DEWATERING OF DOMESTIC WASTE
SLUDGES BY CENTRIFUGATION
STERLING EUGENE SCHULTZ
A DISSE1TATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVRsSITY OF PLORIDA
IN PARTIAL PULRILMENT OF THE REQUUIRMENTS FORl THE
DEGREB OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
SCHULTZ, Sterling Eugene, 1934-
DE WATERING OF DO1VIESTIC WASTE
SLUDGES BY CENTRIFUGATION.
The University of Florida, Ph.D., 1967
Engineering, sanitary and municipal
University Microfilms, Inc., Ann Arbor, Michigan
I would like to express myr sincere appreciation to Dr. H. D.
Putnam, my committees chairman, who provided the guidance, assistance,
and inspiration which were required to complete tlria dissertation. The
support and contribution made by the other members of lag committee,
namely, Dr. C. I. Harding, co-chairmanl, Dr. A. P. Black, and Professor
T. deS. librman, are also gratefully acknowledged.
I am. indebted to H. L. Czewn, vbo assisted me in carrying out this
work. Appaniation is extended to Dr. W. Hi. Morgan, Dr. W. J. Npolan,
J. Bramble, F. Jannaha( Sara Kephart, and other for rendering valuable
assistance when called upon for help. Itr thanks (ao Mrs. J. G. Larson for
tgping the final manuscript.
Sincere appreciation th extended to my wife, Shirlee; for her help
and encouragement throughout the entire project.
Recognition and appreciation are extended to Bird Maohine Company,
South Walpole, Mass., for providing the 6 x 12 inah solid bowl centritage
used in this work; Calgon Corporation. Pitteburgh, Pa.r for supplying the
CAT-FLO)C oatilonio polymer; Dow Chemical Company, Midland, Mi~ch., for
supplylag several cationio and anionic polymers; Rham anrd Base, Philadelphia,
Pa., for supplying Primafloo C-7 oationic polymer s nd the Department of
Environmental EPngineering, University of Florida, for financial support
in the course of the project for equipment and material.
To the United States Air Ikree Acadeqr, Department of Civil
Engineering, for the leave of absence; the Air Fbree Institute of Tech-
nology for support; and Bleadquarters U.S.A.F. for the active duty assign-
ment to the University of Florida for this advanced degree, sincere
appreciation is extended.
lACKNONLEDGMENOTS .r ...... .....
L~IST OFTABLES ....... ....... .......
LfST OF FIOURESS.. ..... .... .... ....
I. INTBaURODCION . ........
TII. HI~STORICAL DEVE10IPMexT . . . . . . . *
First Centritagal Deuatnern in Gerumqa .. .
Centrlifugal Dewatering in the Ulnited Sktate *
III. THEORETICAL CONSIDERATI;ONS OF CENYTRIF1UGA SEPARATION,
SUSlDGE CHARACTERSTICS, AND CREMCICL FLOCCO1ATION .
Centritagal Solid-IJquid Separation by Density
DIfferenc . .. .. .. ..
Derivrtlon oIfTheoretioal Relationhip .. .
Application of the Sigma Concept .. .. ..
Sedimentation Performnano Carve . .. .. .
Separation in the Contianrous Solid Bowl
Centrjitag . . .
Studge Charracteritics .. ..
Colloidal Properties . .
H bdrophilic Colloida . ..
Hlydrophobio Colloida . .
Zsta Potential . .. .
Streaming Current ... .. .
Colloid Stabillity . .. .
Chemical Fplocalation . .. .
IV. EXPEanRIMETA EQUIPamEN MA'TERIALS,
Equipment . . .. . ..
BI~rd Solid Bowl Centritage .
Pilot Plant Equipment ... ... . . .. 38
Equipment Used at Treasure Island, Florida .. 43
Equipment Used at St. Peteraburg, Plorida . .. 44
Katerala . .. .. .. . ... 45
Domestic Waste Sladges .. .. .. ... .. 45
University of Florida Sewage Treatmt~ent Plant . 46
Water Pollation Control Plant, Treasure
Island, Florida . ... ... . .. 46
Northeast Treatment Plant, St. Petersburg,
Florida . . .. .. .. 47'
Po'lyelectroly~tes . .. .. 50
Procedures . . . . . 52
Sledge Preparation .. .. .. .. . 52
Determinations on Feed Sludge, Centrate, and Cake 53
Laboratory Devatering Testa .. . ** **
Experimntal esin ad Evalutio . .. 5
ExperimPental Procedures. .. .. .. . .
Data Recording and Calculations . .. .. 64
V. RESULTS ANlD DISCUSSION .. .. ... .. . ... 66
Introduction .. .. .. .. . .. .. 66
Dewatering Digested Sladge . ... . .. .. 67
Experiment 1 . .. . . ,.. .. 67
Experiment 2 .. .. .. .. .. . .. 71
Experiment 3 .................. 72
Experiment 4 ........********** 75
Experiment 5 ** *. .. . .. . 77
ExJperiment 6 . . .. . .. .. .. 79
Exjperlaenta 7 and 8 . . . . . . . 8
Experimenta 9 and 10 . .. . ... ... 84
Expeime$11.... ....... ... 88
Dewatering Activated Sludge ** ** 90
Experiment12 ...........,****** 90
Experimentl13 .........********* 94
Experimentl14 ..............*** 98
Experiment 15 .. .* ** *. ** 98
ExpeimenI~t l? ............ .... 101
Experimeont 18 .. .. . .. .. .. .10
Experiment l9. .. .. .. .. .. .. .. 105
ExJperiment 20 . .. .. .. .. ... .. 108
Experiment 21 .. .. ... .. .. ... 108
Dewatering Raw Sladge . .. ... .. . .. 111
Experiment 22 . .. . . . . . . . 111
Experiment 23 . . . . . . . . . 117
Disonasion ... ... .. .. . .. ... . 117
M~obine and Praoess Variables . .. . .. .. 117
Sedimentation Performance Comparisons of
Centritages . 119
Comparison of 6x12 ad 24:38 nob Centrirgge. 122
Comparison of 6xl2 and 24x60 Inch Cntri~tage .. 125
Characteristics of Sladges Dewtateed . .. .. 128
Polymer)~( I Dosge. . .. .. .. 130
Centrate Mobility, SCD Readings, and Residrul
Polyser ................... 132
VI. CONCLUSIONS ..... ............ 131)
AFbENDICES .... .... .... 139
1. TABLES 13 THROUGH 34s TABUEATED RESUETS OF ALL
ExERME. .................... 140
2. GEDISSARY . r . ....... 205
LIST OF REFERENCES .. .. .. .. .. .. .... .. 208
BIIOGRALPHICAL SKESTCH . . . ... . .... .. . . 215
LI;ST OF TABLES
1. Performance of Centriauges Devatoring Domaestic Waste
2.. PolyrelectroL~ytes Used in Laboratory Teste .. . .. 51
3. Data Sheet .............,........... 62
4. Calculation Sheet ,. . .. .. ,, ,,. .. 65
5, su1mary of Digested shadge as~periments . . .. . 68
6, Performnane of 24 x 60 Inch Centrifuge Deratering
Digested Sladge, St. Peteraburg, Fla. ... .. . ... qlr
7. Su~a~ry of Acotivated Sludge Experimrents ... .. .. 91
8. Sramaury of Raw Sludge Experiments .. .. .. . .. .
9. Effoot of Antonio and Cati~onio Polyaers on1 DeIwateing
Raw Sludge at the University of Florida Sewage Treatment
10. Sigma V~alues and Volumes for Three Centriflages
Used ...*********************** 121
11. Sigma Values in Square Inches for 6x~12 Inch
Centr~ithge ....... ..************* 122
12. Types of Centrithrges and Their Nanufacturers
in the U.S ..******************** 133
13. Effect of Feed Rate, Polymer Dose, and Location
of Polymer ALddition on Dewatering A~naerobically
Digested Sludge at the University of Florida
Sewage Treatment Plant *... *,.. * ***** 141);
14C. Btfoot of Polymer Dosage and location of Polymer
Addition on Recovery, Cake Solida, and Centrate
Electrophoretio Mobility libn Dewatering ALnaerobically
Digested Sludge at the University of Florida Serage
Treatmelt Plant .. .. 14
15r. Ettect of Gentritugal Force, Polymer Doeage, and
ovationn of Polymer Aiddition on Recovery, Cake
Solids, and Centrate Electrophoretio Mobility
~hen Dewatering Ainaerobically Digested Sludge
at the University of Florida Sewage Treatment
Plant ............ ............14
16. Effect of Three Cationio Polysers an Devatering
Anaerobically Digested 81ndge henl Doeages of
Equal Cost Used to Dewater Ansaerobically
Digested Sludge at the University of Florida
Sewage Treatment Plant ... .. . .. ... 148
17. Effect of Centrifugal Forcer Polymer Dosage,
and location of Polymer Addit~ion on Recovery,
Cakre Solids, and Centrate ELeotrophoretic
NAbility and Streaming Current %hen Dewatering
Amerobioally Digested Sludge at the University
of Florida Sewage Treatment Plant *.. * *. 5e
18. Effect of Centrifklgal Force, Pool Depth, and
Sludge Feed Rate on Deuatering Arnaerobically
Digested 81udge at the University of Florida
Sewage Treatment Plant . .... .. .. ... 153
19. Brfoot of Sludge Concentration, Centriungal
Force, Polymer Dosage, and Deep Pool Depth
on Dewatering AnsaerobioalLy Digested Sludge
at the University of Florida 9swage Treatment
Plant ..................... 157
20. Beet of Sludge Concentration, Centrithgal
Force, Polyser Dosage, and Sballow Pool Depth
an Dewatering Anaerobically Digested Sludge
at the University of Florida Sewage Treatment
Plant......................... 1 60
21. Effect of Pool Depth, Polymer Dosage, and
Location of Polymer ALddi~tion on Dewatering
Anarerabtc~ally Digested Sludge at the Water
Foluntion Control Plant. Treasure Island,
Florida .................... 16)
22. Bifoot of Feed Rate, Polyser Dose, and location
of Polyser A~ddition on Devatering Anaerobioally
Digested Sladge at the Water Pollntion Control
Plant, Treasure ILsland, Florida . . .. * * .
23. Effect of Feed REate, Polymer Dosage, and
Location of Polymer Addition on Deuatering
Return ALctivated SLudge at the University of
Florida Sewage Treatment Plant . .. . . 169
24~. Etreet of Centria~gal Force, Feed Rate, and
Polymer Dosage on Reaovery, Cake Solida, and
Cenrtrate Electrophoretio Mboblity and Streasing
Current When Devatering Return Activated Sludge
at the Unihversity of Florida Sewage Treatment
Plant .............. ......... Ilt2
25. Effect of Centrithsal Force. Feed Bate, and
Polymer Dosage on DewateriLng Return Sludge
from Extended ALeration Unit at the University
of Florida Senage Treatment Plant .. ... .. 175i
26. Effoot of Centrith~gal Force, Feed Rate, Pool Depth
and Polymer Dosage on Dewatering Betr Sludge
at the Water Pollution Control Plant. Treasure
Island, Florida . . .. .. .. . . 178
27. Effect of Deep and Nedium Pool Depths, Feed
Bate, Polymer Dosage, and Polymer Doeing
Bate on Devatering Beturn 81udge at the Water
Pollution Control Plant. Treasure Island, Florida 162
28. Effoot of Centrithgal Force, Feed Rate, and
Polymer Dosage on Devatering Waste Sludge at the
Northeast Sewage Treatment Plant, St. Peteraburg,
Florida ....................... 183
29. Effoot of Polymer Dosage, Polymer Desing Rate,
and Location of Polymer Addition on Dewatering
Waste Sludge at the Northeast Sewage Treatment
Plant, St. Peteraburg, Florida. .. .. . .. . 188
30. Performance of 24 x 60 Inch Centrithge Dewate>
ing Waste 81udge. St. Peteraburg, Florida *.. * 192
31. Effoot of Polymer Dosage and Polyser Dosing
Rate o~n Dewatering.~e~turn Sludge at the Northeast
Sewage Treatment Plant, St. Peteraburg, FLorida ,. 194
32. Effect of Feed Rater Pool Depth, Polyser Dosage,
and2 Polgser Doeing Rate on Performance of
2k x 60 Inch Cenrtrithge Dewatering Return
Sludge, St. Pet~erabrg, Florida .......... 196
33. Effect of Centrithegal Force, Polgser Dosage,
and Locati~on of Polgmer Addition on Dewatering
Raw Sladge at the University of Florida Sewage
Treatment Plant ........ ..... ...... 200
34. Effect of Feed Rate, Polfser Dosage, and Location
Polyser Added on Dewraterng Raw Sludge at the
Water Pollution Control Plant, Treasure Island,
Florida .................. .... .202
LIST OF FIGURES
1. Past and Projected Trends in Domaestio Sludge Production
From Primary andl Combined Secondary Treatment.. . .. 3
2. Dewatered Slaclge VoloPne as Per 4ont of Original Shadge
Volurme at a Given Suspended Solids Conoontration .. 3
3. Cross Seation of a Cylindrical Centritage Bowl Showing
the Separating Mechanian .. . ... .. .... 16
4, Cross~eotionail View of Ogpical Cylindrioal-'4pe
Solid Bowrl Centrifuge . . . ... . .. . 21
5. schematic sketch of a Collo 4 Protein Particle Enoased
in Bound Water (From Rioh3 p. 1361) . ... ... 26
6, Schematia Sketch of the Double Layer Around a Colloid
Particle With a Negative Primary Charge. (From Rich, p.136) 28
7. Potentials Through the Eaeatrostatio Field Surrounding a
Charged H~ydrophilio Colloid Particle. (Froml Rich, p. 131)' 28
8. Bird Solid Bowl Continuous Centrituge MLith Countercurrent
Flow. Courtesy of Bird Machine Company, South Wallpole,
9. Bird Solid Bowl Continuous Centritage With Conourrent
Flow. Coulrtesy of Bird Maobine Company, South WaJlpole,
Mass.. .. .. .. 3
10. Schematic Flow Diagram of Sladge Dewatering Pilot
Plant, adversity~ of Florida *... * * *** 39
11. Bird 6x~12 Inch Solid Bowl Continuous Centrifuge.
Courtesy of Bird Machine Company, South Wallpole, Mass. 41
12. Flowakeet of the Modified Contaot Stabiisation Prooeas
at the Water Pollation Control Plant, Treasure Island,
Florida ........,,....,,......... 48
13. Sohematic Flow Diagrea of the NJortheast Sewage Treat..
ment Plant, St. Petereburg, Fla. * ** ** 49
14, Beourring Molecular Unit of lonis~ed CAT-FLO)C * *.. 2
15. Effect of Sladye Feed Rate, Polymesr Dosage, and
location Polgmer Added on Performance of 6x12 Inch
Centrifuge Deuatering Digested Sladge, Uidversity of
Florida............... .......... 70
16. Effect of Polymer Dosape and location Polgmer Added on
Performanoe of 6x12 Inch Centri~tup Dewatering Digested
Sludge, UI~versitg of Florida . . .. .. . . .. 73
17. Effect of Centritugal Force and Polgomer Dosage on Per-.
formoance of 6xi12 Inch Centrifuge Deartering Diseted
Sludge, Mydersityr of Florida .. ... .. . .
18. Effect of Three Cationio Polgmers and Laocation of Polgmer
Addition on the Performane of 6x1i2 Inch Centrilage
Dewatering Digested Sludge, Ohversity of Florida . .. 76
19. Effect of Centrigl~ Force and Polymer Dosage on Eleo-
trophoretio Nobility and stneming Current Response of
Particles in Centrate blben Doma~tering Dipated Sludge,
Myersit of Florida .. ...... ..... ..... 78
20. Ettect of Centritagal Force, Feed Rate, and Fool Depth
on Pertomaunce of 6x12 Inch Centritape Dearterin
Dipsted Sladge, flbversity of Florida .. . ... 81
21. Ettact of 81udye Conoontration, Centritugl Force, and
Polymer Dosap on Perforannoe of 6osl2 fInh Centrifuge
Demtering Digested Sludge, Qulvoesity of Florida . . 82
22. Effect of 81udye Conoontration, Centritugal Foro*, and
Polgmer Dosage on Performance of 6xi12 Inoh Centritage
Devaterlas Dgested Sltdge, Uhversity of Flodrid . .
23~. Effoot of Feed Rate, Polymer Dosage, and Laocation Polglbsr
Added on Performance of 6612 Inch Centrifuge Dewatering
DIgested Shidge, Treasure Isaond, Fla. * . . 85i
24. Effoot of Feed Rate, Polgyser Dosage, sad location Polgser
Added on Performarce of 24 x 38 Inch Centria~ge Dewatering
Digested Sludge, Treasure Island, Fla. .. .. .. .. 87
25. Effoot of Feed Rate, Polymer Dosage, and Location Polyser
Added on Perfomncure of 6x12 Inch Centritupe Detatering
Return Activated Sludge, Eversity of Florida . * 93
26, Ettect of Centritugal Force, Feed Bat*, and Polymer Dosage
on Performanoe of 6xr12 Inch Centriftip Activated Siuage,
Overadty of Florida .. 96
27. Effect of Centrifurgal Force, Feed Bate, and Po3Jaer
Dosage on Performance of 6x12 Inch Centlituge Deuatering
Return Activated Sludge, Universityr of Florida . .. .. 97
28. Effoot of Centritugal Fore, Feed Bate, and Polgser
Dosage on Performanoe of 6x1L2 Inch Centritupe Sladpe
From Extended aeration Ihift, Universitg of Florida . . 99!
29. Effoot of Centritagal Force, Feed Rate, and Polgser
Dosage on Performane of 6d12 Inch CentriRtup Return
Sludge, Treasure Island, F a. . ... . .. ... 100
30. Effoot of Feed Rate and Polyser Doeape on Performanoe
of 24 x 38 Inch CentritaPg Retara Sludge, Trreasure
Island, la. ...,................ ..., 102
31. Effoot of Centritugal Force, Sludge Feed Rate, and
Polyser Dosage on Performance of 6:12 Inah Centritage
Dewatering Wate 81adge, St. Peteraburg, Fla. .. .. 104)
32. Effoot of Polymrer Doasap, Po2;aer Desing Bate, and
~Loocation of Polgmer Aidditioan on Performance of 6a12 Inch
Centrituge Dewatering Hate Sludpe, St. Petrsburg, Fla. 106
33. Performance of 24 x 60 Inch Centritarge Denctering
Wate 1Sldge, St. Petoerbnrg, Ph. . .. .. .. 107
34. Effoot of Polgnmer Dosage and Polpmer Desing Bate on
Performance of 24 x 60 Inch Centrifuge Dowatering Return
ActiLvated Sladge, St. Peteraburg, Fla. .. .. ... 109
35. arrect of Polaer ALddi~tion Bate, studge Feed Bate, and
Polgmer Dosa~ge on Performanoe of 24 x 60 Inch Centritupe
Dewatering Retura Activated 1Sldge, St. Peteraburg, Fla. 110
36. Effect of Sludgeo Feed Rate and Polgmer Dousag on Per-
formane of 6x12 Inch Centrifuge Dewatering Res Sludg**
University of Florida . . . . . . . . 113
37. Effoot of Feed Rate and Polyser Dosage on Performanoe of
24 x 38 Inch Cenrtritupe Dewatering Raw 81udpe, Treoasure
land, Fla.................. 134.l
38. Sedimentation Performanoe Versos Parameter fior Bird
Solid Boul Centrithges Dewatoring Digested Studge . 120
39* Sedimentation Performance Versua Parameter for Bird
Solid Bowfl Centrithges Dewatering Return Ac~tivated
Sludger Treasure) Island, FZa, . . . . . . 124
0. Sedimentation Performance VersansParaneter fCor Bird
Centrithg~e Deunatering Waste Slodge, St. Petersabrg,
Fla. .. .. .. .. 126
41. Sedimrentation Performane Versusa Parameter .for Bird
Solid Bounl Centrithge Devatering Maste Sludge(
St. Peteraburrg, Fla. .. .. . . . . .. 127
Abstract of Dissrtation Presented to the Iralduate Counnil in
Partial Fafinment of the Requremesnts for the Degree of
Doctor of Philosophy
DWAPTERINGO OF DOMEgSTIC WASTE SLUDGES Ea CENTRIFUGATION
Sterling Engene Schults
Chairmarnt Dr. Hugh D. Paltnam
Major Departments Environmental Engineering
This research determined the effort of machine and process vari-
ables on dewatering domestic waste eludges by means of solid bowl con-
tinuous centritages. The machine variables were centrifugal force and
pool depth. The process variables were sludge feed rate, slodges concen-
tration, type of shudge, type of polymer, polymer dosage, polyser dosing
rate and location of polymer addition. Factorial experiments to test
the main efforts and interactions of these vrariables were randomised
block, split-plot and split-aplit-plot design. Two replications were
made on each treatment combination. The analysis of variance for each
experiment was calculated by the IBM 360 computer at the University of
Florida Computing Center. The AOlV prograM is available at the Center's
The work was originally conducted on a pilot plant scale at the
University of Florida Sewage Treatment Plant, using a 6x12 inch solid
bowl centrifuge. Later in the insestigation, the 6x12 sentritate was
taken to treatment plants having 24x38 inch and 24x60 inch machinesr
Since maxmma clarification and miniman oake moistare cannot be
obtained siamltaneonesly, the use of polyelectrolytes to obtain satia.r
factory clarity and oake solids dryness was an imaportant part of this
Experiments were conducted on the same sludge with the three
maachines. The prupose was to defend the pilot plant work as valid bg
comparison to results obtained with fuill-soale equipment and to compare
the performance of the three machines with the intent of using the
results for anale-up work.
The experiments together with their wnalyses, of variance deter-
mined that the anchine and prooeas variables are highlyr interdependent.
Most of the two-factor interactions can bb readily explained bp graphi-
eal presentation of the data; therefore, the main effects of the vari*
ables can be disonlseed.
As centrifugal force increased, recovery of stlapended solid in-
oreased for digested and raw alrrdges, and the concentration of the cake
solids increased for raw, atiatetd, and digested stadges. An inter-
nsdiate centritagal force increased recoveryI of suspended solids for
activated shtdge. As ppol depth was increased, the reovery of suspended
solids increased for all shdgdes and the cake solid concentration de*
oreased for all shadges. As the slardge feed rate increased, recovery of
suspended solids decreased exponentially for all aludges, and oake solids
concentration increased for raw and digested sladges. Cake solid con-
centration for activated slwdge dooreased when the feed rate increased.
As feet alwdge concentration increased, the recoveryr of easpended solids
decreased and oake solids concentration increased for all sladges.
Sedimentation performanoe, based on the eigma concept, was chosen
as the basia of comparison between the 6 inch and the 24 inch centritages.
This teoohque, although surbject to some criticism, provides a common
denoanantor for comparing the centritages, and it is arpeed that it is
the anct usetal method.
'Whenever comparative tests based on the unoorreoted sigma tantor
could be made, the results showed that the 6x12 inch centritage duplicated
or exceeded the perforancne of the larger machines deatekring digested
and activated shrdges. The aludte dewatering characteristics exhildted
when centrifuged with the smaller centrifuge were very aimilar to those
obtained ~with the larger machines. Hence, results of the pilot plant
experiments can be used to predict readlta with fuill-scale equipment in
the centrifugal force range or 650 to 3,ooo tiess gravitjr.
The research revealed the need for productive Investigation con,
corning the efficiency factors for seale-urp from small to large machines,
since for the comparisons made there was a loss of efficiency as the
machine sise increased.
The centrifuge has intrigued sanitary engineers for many years as
a method for dewatering domestic waste shudges. This interest is a
result of successful centrifugal solid-liquid separations in the chemical
process industries for more than a generation. Dewatering of sewage
shudges was first tried in this country in 1920. The results were on~-
couraging enough that additional experimental installations continued
throughout the twenties and thirties. Nowveer, the cenltrifuge was not
accepted in sewage treatment because of technological problems and
Failure to compete economically with sludge dewatering on drying beds.
The interest in mechanical shudge thickening and deva~teing was
intensified after World War II as the volume of solid to be disposed of
increased exponentially as the population and soonomy expanded. Sludge
disposal quickly became the most troublesome phase of sewage and indus-
trial waste treatment, and today is still the most ditticult problem in
wastewater treatment.1 America' s continued population and industrial
expansion, as well as the national goal for clean rivers and streams,
has amplified Yhe perplexing problem of sludge treatment and disposal.
With an urban population of well over 200 million forecast by
1980, and a national goal of secondary treatment for all annicipal waste-
water, the daily sludge volume will rise to mDor than 150 mIllion gallons
per day.2! This represents a substantial increase whn compared to the
1962 sludge volume of So million gallons per day.3
The past and future trends in domestic sludge production are shown
in Figure 1 where it may be seen that as waste treatment improves with
more advanced methods, a higher fraction ot the soluble impurities is
converted to more voluminous aludges. Because the specitio gravity of
sewage sludge solids is close to that of water, a signifi~cant sludge
olurme rednotion accompanies the lose of water and the sludge volume
changes approximately in the ratio of its solids concentration. Sewage
sludges with a solid concentration in the range of one-half to eight
per cent, undergo a remarkable volume reduction upon dewatering as illue-
trated in Pigure 2. The potential value of thickening voluminous eludges
prior to Labsequent treaktent and of dewatering sludges prior to ultimate
disposal is apparent. Sludge thickening may be defined as the rednotion
in moisture content of the sludge in order to significantly decrease
sludge volume while still maintaining its fluid properties. This defini-
tion excludes sludge dewatering where the purpose is to concentrate the
suspended solid into a relatively dry sludge cake prior to ultimate
The methods for sludge thickening include: gravityr thickenera45
for concentrating primary and secondary shudges or their mixturesr elu-
triation' for thickening digested aludge prior to vacuum filtrations
dissolved-air flotation, for thickening secondary shudges; and centrita-
gation for thickening excess sludge wasted int the operation of activated
The methods for sludge devratering inalade 5 drying bedsl5 and slunge
- 3 -
B I SECONDARY EAEN
B O PRIMARY TREATMENT
1900 1980 1940 1980 1980 2000
Pig. 1 Past and Pagjeoted Trends in Domesktoe Sludge' Pmoducaton
Faom Primary and Combined Secondary Treatment
SO o I 4 I 89
Soid Cnonraho i ewteedSldePe 4n
40. erRtr Sug ou a e e o rgn lde
dolm 2taQvnSseddSld o~nrt~
lagoonal6 for digested aludge; vaonum 11teral? for- raw and digested
shalges; and centritaoges for all types of sewage aludges.182
Sewage treatment is a field in which the potential of centrifuges
is just starting to be realized. Their first application was to assist
existing equipment to forestall a plant expansion.101,92,32
Only the solid bowil continuous centrifuge proved to be successful in
handling sewage sludges. I'ts contitnued success has been reported by
engineers, centrifuge manufacturers, and treatment plant operators.2253
Regardless of the number of articles already oited on centrifuge
performance, a great deal of hesitancy surro~unds the use of the machine
because the effect of the machine and process variables on the centrifuge
performance have not been clearly defined for sewge shudges. In the
past, centrituging has been approached more as an art than as a science
and has been called a "witch daotor's operation."'
If centritages are to achieve their anxirmum potential in the field
of sewage sludge devatering, then the dewatering mpechanian and the effoots
of the factors which influence centrifuge performance must be more clearly
Therefore, this study was made to further centritage toohnology
insofar as the process and machine variables affect the performance of
the solid bowl centrifuge when dewatering domestic waste sludges.
*A statement unde in a seminar at Georgia Institute of Teahnology,
by Dr. W. W. Eohentelder, Jr., Professor of Civil Engineering, Unriversity
of Texas, April, 1967.
First Centritural Dewatering in Germanr
The First attempt to use a centritage for dewatering sewage sludge
was made in 1902 by Herman Schaefer at Cologne, Germeay. This centriftage
ibreed liquid through aieves by centrifugal force and after a dethnite
number of rotations the devatered~ oake was discharged. Later, Schaeter
combined his efforts with Dr. Qutaoter-eMeer and under the latter's
direction a anch improved machine was designed and built. This sachine,
kn~own as the Sohaefer-ter-Meer cenrtritage, was erected at Frankfor6.on-
the-Nain, and was subjected to a series of peribruance tests for krD
The first centritage installation for dowatering sludge at a sewage
treatment plant was at Harburg, Germany, where two Schaeter-te>Meser
centrifuges were installed in 1907. Shortly after this. Sour were
erected at Banover and six at Franktor6.on-the-Main. Later( one machine
vae installed at the sewage worke in Bielefeld, and one at Moscow.35
These early machines were batch ~type, but operated continuously on an
automatic evole of filling, dewatering( and cake discharge.
Centriftlral Dewaterina in the Upi~ted States
ALn improved Schaeter-ter-Meer centrithge was shipped to the United
States and used at Mlwaukee in 1920 for dewatering activated sludge. It
was impossible to get a clear effluent (oentrate) at ea econonioal, feed
rate an~d further improvements were made to the machine as a reanlt of
these tests. The machine was then sent to Baltimore in 1921. Tests were
conducted at Baltimore on raw, seudb-digested, and digested aludge. The
relative ease of dewatering the shudges was in that order, with digested
sludge being the easiest. Results were so proadaing that Further im-
provements were made to the Sohaefer-ter-Meer machine and in 1924 a con-
tinuous, automatically controlled, batch operation mo~del was sent to
Baltimore. Approximately 300 gallons of sludge were centritaged per
batch for 11 minutes. The inet and outlet time required an additional
11 minutes. The average recovery o suspended solids was 65 per cent for
digested sludge. Concern over the high solids content in the effluent
from the centrifuge prompted studies to determine its effect on the BOD
of the raw sewage influent to the plant and chemical precipitation of
the solids. The influent biochemical oxy~gen demand, BODI increased 4.3
per cent. The addition of 0.28 to 0.56 ounces of alum per gallon reduced
the BOD of the centrate from 2815 to 1295 parts per adlllon, but the BOD
of the coagulated centrate was considerably higher than that of the un-
From the centritage performance tests at Baltimlore, it vae con-
oluded that sludge dewatering by centritagation was a process worthy of
consideration and study; however, if the prooeas was to become widely
accepted, a low cost method for further tre sting the centrate would be
RawJ sludge was devatered at thle Collingawood Sewage Disposal
Plant in New Jersey in 1934* After two and one-half years of practical
service, several controlling factors of operation were reported very
vaguely by Peckerr5 (1) chemicals were of no aid in dewatering raw
sludge; (2) best results were obtained with fresh sludge s (3) a reason-
able pH variation was relatively unimportant; and (4) sludge cake pmo-
duction, recovery efficiency, ard cake solid content increased directly
with sludge conoontrat~ion.
In 1939, a continuous feed, high speed (6,000 rpm) De Laval cen~tri-
fuge was installed at Peoria, Illinois, for thickening waste activated
sludge. Five months of testing ensued. The performance test results on
activated sludge may be sumrmed up as follove 5 '9 (1) th ailudge ass
thickened A~m i to 5 per cents (2) the recovery averaged 75 per
cent at an average feed rate of 30 gpm; and (3) the return of the centri-
fuge effluent to the aeration tanks did not seriously affect the operation
of the activated sludge plant. Tests conducted on dewartering primary and
digested aludges determined that they could not be handled because of
their grit content. Approximately four machine shutdowns per day were
necessary for cleaning: dogged orifloes and removing accumulations of
sludge built up on the bowl wall. It was also recommended that an opere
ator be present in the room with the centritage during operation.
In the years just prior to, and during World War II. only light
attention was directed toward centrithgilng of sewage shadges. As Dr. F.
Kiess said in his review of shudge treatment in Germany, "the Germaan
people had other anxieties and urgent tasks than concerning themselves
with the application of new ingenious methods for sludge dewatering.'!
This statement applied equally as well to the United States.
Thre first report of centrituging savage shudges following the war
was in 1950 whren the performance of a De Laval diso centritlage thickening
waste activated sludge was tested at the Sious Falls Senage Treatment
Plant in South Dakota.M Following the successlful testing program, tue
machines were reported to have been installed. The results of the testing
program conducted at Sioux Falls were (1) an 85 per cen~t sludge volume
rednetion was effected; (2) there was no detrimental effect on the primary
clarifiers by returning the centrate to them; and (3) it was necessary to
shut downr for bowl cleaning once every 48 hours. Allowing publication
of the original article there was manch debate as to the benefits derived
from thickenring the sludge. *'3 The Sioux Folls Treatment ln~t does
not currently use any centrithrges for sludge dewatering.
Two unique circumstances provided the dramatic opportunity for
centritages to prove their practicality in dewatering sewage aludges.
The first occurred in 1954 at the Daly City Treatment Plant, Califetnits.18
This plant was of necessity constructed on a site surrounded by dwellings
and business establishments. The consulting engineer selected centriFugal
dewatering of digested sludge as the method most free from odors and
least unsightly In appearance. ~A Bird 18 x 28 inch solid bowl continuous
centrrifuge was installed in 1954 and has operated snooessfullly sinee then.
Recent;ly the 18 inch was replaced by a 24 x 58 inch Bird centritage Jhen
the plant capacity was expanded. A typical analysis of the present oper-
ation is in Table 1. No detrimental effects are incurred by returrning
the cenltrate to the primary olarifier; in fact, settling is enhanced by
a synergism effect.
The second uniqule abocuastance occurred in 1956 at the San Leandro,
California,plan~t when three digesters had gone sour because of overload-
ing. The resultant problem of maintaining plant operation concurrently
with thle handling of some 100,000 gpd of raw sludge, whiile the digesters
were being emptied, was a brlaidable one. A 40 x 60 inch solid bowl Bird
centritate emptied the sour digesters a~nd dewatered the raw shudge.184
These two successful applications of the solid bowl centrithge
revived the interest in their use for deuatering sewage and indnetrial
waste. Bird Machine Company reports more than 50 centritages currently
operating at various sewage treatment plants throughout the United States.3
Table 1 lists the current operating performance of centritages installed
at seven Pnunicipal sewage treatment plants.,
In revriewing the history of centritagal dewaterng of domestic
waste shadges, the factors which contributed to the unsuccessful applioa-
tion of centrithges in thle early installa~tions were (1) the high solid
content in the centrate and an unwillingness to accept it; (2) low
capacity throughput; (3) necessity to screen the sludge prior to centrit-
agation; and (4) a high percentage o downP-time for maintenance.
The recent succeastal applications of the cenrtritage to dewatering
domestic waste shudges can be attributed to a number of factors including a
(1) the accep~tane of a less than clearn centrate and the fact that the
returned fines are not detrimental to some treatment prooeasest (2) im-
provement in the maohinea' operating characteristics brought about by
This information was supplied by plant superintendents in reply
to a questionnaire sent by the author.
location A B C P E F G
of Centrifuges Dewartering Domestic khate Sludges
40 x 60
24 x 38
24 x 38
1. 2-3. 7
24 x 38
Peed rate, gpm ...
Racovery B 71
location* A B C D E F G
Polymaer -- -- --
Times gravity 1333 875 -- 1350 --- 1350
(inches) 2.25 3.0 4 2 1/2 3 7/8 2 1/4
alr/d, D/wk 6;5 24;7 Shr/day 6hr/wk 14chr/uk Varies 24; 6
Cake disposal Land Drying Land Land Drying Barged
Fill beds --l 11111 beds to sea
centrate Ocean Sludge Aeration Pre-
returned Influent outrall --- Influent drying tank aeration
centrate Unknmown --- Beneficial Disastrous Unkcnown None
-operated -- 8 5 11 1/2 7 2 4
A Atlanta, Ga., B Anomymous, C MIanitoba, Canada, D Daly City, Calif, E San. Dist.
No. 6, Saoramento County, Carmichael, Calf., F San ALntonio Leon Creek Plant, Texras, G Joint Plant
countyy of Wes~thester) Yonkers, NJ. I.
*Sise of Bird centrithges installed (bowl aises diameter timesS length in inches).
Table 1 Continued
keen competition among shrdge handling equipment manufacturers; and
(3) the recent increase in research and publication of reanlte on centrit-
ugation which is making more sanritary engineers' aware of this method for
THEORETICAL CONSIDERATIONS OF CENTRIfPUGAL SEPARATION,( SMDGBg
CHARACTERISTICS, ANID CHEMICAL FIDCCULIATfDN
Centrifugal Solid-Lignid Separ~ation byr Dansity Difference
Despite the fact that centrithging is one of the oldest unit
operations, many empirical factors still have to be introduced into the
estimation of centritage performance. The accepted mathematical treat-
ment for evaluating centrithge performance is presented in this section.
Special reference to the assumptions that are generally made in arriving
at that estimate is included. The author dishes to give special recog-
nition to Charles M. Ambler, G. A. Frampton and &. E. O'K[. Trowbidge
whose published articles contributed the 611owing theoretical relation-
Derivation of Theoretical Relatwnship
When a force is applied to a particle in a fluid medium, the pare
tiole is accelerated, F--aa, until it reaches a velocity along the line of
the force at which the resistance to its notion equals the applied force.
In a settling tank this force is gravity. In a cenltrituge, this force is
the centrifugal force, F = am2r dynes, created by the contrithge bowl
spinning ata~o radians per second at a radina of r contimeters from the
axia of rotation, which obviously varies as the particle moved under its
influence. Allowing for particle buoyranoy in the suspending medium~the
absolute mass, m, becomes the effective mass, V(P- *), where anEod
are the density of the particle and fluid medium respectively, and V is
- 14 -
the particle volume.
It the particle is spherical, ite volume is It D /6; if it is not
spherical, its diameter, DI anat be postulated statistically such that
its volume is still V =n D /6. Then the 6rce producing notion is
F~dyns) =(f -)M~r(1)
Movement of the particle under the influence of this force is
immediately hindered by the resistance of the fluid medium to action
through it. Fbr small particles moving at moderate velocities (below
the ~tlr~bulent range), the resisti~ng force is proportional to the relooity
of the moving particle, and for the particular case of a spherical par-
tiole is defined by Stokes' law as
R(dynes) = 35ffDu
in which q is the absolute viscosity of the liquid.
The particle accelerates until it reaches a terminal velocity
where the centritagal force producing action and the viscous resistance
to notion balance each other, then
and solving fbr the particle velocity
a = D2 2r (2)
Since the centritage capacity is manally limited by Its ability to handle
the smallest particles that settle slowly in a given system, the formula
becomes of major importance in the analysis of centrithuge peribraance.
- 15 -
For a simplified approach to a study of this system, consider a
cylindrical centriibge bowrl carrigng a relatively thin liquid layer of
thickness S at a rate Q e.o./sec. through the bowl. See Figure 3. It
is usual to assume the velocity of a particle, p as constant actosa such
a layer, so that the radial distance traveled by the particle in time t
seconds is xoentimetera where
S =ut =D2 Qqo 2C
It is further assumed with much less justification that t = V/Q(
where V = volume of liquid in the bowl at any given aoaent and Q is the
volumetric rate of 3ow in comparable units. The radial distance traveled
by the particle may now be written
x= D2 (94.r) C2r ]1
If a is greater than the initial distance of the particle from
the bowl wall, it will be removed from the liquid phases otherwise it
will1 remain in suspension and be discharged with the centrate. To insure
elmination of the particle from the centrithge, the distance x must be
equal to at least the thickn~ess of the liquid layer in the bowl. Thus
X'Q = 2~ sD = rg-rg = (3)
18 4 Q
18 TL Tg S
which can be written as
Q = KE
- 16 -
SoLID PARTICLES PASSING
OUTWARDS THROUGH LIQUID
90LIDS BEING COMPACTED
AGAINST BOWL WALL
S*THICKNESS OF LIOUtD LAYER
Fig. g Cross Section of a Cylindrical Centrifuge Bowl Showing
the Separating Mechanism.
where k = constant
]tD2 (- &) containing only data relating
TL to the materials being
= c0 ontaining only data relating
gS9 to the cenltrithlge itself
Amler1~ has named this i entity.~P sima wih can be deined as the area
of a gravity settling tank of equivalent separating power to the centritarge.
- 17 -
Its value depends solely upon the centrita~ge parameters, which can be
divided into ~two factors, one concerned with the field force generated
by the machine, and the other with its geometry, ~thus s
r - (6)
This particular approach to centrithxge peribreance has come to be n~own~
as Sigma Theory."6 Fom equations (4) and (6). the naId~aum throughput
at which a particle of a given size will be elindnated from the feed
stream of any given centrifkge can be expressed as
E~qua~tion (7) holds only on the condition that (a) the system be-
haves in accordance with Stokes' Law and (b) that a true vale for A can
be found. A thorough exaemination af these assumrptions and the reasoning
in arriving at the value for correction factors are available.
Existing co~mmercal centrifuges do not approach the idealistia performance
indicated in equation (7). But the amount by which a machine falla abort
of this performance is characteristic to each kind of centr~ithge.%
Appliatieon of the Sippa Conoopt
The assumptions and conditions which were set forth as the basis
of deriving equation 1, 2, 4, and 7 impose limitations to the application
of the At concept. The assumptions concerning the feed material are the
1. Particles are spherical in shape and uniform in siz~e. They
are not to deaggregate, deflooculate, coalesce, or flooculate during
2. Particles are evenly distributed in the continuous liquid
phase and settle as individual particles.
3. The settling velocity of the particles is each that the
Reynolds number does not exceed one.
Assumptions concerning the flow conditions are 3
1. Streamline flow with uniform distribution of the feed in the
full liquid layer.
2. The layer of the deposited solids do not distburb the flow
pattern and rem~ixing of the deposited solids does not occur.
Theoretically, the algma concept al~nva performance comparison
between geometrically and bydzed~ynantoally similar centriauges operating
on the' same feed material. This is frequently made use of in scaling up
to a Afrll size machine from results on a laboratory or pilot plant centri-
fuge. Equation 7 above that the aedialentation performance of any ttD
similar centrithges treating the same anspension will be the same if the
quantityq has the same value for each.
From equation (7) we may write
1 1 D21(~
and for a given degree of separation on a given feed material the exp~ree-
sio n2 n is a constant for the laboratory or pilot centri-
fuge. hence, to carry out the ease degree of separation in centrithges
(1), (2), and (3), they mu~st be operated at a throughput to satisfy the
aondti ~- = Conatant = = (8)
- 19 -
Sedimentation Performance Curve
To transfer the performance results from one centrith~ge to another
the ratio Ql. verous fraction of solids unsedimented is plotted on loga-
rithodo probability paper in the indicated order, where Q is the flow rate
in any convenient unit and sigma is the performance factor in units con-
sistent with1 Q.W
According to the principles presented above, if the system is ideal,
that is to say it has a stable particle size and its clarification proper-
ties are determined by the settling power of the' aentrituge, all centri-
fuge performance data should fall on the same line as QI. versus frao-
tion of solid unsedimented. When the curve for a particular machine
deviates from a straight line, it is an indication that the sigma correla-
tton has broken down, and,therefore, extrapolation zom one machine to
another is recormmended only on the straight-line portion of the curve.
Sepprption in the Contimuous Solid Bold Centrifqge
In the continuous solid bowl centrithge w~ith conveyor discharge,
the mechanism of the separation in the pool consists of sedimentation
hindered by five disturbing factorst5 (1) solid oake layer moving along
the bowl wall, occupying space, and influencing the residence time for
the liquor, and the radial distance of travel for sedimenting particles;
(2) space occupied by flights of the conveyor which reduces the residence
times (3) turbulence created by the relative action of conveyor to bowl;
(4~) a liquid flow pattern which is complex and dittionlt to assess as it
takes a spiral path around the conveyor flights in its passage in an axial
direction toward its discharge points and (5) hindered settling of the
solids as they approach the cake surface in a alarry of high concentration.
- 20 -
Frampton5 calcutlates correctilons for these five disturbing factor for
Sharples Super D.Ca~nter* centrifugos.
Initially ignoring these five disturbing factors, it is possible
to c~alulate a sigm~a value for the pool section of a cylindrical type
bowl with a conical sootion. This type of solid bowl centrifuge is
illustrated in Figure 4. The theoretical value of 4 is given by the
em217*21 (3r2c +2 r2) 2ffe 22 (r2.)21if + r8l A
g 4) g 8
The formulae for sigma values are of utility in comparing the
capacities of mobhines having the same conveying velocity and configura-
tion. Bience, as a design tool within a limited field they have utility,
but in general the treatment represrents too great an over-simfplifatioaon
to be satisfatorily applied to sealing-up.4 Practical opertatin of
continuous bowl centrifuges is guided in general by the above theoretical
considerations but actual performance cannot be predicted by them alone.
For this reason, the experience factor is still basio to proper applton-
The dewatoring aechanisml in the solid bowl conveyor discharge
centrifuge is the most complex; therefore, the scaling-up of th~is class
of equipment abould be treated with great caution by those not widely
experienaed in centrifug technology.
Hakndsiotaved by Sharples-Eqguipment Divison, Pennsalt Chemicals
Corporation, Warminster, Pa.
Fig. 4 CrorSeotional View of Ogpical Cylindrical.
'15pe Solid Bowl Centrituge.
- 21 -
- 22 -
The sludge characteristics which affect centritugal dewatering
are: (1) size, shape, density and charge of the solid particles; (2)
viscosity of the sludge centrates (3) compressibility of the solid pa>-
tioles; (4C) solids canoontrations and (5) chemical composition of the
The size, shape and density of the solid particles have a profound
effect on their sedimentation by centritagal force. From equation (2)
it is apparent that the settling velocity and therefore the clarifying
capacity of a given centrifuge will increase with increased particle shze,
greater solid-liquid density difference, and decreasing liquid viscosity.
The shudte solids are generally amorphous and compressible in the~ aise
range of 200 microns and les.84 SUch IIaterial devaterL prilmarly by
expulsion of water as the solids compact against the bowl wall.i The
larger nonl-compressible solid dewater as liquid flova through the con-
The viscosityr of the sludge water is of academic interest in most
cases, since no advantage can be taken from the inverse relationship of
viscosityr to temperature.
Nevers reports that centriikgal separation is easier with alarries
of high solid concentration. If this effect is tnre for sewage
slurries, it has not been reported.
The sludge chemical composition affects the amount of obasical
required for improving the dewatering characteristic of a sludge. Gente 30
has shown that two factors abould be considered; namely, (1) the bioazrbo-
nate alkalinity of the liquid, anrd (2) the ratio of the volatile matter
- 23 -
to ash. However, a direct correlation between chemical requirements: di
these factors is not always attainable, and in fact, the physiological
characteristics of the sludge and the particle size seem to be more sig-
nificant in determining the chemdoal requirements.T Baraman52 aingled
out particle asle as the anat important meaenrable property di~stinguish-
ing between good and poor filter rates at ten treatment plants using
vacuum filters. Ikudolfa and Balmat claim that the colloidal traction
exerts one-third of the chemical demand but represents only one-sixth of
the total solids on a gravimetric basis. The chenboal composition of
shudge is determined by the source of the sludge ;and the prooeas of
ibrma~tion. Exaaustive investigations have been accomplshed to determine
chemical composition of sludge.535 Regardless of the relative roles
of the chemical and physical properties of the sludge, it is well known
that the various aevage treatment prooeasea produce aludges that have
varying dewatering characteristics. Colloidal behavior is the one common
denominator of all sewage eludges which is responsible for their dewatse-
Domestic waste aludges are complex colloidal systems consisting
of (1) colloidal particle a wth mean diameters from ten ang~s~tmas to one
microns (2) supra-colloidal particles with mean diameters from one to 100
micans; and (3) large particles which are aggregates of hydrophilic
colloide.55 Experiments peribrmed by WW: 6 indicate that anat
colloidal particles separated from raw sewage were hydrophobio. Primary
sedimenltation removed 50O per cent a these hydrophobio colloide from the
- 24 -
raw sewage. They settled as loose aggregates and contributed to the
formation of the sludge. The supra-colloidal solid and larger aggre-
gates of hydrophilic particles comprised approximately 90 per cent of the
total raw sludge solids. The contributionstof particulate fractions to
the total suspended solids in savage are respectively, 52, 42, and 6 per
cent for settleable, supra-colloidal, and colloidal fractions. Micro-
steve analysis of activated sludge by Kennedy 84 4. abowed that between
28 and 72 per cent of the suspended solid had a mean diameter of 45
microns. Wet screen analysis of elat-riated mesophilic digested shudge
showed that 80 per cent had a mean diameter less than 74 micronsr.5
H~ydrophilic colloids in seowge may be proteins, their products of
hydrolytic decomposition, and other organic compounds of biological origin.
The primary charge on the lydrophilio colloid is due to the reactive amino
and carbolgyl groups in the molecules of these biological substances. In
water, the amino group bydrolises and depending on the ph of the system
one or both of the groups dissociates. Wi~th the central rmoleoular
structure represented by the eyabol, R, the dissociation can be depicted
by the following expressions
coonIL coo" Or coo'
R R +- R
L L .'* L,
. 25 -
A ph increase from the isoelectric point depresses the ionisation of the
hydrated amino gmoup and results in not negative charge. A decrease in
ph from the isoelectric point depresses the tonisation of the oarbolplic
gmoup and results in a not positive charge.g
The hydration of hgdrophilic particles is dependent on the same
functional grpaps, -08, -0008, and 482H, and the atmotbare of the selecules.
These knortional groups being water soluble, hold a sheath of water Firaly
around the particle (bound water). Figure 5 is a schematic sketch of a
protein partocle of colloidal size showing the par~tole encased in its
bound water. The particle with its bound water envelope moves as a singl
BydrophobiC colloide in aewage are manally inorganic and negan
tively obarged. ThPey have no afinLdty for water. Therefore, hgdrophobio'
colloids are not encased in bound water.
The primary charge on a h3ydrophobio colloid partiole is thought
to be the result of preferential adsorption of solution ions on the par-
Hiole surfaces. However, since the particle charge oca be reversed by a
change In ph, it has been postulated that adsorption of either H+ or O~f
ions is responsible for the primary charge on the particle.596
The primary charge on either hgdrophilic or hydrophobio colloid
particles attracts solution ions of opposite charge. As a result,
oppositely charged ions increase In the immediate vicinity of the par-
tole. It the primary charge is earticiently large( a compact layer of
counterions ibrma adjacent to the particle, called ~the 8tera layer or
* 26 -
Fis* 5 schematto Sketch of a Colloid Protein Particle Encaseda
in Bound Water, (FromRI~o 1.; p. 136:,)
compact double layer. Brownian movement and indnoed velocity gradients
prevent the Stern layer from establishing electronenrtrality and fluid
mpotion difthaea counterilans into the solution proper.6 The regon be*
tueen the Stern layer and the solution proper is called the Gov.~Chapman
layer or diftnae double 14qer. The counterion concentration in the dit-
inee layer varies from a relatively high level at the Fixed dittheee
bounrdary gradnally out to the ooncenltration of ions in the bulk of the
solution where electronoutralityr with the counterions exists. Figure 6
is a sohematic sketch of the double layer surrounding a colloid particle
* 27 -
vith a negative primary charge.
Concentration differences between oatdonic and anionic species
result in the establishment of an electtostatio field around the partocle.
The potentials through the electfrostatic field of a negatively charged
colloidal particle are illustrated in Figure 7.
The electrochemical potential is the potential across the entire
ionic double layer at the solid-liquid interface.0626 Seta potential
is the potential gradient over the GoupiChapman di3~use layer, or simply
the potential at the plane of shear. This plane bras a boundary between
that portion of the solution around the particle that moves with the par-
tiole and the portion which can more independently of the particle.
The hear PLoan atonnd hydzophilic colloid particles coincides
with the exterior boundary of the bound water, but thr hyldrophobic col-
loide the plane of shear La located near the boundary between the fixed
layer and the diffuse layer.9646 The exact location of shear plane
is debatable. By using electrophoretic mobilitiee, measurable by a
unaber of techniques, the seta potential can be determined.
The wall of a ca~pillaryr or amnnuue through vbich a colloidal ans-
pension is threed to flow will quickly take on the charge characteristics
of the colloidal particles present in the liqurid. This effect is nearly
instantaneonseand reversible." When the particle with its fixed and
difitae layers of charges attaches to the wall and the liquid is breed
to flow past, the mobile counterions separate from the particle at the
shear plane and are physically swept downstream. This movement of like
electrical obarges is an electrical current and is called the streandng
Pig. 6 Schematic Sketch of the Dourble layer AzoPund a Colloid Particle
I19th a Negatr~e Primary Charge- (From Rich, p. 136,.)
OtSTANCE FROM PARTIILE SURFACE -p
Filg. ? PotenrtSala Wizlough the ElectrostateL Field Surrounding a*
Charged Hygdrphilic Collaid Partole. (From RIch, ]p. 138.)
. 29 -
current. If the flow system is made of non-ccndnoting materials, this
current will be forced to reinrn by ohmice condition though the stream.
ing liquid. Polarity is anch that when a pair of Qeletrodes is inserted
upstream an~d downstream of the flowing charged particle, the usptream
electrode has the same sign as the electrical charge on the particle.
The movement of the particles (along with their counterione) through the
capillary or annulae does not contribute to the streaming ourrenlt. The
streandng current results from the physical separation of counrterions
tom charged particle surfaces, and this occurs only in the case of
chargees imobilized on the surface of the caPIllary- or amnnlue. Gerdesi
mathematical derivation for streaking current is excplicitly for the con-
dition of alternating flow.
Colloid stability depends upon the not resultant of. the forces of
attraction and repulsion acting an the colloid. A colloid system is said
to be stable if the colloidal condition is more or less permanent.
The two most important forces of instability are Brownrian movement
and van der Waala' forces of attraction. Brownian movement is imparted
to the suspended particles by their collision with rapidly moving mo~le-
adles of the spending Iliquid. Van der Waala' forces are weak forces
of attraction between atone and/or molecules. They are oanased by perms.
nent or Induced dipole features of the particle adleoules and are similar
to polar bonds. When two parties having dipoles meet with the appro*
private orientation they attract one another.
The forces of repalaion in a )lrdrophobic colloidal system are
attributed directly to the zeta potential. The stability of hydrocphobic
colloide can be destroyed by neutralizing or reducing the seta potential
on the particle surfaces. The stability of hydrophilic colloids is
attributed to the repulsive force of seta potential and the bound water
which acts as an elastic barrier to keep the particles from coming to-
gether. In order to reduce the stability of hydrophilic colloids, one
or both of these factors anat be removed at least partially.
Chemical conditioning is done to improve the dewatering character-
istics of the sludge. The conditioning process implies a flocculation
reaction wherein individual sludge particles are united into rather
loosely bound agglomerates, or floca, thereby increasing the effective
Chemical reagents are classified as flocculants when they react
with anopended matter at the solid-liquid interface and thereby affect
colloid stability. Chemical flooculants are salts, surfactants, colloids,
and natural or synthetic po~pers.
Flocculants alter the sludge particle properties through one or a
combination of three mechanismast (1) rednotion or neutralization of the
seta potential; (2) dehydration of bound vater s and (3) extensive ionic
cross-linking or bridging of particles by synthetic, long-chain, high-
moleanlar eight polymers.
The seta potential of both bydrophilic arnd hydrophobio colloids
can be reduood by ad sting the pH of the system toward the isoelectrio
point. At the isoelectric point the primary charge is zero and no double
layer exists to produce a seta potential.
The seta potential can also be reduced by adding ions or colloide
of opposite charge to the colloidal system. The addition of counterions
serrea to increase the concentration of counterians in the fixed double
layer, anld the seta potential is reduced.
Reduction of the steta potential depends upon the valence of the
opposite charged ions by the Schulse-Hardy rule. According to Schalte;;-
Bardy, a bivalenit ion is 50 to 60 times more effective than a monovalent
ion, and a trivalent lon is 700 to 1,000 times more effective than a mono-
valent ion.70o The colloid stability is ettected only lightly by the
chempical narture of the omoun ion or its valnce charge.a
The bound waterot hydrophilio colloids can be reduced by adding
salts in high concentration. The anions of the salt compete with the
colloid particles for the bound water. The effectiveness of dehydration
depends on the nainre of the anionsadded scoording to the Boiheeiter
series.5 The Hoh~eister series lists the following anions in order of
decrasidng effecivenaasSO ,0 Cl", s0 ". .
Colloide can be floconiated through the addition of polyelectro-
lytes. These materials are high moleonlar weight, long-chained, organic
polymers with a aultitubde' of repeating functional groups along the chain
length.7 The length of the polgser moleanlo extends Into the colloid
size.2 ~Men the polymer is dissolved in water the iunntional groups
ionise. Anionic polyelectrolytes contain acidic groups and when dissolved
in water positive iona xbl1 tonise off the polyelectrolyte chain, leaving
the kunotional sites negatively charged. Cationia polyelectrolytes con-
tain basic groups and when ionised, the negative ions leave the chain
and the sites are positively charged. Nan-charged polymer chains are
- 32 -
called nonionic polyelectrolytes.
The size and shape of a polyelectrolyte in solution depends on the
net charge of the polymer as infinenced by the ph, the nature of the poly-
mer, and the ionic valence in accordance with the Sohnrse.Rardy rule. The
uncharged molecule is like a contracted chain. The electrically charged
chain can be vianal^ sed as a random coil whose length has increased duo
to mutual repulsion of charged sites along the chain.
when used as flocculants, high molecular weight polyelectro'lytes
fthnotion according to the principles of mutual coagulation of sole.5
The polymer rapidly dif~naea to the surface of the oppositely obarged
particles and is adsorbed by the process of ion exchange where it is
dehrydrated and neutralized along with the oppositely abarged surface.
Because of the polymer chain length, the polymer-particle interaction is
more efficient than that of mutual coagulation of spherical particles.
The polymer chain may attach itself to several particles and
establish a bridge between them. This can occur also between particles
held by other polymer molecules, thereby c rose-linking of polymer chains
occurs. aShs process abort-circuite the classical flooculation processes
by building a floo via couloabic rather than by van der Waals' forces.
The flooculation rate is increased, the floo is toughr3~, and the agglom-
erated particles settle more rapidly.
Even though complete neutralization of colloidal charges may not
always be necessary, the optiana polymer dosage depends primarily on the
surface charge density of the colloids which can be determined electron*
phoretically.l Because this value is an average quantity and depends on
the total veighit of solids, dosage can be expressed as weight petr cent
or POundsQ Per ton.
MIPERIMBTAL EqillPWNBT, MATERALS, AllD PRCOCEDIIM
Bird Solid Bowl .Centrimmres
The continuous solid bowil centritages in this study were com~meroial
machines manufactured by Bird Machine Company, South Walpole, Mass. Bird
centrifuges for dewatering domestic wasite aludges are classified as
countercurrent flow or concurrenlt flow. In the first type solid and
liqid pass through the bowl in opposite directions, whereas in the
latter (concurrent flov)~ the solid and liquid paea through the bowl in
the same direction.
Countercurrent flow solid bowl oentrifkes The Bird contianuone
solid bowl centritage with counteronrrent flow is illustrated as a cut-
away drawing in Pigure 8. The two principal elements of this centrituge
are the (a) rotating bowl which is the settling vessel and the (b) con-
veyor which discharges the settled solid. The bowl has (c) adjustable
weire at ite larger end for discharge of clarified effluent, cromrmonly
called the centrate, and (d) solid discharge ports on the opposite end
for discharging the dewatered solida, aiaply called oake. As the bowil
rotates, centritugal force causes the sludge alarry to form an annular
pool. The pool depth is determined by the adjustment of the effluent
weirs. A portion of the bowl diameter is reduced and not submerged in
the poolforming a () drainage deck for dewatering the solids as they
Fig, 8 Bird Solid Bowl Continuous Centrifuge With Countercurrent Flow.
- 35 -
(g) SLUDGE FEED PIPE
(c) ADJUSTABLE WEIRS
(h) CONVEYOR HUB
(i) FLOC NOZZLES
Courtesy of Bird Machine Company, South Walpole, Mass.
(e) POOLo BW
(f)DRAIN AGE DECK
(d)SOLIDS DISCHARGE PORT
(1)CHEMICAL FEED PIPE---
are conveyed across it. The bowl and conveyor rotate in the same direc-
tion, with the conveyor speed approximately 99 per cent of the bowl.
speed, depending upon the particular gear unit ratio employed.
Sludge fed into the centrifuge enters through a stationary (g)
supply pipe and peasses through the (h) conveyor hub into the bowl itself.
As the solids settle out in the bowl, due to cenltrithgal force, they are
picked up by the conveyor and carried along continuously to the solids
outlet; meanwhile. the clarified effinent continuously overflave the
If cheadcal treatment is required to floocoulate fine suspended
solide, the cheadeal solution can be introdnoed into the centrithrge
through a (i) separate cheadcal feed pipe, then through the conveyor hub
into the pool by means of (3) floo nosslea projecting beneath. the liquid
pool surface in the settling sone. Thus the floor nossles gently ads the
chendeal solution with the liquid pool by means of the ditterential
rotating speed between the conveyor and the bowl.
Concurrent flow solid bowrl centrithere The out-away drawing in
Figure 9 above the operating principles of the new Bir concurrent flow
centrithge. The twoc most important ditterences between the conanrrent
flow centritage and the countercurreist flow oentritage are s (1) the point
of introducing shudge alurryr into the bowil and (2) the method of dis-
obarging the clarified effluenrt.
Sludge fed into the (a) rotating bowl of a concurrent flow centri-
thge is discharged through the (h) conveyor hab ahead of the settling
sone. Settled solid are carried along by the (b) conveyor in the same
(a) BOWL------ (b) CONVEYOR
(j) FLOC NOZZLES-7 /(n) SKIMMER
(h) CONVEYOR HUB7/ HOUSING CHAMBER
(g) SLU DGE FE
(o) GEAR UNIT
(f) DRAINAGE DECK
(d) SOLIDS DISCHARGE POftT
(m)SKIMMER ADJUSTMENT CRANK
(1) CHEMICAL FEED PIPE
Fis. 9 Bird solid Bowl Continuous Centrilfuge With Concurrent Flow.
Courtesy of Bird Machine Company, South Walpole, Mass.
direction as the Iliquid, hence the flow pattern through the bowl is
smooth. Settled solid are conveyed over the entire length of the bowl
and they are not disturbed by incoming feed or turbulence caused by the
Clarified (k) oentrate is discharged under pressure by a (1)
akiamer which is adjustable to regulate the pool depth. Adjustmnent is by
an external (a) hand crank and can be made while the machine is operating.
The skimmer is housed in a (n) chamber to confine~ any tulrbalencoe that
aldght disturb the settled solids.
If obea~cal treatment is required it is done in the same manner as
described for the counterourrent machine. The obemical solution is gently
aixed with the liquid pool as a result of the differential rotation be-
tweenr the conveyor and the bowl.
FiloEEEE~~~~EEEt Plant Easipment
The pilot plant for dewatering sewage shudges was located at the
Phelps Laboratory for Environrmental Research at the University of Florida.
The site was located adjacent to the campus sewage treatment plant from
which raw, activated, and digested shudges were obtained for experimentaliiiiio
purposes. The schematic flow diagram for the sludge dewatering pilot
plant is abown in Figure 10.
The equi~mrent list for the pilot plant' included the 611owinpg
major pieces of equipment a Bird 6 x 12 inch solid bowl centriiuge,
positive displacement sludge feed and cheadoal dosing pumps, arladge
holding tank with stirring aechanisas, and metering devices for me~asrr-
ing the rate of polymer addition.
PUMP m + I~u
SLUDGE HOLDING TANKS
Fig. 10 Sohematic Flow Diagram of Sladge Dewatering Pilot Plant,
University of Florida.
- 40 -
Bird 6 x. 12 inch centrithePe The 6 x 12 inch solid bowl centri-
rthge with countercurrent flow is illustrated in Pigure 11. The 6 inch
machine was driven by a 5 horsepower (hp) 3,600 revolutions per minute
(rph) explosion-proof AC motor, belted to the machine. Probalanced drive
sheaves were used to obtain constant speeds of 2,850. 3.750, and 5,810
rpm. The relative centrithgal force (R.0.2.) developed within the bowl
at these three respective speeds was 680, 1(180, and 2,880 times the
force of gravity. The gear unit ration, 10011, fixed the ditterential
rotation between the bowl and conveyor.
The pool depth was adjusted by means of twelve effluent ports made
up of four equally spaced boses on each of three ditterent radit. Three
sets of different diameter bushings were available (3/8, 1/2( 5/8 inch
diameter) burt only the 1/2 inch buahing was need in these tests since it
was representative of the deepest and shallowest available pool depths.
The 1/2 inch buahinggwere placed in corresponding holes 90 degrees apart
around the face of the bowl head. The remaining holes were plugged with
81ardae feed ran The eludge feed pump was a new Nbayno Model
1L4, L Frame. Type CDQ pump having a cast iron inlet boasing a hardened
tool steel chrome plated rotort and a synthetto Buna YI" rubber sltator.
The pamp discharge range vae from 1 gpa to 10 gps. The pump was driven
by a 1/2 hp U.S. Varidrive, Type VAV-HYT-OR, Frame No. 6.56-5. The
positive displacement pumping action provided by the Magno supplied a
M~anuacotured by Robbins and Meyers, Inc., Mopmo hamp Division,
MICanuf~actred by U.S. Electrical Motors, Los Anlgeles, Calif.
11 ,;~:, I
SLUDGE FEED PIPE
CHEMICAL FEED PIPE
-- GEAR UNIT
6x12 Inch Solid Bowl Cont~inuous Centriiuge.
Fig. 11 Bird
Courtesy of Bird Maabine Company, South Halp~ole, Mass.
- 42 -
continuous disobarge of sludge to the centritage without pulsation.
Since the volumetric displacement of sh~dge for each rotor revrolution
was affected only lightly by changes in discharge pressure, metering
of the flow rates was exceptionally acaarate. The variable speed drive
provided a convenient and reprodnoible means of changing the sludge feed
polvser doeinn ~mas The pump used most frequently for adding
polyelectrolytes to the sledge wars a completely rebuilt Moyno Model 1L2,
L Frame, Type CDQ pump construated as described before. The pump dis-
oharge range was from 0.070 spa to 0.70 gpm and was driven by a U.S.
Varidrive, Type VALV-BY-GR and Frame No. 6-56-5 with a 1/4 hp actor.
Botaertera The polymer dosing rate was continuously monitored
by rotameters. The rotameters employed were
1 Flowrator, / Mdel 10A1027A, Serial 67061r5623A1, ma~nzmu
rate 0.4)25 gpm.
1 Flouratorr Model 10A1027A, Serial 670655623&5, marnana rate
81nudae holdiar tanks 81udge from the University of Florida Sewage
Treatment Plant was tracked to holding tanks at the pilot plant. The tank
capacities were 200 and2.500 gallons respectively, each with actor driven
stirring paddles, and baffle arrangements to insure thorough agitation
during the holding period and therefore providing a unifera sludge concen-
tration throughout the tank. The rotational speed a the attrring paddles
Maantac~tured by Pischer and Porter Company, Warainator, Pa.
was 100 rps.
cainimetm Used at Treasure island. Florida
The pilot plant sledge dewatering equipment was track-mounted and
transported to Treasure island so that experiments could be condnated
with the Bir 6 a 12 inch centritage on the same shudges being dewatered
by the at .x 38 inch centritage installed at the Traeaure Island Water
Pollution Co~ntrol Phant.
BirdJ 24 x 8 inch.arrntrifteea The 24 x 38 inch solid bowl con-
tinuous centritgee with countercurrenmt flow was equlipped with a bowl
having a three degree cylindrical emotion and a ten degree conical drying
dook. The gear ratio was 140 51. The centrithge was driven by a 40 hp,
1,800 rps motor and the bowl rotated at a constant rotational speed of
2,000 rps. The relative centrithgeal force produoed at this rotational
speed was 1.335i times the force or gravity. The centritage was con-
atructed and operated as illustrated in Figare 8.
81udae.eed woom Raw and digested aludges were pumped to the
centritage with a Mayno Model ]lt, Type CDQ with a discharge range from
9 gpm to 90 apm. The poap was driven by a U1. S. Varidrive., Type. VAY.BY.
GR. with a 1 1/2 hp actor.
WSiase activated aladge was supplied by gravity to the centrithge
by means of a splitter head box in the returrn shudge 1ln, located 3 1/2
teet above the inlet to the cenrtrithger.
PpPol'a_.drOsinP manm The polymer dosing range was between 0.5 to
10 gallons per abamte. To provide fo r anh a wide pumping range, a P&W
- 44 -
Water Systems+ jet pump, Serial Nlumber P65, was uasd. The pnap was driven
by a 3/4 hp actor and provided a steady uniform disabarge over a wide
range of discharge heads arnd anotion lifts. A pump of this kind was
ideal for the testing program.
Whtame~ters The rotameters used weres
1 Flowrator, Model. 10A11027A1, Serial 67061562313, maxilanrm flow
1 Flowrator, Model 10A11027A, Serial (f/06156251), mazians flow
The pilot plant sludge dewatering equipment vae also transported
to this plant. Exrperiments were conduoted with the Bird 6 x 12 inch
centrithge on the same shudges beint dowatered with a new 24 x 60 inch
Bird centrithge installed at the Northeast Serarge Treatment Plant.
Sidard2 60_ inh centfriftee The Bird 24 x 60 inch solid bowl
continuous centritage was driven by a 75 hp, 1,800 rpmn notor. Bowl
rotational speeds of 1.520 rpm, 2,000 rpm, and 2,400 rpm were obtained
by three aheave sizes. The relative centritagal force was respectively
850, 1.350, and 1.950 times the force of g~ravity. The gear ratio was
14011. A full range of pool depth from a very shallow to a very deep
pool was readily obtainable by emberior adjustment at the effluent
skinager. The meahine was constructed as abownl in Frigure 9 and operated
as previously described.
M~anufactued by Flint and Wallinge K~endall'9lle, Indiana.
- 45 -
83udge fteestmpa The sludge feed pump was a Noyno Pump, Frame
SW128I, with a rated discharge capacity of 175 gpm to 275 gps. The ~pumpppppppppppppppppp
was driven by a U.S. Varidrive operating over a range of 35 rpm to 350
rpm, thes the actual pump discharge range was from 23 to 230 gps.
Polyaer dosinP manII The FANWtaer Systeima jet pump as previonaly
described was used.
Rotaneters A Flourator Nodel 10A1027A~, Serial 6706Af562YL4 with
a mazSiann flow rate of 10 spa was usaed.
Doeaestio Waste 81udaea
The aludges dewatered by cenrtrifkgesr in this study were obtained
directly from processing unite at each sewage treatment~~~aaa~~~aaa plant where the
experiments were condnoted.
Prior to centrithging any sludge in the 6 x 12 inch machine, it
was necessary to screen it through hardware mesh with 1/4 inch openings.
If the sludge twa not careened, the 3/4 inch diameter feed pipe and the
sooelerator within the conveyor hub were quickly plagged. Soreening did
not doorease the easpended solid of the return activated sludge; however,
earee~ag'ot the digested sludge reduced the snapended solids approximately
10 per cent on a dry weight basis. Those items retained on the 1/4 inch
screen would have been readily settled in the cenltrithge and would not
have exerted a chemical demand; however, they would have given the de.
watered sludge a more fibrous consistency. screening had the greatest
e~ffot on raw a~ladge, redoing the solid concentration of the raw aladge
by 50 per~ cent. This was attributed to the fibrous consistency of the
shtdge which quickly covered the openingsc and acted as a fine sroren.
81udges dewatered in the 24 x 38 inch and 24 x 60 inch centrifthges were
not soreened, but were pu~pe~d direat27 from the process unit to the
Since every plant has a different character of aesag eand aladge
to be deuatered, it is diffoutlt to point out average or exceptional
sludge characteristics; therefore a description of~ each plant included
in this study is preseanted.
University of Florida. Swre Treatoant Plant
Located on the camnprta the sewage treatment plant has a capacity
of two million gallons per day and is desigped primarily as a research'
facility. The University Plant has three types o f reatmenrt, including
standard and high-rate filtration and activated sludge treatment. The
digested eludge nreed in the experiments was a mixture of raw, trickling
itlter and wast e ativated shdges~ digested anaerobically in open, tn-
heated digesters. The return activated sludge used was bulkyr and had a
shndge volume index exceeding 200. The raw alndge consisted of raw
settled Manage and anyg digested sludge that settled from the digester
aspernatant returned to the plant int~nent. Since com~pletion of this
atedy the osapua plant is undergoing an exrpuanon progra~ and ambatita-
tion of contact stabilisation for the activated aludge process.r Ther
fore, no ahemratio flow sheet for this treatment plant is included.
Water Pollution Control Plant. Treasure Island. Florida
This plant is located in the center of town and is surrounded by
high-rise hotels commercial esrtablishments, and expensive homess it has
a capacity of 1.5 million gallons per day. A achematic flow sheet of the
Treasure Island Plant is presented in Pigure 12. The method of waste
water~ treatment at this pleat is the contact stabilistion ~prooeas,l
pfeneded by primary sedimentation. Raw, activated returas and aaer~-
obioally digested shudges were dewatered at this plan~t with the 6 x 12
inoh and/or 3kx 38 inch centrithgess. The plant has used aM z8 38 inah
Bird centritage in a dual capacity since 1965 to concentrate waste noti-
vated aladge every other day for three to five hours and to dowrater
digested aladge six to ten hours per month. The operation of the contab*
thge at this uiqulce plant has freuepntly beo presented b others.13,28,30
Northeast T~oreatmnt Plant. 8t. petersbrura. Florida
This plant treated five adlion gallons per day by means of the
high rate activated seration process.7 A schemati flow sheet of the
treatment plant is shown in Figure 13. ALpproimately 30,000 gallons of
waste alndge (2*5 per cent suspended solids) area generated per day. The
aladge used in the ex~perimets was characteristically bulky, bleak and
The aironastances surrounding the installation of the ar& x 4 inch
centrithge are important to understanding the experiments condloted at
The City of St. Petersburg purchased two Bird a x 60 inch solid
boul centrithges with concurrenrt flow for installation at the Albert
Whitton Seange Treatment Plant. The plant was then undergoing expansion
and modifloation to provide seacndary treatment of waste water far the
southeatt sootion of the city. Meanwhile, the Northeast Sewage Treatment
Eg* 12 Flowlhebboit the Mediined Contact Stabilisation Prooeas at the
Water Pollatiton Control Plant, Treasa,Tilland, Faloda
Fig. 13 Schematic Flow Diagrua of the Nlortheast Sewage Treatmrent
~Plant, St. Peteraburg, Fla.
developed an soute excess sludge problem. As a means of resolving the
problem, it was decided to divert one of the tw ocentriihges wihaich er
to be installed at the Albert \hitted Plant to the Ilartheast Plant.A
orash program va instigated in july, 1967, to install and operate one
of the 24 x 60 inch centrihtges at the Northseat Plant. The oentriflge
was to thicken waste aladge before prtmpshg it to the digester and de-
water the digested sludge. Bird Naohine Company conducted the start-urp
testing program in Mich the author participated.
The polyelectrolytes included in laboratory and centritagation
atest in this study are listed in Table 2. Purifloo C-31. Pripafoo C-7,
and CALT-FLOC were selected for embensive use in this study because they
consistently gare the best rresults. Their effootiveness, compared to
each other on an equal cost basis. was nearly the same. Parifloo C-Lp
and Primasthe C-7 are currently used extensively in vastewater treatment
and shudge devatoring; therefore, the results of this stagy can be coa-
pared with work done by others. CAT-ID~LC is a new polyser recently re-
leased for sale, although many years of research and testing have gone
into its dev~elopmna~t. Research conducted at the Universityr of Florida
has found CAT.FID)C to be an extremely effective polymer for ~turbidity
removal in water treatment and for sIo1~Uldelqud separation in seauge
treatment.737 Another factor which influenced the selection of this
polymsr vae thart Carlgon Corparation had developed a colorimetric test
for detcetn CAT-FIDC conontrations as low as ) to 5 ag/1 In water.7
CALT-FIDC: is a high solecular walght, linear homopolymer of diallyl
dimethylammonium chloride. TIhe linear obain hs reoorring 9-anbetituted
Purifloo SA11881 l
Polyelectrolytea Used in Laboratory Tests
Rohnm and E~aas
The Dow Chemical Co.
Naloo Chemical Co.
- 52 -
C C8 CB
Fig. 14 Reourring No~learular niht of lonised CAT-FIDC,
piperidinian halide units alternating with methylene grourps. The ionized
form of the CAT;IDC (cationic pokeear2 moleale i shown in Figure 12,74
Purifloo C-31 and Primafloo C-7 are polyaminoe of high molecular
weight and when ionized are oationic. Since thesle polymers are marketed
as proprietary chemicals, their ohemicoal components and molecular con-
figurations are not described to the trade and only generalired desorip.
tions can be applied to them.
819dge to be used in the pilot plant required collection and
storage of a sufficient volumne to insure completion of the exrperimaent.
The aludge was soreened at its point of colleotion and transported to the
pilot plant. It was then transferred to the holding tanks and stirred to
insure homogeneone concentration. Three Soo at samples were collected
for analyses of the initial aladge properties. A 500 at sample of the
unacreened aladge was obtained to determine the suspended solids cnonn-
tration prior to soreening.
No special sludge preparation was necessary prior to operating
the ten 24 inch centrifugea. Sledge was pumped directly from the process
unit into the centrithge. The sludge was sampled three ditterent times
during the day prior to the start of the test to get the approximate
aladge concentration that could be expected during centriibgation. This
enabled chemical solutions to be prepared in the desired range of dosages.
Deteredinations on Feed Sludae.. Centrate. and Cake
The 611owing deteradnations were made on the initial prope~t~ies
of the shudge, centrate, and oake whe~n applicable a total solids, suspended
solids, volatile suspended solids, alkalini~tyr pB, seletrophoretto mobility,
streaming current, sludge volume index, specifo goravity, temperature, and
residutal CAT-IDUC. The respective deterainations were made as follows
Total solide Total solid deteraibnations were made on the initial
feed shudge and cake in sooordance with Standard Nethods, Part III A.
Residue on Evaporation.M
Total ansPended solids Total suspended srolids~ deteradnations were
made on the intial feed shudge and on the centrate When the conoontra-
tion of the anespnded solids in the feed sludge or centrate wras estimated
to be greater than one per cent, the total easpended solids determinations
wFere made by filtering a 100 al sample through a previonely anabored,
dried (at 103oC for six hours) and weighed 15 on thatman No. I filter
- gai ~
paper in an 11 cm Buchne. thnnel prevented loss of solids around the edge
of the filter. The filters used were dried at 1030C for 24 hours, cooled
to weighing room temperature in a desicoator, and reweighed on a single
pan automatic balancee* The long drying time was necessary becanae of
the large number of samples being dried shooltaneously during a series
of test runs.
then the sludge or centrate snaponded solids caonenltration was
estimated to be less~ than 1 per cent, the total anspended solids deter-
ainations were made by filtering a 50 at sample through a prevrionasly pe-
pared, dried and weight Gooah oracible, containing~ a Fiber glass filter~s
(Rseee Langel 93448H). The crucibles were dried at 1030C for 24 hours,
cooled to weighing room termperaturre, and reweighed on the automatic
balance. The Gooah ortaibles containing a 2.4 cm glass filter were pre-
pared by filtering 100 al of denaineralised water through the filter,
drying at 1030C for six hours, cooling to weighing room temperature in a
desiccator, then weighing on the antomatio balane. The glass fiber 11-
ter was selected for use since it is anrperior to and anarh more convenient
than the old asbestoo mat technique.
Yalattle agnaended olida Volatile suspended solide determaina
tions were made on some of the su~pended solids samples taken of the
sludge and centrate in aodrdaoe with Standard Nethods, Part III D~76
The total alkalinity of the feed sludge wats deteradned in accord-
ance with standard Methods, Parts III and IV.7 The total auclnrtlraln is
Prodcbpt of ALugust Santer of New York, Inc., Albertson, New York.
Produot of Reere ALngel, Clifton, New Jersey.
- 55 -
expressed as ag/ CaCO .
Measurement of ph All ph measurements were nade with a Backmn
Model 0 ph noter.'
Eleetrophoretio makility Electrophoretie sobility deterdnations
were maXde by using a eeta-Meter.* in sooordance with the Zeta-Meter
Manual. The procedure for making a detedr~ntion of the makility bt
feed sludge or centrate sample was as follows. An 11 on Whatman No. 1
filter paper was prepared byr filtering approximately 100 atl of distilled
water through it and disoarding the filtrate. Next, approximately 50 at
samlple of filtered and about 10 al of filtrate was dilated to 100 sa with
distilled water for use in the apparatus. Experience has abovn that fil.
Loring the s~ample does not mraterially alter the electrophoretic makility
of the sample.9 The Zeta-Meter cell was washed with distilled water
and pipe stem : 01eaners followed by rinsing the cell with the filtrate.
of the samlple. The cell was then caretally filled with the filtrate to
avoid babble formation. Twenty~ individual particle makilities, ten in
either direction, were used to obtain an average particle mobility for
each sample. The calculated maklities are tim~averaged rather than
Streamlya current Stresaing current determinations were made bry
using a Strearbag Current Detector in sooordance with the StreadnP
MIanufaotrrred by Beckmran Instruments, Inc., Fullerton, Calif.
An instrument manufactured by Seta-Meter( Inc., New York, Ni. T.
An instrument nanufactured b~Waters Associates, Inc. ,
- 56 -
qurren Detector Insrtruction Manual.8 The procedre for making a deter.
adnartion of the SCD instrument reading of a feed sludge or centrate sample
was as follows. The boat and piston of the instrumaent were scrabbed with
Lava soap and rinsed thoroughl with distilled water. Then the boat and
piston were washed with the samlple itself. The reservoir boat was then
filled with the unfiltered sample and the SCD reading observed on the 10E
linear scale recorded. The boat wase emptied and refilled with additional
aliqurota of the same samle and the SCD read until two consecutive instra-
maent' readings were within f 2 aioroamperes. Usually three determilnations
were necessary for each rample.
Sladae volumle inday The sludge VOl~um index (SVI) determinartions
for activated aladges were made recording to Standard Methods, Part Y C.%
Specific ravity Specific gravity determinations workfacomaplished
in accordarnce with Standard Methods, Part V.76
Tqaperoture The teperature of the sludge was determined in the
holding tank prior to centrifugation by means of a bulb partial imrerson
thermobmeter with a range of -30 to 120oF.
Residual polymer Residual CALT.IDUC polymer deteradnations were
maode according to s procedure developed by Calgon Corporation.75 The
test originally developed for Polymer 261 is applicable for CAT-FIDC
detectionr because its properties are aialar to type 261. The test is
based on reduction in color intensity of indigo caraine in an alkaline
mediumr. Aooording to Kleber,8 the test alght be useful in connection
with the dewdhering of shudge, -but he notes that other nitrogen copounpad
in sewage may interfere. Compounda known to interfere in the test for
Polymer 261 areas shelating agents anoh as Calgon sad EDIA whdeh cause low
reanlts in the test. Bigh concentrations of Nlg prodnoe a turbidity due
to the precipitation of magnesina hydrozide.
Residual CAT-FLDC was determined in only a portion of the cantrate
samples. Samples were prepared by filtering at least 150 atl of the non-
trate through a 0*45 BAW9 Eillipore ritte.+ Deterednations were made
by the procedure outlined by Calgon Corporation with the following mnodi-
floation. The absorbance value for the buffered sample wcas subtracted
from the sample with indigo carmine and butter to correct ibr samp~le
color and turbidity.
Laboratory ]Deuaterina Tests
A part or all of the following laboratory tests were condnoted on
the sludge prior to centritaging to estimate the anat effootive polyser,
polymer dosage, and dewatering characteristics of the sludge.. The respeo~-
tive laboratory tests were made as follows.
Polsmer flocculation Three 250 al g~iraduted oylinders were rulled
writh sludge, the desired polymer dosages, from stook solutions, and dilated
to 25 at with distilled water. The respective polaaper dosages were added
to a cylinder and the cylinders inverted four times. The following was
observeds (1) rate of floo formations (2) degree of separation into dis-
thaot agglomerated floos and clear liquor s and (3) the degree of coapto-
tion upon settling measured by the aludge depth after three minutes.
Bmkacher Amne The speoitic resistanoe of the sludge was deterr
mined by the Buchner thnnel test. The test was seoalplishe in accordance
with the teehnique reommended byJ Paro~n.82 An 1, onl WatmarnNo. 41
filter paper was placed in a Bucohner wthne, moistened, and set in place
*A prodnet of Millipore Corporation. Bedford, Massahusetts.
- 58 -
by applying a vacuum for a few seconds. The desired polynsr dosage was
dilated to a volume of 50 at and added to 200 at of shadge in a 250 atl
gradulated cylinder. The polymer and sludge were then mixed lar inverting
the cylinder four times. The polymer conditioned sludge was poured into
the Buchner funnel and after approximately five seconds for sledge oake
formation, a 20 inch nernary vacuum was applied to the filter. The fl-
trate was collooted in a 250 at graduated syrlinder and its volume was
recorded at 30 second intervals from 0 to 270 seconds. This procedure
was repeated for at least four polymer dosages. The speoitic resistance
was then calcuilated from the data by the well-known method presented by
Eckenfelder and O'Connor.8
Laborato~ry entri~age Thirty at aliquots of sludge were placed
in 50 at beakers. The desired polymer dosage was dilated to 10 al and
added to the sludge. The beaker was inverted tens times to abx the poly,
aer and the sludge and then poured into a 40 al graduated, heavy duty
centritage tube (Kioax Brand, Serial 45404). Four chestarl dosages were
prepared at a single time and then centritaged in an International Centri-
fage / Size 1, Model SBY 12440. The samlples were sooelerated as rapidly
as possible to a apeed equivalent to the relative centrifugal forest to be
used in the solid bowl centritage. The sample was span for 22 seconds
(the approximate detention timejin the Bird 6 x 12 inch centrifuge at a
deep pool and a 1 gpa sludge feed rate) and braked to a rapid stop. The
Ma~nu~actured ~by the International Equipment Compangt, Boston,
relative clarity of the supernatant and its depth were noted between
the different dosages. The solidity of the cake was estimated by probing
it with a 1/16 inch diameter glasa stirring rod.
ExpayiMental Deaggn and Evalruatton
The experiments conducted in this stdtdy were arranged sooording
to an ordered plan in which all the factors were varied in a sgrstematic
way. It was thee possible to determine the Main effoot of each indivildual
factor and the interactions. The experimental designs were of the randea-~
ised blook, spliti~plot( and split-split-plot types, with two replications
mlade of each treatment comb~nation.84'85 The essential teature of the
split-plot experiment is that the supb-plot treatments are not randon-
ised over the whole block but only ovesr the main plots. Randoaksation of
the subf~tretents is different (oharacteristically sMaller) than that for
the Main treatments. The randomaisation was secomp~lisrhed b using a table
of randoamly assorted digits.
The analysis of variance for each experiment was oaloalated by the
IBM~ 360 coapater at the University of Florida Computing Center. The com.
pater program ca'rlrculad the analysis of varianee table by the modiiesd
K~ronecker product method. The program vae written by Sara Kephart, Pro.
grader, University of Florida Complpting Center, and is maintained and
available for use in the Center's library. The program solution assumesr
that the contrasts are antuarlly orthogonal. A aian IPUof seVen factors
is allowed and 15 responses can be run at one pass. The user has the
option of either individual degrees of 1*eedom printed out or just the
total of all degrees of freedom for the particular source of variation.
The program output provide, in say order specitled by the user, each
source of variation along with the degrees of freedom, the ana of square,
the mean square, and it individual degrees o freedom are, required, the
matrix rootor product and its divisor.
In the analysis of variance table the significance of the attects
wnas tested by the F-ratio, where the mean square for the effoot was
divided by the mean aqluare for error, and the result compared with tabu-
lated P distribution values.
The tested were conducted careholly to rednoe the exrperimental
error. Approximatelyr two anonths of operating expearience was necessary
to become familiar with the centrithg~bng procedures, to work out the
experimental procedures and analytical tests, and to improve the mechan-
ieal equipment in support of the project.
Polysmer arep~aration Blmc dea is defined as' pounds of poly-
mer (Ilquid or solid) as received from the manufacturer per ton of dry
suspended solid in the teed aledge, as opposed to aolvaer dosina rate
which is defined as the volnaetric rate of adding a polymer solution to
a sludge. Polymer solutions of desired concentration were prepared in
sufficient roxlume so that they could be added directly to the aladge or
within the centrifthge without trther dilation. This method, although
impracitual on a fall scale operational basia (in acat cases), was ideal
for testing purposea since it eliadnated another source of experimental
error. This procedure was used for all tests in this study except for
the final ones conducted on return shudge at the N9ortheasrt Plant in St.
Peteraburg, Florida. At that time the obesical pumps and dilatian water
arrangement for the 24 x 40 inch centritarge had been installed and were
Order of test rune The order of the test runs was randeadsed
according to the partionlar statistical design of the experiment and
recorded on data sheets as shown in Table 3.
Calibration..of ~ranes Before starting the test runn the sludge
feed and polymer dosint Pumps were calibrated. The shudge pumps were
calibrated by measuring the volume of their discharge over a period of
two pmintes. The variable speed drive setting was marked wrhen the desired
discharge rate was obtained. Consistently reproduoible results were ob.
tained by this technique.
Polyser dosing rates were determined by measuring the solution
drawedown rate in the polymer holding tank or collooting the discharged
volume over an interval of two adnates. The ro~tametess installed in the
polymer discharge line were calibrated stpllanltaneonly for enoh solution
to be used.
E~stabliabine easilibrium The time requPired for establishing
seqilibrium within the centrifkge depended upon the aira of the machine.
Equilibriuma can be defined as that condition when the centrate and cake
being discharged from the centritage are consistently the same for the
existing conditions of the test run. The time required to reach equi-
libriumm is much greater than the flow through time. Time to reach
SSolids in feed. TS
Soisin centrate S
Solds in Cake T
Rate ceadd p
he dosage. lbs/ton
RC.F.. x rvt
- 62 -
- 63 -
equilibrina in the three machines used was s (1) 4 minutes for the 6 x 12
inch centritage; (2) 10 mndutes for the 24 x 38 inch centrithge; and (3)
15 minutes for the 24 x 60 inch centrithge. TIhe cake and centrate were
sampled after equrilibrium had been established. A stop watch as well as
visual inspection were used to determine when equilibrina had been
Saelina the centrate The centrate was sampled by a volumetric
technque. If the centrate contained a high percentage of suspended
solid, a 100 at volume sample wras collected directly in a 100 at gradul
ated aylinders otherwise a 50 m1 graduated cylinder was used. The
cylinder was rapidly passed back and fo~rth through the centrate stress
over a period of 30 seconds. The anaponded solids analysis a m comp~leted
immediately on the vcolumetrio centrate sample. Centrate samples collected
in containers and analysed at a later time were in error by as unach as
10 to 20 per cent in some cases, because the easpended solid aggaomerated
and stuck to the side of the container, thaus rendering the samqple non-
A second sap~le of the centrate was collected in a liter jar and
was used for residual polymer, electrophoretio mobility, and streandng
current determinations as soon as the centrithge test runa were comple~ted.
Samp~line the eludae cake A sludge oake sample of approximately
four ounces ~wa collected in a paper cup. The total solid analysis was
scoooplished immediately by transferring a portion of the cake to a tared
dish, weighing the wet cake, and then drying it at 103oC. Cake samples
collected in sample 3ara and stored for analysts later were found to be
in error from 5 to 10 per cent because water continued to be expelled
from the cake by action of the polymer.
Data Rlecordina and Caladlations
The centrate and cake samples, crucible, filter, and dish numbers
and caloadations were recorded on the Calculation Sheet abown in Table 4.
Suspended solid concentration in the centrate samples was corrected for
the volume of polyser solution added by saanaing that the liquid valence
is discharged as centrate. 1his corrected value for solids in centrate
was used in calculating the recovery of suspended solid.
The per cent recovery of suspended solids ~from the feed was cal-
oulated by the following formulas
Recovery, = 100 100~~-~ 100i~~~;sn'oi x 100
# Solids in Cen- g Solids in Cake
This formula, derived from a materials balance, corrected the per cent
recovery for the moisture content of the cake.
Sample Volume, ml
SS9 in Centrate
SAsh SS = -- x 100 =
Dry Cake + Tare
Dry Cake Weight
ckTS Cake = ----- x 100 =-
(kAsh Cake =: ---- x 100 =
Sample Volumre, al
Dry SS + Tare
Centrate Crucible No.
Dry SS + Tare
SS Ash + Tare
SS Ash Weight
Cake Dish No.
Wet Cake + Tare
Wet Cake Weight
Ash Weight + Tare
Feed Filter No.
Drg TS + Tare
BBESULTS AND0 PISCUSSION
The data and analysis of variance for each experiment in this
study are tabulated in Appendix 1. A glossary of teras used in this
study is presented in Appendix 2. Results of centrifkgal dowatering of
domestic waste aludges are presented In the folloxdng order 3 anaerobically
digested aludges activated sludge and raw sludge.
In this research the effoota of machine and process variables on
dewatering domestic waste shudges were evaluated. These wee centriiutaal
force, pool depth, sludge feed rate, sledge concentration, polymer, poly-
mer dosage, and location of polymer addition.
The split-plot and split-aplit-plot design for the factorial
experiments were chosen for practical reasons related to time and ettort
required in changing the machine speed and pool depth, The experiment
became too large ii more than three variables were evaluated at three
levels. Therefore, the approach sas a systematic investigation of two
or three machine and/or process variables at a time. Esafh treatment
combination was applied twice, that igatreate~ntsl ware replicated two
times. The replicates were established as blocks ainee the sludge de-~
#atering properties were likely to be more homogeneous within each set
of replications than over the whole experiment. The block error can be
attributed to changes in the sludge characteristics over the length of
- 66 -
. 67 .
time to complete the experiment.
The signifloance of the ettects was tested by the F ratio. The
asterisk in the analysis of variance tables indicate the probability
that such a result would be obtained if the distribution of the parent
populations were the samle.
DevateriLna Direated 81uae
The results of empperiaents condnoted on digested sludge are shown
in Figure 15i through M~. For ease and crlarity or understanding the ob-
jective of each experiment, the variables Lnves~tigarted and the responses
obsered ar e hown in Table 5. Exqpsiments 1 through 4 were condnoted
with the samhe aladge sample (1.500 gallons).
The first experiment evaluated the effort of the sludge feed rate,
polymer dosage, and location of polymer addition. The experimental design
was a splt-plot with two replicates made of each treatment combination,
Bert The analysis of r ariance thr this exrperiaent indicates
highly signitioant two and three fPactor interactions between feed rate,
polymer dosage, and location of polgmer addition. The presence of inter
action destrove the additivinty of the main effoote. For cexaple, the
effect of polymer dosage levels is not the same for every feed rate. The
diffeencoesbetween the recwovey means at CATFIDCX dosages of 10 and 40
pounds per ton at 1 and 2 gprm are 19.86 and 7.79. The intersetion (12.07)
may be expressed as the amount remaining between the two dittarenaea.
The polym~er dosages, then( ar e aaply mor e othaent at 1 gpm than at 2
sps. Figure 15 clearly saows this intoeration. interpretation of the
Table 5 Continued
a; a x x o oA
6 1 I i o o C 3
*Centriage~ A = del2 in, B = 242t38 in, C = 24r60 in.
Isain 1 t*. fFa 2 = Treasure ldatd 3 = St. Peterabars
- 70 -
10 IGPM T
CAT-FIDC dosage (1bs/toli)
0 to 4 70
Fig. 15- Effect of Sludge Feed Rate, Polymer Dosage, and L~ocation Polymer
Added on Performance of 6x12 Inch Centrifuge Devatering Digested
Sludge, University of Florida.
Centrithge operation: 1180 x gravity, deep pool (0.594e in.)
Feed properties SS = 4.214); VSS = 64.575; pH = 7.1
Polyvmer addit~iop Dosing rate 108 of sludge feed rate.
- 71 -
main effect is now possible and has signifloant meaning.
The mean square for error vae smalls therefore, the F test became
very sensitive and detected all. the interactions. The graphical pre-
sentation of the data in Figure 15 abows the effect of one variable an
the other, sad reliable conclusions as to the main effects can be derived
in the presecoe of highly sigdnifoant in~teractions. Baob point plotted
on the graph is the mean of two replicates fbr ecoh treatment ~omnbination.
The results of Exqperimnct 1 arer
(1) 81udge feed rate had the greatest effect on recovery and
cake dewatesring. As the feed rate increased the recovery decreased exrpo-
nontially but the oake solids concentration increased exponentially.
(2) Becovery of suspended solids was more complete for a given
CAT-FLOC; dosage when the polgmer was added within the centrithge as
opposed to addition outside the centritiage. Recover increased from 64
per ocnt to 95i pr oct with CAT-FLOC dosages from 0 to 30 lbs/ton, but
in order to recover the reai~ninsn 5 per oct anspended solid in the
centrate, the CAT-FID)C dosage had to be doubled. Polysnei* addition
was ineffootive at sludge feed rates of 2 and 3 gpm. For example, at
2 gpm and CAT-F~IDC dosage of 70 lbe/ton the recovery increased only 13
(3) Cake solids concentration was inverrsely proportional to the
This experiment was designed to study the sortace charge propertiesr
of the particles remaining in the centrate as polymer dosage and location
of addition were varied. The experimental design was a completely
Rdgggg The results shown in Pigure 16 may be anamarised as
(1) Polyser addition within the centrihrge was moPre effective in
increasing the recovery efficiency than when the polymer was added either
before or after the sludge teed pump.
(2) Electrophoretio mobility determinations on particles in the
centrate showed that as recovery continued to improve, the electronega-
tivity of the particles was rednoed and approached tero. Recovery
exceeding 99 per cent was attained when the particle mobility was rednoed
to sero. Polymer added in excess of this dosage reversed the mobility of
the particles remaining in the centrate.
(3) The electronogativity of the aladge particles in the centrate
increased upon centritagation w$.thout say polymer addition. Support for
this phenomenrongs Dstrengthened by electrophoretic mobility deteradnations
prior to centritaging which showed the average mobility of twro samples to
be -1.5 >/aee/v/ams but, after contritugation the mobility at twso samples
was -2.0 and -3.0 respectively. To determine if this was an event. de-
pendent on the centrli~gal force, the following experlaegnt was designed.
In this experiment centrih~gal force and polymer dosage efforts
on dewatering digested sludge were deteradned. The experimental design
wras a split-plot.
Rsggtgg The result abovn in Figure 17 may be summarised as
f ol1ows a
- 73 -
+ 1p~ IN CENTRIFUGE
e- 1 .PUMP
e e DISCHARGE SIDE OF
e FEED PUMIP
0 10 40 70
CAT-FLOC dosage (Ibs/ton dry solids in feed)
A SUCTION SIDE OF FEED
*DISCHARGE SIDE OF
J~ I FEED PUMP
0 10 40 70
CAT-FLOC dosage (1bsfton dry solids in feed)
aR LOCATION POLYMER ADDED
44 o 0- IN CENTRIFUSE
oa SUCTION SIDE OF FEED
t8~ D ISCHARGE SIDE OF
8 a FEED PUMP
0 10 40 70
CAT-FLOC dosage (1bs/ton dry solid in feed)
Fig. 16 Effect of Polymer Dosage and Location Polymer Added on Per-
formance of 6x12 Inch Centritage Devatering Digested Sludge,
University of florida.
Centrifuge ope~a~tions 1180 x gravity, deep pool, feed rate
Feed properties: ss = 4.2kfk, pH = 7.1, Mobiity=.1*5 p/soo/v/om.
Polymer additions Dosing rate 0.10 gpm.
I0 40 TO
CAT-FLOC dosage (1bs/ton dry solids in feed)
CAT-FIDC dosage (1bs/ton dry solids in feed)
0 1040 70
CAT-FLOC dosage (1bs/ton dry solids in feed)
Fig* 17 Effect of Centritugal Force and Polymer Dosage on Performance
of 6312 Inch Centritage Dewatering Digested Sludge, University
Cen~tr LtPupeoperation a Deep pool, feed rate 1 gpm.
Feed properties: SS = 4.2 f, Mobilitr= 1.5i Ir/seek/olm.
Polymer addi~tion Do sing rate 0.10 gpm within centritage.
- 75i -
(1) Hecovery of suspended solid and cake solids concentration
increased linearly as the centrithgal force increased. She analysis of
variance supports this fact at the 1. per cent level of signflaance.
(2) An interesting phenomenranoocurred; namely, the electro-
phoretto mobility of the anspended solids in the centrate sooompanying
dewatering without polymer treatnet was dependent on the level of cen-
trithgal force. The mobility was lesa when~ centritarged at 1,180 times
gravity than at 680 and 2,880 times gravity.
(3) Particle electronegativity was a thnotion of polymer dosage
only and appeared to be directly related to the initial particle charge.
For example, wrhen the particle nobility was -2.54* /se/v/om after cen-
tritagation at 1,180 times the force of gravity, the mobility was reduced
to 0.05i r/soo/v/cm at an applied polymsr dosage of 40 lbs/ton; but for an
initial mobility of -4.0 4/sec/v/amo at 680 times the force of gravity,
the mobility was reduced to only 13.0 W/s~oof/om.
In this experiment the dowatering etticiency of three oationric
polyelectrolytes was compared on the same shrdge. The experiment was
designed as a completely randonised block. The comparison basis waps equal
cost polymer dosages. She prices usred for the polymers in this experiment
were Primafloo C-7 a $1.07 per pound; Purifloo C-3L $0.34 per pounds
and CAT-FLOC was assumed to be $0.34 per pound, since CAT-FIC was not
then commeroially available.
]|@Blts The results of this experiment are abown in Figure 18
and say be aunaarised as follotrs a
m (i\\ \CAT
as A OUT
vi m IN.
0 3 It at
Polymer dosage ($/ton dry solids in feed)
i- ~CAT o(
(p O OUT
(4 r C-51
Q a OUT
Polymer dosage ($/ton dry solids in feed)
Fig. 18 Effect of Three Cationio Polymers and Location of Polymer Addition
on the Performance of 6x12 Inch Centrifuge Devatering Digested
Sludge, University of Florida. .
Centrifuge operation: 1,180 x gravity; deep pools teed rate
Feed properties: SS = 412 $g alkalinity = 1810; pB = 7.1
Polymer addition: Desing rate 0.10 gpma.
(1) All three polymers increased anspended solids recovery more
ettentirely wihen added within the centrii~ge.
(2) CAT-FUIC and Parrifloo C-Lp performed nearly the same on the
basis of recovery and oake solids concentration.
(3) Pripasthe C-7 was better than the other two polymers at lower
dosages, giving higher recoveries and drier cakes. : A given dosage.
beyond that required for 90 per cent recovery was aceampanied by a de-
orease in moisture content of the cake.
This experiment rws prompted by the results observed in Experimeant
3 as shown in Plgare 17 where the particle mobility in the centrate did
not reflect the recovery effiatency. For example, at a centritagal force
of 1,180 tmaea gravity the recovery was 98 per cent and the centrate pare
tioles had an average mobility of *0.5i / eIJs/ov/ams however, 99.9 per
cent recovery at a centrthigal force of 2,680 times gravity was sccoom
panned by centrate particles having an average mobility of -3.18/8secv/om.
To provide additional information on the centrate particle charge and
centrate clarificoation streaming current deteradnations were made on aen*~
trate samaples. In addition, the residual-free CAT-FLOC polymer in the
centrate was determined by the colorimetria technique. This experimesnt
vae designed as a sp~litplot.
EasuAts FiLgure 19 abows the results which may be anamarised as
f ol1ove s
(1) The electrophoretic mobility of the suspended solid in~ the
centrate increased negatively upon centrifugation rdthout polymer
CAT-FLO~C dosage (1bs/ton drly solid in feed)
0980 X GRAVITY
A2080 x RA~VITY
(Ibs/ton dry solids in feed) .
O 80 40 60
CAT-E100 dosage (Ibs/ton dry solids in feed)
Fig. 19 Effect of Centrifugal Force and Polymer Dosage on Electrophoretio
Mobility and Streaming Current Response of Particles in Centrate
1Jhen Dewatering Digested Sludge, U~niversity of Florida.
Centrifuge operation: Deep pool (0.594 in); feed rate 1 spm.
Feed pjropsertiess SS = 3.45k; pH = 7.0; Mobility =-1.8 p/sealv/ca.
Polyvmer addition: Dosing rate 0.10 gpm within centrifutge.
- 79 -
treatment. The average intial Peed sludge mobility wass -1.8 Ls/sealvion,
but after centritagat~ion without polymer treatment the average mobbllity
was -3.2 and -3.8 r(/seofvja at 680 and 1,180 times the three of gravity
respectively. Particle mobility was reduced to~ sero- at a CAT-FLOC)
dosage of 40 lbsr/ton and recovery was approximately 98 per cent. The
results contradict those obtained in Experiment 3 but support those of
(2) The streaming current instrument recorded sero streaming our-
rent at polymer dosages different than those determined by electro-
phoretic mobility determinations although the trand in charge rednotion
wast sinklar. No additional information of value can be extracted from
this use of the Streaming Current Detector instrument in this exrperiments
however, it was~ used in subsequent experiments.
(5) Deteraduationr for CAT-FLOC did not detect any residual at
dosages up to 40 lba/tons however, at doeages of 70 lba/ton the average
residu~al were 4 ag/1 for centrate collected at 680 times the fore of
gravity and 20 ag/1 for samples centriitaged at 2,880 times the force of
The experiments up to this point have evraluated the effect of
centrithgal force and feed rate at a constant deep pool depth. Pool
depth is the second and the last machine variable possible to invesrti-
gate in these pilot-plant experiments with Bird 6 x 12 inch solid bowl
The effort of centrithrgal force, pool depth, and sludge feed rate
* 80 -
were detertt~~~ttt~aedtt~~~ttt in this ex~periment. The experimental design was a split.
Br&it Pigure 20 shows the results obtained. These wre st
(1) Pool depth affects recovery of suspended solid eand oake
solids concentration to an extent greater than centritagal force and feed
rate. Pool depth aota linearly in its effort on dooreasing recovery anrd
oake moisture content.
(2) The effects of feed rate were the same as previonely found
in Experiment 1.
(3) Thre effort of increasing the centrifugal force was sign~it
cant at the ;j per cent level in increasing recovery as the centriugal
force increased, but insignificant in accounting for a drier cake prodct.~t
Enerimnta 7 and 8
The result of Expeimento 1 through 6 have shown that centritagac
tion can achieve clarity of centrate or oake dryneas, but not both of
these desired responses shaultaneously. The two following experiments
were designed to evaluate the effort of the feed sludge concentration at
conditions which would give good clarity or a dry oake. For the best
clarity a deep pool depth and a low feed rate was used. For the driest
cake a shallow pool depth and a high feed rate was used. The centrith~ge
wase operated at centrithgl forces of 680 anrd 2,880 times the force of
gravity and the polymer was introduced within the centrithge bowl at a
volumetric rate equivalent to 10 per cent of the aludge feed rate.
The experiments were of the split-split-plot design. The results
of the deep pool exp~eriment are shown in Figure 21 and the reanlta of the
shallow pool experiment are shownr in Pigure 22.
I I I
>" o a
oy sp o
I Ie~ 1F afo
- 82 *
*I e 15
) ** *
~p peeS o 1 eScoe
4 'eywD uy sPT~og
14 Peo 0f
I e 2 *