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
------- |