yield  (for 35.8%
 solids) was 1.46  kg/hr/m2.

      Using the  above yield data,  the filtration area required to  process a
 given quantity  of sludge was then computed for each of the press  types.  For
 example,  to process 1000 kg/hr  of dry  sludge  solids,  the NGK  press would
 have required:

                    1000 kg/hr = 302  m2  of  filtration area.
                    3.31  kg/hr/
Likewise, the Passavant and Nichols presses would have required 490 m2 and
685 m^ of filtration area, respectively.  Thus using the NGK press as a base,
the Passavant unit requires 62.3% more filter area, and the Nichols press
requires 126.8% more filter area than the NGK press to dewater the same
quantity of sludge.  It must be noted that this relationship was derived
specifically for the Blue Plains sludge.  It is further noted that the
relative filter areas refer only to the largest press sizes available from
each of the manufacturers:  NGK - 500 m2; Passavant - 1080 m2; and Nichols -
628 m2.  For comparisons of smaller size presses,  a different mechanical
time must be used and a new full-scale yield must  be calculated for each
unit.   For convenience,  the following table shows  the process cycle time
used to derive Figure 23:
                                      73

-------
            Date
    Process Cycle Times
           (min)
 NGK             Pass.
                Nich.
            8-2
            8-2
            8-17
            8-18
            8-19
            8-24
            8-25
            8-25
21/9
20/9
16/11
18/11
20/8
21/12
18/7
17/10
120
150
200
190
180
190
160
180
100
120
110
150
140
170
130
140
See Table F-l  in Appendix F for specifications of the large-scale presses
available from each of the manufacturers.
                  a.a r-
                  3.4
                  3.8
                  3.0
                 B
                 I

                 0
                 a
                 J BB
                 ,,8.0
                 j
                 U i.m
                   1.1
                   1.O
                                         IMOK YISLD 6 3B.3 "to SOI-IDB
                                               NICHOLS YISLD 8
                                                3B.B "to •OLIOS
                         •/a a/a s/17 s^is s/is s/aa S/BB S/HB
                                       DATS
                     Figure 23.  Comparative yield data.
                                       74

-------
 CONTINUOUS BELT FILTER PRESSES

      Continuous belt presses for sludge dewatering were originally developed
 in Europe and have generally found wide acceptance in many countries.   Several
 companies in the United States have purchased the technology and are now
 marketing these presses in this country.   Two manufacturers'  units were tested
 in the study - a Parkson Magnum press and a Komline Sanderson Unimat press.
 The units were used to dewater both thickened sludge and cake from a vacuum
 filter.

 Parkson Magnum Press^

      The Magnum press was equipped with two continuous screens made of poly-
 ester monofilament cloth (air permeability of 228 cm?/sec/era2 @AP = 12.5  mm
 water) which ran through a system of guiding and pressing rollers that were
 perforated to allow for water drainage.   There were three dewatering zones;
 a gravity drainage stage,  a low-pressure  stage - 0.5 kg/cm2  (7.5 psig), and
 a high-pressure stage - up to 7.7 kg/cm2  (110 psig).   Their  full-scale press
 has a range of belt speeds from 1.2 to  5.7 m/min.   (See Figure 24.)

 Facilities—
      A 0.25 meter wide laboratory press was tested in May, 1977.   A 1.0
 meter wide trailer-mounted demonstration  unit was  tested in  October, 1977.
 A hopper and Moyno pump were supplied with the demonstration unit to test
 the vacuum filter cake as  feed.

 Operation—
      The laboratory unit  (0.25 meter), pictured  in Figure 25, was used to
 provide  basic information  on dewatering polymer  conditioned  thickened  sludge
 blends.   Data was  also  collected  for  the  further dewatering  of vacuum  filter
 cake.  Various ratios  of secondary  to primary  sludge  were blended in
 laboratory glassware,  and  the polymer was  added  and mixed in.  The  conditioned
 sludge was  placed  on the drainage section  of  the belt  press and,  after a
 suitable drainage  time,  the  belts moved the sludge through the pressure zones.
 Yield was  computed  from measurements of cake weight and belt  speed.  Solids
 recovery on  this unit was  estimated from experience.   When testing vacuum
 filter cake,  samples were  taken from the plant's full-scale units and  manually
 placed on  the  press.

     When  testing on the demonstration size press, vacuum filter  cake was
 collected  from the  full-scale filters in a truck, dumped  on the ground, and
 then loaded  in the  feed hopper with a front end  loader.  An open-throat
Moyno pump fed the  sludge  through a six-inch hose and a variable  orifice feed
nozzle onto  the drainage section of the press.

 Test Data—

     Laboratory (0.25 meter) unit—Results of the tests with varying ratios
of secondary to primary sludge are presented in Figures 26 and 27.  With a
feed range of 5.5% to 9.5% total solids, the final cake solids increased
linearly from 25% to 41% as the percentage of primary increased.  The press
                                      75

-------
                                STAGE  3
                            HIGH   PRESSURI
CAKE
 OUT
                                                           STAGE 1
                                                          DRAINAGE
                                                                                       SLUDGE
                        Figure 24.  Schematic of Parkson belt press.

-------
                   Figure 25.  Parkson Laboratory Belt Press.
 capacity  exhibited  an  S-shaped  curve,  ranging  from  248  to  1230 kg/hr/meter
 of belt width  (547  to  2712  Ibs/hr)  for pure  secondary and  pure primary,
 respectively.  Figure  27  shows  that the polymer  (Percol  721  @ $1.70/lb)
 consumption  decreased  from  5.5  to 1.6  Ibs per  ton dry solids, and estimated
 solids recovery  (reflecting losses  in  both filtrate and  washwater) increased
 from  95%  to  98% with increasing percent primary.  With high  primary sludge
 (greater  than  84% by weight) the belt  speeds were near maximum of 5 meters/min
 and high  pressures  of  7 kg/cm2  were attained.  As the percent primary
 decreased, the belt speed was reduced  to 3 meters/min and  pressures of only
 1.8 kg/cm2 (25 psig) were applied.   With the high primary  sludges, cake
 release from the cloth was  excellent as is pictured in Figure 25.  As the
 percent primary decreased,  a sharp  scraper blade was needed  to remove the
 cake.  Some  solids, however, usually remained imbedded in  the cloth and high
 pressure washing was required to remove them.

     The tests with vacuum  filter cake  showed excellent  results.  At that
 time, average  conditioning  chemicals of  19% lime, 6% FeCl3, and 0.14%
 polymer were added to the vacuum filter  feed.  Cake solids from the vacuum
 filter averaged 20%; no additional chemicals were mixed with the belt press
 feed.  The following table  shows how the cake solids varied with throughput
 rate:
Capacity (kg total solids/hr/m)
% Cake Solids
378
 42
702
 39
972
 36
1260
  35
                                     77

-------
                       MAGNUM PRESS TEST RESULTS
                       Blue Plains Plant. Washington, D.C.
             0


            100
10  20  30  40  50  60  70  80  90  100
         %Primary(wt.% dry solids)

90  80  70  60  50  40  30  20  10
        % Secondary (wt.% dry solids)
Figure 26.   Results  of  tests with varying ratios
               of  secondary to  primary  sludge
               (Parkson Corporation).
                    MAGNUM PRESS TEST RESULTS
                    Blue Plains Plant, Washington, D.C.
         10   20   30   40   50   60   70   80   90   100
                     %Primary(wt.% dry solids)

    100  90   80   70   60   50   40   30   20   10    0
                    % Secondary (wt. % dry solids)
Figure  27.   Polymer  dosage and solids  recovery for
varying ratios  of secondary  to primary sludge
(Parkson Corporation).
                           78

-------
 Solids retention on the belt was estimated at 99%.  Cake release was
 excellent, similar to the release with 100% primary sludge.  These tests
 clearly indicated that a belt press retrofitted to a vacuum filter could
 produce cake solids in the desired auto-combustible range.  Further tests
 were therefore conducted on a full-scale unit.

      Demonstration (1.0 meter) unit—Vacuum filter cake tests on the full-
 scale unit encountered difficulty and the good results with vacuum filter
 cake on the laboratory unit were not duplicated.  Test results are presented
 in Table 20.  All the problems were related to the feeding and distribution
           TABLE 20.  PARKSON PRESS AS A RETROFIT TO VACUUM FILTERS
CAPACITY

RUN NO.
1
2
3
4
5
6
7
8
BELT SPEED
(m/min)
2
3
3
3
3
3
3
1
HIGH PRESSURE % CAKE TOTAL SOLIDS
(kg /cm2)
4.4
3.9
2.1
2.3
1.8
2.1
2.1
2.3
SOLIDS (kg/hr/m
35.
35.
35.
28.
29.
29.
30.
35.
width) REMARKS
5 ~®' 1 Matl. directly
8 323 f from filter
1 376J
0 3391
4 316
0 316
6 303
1 115J



Matl. from
screw conveyor

 of  the  vacuum filter  cake  (cake  solids  at  20%).   In Runs  #1  through #3  in
 Table 20,  cake directly  from the vacuum filter was  used.   The  feed  system
 was that previously described.   The  sticky nature of the  sludge  caused
 it  to hang up on  the  walls  of the hopper and  form a bridge across the pump
 inlet.  Some  wash water  was added to  the hopper  to  facilitate  feeding the
 pump, but  interruptions  of  flow  to the  press  were numerous,  and  the cake  had
 to  be forced  manually into  the bottom of the  hopper.   In  Run #1, the feed
 layer was  too thick   and a  shearing and rolling  effect at the  beginning of
 the high-pressure section resulted.   This  condition caused the screen to
 wrinkle and crease under pressure.  In  Run #3, the  pressure  was  lowered to a
 point (approximately  2.1 kg/cm2)  where  the material would not  extrude from
 the sides  at  the  high pressure roller.   Because  of  these  feed  problems, the
 yields were low in these first three  runs.  Filtrate suspended solids were
 measured at 1328  mg/1 in Run  #1,  thus giving  only a 95% solids retention
 on  the press.

     The vacuum filter cake was  then processed through a  screw feeder in
 order to make the cake more fluid.  This material was easily fed through
 the hopper/pump arrangement.  Runs #4 through #8 show that the material,
 although more fluid,  also became more difficult  to press as  the floe
 deteriorated with the screw action.   The speed on the press had to be reduced
by one-third in order to achieve the 35% cake solids.  Filtrate suspended
 solids increased to approximately 1900 mg/1 and solids retention in the
press was only 93%.   Much more work in developing an acceptable feed system

                                     79

-------
is required in order to use the belt press in this application.

Komline-Sanderson Unimat Belt Press

Facilities—
     The Unimat press was similar in concept to the Parkson press.  The
Unimat press, pictured in Figure 28, had four dewatering zones:  a gravity
drainage stage; and low, medium, and high pressure stages.  Pressures in the
high-pressure section were in excess of 2.1 kg/cm2 (30 psig).   The trailer
mounted unit tested was their GM2H -  5/7 pilot plant model, with an
effective width of 0.5 meter.

Operation—
     For the thickened sludge tests, the primary and secondary sludges
were thickened separately and blended in an 11.4 nr* (3000 gal) tank to
produce a 2/1 secondary to primary sludge solids ratio.  The blended
sludge was metered to a flocculation tank and polymer was added prior to
feeding the press.  The final cake and filtrate were analyzed for total
solids.  Yield was determined from measurements of the total solids and
flow rate of the feed.

     For the vacuum filter cake tests, a truck-load of 20% solids cake was
taken from the plant's full-scale units and delivered to a point adjacent
to the trailer.  The cake was manually fed to the low-pressure section of
the belt press in bucket loads.  The yield was estimated by counting buckets
per unit time.  During these tests with the vacuum filter cake, problems
were encountered with the motor drive on the press.  At times, the motor was
overloaded and kept kicking out;  a slightly larger motor and drive probably
should have been used to handle this feed.

Test Data—

     Thickened sludge feed—(2/1 secondary/primary).  Table 21 summarizes
the results of the tests run with the thickened sludge.  During this test
period, the plant experienced some upset conditions and the sludge was
septic when received.  The sludge characteristics varied considerably from
days when the polymer would not flocculate the sludge to days when the
same polymer worked very well.  Thus, the results were quite inconsistent.
Laboratory tests were made each morning to determine which of the two
available polymers would work.  These results show the range of cake solids
that were achieved, depending on whether the sludge would respond to the
polymer at that time.  The first two runs on 7/26 show that a doubling of
the polymer rate had only marginal results on the final cake solids.  Unlike
other chemical conditioning agents, the polymers appeared to be quite
selective and worked only within a very narrow range.  The last three runs on
7/26 were the best for the entire series, and seem to be representative of
what the belt press can produce with the proper polymer conditioning.  The
overall average results were cake solids of 31 to 33% at a rate of 307
kg/hr/meter of belt width with a polymer cost of approximately $9.00
per ton of sludge solids.  Unfortunately, because the Blue Plains sludge
is so variable, these results would not be obtained every day.  With our
type of sludge, a number of polymers would have to be readily available for

                                      80

-------
                                                         T8
OQ
N3

00
W
fD
m
to
CO

-------
                              TABLE 21.  UNIMAT BELT PRESS RESULTS ON 2/1 SLUDGE
00
K3

DATE
7-19-77
7-20-77
7-21-77
7-26-77
7-26-77
7-26-77
7-26-77
7-26-77

% FEED
SOLIDS
7.1
6.2
5.2
4.5
4.5
5.2
5.2
5.2

POLYMER
Ib/ton D.S.
7.46*
73.0**
68.7**
70.8**
136 **
5.84*
5.84*
5.84*

POLYMER
$/ton
$11.19
$ 9.49
$ 8.93
$ 9.20
$17.68
$ 8.76
$ 8.76
$ 8.76

% CAKE
SOLIDS
24.3
27.6
22.7
27.3
29.2
31.3
32.8
31.9

YIELD
kg/hr/meter width
370
385
322
292
292
307
307
307
FILTRATE ***
SOLIDS (mg/1)
TOTAL/ SUSP END ED
1700/ -
1800/ -
300/ -
900/ -
880/384
880/384
880/384
           * Percol 776
          ** Calgon 2820
          '** includes filtrate and wash water

-------
 immediate use  as  the  sludge  characteristics  changed.  Washwater  flow rates
 were  not  measured during  any of  these  tests,  therefore  solids  recovery
 was not computed.


            TABLE  22.  UNIMAT PRESS AS  A RETROFIT TO VACUUM FILTER



DATE
7-13-77
7-13-77
7-13-77
7-14-77
7-14-77
7-25-77
7-25-77

% FEED
SOLIDS
22.1
22.3
22.3
22.0
22.8
23.0
23.6

% CAKE
SOLIDS
34.3
35.5
30.4
37.6
35.3
33.2
34.1
TOTAL SOLIDS
YIELD
kg/hr/meter width
655
595
947
543
585
613
1399
FILTRATE SOLIDS
mg/1
TOTAL/ SUSPENDED
- / -
- / -
2800 / -
- / -
- / -
1776 / 1244
- / -
     Vacuum filter cake—Table 22 shows results for the vacuum filter cake
feed.Because the vacuum filter cake was originally conditioned with lime
and ferric chloride, and because it was hand fed to the press, these results
are consistent.  The runs on 7/13 show that as the yield (i.e. belt speed)
was increased, the final cake solids decreased. Cake samples taken from the
intermediate-pressure zones during these runs showed that the first two
zones increased the cake solids from 22% to approximately 27%.  The high-
pressure zone, on the other hand, increased cake solids from about 27%
to approximately 35% and was responsible for the majority of the dewatering.
These results look promising and indicate that further tests, perhaps using
only a high-pressure section of the press, are warranted.
                                     83

-------
VACUUM FILTER RETROFIT - ENVIROTECH HI-SOLIDS FILTER

      Envirotech Corporation has developed a retrofit unit for a belt-type
vacuum filter.  This "Hi-Solids" filter is an expression device which
extracts additional moisture from a vacuum filter cake.  It is equipped with
an air compressor and associated controls, in addition to the standard
auxiliary equipment needed for a vacuum filter.  The unit is pictured in
Figure 29.  The filter cloth leaves the drum at its uppermost point and
travels over a stationary grid.  Above the grid is a rubber diaphragm which
applies pressure to the cloth and the filter cake.  A vacuum is pulled on
the bottom  of the grid to carry away extracted moisture.  The operation
of the unit is on a discontinuous cycle.  A typical cycle takes either 5.2
or 7.2 minutes per revolution, corresponding to a 20 second or 40 second
press time.  For a 20 second press time the following sequence occurs:  the
cake forms on the drum for 20 seconds, the drum progresses 1/5 a revolution
                                                 Cloth
                            Hi-Solids
                            Assembly
              Figure 29.  Schematic of Envirotech Hi-Solids Filter
                                      84

-------
 and stops; the cake dries under vacuum for 20 seconds and the drum progresses
 another 1/5 revolution, etc.  When the cake reaches the press zone, it is
 squeezed by the diaphragm for 20 seconds; then it progresses 1/5 revolution
 and is discharged.  The cloth is subsequently washed and the cycle repeated.
 Diaphragm pressure is a maximum of 10.5 kg/cm2 (150 psig).

 Facilities-
       Test work was conducted in April, 1976 on Envirotech's trailer-
 mounted, 3 ft.   diameter by 3 ft. face filter.  Auxiliary equipment included
 a sludge feed tank, three chemical feed tanks, flocculator,  air compressor,
 transfer pumps, vacuum pump, filtrate pump and receiver.

 Operation—
       A vacuum filter leaf apparatus was used to evaluate the proper filter
 media and optimum chemical dosage for each of six different  sludge ratios.
 Candidate filter media were evaluated by conducting five consecutive top
 feed leaf tests.  Selection was based on good cake discharge, filtrate
 quality, and apparent resistance to cloth blinding.  Chemical dosages were
 optimized by determining the minimum concentrations which gave the maximum
 filtrate volume for the total dewatering time.

       For the tests on the Hi-Solids Filter,  primary sludge  was obtained
 from a pilot plant primary clarifier,  and secondary sludge was obtained from
 the plant's secondary clarifiers.   Both sludges  were thickened and delivered
 to  the trailer  feed tank for blending.   Using the predetermined chemical
 dosages, the Hi-Solids Filter was operated at two different  test conditions
 for each sludge:

       Condition                   Press Time                 Cycle Time

           1                       20 seconds                 5.2 min/rev.
           2                       40 seconds                 7.2 min/rev.


 Test  Data—
      The results  of  the chemical  conditioning tests are  plotted vs percent
 secondary sludge  in  Figure 30.  Using  these chemical dosages  for each sludge
 ratio,  performance tests on the Hi-Solids  Filter  were conducted.   The results
 are presented in  Table 23.   Once  again  the results  show  decreasing cake
 solids  and yields  with increasing  ratio of secondary sludge.

      The 2/1 ratio was used  for comparison with a standard vacuum filter
 installation.   Review of the  data  showed that  a vacuum filter  gave a  17%
 solids  cake at  5.2 MPR,  and  17.5%  solids at 7.2 MPR cycle time.   The  Hi-
 Solids  Filter increased  this  to 24 and  25% solids,  but was'not  able to
 achieve  the desired  35%  cake  solids.  Full-scale  yield on a vacuum filter
 with  optimum chemical  conditioning is expected to be 14.6 kg/hr/m2.
 Because  of  the  Hi-Solids Filter attachment, this  yield was not  achieved
 at  the  20  second press  time  (5.2 MPR)  (14.0 kg/hr/m2) and was  even further
 reduced  (11.3 kg/hr/m^) with  the 40 second press  time (7.2 MPR).   Because
 cake solids above the  25% range were not produced, the Hi-Solids Filter
was not  considered as a dewatering option  for the plant.

                                      85

-------
  11
 ID
I
u
«
IL
                 CaCOHl
                                                            18
                                                            IB
                                                                 o
                                                                 0
IS
                                                            10
         10   aa   30  40   so  BO   70   so   so   100

                   »lo SECONDARY  SLUDGE
       Figure 30.   Chemical dosages vs percent secondary sludge.

                         Envirotech tests.
                               86

-------
                                     TABLE  23.   HI-SOLIDS FILTER RESULTS
oo

RATIO
SEC/PRIM
0/1
1/3
1/1
2/1
3/1
1/0
FULL-SCALE YIELD
% FEED
SOLIDS**
6.9
5.4
4.9
3.9
4.1
3.9
% CHEMICALS
LIME/FeCl3
6.5/2.6
11.9/3.3
18.5/5.8
21.0/7.2
19.8/8.1
19.0/8.0
% CAKE SOLIDS
5.2 MPR** 7.2 MPR
40.7
31.0
25.9
24.0
23.8
N.R.***
N.R.
31.6
27.1
25.0
27.0
N.R.
kg/hr/m2
5.2 MPR 7.2 MPR
25.2
21.2
13.0
14.0
14.3
10.1
20.7
17.7
10.4
11.3
11.6
8.4
         *    before conditioning
         **   MPR = minutes per revolution
         ***  N.R.  = not reported
         Vacuum Level for all tests-20 in.  Hg
         Submergence for all tests  -  15%

-------
VACUUM FILTER

     During the August, 1977 comparison runs on the filter presses, samples
of the conditioned sludge were also used to run a series of optimizing tests
on a vacuum filter leaf.  Tests were conducted according to the procedure
described in the Komline-Sanderson Engineering Corporation instruction
manual entitled "Test Leaf Instructions - Rotary Drum Vacuum Filter"
(Document Number KSM029).  The 0.1 ft2 leaf, when used properly, can give
excellent correlations with actual full-scale vacuum filter operation.

     For each sludge sample, at least five different form and dry times
were run to show a range of possible operating conditions.  The best of
each of these runs is shown and compared to the corresponding NGK run in
Table 24.  Generally, vacuum filter performance was adversely affected by
high ratios of secondary sludge; cake, solids above 20% were easily achieved
when the percentage of secondary sludge was 50% or less.  The filter press
usually gave approximately twice the percent solids achieved with a vacuum
filter; however, because of its continuous operation, the vacuum filter
yields were much higher than the filter press yields (i.e. the total filtra-
tion area required was much less for a vacuum filter). In all cases, and
especially with the higher percentages of secondary sludge, the filter
press could more easily dewater varying sludge feeds.  Marginally conditioned
sludges, or the difficult 100% secondary sludges, gave poor results on the
vacuum filter  but gave acceptable yields and cake solids on the filter
press.
                                     88

-------
                             TABLE 24.   COMPARISON RUNS - VACUUM FILTER/FILTER PRESS
00
VO
DATE
8-11-77
8-9-77
8-10-77
8-10-77
7-27-77
8-2-77
8-4-77
8-4-77
8-5-77
8-5-77
8-8-77
8-8-77
8-9-77
8-3-77
8-3-77
RATIO
SEC/ PRIM
1/2
1/1
1/1
1/1
2/1
2/1
2/1
3/1
3/1
3/1
4/1
4/1
4/1
1/0
1/0
% CHEMICALS
LIME/FeCl3
10.9/3.6
19.6/6.5
15.1/5.0
12.1/4.0
26.8/8.9
22.1/7.4
14.7/4.5
25.8/8.6
18.8/6.3
25.5/8.3
29.6/9.9
19.5/6.5
15.2/5.1
22.9/7.6
18.4/6.1
% CAKE
Vac. Filter
23.5
19.4
21.4
21.5
22.2
17.1
21.2
18.6
21.8
19.6
18.2
19.5
18.5
14.2
12.7
SOLIDS
NGK Press
47.9
46.8
47.0
48.5
44.9
41.4
40.2
40.6
38.0
42.4
44.9
43.4
41.0
39.2
37.6
FULL-SCALE
kg/hr/m2
Vac. Filter
19.5
25.9
22.5
21.0
13.2
7.8
13.2
40.0
18.6
27.3
25.4
38.6
6.8
7.3
4.4
YIELD
NGK Press
3.06
3.40
3.64
3.35
3.71
3.06
2.84
2.53
2.24
2.52
3.36
3.25
2.57
1.87
2.79

-------
                                 SECTION 7

                               SPECIAL TESTS
CORRELATION WITH SPECIFIC RESISTANCE

     During the latter part of the study, bench-scale filterability tests
were performed in conjunction with experimental work on the pilot filter
presses.  Capillary suction time (GST), modified Buchner funnel (Rv), and
high pressure (Rp) methods were used to determine the average specific
resistance to filtration of the conditioned sludge mixture and provide
correlations with press performance.  Detailed descriptions of these test
methods are given in Appendix C.

     Samples of the conditioned mixture were taken directly from the NGK
mix tank to insure that the bench-scale tests were made on the same sludge
mixture that was fed to the pilot filters.  The filterability tests were
begun simultaneously with the start of the press cycles.  The GST was
measured first,  followed by the pressure and modified Buchner funnel
determinations,  respectively.

     In Figures 31-33, the results of these tests for the NGK press are
plotted.  These graphs show that a definite empirical correlation existed
between the average specific resistance of the conditioned sludge mixture
and press performance.  In general, press yields decreased as the resistance
to filtration increased  and minimum acceptable filtration for this press,
i.e. a process yield of 3.17 kg/hr/m2 to give cake solids of 35%,  occurred
at Rv = 27 x 1010 cm/g, Rp = 7 and GST = 15 seconds (these values were
obtained using a least squares linear regression analysis of the data in
Figures 31-33).   Results similar to this were also found for each of the
fixed volume presses.

     At a plant similar to Blue Plains where sludge characteristics and,
hence, chemical conditioning demand varies daily, correlations such as
these can provide an invaluable tool for controlling full-scale press
operations.  Tests on the pilot press units showed that the quantitative
measure of the specific resistance gave a good indication of press perfor-
mance, regardless of either sludge blend ratio or quantity of conditioning
chemicals added.  In the calculation of the specific resistance parameter
(See Appendix C.), the effect of these variables is. minimized so.that
consistent values will result.  Notice in the following table that
comparable resistance values have  comparable press yields.  For example,
the run on the 4/1 secondary/primary sludge required a conditioning dosage ot
                                      90

-------
n 10
§


J
m
u
0
E
a
                      I    I    I   I  I  I  I  I
                     3    4   BB7B910


                              H. .1 CIO10 CM/Q]
 I    I    I   I  I

3O  4O BO BO TO
   E
   3
   H  3


   n
   n
                     Figure  31.   Process yield vs Rv.
                          A  .B .B .7 .B JB JO
                                                               B  B  7 8 9 10
                     Figure 32.  Process yield vs Rp.
                                     91

-------
      RATIO     % CHEMICALS                 Rv      GST     PROCESS YIELD
    SEC/PRIM    Lime/FeCl3     Rp      1010 cm/g    sec       kg/hr/m2

      2/1        18.3/6.2    3.02       32.76       12.4        3.12
      2/1        16.6/5.5    3.10       28.71       12.5        3.27
      2/1        14.2/4.8      -        25.86       15.3        3.12
      4/1        29.6/9.9    1.33       10.74       11.0        5.42
      1/1        15.2/5.0    1.28        9.57       12.8        5.81
      1/2        15.5/5.2    0.86       5.20         9.3        6.79
      1/1        18.9/6.4    0.69       4.04         8.4        6.20


29.6% lime and 9.9% FeCl3; specific resistance values of Rp = 1.33, Rv =
10.74 x 1010 cm/g,  and GST =11.0 sec were obtained; and a yield of 5.42
kg/hr/m^ resulted.   In the subsequent run on the 1/1 sludge mixture,
however, a chemical dosage of only 15.2% lime and 5.0% FeCl3 was required,
yet nearly identical resistance values of Rp, Rv, and CSX equal to 1.28,
9.57 x IQlO cm/g, and 12.8 sec, respectively, were obtained.  The resulting
press yield, therefore, was also nearly identical at 5.81 kg/hr/m2.
Because of correlations of this type between the specific resistance and
the press yield, these bench-scale resistance tests provide a quick method
whereby press output can be approximated prior to filtration.  And
within the span of only a few moments for testing, hours which would be
wasted on inadequate filtration can be avoided.

     It is evident 'from the above table, however, that the CST, although
a good indicator of filterability, does not give consistent results.
Other researchers2 have found that the test is extremely sensitive to
the feed solids concentration of the sludge and, hence, is most useful
only when correlated with results from the pressure and modified Buchner
funnel tests.  In Figures" 34 and 35, these correlations, in which the CST
has been corrected for feed solids, are shown.  While the CST would be
the preferred method of determining filterability since it requires only
a few seconds to perform, the considerable amount of data scattering
suggests that this method introduces a significant  error in resistance
determinations.

     Several manufacturers have indicated that for  high pressure filtrations,
the pressure method is preferred.  In Figure  36, however, the correlation
between Rp and Rv shows that the modified Buchner funnel and pressure tests
produced comparable results  (correlation coefficient = 0.902) for our
particular sludge.  This  indicates that the  Buchner test, which requires
less time and is much  easier to operate, can  at  times be used with  equal
accuracy in high-pressure filtration work.  Moreover, where precise
determinations of actual  resistance values are not  required, the modified
Buchner test  can be reduced  to a  simpler form in which only  the quantity
of  filtrate collected  within a given period  of  time is noted.  Compilation

TiBaskerville, R.C.  and R.S. Gale, "A  Simple Automatic Instrument  for
     Determining  the Filterability of Sewage  Sludges,  "Water Pollution
     Control,  67_, 233  (1968).

                                     92

-------
                       I	1—L_l	1  I I  i I  I I I I I I i i I I
                              1-5      a    S.B   3      4

                           CST/flo SOLIDS OF CONDITIONED FEED)-(BEC)
Figure  33.   Process yield vs CST/(percent solids  of conditioned  feed)
 u
 ui
 n
 D
 ui

 "•  4
 Q
 0

 0
 U
(D
Q

0 1.0
0)  .9
ID  .e
U
                                                      J	I	I  I   I  I I
                        R  - CIO10 CM/Q J
                                                SO
                                                      30  40  BO BO 7080 8O
    Figure 34.  CST/(percent solids of conditioned feed) vs  Rv.
                                   93

-------
of all the data collected for the Blue Plains sludge showed that if 80 mis
of filtrate were collected within two minutes, the filter cake would com-
pletely form and a resistance value of Rv = 27 x IQlO cm/g would result.
     The specific resistance tests were also used to evaluate the effect of
varying chemical dosages on the different sludge mixtures.  As stated
previously and shown in the following table, the filterability of the
conditioned sludge generally increased with increasing chemical addition;
this change was reflected in decreasing values of the specific resistance:
DATE
8-19
8-10
8-10
% CHEMICALS
LIME/FeCl3 Rp
19.6/6.5 1.16
15.2/5.0 1.28
12.1/4.0 2.06
Rv
1010 cm/g
6.84
9.57
19.49
CST
sec
10.5
12.8
12.6
PROCESS YIELD
kg/m2/hr
6.69
5.8.
5.17
                         (Ratio secondary/primary = 1/1)

Theoretically, the point at which optimum chemical conditioning occurs, i.e.
the greatest increase in press yield per unit addition of chemicals, can be
obtained from resistance measurements.  In a full-scale installation, with
a determination of this type, a substantial cost savings in chemicals can
be realized since the unnecessary addition of conditioners would be avoided.


DEWATERING OF VARIABLE SLUDGE CONCENTRATIONS

     Throughout the study, the unconditioned sludges averaged 5% total
solids.  Conditioning with lime and ferric chloride raised the total solids
to 6.0% - 6.5% for feeding to the press.  Because gravity thickening and
air flotation thickening will produce a consistent 5% solids feed, no
special tests were run to determine quantitatively the effect that variable
feed concentrations had on filter press results; however, the NGK press was
capable of handling a range of feed solids from a low of 2.4% (1.8% before
conditioning) to a high of 10.0% (8.4% before conditioning).  In the low
solids region, the press yields were slightly lower because more water had
to be processed; but because of the separate squeezing cycle in the diaphragm
press, cake solids were not affected.  In the high solids region, two
adverse effects were noted:

     1.  Conditioning in the mix tank was difficult because the high sludge
         solids were very viscous and chemical dispersion was hindered.

     2.  The sludge pump and the feed ports in the press were more easily
         plugged with trash and heavy solids.

Total feed solids of 10% appear to be the upper concentration that the
diaphragm press can handle.  Tests to evaluate feed concentrations were
not run for the fixed volume presses.
                                     94

-------
D

S  •
0
h

§«
0
u


0

ffl
D
 i.o
   .8
U
\     iii   tiii
i    i   i   i  i  i  i
            .a   .S  .8 .7.8 .S I.O
                                                    4  S  8 7  8 8
      Figure 35.  CST/(percent solids of  conditioned feed) vs Rp.
                       J	1	1	I  I I I I I
                               B a 7

                                Bv-C10n0 CM/01


                        Figure 36.  Rp vs Rv.
                                          j	i  i  i  i i i
                                 95

-------
MATERIAL BALANCE

     The NGK press was the only unit tested for which all input and output
streams could easily be measured.  Test data from a run on 10/18/77 were
used to calculate a sample material balance.  The calculations, detailed in
Appendix D, show a very good balance between input and output total solids.
CONDITIONING WITH POLYMER

     Over thirty polymers were screened in an attempt to find a polymer
that could adequately condition the sludge for dewatering on a filter press.
Allied Colloid's Percol 776 was found to give the best results in the
conditioning step.  This polymer, a high activity cationic formulation,
worked quite well in conjunction with ferric chloride on a vacuum filter.
Tests on the Buchner funnel also showed good filterability.  Full-scale tests
on the NGK press, however, gave rather poor results.  Cake solids no greater
than 28.6% were achieved, even with extended squeezing cycle times.  Yields
were, therefore, quite low.  The largest problem, however, was the almost
immediate cloth blinding.  High pressure sprays were ineffective in cleaning
the cloths, thus requiring them to be removed for acid washing.

     It is believed that the floe formed by the polymer was too weak to
withstand the high pressures in the press.  But, with proper cloth selection
and care in conditioned sludge handling and feeding, the filter press can
probably be adapted to dewater a polymer conditioned sludge.  The sludge
at Blue Plains, however, varies to the extent that one polymer that will
work effectively 100% of the time has not yet been found.  In contrast, the
lime/FeCl3 conditioning system can be adjusted to always give satisfactory
results.
TESTS ON PRESS CAKE PROCESSING

     The automatic operating mode on the NGK press allowed the production
of relatively large quantities of filter press cake for other purposes.
Throughout  the entire  study, the filter press cakes were used for composting
trials  at the Beltsville, Md. compost site.  A number of cakes were
analyzed for their calorific value and used in incineration tests conducted
in both a multiple hearth incinerator and a rocking grate solid waste
incinerator.  A  local  power utility also analyzed the press cake for possible
use  in  coal fired boilers.

Cake Physical Properties

     The cake, when discharged from a diaphragm  press, resembles a large
waffle.  Generally, it is rigid and free-standing but breaks up easily
 (see Figure 37).  The  density ranges from 1121 to 1185 kg/m3  (70-74 lb/ftj
This press  cake, when  conditioned with lime and  FeClS, dries out in several
days.   Sludge cakes that were conditioned with 20% lime/6.7% FeCls were
exposed to  ambient weather conditions.  One cake was placed in  the open,
exposed to  sunlight, rain, etc.  Another cake was placed in the center of a

                                     96

-------
 61  cm  (2  foot) high "pyramid"  of  sludge  cakes.  Back day  a  portion  of  the
 cake was  analyzed  for percent  solids.
                        TEMPERATURE     %  SOLIDS        %  SOLIDS
DAY    WEATHER          WHEN  SAMPLED       OPEN          PYRAMID

  0     Cloudy               27 °C           39            41.6
  1     Rain/clearing        14 °C           44.7          41.1
  3     Sunny                20 °C           72            40.8
  4     Cloudy               24 °C           77.5          40.7
  5     Rain/clearing        24 °C           77.2          41.7


     These results show that  the press cake does air dry when spread in
a thin layer or if stacked  vertically.  As the cake dried it became
impervious to rain and was  very hard and brittle.  The  material in the
center of the pyramid did not dry; however, further observations of the
cake in the interior of taller piles up to 122 cm  (4 feet) showed some
self heating after 3 to 4 weeks as aerobic decomposition  (composting)
proceeded.  However,  when the cake was broken up into 5 cm (2 inch) pieces
prior to piling outside, it was easily rewetted by rain and became
difficult to handle.   These observations were quite useful when conducting
the composting trials.

Cake Breaking

     In a large-scale installation, some type of cake breaker will be
required prior to any further processing.   Fortunately, the diaphragm press
gives a fairly uniform product which can be easily handled in a controlled
situation.  Test work centered around finding acceptable methods of cake
breaking and establishing the parameters that affected this step.

     Three types of units were found to work:

     1.   A small tree and branch chipper,  operated at high speeds,  was
         capable of breaking up fresh press cakes.   The high speeds,
         however,  caused the machine to gum up easily.   The unit also
         reduced any  partially dried cake  to  dust.

     2.   A garden rototiller,  run through  the cakes while  piled on the
         ground,  was  used to prepare the sludge prior to composting.   This
         slow speed unit did an acceptable job for small test  quantities
         of  cake.

     3.   A make  shift  variable speed screw, pictured in Figure 38,  worked
         quite well to produce chunks  in the  5 cm (2 inch)  range.   Tests
         on  this  unit  showed that  slow speeds  gave  the  best  results.   The
         effectiveness of  this machine was  found  to be  a function of  percent
         cake solids.  Cake  solids  below approximately  27% tended to  stick


                                    97

-------
Figure 37.  Cake from NGK Diaphragm Press.
           Figure  38.   Cake  Breaker.
                     98

-------
         to moving parts.  Above the 27% level, the cakes responded
         to mechanical handling with no sticking.  This unit was also
         effective in breaking up the cakes from the fixed volume presses.
         The wetter inner sections of the cakes from these presses caused
         no problems as long as there were sufficient guantities of the
         dry outside cake sections to help scour the internal screw.

Compost Trials

     Tests were conducted at U.S.D.A.'s compost research facility at Belts-
ville, Maryland, in which the filter press cakes were composted via the
static pile method.  This method is quite successful for composting vacuum
filter cake at 20% solids.  For the vacuum filter cake, the sludge is mixed
with wood chips (2:1 chips to sludge volumetric ratio; 1:1 weight basis)
and stacked to a height of approximately 2.4 m (8 feet) over a perforated
pipe.  A layer of finished compost blankets the pile.  Air is drawn through
the pile for a period of 21 days, causing temperatures to reach a normal
70 °C.  The mixture is sufficiently deodorized in this time period and the
pile is then moved to a stationary curing pile approximately 4.6 m (15 ft)
tall for 30 days.  The curing period ensures maximum pathogen kill.  After
screening out the wood chips for reuse, the product is ready for distri-
bution.

     The wood chips are needed to reduce the initial moisture of the mix,
to provide air passages, and to provide an additional carbon source.  The
wood chips are the major operating cost of this operation.  It was hoped
that with the filter press cake, the wood chips could be eliminated or
substantially reduced in quantity.  Initial composting tests on the press
cake without wood chips did produce the required temperatures in the pile,
but  complete deodorization of the mass was not achieved;   the larger
chunks of cake had crusted over and contained an anaerobic inner core.

     Good results with the filter press cake were obtained by breaking up
the cake to a size of 7.6 cm (3 inch) or less with a rototiller, and mixing
with wood chips to a volumetric ratio of 0.5:1 chips/sludge cake (approxi-
mately 0.2:1.0 weight basis).  The same time periods of about 21 days
composting and 30 days curing were required for the process.   These resultsj
though, are only preliminary since much larger quantities of press cake
are required for a full-scale demonstration test.  It is projected, however,
that if filter press cake is available, cost savings of up to 60% over the
vacuum filter operation can be obtained.

Incinerator Tests
     A number of the press cake samples were analyzed in a Parr adiabatic
oxygen bomb for their calorific value.  An average of 32 samples of 2/1
secondary/primary sludge press cake showed that the press cake can be
considered to.be a low-value fuel that will burn without auxiliary fuel
oil.
                                    99

-------
     Dry solids basis                  3228 cal/gm (5806 Btu/lb)
     Wet solids basis                  1225 cal/gm (2204 Btu/lb)
     Dry volatile solids basis         6293 cal/gm (11318 Btu/lb)

Multiple hearth unit—
     Samples of the filter press cake were incinerated in a 45.7 cm dia-
meter (18 inch) single hearth furnace at the Nichols Engineering Research
Facility.  The purpose of the tests was to determine if the high chemical
content of the Blue Plains filter-pressed sludge would cause any clinkering
problems.  Prior to one of the tests, the Blue Plains secondary plant was
overdosed with FeClo to simulate the approximate iron and phosphate con-
tent in the sludge that is expected when the advanced waste treatment
facilities are completed.  This sludge was then overdosed with lime and
FeClg for conditioning prior to filter pressing.  A second sludge tested
had the normal amounts of iron and phosphate that were available from the
plant at that time.  The following table identifies the sludges tested.
              % CHEMICALS                       % VOLATILE    BTU/lb DRY
SLUDGE CAKE   LIME/FeCl3   % Fe*    % SOLIDS       SOLIDS       SOLIDS

     1         24.4/8.2     8         42.8          45.4        4872
     2         15.6/5.1     8         39.3          50.1        5373
     3         25.9/8.7     5         39.6          47.2        4995
     4         13.9/4.6     5         37.7          53.5        5789

          ^estimated by calculation
     Each of the sludge samples were incinerated to complete burnout at
temperatures from 927 °C to 1038 °C (1700 °F to 1900 °F).  Particle size
fed to the furnace ranged from 2.5 to 7.5 cm (1 to 3 inches).  Excellent
burnout was achieved with no clinker formation.  It was concluded that
the filter press cake and the high chemical addition would pose no special
problems for the incineration of the Blue Plains sludge.

Solid Waste Incinerator Tests—
     A qualitative test was conducted to determine if the filter press
cake combined with solid waste would burn in a solid waste incinerator.
The test unit was a Flynn & Emrich rocking grate design with underfire
and overfire air controls.  The furnace, fed by cranes from a storage
pit, had an average solids detention time of 45 minutes.  Approximately
6600 wet kg (3000 Ibs) of press cake at 35% solids were dumped into the
furnace along with solid waste.  The temperature in the combustion chamber
above the furnace dropped from its normal 677 °C (1250 °F) to 593 °C
(1100 °F) when the sludge was in the burning zone.  Examination of the
residue, however, showed no signs of the sludge cake  and it was assumed
to be completely burned.  These tests indicated that the press cake would
burn well in a co-disposal scheme with solid waste.  Because of the press
cake moisture, however, there is a limit to the amount that can be blended.
                                     100

-------
 Calculations show that with a 35% solids cake, approximately 20 to 40% of
 the wet feed to the incinerator can be sludge cake.

 Evaluation of Press Cake in a Coal Fired Boiler

      Samples of filter press cake were given to a local utility, Potomac
 Electric Power Company, for their routine fuel analysis.  The purpose was
 to determine if a filter cake-coal mixture could be fed to a boiler to
 produce electricity.  They analyzed the cake sample for ash, sulfur,
 moisture, and calorific content:

                                As received               Dry basis

 Ash                               12.7%                    33.9%
 Sulfur                             0.14%                    0.37%
 Water                             62.5%
 Cal/gm (Btu/lb)                   1171 (2108)              3123 (5621)

 This analysis caused them to reject the press cake as a fuel.   They stated
 that,  "Although there are no chemical reasons that we can see  which would
 preclude the use of this sludge as a fuel,  the amount of ash is extremely
 high and would considerably increase our ash handling problems."  Because
 of possible pluggage problems,  "our suppliers of  coal mills express
 concern with the fibrous material in the filter press cake,...   The amount
 of gas flow handled by the induced draft fans would increase because of
 the high moisture content of the sludge.   Considering the additional costs
 for fan power and ash handling  and the additional expense of the added new
 equipment for handling the sludge,  it is doubtful if there is  any economic
 benefit to be gained from the burning of sludge.   Further study would be
 required to confirm this preliminary cost estimate."3   Sufficient  quantities
 of filter press cake were not available for  a full-scale test.
3Letter R.C.  Ungemach (Pepco) to R.C.  McDonell (Montgomery County Council)
 June 3, 1977.                                                 -
                                    101

-------
                                  SECTION 8

                               PROCESS DESIGN
     The purpose of the study was to evaluate the various dewatering devices
that are capable of producing an auto-combustible cake and to develop design
parameters for those units that actually achieved this goal.   For auto-
combustion, approximately 35% total solids in cakes containing lime and
FeCl3 conditioning  and 30% solids in polymer-conditioned cakes are required
from the dewatering process.  The continuous belt press and each of the
filter presses met these requirements.
CONTINUOUS BELT PRESS

     A continuous belt press can produce a 30% solids cake with polymer
conditioning.  While theoretically this is acceptable for auto-combustion,
some practical problems negated consideration of this press for use at the
Blue Plains plant:

     1.  The need to rely solely on polymer conditioning is unacceptable.
         During the testing on the belt press, the variability of the sludge
         feed was so pronounced that no single polymer was found that
         properly conditioned the feed at all times.  Apparently the high-
         rate secondary process produces a variable waste sludge that has
         a highly variable response to polymer conditioning.  While the use
         of lime as a conditioning agent would reduce this variability,
         scaling and cloth plugging problems normally associated with using
         lime have made belt press suppliers somewhat reluctant to rely on
         it for conditioning.

     2.  The high solids content of the filtrate can cause recycle problems
         in a plant where effluent suspended solids must be controlled to
         very low levels.  Suspended solids capture in the belt press was
         estimated at 95% for the 2/1 secondary/primary sludge.  However,
         the poor cake discharge generally experienced with this sludge led
         us to believe that this figure could be an over-estimation of the
         actual recovery.  When advanced waste treatment facilities are
         completed at Blue Plains, the wastewater effluent must meet a
         required 7 mg/1 suspended solids and 0.22 mg/1 total phosphorus
         standard.  Because of the necessity to recycle all water streams
         from solids processing, the thickening and dewatering systems must
         have a high degree of solids capture.  The entire plant is self-
         contained with only two effluent streams; the wastewater discharge


                                     102

-------
       to  the  river  and  the  sludge.   If  solids  are  not  removed  via  the sludge
       stream,  they  will be  recirculated through  the  treatment  system and  even-
       tually  discharged to  the  river.   Calculations  show that  if only
       a 95% solids  recovery was achieved in  dewatering,  the  filtrate stream
       returned to the head  of the plant would  raise  the  suspended  solids  level
       in  the  influent raw wastewater by approximately  15 mg/1.  This influx of
       fine solids,  together with the recycle solids  from thickening  operations
       (which  would  raise the influent level  another  22 mg/1),  would  pose  an
       additional burden on  the  wastewater treatment  train.   Because  of the
       uncertainties in  operating advanced waste  treatment with multi-media
       filters,  the  authors  believe  that unless a 98-99%  solids capture is
       achieved in the dewatering operation,  thus minimizing  recycle  fines
       buildup in the system, the plant  may have  difficulty in  achieving its
       effluent standards.

       For these reasons, the belt press was  considered unsuitable  for use at
 Blue  Plains to dewater  thickened sludge.   The  unit,  though,  has many
 advantages that warrent full investigation at  other  facilities.  In  a plant
 that  has  a fairly consistent sludge that responds  well to polymer  condition-
 ing,  the  press can  provide  a low capital,  low  operating  cost process for
 producing an  auto-combustible cake.

       When used as  a retrofit device to a vacuum filter,  test  work showed that
 the high-pressure section of the belt press  can  further  dewater the  vacuum
 filter cake to the  same final cake  solids  as a filter  press.   In this case,
 lime  and  FeCl3, rather  than polymer, were  used in  conditioning the thickened
 sludge.   No additional  chemicals were used to  condition  the vacuum filter cake
 prior to  dewatering on  the  belt press.   Thus,  if the problems  encountered
 during the demonstration of this process  can be  overcome, the use  of the  press
 as a  retrofit  unit  can  be a very cost-effective  alternative to the filter
 press, especially for existing  vacuum filter installations.  Further test
 work,  however,  is needed to evaluate possible  feeding  and distribution
 systems.  Long-range problems associated with  the  use  of  lime and  FeCl3 on  the
 add-on device  should also be assessed.

 FILTER PRESS

 Chemical  conditioning

       The addition  of lime  and  FeCl3 to the  sludge is  necessary for  the
 operation of  the filter  press.   Throughout the study period the chemical
 dosages required for good filterability varied with the sludge character-
 istics.   In a  full-scale continuous dewatering operation one of the  highest
 priorities should be placed on  defining the variables  in the wastewater
 processing train that most  affect the sludge characteristics and operating
 the treatment plant in a manner  that will minimize their effect on sludge
 filterability.  Doing this will not only provide a smoother dewatering
operation, but will also save many dollars in chemical conditioning  costs.

      Another high priority should be placed on the conditioning step itself.
During the course of the study several important large-scale design
considerations evolved:

                                     103

-------
1.  Because of (a) the cost of chemicals and (b)  the increase in final
    disposal costs due to the addition of inert conditioning chemicals,
    optimization of the dosage is necessary.  A method of predicting
    this optimum dosage was not found; however, bench-scale methods
    (CST, Buchner funnel, pressure tests) were developed which gave
    an indication of how well the conditioned sludge would dewater on
    a filter press.  To avoid the necessity of running a large number
    of these tests, the authors feel that a small pilot-model horizontal
    vacuum filter could be used as a Buchner funnel to continuously
    monitor the specific resistance filtration parameter.  A small slip
    stream from the conditioning tank would be fed to this filter and,
    with the unit running at a constant speed, the time needed to
    produce a dry cake would provide an indication of the dewaterability
    of the conditioned sludge.  The bench-scale tests had shown that if
    the conditioned sludge could be filtered down to a good cake within
    3 to 4 minutes on the Buchner funnel, that same sludge would also
    filter easily on the pilot press  and a 35% solids cake would be
    produced.  By adjusting the feed rate and belt speed of the
    horizontal vacuum filter, this correlation could be established
    for the full-scale press.  If the cake dries too quickly, the
    sludge has been over conditioned and the chemicals can be cut back
    slightly.  If the cake takes too long a time to form, the chemical
    dosage is insufficient and can be increased.  The unit would
    obviously have to be calibrated in the field under continuous
    operating conditions.  A small unit, costing less than $20,000,
    could provide the necessary information to control a multi-
    million dollar filter press installation.

         A horizontal vacuum filter was not obtained in time to be
    tested in the study.  When a unit becomes available, however,
    tests will be conducted to prove this concept.

2.  Because of the wide range of sludge feed rates to the press,
    better control of the conditioning chemicals could be obtained
    by conditioning at a constant flow rate.  The arrangement used
    during the continuous run on the NGK press (depicted in Figure 16)
    is a good example.  In addition to the conditioning tank, a small
    surge tank was used to hold the conditioned sludge for feed to the
    press.  The sludge leaving the conditioner could then be sampled
    and  checked for filterability.  The conditioning tank must be
    designed to provide good mixing without shearing the floe.  The
    surge tank should be sized for a maximum of 30 minutes detention
    time with only enough agitation to keep the solids in suspension.
    Both tanks must be designed so that miscellaneous trash and small
    fibers do not build up on the moving parts.  A shredder  (mazorator)
    installed in  the  sludge  feed line to the  tanks will keep the
    trash to a manageable size and avoid plugging in the filter press.

3.  Careful handling  of  the  conditioned  sludge at all times is a
    necessity if  chemical costs are to be optimized.  The filter  press
    feed pump must be of a design  to  minimize  shearing of the  sludge
                                104

-------
          floe as the material is delivered to the press under pressure.

      4.   The corrosiveness of the FeCl3 solution is  an  important
          consideration in selecting materials of contruction for  both the
          filter press system and the final disposal  system.   Sludge cake
          that has been conditioned with FeCl3 is generally mildly corrosive.

      5.   During the design phase,  close attention must  be given to lime
          slurry handling.   Experience at Blue Plains has shown that scaling
          will occur not only in the lime slurry  lines  but also in the sludge
          feed lines and filtrate lines.   Injection of an anti-sealant
          solution in the lime slurry can assist  in alleviating these problems.
          Careful design of all piping systems including access for periodic
          clean-out is a necessity.

 Filter Press Design

      The  following section on costs shows  that the three types of  presses—
 diaphragm, high-pressure,  and low-pressure -  can all provide  the  required
 cake  solids  at  approximately the same unit costs,,  Several advantages and
 disadvantages of each type are not,  however,  readily apparent  from these
 tables.

      The  diaphragm-type press generally  gave  higher  cake solids,  shorter
 cycle times,  and a more uniform cake than  the fixed  volume presses.
 Essentially  this improved  performance was  related  to cake  thickness.   The
 diaphragm press  operates best with  a 13  to 19 mm (1/2 to  3/4 inch)  cake;
 the low-pressure press  with a 25  to  32 mm  (1.0 to  1.25  inch) cake;  and
 the high-pressure press with a 30 to 40 mm (1.18  to  1.57  inch)  cake.   The
 low-pressure  press  (100 psig)  generally gave  shorter cycle times than the
 high-pressure press  (225 psig),  although the  final overall yield was  greater
 with  the  high-pressure  unit.   Apparently,  the higher pressures  have  the
 advantage of  being able to  handle the  thicker  cakes and, therefore, give
 higher yields per  press cycle.

      Cake discharge with the  high-pressure press was not, however,
 completely acceptable.  Either  the  selection  of a filter media which would
 improve the discharge or the  use of  a precoat would  be  required for a  full-
 scale installation.  Cake discharge was generally very  good on  the low-
 pressure press   and no precoat would be required.  The  diaphragm press used
 low pressures to  feed the sludge  (<100 psig), but squeezed at pressures of
 213 psig.   Cake  discharge was  always good when the sludge was well condi-
 tioned and no precoat would be required.

     An additional advantage of the diaphragm-type press is that it is the
only type of press that can successfully dewater marginally conditioned
sludges.   Therefore, during periods of low sludge production, when extra  fil-
ter capacity is available, less chemicals could be used to marginally
condition the sludge.  Longer squeeze times would be needed to achieve
the required 35% cake solids.  Thus, the squeezing diaphragm can be used
to minimize the overall chemical costs.
                                     105

-------
     This increased flexibility of a diaphragm-type press can also be used
to give any desired cake solids on normally conditioned sludges (by using
extended squeeze times) up to the limit of filterability (approximately
40-45% total solids for a 2/1 sludge).   The high pressures in the diaphragm
press are developed by the squeezing water pump, a relatively low maintenance
item.  High pressures in a fixed volume press are developed by the sludge
feed pump, generally a higher maintenance item.

     The production of sludge cake with a diaphragm-type press gives the
process advantages as mentioned above;  however, from a mechanical standpoint,
this also means more mechanical movement of the press components per ton
of sludge filtered as compared to the fixed volume presses.  This increased
mechanical movement could mean not only higher maintenance costs, but
also increased instrumentation to control the cycle.  In addition, a
diaphragm-type press will be discharging cakes every 30-60 minutes, while
the fixed volume presses will discharge thicker cakes every three hours.
Unless these discharge operations can be made totally automatic and trouble-
free, an operator should be present.  Accordingly, in the cost estimates
for the NGK press, the automatic shaker was eliminated and was replaced by
increased manpower.  The discharge operation on the diaphragm-type press
that releases all cakes at one time  (e.g. Ingersoll-Rand, Lasta design)
appears to have some advantages over the single or two cake discharge
operation.

     Diaphragm and cloth replacement costs represent another disadvantage
for the diaphragm-type press.  Cloth wear occurs in the initial sludge
feed portion of each cycle.  Once the cake is formed, filtration occurs
through the cake and the cloth essentially sees only relatively clear water.
Cloth life for all types of presses  is estimated at 3000 cycles.  With a
fixed volume press this equates to changing a set of cloths each year; with
a diaphragm press, every three months.   Cloth replacement with some filter
designs is quite difficult and long  periods of downtime are required.
Considerably more field testing is required to find more wear resistant cloths
and to define the parameters that extend cloth life.

     Based on the manufacturer's recommendation, diaphragm life is estimated
at 20,000 cycles.  Verification of this cycle life must also be established
under full-scale operating conditions.

Design Parameters—

     NGK diaphragm press—For the 2/1 secondary/primary sludge ratio, Table
10 in the report shows an average of 95 runs over a seven month period.  It
is assumed that these results would  be representative of a full years
operation.  On the average, the chemical dose would be 20% lime and 7% FeC^^
A 17-minute pump time, 18-minute squeeze time, and a 19-minute mechanical time
would be required per cycle  (54-minute total).  Cake solids of 35-40% would
be attained at an average yield of 2.39 kg/hr/m2  (0.49 Ib/hr/ft2).  The cloth
shaker is not included in the mechanical time; it would add another 6 minutes
to the cycle for cake discharge.  The cloth wash is assumed to be required
every 20 cycles.


                                     106

-------
       When more  production is  required,  increased  chemical  dosages  (another  5
  percentage points  of  lime and 2  of  FeCl3)  can  increase  the yield by  a
  factor  of 1.2  (based  on  full-scale  yield data  derived from Table 5).  A
  diaphragm-type  filter press installation can therefore  be  sized for  average
  sludge  production  figures, but it has a built-in  capability to increase sludge
  throughput by slightly overdosing the chemicals.
 7 W2nnn.                    nCUe * PWP t0 dellver pressures u? to
 /kg/cm   (100 psig).  A feed pressure recorder would provide a useful indica-
 tion of whether the rate of pressure rise is too great  (indicating pluggage
 or underconditioned sludge).  A sludge flowrate meter, correlated with toLl
 solids of the sludge feed, could be used to stop the feed pump at the optimum
 teed rate.  The diaphragm pressurization system should include a squeezing
 water pump that will deliver variable pressures up to 17.6 kg/cm2 (250 psig).
 The filtrate during both the pumping and squeezing cycles shSuld be monitored
 for flowrate and total flow per cycle.   The rate monitor would signal the end
 of either the pumping or squeezing cycle.   The flow totalizer would be used
 to indicate differences between runs and provide a monitor on cloth pluggage.
 TP ,9nBaSed /° the lnformation Available,  the cloth of choice would be the
 TR 520 type (described  in Table 9).   The cloth wash system needs further
 evaluation.   The full system pressure of 70 kg/cm2 (1000 psig) was  never
 achieved during the study.   The operating  pressure available was only 24.6
 kg/cm/ (350 psig)  and this  was not always  sufficient to clean the cloths.   An
 acid  wash system may be necessary for units that  use lime for conditioning.
 Acid  washing is quite effective at removing calcuim carbonate and lime
 deposits both from the  cloth and the  filtrate passages on the plates.   For
 ihnn^T     rSlatlr^  b°th aCld  Wash±ng and  high-Pressure spray washing
 should be available.  Acid  is used to free  the system of  lime deposits;  high
 pressure sprays are  used  to remove imbedded sludge particles  from the filter
 media.

       The large number  of electrical  functions necessary  for  the diaphragm-
 type  press would best be  served by using solid state components  which could be
 programmed to  indicate  malfunctions in both machine and circuit  operations.
 Some  problems  were encountered with the  limit switches and  relays that were
 not easily located and/or correctable.

       The NGK  pilot  press was provided with a 25 mm (1-inch)  filtration
 chamber.   Because  the sludge  was easily  filtered on the fixed  volume presses
 with much thicker  cakes,  we recommend  that  a  thicker  chamber be  provided on
 a  full-scale design.  Chamber thickness  up  to  38 mm (1.5  inches)  should
 increase  the overall  yield  substantially.  A  larger  chamber would not
 compromise the  advantages of  the diaphragm- type press.  With easily  fil-
 terable  sludges, more cake  per  cycle  can be discharged. If, however,  the
 sludge filterability  is poor,  short pump times could  still be  used to
 provide a  thin, dry cake.   Diaphragm and cloth life must  each  be  evaluated
with a thicker  chamber  because of the increased distance of diaphragm
movement.  An added benefit of a larger  chamber is  that the feed opening
would be less likely to be plugged with  rags and trash.
                                      107

-------
     Lasta Diaphragm Press—Comparison testing showed that the Lasta press
would give both equivalent cake solids and somewhat higher yields than the
NGK press.  An average of seven runs on both presses indicated that the full-
scale Lasta yield was 22.6% higher than the NGK yield for the Blue Plains
sludge.  Because of this limited amount of data, design parameters for this
press were developed by scaling results from the NGK press.  Using the seven
month average on the NGK press and applying the 22.6% factor, the full-scale
average yield for the Lasta press is then 2.93 kg/hr/m2 (0.60 Ib/hr/ft2).
The number of Lasta filter press units required for installation, though,
will be greater than that for NGK.  The largest Lasta press has only 204 m2
filtration area, whereas the largest NGK press has 500 m2.

     The main advantage of the Lasta design is the shorter mechanical turn-
around time (10.5 minutes for their 204 m^ press), since all chambers
discharge at once.  Additionally, the cake discharge and cloth washing
operations are almost completely automatic and the operator attention
required would be minimal.

     Because of Lasta's shorter mechanical time and, consequently, higher
yield, the optimum pumping cycle  (as determined from the solids addition
rate) is slightly shorter.  This provides for a thinner cake and, therefore,
somewhat shortened squeezing times.  Sufficient data was not collected to
compute average cycle times; however, an estimate would be in the range of
30 to 40 minutes total.  This assumes that during cloth washing, accomplished
by low-pressure (100 psig) sprays, only 1/4 of the filter cloths will be
washed each cycle.

     The same type of controls as discussed for the NGK press would also
be required for the Lasta design.  Because of the shorter cycle times
in this press, however, the main disadvantages of a diaphragm-type press,
i.e. filter cloth and diaphragm replacement costs, could possibly be even
more pronounced with the Lasta-type design.

     High-pressure press  (Passavant)—The full-scale design for the Passavant
press  is based on the comparison runs in August, 1977.  During that time
the results showed that the high-pressure press could process the same
quantity of sludge but would require 62.3% more filtration area than the
NGK press.  Design parameters for this press were also developed by scaling
the results from the NGK diaphragm unit.  Taking the seven month average of
data on the NGK press and applying the 62.3% factor gives a Passavant design
yield  of  1.51 kg/hr/m2  (0.31 Ib/hr/ft2) with a 40 mm (1.57 inch) chamber
thickness.  A mechanical  time of  20 minutes is required for their Model  20
press  (11,625 ft2).  Using 20% lime and 7% FeCl3 for conditioning, cake
solids of  34 to 37% will be produced in an average cycle  time of 3-1/3 hours.
Increasing this chemical  dosage  (another 5 percentage points of lime and 2
of FeClo)  should result in an increased yield of approximately 20%.  As with
the NGK press, a built  in capacity exists for handling increased sludge
production by increasing  the chemical dosage.

     The  sludge feed system should include a pump  to deliver pressures
up to  15.8 kg/cm2  (225  psig).  A feed pressure  recorder is necessary but a
flowrate  indicator would  not be  needed with this press.   Filtrate rate and

                                     108

-------
  total filtrate flow is the preferred method of monitoring the operation.
  The sludge feed system surge tank should be shared with several presses'
  so that the conditioned sludge detention time does not increase above
  the 30 minute limit (to avoid floe deterioration).

      Low-pressure press (Nichols)—The full-scale design for the Nichols
  press is based on the comparison runs in August, 1977.  During that time
  the results showed that the low-pressure press could also process an
  equivalent quantity of sludge but would require 126.8% more filtration area
  than the NGK press.  Again, design parameters were developed by scaling
 NGK press results because of the limited amount of data available.  Taking
 the seven month data on the NGK press and applying the 126.8% factor gives
 a Nichols design yield of 1.07 kg/hr/m2 (0.22 Ib/hr/ft2),  with a 32 mm
  (1.25 inch) chamber thickness.   A mechanical time of 20 minutes is required
 for their largest press (6760 ft2 with 115 chambers).  Using 20% lime and
 7% FeCl3 for conditioning, cake solids of 34 to 37% will be produced in an
 average cycle time of  3 hours.   Increasing the chemical dosage will give an
 increase in yield similar to the high-pressure press.  With the exception
 of using a feed pump of 7.0 kg/cm2 (100 psig),  all other comments  for process
 control are identical  to  those made  for the high-pressure  press.


 MULTIPLE-HEARTH INCINERATOR DESIGN

      Tests run on the  single-hearth  incinerator showed that  there  would be
 no clinkering  problem  with the  filter press cake.   We were unable  to run
 any tests  to prove that the press cake was  auto-combustible;  however,  the
 calculations are  fairly well known and have been verified  in large-scale
 installations.  Figure 39  is a  plot  showing incinerator  outlet  temperature
 as a  function  of  cake  total solids and percent  conditioners.  Any
 combination that  gives an  outlet  temperature above 800 °F  is  auto-combustible.
 The graph  was  provided by  Whitman, Requardt and  Associates,  Baltimore,  Mary-
 land.  Assumptions were:

     Feed  rate:  507,000 Ibs/day  sludge solids
     Volatile  solids before  conditioning:   60%
     Heating value:  10,000  Btu/lb V.S.
     Excess air:   75%

     The figure shows  that with 27% conditioners (20%  lime, 7% FeCl3) and
 35% cake solids the feed is auto-combustibile with an  800 °F outlet  temper-
 ature.  If the quantity of conditioners increases, then there must be a
 corresponding increase in percent cake solids.  For example, if the  dosage
 rate is increased to 33% (25% lime, 8% FeCl3), then the cake solids must
 increase to approximately 36.2%.  This increase in cake solids is easily
 accomplished with any of the filter presses.  Essentially the figure shows
 that large increases in chemical addition rates require only small increases
 in final cake solids to maintain the 800 °F temperature.  For a given
 chemical conditioning,  increasing the cake dryness by extending cycle times
 in the press has the effect of raising the incinerator outlet temperature.
 If an afterburner is used  immediately downstream of the incinerator, this
increase in outlet temperature may result in some fuel savings.   In
                                    109

-------
11OO
                                                                                         T1OO
                                                                                         1OOO
                                                                                         BOO
BOO
   18
             ao
                                                            3O
                                                                     3E
            Figure  39.
2       24       as       as
         °lo CONDITIONERS
 Incinerator  outlet  temperature vs percent conditioners,
                                                                                       38

-------
actuality  however, it is more cost-effective to remove the water vapor
thus reducing the amount of gas to be heated  prior to raising the outlet
gas temperature (for toxic pollutant control @ 1350 «F).  There is then,
little benefit to achieving a very dry cake above the auto-combustible
range.  However, if a waste-heat boiler is used for steam generation, the
higher outlet temperatures could increase the steam production, and In this
case there may be some benefit to increasing the cake solids from the press.
The alternatives are too complex for any generalized calculations to show
the tradeoffs; hence each design must be evaluated on an individual basis.
                                   Ill

-------
                                  SECTION 9

                        DEWATERING AND DISPOSAL COSTS
     Estimates of capital and operating costs are presented in Tables 25
through 29.  These estimates are for a large municipal wastewater treatment
plant generating 250 dry tons of sludge per day (roughly equivalent to a
wastewater flow of 200-250 MGD).  These estimates are purposely generalized
and not specific to the Blue Plains  plant.  The dewatering options costed
are vacuum filters, filter presses, and belt presses.  Final disposal costs
for both incineration and composting are included.

     The following general assumptions were used in developing the tables.
The reader is referred to Appendix E for details of all calculations.

     1.  Sludge:  500,000 Ibs/day dry incoming sludge solids @ a
         concentration of 5% (before conditioning); 2/1 secondary/primary
         sludge solids ratio.

     2.  Chemical conditioners:  For vacuum filter and filter presses,
         lime @ 20%, FeCls @ 7% of dry sludge solids.  Lime cost @ $0.022/
         Ib; FeClo cost @ $0.065/lb.  Anti-sealant needed to help prevent
         lime deposits.  For belt press, polymer costs @ $15.00 per ton
         of sludge solids..

     3.  Yield:  Based on test data and expressed as pounds of sludge solids
         per hour per square foot of filtration area.  On belt press,
         expressed as pounds per hour per meter of belt width.

     4.  Number of Units:  Based on largest size unit available.  (See
         Appendix F for specifications.)

     5.  Capital cost  (1978 dollars):  Includes chemical feed system, sludge
         feed pumps, dewatering unit with  all necessary accessories,  and
         conveyor system to transport cake  to next process.  The total
         capital cost was obtained by multiplying the manufacturers'
         equipment cost by a factor of 3 to include  installation, piping,
         utilities, building and engineering.

     6.  Amortization:  Computed at 6-3/8%  and  20-year life.
                        Capital cost x 0.09 = annual amortization cost.

     7.  Power:  Cost  at  $0.04  per kwhr.
                                     112

-------
                                   TABLE 25.  DEWATERING COSTS


Sludge Solids (tons/day)
Yield (Ib/hr/ft2)
% Cake Solids
Unit Size, ft2
No. of Units*
Capital Cost, $1,000
Annual Costs, $1,000
Amortization
Chemicals
Power
Water
Operating Labor
Maintenance
Total
Unit Cost, $/ton
Vacuum
Filter
250
3.0
20
600
13
8,700

783.0
1,663.4
351.9
94.6
588.0
87.1
3,568.0
39.10
Diaphragm
Press
250
0.49
35
5380
9
23,000

2,070.0
1,663.4
248 .'9
6.4
504.0
496.5
4,989.2
54.68
High-Pressure
Press
250
0.31
35
11,625
7
25,350

2,281.5
1,663.4
210.0
1.6
420.0
450.0
5,026.5
55.08
Low-Pressure
Press
250
0.22
35
6760
15
21,800

1,962.0
1,663.4
324.5
1.0
756.0
218.0
4,924.9
53.97
* Includes one standby unit.

-------
                        TABLE 26.  BELT PRESS COSTS
Sludge Solids (Tons/day)
Yield (Ib/hr/meter of width)
% Cake Solids
Unit size
No. of units*

Capital Cost, $

Annual Costs, $
Belt Press

   250
   675
   30
   3 meter
    12

  7,050,000
Vac. Filter
     +
 Belt Press

    250
3.0 - 1180**
20 - 35 **
600 ft2 - 2 meter **
  13 _ 10**

 12,400,000
Amortization
Chemicals
Power
Water
Operating Labor
Maintenance
Total
Unit Costs, $/ton
634,500
1,368,800
81,300
108,400
672,000
90,500
2,955,500
32.39
1,116,000
1,663,400
398,240
153,700
840,000
143,750
4,315,090
47.29
* Includes one standby unit.
** First entry for vacuum filter; second for belt press
                                     114

-------
                     TABLE 27.  INCINERATION COSTS
                               Vacuum Filter          Filter Press
                                  Feed                   Feed

Total Feed, tons/day             317.5                  317.5
% Feed Solids                     20                     35
% Volatile Solids                 47.2                   47.2
Furnace Diameter                 25' - 9"               25'- - 9"
No. Hearths                       12                     12
Furnace Capacity,
  Ibs wet feed/hr/ft2             10                     10

Capital Cost, $1,000              20,000                 10,000

Annual Costs, $1,000
Amortization
Power
Fuel
Operating Labor
Maintenance
Ash Disposal
Total
1,800.0
899.2
4,110.0
504.0
400.0
610.0
8,323.2
900.0
405.2
630.0
336.0
200.0
610.0
3,081.2
Unit Cost, $/ton Of sludge solids     91.21               33.77
                                 115

-------
                               TABLE 28.  LAND DISPOSAL COSTS
                                     Vac. Filter Cake
                                       @ 20 % solids
                                 $/wet ton      $/dry ton of
                                               sludge solids
                                       Filter  Press  Cake
                                        @  35%  solids
                                    $/wet  ton      $/dry  ton of
                                                  sludge  solids
Hauling (25 mile distance)
Composting
Total
9.40
8.84
18.24
59.69
56.13
115.82
6.25
6.79
13.04
22.68
24.64
47.32
                  TABLE 29.  TOTAL DISPOSAL COSTS  $/ton of dry sludge solids
Incineration
Composting
Vacuum Filter Feed

     130.31
     154.92
Filter Press Feed

       88.45
      102.00

-------
      8   Water;  City water, at a cost of $0.53 per 1000 gallons,
          required for high-pressure washes on the filter presses.
          Filtered and disinfected water for low-pressure sprays such as
          required on a vacuum filter and belt press is supplied at a
          cost of $0.25 per 1000 gallons.  Chemical makeup water supplied
          at no cost from filtrate or plant effluent.

      9.  Operating labor;   $21,000 per man year; 4 crews required per week
          to cover a 7-day operation.  Includes supervision.

     10.  Maintenance:   Based on a percentage of equipment purchase costs
          plus cloth replacement costs.

     11.  Unit cost;   $/ton of incoming dry sludge solids.

      In Table 25,  costs  for  the vacuum filter  and  the  high-pressure and
 low-pressure press are based on the costs for  actual operating installations
 in the United States and are considered fairly accurate.   No large scale
 operations of the diaphragm-type  press are currently on line in the United
 States; hence,  the costs for these units are based on  the best information
 available from the manufacturer.   Operating costs  for  the three types of
 filter presses are essentially equal at $54 to $55 per ton.   Selection of
 one type versus the other  can,  therefore,  be based on  the operating
 parameters desired and/or  competitive bidding.   For dewatering only,  the
 vacuum filter provides a cheaper  alternative than  filter  presses.   However,
 the major differences  are  essentially due to the amortization costs.   Out-of-
 pocket annual operating  costs,  exclusive of amortization,  for either the
 vacuum filter or any of  the  filter press types are approximately $30 to  $32
 per ton.

      Table 26 presents some  cost  estimates for the belt press,  both as a
 single unit  or  as  a  retrofit  to a vacuum filter.   As with the diaphragm
 press,  no large-scale  belt press  installations  are currently on line  in
 the United States  to provide  actual  cost data.   Therefore,  the  estimates
 given are based on information  available from  the  manufacturers.   It  is
 assumed that  a  suitable  polymer at a reasonable cost can  be  provided  for
 sludge  conditioning.   The estimates  show that  the  belt press  has  the
 potential for providing  a very  reasonable  alternative   ($32.39  per  ton)
 to  either a vacuum filter or  a  filter press.  However, because of problems
 detailed  in the previous section of  this report, the belt press was not
 considered suitable  for  Blue  Plains.

      The  use  of a belt press  as a retrofit to a vacuum filter installation
 shows a reasonable cost  ($47.29 per  ton).  This estimate assumes the  full
 price for a new vacuum filter installation; enough information  is
 presented in Appendix E  to fully cost this option for a specific existing
 facility.  As detailed in the previous section, however, further work
must be initiated to develop a workable  system prior to implementing  this
 option.
                                    117

-------
     Table 27 shows approximate costs for a multiple-hearth incineration
facility.  A single train includes a 12-hearth incinerator, electrostatic
precipitator, sub-cooler, and fume furnace.  Emissions are controlled to
the EPA limit of 1.3 pounds of particulate matter per dry ton of solids
input.  Fuel costs are based on producing an 800 °F exhaust temperature
from the furnace and then further raising the stack gases to 1350 °F.
For the 20% feed, considerable fuel is required in the furnace; for the
35% feed, the sludge is auto-combustible and fuel is required only to raise
the stack gas temperature.  Table 27 shows considerable savings when
incinerating a cake in the auto-combustible range.  Along with the savings
in fuel requirements, fewer furnaces (2 vs 4 units) are required, thereby
realizing additional savings in power, labor, and maintenance costs.

     The hauling costs in Table 28 are based on a 25 mile haul distance
to a processing or disposal site.  The composting costs are based on the
open-air static pile method developed at Beltsville, Maryland.  These costs
are for processing only, and do not include any costs or revenues derived
from the marketing/disposal of the final product.  Because of the long
transport distance, the costs per dry ton for hauling are nearly equal to
the composting costs.

     The total disposal costs in Table 29 show:

     1.  Total disposal costs for filter pressing and incineration are
         approximately $88 per ton.  This compares to the total cost for
         vacuum filtering and incineration at $130 per ton.  Therefore,
         savings of nearly $4,000,000 per year for a 250 ton-per-day plant
         are possible by selecting filter presses for dewatering.

     2.  Total disposal costs for filter pressing and composting  (including
         the cost of hauling the press cake 25 miles) are approximately
         $102 per ton.  This compares to the total cost of vacuum filtering
         and composting (including hauling) of $155 per ton.  Choosing a
         filter press rather than a vacuum filter, therefore, will result
         in annual savings of nearly $5,000,000 for a 250 ton-per-day plant.
                                     118

-------
                                 APPENDIX A

                             LABORATORY ANALYSES
     The following routine laboratory analyses were performed.  Detailed de-
scriptions of individual procedures can be found in Standard Methods for the
Examination of Water and Wastewater, 14th edition.
     Primary, Secondary
     Sludge

     Total Solids
     PH
     Specific Gravity
     Sludge Feed

     Total Solids
     Volatile Solids
     PH
     Specific Gravity
     Fe"* ' '
    Filter  Cake

    Total Solids
    Volatile Solids
    Density
    Fe
    BTU
 Method

 O'Haus moisture balance
 Glass electrode
 Referred to weight  of 1 liter of
 water at room temperature
 Dried  at  103-105  °C  overnight
 Ignition  of  dried residue  at 610°C
 Same as above
 Same as above
 Atomic absorption spectroscopy
 using  Varian AA-6 Spectrophotometer
Dried at 103-105 °C overnight
Same as above
Variable-volume press - determined
from volume of water displaced
by a known weight of filter cake
Fixed volume press - determined
from the total weight of filter
cake divided by total chamber
volume
Same as above
Determined using Parr adiabatic
Oxygen Bomb Calorimeter
    Filtrate
    Total Solids
                                          Same as above
                                    119

-------
Suspended Solids

pH
Cl"

COD

Total Phosphate

Total Nitrogen, Nitrate
Determined according to Standard
Methods
Same as above
Determined according to Standard
Methods
Determined according to Standard
Methods
Determined according to Standard
Methods
Determined according to Standard
Methods
                                 120

-------
                                 APPENDIX B  -  DATA SHEETS
 DATE- 6/23/77
                    PRESS- NGK
                                          DATA SHEET  1
                                     SQUEEZING PRESSURE-   213
                                                                    CLOTH  TYPE  NY51-4
SLUDGE TYPE- Prim & Sec APPEARANCE- Brown color
TEMPERATURE- 25°c_ 	 GRINDING Yes PURPOSE OF RUN- V9rvinB chDnl-pn1 _,<,,,.,..,...,
RUN S
PRIMARY/SECONDARY (pH) ..
	 	 — 	 C 	 p^ 	 J_
COMBINED SLUDGE-uncond.-cond.
LEVEL-beforeS after cond.(INCH
LEVEL (after pumping)
LIME ADDED (GALS)
FECL3 ADDED(GALS)
SLUDGE PUMPED (GALS)
FILTRATE COLLECTED (PUMPING)
FILTRATE COLLECTED (SQUEEZE)
FILTRATE COLLECTED (TOTAL)
FILTRATE : pH
FILTRATE APPEARANCE
CAKE .-WEIGHT (WET)
CAKE :CONSISTENCY/DISCHARGE
PUMPING TIME
PUMPING PRESSURE (TERMINAL)
SQUEEZING TIME
CLOTH WASH: before run
CAKE THICKNESS
CST (of conditioned sludge)
CAKE DENSITY
TANKED DRAINED
PRIMARY- SPECIFIC GRAVITY-
SECONDARY-SPECIFIC GRAVITY-
COMBINED -UNCOND.(SP.GR.)
COMBINED-COND. (SP.GR.)
#1
- / -
- / -
232/3/211A
30 7/R
16-3/4
5.6
61.6
35.7
10.9/46.6
52.9
-
clear
74
excellent
15
58
18
yes
1/4-1/2
14.8
1.1757
yes
1.027
1.014
1.0114
1.0208
#2
- / -
— / _
23 2/3 / 21
™-i/7
20
6.7
60.8
39
9.9/48.9
53.9
-
clear
83
excellent
16
53
18
no
1/4-1/2
10.8
1.18
yes


1.0058
1.007
#3
- / -
/
232/3/213/4
2Q-1/R
13-1/4
4-1/2
47.2
19.3
11.2/30.5
36.5
_
cloudy
64
wet
11
95
18
no
1/4-1/2
24.2
1.12
yes


1.0027
1.0065
#4
- / -
/ -
/












65
- /
- 1 -
i












'




















PRIMARY-INCHES- 4-1/3
REMARKS:   ph meter inoperative
SECONDARY-INCHES-  20
                                   RATIO-///# PRI./SEC.  32/68
                                             121

-------
Explanation of Data Sheet 1

1.  Information at top of page is self-explanatory.   The grinding entry indi-
    cates if the primary sludge mazorator was used.

2.  pH - was normally measured on each of the sludges, both before and after
    conditioning,, and on the filtrate.  The pH probes failed quite often
    when used with sludge and were inoperative on this particular day.

3.  The "LEVEL" entries show the amount of sludge in the tank at various
    intervals, measured from a reference point at the top of the tank.
    The total tank depth from this point was 48 inches.  Tank calibration
    was 6.4 gallons/inch.

4.  The lime and FeCl3 gallons added are also equivalent to the pounds of
    each added.  Note that lime and FeCl3 used were both-1 Ib/gal for easy
    calibration.

    Lime makeup in the lime slurry tank was as follows:

          An 80 Ib bag of lime was added to 75.5 gallons water
          The resultant total volume was 80 gallons.

    FeCl3 makeup in the holding tank was as follows:

          28.6 gallons FeCl3  (30% by weight; 3.5 Ibs FeCl3 per
          gallon; Specific gravity = 1.362) was added to the vat.
          The vat was filled to 100 gallons with water and mixed well.

5.  The  "filtrate collected  (pumping)" was measured in gallons by a dipstick
    reading in the 100 gallon collection vat.

6.  The  "filtrate collected  (squeeze)" was measured in gallons by a dipstick
    reading in the 15 gallon collection vat.  The first entry shows the
    quantity  collected during the squeeze cycle only; the second entry
    shows the volume collected in the pumping and squeezing cycles combined.

7.  The  "filtrate collected  (total)"  shows the total  amount in the vats
    after the pump and squeeze cycles plus the filtrate and sludge blowing
    cycles.   The blow cycles  contributed approximately six  gallons.   The
    filtrate  samples were  collected prior to  the blowing cycles.

8.  Cake weight was measured  in pounds by collecting  all the discharged
    cake and  weighing on a beam scale.
                                      122

-------
  9.   Cake:  consistancy/discharge is  the operator's opinion on the hardness
      and  quality of the  discharged cake.

 10.   Pumping time is the total time that the sludge pump was  running.   It
      takes approximately 1.5 to 2.0 minutes to  fill the chambers  in  the NGK
      pilot press.  Some  press manufacturers refer to pump time as the time
      of filtration after the chambers are filled.

 11.   Terminal pump pressure is the reading taken on the discharge of the
      diaphragm pump at the end of the cycle.  Because this was a piston
      pump, the pressure  gage pulsed and the reading is only approximate to
      + 5  psig.

 12.   Squeezing time is the total time that the  squeezing pump was running.
      It took only 15-20  seconds to fill the chambers on the pilot press.
      Larger presses may  require 2 to 3 minutes  to fill the chambers, thus
      extending this time in actual large-scale  operation.

 13.   The  "cloth wash" is self-explanatory.  The automatic system was used.

 14.   Cake thickness was  measured in inches at various points on various
      cakes.

 15.   CST  (capillary suction time) in seconds shows the relative filterabil-
      ity of the conditioned sludge.

 16.   Cake density (gm/cc) was measured by placing a one liter graduated
      cylinder on a balance and measuring both the weight of cake added and
      the volume of water displaced.

 17.   "Tank Drained" refers to the status of the mixing tank at the end of
      the run.

 18.   The specific gravity of the primary sludge, secondary sludge, uncondi-
      tioned,  and conditioned combined sludges were all measured by weighing
      one liter of the sample.   Because of gas bubbles and pieces of trash
      in individual samples, this method was not completely accurate.   How-
      ever, the averages of many samples can be used for most calculations
     with little error.

19.   The primary and secondary "inches" refer to volume in the mixing tank.
     Note the high volume ratio (4.6/1) in contrast to the low weight ratio
      (2/1).

20.  The specific gravity of the 1 Ib/gallon chemical solutions were measured
     periodically.
                                    123

-------
                                          DATA SHEET 2
DATE: 6 / 23 / 77
TIME
(MIN)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
COMMENTS :
SLUDGE PUMPING
INCHES GALLONS PRESSURE
21 1/4
24 5/8
25 7/8
27 1/4
27 3/4
28 3/8
28 5/8
29
29 1/4
29 1/2
29 7/8
30 1/4
30 1/2
30 5/8
30 3/4
30 7/8






0
21.6
29.6
38.4
41.6
45.6
47.2
49.6
51.2
52.8
55.2
57.6
59.2
60
60.8
61.6





0
12
21
35
49
49
50
49
54
54
54
56
56
58
56
58





Run #1
FILT
PUMPING
INCHES GALLONS
0
0
1 5/8
3 1/2
5 3/8
6 3/8
7 3/8
8 1/4
9
9 3/4
10 1/2
11 1/4
11 3/4
12 3/8
12 7/8
13 3/8





0
0
4.3
9.3
14.3
17
19.7
22
24
26
28
30
31.3
33
34.3
35.7



BLOWDOW

UTE VOLUME
SQUEEZING
INCHES GALLONS
0
3 1/4
5 3/8
7 5/8
9 1/2
10 7/8
12
13 1/4
14
14 7/8
15 1/2
16 1/8
16 5/8
17 1/8
17 1/2
17 7/8
18 1/4
18 1/2
18 3/4


0
37.6
38.8
40.1
41.2
42
42.7
43.4
43.8
44.3
44.7
45.1
45.3
45.6
45.9
46.1
46.3
46.4
46.6
52.9
t ,

                                                                       total cumulative gallons
Pump filtrate in 100 gallon vat  -  2.67 gal/inch
                                                                           of filtrate
Squeeze filtrate in 15 gallon vat  -  0.58 gal/inch
                                               124

-------
 Explanation of Data Sheet 2

 1.   During  each run,  dipstick readings  of  the  sludge  and  filtrate  levels
     were  simultaneously  taken by  two  operators.

 2.   Column  1 shows  the time in minutes.

 3.   Column  2 shows  the inches of  sludge in the mixing tank as measured from
     the top reference point.

 4.   Column  3 shows  the volume of  sludge pumped, corresponding to the measure-
     ment  in column  2.

 5.   Column  4 shows  the sludge feed pump discharge pressure in psig.

 6.   Column  5 shows  the inches of  filtrate  collected in the 100 gallon vat
     during  the pumping cycle  as measured by dipstick  from the bottom.

 7.   Column  6 shows  the volume of  filtrate  corresponding to the readings in
     column  5.

 8.   Column  7 shows  the inches  of  filtrate  collected in the 15 gallon vat
     during  the squeezing cycle, as measured by dipstick from the bottom.
     This smaller vat was used  so  that low  filtrate readings could be observed.

 9.   Column  8  shows  the volume  of  filtrate  corresponding to the readings in
     column  7  and added to the  total volume collected during the pumping cycle.

     Note that the cycle times for this run were 15 minutes of pumping, fol-
 lowed by 18 minutes of squeezing.   Normally,  the pumping cycle was terminated
when the level in the sludge tank dropped to 1/8" per minute for three conse-
 cutive minutes.  The squeezing cycle was terminated when the filtrate rate
 dropped to 1/4" per minute for three consecutive minutes.
                                     125

-------
                                     DATA SHEET  3
                              ANALYSIS  REQUEST  AND REPORT

                            Analytical  Services Laboratory
                                 FWPCA-DC  Pilot Plant
                                   Washington,  D.  C,
                                                    Analysis Requested
Submitted by:   W. Ruby
Date Submitted: 6/23/77
Date Reported:  6/24/77
LAB NO.
6/23/77





















SAMPLE IDENTIFICATION
Primary
Secondary
NGK Rl Pump
NGK Rl Squeeze
NGK R2 Pump
NGK R2 Squeeze
NGK R3 Pump
NGK R3 Squeeze















9.26
4.34
5.15
-
5.68
-
5.90
-















71.0
60.5
67.6
-
63.7
-
64.6
-

















6.59
-
7.30
-
6.93
-

















47.5
-
43.8
-
48.9
-

















39.2
-
38.4
-
29.4
-

















47.1
-
44.5
-
47.9
-

















292
41
172
27
1064
422

















9250
8955
10457
11241
9863
9001





































                                          126

-------
 Explanation of Data Sheet 3

      The following analyses were determined for each filter press run.
 Laboratory procedures  are explained further in Appendix A.

 1.   Primary sludge - % total solids
                     % volatile  solids

 2.   Secondary  sludge - %  total solids
                        %  volatile solids

 3.   Combined sludge -  unconditioned feed
                        %  total solids
                        %  volatile solids

 4.   Combined sludge -  conditioned feed
                       % total solids
                       % volatile  solids

 5.   Filter  Cake -  % total  solids
                   % volatile solids

 6.   Filtrate - total solids as mg/1
               suspended solids as mg/1

     The primary and secondary sludge samples were composites taken either
 from the NGK mixing tank or directly from the discharge of the thickeners.

     The combined sludge samples, both unconditioned and conditioned,  were
 dipped from the NGK mix tank.

     Filter cake samples were taken at random from various sections of at
 least four of the six cakes.  Tests were run to show that the cakes were
uniform throughout with respect to % solids.  No appreciable difference in
 % solids of the six cakes was ever observed.

     Filtrate samples were dipped from the filtrate collection tanks after
agitation to ensure a representative sample.  For some runs the filtrates
from the pumping and squeezing cycles were analyzed separately.
                                     127

-------
                                            DATA SHEET 4
                                     FILTER PRESS  DATA - NGK
DATE- RUN* - 6/23/77
TYPE SLUDGE-
RATIO- J?///-PR./SEC.
FEED SOLIDS-%SOL./%VOL-(uncond.)
PRIMARY-7.SOL . /%VOL/ pH . -
SECONDARY-%SOL./%VOL./ pH.-
FEED SOLIDS-%SOL./%VOL.-(cond.)
PRIMARY- SP. GR. (gr./cc.)
SECONDARY-SP . GR. - ( gr . / cc . )
FEED SOLIDS-SP.GR. -(uncond. ) -_pH . -
FEED SOLIDS-SP.GR. -(cond.)- pH.-
LIME (added) %-
FECL3 (added) Z-
VOLUME-(feed to oress)-GALS.
PUMP TIME-(Mlns.)
SQUEEZE TIME-(Mins.)-
TERMINAL PRESSURE -Pump psig.-
SQUEEZING PRESSURE-osiE
FILTRATE VOLUME-(gals . )PUMPING-
FILTRATE VOLUME- ( Gals .) SQUEEZE-
FILTRATE VOL. -(Gals.) TOTAL+B.D.
FILTRATE.. pH.
pump/squeeze
FILTRATE (mg/1) TOEAL SOLIDS -
pump/squeeze
FILTRATE (mg/1) SUS. SOLIDS-
FILTER CAKE-(Wet weight)
FILTER CAKE-(%Sol./%Vol.)
FILTER CAKE (Dry Weight) corr.
CAKE THICKNESS-(Inches)
YIELD (lbs./ft.2hr.) process
' 1
Prim & Sec
31.7/68.3 	
5.15/67.5
9.26/71/-
4. 34/60. 5/-
6.59/47.5
1.027
1.014
IJUlALr
1.0208/-
24.7
8.3,
61.6
15
18
58
213
35.7
10.9/46.6
52.9

9250/8955
292/41
74.
39.2/47.1
21.8
1/4-1/2
0.63
2

	 >• —
5.68/63.7
9.26/71/-
4. 34/60. 5/-
7.30/43.8
1.027
1.014
1.0058/-
1..007/-
27.0 '
L?.0
60.8
16
18
53
213
39.0
9.9/48.9
53.9

10457/11241
177/27
83
38.4/44.5
23.4
1/4-1/2
0.66
3
v
	 ^,
5.9/64.6
9.26/71/
4. 34/60. 5/
6.93/48.9
1.027
1.014
1.0027/-
1.0065/-
17.2
5.9
47.2
11
18
95
213
19.3
11.2/30.5
36.5

N.S.
M S
64
29.4/47.9
15.3
1/4-1/2
0.50





























* NS indicates no sample was  taken




CLOTH TYPE-NY51-4 TEMPERATURE-  25"C   PRESS MECHANICAL TIME- 5 min.   PRESS FILTER AREA-  62.4 ft2
CONCLUSIONS-   Tests show good filtration and high yields with high
                                                                             dr.oago,
      Low lime  (17.2%) would not dewater effectively.  Note low yields and IOTJ
                                                                                    anUHg.
                                               128

-------
Explanation of Data Sheet 4

     Data Sheet 4 combines the data collected from the run (Data Sheets 1 and
2) with the laboratory results (Data Sheet 3) and several calculations to
summarize the series of tests run.  All examples use Run #1.

1.  The actual weight ratio of primary to secondary sludge was calculated
    from the measurements taken of the volume, specific gravity, and laboratory
    % solids of each sludge.

    Ibs solids = (inches in tank) x (tank calibration) x (specific gravity) x
                 (density of water) x % solids
                                         100

    Ibs primary sludge = 4.33 in x 6.4 gal/in x 1.027 x 8.3453 Ib/gal x |^-
                       =22.0 Ibs                                       10°

    Ibs secondary sludge = 20 in x 6.4 gal/in x 1.014 x 8.3453 Ib/gal x ^-^-
                         =47.0 Ibs                                     10°

    Ratio:  Ibs primary/lbs secondary = — or 31.7%
                                        47    68.3%

2.  Percentages of total and volatile solids, pH (if available), and specific
    gravity were summarized.

3.  The percentages of lime and FeCl3 were calculated as follows.  In comput-
    ing the chemical dosages,  the total Ibs of solids in the tank were
    determined from the feed solids (unconditioned) measurement and the vol-
    ume.

    Ibs solids = 24.33 in x 6.4 gal/in x 1.0114 x 8.3453 Ib/gal x |
               =67.7 Ibs

    Lime added - 16.75 gal x 1 Ib/gal = 16.75 Ibs

        % lime = 16-75  x 100  =24.7%
                  o/. 7

    FeCl3 added = 5.6 gal x 1 Ib/gal = 5.6 Ibs

                       x 100  = 8.3%
                  67.7

4.  Filter cake corrected dry weight is given.  The final cake weight was
                                     129

-------
    corrected for the chemicals added.

                                     solids
    Cake dry weight = wet weight x

                    = 74 Ibs x
      100
39.2
                               100

                    = 29 Ibs

    Corrected dry weight = (cake dry weight )- (chemical weight)
    The chemical weight is known as a percentage of the incoming sludge
    solids; the proper formula is then:

    Corrected dry weight • "** dry weight
                3    °      1  + %  (lime +
                                       100
                                 29
                           1 + (24.7 + 8.3)
                                   100
                         - 21.8 Ibs

5.  Yield is reported as:

          	corrected dry weight
          (cycle time) x (filtration area)

    Corrected dry weight = 21.8 Ibs

    cycle time = 15 min + 18 min =.^3 ml"   = 0.55 hr
                                  60 mxn/hr

    Filtration area =62.4 ft2

    Yield =      21.8 Ibs
            (0.55 hr) x (62.4 ft2)

          - 0.63 lb/hr/ft2

    This yield is called the process yield, since it includes only pump and
    squeezing times.  It is a good figure for comparing runs in a set to
    establish trends.  For scale-up, the mechanical time of the full-scale
    press must be included in the "full-scale yield".  For the 500 m3 NGK
    press, this mechanical time is 19 minutes.  It includes time to open
    and close the press, discharge cake, and fill the squeezing chambers, etc.
    The full-scale cycle time is then:

          33 min + 19 min = 52 min  = 0.87 hr
                            60 min/hr

    Full-scale yield = 	21>8 lbs	9-            ,   ,   7
                       (0.87 hr) x (62.4 ft'1) = 0.40 lbs/hr/ft2


                                      130

-------
DATA SHEET 5
DATE 8/25/77
RUN $
REFERENCE RUN-(NGK)
CLOTH TYPE
SLUDGE TYPE
RATIO-PR. /SEC.
TEMPERATURE
CST (COND. SLUDGE)
SLUDGE pH (UNCOND./COND.)
LIME %
FERRIC %
FILTRATE VOLUME (TOTAL)
FILTRATE pH.
FILTRATE APPEARANCE
CAKE (DISCH./ CONSIST.)
CLOTH COND.CAFTER DISCH.)
CAKE WEIGHT
CAKE WEIGHT TOTAL
CAKE THICKNESS
TOTAL CYCLE TIME (MINS.)
PASSAVANT PRESS - OPERATIONAL DATA

1
#1

Sec + Prim
31.6/68.4
28° C
- '
-
20.1
6;7
39.7 gal
_
-
good
some sticking
#1 *2
87.5 Ibs
1.5 inches
160
REMARKS: Comparison with NGK and Nichols.
TIME
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120

125
130
135
140
145
150
155
160

PRESSURE
0
110
170
215
225
225
225

225

225

225

225

225

225

225

225

225

225

225

225

225

FILTRATE
VOLUME
inches
3-5/8
5-1/2
6-3/4
7-7/8
8-5/8
9-1/4

10-3/8

11-1/8

11-7/8

12-1/2

12-3/4

13-1/4

13-1/2

13-7/8

14-1/4

14-3/8

14-5/8

14-7/8

14-7/8

EILTRATE
RATE
.. .0 (sal)
9.7
14.7
18.0
21.0
23.0
24.5

27.7

29.7

31.7

33.4

34.0

35.4

36.0

37.0

38.0

38.4

39.0

39.7

39.7

    131

-------
Explanation of Data Sheet 5

     Data Sheet 5 was used to collect information during each run on the
Fassavant press.

1.  The data in the column on the left is self-explanatory.  Some information
    was taken from the referenced corresponding NGK run.

2.  Pressure and filtrate readings were normally taken every ten minutes
    during the run.  Filtrate was collected in a 100 gallon plastic vat.
    Vat calibration was 2.67 gal/inch.
                                        132

-------
 DATE    8/25/77
                                            DATA SHEET 6
                                    PASSAVANT  PRESS-DATA  SHEET
 RUN
 TYPE SLUDGE:
                                         Prim/Sec 1/2
 RATIO;  _f Pr./ Sec.
          31.6/68.4
 FEED SOLIDS:%SOL./% VOL.(uncond.)/pH.
         '5.65/60.A/-
 FEED SOLIDS:%SOL./% VOL.(cond.)/ 6H.
          7.59/42.9/-
 LIME  %:
                                         20.1
 FERRIC %:
                                         6.7
 CYCLE  TIME:(MINS.)
         160
TERMINAL  FEED PRESSUREfpsigj_:_
FILTRATE  VOLUME:  (gal.)
         225
                                         39.7
JFILTRATE pH.;
FILTRATE:(mg/1) TOTAL  SOLIDS
                                        8831
FILTRATE;(mg/1)  SUS.SOLIDS
                                        29
FILTER CAKE:(WET WEIGHT)
         87.5 Ibs
FILTER CAKE;(%SOL./ %VOL.)
         36.2/44.2
FILTER CAKE:(CORR. DRY WEIGHT)
                                        25.Q lt,g
CAKE THICKNESS;( INCHES )
         1.5
CAKE DENSITY: (Ibs./ft.3 )
         76.8
YIELD: (lbs./ft.2/hr.) Process
         0.52
YIELD:  (full scale)
                                        0.46
Terminal filtrate rate (eal/hr/ft^l
CLOTH TYPE:   T-167
TEMPERATURE: 28 °C
                                                               PRESS AREA:
                                                                            18.0 ft''
CONCLUSIONS:
                       Good run for 2/1  sludge
                                               133

-------
Explanation of Data Sheet 6

     Data Sheet 6 was used to summarize all data collected on each run
on the Passavant press.

1.  Data on the sludge ratio, the % feed solids, and the chemical rates
    were taken from the corresponding NGK run.

2.  Filtrate solids were determined from a composite sample of the filtrate.

3.  Cake solids were determined by averaging the % solids for each of the two
    cakes sampled.

4.  Corrected dry weight was determined by the same procedure described in the
    explanation for Data Sheet 4.

5.  Process yield was reported as:

        Corrected dry weight	
      (cycle time) "x (filtration area)

    Yield =  	25.0 Ibs	 •
            (160/60 hr) x (18.0 ft2)

          =0.52 lb/hr/ft2

6.  Full-scale yield was calculated by adding 20 minutes to the process
    cycle time (turn around time required for a large-scale press).

    Full-scale cycle time = 160 min + 20 min = 180 min   " 3.0 hr
                                               60 min/hr
         T, 11    i   *  i j        25-0 Ibs
         Full scale yield =
                            (3.0 hr) x (18.0

                            0.46 lb/hr/ft2
                                     134

-------
DATA SHEET 7
DATE 8/25/77
RUN t
REFERENCE RUN-(NGK)
CLOTH TYPE
SLUDGE TYPE
RATIO-PR. /SEC.
TEMPERATURE
CST (COND. SLUDGE)
SLUDGE pH (UN COND. /COND.)
LIME %
FERRIC %
FILTRATE VOLUME (TOTAL)
FILTRATE pH.
FILTRATE APPEARANCE
CAKE (DISCH. /CONSIST.)
CLOTH COND. (AFTER DISCH.)
CAKE HEIGHT crams
CAKE WEIGHT TOTAL
CAKE THICKNESS
TOTAL CYCLE TIME (MINS.)
NICHOLS PRESS - OPERATIONAL DATA

1
#1
4709/40
Sec +'Prim
31.6/68.4
28° C
_
_
20.1
6.7
23.250 ml
_
clear
excellent
pj-ean
i?l 2160 #2 2175
4335 sms - 9.55 Ibs
1"
130
REMARKS: Comparison with NGK and Passavant
TTME
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
Time Pressure Filtrate rate Total Volum
(min) (psig) (ml) (ml)
1 15 850 850
2 30 2500 3350
3 50 1500 4850
4 75 1950 6800
5 100 1300 8100
125
130
135
140
145
150
155
160
165
170
175
180
PRESSURE

100
115

95

110

85

107

93

100

103

95 ,

101

107

100

101










TILTRATE
VOLUME

8100
12300

16250

18600

20050

21100
*
21650

22090

22430

22690

23890

23030

23150

23250










EILTRATE
RATE

8100
4200

3950

2350

1450

1050

550

440

340

260

200

140

120

100










  135

-------
                                           DATA SHEET 8
                                  NICHOLS PRESS-DATA  SHEET
DATE 8/25/77
RUN #
TYPE SLUDGE:
RATIO: #/# Pr./ Sec.
FEED SOLIDS :%SOL./% VOL. (uncond.) /pH.
FEED SOLIDS :%SOL./% VOL.(cond.)/ gH.
LIME %:
FERRIC %:
CYCLE TIME:(MINS.)
TERMINAL FEED PRESSURE(psig) :
FILTRATE VOLUME : (MLS .)
FILTRATE pH. :
FILTRATE : (mg/1) TOTAL SOLIDS
FILTRATE : (mg/ 1) SUS . SOLIDS
FILTER CAKE: (WET WEIGHT)
FILTER CAKE:(%SOL./ %VOL.)
FILTER CAKE:(CORR. DRY WEIGHT)
CAKE THICKNESS :( INCHES )
CAKE DENSITY: (lbs./ft.3 )
YIELD: (lbs./ft.2/hr.) process
Terminal filtrate rate ml/m±n
Full scale yield (Ibs/ft2/hr)

I
Prim/Sec 1/2
31.6/68.4
5. 65/60. 4/-
7. 59/42. 9/-
20.1
6.7
130
101
23250
_
8013
49
9.55 Ibs
36.2/47.3
2.7 Ibs
1
71.5
0.31
10
0.27


































































CLOTH TYPE; 4709/40
TEMPERATURE:  28° C
PRESS AREA: 4 ft2
CONCLUSIONS:
               Good run  for 2/1 sludge
                                               136

-------
Explanation of Data Sheet 7
     Data Sheet 7 was used to collect information during each run on the
Nichols press.

1.  This data sheet is nearly identical to Data Sheet 5  and the same remarks
    are appropriate.

2.  The first five minutes of the run were programmed to increase the
    pressure in the feed tank slowly by using a regulator valve.

3.  Pressure and filtrate readings were normally taken every ten minutes
    during the run.  Filtrate was collected in a graduated cylinder and
    composited during the run.  The pressure fluctuated slightly because
    of other demands on the air supply system.


Explanation of Data Sheet 8
     Data Sheet 8 was used to summarize all data collected on each run on the
Nichols press.  This data sheet is identical to Data Sheet 6.
                                     137

-------
                                 APPENDIX C

             DETERMINATION OF SPECIFIC RESISTANCE TO FILTRATION


     From June through October 1977, laboratory measurements of the
filterability of the conditioned sludge mixtures were made using three
different techniques:

     1.  Modified Buchner funnel method

     2.  Positive pressure method

     3.  Capillary suction method

The modified Buchner funnel and pressure methods give quantitative
measurements of the filterability of a sludge by determination of its
average specific resistance to filtration.  The specific resistance is
defined as the pressure difference required to produce a unit rate
of filtrate flow of unit viscosity through a unit weight of cake and
is calculated from the following equation:

               R = 2PA2b
Where          R = average specific resistance
               P = the filtration pressure
               A = the filter area
               /*• = the viscosity of the filtrate
               c = the weight of cake solids per unit volume of filtrate
               b = slope of filtrate discharge curve, i.e. time/volume
                   versus volume

A more detailed discussion of the theoretical derivation of this equation
can be found in the literature.4

     The capillary suction test was developed as an alternative to the
Buchner funnel and pressure tests.  It measures the time required for
filtrate to drain from a given volume of sludge and is easily correlated
4.  Carman, P.C. "Fundamental Principles of Industrial Filtration (A
    Critical Review of Present Knowledge)."  Transactions of the- Institution
    of Chemical Engineers, 1938, 16, 168.
                                     138

-------
with the specific resistance parameter.
MODIFIED BUCHNER FUNNEL METHOD

     This technique is used most frequently in studies of sewage sludge
filtrations.5,6  Essentially, it consists of effecting the filtration
process by the use of a vacuum pressure.

Facilities

     The filter consisted of a white, porcelain Buchner funnel  (9.1 cm
plate diameter) .  Standard laboratory paper (9 cm diameter) graded for
a fast filtering speed was used as the filter media.  Suction was provided
by a 24 inch gage vacuum pump.  Accessory equipment included a  250 ml
graduated cylinder with side arm, stand, stopwatch, and thermometer.  The
funnel, with accessories, was assembled as shown in Figure C-l.

Operation

     A filter paper was wetted and placed in the bottom of the  funnel.  The
conditioned sludge was prepared by pouring twice from one beaker to another
to resuspend any solid particles which had settled out.  The temperature
of the sample was noted and a 200 ml portion was measured into  the funnel.
The vacuum was then started immediately, and simultaneous readings of
filtrate quantity and time were recorded as the filtrate collected.  These
readings were taken at 10 or 15 second intervals for an elapsed time of
240 seconds or until the vacuum broke, i.e. the filter cake was completely
formed .

Analysis

               R = Rv = 2PA2b
Experimental parameters were measured in the metric system and Rv reported
in units of cm/g.  A filtration vacuum differential of 24 inches Hg (81360
5.  Coackley, P. and B.R.S. Jones.  "Vacuum Sludge Filtration I.
    Interpretation of Results by the Concept of Specific Resistance.
    Sewage and Industrial Wastes, 1956,^, 963.

6.  Swanwick, J.D., F.W.  Lussignea, and K.J. White.  "The Measurement
    of the Specific Resistance to Filtration and Its Application in
    Studies of Sludge Dewatering."  Journal of the Institute of Sewage
    Purification, 1961, 6 487.
                                    139

-------
N/m^) Was used.  The filtration area was taken to be that of the filter
paper, 63.62 cm .   The filtrate viscosity.it, was assumed to be that of
water, measured at the temperature of the sample and converted to units
of Ns/m^.  The weight of dry cake solids per unit volume of filtrate, c
(g/cm3), was approximated from the sample feed solids concentration
according to the relationship:
                      TSS
                   1000-TSS

where TSS is the total suspended solids of the feed, g/1.

     The slope of the filtrate curve, b (s/cm6),  was obtained by plotting
6/V vs V, where V is the filtrate volume collected in time fi.  Carman^
noted that in the determination of b, the initial readings of 6 and V
represent the initial resistance of the filter medium rather than the
specific resistance of the solids.  Hence, -9 and V should not be measured
from the beginning of the filtration; rather, the filtration pressure
should be raised slowly to its full value to minimize the effect of this
initial resistance.  Once the pressure reaches constancy, the readings
should then be taken.  In our determinations, this procedure was not used;
pressures were raised immediately to full value and measurements of 0 and
V were taken from the start of the experiment.  However, in order to
circumvent the problem posed by resistance of the filter medium, these
first initial readings were not used in computing the slope of the
filtrate curve.  By doing this, the procedure was simplified and
standardized for all operating personnel, yet experimental accuracy was
still maintained.

     A sample of the data collected and its analysis is shown for this
method in the following data sheet.


POSITIVE PRESSURE METHOD

     This method has been used by several researchers"»^ during studies
of high-pressure filtration.  It is similar in operation to the modified
Buchner funnel method, except that the filtration pressure is provided
by positive instead of vacuum pressure.
    Carman, P.C. op. cit.

    Coackley, P and B.R.S. Jones.  "Vacuum Sludge Filtration I.
    Interpretation of Results by the Concept of Specific Resistance."
    Sewage and Industrial Wastes, 1956, 28, 963.

    "Pressure Filtration of Waste Water Sludge with Ash Filter Aid."
     Environmental Protection Agency (EPA) Technology Series, 1978,
     EPA-R2-73-231.

                                     140

-------
       Figure C-l.  Buchner Funnel Apparatus.
Figure C-2.  Passavant Series 275 Resistance Meter.
                        141

-------
Facilities

     A Passavant Series 275 Resistance Meter was used.  It included a 7 cm
diameter stainless steel body and support screen.  A 7 cm diameter filter
cloth together with a standardized laboratory filter paper served as the
filter media.  Required additional equipment included a cylinder of
compressed nitrogen gas, a 100 ml buret, stand, and.stopwatch.  The
instrument was assembled as shown in Figure C-2.

Operation

     The filter cloth was wetted and placed in the meter over the support
screen.  One of the laboratory filter papers was wetted and placed on top
of this cloth.  A 250 ml sample of the conditioned sludge was measured
into the meter, and the top was attached and secured.  The 100 ml buret
was initially filled to its lower 100 ml mark; and the meter was then
pressurized to 225 psig with the compressed gas.  Simultaneous readings
of filtrate quantity and time were subsequently recorded.  Readings were
taken at 15 second intervals for a period of 240 seconds or until the
filtration was completed and blow-by occurred, with gas passing through
the filter.

Analysis

     The specific resistance equation was revised by Passavant Corp. to give:

               R = Rp - 2PA2b . Kb
                        yW-C     c

In this equation, Rp is measured as a dimensionless quantity.  K is an
index constant developed by Passavant, measured in arbitrary units of
g-cnrVs.  It is a function of the pressure, temperature, and viscosity
of the sludge being dewatered.  The slope, b (s/cm°), of the filtrate
curve and the weight of dry cake solids per unit volume of filtrate, c
(g/cm^)» were obtained as described in the previous section.  It was noted
earlier that this pressure test is essentially a Buchner funnel test run
under positive pressure.  Results, therefore, could also be interpreted
by the same formula used for the Buchner funnel test.

     A sample of the data for this test and its analysis is shown in the
accompanying data sheet.
CAPILLARY SUCTION METHOD

     The .use of this method for determining the filterability of sewage
sludges was developed as the result of research studies at the Water
Pollution Research Laboratory in Stevenage, England. "»-^  It was developed
as an alternative to the modified Buchner funnel test to permit rapid
assessment of the filterability of a sewage sludge.
                                     142

-------
The principle of the method is that filtration occurs by the suction applied
to the sludge by the capillary action of a standard grade, absorbent
filter paper.  The rate at which the paper becomes wetted gives an
indication of the filterability of the sludge.

Facilities

     A capillary suction time meter, which consisted of two transparent
plates separated by a filter paper and an automatic timer, was used.  The
lower plate measured 9 cm x 9 cm x 0.6 cm high.  One edge of this plate
was raised to a height of 1.2 cm and served to position the filter paper.
(The type of filter paper used was the Whatman No. 17 chromatography grade).
The upper plate measured 7 cm x 9 cm x 2.3 cm high and contained a central
hole approximately 1.9 cm in diameter.  On the under side of the plate,
concentric with the center hole, were two circular marks of diameters 3.2
cm and 4.5 cm; these marks were connected electrically to an automatic
timer.  A stainless steel cylinder, 2.5 cm high and 1.8 cm inner diameter,
fitted into the central hole of the upper plate and served as a reservoir
for the sludge sample.

Operation

     The filter paper was positioned on the lower plate along the raised
edge.  The upper plate, with electrical  connections touching the filter
paper, was placed on top of the filter paper along the raised edge of the
lower plate.  The stainless steel cylinder was placed in the hole in the
upper plate and a small volume of the sludge sample poured into it.  As
the suction pressure of the filter paper drained the filtrate from the
sample, the automatic timer started when the outward progression of the
filtrate reached the first connection and stopped when it reached the
second.  The capillary suction time or CST (in seconds) was then read from
the timer.

     A picture of the CST instrument is shown in Figure C-3 and sample
data is shown in the data sheet.

Analysis

     The CST only provides an indication of the filterability of the sludge.
Through calibration with the modified Buchner funnel and pressure methods,
however,  it can be correlated with the specific resistance parameter.
(See Section 7,  Special Tests).
10.  Gale, R.S. and R.C. Baskerville.  "Capillary Suction Method for
     Determination of the Filtration Properties of a Solid/Liquid
     Suspension."  Chemistry and Industry, 1967, p 355.

11.  Gale, R.S. and R.C. Baskerville.  "A Simple Automatic Instrument
     for determining the Filtrability of Sewage Sludges."  Water Pollution
     Control, 1968, 67,  p. 233.
                                     143

-------
Figure C-3.  GST Instrument
            144

-------
Date:  7/28/77                                   Run: #1
Sludge Description:  Ratio Secondary/Primary @ 2/1
Conditioning Additives:  FeCl3 - 3.67 gals
                         Lime -11 gals
Test #1

Test #2
CST = 10.9 sec.

Pressure Method
       Time (9)
           s

          0
         15
         30
         45
         60
         75
         90
        105
        120
        135
        150
        165
        180
        195
        210
        225
        240
                 Test it 3
29 °C
Reading
3
100.0
84.0
75.8
68.8
63.2
58.0
53.2
48.6
45.0
41.0
37.2
34.0
31.0
27.8
24.6
21.8
18.8
Modified
Temperature :
Time (6)
s
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
Pressure:
(v) Volume (V)
cm3
0
16.0
24.2
31.2
36.8
42.0
46.8
51.4
55.0
59.0
62.8
66.0
69.0
72.2
75.4
78.2
81.2
Buchner Funnel Method
29 °C Pressure: 24 " Hg
Volume (V)
cm3
0
34
48
54
64
72
76
86
90
96
100
104
108
112
116
120
Vacuum broke @ 2'
225 psig
e/v
s/cm3
—
0.938
1.240
1.442
1.630
1.786
1.923
2.043
2.182
2.288
2.389
2.500
2.609
2.701
2.785
2.877
2.956


e/v
s/cm
_
0.294
0.417
0.556
0.625
0.694
0.789
0.814
0.888
0.938
1.000
1.058
1.111
1.161
1.207
1.250
38"
                                     145

-------
                                 ANALYSIS

Test # 1     GST = 10.9 sec

Test # 2     Pressure Method

     a)  Calculation of c

         Unconditioned sludge:

              21.0 inches x 6.4 gal x 3.785 I = 500.76 1
                               inch        gal

         500.76 1 x 5.96% x 0.995 g x 1000 ml - 29725.65g
                    100%          ml       ~
         Lime:
             11 gals x 3.785 JL = 41.64 1
                            gal

             11 gals x 1 Ib x 454 & = 4994.Og
                        gal      Ib
         Fed :
             3.57 gals x 3.785 ^= 13.89 1
                              gal

             3.67 gals x 1 Ib x 454 jg_ = 1666.18g
                          gal      Ib
Sludge
Lime
FeCl3
TSS =
TSS
29725.65
4994.00
1666.18
36385.83
36385.83
556.29 "
- n n?n ol,
500.76
41.64
13.89
556.29
65.41 g/1
            1000 - TSS          °'

     b)  @ 29 °C and 225 psig,  K = 5.43 g-cm3  (see chart)
     c)  from graph, b = 0.0235 s/cm
     d)  RP=
              c
                                 146

-------
               = 5.43 g-cm3/s x 0.0235 s/cm6
                         0.070 g/cm3

               = 1.82 (dimensionless)

e.  Alternatively, if Rp is analyzed according to the equation

                            Rp = 2PA2b
                                /*c

             A = 38.5 cm2; A2 = 1482.3 cm4

             P = 225 psig = 155.133 x 104 N/m2

             @29 °C,yu<.H20 = 8.21 x 10~4 Ns/m2

             Rp = 2PA2b
                  ^c

             = 2 x 155.133 x 104 N/m2 x 1482.3 cm4 x 0.0235 s/cm6
                            8.21 x 10~4 Ns/m2 x 0.070 g/cm3

             = 1.88 x 1012 cm/g

    Test #3  Modified Buchner Funnel Method

             a)  c = 0.070 g/cm3

             b)  P = 24"Hg = 81360 N/m2

             c)  @ 29°C,^.H20 = 8.21xlO~4 Ns/m2

             d)  A - 63.6 cm; A2 =4047.5 cm4

             e)  from graph, b = 0.00775 s/cm6

             f)  Rv = 2PA2b
                      ,xc


                    = 2 x 81360N/m2 x 4047.5 cm4 x 0.00775s/cm6
                                   8.21xlO~4 Ns/m2x 0.070  g/cm3

                    - 8.88 x 1010 cm/g
                                 147

-------
DATE-  7-28-77
NGK
                                     SQUEEZING PRESSURE-
                                 213
                                         CLOTH TYPE  NY 51-4
SLUDGE TYPE-   Secondary/Primary
                    APPEARANCE-   Black and grainy
TEMPERATURE- GRINDING-n»azorated PURPOSE OF RUN- Comparison Study
RUN t
PRIflARY/SECONDARY (pH) „ ,
pll
COMBINED SLUDGE-uncond.-cond.
LEVEL-before& after cond.(INCH)
LEVEL (after pumping)
LIME ADDED (GALS)
FECL3 ADDED(GALS)
SLUDGE PUMPED (GALS)
FILTRATE COLLECTED (PUMPING)
FILTRATE COLLECTED (SQUEEZE)
FILTRATE COLLECTED (TOTAL)
FILTRATE:pH
FILTRATE APPEARANCE
CAKE .'WEIGHT (WET)
CAKE: CONSISTENCY/DISCHARGE
PUMPING TIME
'PUMPING PRESSURE (TERMINAL)
SQUEEZING TIME
CLOTH WASH: before run
CAKE THICKNESS
CST (of conditioned sludge)
CAKE DENSITY (g/cm )
TANKED DRAINED
PRIMARY- SPECIFIC GRAVITY-
SECONDARY-SPECIFIC GRAVITY-
COMBINED -UNCOND. (SP.GR.)
COMBINED-COND. (SP.GR.)
fl
- 1 -
- I -
n
/
I
27/241/2/317/8 /
44 7/8
11
3.67





yellowish
105 #
Excellent
17
94
19
yes
1/2 to V4
12.6
1.175
ves
1.0154
1.0
0.996
1.0205























#3
/
/
/























#4
/
/
/























#5
/
/
/























 PRIMARY-INCHES-
                             SECONDARY-INCHES-  17
                                                                 RATIO-!?/* PRI./SEC.   32/68
 REMARKS:
             Primary  Solids  =  10.8%
             Secondary  Solids  • 5.76%
             Combined  Solids = 5.96%
                                              148

-------
                                  TABLE 7.  EXTENDED RUNS - 3/8/77
                                       FILTRATE  mg/1
% CHEMICALS
LIME/FeCl3
26.7/7.8
26.8/7.8
26.8/7.8
29.3/8.5
29.3/8.5
29.3/8.5
30.0/8.8
30.0/8.8
pli
11.9
11.9
11.9
11.7
11.7
11.8
11.8
11.9
BOD
554
916
694
684
682
684
420
374
COD
1560
2067
2061
1640
1668
1785
1730
1867
P°4
35.8
110
63.6
44.7
56.7
69 .-2
47.5
64.0
TKN
173
253
214
244
240
296
226
258
NH3
85.3
Ip7
105
106
96.9
99.2
96.9
110
N03
.70
.'74
.80
.75
.77
.79
.74
.72
TOTAL
ALKALINITY
1801
2367
1987
2021
1961
1959
2063
2008
TOTAL
SOLIDS
7286
9431
8575
8307
8203
8140
8471
8474
SUSPENDED
SOLIDS
144
1540
464
151
62
55
127
47
28.5/8.3      11.8    626    1923   61.4   238   101    .75      2020       8361

                              TABLE 8.  FILTRATE QUALITY VS. CHEMICAL CONDITIONING
323  (Average)
DATE
3-10-77
3-10-77
3-10-77
3-10-77
7-12-77
7-12-77
7-12-77
11-1-77
11-1-77
11-1-77
% CHEMICALS
LIME/FeCl3
45.5/13.3
34.4/10.0
22.8/6.7
17.1/5.0
16.8/5.6
14.4/4.9
12.2/4.1
25.2/8.3
19.7/6.6
14.8/5.0
% CAKE
SOLIDS
39.6
38.5
36.0
31.6
41.8
39.9
29.0
38.5
36.8
34.5
CAKE
DISCHARGE
excellent
excellent
excellent
good
excellent
excellent
poor
excellent
excellent
good
TOTAL SOLIDS
mg/1
10120
8910
8189
9404
9815/10130*
9736/10024*
8963/8057*
6773
5879
5546
                                                                      SUSPENDED SOLIDS
                                                                          mg/1               pH

                                                                           87                11.5
                                                                           70                11.6
                                                                           80                11.5
                                                                         2604                11.5

                                                                         120/18*             11.5
                                                                         216/45*             11.5
                                                                         438/198*            11.5

                                                                           28
                                                                           22
                                                                          189
            * First entry is from the pump cycle;  second entry from the squeezing cycle.

-------
          3.O
          E.O
        n
         o
         N
         0
           1.O
u-  TBBT  S  PRBBBURB
             MBTHOO
                                            bi 0.0300  S/CMB
                                 BO
                               V CCM3)
                                                      100
              3  MODIFIED  BUCHNER
                 FUNNEL  METHOD
  1.O
u
X
O
                                                            s/cM
                                         1OO
                                                            ISO
                                 150

-------
 K-Factor as a Function of Temperature
Temperature
°C °F
0

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
32
33
34
35
32.0

33.8
35.6
37.4
39.2
41.0

42.8
44.6
46.4
48.2
50.0

51.8
53.6
55.4
57.2
59.0

60.8
62.6
64.4
66.2
68.0

69.8
71.6
73.4
75.2
77.0

78.8
80.6
82.4
84.2
86.0

87.8
89.6
91.4
93.2
95.0
K
2.48

2.57
2.66
2.74
2.83
2.93

3.02
3.11
3.21
3.30
3.40

3.49
3.59
3.69
3.80
3.90

4.00
4.10
4.21
4.31
4.43

4.53
4.64
4.75
4.86
4.97

5.09
5.20
5.31
5.43
5.55

5.67
5.79
5.91
6.03
6.15
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Temperature
°C °F


36
37
38
39
40

41
42
43
44
45

46
47
48
49
50

51
52
53
54
55

56
57
58
59
60

61
62
63
64
65

66
67
68
69
70


96.8
98.6
100.4
102.2
104.0

105.8
107.6
109.4
111.2
113.0

114.8
116.6
118.4
120.2
122.0

123.8
125.6
127.4
129.2
131.0

132.8
134.6
136.4
138.2
140.0

141.8
143.6
145.4
147.2
149.0

150.8
152.6
154.4
156.2
158.0
K


6.27
6.40
6.52
6.65
6.77

6.90
7.03
7.16
7.29
7.42

7.55
7.68
7.82
7.95
8.09

8.23
8.36
8.50
8.63
8.77

8.91
9.05
9.20
9.34
9.48

9.62
9.77
9.91
10.1
10.2

10.3
10.5
10.6
10.8
10.9
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Temperature
°C °F


71
72
73
74
75

76
77
78
79
80

81
82
83
84
85

86
87
88
89
90

91
92
93
94
95

96
97
98
99
100








159.
161.
163.
165.
167.

168.
170.
172.
174.
176.

177.
179.
181.
183.
185.

186.
188.
190.
192.
194.

195.
197.
199.
201.
203.

204.
206.
208.
210.
212.








8
6
4
2
0

8
6
4
2
0

8
6
4
2
0

8
6
4
2
0

8
6
4
2
0

8
6
4
2
0






ic


11.1
11.2
11.4
11.5
11.7

11.9
12.0
12.2
12.3
12.5

12.6
12; 8
12.9
13.1
13.2

13.4
13.5
13.7
13.8
14.0

14.2
14.3
14.5
14.6
14.8

15.0
15.2
15.3
15.5
15.7






This table subsumes a constant  pressure of 225 psig, a filtration area
equal to that of %he PASSAVANT  Series 275 Resistance Meter, and the
dynamic viscosity of the  filtrate  to be equivalent to that of water.

Source:  Passavant Corporation
                                   151

-------
                                APPENDIX D

                            MATERIAL BALANCE
Data for this test was collected on 10/18/77.  Refer to Data Sheet at end
of this section.
             Sludge
             Chemicals
                                             Filtrate

                                             Slowdown
                                   Filter
                                    Cake
   INPUT,  Ibs total  solids

   Sludge     37.5
   Chemicals  10.0
                        OUTPUT,  Ibs total solids

                        Filter Cake
                          Sludge            32.7
                          Chemicals          8.7
                        Filtrate             4.2
                        Sludge Slowdown
                          Sludge             1.6
                          Chemicals          0.4
                        Filtrate Slowdown    0.2
   Total
47.5
Total
47.8
   Error in Analysis = 47'8~47'5 x 100  = 0.63%
                         47.5
                                    152

-------
Calculations

1.  Input

    Solid input to the press consists of the sludge and chemical solids in
    the feed.  The following parameters were measured:

              i.   feed volume to the press - 77.6 gals
             ii,   solid content of feed - 7.10%
            iii.   specific gravity of feed - 1.033
             iv.   chemical content of feed - 26.7% of dry sludge solids

    The total mass of solids in the feed is calculated as:
    Mass total solids - Mass Feed x % solids
                      - (77.'6 gals x 8.345 -f- x 1.033) x
                                           gal            100/i

                      = 47.5 Ibs

    and consists of both chemical and sludge solids.

    The mass of chemical solids is:
    Mass chemical solids = Mass total solids x % chemical solids

    Chemical content of the feed is  26.7% of the dry  sludge solids;  therefore
    based on the total feed solids,  i.e., sludge + chemicals,  the chemical
    content  is :


                                  x  100  - 21.1%
                            126.7


   Mass  chemical  solids  =  47.5  Ibs  x  '
                         =  10.0  Ibs

    The mass  of  dry  sludge  solids  is:
    Mass  sludge  solids =  Mass  total  solids  - Mass  chemical  solids
                      =  37.5  Ibs
                                     153

-------
INPUT SUMMARY

    chemical solids         10.0 Ibs
    sludge solids           37.5 Ibs
                            47.5 Ibs

2.   Output

    Output from the press consists of solids in the cake, filtrate and blow
    down.  The following output variables were measured:

             i.  Filter cake weight - 101.0 Ibs
            ii.  Solid content of filter cake - 41.0%
           iii.  Volume of filtrate collected - 58.3 gals
            iv.  Total solids concentration of filtrate - 8569 mg/1
             v.  Volume of blowdown - 6.4 gals

    a.  Filter Cake

        Total mass of the cake is 101.0 Ibs; total mass of the solids frac-
    tion of the cake is calculated as:

    Mass total solids = Mass cake x % solids

                      -101.0 Ibs


                      - 41.4 Ibs

    The mass of chemical solids in the cake is calculated as:
    Mass chemical solids = Mass total solids x % chemical solids

                           /i / i i_    ^ J. • J_ /o
                         = 41.4 Ibs x


                         =8.7 Ibs

    and  the mass of sludge solids is:
    Mass sludge solids = Mass total solids - Mass chemical solids

                       =32.7 Ibs

    b.   Filtrate

         A  total volume of 58.3 gallons of filtrate was  collected during  the
    filter press cycle.  The mass of  solid  particles within  the filtrate
    is calculated  as:
    Mass filtrate  solids = Volume filtrate x Concentration filtrate  solids
                            58.3  gal  x 8569  mg/1  x 8.345  x 10~6  mg/f&

                            4.2 Ibs
                                     154

-------
c.  Slowdown
     The blowdown volume consists of sludge and filtrate collected during
the blowdown cycle.  It is assumed that the blowdown is composed of
equal volumes of each.

    Sludge Blowdown
     The volume of sludge collected during blowdown is assumed to be 3.2
gallons.  This sludge has a total mass of:
Mass sludge blowdown = Volume sludge blowdown x density sludge
                     - 3.2 gal x 8.345 Ib/gal x 1.033
                     = 27.6 Ibs

The total mass of solids within the blowdown is, therefore:
Mass total solids = Mass sludge blowdown x % solids

                  , 27.6 Ibs *

                  = 2.0 Ibs

The mass of chemical solids is calculated as:
Mass chemical solids = Mass total solids x % chemical solids

                       0 n ...     21.1%
                     =2.0 Ibs x
                                 100%
                     = 0.4 Ibs

and, the mass of sludge solids is:
Mass sludge solids = Mass total solids - Mass chemical solids
                   =1.6 Ibs

    Filtrate Blowdown
     The volume of filtrate collected is assumed to be 3.2 gallons.  The
mass of solid particles within the filtrate is, therefore, calculated
as:
Mass filtrate solids = Volume filtrate x Concentration filtrate solids

                     =3.2 gal x 8569 mg/1 x 8.345 x 10~6

                     =0.2 Ibs

OUTPUT SUMMARY

      Filter Cake          Ibs
         Sludge solids     32.7
         Chemical solids    8.7

      Filtrate              4.2
      Sludge Blowdown
         Sludge solids      1.6
         Chemical solids    0.4

      Filtrate blowdown     0.2
                   Total   47.8


                                     155

-------
                                    FILTER PRESS DATA
DATE- RUN* - 10/18/77
TYPE SLUDGE-
RATIO- tf/#-PR./SEC.
FEED SOLIDS-%SOL./%VOLr(uncond.)
PRIMARY-%SOL./%VOL/ pH.-
SECONDARY-%SOL./%VOL./ pH.-
FEED SOLIDS-%SOL./%VOL.-(cond.)
PRIMARY- SP. GR. (gr./cc.)
SECONDARY-SP.GR.-(gr./cc.)
FEED SOLIDS-SP.GR. -(uncond.)- pH.-
FEED SOLIDS-SP.GR. -(cond.)- pH . -
LIME (added) %-
FECL3 (added) %-
VOLUME-(feed to Dress)-GALS.
PUMP' TIME- (Mins.)
SQUEEZE TIME-(Mins.)-
TERMINAL PRESSURE -Pump psig.-
SQUEEZING PRESSURE-psig
FILTRATE VOLUME -_£gals .) PIMPING-
FILTRATE VOLUME-(Gals-) SQUEEZE-
FILTRATE VOL. -(Gals.) TOTAL+B.D.
FILTRATE pH . 	
FILTRATE (mg/1) TOIAL SOLIDS-
FILTRATE (mg/1) SUS. SOLIDS-
FILTER CAKE-(Wet weight) Ibs.
FILTER CAKE-(%Sol./%Vol.)
FILTER CAKE (Dry Weight) Ibs.
CAKE THICKNESS- (Inches)
YIELD (lbs./ft.2hr_.) 	

Pri/Sec @ 1/2
30.7/69.3
5.77/65.2
11.1 - 68.8
5.1 - 61.6
7.1 - 42.4
1.034
1.014
1.008
1.033
20.1
6.6
77.6
14
15
75
213
52.0
6.3 - 58.3
6.4 - 64.7

8569
59
loi.o
41.0 48.3
32.7
3/8 3/4
1.08























































































CLOTH TYPE-NY 51-4TEMPERATURE-
                                       PRESS MECHANICAL TIME- Q min.   PRESS FILTER AREA-fi?.A  ft-2
CONCLUSIONS
                                              156

-------
                                 APPENDIX E

                             DERIVATION OF COSTS
     Costs are derived for a plant generating 250 dry tons/day (500,000
Ibs/day) of sludge solids (roughly equivalent to a wastewater flow of
200-250 MGD).
VACUUM FILTER

Number of Units

Full-scale yield =3.0 lb/hr/ft2
Filtration area = 600 ft2/unit

     Number of units - 	500,000 Ib/day	
                      3.0  Ib/hr/ft2 x 24 hr/day x 600 ftz/unit

                     = 11.6 units
                     or 12 units + 1 spare = 13 units

Capital Costs

1.  Filters—from Komline-Sanderson, the cost per unit is $158,000.

         Total installed cost =  $158,000  x  13  units  x  3 =  $6,162,000

2.  Lime System—for feeding an average of 50 tons/day

         Total installed cost = $1,000,000

    and includes conveyors, bins, slakers, pumps, etc.

3.  Ferric Chloride System—for feeding 16.7 tons/day of 10% solution

         Total installed cost = $500,000

    and includes storage tanks, pumps, etc.

4.  Conveyors—to transport filter cake to next process.

         Total cost = $1,000,000
                                      157

-------
5.  Total Capital Costs—
         Filters                           $6,162,000
         Lime                               1,000,000
         Ferric Chloride                      500,000
         Conveyors                          1,000,000
                              Total        $8,662,000
Annual Costs

1.  Amortization—

         Amortized cost = $8,700,000 x 0.09 = $783,000

2.  Chemicals—chemical usage consists of lime @ 20%, ferric chloride
    @ 7%, and anti-sealant (to counteract lime scale).

         Lime cost = 100,000 Ib/day x $.022/lb x 365 days/yr = $803,000

         FeCls cost = 35,000 Ib/day x $.065/lb x 365 days/yr = $830,375

         Anti-sealant cost = $30,000

         Total chemical costs = $1,663,375

3.  Power—power co'sts assume 100% duty cycle usage.

         Filters - 12 units @ 90 Hp/unit                 1080 Hp
         Sludge pumps - 12 @ 10 Hp/pump                   120
         FeCls system                                      22
         Lime system                                       87
         Conveyors                                         37
                                             Total       1346 Hp

         Power cost = 1346 Hp x .746 Kw/Hp x $.04/Kw-hr x 8760 hr/yr
                    = $ 351,900

4.  Water—the cloth washing system requires 60 gpm/unit.

         Water cost = 60  gpm/unit x 12 units x $.25/1000 gal x 525,600 min/yr
                    = $94,600

5.  Operating Labor—labor costs assume each crew consists of 1 supervisor,
    5 men to operate the  filters and 1 man to operate the chemical -system.
    To  cover a 7 day/week operation, 4 crews will be required, and  28 man-
    years  (7 men/crew x 4 crews) will be  expended.

         Labor cost = 28  man-years x $21,000/man-year =  $588,000
                                      158

-------
 6.   Maintenance—maintenance costs consist of the costs for both normal
     maintenance (materials and labor) and filter cloth replacement.   Normal
     maintenance is based on 2% of the purchase price of all equipment,  i.e.
     (Total capital cost)/3.   The filter cloths must be replaced once every
     2000 hrs at a cost of $550 per cloth.   The labor costs for changing
     the cloths are included in the cost for operating labor.

         Normal maintenance cost =  $8,662,000 x $.02 = $58,000


         Cloth replacement cost = 8760 hr/yr    	 x 12 units x $55o/cioth
                                  2000 hr/cloth/unit
                                = $29,150

         Total maintenance costs = $87,150

 7.   Total Annual  Costs—

         Amortization                     $   783,000
         Chemicals                        1,663,000
         Power                              351,900
         Water                               94,600
         Labor                              588,000
         Maintenance                          87,150
                               Total      $3,567,650


Unit Cost

For processing 250 tons/day of dry sludge solids.

        Unit  cost = 	$ 3,567,650/yr	 m $39>10/ton
                    250  tons/day x 365 days/yr   *jy-lu/ton


FILTER PRESS—VARIABLE VOLUME UNIT

Number of Units

Full-scale yield =0.49 lb/hr/ft2
Filtration area = 5380 ft2/unit

    Number of units = 	500.000 Ib/day	
                      0.49 lb/hr/ftz x 24 hr/day x 5380 ftz/unit

                    =7.9 units
                    or 8 units + 1 spare = 9 units
                                    159

-------
Capital Costs

1.  Presses—from Envirex, the cost for 9 units is $6,500,000.

        Total installed cost = $6,500,000 x 3 <= $19,500,000

2.  Chemical System—for lime and FeCl3J see vacuum filter costs.

        Total installed cost = $1,500,000

3.  Flight Conveyors—

        Total installed cost = $2,000,000

4.  Total Capital Costs—

        Presses                           $19,500,000
        Chemical System                     1,500,000
        Flight Conveyors                    2,000,000
                                  Total   $23,000,000


Annual Costs

1.  Amortization—

        Amortized cost = $23,000,000 x  .09 = $2,070,000

2.  Chemicals—for lime, Fed,, and anti-sealant; see vacuum filter costs.

        Chemical cost = $1,663,400

3.  Power—costs assume 100% duty cycle usage

Press j>lus accessories—

        Power usage = 37 Kw-hr/ton

Associated systems—

        Lime system                       81 Hp
        FeCl3 system                      14
        Sludge pumping-4 pumps @
          15 Hp ea.                       60
        Conditioning system               20
        Conveyors                        260
                                         435 Hp

        Power usage = 435 Hp x  .746 Kw/Hp x 24 hr/day =  31.2 Kw-hr/ton
                                    250 ton/day
                                     160

-------
Power Costs—total power usage is 68.2 Kw-hr/ton

        Power cost = 68.2 Kw-hr/ton x 250 ton/day x 365 days/yr x $.04/kw-hr
                   = $248,900

4.  Water—filter cloths will require washing once every 20 press cycles.
    With a 54-minute cycle per press, the filter cloths will be washed
    1.35 times per day and each wash will consume 3000 gallons.  City
    water will be used.

        Water cost = 3000 gal/cycle x 1.35 cycles/day/unit x 365 days/yr

                              x 8 units x $.5267/1000 gal
                   = $6400

5.  Operating Labor—costs assume each crew consists of 1 supervisor, 4 men
    to operate the presses, and 1 man to operate the chemical system.
    Four crews will be required for a 7 day/week operation and 24 man-years
    will be expended.

        Labor cost = 24 man-years x $21,000/man-year = $504,000

6.  Maintenance—maintenance costs consist of the costs for normal equipment
    maintenance and filter cloth and diaphragm replacement.  Normal
    equipment maintenance is based on 2% of the purchase price of all
    equipment.   Filter cloths will require replacement once every 3000 cycles
    at a material cost of $6500/unit.   Diaphragms will require replacement
    once every 20,000 cycles at a material cost of $26,000/unit.   Operating
    labor will change the filter cloths and diaphragms; 1.5 man-years will
    be expended.

        Normal maintenance cost = $23,000,000/3 x .02 = $153,000

        Cloth replacement cost =  27  cycles/day x 365  days/yr x $6500/unit
                                 3000  cycles

                                               x 9 units

                               =  $195,000

        Diaphragm replacement  cost = 27  cycles/day x  365 days/yr
                                    20,000 cycles

                                                  x  $26,000/unit  x 9  units

                                  = $117,000

        Labor cost =1.5 man-years x $21,000/man-year = $31,500

        Total maintenance cost =  $496,500
                                    161

-------
    Total Annual Costs

        Amortization                      $ 2,070,000
        Chemicals                           1,663,400
        Power                                 248,900
        Water                                   6,400
        Operating Labor                       504,000
        Maintenance                           496,500
                                  Total   $ 4,989,200
Unit Cost
For processing 250 tons/day of dry sludge solids

        Unit cost = 	$ 4,989,200/yr	 0 $54.68/ton
                    250 tons/day x 365 days/yr
FILTER PRESS—HIGH-PRESSURE FIXED VOLUME UNIT

Number of Units

Full-scale yield = 0.31 Ib/hr/ft2
Filtration area = 11625 ft2/unit
Chamber size = 40 mm (1.57 inches)

        Number of units =              ,   .„ ,50?'000         >;
                          0.31 Ib/hr/ft2 x 24 hr/day x 11625 ft2/unit
                        =5.8 units
                        or 6 units + 1 spare = 7 units

Capital Costs

From Passavant, the cost for 7 units, including all chemical systems,
conveyors, etc., is $8,450,000.

        Total installed cost = $8,'450,000 x  3 = $25,350,000

Annual Costs

1.  Amortization—

        Amortized cost = $25,350,000 x  .09 = $2,281,500

2.  Chemicals—for lime, FeCl3, and anti-sealant;  see vacuum filter  costs.

        Chemical cost = $1,663,400

3.  Power—for 6 presses in operation, power usage is 57.3 Kw-hr/ton

        Power  cost =57.3 Kw-hr/ton x 250 tons/day x 365 days/yr  x $.04/Kw-hr
                   =  $210,000

                                     162

-------
 4.  Water—filter cloths will require washing once per month;  each wash
     will consume 36,000 gallons of water.

         Water cost = 36,000 gal/cycle x 1  cycle/mo/unit x 12 mo/yr

                                 x 7 units  x $ .5267/1000 gal

                    = $1600

 5.  Operating Labor—costs assume each crew consists  of 1 supervisor,
     3 men to operate the presses, and 1 man to operate the chemical system.
     For a 7 day/week operation,  20 man-years will  be  expended.

         Labor cost = 20 man-years x $21,000/man-year  = $420,000

 6.  Maintenance—maintenance costs consist  of the  costs for normal
     equipment maintenance and filter  cloth  replacement.   Equipment
     maintenance is based on 2% of the purchase price  of all equipment.
     Filter  cloths will  require replacement  once per year at a cost of
     $240,000 for materials and $40,000 for  labor (includes  2 men on
     day shift year-round).

         Normal maintenance cost = $25,350,000/3 x  .02  = $170,000

         Cloth replacement  cost =  $240,000

         Labor cost  =  $40,000

         Total maintenance  costs =  $450,000

 7.   Total Annual  Costs—

        Amortization                      $2,281,500
         Chemicals                          1,663,400
        Power                                210,000
        Water                                  1,600
        Operating Labor                      420,000
        Maintenance                          450.000
                                 Total    $5,026,500


Unit Cost

For processing 250 tons/day of dry sludge solids.

        Unit cost =	$5,026,500/yr.       = $55.08/ton
                    250 tons/day x 365 days/yr
                                    163

-------
FILTER PRESS - LOW-PRESSURE FIXED VOLUME UNIT

Number of Units

Full-scale yield =0.22 Ib/hr/ft2
Filtration area = 6760 ft^/Unit
Chamber size = 32 mm (1.25 inches)
                            	500,000 Ib/day	
          Number of units = TJ.22 Ib/hr/ft2 x 24 hr/day x 6760 ft2/unit
                          = 14 units
                          or 14 units + 1 spare = 15 units


Capital Costs

1.  Presses—from Nichols, the cost for each press is $400,000.

          Total installed cost = $400,000/unit x 15 units x 3 = $18,000,000

2.  Chemical System—for lime and FeCl3, see vacuum filter costs.

          Total installed cost = $1,500,000

3.  Conveyors—

          Total installed cost = $2,300,000

4.  Total Capital Costs—

          Presses                         $18,000,000
          Chemical System                   1,500,000
          Conveyors                         2,300.000
                              Total       $21,800,000


Annual Costs

1.  Amortization—

          Amortized cost = $21,800,000 x .09 = $1,962,000

2.  Chemicals—for lime, FeCl3, and anti-sealant; see vacuum filter costs.

          Chemical cost = $1,663,400

3.  Power—

Press—includes sludge and chemical feed systems.

          Power usage = 65.3 Kw-hr/ton


                                    164

-------
 Transfer conveyors—power usage is  330 Hp.

      Power usage = 330 Hp x .746 Kw/Hp x 24  hr/day  =23.6  Kw-hr/ton
                             250 tons/day


 Power costs -  total power usage is  88.9 kw-hr/ton

      Power cost  =88.9 Kw-hr/ton x  250 tons/day x 365  days/yr x  $.04/Kw-hr
                 = $324,500

 4.  Water—filter cloths  on one  press  only  will be washed each  day.  Each
    wash will consume  5000 gallons of  water.

           Water  cost = 5000 gal/cycle  x 1 cycle/day x  365 days/yr
                                       x $.5267/1000 gal = $1,000

 5.  Operating Labor—costs assume each crew consists of 1 supervisor,
    7 press  operators,  and 1 man to operate the chemical system.  For
    a 7 day/week  operation, 36 man-years will be expended.

     Labor  cost =  36 man-years x  $21,000/man-year = $756,000

 6.  Maintenance—maintenance costs consist of the costs for normal equip-
    ment  maintenance  and  cloth  replacement.  Equipment maintenance is
    based on 2% of the purchase price  of all equipment.  Filter cloths
    will require replacement once per  year at a cost of $4600/unit for
    materials and  $6,000  for labor (600 man-hr/yr).

     Normal maintenance cost = $21,500,000/3 x .02  = $143,000

     Cloth replacement cost = $4600/unit x 15 units  = $69,000

     Labor cost = $6,000

     Total maintenance costs = $218,000

7.  Total Annual Costs—
          Amortization                    $1,962,000
          Chemicals                        1,663,000
          Power                              324,500
          Water                                1,000
          Operating Labor                     756,000
         Maintenance                        218,000
                                 Total     $4,924,500
                                    165

-------
Unit Cost

To process 250 tons/day of dry sludge solids.

     Unit cost =    $4»924 500/yr    	= $53.97/ton
                 250 tons/day x 365 days/yr


BELT PRESS

Number of Units

Full-scale yield = 675 Ib/hr/m width
Belt width = 3m/unit

     M  ,     ,          	500,000 Ib/day	
     Number of units =  ,   T^/U—7	o7~~u—T<—	5—7	TT = 10 3 units
                        675 Ib/hr/m x 24 hr/day x 3m/unit   J-".j UU.LLS>
                                                          or 11 + 1 spare=12
Capital Costs                                               units.

1.  Presses—from Komline Sanderson, the cost per unit  is $147,000.

     Total installed cost = $147,000/unit x 12 units x  3
                          = $5,300,000

2.  Polymer Feed System - includes storage, mixing, pumping, etc.

     Total installed cost = $750,000

3.  Conveyors—

     Total installed cost = $1,000,000

4.  Total Capital Costs—

          Presses                         $5,300,000
          Chemical System                    750,000
          Conveyors                        1,000,000
                               Total      $7,050,000


Annual Costs

1.  Amortization—

     Amortized cost = $7,050,000 x .09 = $634,500

2.  Chemicals—test work indicated $9.00 per ton of  sludge processed.
    However, because of uncertainty of polymer suitability, assume
    $15.00 per ton.

     Chemical cost = 250 tons/day x 365 days/yr x $15.00/ton
                   = $1,368,800

                                     166

-------
  3.  Power—

      Press - 12.75 Hp/unit x 11 units                140 Hp
      Sludge pumps - 10 Hp/unit x 11 units            HO
      Polymer system                                   31
      Conveyors                                        30
                                           Total      311 Hp

      Power cost = 311 Hp x .746 Kw-hr/Hp x 8760 hr/yr x $.04/Kw-hr
                 = $81,300

 4.  Water—each unit will consume 75 gpm.

      Water cost = 75 gal/min/unit x 11 units x 525,600 min/yr
                                   x $.25/1000 gal
                 = $108,400

 5.  Labor—costs assume each crew consists of 1 supervisor, 6 men to
     operate the presses,  and 1 man to operate the chemical system.   For
     a 7 day/week operation,  32 man-years will be expended.

     Labor  cost = 32 man-years x $21,000/man-year = $672,000

 6.   Maintenance—raantenance  costs  consist of  the costs  for  normal
     maintenance and belt  replacement.   Normal maintenance  costs  for
     materials and labor are based  on  3% of the purchase  price of all
     equipment.   Belt replacement costs  will total $20,000 per year.

     Normal maintenance cost =  $7,050,000/3 x .03 = $70,500

     Belt replacement cost = $20,000

     Total maintenance costs =  $90,500

 7.   Total Annual  Costs—

     Amortization                       $  534 500
     Chemicals                           1,368,'800
     Power                                  81,300
     Water                                 108,400
     Operating Labor                       672,000
     Maintenance                            90 500
                                        $2,955,500


Unit Cost

For processing  250 tons/day of  dry sludge solids.


                                    days/yr '  «2.39/t«
                                    167

-------
VACUUM FILTER PLUS BELT PRESS

Number of Units

Vacuum filters - 12 units H- 1 spare = 13 units
Full-scale yield for belt press = 1181 Ib/hr/m width (Parkson tests)
Belt width - 2 m/unit

     -T  ,     f   ._              500,000 Ib/day
     Number of units =         .,—;	' 0. ,   ,, f	z—-.	—
                       1181 Ib/hr/m x 24 hr/day x 2m/unit
                     = 8.8 units
                     or 9 units + 1 spare = 10 units

Capital Costs

1.  Vacuum Filters—

          Total installed cost = $8,700,000

2.  Belt Presses—from Parkson, the cost per unit is $72,000

          Total installed cost = $72,000/unit x 10 units x 3
                               = $2,160,000

3.  Distribution and Feeding System—

          Total installed cost = $1,000,000

4.  Additional Conveyors—

          Total installed cost = $500,000

5.  Total Capital Costs—

          Vacuum Filters                  $8,700,000
          Belt Presses                     2,160,000
          Distribution/Feed System         1,000,000
          Additional Conveyor                500,000
                                 Total   $12,400,000

Annual Costs

1.  Amortization—

          Amortized cost = $12,400,000 x .09 = $1,116,000

2.  Chemicals—same as costs for vacuum filters.

          Chemical cost = $1,663,400
                                    168

-------
 3.   Power—
           Vacuum Filter System                       1345
           Belt  Presses  - 12.5  Hp/unit  x  9  units        112.5
           Distribution/Feed  System                     50
           Additional  Conveyors                         15
                                           Total       1523.5  Hp

          Power  cost  =  1523.5 Hp x  .746 Kw/Hp x  $.04/Kw-hr
                                     x 8760 hr/yr

                      =  $398,240

4.  Water—the vacuum filter system will consume 60 gpm/unit; the
    belt press system will consume 50 gpm/unit.

          Water  consumption = 60 gpm/unit x 12 units
                                    + 50 gpm/unit x 9 units
                            = 1170 gpm

          Water  cost  = 1170 gal/min x 525,600 min/yr x $.25/1000 gal
                      = $153,700

5.  Operating Labor - costs assume a 7-man crew will operate the vacuum
    filter system and a 3-man crew will operate the belt press system.
    For a 7 day/week operation,  40 man-years will be expended.

          Labor cost = 40 man-years x $21,000/man-year
                     = $840,000

6.  Maintenance—maintenance  costs consist  of  normal maintenance costs
    on both the vacuum filters and the  belt presses and  belt  replacement
    costs on the belt  press.  Vacuum filter maintenance  costs will  total
    $87,150  per year (see vacuum filter  costs).   Belt  press maintenance
    costs are based on 3% of  the purchase price of  the additional
    equipment associated with the  belt  press.  Belt replacement  costs
    will total $20,000 per year.

          Vacuum filter  maintenance  costs = $87,150

          Belt press maintenance costs = $3,660,000/3  x  .03
                                      = $36,600

         Belt replacement  costs = $20,000

         Total maintenance costs = $143,750
                                   169

-------
7.  Total Annual Costs—
          Amortization                    $1,116,000
          Chemicals                        1,663,400
          Power                              398,240
          Water                              153,700
          Operating Labor                    840,000
          Maintenance                        143,750
                                          $4,315,090
Unit Cost
For processing 250 tons/day of sludge solids.

          Unit cost	$4,315.090	   .  $47.29/ton
          Unit cost - 25Q tons/day x 365 days/yr
 INCINERATION
    The  costs  in  this  section are rough  approximations developed  from
 on-going design work for  the District  of Columbia.

 Number of Units
                                          2
 Incinerator  rating  = 10 Ib wet  feed/hr/ft  of  burning area
 Burning  area for  a  12  hearth unit =  4584 ft2/unit  (25.75  ft  diameter)
 Feed  capacity  = 45,480 Ib wet feed./hr
 Feed  rate =  317.5 tons/day of dry solids (250  tons/day of dry sludge
                                              solids  + 27% chemicals)
 Availability factor =  85%

 For a 20%  feed
                                317.5 tons/day x 2000 Ibs/ton	
      Number  of units  = 45<48o lb wet feed x .2 Ib  dry  feed  x 24_hr x 0.85
                           hr/unit              lb  wet  feed     day
                      = 3.4 or 4 units

 Similarly,  for a 35%  feed

      Number of units  = 1.9  or  2 units

 Capital Costs

 Includes air pollution control equipment (electrostatic precipitator)
 to meet emission requirements,  installation, building,  utilities, and
 engineering.

           'Total  installed cost = $5,000,000/unit
                                      170

-------
 Annual Costs

 1.  Amortization—

      For a 20% feed, amortized cost = $20,000,000 x .09 = $1,800,000

      For a 35% feed, amortized cost = $900,000

 2.  Power—

      For a 20% feed, power usage is 860 Hp/unit

           Power cost = 860 Hp/unit x 4 units x .746Kw/Hp x 8760 hr/yr
                                 x $ .04/Kw-hr
                      = $899,200

      For a 35% feed, power usage is 775 Hp/unit

           Power cost = $405,200

 3.   Fuel—the  incinerator  will produce an 800 °F  outlet temperature.
     A fume furnace will raise all stack gases to  1350 °F before discharge.
     Water vapor will be removed in a subcooler, prior to reheating  the
     stack gases.  With a 20%  solids feed,  21,000  gal/day of  #2  fuel
     oil  will be required for  incineration,  and 7,133 gal/day of #2  fuel
     oil  will be required for  the  fume  furnace,  for a total fuel usage of
     28,133  gal/day.   With  a 35% solids feed,  4306 gal/day of #2 fuel
     oil  will be required for  the  fume  furnace only.

      For  a  20%  feed,  fuel  cost  =  28,133  gal/day x 365 days/yr x $ .40/gal
                                =  $4,110,000

      For  a  35%  feed,  fuel  cost  =  $630,000

4.   Operating Labor—costs assume  each crew consists of  1 supervisor, 1
     operator per unit and  1 helper.  Four crews will be  required to cover
     a 7 day/week operation.  For a 20% feed, 24 man-years will be expended;
     for a 35% feed, 16 man-years will be expended.

     For a 20% feed,  labor cost = 24 man-years x $21,000/man-year
                                = $504,000

     For a 35% feed,  labor cost = $336,000

5.   Maintenance—costs assume $100,000/unit per year.

     For a 20%  feed,  maintenance cost = $100,000/unit x  4 units
                                      = $400,000

     For a 35%  feed,  maintenance cost = $200,000
                                    171

-------
    Ash Disposal—hauling  and  disposal  costs will  total  $10.00/ton of
    ash and  are computed based on  40% of  the incoming  feed  to  the
    ciewatering process  plus  100% of  the inert  chemicals  added.

          Total ash quantity = 167.5 tons/day

          Ash disposal  costs = 167.5 ton/day x 365 days/yr  x $10.00/ton
                            = $610,000
    Total Annual Costs—

          Amortization
          Power
          Fuel
          Operating Labor
          Maintenance
          Ash Disposal
                            Total
      20% Feed

     $1,800,000
        899,200
      4,110,000
        504,000
        400,000
        610,000
     $8,323,200
  35% Feed

 $ 900,000
   405,200
   630,000
   336,000
   200,000
   610.000
$3,081,200
Unit Cost
     For processing 250 tons/day of sludge solids.

           nn« *  ,    •              $8,323,200/yr
     For a 20% feed, unit cost =	—	*	LJ!—
                                 250 tons/day x 365 days/yr
                           = !?91.21/ton
     For a 35% feed, unit cost = $33.77/ton

HAULING

     Hauling costs are based on actual costs now incurred at Blue Plains
to haul sludge cake a 25 mile distance.  Undigested vacuum-filter cake
must be transported in enclosed vehicles, e;g., a concrete mixer; hence
costs are $9.40 per wet ton.  Filter-press cake is assumed to be dry
enough to carry in an open dump truck; hence costs of $6.25 per wet ton.

     The costs per dry ton of sludge solids were developed by correcting
the above figures for the percent cake solids and quantities of chemicals
added.  For example, the cost of hauling vacuum filter cake at 20% solids
is calculated as
          Hauling cost
                         Q>2
	   1.27  tons  total  solids
 ton/wet  ton X  1.0  ton  sludge solids
                           $9.40/wet ton
                       =  $59.69/dry ton of sludge solids
                                     172

-------
COMPOSTING
    ^Costs were obtained from a paper entitled "Composting Filter Press
Cake"; presented at Compost Science Meeting, April, 1978 at Omaha,
Nebraska;  G. Wilson, D. Colacicco, and D. Casey, USDA, Beltsville,
Maryland.  Costs in this paper are presented in $/wet ton of sludge as
received.  To convert to $/dry ton of sludge solids, use the procedure
as described under hauling costs.
                                   173

-------
                                            APPENDIX F
                                  FULL-SCALE UNIT SPECIFICATIONS
                              TABLE F-l.   FILTER PRESS SPECIFICATIONS

2
Filtration area, m
No. of Chambers
Plate dimensions, m x m
Chamber thickness, mm
Yield, lb/hr/ft2
Avg cycle time, min
No. of units for 250 TPD*
Budget purchase price, $1000/unit
Power usage, Kw-hr/ton
NGK
500
130
1.5x1,5
30
0.49
54
9
722
68.2
LASTA
204
32
2x2
25
0.60
^v40
17
775
62.5
PAS SAVANT
1080
150
2x2
40
0.31
210
7
715
57.3
NICHOLS
628
115
2x1.5
32
Q-22
180
15
400
88.9
*Includes 1 spare

-------
TABLE F-2.   FILTER MEDIA SPECIFICATIONS

TYPE
NGK NY516
NGK TR520
NGK NY51-4
LASTA P920
LASTA P891
LASTA P940
PAS SAVANT T167
NICHOLS 4709/40

CONSTRUCTION
plain
herringbone twill
twill
2x2 twill
2x2 twill
2x2 twill
twill
2x2 twill

MATERIAL
polyaraide/polypropylene
polyester/polyester
polyamide/polyester
polypropylene
polypropylene
polypropylene
nylon
polypropylene
AIR PERMEABILITY
at A p = 12.7 mm HO
cm3/sec/cm2
4.0
11.0
93.0
13.3
25.0
40.0
76.7
20.3

-------
                             GLOSSARY OF TERMS
Sludge Solids:                -         Sewage sludge solids only.
Total Solids:                          Sewage sludge plus chemical solids.
Process Yield:                         Calculated as kilograms of sludge
                                         solids per hour of filtration time
                                         per square meter of filtration area.
Full-scale Yield:                      Same calculation as process yield
                                         except that cycle time includes both
                                         filtration and mechanical cycle time
                                         for a full-scale press.
                                      176

-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-79-123
2.
4. TITLE AND SUBTITLE
EVALUATION OF DEWATERING DEVICES FOR PRODL
SOLIDS SLUDGE CAKE
7. AUTHOR(S)
Alan F. Cassel and Berinda
P. Johnson
9. PERFORMING ORGANIZATION NAME AND ADDRESS
District of Columbia Government
Department of Environmental Services
Water Resources Management Administration
5000 Overlook Avenue, Washington, D. C. 2
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laborator
Office of Research and Development
U. S. Environmental Protection Agency
Cincinnati, Ohio 45268
15. SUPPLEMENTARY NOTES
Project Officer: Roland V
3. RECIPIENT'S ACCESSION- NO.
5. REPORT DATE
August 1979
ip-rvrp HTPH
(Issuing Date)
UJ.INU nxun- 6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1BC821, SOS #1, Task A38
11. CONTRACT/GRANT NO.
68-03-2455
0032
13. TYPE OF REPORT AND PERIOD COVERED
v Research StuHy
14. SPONSORING AGENCY CODE
EPA/600/14
. Villiers (513) 684-7664
16. ABSTRACT
Pilot-scale dewatering tests were made to establish design and operating
parameters for dewatering municipal wastewater sludges on recessed plate filter
presses (both diaphragm and fixed volume types), continuous belt presses, and
retrofit units for a vacuum filter. Results from the 1.5-year study showed that
when dewatering lime and ferric chloride- conditioned sludges, the recessed plate
presses consistently produced a 30-40% solids filter cake. Feed solids to the
units averaged 5% total solids with a range from 2.4 to 10%. Various ratios of
waste-activated to primary sludge solids, with emphasis on the 2/1 ratio, were
tested. Belt presses produced cake solids from 25-30% when the polymer condi-
tioning dosage was optimized. When used as a retrofit device to a vacuum filter, •
the^belt press gave cake solids in the 30-40% range during laboratory-scale tests.
Design parameters are developed to dewater a mixture of 67% secondary and 33%
primary sludge in a full-scale plant installation. The estimated costs for
dewatering plus final disposal by either incineration or composting are also
presented.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Sludge
Dewatering
Sludge disposal
Waste treatment
Economic analysis
Cost estimates
13. DISTRIBUTION STATEMENT
Release to public
EPA Form 2220-1 (9-73)
b.lDENTIFIERS/OPEN ENDED TERMS
Sludge processing
Pilot study
Performance data
Design guidelines
Sludge conditioning
Sludge dewatering
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page)
c. COS AT I Field/Group
13B
21. NO. OF PAGES
191
22. PRICE
177

-------