7€¦ · experiment 5-1: chemical oxygen demand . . 92 · ,experiment 5-2: refiltration time........
TRANSCRIPT
_ 7 Q TABLE OF CONTENTS 37gä 7 Page 4
7 7 7 _ 7 .77 Acknowledgements . . . . .............. ii
List of Tables . . . ........'........ vii 1
7 1 List of Figures ....7.............. ix 1I. Introduction.................. 1
II. Literature Review ............... 3 6. Modern Leather Production .......... 3 Q Q
_. 4Vegetable Tanning Process ......... 5 ~Leather Processing Wastes7......... 73 Treatment of Spent Vegetable Tannin Liquor. . 8‘ Q Mixing with Domestic Sewage . ‘....... 8QQ 3 7Mixing with Alum Tanning Solution ..... 10 Q4Biological Treatment............ 10Q Chemical Treatment.Q............ 11Nature of Colloids. . .”.
.·......... 13N 7 · 33 7 Colloid·Structure ...... .·....... 13T 1 Electrical Double Layer . . .r....... 14Q }· ‘ 7 I Attractive Forces Between Sol Particles . . 163 7 Development of Zeta Potential—Stern Model . 17 7 Q17 QColloidal Stability ............ 21 1 7Mechanistic Theories of Coagulation ..... 2477 1 Q Chemical Theory .....·.......‘. . 24
7Q- Adsorption Theory . ._........·. . . 24‘ · Physical Theory . . . . ......·.... 247 3Q« ‘_”
3 Q Perikinetic and Orthokinetic Flocculation . 26 3 7 -_ ·· Flocculation and Coagulation. . . ...·. . 26 7 7· · Sweep Floc. . ....·. . .-........ 27 ° 6 r‘ Hydrolyzed Metal Ions . . . ... . ._..... 28 §_ 3 > Aluminum Ions ......·......... 28 @Ferric Ions . ..............7. 32 3 77 Polymeric Flocculants 344Cationic Polyelectrolytes ....7..... 34 1 1Anionic Polyelectrolytes.......... 36 Q_ ·Nonionib Polymers ............. 37 >§Q QFlocculation Mechanisms of Polyelectrolytes 37 7 3 j_ 3 ‘ ‘ Applied Polyelectrolyte Research...... 39 _·‘· 3 Independent Variables Affecting Colloidal 33 7· Stability . . ............. 44 ·Q‘—3Inorganic Ion :3....· . ....3..... 44 3}LQ pH Q.........·........... 44 é3 3 ‘Time Lag Between Inorganic Ion and Polymer. 46Q
5
‘ 3 _ 7 7 iii Q 3 .;4 Q 7 . ‘ 4 Z _ .
-1 7 3 7 ‘ ·„ 7 ;
9 Page
Salt Concentration of Colloid System.... 46Method of Mixing.............. 47Age of Hydrolyzed Metallic Ion....... 47Temperature ............... . 47
Dependent Variables Used toMeasureCoagulation....... . ...... 47
Turbidity ................. 48Zeta Potential............... 48Filtration and Refiltration Time...... 50Colloidal Titration ............ 51Specific Resistivity............ 51· Solids Concentration. . L ......... 51
Summary ................... 52III. Materials and Methods ............. 53
Experimental Materials. .4.......... 53_ Experimental Equipment............ 55Experimental Methods............. 59 ~
Experiment One: Characterization of Waste. 59Experiment Two: Deterioration of Tannin
Waste and Storing Procedures......60Selectionof Independent Variables
to be Tested. . ............ 61Experimental Design ............ 62Experiment Three: Gravimetric Mass4 Balance ................ 66Experiment Four: Selection and
Comparison of Dependent Variables . . . 66Experiment Five: Selection of Procedures
for Determining Dependent Variables . . 67Chemical Oxygen Demand.......... 67Filtration Time ............. 68
9 Light Scattering Turbidity......._. 68‘ Light Transmission Turbidity.‘...... 68 4. Refiltration Time ............ 70
Subsidence Volume and Rate. . ...... 71Tannin Concentration........... 71Total Volatile Solids . . _. .... 724 Volatile Suspended Solids ........ 73
Jar Testing Procedure . . . . . ..... . 74Experiment Six: Polyelectrolyte Testing
and Evaluation ............ 78 gExperiment Seven: Expanded Testing 4 4- Procedure for pH Evaluation ...... 78
_ iv
· Page ·
IV. Experimental Data and Results ......... 82”Experiment One - Characterization of Waste. . 82Experiment Two - Deterioration of Tannin
Waste During Storage...... l....· 84Experiment Three - Gravimétric Mass Balance . 87‘ - Related Gravimetric-Testing . . ...... 896 Experiment Four - Accuracy of Dependent
Variables ............... 90Experiment Five - Calibration of Dependent
· .. Response Procedures .......... 92Experiment 5-1: Chemical Oxygen Demand . . 92
· ,Experiment 5-2: Refiltration Time..... 92Experiment 5-5: Tannin Concentration .1. . 94
· Experiment Six — Polyelectrolyte Evaluation-. 94Dependent Variable-Volatile Solids
· Reduction Relationship......... 96 .Evaluation of Independent Variables .... 101Selection of the Best Polyelectrolyte-
Aluminum System 116Experiment Seven - pH Evaluation in Expanded
Test ................. 116‘ Aluminum and PolyelectrolyteConcentration Variables ........ 117
_ 1 Time Interval Variable .......... 117*pH Variable ................ 121Overall Parameter Evaluation........ 121
V. Discussion of Results ............. 125 '- Nature of Spent Vegetable Tannin Liquor . . . 125
Gravimetric Relationship of Treated Waste . . 127 „J Response Variable Evaluation. ._....... 128 ·' Color Removal ........‘.·...... 129 ‘
Chemical Oxygen Demand........... 150Settled Volume and Filtration Time .... 150Turbidity .................. 151
Independent Variable Evaluation ....... 151 ’Aluminum Sulfate Concentration...... . . 152"Electrolyte Concentration and Type..... 155Time Interval ............... 154pH. . . . ..........* ....... 155i Optimum Values·of Independent Variables
_ for Colloidal_Destabilization—Economic° · Considerations............. 156
Theoretical Consideration of Polyelectro-lyte Mechanism. ............ 158 · —
Jar Testing Procedure . . . ........ 142Limitations of the Investigation...... 142
V .
Page ° .VI. Couclusions ................. 143
VII. Bibliography. ·................ 145VIII. Appeudix 160
IX.° Vita. . 5..........1 .... g..... 173
vi
3 LIST OF TABLES ·
Table p · Page‘
I. Reported Results of Spent Vegetable Tan ·I I I I I I I I I I I I I I I I I II_
II. Regression Equation with Independent Variablesand Their Code Value Range ......... 64
III. Experimental Design Test Sequence Using CodedVariable Values....... ; ....... 65
IV. Expanded Regression Equation with IndependentVariables and Their Code Value Range . ._. . 80
V. Expanded Experimental Design for pH Evaluation 81VI. Results of Experiment Onezq Characterization
of the Tannin Waste............. 83VII. Results of Experiment Two: Deterioration of
Tannin Waste ge............... 85VIII. Results of Experiment Three: Gravimetric_ Mass Balance ....... . .... . . . . 88-IX. Results of Experiment Four: Accuracy of
Dependent Variable .........._. . . p 91EX. Regression Equation Coefficient Results for
Coded Variables in Experiments 6A to 6C . . 102 pXI. Regression Equation Coefficient Results for pCoded Variables in Experiments 6D to 6J . . 103O
XII. Regression Coefficients for Experiment Seven,· Extended Test for pH . .'........... 118XIII. Data for Experiment 6A, Cationic Nalcolyte
· I I I I I I I I I I I I I I I I I I IIR
XIV. Data for Experiment 6B, Cationic Nalcoylte’ I I I I I I I I I I I I I I I I I I! IIXV.
Data for Experiment 6C, Cationic Nalcolyte. I JI I ,I I I I I I I I I I I I I I I II„
* ‘ viil
M
TableI
V Page
« XVI.. Data for Experiment 6D, Cationic NalcolyteI I I I I I I I I I • I I I I I I I I I
I·XVII.Data for Experiment 6E, N0ni0nic.Nalc01yteI —
I I I I I I I I I I I I I I I I I I I
IXVIII.Data for Experiment 6F, Cationic NalcolyteI I I I I I I I I I I I I I
AI I . I I I
IXIX.Data for Experiment 6G, Anionic Nalcolyte- I I I I I I I I I I I I I, I I I I I ' IIIn
XX. Data for Experiment 6H, Anionic Nalcolytel
~ I I I I I I I I I I I I I I I I I I I ·
XXI. Data for Experiment 6I, Anionic Nalcolyte-
I I I I I I I I I I I I I I I I I I I
IXXII.Data fer Experiment 6J, Cationic Nalcolyte~
- -I I I I I I I I I I I I I I I I I I III
7 XXIII. Data for Experiment 7, Extended Parameter I’ Experiment Using Anionic Nalco 675 .... 172
viiil
~
LIST QF FIGURES
Figure _ Page1. Flow Sheet of Sole and Heavy Leather
Vegetable Tanning Process........... 4
„ 2. Structure of Guoy—Chapman Double Layer. .... 15 °V ‘ 3. Repulsive and Attractive Energy as a Function
of Particle Separation_at Three ElectrolyteConcentrations ................' 18 ° ·
4. Variation of Potential with Distance where·
the Potential Determining Layer is Distinct_from the Solid Surface ............ 20
5. Solubility of Aluminum Hydroxide as a FunctionÜf PHO O O O O O O O O O O O O O O O O O O OO'
6. Swollen Ionic Gel in Equilibrium with · '-“ 'Electrolyte Solution ............. 35
·6
7. Flow Diagram of Auto-Analyzer for ChemicalOxygen Demand Determination.......... 69
_ · 8. Schematic Diagram of Jar Test Apparatus .... 76
9. Deterioration of Waste with Time..... . . . 86e
10. Calibration Curve for Chemical Oxygen Demand. . 93
11. Standardization Curve for Tannin_Concentration. 9512. Relationship Between Response Parameters Total
nVolatile Solids, Volatile Suspended Solids,and Turbidity as Light Scattering. ...... 97
13. Relationship Between Response Parameters Total_ Volatile Solids, Settled Volume, and Turbidity
_ ° as Light Transmission ...... . ..... 98
14. Relationship Between Response Parameters TotalVolatile Solids, Filtration Time, andTannin Reduction......... . ..... 99 ·
15. Relationship Between Response Parameters, TotalVolatile Solids and Reduction (%) ChemicalOxygen Demand................. 100
ix
, W
Figure 1 Page16. Variation of Volatile Solids with Aluminum
Sulfate Concentration, Exp. 6A, 6B, 6C...-. 10617. Variation of Volatile Solids with Aluminum
Sulfate Concentration, Experiments6D 6JI I I I I I I I I I I I I I I I I I
I18.Variation of % Reduction of Total Volatile JSolids with Aluminum Concentration for
V Polyelectrolyte Type............. 109
19. Variation of % Reduction of Total Volatile .R Solids with Polyelectrolyte Concentration,
~ Experiments 6D to 6J. . ............ 110 V20. Variation of % Reduction of Total Volatile
Solids with Poly. Conc. for Average of ·V Best Poly. Type ............... 111
V 21. Variation of % Reduction of Total Volatile ‘Solids with Time Interval for
V Experiments 6D to 6JV............. 11322. Variation of % Reduction of Total Volatile
Solids with Time Interval for Best4 Polyelectrolyte Type............. 114E
23. Variation of Volatile Solids with Aluminum _1 Sulfate Concentration for
SpecificConcentrationsof Anionic Polyelectrolyte ·V (Nalco 675), Experiment 6G.......... 115
24. .Variation of Volatile Solids with Aluminum_Sulfate Concentration for Nalco 675 ..... 119 ·
25. Variation of Volatile Solids with Poly-E
electrolyte Concentration for Nalco 675 . . . 120 F
26. Variation of Volatile Solids with TimeV Interval for Nalco 675. . .......... 122
27. Variation of Volatile Solids with pH forI I I I I I I I I I I I I I I I I ‘ .
28. Variation of Volatile Solids with Al2(S0q)3_Concentration at SpecificNalco 675 Concs., Exp. 7........... 124
~
' l
If
I. INTRODUCTION
. One of the more challenging problems confronting the
water pollution control field is presented by the tanningl
industry. Tannery wastes are varied and complex in nature. IThey are characterized by high values of biochemical oxygen
· demand, chemical oxygen demand, total.solids, suspendedI
solids, volatile solids, and color. The odor of tannery
wastes is particularly obnoxious and persists on both cloth-I Iing and equipment. In considering wastewater treatment for
tanneries, it should be recognized that the leather indus-
I try has been placed in a relatively unfavorable economic
position on the one hand by the development of the synthetic
chemical industry and on the other hand by the increased· ~emphasis on water pollution control. .Because of this
position methods for successful treatment of these wastesf
lmust be developed which can be accomplished at a minimum
cost if they are to be implemented. . I· The tannery waste treatment problem is complicated by
the fact that tanning wastes vary widely from process to
process and even from plant to plant where identical pro-_
cesses are used. These variations are due mainly to the_ I individuality of the tanner which has prompted the observa- ·
tion that tanning remains a combination of art and science._ This diversity is particularly true in tanneries using the A
I
I1
. °f
T 2 D
vegetable tanning process for the manufacture of sole
leather. The three processes presently used are vegetable,
alum and chrome tanning,i
Optimal treatment of spent chrome tan liquor is accom-
p plished by mixing this waste with„effluent from the beam
house, This process is not applicable to spent vegetableT tan liquor and initial segregation of this wastewater stream
is typical. The purpose of this investigation was to develop
a satisfactory method of chemical pretreatment of spent
vegetable tan liquor; 'This waste stream is primarily respon-
sible for the color associated with vegetable tanning wastes
and also eontributes a high percentage of the organic matter
as measured by BOD or COD in the factory effluent. An ef-
fective means of chemical pretreatment would substantiallyreduce the color and decrease the organic matter loading on
a subsequent biological process unit,.
II. LITERATURE REVIEW
Perhaps the oldest technological art of man is the pro-
cessing of skins and hides to produce leather goods and furs.By the time of literate history textiles were the primary ‘
' ‘”form of clothing, and this particular use of leather was al-
ready in descent. The leather industry was highly developedat that time, and natural tanning processes remain substan— ‘
ltially the same today as then. Early tanneries were located
on water courses, and because of the unpleasant odors associ-
_ ated with tannery operations and because of social custom _·“ the tanners congregated in their own districts. (68) Rudolfs
Wl
describes the tanning industry as an operation derived from' ·"prehistoric times with wastes of Neolithic characteristics.”
(28, 91) · E‘
‘· godern Leather Production6
The production of leather consists of 10 majorprocess-ingsteps. These steps are in process order: 1) storage
and trimming, 2) washing and soaking, 3) green fleshing,
4) unhairing and lime fleshing, 5) lime splitting, 6) bating,i 7) piekling, 8) degreasing, 9) tanning, and 10) finishing.
(106) A'’'°typical flow diagram for a vegetable tanning process .
is shown in Figure 1. (29) Steps one through five are knownW as beamhouse operations; steps six through nine are termed
tanhouse operations. Eye (31) lists characteristics and
„3 p p
-‘#_ 1
‘Wash Hides .
.... and Flesh . ————V- Soak _ _
I I,-I
. I g |I I . . I IIn--- Lime Unhair —·•>'I
_ I — . I· II I
I4--- Bate ‘ Wash """"'°)'I
. I IIntermittent „ IWastes 4---- Veget_able tan . VÄI ContinuousI I I wastes
ISole and heavy leathers _
I. |
I I I‘ Bleach · _ _I
~ 011 — _
FinishI
_ II
Figure 1. Flow Sheet of Sole and Heavy LeatherI
Vegetable Tanning Process. ‘
65
quantities of waste flows of a tannery in a summary paper.Typically, treatment of a hide converts readily putre—
scible material into leather, a tough, durable substancewhich will not putrefy for hundreds of years even whenwet. (106) All of the various steps prior to the tanning
— operation are simply preparative to this key·process.
Vegetable Tanning Process '· (V · r”
Processing prior to the tanning operation removes all _(fatty and epidermal material from the hide, leaving onlythe corium (protein collagen, C102H149N31038). (29) In thetanning process the tannin substance combines with the cori—‘
I um protein to form an insoluble resistant compound. (39, 13ä)Many naturally oceurring substances are able to accomplish
p this step, and this ability defines the word tannin. Tan-, nins are classified as being either hydrolyzable (pyrogallol
group) or condensed (catechol group). ‘Basic work byFischerandBergmann led to the synthesis of a compound similar to
Chinese nutgall tannin, a typical hydrolyzable tannin. ThisV
Nsubstance, pentadigalloylglucose, has the structure: (39,
164) — g y _
46
I |-——————l—0———lCH — CH — CH - CH — CH ··· CH2
I I I I I IO O O O OI I I I I
CO R R R ~ I R _
IH) E; Ol
R OH CO
II, ilOH
The most important hydrolyzable tannins are obtained fromchestnut wood, sumac, valonia, and myrobalan. I ‘
The condensed tannins are made up of mixtures of pro-ducts of polymerization of catechin or catechin—1ike sub- '
stances. Condensed tannins are obtained from wattle, que-
bracho, mangrowe, and hemlock barks and woods. (47) The
structure of catechin is: I °I-
~ OH IW HO O /g ·
. C' ‘. l
>(2/ H‘o¤ ~_ OH H2 IOne method of identifying the two different groups of tan-
nins is through their common reaction with ferric salts: ‘
Condensed tannins yield a green—b1ack color; hydrolyzabletannins produce a b1ue—black color. (39) A
The stoichimetry of the tanning reaction is not clear,' but there is general agreement that the most common protein
group, the keto—imide link, is able to interact and bondwith the tannin molecule. (40) In McLaughlin and Theis' (74)book a number of theories of tannin collagen interaction arelisted. However, Gustavson (41) emphasizes that "the fixa—tion of vegetable tannins by multipointal hydrogen bondingof the phenolic hydrogen atoms on some reactive groups ofthe protein is the type of mechanism . . . which is favoredat this time." l
' (In the physical tanning process it is not desirable for
a high concentration of tanning solution to come in contact.with raw hides. The rate of combination of the tannin withthe skin of the hide would be much more rapid than the rate
4
of diffusion, resulting in a permanent distortion of the’hide. (135) To prevent this occurrence the hides and tan Aliquor are moved countercurrent to one another, with freshhides coming in contact with weak tannin solutions whichhave a low tannin—nontannin ratio.- The process may takeupto
one month. ·
Leather Processing Wastes I i‘Two major wastes are produced in leather production:
U
l8
1) the.waste from the unhairing process in the beam house 3
°and 2) the spent tannin liquor. The beam house waste is
alkaline, milky colored, sulfide laden, and contains about
one—half of the plant waste volume. The spent tannin liquor
- is only about five per cent of the plant waste volume, but 'n ”
is both highly concentrated and colored. (106) The deep lbrown color is particularlylobjectionable in its inhibition° ·
of recreational water uses. In Table I the ranges of the E
tannin waste indices are listed. (3) The effects on the
. New River of wastes from leather processing plants in South-
western Virginia have been studied by two chemical engineer- „'U ing students, Sentz (93) and Johnson (48).
b
l‘ _ Treatment gf Spent Vegetable Tannin Eiquor
Because of the unique nature of the vegetable tannin
liquor, waste segregation and individual treatment is Äq desirable. (107) The summary review paper by Eye (33)
7·
lists research accomplished in the treatment of tannery( wastes. Indicated in this publication are the four treatment
methods generally used. ~ll.
_
l‘Mixing Eith Domestic Sewage
ll ·
Many researchers have investigated combined treatment(
of domestic and tannery wastes. The purpose of these in-\
vestigations has been to define the maximum tannery waste-
domestic waste_ratio which still results in satisfaetory
9
/\OO
•/\AOI er4 O N II-IO I-II I-I I-I I-IB >·! \•~ \O \ \ \4 F:} bßßl b0O hn bi) hbI-I •~ E El? Elf: E E E
Ä · P-! Il) OO! OO\ O I O N O ' IIO O Erl ON O—·I O I O • OD E 1 A O O lü I~H :> ¤¤ ooe: «-ne: In on 6A Q E NE mg NN «-IE · _ ’
hl AI
I-I I-I I-I I-I I-IA 4 \ \ \ \ \ .III 4 hD bu bi) bb bi)ga A I-I E ~ A E E B E E ‘Cr] 4* O O O O O N IO O •·I ¤‘\ u‘\ O 4* O O • I
E O v I~ O O O ux ux> IH O on In cz: ·:.4 In H cu vw
¤«I I
E E 2 .. ..OlOcnI:} ^ E I EE I-I 4 co I I I 4* IA > FN ux. I I 6 I • ID oö -« ol cx mE2 W 3 2 *2 ··2 ·°“ Ä ä S "
‘
Q •
3 ÄEI I-I I-I «-I I-IBJ ^ \ \ \ ‘ \O M bo bu ho Ian• D •· O E E E I E O II-··I · E-I Z vl I • IO v O O O O u‘\
Bl E U2 I O O 4* O1 I E Q Q Q Q
I-] O 4* Ig · Z IH N
,4 lv!-{äO I-IE4 In EI
U) Q äQ H EII-I A z ·‘ S 2 Ä 33E-nI-I In Z4 Erl A Q OZ O 6 6 ca · 6O I-! I'!) QE Ü: I-: E ä ES E EE S S 3¤‘^6z ca 6 6 I: 6. 6 ¤¤· A I-I III E-! I> VJ III O ¤« EI
10biologicaltreatment using the trickling filter or activatedsludge processes. Consensus conclusions are that hair andfleshings must be removed by pretreatment and that the maxi-mum flow of the tannery waste be no more than 30 per centof the total waste flow. (54) It is generally true, however,
·that this method of treatment is not available to manytanneries due to overloaded municipal facilities or theremote location of the tannery.
lv
Mixing wtth hlhh Tanning SolutionA
Large tanneries often have several tanning processesoperating simultaneously. Tomlinson (104) has describedhtreatment of the spent vegetable tannin waste by mixing with.the spent alum tanning waste at the Caldwell Lace LeatherCompany. In this study the volume ratio of vegetabletoalumwastes was varied from nine to one through one to one.Significant reductions in color and solids concentrationwere obtained using the one to one ratio. However, long ~settling times of up to 18 hours were required to obtaindesirable sludge densities„
l Biological Treatment
pTheprimary advantage of biological treatment of tan-nery waste is the large reduction of BOD. Eye and Aldous
p (32) studied aerobic-anerobic biological treatment in astratified lagoon, with the two zones separated by slotted
11 ‘
baffles. Feed was introduced into the anaerobic region and „withdrawn at the surface of the aerobic zone. Using an
eight day detention time and neutralized feed, BOD reductions(of85 per cent were obtained. Poor color removal was
observed. V ‘·
Parker (80, 81) has also investigated biological treat-_ A ment and reported on the kinetics of spent vegetable tannin
liquor decomposition in an aerobic system. In his latest
study Parker (81) observed BOD reductions of up to 90 per
cent. Color removals of 90 to 95 per cent were indicated,
, but color standards were not obtainable, and it is unclear
how the Values were determined. Also, very large sludge
Volumes resulted following dilution and treatment with lime
wastewaters, being equivalent to the initial Volume of the
tannin waste. . . ve Foreign researchers have also been active, particularly
in India. Tharbaraj, et al. (101) obtained greater than
90 per cent.BOD reductions using an aerobic biologicalU (
V system. Chakrabarty (18) also indicated good BOD reductions
using the activated sludge process with an acclimated seed.
Color removal, however, was not_significant.(
Chemical Treatment
The spent vegetable tannin waste is colloidal. For
this reason several chemical treatment processes have beentested. Toyoda, et al. (105) found that ferric chloride
y p12was
an effective coagulant and could be used over a large ‘
pH range. Later investigators have specifically attempted
to improve color removal which was generally lacking in(
biological processes. Tomlinson, et al. (104) used massive
_ doses of activated carbon and obtained almost no color re-
moval. They observed that color intensity was a direct
function of pH. In a more recent paper Tomlinson
(105)observedcolor removal using a long chain synthetic cationic
polyelectrolyte as a coagulant on a laboratory prepared
colloidal tannin solution. As the pH was decreased, increased
color removal was uoted. This work is of doubtful practical(
value because: l) the waste must be returned to a higher
, pH before discharge to the receiving water, and 2) results y
from a spent and unspent tannin solution may not be similar.
· Edwards (27) and King, et al. (54) have investigated
· the chemical treatment of tannin waste using alum as a
primary coagulant and a cationic polyelectrolyte as a psecondary coagulant. In this work, the quantitative nature
of the investigation was secondary to the basic purpose of
determining the feasibility of the process. These research-
ers also tested biological processes in conjunction with(
chemical treatment, observing color removal up to 80percent.
King (55) concluded that "little color removal may
be anticipated in a purely biological process." In a later
work King (57) observed that the advantage of using both
.13ß (
chemical and biological steps in a colored colloidal wastewas\that the chemical phase attacked the color problem, thebiological phase the removal of organic material.
(Qäture gf Colloids
· ' . Naturally occurring colloidal material is generallynegatively charged. Positive metal ions have long been usedas primary coagulants, and, in accordance with the Schulze-Hardy rule, multivalent ions have proven more effective thanmonovalent ions. ThuS,.ferric and aluminum ions were ob-served in early times to be efficient coagulants. Calcium,with its lower charge is less effective, but being cheaper,has also been used. In order to develop a satisfactory A
theory of coagulation an understanding of the nature ofcolloids is essential. h
1 Colloid Structure( (
_
In the colloidal state, the suspended substance will ‘
have particles in the size range from 10-7 to 10-qcm. (126)
\.
Colloidal particles move toward an electrode in an electri-cal field, indicating that the particles are charged. Col-loidal particles may exhibit either positive ornegative(charge. This movement of the particles is termed electro-phoresis. (110) As in ionie solutions, a sol (a colloidalSuspension of solid particles in a liquid) does not have anet electric charge.
1ä
Electrical Double Bayer. Colloidal charge compensation isbest explained through the model of the electrical double·layer. The double layer consists of the charged particleand an equal ionic charge formed by ions in the surroundingliquid near the surface of the particle. (121)ul
Two methods are suggested by van Olphen (110) for the "e'
acquisition of the surface charge: 1) crystal imperfectionsand 2) preferential adsorption of specific charged ions onthe colloid surface. Verwey and Overbeek (121) list only
”
the latter-reason. These explanations are apropos to a W
hydrophobic system in which the solvent is not greatly at-
tracted to the colloid surface. In lyophilic systems the” solvent molecule sheath lends added stability to the colloid.
.The double layer charges on a specific colloidal systemare, of course, repulsive in nature. While there may be aBoltzmann type of distribution, a stable system will be
p either positively or negatively charged. As one colloidalparticle approaches another, repulsive forces, due to their
I
similar charge, keep them from coagulating. This typeofcolloidstability was recognized by Guoy (37) and Chapman
(19). A diagram of the diffuse double layer as they de!
picted it is shown in Figure 2. (111) Verwey and Overbeek
(120) gave the following equation describing the repulsive
potential between double layers of spherical particles: Y
15 g
Sol interface .
FE?’(//’Counteri0ns
. 4- + + —
g4
_ — + - . Bulk of solution—- 4- é 4- _ 4- .. + .. ... — . l‘- + +
4
-y — + +.... + ·” — + -|- 4 ‘
4- + + -I·- .|. - ..
.. - + - ' ~4
· + „". . +
Figure 2. Structure of Guoy—Chapman Double Layer.
hl A ~ 16
(
Da2I VR exp E-K(r—2a)) (1)
where(
·V ·
D = dielectric constantofm
= ionic concentrationA
e = electronic charge ·r
Z = valency of the ions g _ 1 7h = Boltzmann constant _ n
eT = temperature, °K
rl l
-qfa = surface potential .
a = particle diameter
p „ r'= distance from the particle center· 1/K
L:
(om/enme2z2)%AttractiveForces Between gel Particles. In order to achievercoagulation, there must be attractive forces between sol - UPparticles. London described the weak, non—polar Van der h
_Waa1s attractive forces acting between all particles from a
quantum mechanical viewpoint in 1932. (122) Shortly there-
after Hamaker (123, 43) described both attractive and repul-
rsive forces for two particles. It might seem that the Van( der Waals forces would be insignificant and thus not have
ab appreciable effect on the stability of a colloidal system. nHowever, this attractive force between atoms is additive;
thus, for systems with large numbers of atoms per sol parti— ,
rcle, the force becomes significant. (112) A graphieal
n 7 17
2representationof repulsive and attractive energy as a
function of particle scparation is given in Figure 5. (75,
112) By summing the repulsion and attraction curves, net
attraction occurs only at very small particle separations
with high electrolyte concentration. Hamaker's (43)
equation describing sol particle attraction is:
‘ 2 2 2 ' 2 2 lln (2)
where:” a = radius of particle .’ r = distance between center of particles
Hamaker constant(nv10—12
ergs)
Following the work of Hamaker, many investigators at- 1
tempted mathematical refinements to improve his model.
(Chief among this group were oerjaguin and Landau (20) and 7 ‘
Verwey and Overbeek (120). These two pairs of researchers, V
working independently during the last World War, made simi-
lar evaluations and arrived at much the same mathematical
developments. -4
Development pi ggpä Potential--Qpgpn Mpggl '
The Guoy-Chapman model of the electrical double layer
has a serious defect in that the finite dimensions of the ‘ ·
ions in solution were assumed to be zero.- In dilute solu-( ° tions this assumption is permissible; however, it leads to '
F · _18 „_
‘ VR1 A T „
t Low
V Intermediatear 10 e ->-
_ _ Separation‘V van der Waals
V A.; Attraction _ ‘ ·
Figure 3. Repulsive and Attractive Energy as a Functionof Particle Separation at Three ElectrolyteConcentrations.
1
absurd predictions for counter ion concentration in concen-trated solutions. (124) An early attempt to create a more grealistic doub1e—layer model was·made by Stern (97) when heassumed that the position of closest approach of a counter-ion to the charged sol surface was limited by the size of,the ion. (114) In this description the counter ion chargeis separated from the surface charge by some thickness,$,,in which there exists no charge. Thus, a molecular con-denser exists between the counter-ion charge and the solsurface charge. In this region the potential drops linearly,but in the diffuse region (Guoy—Chapman) the potential drop
d is exponential. A graph showing the potential function of· a sol particle as seen by Stern is shown in Figure 4.
(1)~
_ The effect of potential causing ions is also shown. In_ this picture the Stern layer is assumed to be immobile. At
some point away from the sol particle surface and at least”
1at the limit of the Stern layer there exists a shearlayerwhererelative movement of counter-ions and the sol parti-cle exists. The potential at this shear layer is known asthe zeta potential.
Several authors (115, 125, 127) emphasize the diffi-culty from both practical and theoretical viewpoints inworking with zeta potential. For example Verwey and Over- . .beek (125) stated in 1948, "there are practically no casesstanding the test of even moderate criticism in which {
20 I
\‘
I
I I"
Potential-determining Layer
I. I I ///// ///Stern LayerI4;
II Diffuse Layer. · I
0
I Ä I I· I I?
‘ Shear Layer
I I SI I II I II II I III
'
IFigure4. Variation of Potential with DistanceWhere the Potential Determining Layer „is Distinct from the Solid Surface.
I I21·
determinations of zeta potential for one system with dif-
ferent methods give concordant results.“ Vold, (127) pub-— _lishing in 1964, took a somewhat less critical, but still I
negative Viewpoint. van Olphen (115) stated that it was
_ not surprising that a relation exists between colloidal
stability and the magnitude of the zeta potential. He then
further stated, "because of its ill—defined character, the
zeta potential is not a useful quantitative criterion ofstability.” Clearly there is a wide breach of opinion on
the usefulness of zeta potential. Colloidal theoreticians
avoid its use; practicing colloidal chemists, such as Rid-
dick, rely on it as the fundamental parameter.
.Zeta potential is calculated indirectly by measuring(the
electrophoretic Velocity and applying the following
equation developed from the work of Smoluehowski and
— Helmholtz: (22, 95, 128, 129)· · V(
_
I · v Äh x /l1TI’lL Iii ( — (3)
— where: v = electrophoretic VelocityV D = dielectric constant ‘ . I
X = applied potential gradient·
WL:=viscosity of the liquid phase
Colloidal ätabilityAll sols are unstable, and thus it is perhaps unfor- V
tunate that it is common to describe some as being stable.
"Stable" and "unstable" sols differ only in their rate of
coagulation. It is customary to describe the coagulation
v process by the term flocculation‘time (time for half of the
sol particles to disappear). Thus(coagulation is assumed
to be similar to a second order chemical reaction. (11)(The basic concepts in the kinetics offlocculationwereset forth by von Smoluchowski. (131, 132) He first
considered rapid coagulation and developed the equation for
coagulation time:(
·
T=’
where: D1 = diffusion constant(
. R = distance of interaction .„ vo = initial number of particles.
By substituting the definition of_the diffusion coefficient — T
he obtained, for room temperature:
2 XV1011 (5) (°
B oThus, Smoluchowski showed that the time of coagulation
(for a rapidly coaéulating system was inversely proportional
only to the number of sol particles. Overbeek verified the
correctness of this conclusion in a later summary text.
(58)Theapproach taken by Smoluchowski in this development was ·that every bimolecular sol collision resulted in coagulation.
23
It is not surprising then that in his effort to describeslow coagulation he added only a fractional factory 64 , ·
which converted total collisions to collisions resulting
in successful agglomeration. He obtained an expression
similarly to equation (4):
T = *1:-äävtc (6)1 o
Using this approach all coagulation-time curves should be
interchangeable by manipulation of the time scale. Regret-
tably, there was no theory relatingxwith double layer —
potential, concentration of electrolyte, or valence of „
electrolyte. ' —
Another fundamental approach for slowcoagulation,more
applicable to the double layer theory, was developed
by Fuchs (59). In this work Fuchs assumed that there was
an energy barrier to coagulation similar to the activation
of a chemical reaction. As coagulation proceeded fewer
and fewer highly energetic sol particles remained;
thus,coagulationmust slow down. In rapid coagulation this‘ energy barrier was either very low or non-existent, allow-
ing coagulation to go to completion. This concept lent
itself very well to the double layer theory of colloidal
structure. AVerwey and Overbeek (61, 120) used this approach in
reconciling the Schulze-Hardy rule with their theory and
4
24evaluationsof colloidal stability. They pointed out
graphically that addition of electrolytes lowered the po-
tential of the sol particle; the greater the positive ion
charge, the greater the lowering of the charge of the sol
particle. Verwey and Overbeek believed, as part of theirtheory, that this lowering is caused by addition of more
ions to the Stern layer and with it compaction of the double4
layer. This sol potential lowering from Fuchs' viewpoint· (with Verwey and Overbeek's agreement) caused a decrease in ‘
the energy barrier separating the colloids and resulted in
coagulation. Verwey and Overbeek (62, 120) showed that:
c = 6 x 10-22 -3-6 (7)A Z X .X where: C = flocculation value ' _ _
X = surface tensionA = van der WaalsconstantZ
= valence of ion ( h
Thus, flocculation effectiveness of monovalent, divalentand trivalent ions were predicted to be in the ratio of
100:1.6:0.13, values which are in reasonable agreement with
experimental evidence. (62) °
Mechanistic Theories of CoagulationThe accepted theory of colloid‘structure and stability
6 is termed the D-L-V-0 theory from the initials of the-
25 ·
Aauthors. (20, 120) Verwey and Overbeek also explained themechanism of colloidal coagulation, as noted in the previousparagraphs. However, there are several competing viewpoints,°overlapping explanations, and semantical differences preva-lent in the literature which need elarification.
4 Chemical Theory
Duclaux and Pauli (61), among others, proposed thatfloceulation was the result of formation of an insolublesalt or, in a modified form, as the result of formation ofa non—ionized compound in the double layer. The chemieal —
' ’theory also stated that the primary charge of colloid parti-cles arises from the ionization of complex ionogenic groupspresent on the surface of the dispersed partieles. (98)The weakness of the theory is in the similarity of coagu-lating ability of ions of the same charge.
l-° Adsorption Theory V ‘ _
_ Freundlich, et al. (36, 61) proposed that diseharge ofthe sol particle resulting from preferential adsorption ofthe floeculating ions caused coagulation. This was disprovedby experiment
evidenee.PhysicalTheory — _
y This theory was developed over a forty year period byl Guoy, Chapman, Fuchs, Stern, Hamaker, Verwey, Overbeek,
Derjaguin, and Landau and has been considered in detail. lt
y I
26
emphasizes counter—ion adsorption, reduotion of zeta po- ‘
tential, and compaction of the double layer in the coagula-tive process. ‘
Perikinetic ang Orthokinetic Flocculation-. These terms describe two different phases of coagulation I
in accordance with the physical theory. (25, 60) Peri-kinetic flocculation occurs when repulsive foroes betweensol particles can be overcome by the Brownian movement of.the particles. Perikinetic flocculation has been describedas consisting of four phases: 1) hydrolysis of coagulant,( -2)
spontaneous crystal growth, 5) prooipitation of particles,land 4) agglomeration. (55) Orthokinetic flocculation isflocculation caused by systematic movement (velocity gradi-ents). Insuffioient energy is present for Brownian movement)to overcome repulsion potential between sol particles; this
‘mixing energy must be supplied for rapid coagulation. (60)
Flocculation and Coagulation( (
· •van Olphen (109) stated that these two terms are normally
used interchangeably. LaMer (65, 64, 65, 98) has chosen todifferentiate between them. He defined coagulation as pro- lcesses which affect a reduction of the total potential energyof interaction between the double layers of two similar par-ticles. Flocculation was defined as processes which enmeshthe colloidal particles in a three dimensional floc network
by the formation of chemical bridges. While this distinction· seems to be useful, regrettably for the case of hydrolyzed
(
metallic ions, the mechanism is not entirely clear. De-lstabilization of a hydrophobic colloid by the addition of anindifferent electrolyte is clearly coagulation. Destabili-
·-zation of a negatively charged dispersion by an anionicpolyelectrolyte is an example of flocculation. W
_ As can be seen, the term perikinetic flocculation is ‘
comparable to LaMer's term coagulaticn while orthokineticflocculation is more like LaMer's term flocculation. Be-
cause of the wide use of trivalent hydrolyzed metal ions incoagulation processes this topic will be considered in moredetail.°
Stumm (98) has described colloidal destabilization oc-’ ~curring close to, but not exceeding the so—called critical
stabilization concentration, as being an area in which thehydrolyzed metal ion acted as "sweep floc." In this regionthe sol particle concentration is very low, as most has al-ready been destabilized. Floc formation is large withextension bridging. Stumm thus described this range of
coagulant floc as being comparable to the sawdust used by '
a floor sweeper to collect dust. Hence, he coined the word
sweep floc. g
‘ 28
Hydrolvzed Metal Ions° The two most important hydrolyzed metal ions are the
ferric and aluminum ions. The following conclusions have
been made from previous_w0rk concerning hydrolyzed metal
A ions. (98) 1) Hydrolyzed metal ions are generally more ’
efficient in destabilizing colloidal particles than un-
hydrolyzed ions. 2) Hydrolysis of metal ions can be inter—' ·
preted as progressive replacement of coordinated water mole-· cules by hydroxide groups. 3) Simple hydrolysis species
can condense and form polynuclear hydroxo complexes, parti-
_ cularly when the metal hydroxide is oversaturated. 4) Poly- .*’ nuclear hydrolysis products may therefore occur as soluble
Vkinetic intermediates in the transition to the precipitation
of the metal hydroxide. _- _
Aluminum Ions ~ .' The hypothetical conversion of the trivalent aluminum
ion to the aluminate ion is postulated as follows. (99)
A1(11205011]‘
n · 4+ +++°A1 (OH), -4-— Al (011)[ 8 ac] aq [ 6 15] aq{PI)OH- —(S)
Matijevic (71) agrecd with at least part of this
hypothesis as he concluded from his own work that hydrolysis
of the aluminum ion in the pH range four to seven leads to-the formation of a soluble polynuclear tetravalent aluminum
complex. He postulated the structure [Al8(OH)éä 4+ as the
most likely formula of the complex.
·Itwas known that the pH of an aluminum ion solutiondecreased with age and that aged solutions gave different
c coagulation characteristics than unaged solution. For a ’
given metal ion concentration the pH is the most important
factor determining which particular polymeric species pre-I
i dominates. .Thus, the pH of the medium is of primary impor-
tance in establishing the mean charge of thehydrolysisproducts
and of the rate of coagulation. (99) Matijevic (72)
_ substantiated this opinion in a later paper in which he(
stated that coagulation and sol particle charge reversal
are functions of metal ion concentration andpH.Figure5 shows a graph of the pH—solubility curve for
aluminum hydroxide. As can be seen in this figure, the4
(hydroxide is least soluble at about pH 6. (99) On this
graph the different species of the aluminum ion existing at
various pH levels are listed. This compilation by A. P.
Black (14) lists the range of existence of[}l8(0H)2ä +4 as H
_ from pH 4 to 6, thus, differing from Matijevic's 4 tc 7.
In studying coagulation using aluminum as the coagu- -
lating counter—ion Stumm noted that as the coagulant dosagc
H I 30
I I O
IIIIII I·¤
6A E4 +3 A1(011) ·Al6(OH)15 3(S)H H * ÜéP-4I-‘-6 +l£ I 1A Al (OH) Mater treatment
< 2 2 ‘ _ plant operation‘ region¤o · .~..1 .H‘ ‘(
+2
I -10 — 1 H I2 4 » 6 8 1O H 12
· pH H
Figure 5. Solubility of Aluminum Hydroxide as aFunction of pH.
3 1 1
increased there was always some critical value beyond whichcoagulation was produced. This value was defined as thecritical coagulation.concentration. Matijevic (69, 70) de-lscribed several methods of determining the c.c.c. value.Then, as coagulant concentration increased, increased tur-bidity, or restabilization of the coagulated sol particles,was observed. This value was defined as the critical sta~bilization concentration, c.s.c. Thus, between these two Iconcentrations coagulation was-observed. As the coagulating
11
counter-ion concentration was increased the negative chargeon the sol particle gradually decreased until coagulationwas possible. Further concentration increases of the —counter-ion eventually caused sol particle charge reversalby forced adsorption and subsequent colloidal restabiliza-
‘ tion. Matijevic (73) observed that zeta potential did not
V attain the same constant value_at the critical coagulationconcentration. Therefore, it was felt that zeta potentialalone was not sufficient criterion for stability._ Matijevicbelieved that turbidity determinations were more effective
in evaluating coagulation with the hydrolyzed aluminum ion.
T. M. Riddick (87) has studied industrial applicationsF of the aluminum ion over a long consulting career. He be-
lieved that pure aluminum oxides were of little value for x_ forming floc on account of the resulting floc fragility,
which reflected their low degree of polymerization and a
32lackof chemical or mechanical interlocking. Alum (or pure, _
hydrated aluminum sulfatc) produced hydrous aluminum oxides
which formed the basic unit of a tough floc. He believed
nthat the following steps occurred when alum was added to wa-
ter: 1) Hydrous aluminum oxides were instantly formed withV
a resulting drop in pH. 2) Polymers or gels consisting of
aligned polymers (in the raage of a few hundred angstroms)
were instantly adsorbed on the first available and acceptable
colloid. Thus, the colloid was literally coated with coagu—
_ lant. ‘5) If, at the moment of hydration formation, a colloid
was not immediately available, the polymers underwent further
buildup on other floc. 4) In order to obtain good floc—· '
colloid contact it was necessary to have violent agitation
when the alum was added. The period of rapid mixing was
critical as excessive shearing of the newly formed metallic
A polymer would result in inefficient coagulation. 5) The
smaller the size of the floc, the better was the colloid Vremoval. This latter observation is particularly cogent as
many applied colloid chemists seem to feel that "the bigger
the floc, the better the removal." Both theory and practiceh
would seem to dispute this viewpoint. ° »
Ferric long ' .°V
.I
.
The ferric is the second of the most commonly usedhydrolyzing metallic ions. Generally, it has not found as
wide use as has alum. Edwards (27) compared the use of
n
. 35 e
_a1um with the ferric ion in earlier werk on the chemieal g- treatment of tan liquor wastes and found the alum tc be
superior. Matijevic (71) has studied the hydrolysis of
the ferric ion very extensively, but due to Edwards' lackh
of success, it will not be covered in this review.
. Recalling that one of the problems with tannin waste
is color removal, it is not encouraging to note that 8tumm .
(98) and Black (10) report that many reactions between r‘ tannins and similar substances with ferric ions have . ·
yielded intensively colored solutions., Hem (45) has studied‘ both natural and synthetically prepared tannin solutions
(
„„ in their reaction with ferric ions to produce color. Benoitn
‘ (4, 5) reports that due to the stabilizing soluble complex
formed by tannin and iron in soil, iron is able to migrate(
readily in soil systems. These reports and more all point
to the inadvisability of adding ferric ion to a tannin
system. The only disagreement with this logic is presented
by Hommon (46) in which he describes the chemical treatment
.0f tannin waste with.ferrous sulphate to produee a waste‘
with the consistency of weak tea. ”In this treatment, how-
ever, the waste was diluted to seven times its initial
wlvolume. . yA (
· Stumm and 0'Melia (98) have concluded that the colloid
destabilizing ability of the ferric ion is similar to that
of the aluminum ion in that it acts by both coagulation
and flocculation and that process efficiency is dependent ,
on concentration and pH. p
Polymeric Flocculants4The incorporation of ionic substituents in a polymeric
6chain produces substances which have the properties of both
polymers and electrolytes. Flory (35} reported that the
configuration of the polyelectrolyte molecule in a dilute
solution may be magnified by electrostatic repulsion between.
charged portions of its own structure. A possible configura-
tion is shown in Figure 6. This swelling results in high
viscosity, even in dilute solutions. Other evidence of the‘ ionic nature of the polyelectrolytes is seen in the high
conductivity and heats of solution of these substanees. The
. charge on the primary chain can be either positive or nega-
tive and thus these substances are designated as cationic or
anionic polyelectrolytes,.respectively.I
p
Cationic Polyelectrolytes In
O TWhen cationic polymers come into contact with a polar
solvent the inorganic anion dissociates. The concentration
gradient results in diffusion or migration of the anion VV
away from the polyelectrolyte chain. The result of the
anion migration is that the polymer chain becomes positive-T
ly charged, thus making further anion migration more diffi-
cult. Steady state is eventually reached. lnorganic cation
I35
U_
I I Q Q Q I . I D E
9Ä I _» QI . ~. E
I I9 I‘
Q) I <—-> SolventIe
I<>eI IQ Q I·<——> QB
M G): Anion —G) G) F I IQ QI · „G)= canon I· @ » I Q.}: Fixed Cge. ·Figure 6. Swollen Ionic Gel in Equilibrium with
Electrolyte Solution. ‘
Acounter-m1grat1¤n will only oceur if this ion is introduced
from extra system sources. Thus, the very high osmotic
pressure following ionization will result, at equilibrium,
in a highly charged polyelectrolyte. An example of al
eationie polyelectrolyte would be a substituted amino group
or an organic salt. Flory (34) gives the following example.
„\\
p‘ N N
~+ éu +\\Bu _
’p
lBr- Br- ·
In a eoagulation system the positively charged polyeleetro—
ilyte would be used to destabilize negatively charged eolloidal
sol particles. T U V l ° .( T
Anionie Polyeleetrolytes . _
The ionization and equilibration of the anionie poly—Q
V electrolyte acts similarly to the eationie polyeleetrolyte,
excepting the reversal of eharges._ The anionie p0lyeleetro—— lyte may be used to destabilize a positively charged col-
loidal system. An example of this type of polyelectrolyte .
is illustrated. (35) pA
\
37
COO- COO-.( A
ANa+
Na+
Nonionic PolymereA
Nonionic polymers are macromoleoules with different xApolar functional groups distributed along the polymer chainA
resulting in no net charge on the polymer. (118) Examples
_ are polysaccharides and synthetic polyoxyethylene. At
present little_is understood about the mechanism of inter-
action between sol particles and the polymer. It is known
that they are absorbed on the sol particle surface, but how
the adsorption affects ion distribution in the Stern andA Gouy layers is not known. (118)
A °
AFlocculation Mechanisms gf Polyelectrolytes
A
_ One of the advantages of polyelectrolytes is that they
are effective in very low concentration. Typically, in- ·
dustrial recommendations are a maximum of five milligrams
per liter. (76) The reason for this was given by van Olphen. '
(116) If the polyelectrolyte content of a hydrophobic sol
ranges from a few tenths to several percent, the salt
tolerance of the hydrophobic soli is increased. The opposite-
ly charged salt ions are attracted to the polymer "membrane"
and greater stability is given to the sol particles. van
Olphen (116) also suggested that "the polyelectrolytes
, T
envelop the hydrophobie sol particles and lend them the »° special characteristics of the hydrophilic sols, i.e.,
their stability and high salt resistan0e." This protective
action is independent of the sign of the charges of the
. hydrophobic and hydrophilic colloid. When the concentration‘ of the polyelectrolyte is very low, however, the sol is
apparently sensitized towards flocculation by addition of ' —
salt particles. van Olphen (117) suggests that the polyion
is now adsorbed on several different sol particles and inthis manner forms a series of interlocking bridges, thus
y prompting flocculation. He·further states that in the ab- rsence of salt, polyion adsorption does not occur. "A cer-
°.tain amount of salt must be present to establish particlei
(bridging by simultaneous adsorption on different sol parti- _
licles." (117) This viewpoint is reinforced by the reali-
_ zation that salt addition reduces interparticle distanceand increases the possibility of dual adsorption. van ·
Olphen states that the addition of the polyelectrolyte
before the salt increases the possibility that the poly-
electrolyte ions will become adsorbed on sol particles.
p ·This theory is particularly interesting in light of the
fact that work published by practicing chemists and engi-lneering consultants has recommended the reverse order. (8, ‘
26, 27, 89)
\·39 R
In more recent work Walles (193) postulated three ·° separate mechanisms for eoagulation by organic water-soluble
polymers. The first method was termed charge effect neutra-“ lization in which the effective charge of the sol particleR
. was reduced by double layer reduction, In this method, 'R
‘ flocculent molecular weight has little bearing. Polymer
bridging was the second method. The flocculent polyelectro- ·
lyte interferes with the free movement of the particle,. '
either by adsorptionofla
number of particles on a polymer
network or by formation of a long bridge between two parti— pcles. Flocculent molecular weight is of critical impor— _
“' ° tance. The third mechanism postulated was mutual dehydration.R ”
Hydrophilic colloids or true solutions are flocculated by
non—ionic polymers. Chemical reaction or complex formationW
results in a non-soluble product with flocculent molecular
V weight being important,. . —
Ries (90) agreed with the first two mechanisms postu—R
·
lated by Walles, and he indieated that the determination of
their relative importance is highly controversial and proba-bly not possible at this time.
A R
'Applied Polyelectrolyte Research l
Birkner and Edzwald (8) have studied the treatment of
raw water with both alum and polyelectrolytes to remove col-
loidal material. They concluded that these coagulants were
successful, but stressed the idea that costs were high.
40
The concentrations recommended were very low, 1.5 mg/l .l ° polyelectrolyte and 20 mg/l alum. In a later work these
same authors (7) studied the destabilization of a clay col-’ loidal system from an equilibrium viewpoint. Adsorption of
_ the long chain polyelectrolyte was assumed, and Langmuir ·{ ( -adsorption graphs were prepared using various counter-ions.
Superior polymer adsorption (and subsequent colloidal de- ·
stabilization) was observed when the monovalent counter-ion
was present. It was also found that there was a critical
polymer concentration either above or below which the degree(
of destabilization decreased.l'«”
r Probably the two most prolific authors on practical’(
application of chemical coagulation of colloidal material
and the influence of polyelectrolytes are T. M. Riddickl_ and A. P. Black. T. M. Riddick described in·a series of p
_. —articles and a book (85_- S9) his increasing involvement .H .‘ with the basic control factors involving colloidal stabili-
U~
’ · ty. In an early article_(85) he stated that color and oh-
jectionable taste could only be effectively removed in—a U. water treatment plant by the use of a cationic polyelectro-
l
_ lyte. ”In a sequential article (86) he argued that the most
effective method of determining colloidal destabilization
* is by measuring the zeta potential. This concept was cer-
tainly not novel as Burton (17) had noted the necessity of‘
zero zeta potential in working with gold sols in 1906 and
41
1909. Riddick suggested that colloidal destabilization can
· be accomplished more completely by simultaneous addition of
an inorganic coagulant and a polyelectrolyte. Prior to this
time coagulation was normally accomplished by a single coagu-·
_ lant with the quantities being determined by jar testing. »j V Jar testing involved visual estimation of the best floccu— 4
·lation quality for various coagulant dosages. Riddick — _stated that good flocculation producing dosages were still »
too electronegative to obtain effective colloid removal.
Thus, the visual observation of a floc would tell little· concerning the key stability parameter——the colloidal sur-
„. face charge. _Proper coagulant dosages not only cause floc”
formulation but also reduce surface colloid charge to neare _ zero. ·At these conditions gentle agitation will cause col-
~loidal collisions; and, there being no charge repulsion,
h coagulation will then occur. p‘ Riddick listed three stages of flocculation: 1) rapid
j.
‘ “mixing and microfloc formation, 2) colloid gathering, and
3) agglomeration. The degree of agitation was noted to be
important. There must be sufficient Velocity gradients to
- cause bumping, but low enough Velocity levels to prevent
lfloc destruction. Riddick suggested that thick wire screens
serve more efficiently than solid paddles in producing the
proper Velocity gradient conditions. For poorly settling
flocs he recommended the addition of weighing agents such
as barium sulfate or calcite. ‘ « l
4 42 '
A. P. Black (16) has also reported on the use of zeta°potential as the primary control parameter in removal of '
colloidal organic color. In this work he used alum aloneas a coagulating agent, but stressed the importance of zero
_ zeta potential rather than floc appearance. In later ef- 4
·- forts Black, et al. (11) described the effectiveness of or-ganic polyelectrolytes in destabilizing very dilute colloidal.
clay suspensions. He also found that a log-log relationship 4
pexisted between the suspended concentration of silica andpolyelectrolyte concentration. In these investigations
~ there was either a need or an attempt to measure coagulation _~~ by some numerical, indirect variable. °
lIn other work Black, et al. (12) investigated the floc-
culation of kaolinite by an anionic polyelectrolyte. Thekaolinite was negatively charged and it was necessary to
l. add calcium to reduce interparticle repulsion before floccu-lation occurred. This work was particularly interesting as
(2
I negatively charged polyelectrolytes were used to coagulateinitially negatively charged colloidal particles. ·
‘ 2 Teot and Daniels (100), in a paper that was published
p after the experimental work of this research was completed,made the following conclusions. Anionic sewagecolloidswere
not significantly flocculated by anionic polyelectro-
lytes in the absence of multivalent cations. The cationalone, in sufficient.concentration, will coagulate the
43.
Vsewage colloid. However, in the presence of an anionic V
-polyelectrolyte a lower concentration of multivalent cationwill accomplish the same degree of coagulation. Teot in
Vthis same investigation tested the order of adding cation
VV and polyelectrolyte and found that the order was dependent 7
(-V .. on the system being tested. When a synthetic colloidal_ system (latex) was used, the order was either unimportant - _
or slightly favored the addition of the po1yelectrelyteV ~
V first. In a natural colloidal system (sewage), however,there was a definite advantage in adding the cation first.Teot stated that it was not possible to prove or disprove _
_. any particular mechanism of flocculatidn with informationpresently available.
V~
In his most recent published work, Birkner (7) hasV
studied flocculation of colloidal clay suspensions with
nonionic polymers. Again it was determined that combinedV[use of cations and polymers provided superior performance V
.to either coagulant alone. He further noted that there wassome optimum polymer concentration resulting in minimum.colloidal turbidity. 0n clay sols of 60 mg/l, the optimum
coagulating polymer concentration was less than one mg/l.V VDixon, et al. (21) also noted this restabilizing capa-
bility of the polymers, but noted a limiting value of 10
mg/1 polymer concentration when working with bacterial
systems. Successful flocculation of bacterial systems of
44 (
concentrations ranging from 50 to 2700 mg/1 was obtained ·
°with cationic polyclectrolytes. VLaMer and Smellie (66) have successfully flocculated
‘
negatively charged clay-type particles using a negatively
. charged polyelectrolyte. These investigators emphasized '( ' that they achieved bettcr results with the polyelectrolyte
alone than with a metallic cation. They pointed out that” r
zeta potential as a measure of colloidal destabilisation ‘
would not have been appropriate in this work.
Independent Variables Affecting Colloidal Stability .·" There are many variables affecting coagulation effi-
4ciency. The purpose of this section is to briefly discuss
these factors. In accordance with an evaluation of theliterature reviewed, they are described in an approximate
_ order of importance. „ „ .
Inorganic Ipp IU
.·
Clearly, the most important variable affecting colloidal
stability is the inorganic ion. As emphasized in this re-
view, for the negatively charged sol, a multivalent ion ismost effective in destabilization. ‘Thus, charge, type, and
i concentration of the ion all affect sol stability.
, The structure and charge of hydrolyzable metallic
5
cations is dependent on the pH. Therefore, sol stability ~° will be affected by pH. Stumm and Morgan (98) recommended
that pH be held constant in any systematic coagulation in-
vestigation. These investigators believe that the best pro-A. cedure for coagulation testing makes use of an automatic ‘
Utitrator, using sodium hydroxide as the titrant. It was
pointed out that addition of a hydrolyzable cation decreases ~
the pH in its reaction with available alkalinity. .Riddick
_ (89) also supported this viewpoint, but suggested that
potassium hydroxide will have less effect on sol stability
than sodium hydroxide, therefore providing more constant _‘” conditions for varying cation concentrations.
l
There is some disagreement among investigators as to
the best pH range for alum coagulation, but the differences
are small.- Black and Willems (16) recommended a range of.
_ pH from 5.2 to 6.7 as being satisfactory, but later Black
and Chen (14) specified a pH of 5.6. Rebhun (82) found ·
that a pH range of 5.6 to 6.2 produced optimum flocculation.
LaMer and Smellie (66) noted that the structure (coiled
or threadlike) of polyelectrolytes was a function of pH as
_ gobserved by the change in viscosity with pH. They remarked
that it was not surprising that pH thus affects molecular
bridging and the flocculation power of a polyelectrolyte.
Polymer-· ·
The charge, or lack of it, and concentratimaof the
. [*6
polymer affect colloidal stability. More difficult to evalu-ate perhaps is the synergistic effect of the_different possi-
[ ble combinations of metallic ion'and polymer.
’Time Bag Between Inorganic lon_and Polymer
‘ ‘ There is obvious disagreement among previous investi-
a gators on the time variable. This is emphasized from two
viewpoints. First, there is no general agreement ontheorderof addition of the polymer and the inorganic ion. ·
Secondly, no specific time lag is normally stated in the
literature. There is either a simple statement on the orderof addition or an indefinite recommendation of ”several .
_ minutes" or "some time after." (8, 23) °
äalt Concentration of Colloid System ·B — As noted by both van Olphen (1Qe) and Riddick (89), the
salt concentration of a colloidal system affects the double4
layer structure and potential. Any dilution with distilledl
water of a colloidal system in order to achieve some
commoninitialsolids concentration causes some error. This isbecause the salt concentration and the colloid-salt concen-tration ratio changes. The only proper dilution must be
l
accomplished with the salt solution separated centrifugally
from the colloid system.)
y - I
**7Methodgi Mixing »Jar testing procedure is discussed in detail in the
procedure section, but it would seem reasonable that the
manner of contacting the colloid system with the flocculent
. will be of considerable importance in resulting destabili-
zation. LaMer and Smellie (66) have stated that in order
to obtain reproducible flocculation it was necessary to
carry detailed systematic studies of the effects of order ·
of mixing and the intensity and length ofagitation.ggg
gi Hydrolyzed Metallic ggg_(Riddickhas noted repeatedly in his papers the impor-
_ tance of contacting fresh alum with the colloid system.(
Many investigators apparently prepare cation—water solutions
which are used for indefinite periods. Floc formation in·
the absence of the colloid system to be treated is less
(efficient in the destabilization process.( ‘ ”
Temperature ( . d· Temperature should be held constant to insure repro-ducibility of results. I
’De endent Variables Qggg gg Measure CoagulationMany different variables have been used in the litera— -
ture to characterize and evaluate coagulatien processes.
They are listed in the order of frequency encountered in "
the literature. d
p A8 '
Turbidity l U.
Early work by Smoluchowski lcd to the use of optical
density as a function of time to measure the rate of floc-culation. It was theorized that as flocculation proceeded
V
the number of particles would decrease, less light would_be scattered or absorbed, and the optical density would
decrease. Dole (22) has pointed out that this basic con-
cept has severe limitations. 1Flocs form in complex shapessand light scattering properties vary with floc size.Also,the
flocculating system has a memory of previous or prior
.shear history. Thus, two systems with previous differing
lshear histories but with otherwise identical characteristicsdo not produce the same turbidity results even though given
the same coagulation treatment. For these reasons, investi-gators accept the fact that it is unlikely that they will
A be able to reproduce one another's turbidity results. How-
ever, for a given test sequence turbidity data is easily
obtainable and readily analyzed. ( _l
Turbidity is an optical property of a sample which
causes light to be scattered and absorbed rather thantransmitted in straight lines through the sample. Two ‘
types of turbidity measurements are presently used and are
termed nephelometry and absorptometry. In nephelometry‘
the light scattered by colloidal particles at right anglesto the incident light is measured. The advantages of this r
_ type of turbidity measurement is that it is sensitive to very ,
low colloidal concentrations and does not record color as
p turbidity. (42) The principal disadvantages are that the
turbidity response curve is bell—shaped, giving the same tur-
bidity reading for two different concentrations, and thati
data are affected by stray light._ The reason for the former
phenomenom is that at low colloidal concentrations, there is
little light scattering due to the-low number of colloid
particles. At higher colloidal concentrations there is(
much greater light absorption, and, consequently, mueh less
light available to be scattered. Less error is involved in P
using the rising portion of the bell—shaped curve. ° V
’ Absorptometry is the measurement of light passed through
colloidal sample in straight lines parallel to incident l
.light. This type of measurement is identical to spectro—
phometric methods, The principal_advantage of this type of °
turbidity reading is that there is no upper limit to colloid
concentration as with nephelometry. The main disadvantages
are that this measurement is_insensitive to low turbidityP
”and that color is measured as turbidity.
( ”Zeta Potential
Ul ( _·
Pespite sharp controversy as to its relevance to floc— c
culation either zeta potential or electrophoretie mobilityI
is consistently-reported in the literature. Zeta potential Vis as elose·as a researcher can get to a direct measurement —
i50 (
of colloidal stability. All other measurements are in- .
direct and therefore subject to other independent variablesnot related to coagulation. Unfortunately, electrophoresis
'equipment is expensive, and data are difficult to obtain.
~ Filtration ang Refiltration TimeLaMer in many articles (64, 77, 94) has put forth the
concept that coagulation is only successful when it has pro-
duced a precipitate which was easily settleable and filter-
able. He believed that cross linking and hydrolysis in-
creased as filtration time decreased. However, during fil-
(tration the filter cake developed, thus changing cake porosi-
ty and thickness, both filtration time variables. Thus, thebest method is to determine the time of refiltration of the
filtrate through the already formed filter cake. It has
been determined that the minimum refiltration time falls‘
within the limits of the critical coagulatien concentration
and the critical stabilization concentration. Only when re- a
filtration is too long is the less accurate filtration timerecommended. p · ’
Settling gates . lh
‘ . The use of settling rates to measure coagulation has7
as its basis the same viewpoint that led to the developmentof refiltration time, namely, that coagulation is successfulonly when the precipitate is easily removed. Basically the
settling rate is determined by following the change of the -
height of the solid—liquid interface with time. (24) Graphi-
cal results show zones of settling, transition, and com-
pression. (24)e_i'·
Colloidal Titration UKawamura, et al., (48 - 52) have described a method of
determining null zeta potential using a titration technique. 4
Oppositely charged oolloids are used to neutralize the e
charge on the sample colloid. The reaction is reversible,
allowing back titration. A major advantage of this method
is elimination of expensive electrophoretic equipment. _
l Specific ResistivityI U 4
‘ An involved graphical calculation is necessary to ob-
.tain this parameter describing the resistance of a filter
cake to liquid flow. (2) ·This parameter has to some extent
been replaced by filtration and refiltration time.n
Solids Concentrations e· ’
Any number of different parameters deseribing reductions
of solids concentrations in the clarified phase as compared
to the original colloidal suspension are possible. Examples
include total solids, volatile suspended solids, and selected
chemical component concentrations. .«
52
SummaryT
„‘ i Destabilizing chemical treatment of colloidal systems
u'
has been investigated with increasing thoroughness in recent
years.“ This type of treatment appears applicable to thei
_ colloidal spent vegetable tannin waste, and indeed qualita— ·T · tive and limited quantitative work has been completed.
There remain many unanswered problems concerning the criti—· 5
cal, independent parameters affecting destabilization, the
coagulant—tannin waste ratio, the best a1uminum—polyelectro—
lyte system, and the best response to measure colloidalu
coneentration. This investigation will attack these _W~# problems. T
5 1
IlI.‘ MATERIALS AND METHODS
The following sections contain information related to
Uexperimental materials, equipment, and methods.
·Experimental Materials ‘ °
' The following paragraphs contain a list of the materials
used in this investigation.(
1 •
Aluminum Sulfate. (Al2(S04)3•18H20) crystal, Fisher
certified reagent, (Fisher). Used as a primary coagulant._ Anionic Polyelectrolytes. Nalcolyte 672, 674, 675 coagu-
lant polymer (Nalco). Used as a coagulant aid.6
Buffer Salt, pH 4.01. ‘(Potassium Biphthalate), dry,
Fisher certified reagent, (Fisher). Used in preparing con-
stant pH solution for standardizing pH meter.
A Buffer Salt, pH 6.86. (Potassium Phosphate—Monobasic
Disodium Phosphate), dry, Fisher certified reagent, (Fisher).
Used in preparing constant pH solution for standardizing pH
meter. ·
Buffer Salt, pH 10.4.. (Tris (Hydroxymethyl) Amino
Methane), dry, Fisher certified reagent, (Fisher). Used in
preparing constant pH solution for standardizing pH meter.
_Cationic Polyelectrolytes. Nalcolyte 605, 605, 607,
610, 652, 654 polymer (Nalco). Used as a coagulant aid.1 1 53 . g
Mercuric Sulfate. (HgSO4), Fisher certified reagent,. .' ’ (Fisher). Used to eliminate interferences due to chlorides
in the tannin waste in the Technicon Auto—Analyzer Chemical
Oxygen Demand analysis._
_ Mercury, Metal. Distilled, technical grade, technicali · reagent, (Fisher). Used as working fluid in U-tube manometer.
·Molybdophosphoric Acid. (20Mo03·2H3P04•48H2O), powder, .
analytical grade, (Fisher). Used as a component in tannin
reagent._
°;
V Nonionic Polymer. Nalcolyte 670 coagulant polymer,(
(Nalco). Used as a coagulant aid. ° _«~ Phosphoric Acid. (HBPO4), 85%, liquid, (Fisher). Used _
to prepare_tannin reagent.‘ Potassium Dichromate. (K2Cr207), fine crystal, ACS
(certified reagent, (Fisher). Used as oxidant in the Techni-
U con Auto—Analyzer C.0.D._analysis. v
Potassium Hydroxide. (KOH), pellets, ACS certified(
i
reagent, (Fisher). Used to control ph in jar testing.
Silver Sulfate. (Ag2S04), powder, ACS certified reagent,
·(Fisher). Used as catalyst in oxidation reaction in the
j \Technicon·Auto—Analyzer C.0.D. analysis.·
Sodium Carbonate. (Na2CO3 and Na2C03·10H2O), anhydrous(
and crystal, respectively, ACS certified reagents. Used in‘ the tannin determination analysis. _
1
(i 55.
t _ Sodium Tungstate Dihydrate. (Na2W04·2H20), ACS certi- 1
fied reagent, (Fisher). Used to prepare tannin reagent.
Spent Vegetable Tannin Waste.- Mixed protein, tanninY
composition, three per cent eolloidal suspension-solution.
_ Obtained from Leas—McVitty Corporation, Pearisburg, Va.
Used as the eolloidal waste material.(
_
Sulfamic Acid. (NHQSOZH), Fisher certified·reagent,
(Fisher).. Used to eliminate interference due to nitrates
in the Technicon Auto—Analyzer C.0.D. analysis.4
U Sulfuric Acid. (HZSOQ), Technical grade, 66 Baume,
(Fisher). Used to digest waste tannin liquor in Technicon
Auto-Analyzer C.0.D.analysis._
— Experimental Equipment‘
·· The following paragraphs contain a list of the equip-
ment used in this investigation. - .n~
Auto-Analyzer, Technicon. (Technicon). Used to deter-
mine chemical oxygen demand.( A-
Analytical Balanee. Accuracy 0.0001 grams, (Sartorius),
used for determining weights accurately. ‘
Barrel, Plastic. (Thirty gallon (obtained from Chemical
Engineering Dept., V.P.I.). Used to store waste in chemistryt
cold room. ' ·(
Beakers. 50, 100, 250, 500 ml, Pyrex brand glass
(Corning). Used in various laboratory tests.l
o ,,56 O
Bottles. Plastic, 20 liter, (Nalge). Used to store
tannin waste. . j- _
I Burettes. 25, 50 ml, (Fisher). Used for preparing
exact concentration and for titrations.
Condensers. Water, 24/40 ground glass joint, Pyrex
brand glass, (Corning). Used to prepare tannin reagent.
Dessicators. (Fisher). Used to cool gravimetricI
samples under dry conditions following oven and furnace
heating.,
Evaporating Dishes. Coors Porcelain, capacity 90 ml,‘ (Fisher). Used in gravimetric testing.
Flasks, Boiling. One liter, 24/40 ground glass joint,
Pyrex brand glass, (Corning). Used to prepare tannin
reagent..”
Flasks, Volumetric. 100, 250, 500, 1000 ml, (Fisher).
. Used to prepare exact solution concentrations in standardi-
zation and tannin analysis.W
Filtration Apparatus. 300 ml funnel with ground glass
· or teflon sealing surface, spring action holding clamp,
ground glass or teflon coated base with neoprene stopper, °
and one liter filtering flask, (Millipore). Used in analy-
tical tests to determine suspended solids, filtration times,
and refiltration times. .
Filter. Cellulose and fiber glass, 47 mm diameter,
(Millipore). Used as filters in all filtration work.
„A
* 57
l ‘ Furnace, Electric Muffle. Six hundred degree centigrade
temperature, (Precision). Used to volatilize all organic
material to determine volatile and fixed solids concentrations.
Graduated Cylinders. 25 and 5O ml, Pyrex brand glass,{
(Corning). Used for liquid volume measurement and for fil-
tration rate determinations. ‘
Graduated Cylinders.! One liter, Pyrex brand glass,
(Corning). Used for subsidence testing.n
· .
'Jars, Square. Double strength 1/8 inch plate glass,4
dimensions approximately 4 x 4 x 55inches, volume about1.25,(
liters, made by author. Used as container in jar testing.”· Mixer, Electric. One horsepower, (Lightning). Used to
‘ mix stored waste to insure homogeneity of stored colloidal
waste samples. U ._ Mixer, Magnetic. Capable of heating, (Corning). Used(to prepare solutions. ° ·
,·i, y
U
Nephelometer. (Hach). Used to determine turbidity by
light scattering. wE
_
Oven, Drying. Adjustable to constant 102°C, (Precision).S
Used to vaporize water and volatile organies.E
pH Meter. (Corning). Used to measure pH during jar
_ testing.WPipettes.
1, 2, 5, 10, 5O ml capacity, Pyrex brand, _
(Corning). Used to obtain samples during jar testing.
Pump, Vacuum. Welch Dist-0—Pump with motor, single° stage, Vented exhaust, Vacuum limit 15 microns, (Welch).
E Used to control pressure drop during filtration.(
Fipet Filler. (Fisher). Used to fill pipettes.
Shaker. Laboratory bottle shaker, 115 V, 60 cycles,5 ( (Eberbach). Used to prepare solutions of polyelectrolytes.
Spectrophotometer. Spec420, (Bausch and Lomb). Used ‘ .
in tannin and turbidity tests. · _ · .
Spectrophotometer. Model B—4, (Beckman). Used
ininitialtannin analysis experimentation._)
Steam Bath. Eight Variable sized openings, stainless _~‘ steel, electrically heated, 220 Volts, 2000 watts, (Fisher). 1
‘Used to evaporate most of water before placing evaporating
dishes in drying oven. _
Stopwatch. Graduated in 1/5 second, jeweled movement, ·
, non—magnetic, (Fisher). _Used to measure timed operations
and analyses. a · .
Stirrer, Multiple. Water analysis, 6—unit, complete
with stirrer floc illuminator and base, (Phipps and Bird). n
Used to produce low Velocity gradients in jar testing.
, Stirrer, Variable Speed. .0ne—tenth horsepower, 110 V j. A.C., (Fisher). Used to obtain Various Velocity gradients
during rapid mixing period of jar testing.
Timer—TimeswiQgh. (Eight inch dial, (Fisher). Used
to measure operations where less accuracy was required.
$9 ‘
_ Weighing Dishes. Aluminum, weight about 1.5 grams,i ·
60 mm diameter, (Fisher). Used for weighing on analytical ·
balance. °
·hi _ Experimental Methods
The overall purpose of this investigation was to develop
a chemical method for treating spent vegetable tannin liquor.
To accomplish this objective the effects of selected inde-
j pendent variables on the stability of the colloidal wasteI
material were quantitatively analyzed. A discussion of the
major experiments involved in the research is given in this E
section. U
Experiment One: Characterization of Waste I
Initial laboratory work centered around characterization
g and measurement of variation of the spent vegetable tannin 9
liquor. Using total solids, total volatile solids, and9 total fixed solids as the basic responses, fresh batches of
tannin waste were obtained at the same time daily for a ten
day period and then analyzed. Duplicate or triplicate sam-
ples, depending on laboratory convenience, were taken and
the analytical work listed above was performed starting at
the same time daily. All solids tests were run in accordance
with Standard Methods (96) with minor modifications.·
”60 ‘ )
Experiment Two: Deterioration of Tannin Waste and StoringProcedures
Selected batches of waste were tested beyond the firstday in order to determine the degree of deterioration as
_the waste aged. The same test responses used above werel
performed daily for a seven day period. The first testseries was made using room temperature_storage facilities.6A second series of tests was performed while storing the
l
. waste at twenty degrees centigrade.; A third and finalseries of tests was completed while using the cold roomstorage facilities in the chemistry building, in which thetemperature is maintained near zero degrees centigrade.
T Also, mixing equipment and procedures were tested toinsure waste homogeneity when sampling stored tannin waste.
~ The final method chosen resulted in excess of one hundredgallons of tannin waste being stored in the cold room inl
—the Davidson Hall, V.P.I. A lighting mixer stirred thevwastes in the 30 gallon tank for thirty minutes prior to
6
siphoning a two day supply of twenty liters. After removalof the 20 liters from the tank, this quantity was immediatelyreplaced from smaller containers located in the cold room. 4 pIn this manner tannin waste removed bi—daily was assured to
6
remain uniform. · 4A
61 *
Selection of Independent Variables tp_pe Tested iAs seen in the literature review section on variables
affecting colloidal stability, there were a large number offactors which had a possible effect on destabilizing colloids.From Edwards' (27) prior research, aluminum was chosen as‘the metallic cation to be used as a primary coagulant. Thetest variable was alum concentration. _
° Current research had indicated the suitability of usingpolyelectrolytes in conjunction with the aluminum. From
· *literature surveyed prior to the beginning of experimental·work, it appeared that cationic polyelectrolytes would bemost suitable for destabilizing a negatively charged colloidalsystem. It was tentatively decided to test ten differentpolyelectrolytes. And, in order to consider all charge
— types, six cationic, three anionic, and one nonionic poly-electrolytes were selected. Two test variables were, there-fore, introduced, charge type (qualitative) and polyelectro-lyte concentration (quantitative). ”
u _· ~ The pH of a solution affects the efficiency of a coagu-·lation process. However, because of the necessity to limitthe number of independent variables and because pH and ef-fective aluminum ion concentration are interrelated, it wasdecided to hold pH constant at 6.0 during first stage poly-electrolyte testing. In the final, expanded testing a pHrange from 5.0 to 7.0 was tested, covering the most effectivepH range for aluminum.
62 ‘
The literature showed that researchers were least con- .
‘fident in their procedure in describing the best time lag
between aluminum and polyelectrolyte addition. This was
the final independent variable chosen, and the remaining0
0. possible variables were either maintained constant (e.g., ‘ ‘
0 ‘l jar test procedure, concentration of waste) or assumed con-
stant (e.g., quality of waste,0temperature)._ ‘ .
Experimental Design0
j0 0
It was decided to test each of the polyelectrolyte _
types separately, thus necessitating~ten.separate tests in-
Ä volving the three remaining independent variables, aluminum0 0
. concentration, polyelectrolyte'concentration, and time lag.
It was felt that a minimum of five levels across the range‘ ° of each variable was desirable; thus, a total of five cubed,
or 125, tests would be required if every possible variable0
combination were performed. As each test had from two to s
ten separate response variables (dependent variables) the0
number of tests involved would require reduction of the
number of polyelectrolytes tested. (If each of the ten0
polyelectrolytes had ten responses, the total number of ·
0tests would have been 12,500, an impossibly high number). i0
For these reasons, Dr. R. H. Myers, Department of
. Statistics, suggested an experimental design which would
reduce the number of tests from 125 to 20. The data ob—
tained from these tests could then be used to produce a
regression equation using the BMD regression program at «' the V.P.I. computing center. Graphical interpretation would
6 then be possible. ‘ ‘
The value range of_the variables was more difficult toVn select. For aluminum the range of zero to 3000 mg/l was 3
E ‘ chosen initially, but preliminary testing indicated that a
range of zero to 6000 mg/1 was preferable, allowing maximum‘ .
colloid destabilization to be obtained. The maximum recom-mended concentration of some polyelectrolytes was fifty
mg/1. So that the time lag variable would always have a
real value the minimum polyelectrolyte concentration chosen _~“ was 2 mg/1, resulting in a range of two to 50 mg/1. The
4range for the time lag between the aluminum and the poly-
electrolyte was minus 0.811 to plus 3.189 minutes. This—— range allowed the polyelectrolyte to be added before, at
T
p the same time, and following the aluminum addition. Afterthe range was chosen for each variable, the levels were
6
determined for the particular experimental design, and were
assigned code values. This coding and the form of the»
resulting regression equation is illustrated in Table II.6
_ Table III gives the experimental design of the 20 component
_ °l.sub-tests for the testing of a given polyelectrolyte.
At the time experimental work was begun and completed
Vthere was no reported use of regression equation analysis‘ to predict coagulant dosages. Recently, however, Otts and
64 4
E .cb 2;;Q: s-I
E .4 C N cz;Ä
' I I' V
%Ü]Q · . no 1~O CN] g. «-·| 1-I • QO X I _cwI «-1
1-I · w-I «-IM N 4 4‘ *"I ‘ N . .ä - 1-I
4 CDE-I · · O O \O N+ . O N •
Q ‘ Q -1E
N . _ ' FNvw .
P4 ‘ —{I) FN *Ü] ' FN+-] · CD 'Q, 4 4 1-1 cr vw 4*
,<I + + CD • •I-I · ‘ 5 O NXFN - 4* 4* -:> vw
~ mE, .Z +Ü]Q ‘ 4 N N _Z N CD O. O N\D O UN •9-1 N • O FNIE:] N 1-I \O _Q ' CG +Z,
. I-I · ‘+
III N · °— E.; . M _ .F-I - N •• · rg
· FN U} In · Oz + X cb m +-1, -60 cu +-1 0 ‘\_ -6I-I N X Q r-I bi)•~E-•
«-I · FN cd ¤ E • GJ .<1 ><: cw: ·-1 6 · o 6+-*»D w—I CQ $-1 -+-4
•~ Q +9:50* «-• cvs. m 1-1 6 :>»¤M Q + > ¤ N O +-1-+-1
O OEZ + FN +9 E •~ O $-1•• Q-{ I
¤•p¤{ • gp +) A
I-I _Q v-I X O Q O >~„ O •U') _O X FN 'O ° Q _ +-—I OE _W ·H 1-I w-I Q •~ O O 1--I5L1-] +9 CQ Q2 O O O Oe-ICÜ1 _f-U ~ B +°\ >><Äcb Z5 + + O3 +-I E Obi) +-4 V_ Ü] _ 'O O Z5 oa Ü$·IIII Q 0 ::1 :> s: +-1 1:41+:‘m I-I -1-I O •~ +9
Q O E §>~, N O<+-1• o II <•-1 -6 6 ·-1>< E<¤|···I ·¤··* O O 1--I O -+-1+-1 cn I>-¤ O <¤ Q E1.
Ü] O QQ $-1 « . _·+-1E2 O
lg
EI Q'] O
1
65
TABLE III. EXPERIMENTAL DESIGN TEST SEQUENCE USING
· QQDED VARIABLE VALUES
Subtest No. Aluminum, X1 Polxelectrolyte, X2 Time, X37+ 1 -1 1 -1 -1 6
2 7 +1 . -1 -13 -1
1 ' +1 +1 · _
4 -1 .1
-11
+1 -
5 +1 - +1 -16 71
7 +11
-1 +1- 7 -
1-1 +1 1 +1 7
-- 78 . +1 +1 +1’ 9 -1.682 1 6 0 . 010 «+1.682 70 01
111 0 -1.6821
1 _ O .
7 12 « 071
7 +1.682 1 013- 0 0 -1.682 1 114 0 0 1 +1.68215 ° 0 0 016 E 01 10 1
01177 0 0 O1
18 0 · 01
7 0
19 · 0 , ·0 020 0 7 0 O
n‘ r · 66 °
Chicca (78) have used regression analysis to predict thealum dosage in treatment for a potable water supply. The {work was successful in that a dependent response variable,Vturbidity, was related to alum concentration in a computergenerated regression equation.
Experiment Three: Gravimetric Mass Balanpe_
» Because of the author's particular technical backgroundit was believed that a gravimetric measurement, though tedi- _ _ous, might prove an accurate method of determining the effi-ciency of the various coagulant systems. For this reason amass balance was made on the treated waste by measuringtotal, total volatile, and total fixed solids of mixed,(clarified, and concentrated portions of waste following atypical coagulant treatment. In addition the same data wasE
‘obtained for the suspended solids test sequence. From this _experiment the reliability of gravimetric_methods in measur-
E ing mass balances before and after subsidence wasassessed,.and the particular gravimetric method which showed thegreater per cent change following coagulant treatment wasdetermined.
(Experiment Four: Selection and Comparison of Dependent_
Variables { . .Initial testing was undertaken to determine the most
‘ accurate response variable. Ten tests were performed on the ·
tannin waste using the same independent variable values andNalcolyte polyelectrolyte 605 as the coagulant aid. Thefollowing dependent responses were measured: light scatter- '
'ing turbidity, light transmission turbidity, COD, totalvolatile solids, volatile suspended solids, filtration time,refiltration time, substance volume and rate, and tannin ‘
concentration. It had been hoped to measure zeta potentialas a response variable, but the waste system and obsoleteelectrophoretic mobility equipment proved cantankerous. Theresponses were thus limited to those previously listed. The
. pdata·obtained from these tests were analyzed according to“—
range and standard deviation. h ·
'Experiment Five: Selection of Procedures for DeterminingDependent Variables .
6
‘ ‘ In the following sections the specific laboratory pro-· cedures used to obtain the different response variables are
ldescribed orreferenced.Chemical
Oxygen Demand. 0.0.D. determinations were performedusing the Technicon Auto-Analyzer. Specific operation
E
directions for this equipment are listed in the operatingmanual accompanying the Auto-Analyzer. Generally, the chemi-cal reagents oxidize and digest the oxidizable material sup-plied by the waste. During the oxidation, depletion of the
p oxidizing agent reduces the amount of color of the solution.
J J 68
This change in color is measured by a colorimeter. A stan— .
°dard curve was obtained using glucose concentrations of
known C.O.D. From this standard curve the C.0.D. of the
J waste was determined. A flow diagram of the Auto—AnalyzerA
_ is shown in Figure 7. (30) ' ·A
Filtration Time. Filtration time is the elapsed time betweenthe pouring of the sample onto a filter and the appearance
A A
Aof the first dry spot on the filter cake in the suspended T
J solids test. The procedure for this test is given in more
detail in the discussion of the suspended solids test. The.. basis for the filtration time response is given by LaMer, '
A_ et ai. (67, 94)
A
Light Scattering Turbidity. This turbidity response wasA_ made on a Hach nephelometer. A plastic cylinder was supplied
AJto calibrate the instrument in Jackson Turbidity Units
A(J.T.U.) following a 15 minute warm up period. The undiluted
samples were poured into empty, unscarred glass cylinders to
at least a minimum depth and placed in the instrument.« The A
glass cylinders were then covered to prevent stray light J
J — linterference, and the J.T.U. reading was made.{
Light Transmission Turbidity. The Bausch and Lomb Spectronic
Twenty colorimeter was used to determine light transmissionA turbidity. Specific operating instructions for this instru-
ment can be found in the operating manual. The general '
69
SEC Ol O O Q\ 1;\ Q Q1-1 -01 ***0_ ng *** Ci! _ Q Cd Ol G1011
0:1. 1., Iii ***7* .:2¤DE 1-1 |1g 1 S$-1aß gl L, E
1 1->Sm , lg 1 g ¤> :-1
l"*"aß1->I |m 3 3 · <•-1
° I 'ä’„ 13 *== EEI.:l‘*' 1 ·—• -3. 2 1 gf 1I-: Q '"° °’ "' * ¤>
l °H '*I '**‘ B4 QI *—‘· += 1a
W >·· E w :s S.,-4 .,..1 Q.,.4Q
<C/2 Q__ _ 1:0V ** **1 * · ·1ä“ igI 1 |1; 0
.1ä
0 II I! II 1 II II 0 II I3 ,1 1-1
L l- llE¤— cu1 2 :11
l/ 11 \ IL jf
,1:: <1-11cu -
· 1:- 5., _ ä°’ Q Q7 #1 .:1 N° 1-I
1 5,, E 7* B cß(D
.1 Q .,.4¤15 ° <¤1 __‘
\ ¢ _ ä 2 1E B‘ Z312: 1 ".¤>¢ °§ .
5 -1-1Z In CD gg 1
1 0 *3*** *·• ¤¤¤-1-* Q) .,.4.1 @1 Q E1 sg ¤§
E ¥>~1 0- 0 ggg5 1 1 Uv Z ·S --7 111Q l O<¤ E 1 Y 1 *‘> Ü Q) '• Q)
0 I" 7**E- >.::>1 "" °’ ‘° ¤
Q *1 EO· 1 Q .,.4 cn$ Q
70
procedure was as follows: 1) one milliliter of the clari-•
fied portion of the treated waste was diluted with distilled
water to 10 milliliters, 2) the sample's per cent light
transmission was determined at a wave length of 700 milli-
·. microns. It should be noted that this wave length was not
U ’ Ithedominant wave length as described in Standard Methods for
q the Examination of Water and Wastewater. (96) Another test
performed on the Spectronic Twenty was tannindetermination,and
this test required operation at 700 millimicrons.
Operation of the instrument above 675 millimicrons requires4
t a different light photometer than below this wave length. _””
For accuracy and ease of operation it was decided toutilize·
the wave length of the tannin determination, the more accu-
h rate of the two tests. Dilution of the tannin waste was
necessary to obtain an initial per cent light transmission·
_ of other than zero- . _ W —
(Refiltration Time. The experimental apparatus for the vola—
*2
tile suspended solids test can be modified so that the fil-6trate can be collected. The time that 10 milliliters ofthis filtrate takes to refilter through the already formed
”filter cake is termed refiltration time.’ LaMer, et al. (67),
(94) have described the theory and modified procedures for
refiltration time.
q ·. V 71U
Subsidence Volume and Rate. Following the completion of the
'mixing and stirring sequence in the jar test procedure, the
entire mixture was poured into one liter graduated eylinders.
The level of the interface between the clarified and concen-~
_ trated phase was measured at a regular, staggered time 'I v sequence for the next 18 hours. The level of thisinterfaceat
18 hours was termed the subsidence volume. The rate of ‘ ~
change of the interface with time over the initial constant
period was termed the subsidence rate. These procedures, inamended form, are described by Eckenfelder (24), and also
_ in Standard Methods (96) under the title of settleable _j” solids. A j · '
_Tannin Concentration. Tannin concentrations were determined· in accordance with the procedure described in Standard Methods
(96). The instrument used was the Bausch and Lomb Spectronic1 Twenty. The method described required very weak solutions,
necessitating the dilution of both treated and untreatedA
(waste prior to analysis. Initial work in this area was two
fold. First different concentrations of pure tannic acidwere prepared in an attempt to establish a Beer's Law type
relationship. As the author consistently failed to weaken
the solutions to a sufficiently dilute condition, this
undertaking was overly time consuming. Secondly, testing
was done on raw tannin waste to determine the necessary ex-
tent of dilution. This level was found to be 1 to 1000.
. 72
The dilution was accomplished in two stages. One ml ofwaste was diluted to 10 ml in a 10 ml graduated cylinder.After inversion mixing, one ml was withdrawn with a one mlpipette and diluted to 100 ml in a volumetric flask. Fol-lowing the determination of the proper dilution range,tests were made to insure that no interference by non—tanninswas observed. This procedure was accomplished by preparingvarious dilutions in the proper range, doing tannin deter-minations, and then graphically evaluating Beer's Law _ 1
relevance.Berks,
et al. (6) have described the critical sectionsof the tannin determination procedure. It was stated that—the amount of tannin reagent was not critical to the accu-racy of the analysis, but that both the amount of sodium
— carbonate and the time interval between addition of sodiumcarbonate and making the spectrophometric transmissionreading were very important.
”(
The final pre-testing related to tannin determinationwas standardization of the absorption cells using cobaltchloride solutions to measure correction factors for the
l
individual cells. This procedure is described in the_ spectrophotometer manual. _
Total Volatile Solids. Total solids, total volatile solids
and total fixed solids are all determined in the sameanalytical test. Following settling of the treated waste y
g vz ( 1
for 18 hours a 10 ml sample of the clarified portion wasplaced in a weighed, dry 90 ml evaporating dish, The eva-porating dishes were weighed on an analytical balance,
vevaporated on a steam bath for 90 minutes, and placed in a
103°C oven for a minimum of three hours and usually over-A
night. Following removal from the oven, cooling, and re- yweighing the evaporating dish, it was placed in a muffle
furnace at 6oo¤c. The minimum of residence time in the
furnace was three hours. Cooling and weighing produced
data allowing the calculation of total volatile solids.This procedure has been modified for a specific application
from the more general directions given in Standard Methods(96). g ,
UVolatile Suspended Solids. .Suspended solids tests were yperformed using Millipore filtration equipment and fiber-
glass filters rather than the Cooch crucibles and asbestos
_ filters recommended by Standard Methods (96). The total
suspended solids tests performed using Millipore equipment
are described generally in their methods manual. Speci-
fically, a 25 ml sample was filtered under a constant q
vacuum of 60 cm of mercury through a 47 mm diameter, 35 ‘
micron pore size, fiberglass, millipore filter. After
completion of filtration, the filter and filter cahe were
placed in aluminum weighing dishes. »Then, using the same
technique as was employed in total volatile solids procedure,
‘ A74A
theAaluminum dishes were 1) weighed on an analytical balance,2) placedAin a 103°C oven for a minimum of three hours, 3)cooled in a dessicator and reweighed, 4) placed in themuffle furnace for a minimum of three hours and 5) cooledin a dessicator and weighed on the analytical balance. From A,these data the total suspended solids, volatile suspendedsolids, and the fixed suspended solids_were determined.
« Jar Testing Procedure(
AV
There is no standard jar test procedure. Sawyer andMcCarty (92), Birkner and Edzwald (7, 8), Black and Chen(14), E1a6k Gt a1. (15), E6k6dr61d6r (25), Edwards (27), A
A and Hannah et al. (44) all describe coagulation jar testingprocedures, and while the methods are similar, they are not
- the same. The procedures have been modified to suit the' nindividual investigator's system, equipment, requirements .
(such as constant pH), or prejudice (square versus roundjars). The author's procedure is a modification of thepreviously mentioned investigators' methods and is identicalto none of them. ‘ (
EOnce the author had familiarized himself with the
necessary laboratory techniques, a maximum of 10 jar tests 7
were started on a given day. A particular run extended inits analytical phase over a three day period. Two separate
·
apparatus arrangements were used for the rapid and slowmixing phases of jar testing procedure. The liquid waste A
p _ 75l ”
containers were square jars with a total volume in excess a
of 1200 ml. The rapid-mix apparatus is shown in Figure 8.The jars were filled with one liter of waste and the stir-
ring rod of the variable speed stirrer, the pH electrodes,
and the graduated cylinder for pH control were placed in*’
correct proximity as shown. The initial stirring rate was V
set at 100 RPM and the pH was adjusted.to the desired level I(normally 6.0) using concentrated potassium hydroxide. The
correct dosage of Al2(S04)3·18H20 was added in the fine
crystal solid state and the resulting pH drop was corrected
by rapid addition of KOH. The concentration of the KOH
wassuchthat rarely was more than 10 ml of the solution neces-‘
sary. This volume variation was neglected in calculating·
concentrations.·— _ Rapid mixing was continued for exactly one minute and
y the speed was then reduced to the minimum possible with
the stirrer, about 35 RPM, for the time lag required by
· the experimental design. At this time the stirrer speedt
was increased to 100 RPM again and the polyelectrolyte con-
currently added. The concentration of the solution wasprepared so that a 100 to.1 dilution would always produce
1 the concentration desired in the waste. The polyelectro-
lyte was discharged from a 10 ml pipette_at a constant flow
rate using a pipette filler for a period of 30 seconds.
Rapid mixing was continued for an additional 30 seconds.
76 4 °1 - $.1,-1 _
_ rx Q) Q.$-1 N $-14) ~1-4-49·*·’ E S·«-1 O
1-1 'SOcn GJ-1 cu mm4- '¤ ¤«¤«$-1 O U) -_ ¢‘6>~.$-1 M
·¤ •1+>+-v ¤> Q«,-:, ·„-1 0 1--19:4GJ GO GJ „¤ -$·4¢Gr——i CG-49ä>4<¤¤1$:J ·•-4 QD•-I S65 $-1 $-1
vom cvs SU)-/S1 >€¤ 1Q) QB ‘ .@_td1$-1
" ¤-1@ =¤,’vQ
”\ <-49méi$-1cß*1
_ 9-4‘ l>I¤·° °‘ E. . „ bl)‘ . ' cd
-1-4
. Ü "-1-4. · _ .+Q
Dlv4**ä
._ $-1
$5bi)
77 0
lU The square jar was then removed from the rapid mix .
stirrer and placed on the Phipps and Bird multiple stirrer.
Here stirring was continued at 30 RPM for a period of 20
minutes. After removal of the first gar to the multiple
stirrer a second test was begun, and the procedure was
_ repeated. The timing on each of the io separate jar tests
was recorded individually. After the_stirring periods
were completed the contents of the square jars were poured
into one liter graduated cylinders and allowed to settlel
for a period of 18 hours. This period_was chosen because
Ä- of the slow rate of subsidence of the heavy phase-clarified
phase interface and because of the convenience of this ‘
Ä period in beginning the second day's analytical work. All0
analytical work was performed on the clarified section with
.samples being taken 1% inches beneath the surface.i
Some applied chemists use physical appearance as al
measure of coagulation efficiency; however, the author didi
not use this index. Nonetheless, any unusual clarity ob-
tained during or at the completion of mixing was noted and
special attention was paid to the results of these tests.
The pH meter was used to measure and control the pH
during jar testing. It was standardized daily with buffer
solutions of approximately pH 4, 7, and 10. Considerable .
fouling of the electrodes was observed and this problem
was remedied with cleansing solutions of concentrated HC1. ‘
78 A
Anionic polyelectrolytes, in particular, were difficult° to prepare in the high concentrations necessary for parti—
cular subtest requirements. This was overcome finally by
use of volumetric flasks (3/4 filled) and alternate use ofi, thermomix (heating and stirring) equipment and the labora—"
‘ tory shaker (violent mixing). Some preparations required
24 hours.” ° -
Experiment Six: Polyelectrolyte Testing and Evaluationi
Using the jar test and the experimental design described
in the previous sections regression equations were obtained
_(
for each of the 10 polyelectrolyte system. The coefficientse
I,of the polyelectrolyte variable, B2 and B22 (Table II) were
' compared and the polyelectrolyte showing the highest removal· value was chosen for further testing in connection with a
new variable, pH. In addition, the accuracy of the various
dependent variables was evaluated by comparing these vari-
ables with the reduction of total volatile solids usingU
results compiled in testing the a1uminum—polyelectrolyte
systems. Also, the effect on removal efficiency of eaeh of
the three independent variables for each alum—polyelectrolyte
'system was evaluated and depicted graphically.
Experiment Seven: Expanded Testing Procedure for pH
Evaluation ( TA 4 _
Following selection of the best polyelectrolyte a new
_79 e
experimental design was developed to evaluate an additionalindependent variable, pH. This design and coding is shownin Tables IV and V. A regression equation was developedfrom the data acquired using this experimental design, andgraphical evaluation of this equation in a manner similareto that described previously allowed interpretation of theeffect of the individual independent variables. p 1
It may be noted that the time lag variable range wasA
doubled from the previous experimental design. This in-Q
crease in range was based on results of initial coagulationtesting with the 10 polyelectrolytes and will be discussedfully in the data and results section. ”
80
' — N O N O O‘ I H •UN
· II/2 _ ‘ ' CJ 4* UN UNtz] _ 4* -4 -4 -4 • •:-1 M ° 4 UN -4 UN
. Q 01 -4<T‘ N- ._I-4 · ‘M 4* -
M · 4* 01 7<Ü CQ N ' '¥> » 0 ¤o 0 0‘ + _+ O H N • •E4 · O NN kOZ N NN . NN ·JF:] NN M _D N' NZ . VN NE1 NN NN _9-4 ID N I -Q ° Q 0 00 UN UN
+ — T'I 1'Ä [*r\ • •
E_ + + UN 4* \O4NNN gr 4* .· VN N
E ° I m MH I ‘, 3
‘+ 4* '
C:] H' .Z D N Q °O :-3 N O O O O|"'I N + N H \.¢°N • • ”E-4 ·4> 01 -4- 0 xo 4-4 N NN \OD Fil N NQ -4 · IC1:] O +· ·N ·
. O NNZ N H U1‘ 0 4:: 4><4 Q - ··¤·¤·4·-4 4--4 N _ ~ ·— ggU) Lil Q _+ ‘ _ •• :60 ·rn Q ~ 40, 4 0Q E4 + 01 0 ur E0M . M :-4 cn *:-4 ISmfb O H :¤ GJ QEl N M :6 :-4 bi) •— -:-4 •~M --4 01 -:-4 g a · E NN 0
I · M -4 $-4 0 0 SM rnQ ‘ H CQ Q ·•·* •~ $1 · 4-4 GJ::1 -4 {> rn -:-4 0 qg .. ,.4Q Q + · S >< 0 0 *2
4* -4-* GJ :.14-* 0ä •• + N N ¤ E ^ GJ G->¥>> --4C4-: g 4* NN 0 -:-4
• 4-*4 0:-4 anN ’ .0 *'* N N "Ö D O >~: 30 SZ!E4 _·:-4 M 4* 4* g _ gz :-4 -4-*$-4 0
-4-* H 4* NN GJ •~ 0 0 GJ-P EQ 'CQ CG Q . GJ O $-4:-4 ::¤O -:-4• S 0 S 4-1\ 0 Q
4> _ + + + .-0 :-4 E 0 vw wa:-4H- BT] _ . S1': Q 5 GJ E QGJ •~
0 4-4 > g :-4 :-J:>> 4*[Ü C -4-4 G) •~ 4-4 MA 0 _ *4-4 0 E :>„ 01 00Q -:-4 II 0 *0 S :-><En-:4
co . 0 :-4 0 -:-4 :4;*,E-4 — - cn :>-4 °cJ 4 ¤-: E4 Q
0 C$-:_ -:-4· bl) "C$_ .
GJ 0 ·II} O
81
TABLE V. EXPANDED EXPERIMENTAL_DESIGN FOR EH EVALUATION
Subteet N0. Aluminum, X1 Poly., X2 Time Lag, X3 EH,X41 -1 V -1 - ··-1 -11A -1 -1. -1 -1.2 +1 -1 -1 -12A .+1 -1 -1 -14 ' 3 - -1 +1 -1 -13A -1 . +1 -1 -14 -1 E — -1 · +1 -1
_ 4A -1 -1 . +1 - -1_ 5 -1 -1 -1 · +1
. 5A . -1 -1 -1 4+1‘6 _ +1 +1 +1 -1 —
*6A . +1 _ +1 +1 -1. 7 +1 +1 -1 +1
7A +1 . +1 -1 +1 —8 +1 -1 +1 +18A +1 -1 +1 . +19 _ -1 +1 +1 +1 _ ’9A -1 +1 +1 +1
10 +1 +1 +1 +1 4
10A +1 +1 +1 +1‘ 11 +1 ' +1 -1 . -111A 4 +1 . +1 ‘ ·‘ -1 -112 +1 -1 +1 -1
· 12A +1 --1 . +1 -1· 13 4 - -1 M -1
” +1 +1 -13A --1 4 ” -1. - . +1 +114 V -1
4+1 ‘ +1 -1
14A -1 ' +1 +1 -14
15 _ +1 -1 -1 +115A « +1 -1 -1 · +116 -1 +1 _ -1 +1
4
16A _ -1 _ +1 - -1 4 +117 -2 0 O 018 +2 0 0 0
— 19_4O -2 ° 0 0
20 0 ‘ +2 ‘ 0 O4 21 0 0 -2 0 4 _224V 0 0 +2 0
23 O 0 O -224 0 0 0 +2 ·25 - 30 0 O · 0 0
IV. EXPERIMENTAL DATA AND RESULTS —
E-The experimental data and results for this investigation
are presented in the succeeding sections.
jlg l
Experiment One - Characterization of Waste _
The data and results for_the characterization of the .U
tannin waste are shown in Table VI. The results of this ex-·
l pperiment showed an average total solids concentration of
30,151 mg/l with an average total Volatile solids content of .
20,815 mg/l. The range experienced for the total solids
h6
concentration was, excluding data cn 6-6-69, from 28,320 tol
„ 32,360, a Variation of 4,040 mg/l or about 13 per cent. ’
Since the experimental design required constant tannin waste‘ . concentration, it was eoncluded that this Variation was too 1
large and could not be accepted. Subsequent testing showedl
these Variation Values were higher than normally obtained .
in a sequential three month period. While total Volatilen U6
solids remained relatively constant at 21,000 mg/l, total” solids dropped to about 28,000 mg/1. ’ l
·
It was noted that the Very high concentrations obtainedl
Ion 6-6-69 had a distinctly different color from the normal
milky brown waste color. Such waste batches were thereafter
avoided. _
82
83 ‘
S S S S S¤> § 3 3 3 é § 3 3 § § §O O O O O O O O O O O O¤$-1 $-1 $-1 $-1 $1 $-1 $-1 $1 $-1 $-1 $-1 $1
. Iiü *6:0 S S S S S S S S S S S ·E1 E-11-1cn 6:0 >. :>. t>» :>> :>. :>.$-1<‘1.¢DO M M M M M M 46 M M M M3 _ 1-4 _1—4 1-4 1-4 1-4 1-4 CD 1-4 1--4 1-1 1-4
. -1-1 -1-1 -1-1 -1-1 -1-1 -1-1 1-1 -1-4 -1-1 -1-1 -1-4. E 4 EI E2 E E EI E! Q E E E3 E!
E··I(D
III] "U '.III ""I ‘ .
· E1 _1-4 4* LO O O O O O O LO O l‘- IO OOr-1 CD IO ON LO ON <I) LO LO 1-1 IO CO 1-4 O
ä (Obi) O ·4* IO IO 4* O\ CD LO «-I LO N -4 O' "UE ON ON ON ON ON- ON. O l‘ CD ON ON ON l\
Z <D 1-4O N
1 I-4 -1-1 . —
H .II}E1 rnO "Ö 1ä -1-1 .
,-4 1 .
ä O ON O O O O O O O O O l‘ CI) O(O1—I LO l\· l\1CD N Q) IO IO IO GN IO IO OO who
l‘. LO ON IO N O O O LO l‘ l‘ O O-
1-IE N N N O O O (D :-1-O CD *1 =··I FI•• -1-1 N N N N N N N N N 1-4 N N N
Z GO _1-1 ·E4 :>Z ‘ .CI]E . .I-4 1 · ·
C!] (0‘
B-1 'Ukg .,.1 .Erl 1-1 xn LO O O O O O O LO O 4* 1-4 O
O1-4 LO O MO NN -1-4 \O‘ CD CD 4* N N LO O
ä (Oben (I) O IO l‘ l‘ O CD MO \O IO O 1-1 O1-IE -1-4 N N O\ O\ O CD CD CD 'CD 1-1 O (I) ‘
(O EU IO IO IO N N IO IO N N ~N NN IO NE-1 -1-* . . · ·1J 0 -
. ID E·1- '· ·(O ,EI} $-1 hl) O
"CSEU Z TO• <D<D ··-1 I
I--4 1—4P> ON CN ON ON O\ ON ON ON ON ON GJ 'OON O> I MO LO MO MO \O \O © \O RD \O ¤D 5*.0 -1-1
EP. I I I I- I I I I I I CU 1-IICr] (SGS (I) O N IO 4* LO NO l‘ CD O $-1 OMO1-1 mr: 01 rn 1-1 cu :11 ¤G I I I I I I I I I I I P ¤>I +><1: 0.:: Ln nn xo Lo LO LQ O Lo LO LD 4 LO .
- E-1 -1->·+¤•
.63:: bo 1-1° GO P IE .1 <¤ 1‘
I a·
84 _
‘ Experiment Two — Deterioration of Tannin Waste .A
‘ l E During Storage L
The purpose of this experiment was to determine therate of deterioration of the tannin waste during storage.
. Three batches of waste were tested over a one week period
at room temperature, 20°C, and approximately zero °C, using
solids concentration as the response. ”The results.are shown
in Table VII. In Figure 9, the change in Volatile solids oi
with time is shown for each of the three temperatures. _
While the Variation was not identical at 20°C and room tem-perature, there is consistency in that the Volatile solidsdecreased significantly with a tannin waste age of either·
two or three days. Further, the appearance of the wasteE
underwent a striking change at roughly the same time.
°Floating~white curds began_to appear in the waste containers,and eventually covered the surface of the waste. On theiother hand, storage at near O°O prevented significant change
.as shown in Figure 9. The Variation at this temperature is ‘believed to be due to error in the analytical technique alone.
_ The results obtained in Experiments One and Two required
that a single large batch of Vegetable tannin waste be col-‘
lected at one time, stored at about O°C, and used for themajor polyelectrolyte experimental sequence. This procedure
r
u was followed for the larger portion of the tests as an ex- jcess of 100 gallons of waste was stored in the chemistry
i
department cold room.
85
. Fl A _- bl) 2
E1-—l1—I „ '.
· ¥lJ¤Db1JBBEA -P1-·IU1U} OOO /t"\ONN NON <1‘N\O NIDO
ONNNN <I‘\Dv·l 41"UNON l€N&O-·l l‘lwLfN NNNIIN NN<1)vlONUNLIN OLINUN <Du’N·:1"' CDONCD ON<T‘NN N<I"l‘·gfplj ggg ggg ggg ggg ggg ggg ggg
U) o QOO CDONQ) CDONGO ONONCD CDCNC0 CDONCD Q)ON(1) ONONCDgd O1-ltßw vl N vl' N vl N vl N vl N vl N vl N ,· S-11->·¤«-1 ' A 2 ‘ .
° IZ ßtöüßtß . . ‘ · r‘ I-I, ¤>1-1N-1=> ‘ -‘ . ZG-:-IO
E><1—l+> ,
p-I 2 ' .
Er-1O1 bi) ‘—§1-l1·—·I ·
Z \\ ° .¤ www .‘I·-l ··¤ES · „E-! U21-lüiln ~
<DO'¤"C5 OOO CIJNO NNl\O d"vlLfN vlOv4 NNvl<I" NONO E-lm--1·-1 womo mm-1 .1-1Lnw vlOO 1-lmm www C>®C>E-I: 1-—l1—l ONNNNN C0<I"NN \!\<I‘ON -=*1‘©O LINLINO CD<I"t~'N <I"LfNON
O ggg ggg ggg ggg ggg ggg ggg
. FT] _Or-lUJUJ' NONN NONN NONvl NONN vlONvl OONO OONONE-! N-:-l N NN N NN N NN N NN N NN N NN N N ·EI GGJCG °
1-lN+-7* .2 I o--16
>‘I-·I+-’*3 .
+*1-I _· El ll) · ' „
Z <DbD ' .BJ Ela-11-1 · .E \\I-I ·_<DU1bD¤DE SEMG1 ~P1—-lllätfi GNd"NN UNLFNO l‘vlCD 4I‘N\O ©O\O \O4T‘O UNOUN
' &GO"¤"¤ ©CDu’N ONvlvl Lf\kOvl ONMOLIN <I"LlNON ONl‘-lw -:1*‘<I"d)Il'] S60)-:-l•1-l l\OCI) vlN<I" \O\ONN CN1-lv-! l‘Ol‘ 4I"CIJNN CDl‘LlN·_ ggg ggg ggg ggg ggg ggg ggg
El-1 CDOO NONvl NNONN NNCDN NONN OICNH C\]®vl vl<1)OO E1—lU}U} N NN N NN N NN N NN N NN N NN N NN
®·1-Ir/J' E-1+-*1*::5-1 ~El 2 CUGJEÜ -Ä E•"*N+° — , ·zu oo--lo · .U) o><1-11¤ · _Fl'] D'-}Cd
GJ. _
I-l In AI-I G · . „> 3 ·CT] S:A ·1-1 -CQ<ü
¤>>E·* GSK} O2 vl · N NN <|" UN \O
EO „*I-l ' ·Ö .G) _ ~ „
<: „
86 H
l I O IQ) *°I II
EJ Gé ·I^ II „¤
I IU}I:>„
I Ü I 09 HSIZI II
3 I¤ ICDIG)
I Q{Ä
$$*6 I H II-I-7+9+9 _
Icöcßß _ ;.„$$$ I °’
I I I I I w *"am I@ ° I2@@0)I 6;-·&-·&« IEco I I 6Ä IOco _OOO I qg
II ue "‘ ÜbI
I
(ÄII
I6 6 6 6 6 6 6 6 I
O O O_O O O O OIQ uw Q LI’\ Q ux ux ONN Ol O] 1-4 1-4 I I TI/gw ‘¤lHZ KHG WHJJ Spxtos a1;1c1dA u; ascaabaq
U The remaining testing in this experiment was concerned
with the determination of a sampling procedure from the .
stored.waste that would insure sample uniformity. The final
procedure adopted is described in the previous chapter.
· · . Experiment Three - Gravimetric Mass balance.
The purpose of this experiment was to determine the
gravimetric index most responsive to change following.coagu-.
lant treatment. From the results in Table VIII it was con-.
cluded that for the solids analysis either the reduction in
volatile solids in the clarified portion or the increase in
volatile solids in the concentrated portion would be respon-.
sive indices to coagulant treatment. Because of the fact.
that the concentrated portion concentration is a function of
. time while the clarified portion concentration remains con-
_stant, the analysis involving the concentrated portion was.
rejected. The suspended solids analysis showed much greater
percentage changes but were inherently less accurate due to I.
the smaller weights.considered and the more involved experi-
mental procedure. Once again the superior index was the
? reduction of the volatile suspended solids. Fixed solids
were not considered as this index is a measure of quantity
of Al2(SO4)3 · 18H2O added. ._ There were several interesting.relationships noted in p
the results in Table VIII. In terms of a mass balance it
88
bb 1-I MO t~‘N ON E- 4* 'O O} I I I I
'° G O Ü [‘ 1-4 \O \O l~"N1-4 C1) N ON 1-4 1-4
4.¤o° + + -4- + NN 1-4$-4-4-4 + +-4-9+9 ·Q$-4
td 00 • 141 1.41 Q Q Q QO 'O¤¤(O O Ü _ ON _N UN l\. ä :1 ::1 0 1~ 1- 01_ 1~ ON
Q Q A A A A A A
_ DVI .• _ .
'U O Ü 4* FN \O l\ U‘N ·O GQ 1-4 N. 4* + ·I· 'I-4 ··4-4o° I + I I ·Q ‘I-·I•r··I ‘ ‘
E4 ·•4*4+°C!] $-4$-4 • O O O O ‘O OE GOG N N 4* ON UN 4*4-4 ·4-4¤-4::1 ON 4* 141 1~ 4* 141> O Q A A A A l A. Ä · 0 141 01 oo 01 01 -
v4 1-4 NC5 . ·I . ·
· BJ • 1~ 1-4 141 0 01 ‘E Fc QD • • • • •-4::1 0 ..:1 0 10 ON 1-4 4* 0Q +9 O N 4* ON \OE4 445 + + + N +49 ° +E4 -F-4G.’ Z 5449 . .‘ Kü _VJ l .EI 400 • 0 0 0 141 141 141 ·4-4 0;:0 1~1 4* 1-4 141 UN oo · 14::1 ><1 ::1 4* 1~ 01 141 10 ON
'I" O G A A A A A
¤-4 2: 0 ON 01 01 1~. -4 ooB1
·Q •
U) "Ö SE4 ·G O I I I I I IA -4-ÜIGD ß+°° —U) GUI — .
Q +9fZ • O O O O O OQ 0 oo 01 0 ON 01 1-4{D CI 4* O UN 1-4 4* 10
O Q A A A A A
” 4-4 +-4 1-4 N ‘I-I .¥> · 1Q A .•m 4-4 4-4 4-4 U14-I 4-4 _«-4<Ü •':•‘\ \.E4 bb bb bb Übb •bb •bb
E E E' E VIE VIE_ G . G ZS 5g—{ A A A p¤{ A U} A U) A
-4-lm U1 U1 ·-4m' cn U)~4—9"O 'O"O 4-4"O +9'O "O"O 4-4"Ocd-:-4 G-4-4 mj-4-4 cd-1-4 G·4-4 CG-4-4r-41-4 :144-4 -4-*44--4 4-44-4 Mr-4 +-9r-IOO -4-4O OO OO ·«-4O OOt>u1 mm E4Cl} :>m mm E-4m
t 89 9
was seen that the average of the total solids in the clari-
fied and concentrated portions equaled the total solids of
the mixed state, 32,065 versus 32,210 mg/l, respectively.
This correlation was excellent and expected, as the samples
1 were taken when the concentrated phase had settled to exactly
half of the initial volume. No such relationship was foundin the suspended solids calculations, indicating the decreased
reliability. 1 l
The increase and constancy of fixed solids in all samples
was significant. The increase was clearly due to the addition
of the Al2(S04)3·18H20. Further, the constancy of the various8
fixed solids indicated that this salt was in part either
suspended or dissolved and had not precipitated with the« destabilized colloidal substances. However, this material
y was not largely suspended as the fixed suspended solids data
. indicated. There was essentially no increase in fixed sus-
pended solids from the untreated tannin waste to the clarii- I' fied portion of the treated waste. This result meant that
substantially all of the hydrolyzed metallic ion in the
clarified portion was passing through the_filter cake and_
was therefore dissolved in solution.Q
Y
Related Gravimetric Testing “ I
lIn the gravimetric analysis development there were a
number of related, relatively less important tests that
were performed to assure the accuracy of the overall
gv . p 90 I
procedure. On this basis the following decisions were ·
made. 1) Fiber glass Millipore filters were adopted for .
suspended solids testing because cellulose filters tended
_ to burst into flame in the muffle furnace, scattering
n filter cake and giving poor results. 2) The sample size—·° of 10 ml for the total solids testing was chosen since the
_ 50 ml size recommended by Standard Methods did not take
into consideration the high waste concentration and exces-
sively long drying time. 3) Minimum drying times for solids‘ testing were determined·for the steam bath, oven, furnace,
·and·dessicator. 4) Twenty ml suspended solids samples were
found to give reasonable filtration times at a vacuumof)
60 cm of Hg.V
~°Experiment Four: Accuracy of Dependent Variables
VThe purpose of this test was to determine which of the
6dependent variables could be determined with the greatest
accuracy. The results are shown in Table IX. °
In comparing the coefficients of variation„ Ö/x, of
each of the dependent variables it was evident that the
most accurate responses were reduction in volatile solids y
_ and the spectrophometrio tannin concentration. Because the
‘spe¤tr¤pu¤t¤metér was used by several other students regu-
larly and because the photometric tube was also changed
periodically, it was decided to use the per cent reduction
of volatile solids as the primary response. The other _w
91
InSIC! . _
U ¢U·4·l O U'N CD 4* O UN l\ ON ON NO C1) NN NNkg O O I ll I I O O I C O ‘-I
NES NN NN N N NN N N 4* N l~'N N O 41) I ·td \O \O MO ‘0 NO RO \O MO *0 *0 \O • • 1-4§E-4
Eil .•. CQ "d&
<l '*"* O O UN O O O O O O O N kO MOI-•I • • • • • • • • • • • • g-{ {ijII} 04-4 CD ON CD UN UN UN O N vl (D MD 1-4 • Q)<¢ {JO 4* 4* 4* 4* 4* 4* \O 4* 4* 4* 4* • UN:> v2:>
, gg · .Q gw-4 -Z ···!E ‘Iii} -I-P 4* UN 1-4 4* Q) UN UN ON IN. N MO 1-4 1-4Q (Ü •• • • • •. • • • • • • • 1-{ [_{\Er] $-ed) NN N 4* NN 4* UN N 1-4 N'N 4* NN 1-4* • l\Q PE vl vl vl vl vl vl vl vl vl vl 1-l • vl
, 4--l·4·-l[7.; •n-IE4 .O F!-4‘ ·!>·• 4 'O ·Ä
• •~ O l\ CD 4* -1-4 \O (I) N 1-I UN ON 1-4 N'N 4*FUQ
‘• • • • • • • ·• • • • [IN [I'} I_
Q QJO O 4* NN N NN N 1-4 4* 4* N N O • NNO CGO 4* 4* 4* 4* 4* 4* 4* 4* 4*_ 4* 4* • 1-4O4 _ _
4 rnSl ‘ ·-‘ 4:1 ¢6§’
ID $-4 O O O UN UN O UN O O UN l\· 4* \OQ • • • • .•_ • • • • • • [§• Q] QEr-4 ' $-4 O ON O UN NN [N 4* 4* UN 41) IN O ••5 MO UN UN'UN \O UN UN NO UN UN U'N • 4*
. E4 ·-PE-l · ’·Z •·-l — ·E4-4 ·+>QC1'} GSE-4 - .
ß-4 U) ·~ UN UN UN O NN O UN U'N NN O ON NN UN 4*M „-Q UN N \O -UN 4* 4* UN ,4* UN \O 4* NN • IL11 •$-4 NN NN NN NN NN NN NN NN NN NN NN O 1-4 NN-I-*5 • vl‘ »B1 °•-JE-4 'O · . °U) _$°E-I • •«
~A •··! · ..Q Q4-4 ON N l\ NN O O N ON 1-4 N ON N NNU') }>O UN ND 1-4 MO UN ON l\ _N N N'N N \O OE
• • • • • • • • • •_ • Q • [_f\Q:} ·-• ON vl N 1-4 ON NN CD 4* ON N_ ON° • NN"dg 4* UN UN UN 4* 4* 4* 4* 4* U'N 4* ’ ,
Q .• LIZUJ ’ 4 ,><lH ,
'Eil •*' ‘4-J O O ON 4* Yi l\ N NN 4* l\ 4* ON U'N CDQ >UJ O NN N 4* ON 1-4 l\ N ON CD NN 14 ‘NN N4 *¤
• • • • • • • • • • • Q \Q I°E-4 •·r-l N N N NN N 1-4 N N 1-4 N N • • 1-4'O4-4 4* 4* 4* 4* 4* 4* 4* 4* 4* 4* 4* OOO ‘IIICD IN bl). _ E7 Q S2· IN • ··-•+) • · • • • §-( • •'¤
N NN 4* UN UD l\ Q) ON O bl) "¤{> @$-4 Os:Oz 1-4 ow Octl· ‘&• 4 mc: <.>:> 4:::
· dependent variables were also tested, but only to compare4
° their accuracy and linearity with the primary response.l
All regression equation development and, thus, independent
parameter evaluation was based on the primary response.
dV.
Experiment Five — Calibration of Dependent Response Procedures\Calibrations developed for use in the dependent response
‘procedures are described in the following paragraphs..4
Experiment 5-1: Chemical Oxygen DemandA
Chemical oxygen demand analyses were performed on the_ Technicon Auto-Analyzer. The use of this equipment required .
“” a calibration curve of light transmittance versus the knownl
chemical oxygen demand of prepared glucose solutions. The
Ll
results of this test are shown in Figure 10. Use of this
curve allowed direct COD measurement of all tannin wastey nsamples. It was necessary to dilute the waste to four per
cent solutions in order to duplicate the light transmittance- ·r of the glucose samples. Untreated vegetable tannin waste ·
had a chemical oxygen demand of approximately 58,000 mg/l.
_Experiment 5-2: Refiltration Timeh
‘ Attempts to measure refiltration time met with failure.The sample size for the suspended solids test was based on
obtaining a sufficiently large filter cake weight to provide
- reasonable accuracy. Regrettably, this resulted in fi1tra—
tion time ranging up to one hour.- Refiltration tests ran
4 93 ·
-
4 O 4‘ O. UN ‘
CN!
. "Ö_ Q
·— O 65— O E .
0 ^ Q3' Ol :-4 Q
4 QD EE CD
~../ bj)!>>
Q "Ö NC: Oaß
. E r-I4 4 O GJ CU
O D O 14 UN ·1-4
_ GJ CD‘ QD S. Q, Q 49 0 :-1
O1-4 ¢1·-1
- cöS ° E9 0 E s-1--1 ¤> *.5
° _ S CJCD $3
4 . _ Q·1—449CU$-1
4 ,Q ·
' 1-40 3 MUN CU
4 e °" 6
4°
· ,.1‘ 4 Q)
5O QD
. •H
·UN 0 UN 0 UN 0 UN 0 Eu\O \O UN UN dl" Q" UN NN
—(%) qqäyq _
I94
i
for a period of 12 hours without completion. During this·’
_time filtrate evaporation had to occur as the filtration
was performed under 60 cm of Hg. The evaporation would have
_ had a serious effect on·accuracy. For these reasons attempts
‘ to measure refiltration times were discontinued, and filtra—_ Vtion time was substituted.
Experiment 5-3: Tannin Concentration ‘
Tannin concentration determinations was the most accu-E
rate of eight response parameters measured in this investi-1 gation. The concentrations of tannin in the waste were
_„ reported as mg/l of tannic acid, the standardizing tannin.·
‘ ‘This standardization curve is reported in Figure 11. Beer's
Law is obeyed over the range pictured, but fails at concen-i
trations greater than 12 mg/l. As the color problem associ-U
ated with vegetable tannin waste was due to the tannin con-6
centration, all color removal values were based on tannin Üconcentration reductions. The untreated tannin waste had a .
tannin concentration equivalent to 10,700 mg/l of tannic
acid.~ ” l
» . Experiment Six - Polyelectrolyte Evaluation . °
This experiment had three distinct purposes: 1) com-, parison of the relationship of all dependent responses with
o the most accurate dependent response, per cent reduction in
volatile solids over the maximum destabilization range,
I
· I‘ I
„ ° I~ 95
100I I I
I ° 90 °4 .
s80
6
6 0 “ .
_ 60I I ‘
QQ, „
+9I: 59
¤°Q
S-eE-4
_ E 40 0bl)·«-••-J
0 E 9cb .
*6;. ¤-« 30 0
025 6
0
° 201918 6 ·
0 1 2 5 M 5 6 7 8 9 10Tanuic Acid Conceutratiou, mg/1
Figure 11. Staudardization Curve for Tanniu .Concentratiou.
96_
2) evaluation of the importance of each of the three inde— F
pendent variables on waste treatment, and 5)_selection of
the best polyelectrolyte—alum system for further testing.A
All of these purposes were accomplished using the data ac-
. quired from the experimentally designed jar test sequences.
Dependent Variable — Volatile Solids Reduction RelationshipH
The data used in obtaining these comparisons came from
selected tests of the 10 alum—polyelectrolyte systems, de- „
pending on which subsidiary dependent variable was measured
o beside the primary response. The data for the 10 coagulant
system tests are contained in Tables XIII to XXII located in
_ Appendix A. The results of these comparisons are shown in
Figures 12 to 15. As seen in these figures, there was a
functional linear relationship between all of the responseIparameters and the per cent reduction in total Volatilesolids. However, it is clear that most of these parameters
have a relatively large spread from this linear correlation —
and, thus, their use as an analytical response variable 1
would be generally unreliable.’ As predicted in Experiment
Four, the tannin and COD determinations in Figures 14 and
15 appeared to give the best linearity. Thus, it was be-l
lieved that these two responses would produce the same type
of regression equation and response·curve as the per centl
reduction in Volatile solids did. This expectation was con- p
firmed by obtaining a regression equation using tannin
sconcentration as the primary response.l Z
97°
°f1°.L' · ,Q Q3 O
pc äunxnwqnnaog cmäm su Äqnplqvm0
’· - .L
cx 'I2
O O
O •~
Ü- "' vx 0 -6 .
‘ an QO 1 ·*" 0 ""¤¤
3“ Q anGY
Ü 0O-^ Q, Q
A 2 O BIT‘ ca
2 0A32
E1O
2-·‘”
EPQ ® 0 3 **7 Q
Q QÜ
2 Om
2 ÜGG 3 .,.2-0
2
Ü
Ügw
C. nä":.
QQ G Q O Q :66-•
‘vw °"
’{Q -6
°0
0 og
Q Q mCD wnns
.
gg:
V}Q
‘0 2,,,,;
O
~>
‘ Q P 03Q ¤°°"'§
>~» ·.-«
AO Elm
. +° +¤ ’ ~O O ¤>
AS °° 0- 3 33,
•«-• Qä
O-•+>•-5 Q
-9 0
OO‘
ExQ
Qqg Q
Q0.
E-• wgQ
¤Qm‘
W4;
a
22
l .
1-{P
·22
Q(3 Qiää
" ·2
0 QGO [N -3 9 0 O
O
·I°$ papllödgng 8*"
nm
· JlüflpagCu
. 98
· UOTSSTWSUÜJJ1'TLIQTQIOGN cu t~ no UN :1* ° vw Ol v4 •.
·-7 ~—? QQ, °"° 0 " 'SG U,
.. Ö 0 —cnge E 3
¤···¢ CTSo ag
0 *" SE · SQ - Q 11 U, °
O O1-1cd E ·
Q gg GQ G G gg Q2_ o0 Q 2 QS
· (29 GG geQGGG O Q0 S §‘3 -_ Ö O S-1n0 sz: cu rnG 0 0 Q -,-1 :1, 0Q G Ö 0 0 :>,53 S13
. g°
0 C>"¤1 1 · E $**3
0 0 S-,· Q Cd :*5
0 O +¤ CIEU . +-> cc 01 :1 0·:s
- Q S S5**P:CD •~ ' Q
-2 ,2 ¤—.¤ +¤ .:1* 0$-1 +¤ .m>0 ¤2, E4 U) Q G O"C§-1 -1-« 0
Sl?O 0
O O O O O O O O O $-1O GN CD l‘ \D · LFN -:2** t^, Ol :3*'* bi)QUTVVJQS 'SJTI ST JQLTV GSBTIJ ÄABQII JO
99 9
IIO'§’.}‘\2.I1Il8OLl0Q -U.]LLIlI'£?,'[_, uogqodpag C|.l130.I9dU 0 0 0 0 0 0 0 0 0 J
ON Ä CD l\ \O LG 4* t*\ Gl 94 Q -5O -2-aLf\ 2-I
Ü 0% _ 0 .20 S
.® •r·| 22.
r-4 «'-{G ¤ E-U)· ® G Q, 3
2 Ü (gg} G G - IJ G- G G--4 -9GE G 2 2
g C) _ > 6-«G 6 ts 2
· GS B25Q @ Ö @ · ° E oc9 G G 9 OE ES·92 G0 EB 3äGG
$—• S SQ) 0 0-9QJ 9-•<;‘
9 G WSG QJGGEGQ
‘ O "U_ G 9 GT3 EJ :1
. S +9 . 9 S mm.2 2 9 2 2— +-‘ 00
l
G S CQ EQ •,-|
-.,4 .
+° E ¤-•E-•r-I Q ° . .,4
9··-• cd ,: g
9 Er-. E- mo ~9 $3--*
‘ O 0-9
” h U • G-99 - 2-42--4
9OC9 :2-9
v'* O ON I1) l‘ MO LG 4* IG G1 S1-I 1-4 _ hg
· 'I1'[llI ‘Sp'@°[0S papuadsng 8’[‘;113‘[0A Bull], uO11I2.1’.)‘[’;-„;{
100,
‘ l Q}‘ «-4
· _V O ·•-+ °
¤¤ +-¤
‘ td, O
„ _ > · 5.
';’° 5
. ' ‘ 5 O wg .
5“‘ ¤a
O°
5 B;
°„
:1;;:,5 5-+
G Gwwggn
G 0 _··¤+¤;>,
· Ö~«-cqgpq
_ 0°-=¥‘ wg;
' 5 @3 Q rfßß-·_'¤_j%¤9 5 mc:
G G 6 · SSP§
0 ' „䤤’«-/
.
-«-eg;‘ · . +>¤>o
·°§§.’L3‘
_"C$·•¤Q ‘@@:3 .
‘ . O ¤:’C¤"¤0 N ¤>
$$72-«¤2-5-•
. 9 ää. ·
(U
° ‘ ··•-+:0@ 5 ¤ SE V
QJO
Q•*2
Q O O ap -VN
O $.5SDpl1'0t1I9([ LIGÜÜÄXQ Ipglwgqo uoplanpau
6101
Evaluation of Independent Variables6
e”
The data used to generate the regression equations
describing the alum—polyelectrolyte systems are contained in
Tables XIII to XXII of AppendixA.:
The regression coeffi-··
X cients of Experiments 6A, 6B, and 6C were reported separately6 ’ ”from 6D to 6J since the range of the aluminum sulfate concen—
trations in the former experiments was from zero to 3000 mg/1,
half of the value chosen for the final experimental design.
However, this difference did not prevent graphical comparison
of the different polyelectrolyte systems on the same graph.C
X Regression equation coefficient results are presented in _‘” Tables X and XI for 6A to 6C and 6D to 6J, respectively.”
tThese coefficient values are for the coded form of the
ß 4
,‘ regression equation and are not applicable to an equation
- using absolute time and concentration valuesr The regression
t equations were developed using the primary response variable,
percent reduction in total volatile solids. The form ofC
·
this equation was previously described in Table II. As wasl
seen, the equation used first and second order terms foreach independent variable as well as cross products terms
_between variables. In Table II the coefficients of each of Athese variable terms were tabulated. ·In observing the re-
sults in Tables X and XI it was seen that the first orderalum coefficient was about 10 times larger than the best
first order polyelectrolyte coefficient. It was also noted
1 102
'° 1-1- vw uw uw 01 1-1_ 01 OJ O vw . I-1Q • • • l
- · U) I I '0 °C:] S S ‘FJ 0 115Q -1-1 0 Ol Ol bb1-1 vw 1-1_ Ol 0 :1 1-1
· Q q-} • • -• •H •.-| ..Q GQ 01 1-1 I ·¤4 $-1 I 0 Ol cn ·_ ° ° > 0 O CD 0P — ' \O O 1-1 ·
- Q ’ S ¥>~1 • - .-QQ 1-1 Cwl ow 0 0 s-1 1-1 63Q 1-1 ·uw vw vw 0 I E1 ‘O 'Q •\ ~• · •
¤_
O *1** S ‘. E .'·"'¤-'I *1** _· O- 1-1 ‘ *0Er.1 · 0 0
$-1 .-Qtn vw :1* -:1* <IJ‘ ¤.1 vl av -1-1E-1 vw O 4* vw I O $-1A .-¤ • • • bb \O O:0_ 0 1-1 -1 q rncn 0 E -1-1 0Q MO -1-1 3 *0 ·nd E1 · l\ CD &D 0
. 0 vw l\ vw \O 1-1 In _E1 -+9 Q • • •‘ e-I Q 45_ Z I I I O O O .Q 4 <1-1 uw 0I-I ·xO vi E. O <D cdI-1 ·t/J vw Cwl 4* ,:I InEn E1 CN} xD 4* kO P .Q Q] • • • Q[T-I I-Ü ·°1-Q 1-I •~ Q0 2.* 1 . :>~._ +1 ·
- U 1-1 1-1 1-1 _0: 0 :1 -1 Cwl 0
Z C12 F9-1 uw O \O 0 +' Ow $-1-0 ¤-1 Ol 1~ vw 0 vw cc}-1| N Q • • • ” •«
NE4 E _ _ O bb‘ 4 ‘ ¤¤ SIQ Z .,.1 .1-1 \O· CD O *0 .Q 1-1 ow cw vw 1- 01 0 041-{ • • • Q
Z .¤ vw vw vw 4 \O O.O E I I I \D • vw G)I-1 :3 -1-1 Em 1-1 U2 + -1-1U2 4 -1 xo 1-1 1-1 -1-1 ·_ Q q-{ » • ••::1
.0 0 cw 0 0 -0L5 - -1 5 5::Eil C6
®'U •~ <D. . +9 S40 Q) · P _' • $22 Nm 1-1 0 t>>>< 0 -1 oo uw. 00 .0 0 -14-} Q • • • ‘
Q Q QQ rn.0 01 uw vw 1-1m -1-1 0 1-11-] S vw vw vw . 00 $-1 Pcn 0 C¤3 0 E 04 0 :> .0 0E1 S P"•• 'Ö •1-I1-I G)‘ GJ G9 E\ >>•F¤ • • • • 4) FU y-{‘O$ZO P4 PQ PQ 0 0 1-1E 0
Z¢'¤Z 0xO GMO 0xO Z O 41 Q-1• O O O>}Q) • • • •1-101 uw¤.1 4*01 vw¤.• ~
·0:>1>< 0>< vw>< ON0.16-0: 01:1 ¤011=Z1 ¤0Q
. 103 E
vw (D «-I uw 4* 4* O _IS N I w-I l~ l\- ID RH O _
.9 4-I 4-4 0 I-Im I I I I I I·—I
. EI 4-I 4•-lM Q ‘ *0
· 4 o oo uw 0 oo vw QI-I ·•-I N l\ O uw vw Ow vw zuä P vw (D O «—I 4* 03 4* 4*
Q 1•{ • • • • • • • |4—|
I> 22 -4 I I I 4-II Q Q 4 In
Erl P · QQ :.1 uw vw uw .-4O I-I ~ Ow uw O N 4* Ow Q) .9O N 4* vw uw -4 GJ 4* 4-I 46
$4 • • • • • • • E4Q .9 4-I ,4-I 4-IO + + .+ + I I QE:. --4-IU) *0 ‘E-I Ow 0w 4-4 <1) CD Q4-1 0 vw N O 4-I 00 ow 4-rQ vw oo vw 0 I~ 4* -4 oo m[Q Vj [·(\ • • • • • • • •,-4 _Erl O 9 4-4 4-4 4-4Q Q ·
o E vw uw QE-I P -4-I vw N Ow vw N 0w 4* •«-I
‘ z E-4 uw uw 4* uw 0 -4 N .9[rg] ‘ [~l\ • • • • • • • (QI-I_·O 9 N 4-I -4 Q0 I I I I o ·I···I U) -4-4 _Er-4 E-I · I PEv-4 Z O N vw. 0 4* cw 0Erl Q N vw I~ 4-I oo N 4-4 vw .-4
N • • • • • • • QO I-I •.9 +-4 N 4-I -4 ,N N -4 $-4
Ed I I I· Z Erl 4-I Q
O Il-4 O I1) uw O ‘ O l\ N N 0. I-I I><I Q oo vw N vw uw ow uw -.-4
E Q Q] • • •_ • • • • +34 .9 vl I EU
_ Q z I __ I-4Q oo N O vw N 00 0 Q
4-I vw vw vw uw tw- uw Q) Q4-4 • • • • • •. • Q
O .9 4* uw 4* uw Lfw 4* vw QH · E I I I' I ‘ I I I OU) :5 — ~ _ QI/J 4-IQ 4 vw vw cw N cw uw cI>Q 4-I 4-I oo I~ ow oo I- 4* .-4{5 • •I • • • • • QErl CD 0 cw 0 0 cw 4-I ct!Q 4-4 _«-I 4-I «-I -.-4
, $-4• +> . >
I I4 46 0 N N N oo vw 0 -:649 Q • • • • • •
•Erly m.9_ N N N 4* uw N N Q4-l S 4* 4* 4* ·4* 4* 4* 4*Q o an4 O 4-4E·• · .9
. • Q Q O QI ¤
.,.4 .,.4 .,.4 .,.4
I O S S S $-4I
•!¤ • • b • •p( Q Q Q Qo:*.o PQ +>•·¤ —I-¤E¤ ¤E:l -4-40 ·.-4;:4;; -.-44-4 :>ZQZ QO IUO 0O 0O SO QO QOgo 0 0 0 Z 4 4 4 *34-IQ NQ O9-4. NQ O9-4 LIwQ NQ 4*Q *0
. -‘ ¤>>>< vw>< ~·-IN ON NN NN l‘·>< l\>< G¤-IE-4Erl OE:} OC:} Om OIL:} Om Om Om 0
4 __ 104V A
that all of the alum coefficients were about the same, in-
dicating that alum had roughly the same effect on coagulation
'regardless of polyelectrolyte type or concentration or varia-
tion in the time of addition. There was no apparent con- _
V sistency in the polyelectrolyte coefficients. The remaining
consistency observed was_in the time variable. The first
order coefficient was negative for cationic polymersand-
positive for anionic polymers. i l
1 The evaluation of experiments involving one independent
variable is accomplished with no difficulty. When two inde-
_ 4 pendent variables are evaluated, graphical presentations in-” volve three dimensional figures and are difficult to grasp
mentally. Experiments involving three or more independent
·variables, as in this investigation, cannot be graphed simul-
taneously with the dependent response because four or more
l dimensions are required. Added to this complexity is the
fact that this investigation was experimentally designed
with an inherent accuracy greater at the center of the design4 and smaller at its limits., It is possible to hold two vari-
ables constant and plot the third variable against the
response variable, typically at zero concentration of the
two remaining constant variables. This procedure was not“
followed because of lesser accuracy obtained at thelimitsof
the experimental variable ranges. Instead the variables
not being compared were held constant at the center of the
F ·105
I
experimental design, zero code value. Thus, the effects of
the variation of the independent variables as they acted in
concert with constant values of the remaining two variables
were observed. ‘_‘‘
In Figure 16 the effect of the aluminum sulfate con-centration on volatile solids removal is plotted for three
7 cationic polyelectrolytes, Naleo 603,,605, and 634. From
this figure it was not clear that the maximum removal was
obtained at the aluminum sulfate limit of 3,000 mg/l. Forl
this reason the variable range was doubled to 6,000 mg/lf
and a new, final experimental design was developed for the,
polyeleetrolyte evaluation testing. It was planned to later1 retest these polyelectrolytes using the new design. Subse—
quent results showed anionic polymer superiority, rendering
. this work unnecessary. . ·In Figure 17 the effect of alum concentration on totalh
volatile solids removal is illustrated for each of seven
polyelectrolytes, Nalco 632, 610, 607, 670, 675, 672, and
674. Folyelectrolyte concentration and time interval were
once again held constant at the center of the experimental
design, the most accurate position- llt was noted that the
results obtained in Figure 16 were almost identical to that
obtained in Figure 17 over the same variable range. Removal .
of volatile solids increased linearly through an alum con-
centration of 3000 mg/l, then tended to level off to a ° l
’106
31 G Ol ·
~ ~U *4 G xo' NN • <‘D
n . r-Ä 1*4 P' , + G! ,‘ hD .Q .E Q-I ·‘ xx) :-I
.•„GQ • C/2 °
42 III G OG .::3x¤_x¤ no N E#*3 4 V 019 GC:} Q°
• • • Q •¤I Q: Q: ¤« I-: 4-IQ ··-•
\_ N N N +—> bx) E\ "‘ "“ Z ä E; E\ ~ •~ •~
· OO CD <C'ääää S5 Z ,@\ xD ND no NO I-4 +->\ (6 -:-4 •
\ o o o 6 +> zu .‘ \ 0 O O -9 :2 MO
\ :-I :-4 :-4 cd GJ UD\ Ö Ö Ö vc E 'Ö "
~Z Z Z @:3* ‘: :63
- . G:3 6 o ·‘ Ä mm c>¤•m ~4 @@6 N N\ E Cd ¤>~O\ ä Z1
_ ”\ _ _ ·•·-« a) +¤¤-I\ ° E I-I c¤N\ 5 9 ¤'*4LwJ
G-: cu o _\ · . G $4 Q\ «-I :6> 4;:
\ *6 Qu:)\ Q-gas: V
04 ·o 4-);;x v-IL) mid)\ G I ·«-46\\o¤“‘ S8
‘ N•\\ 00 ao
\ no «-IR1 A g ,-4 Q
· ‘ I S-:G G G _ G G :5 ·4* 1-*1 N N 4-I hD
„ >•'-|
SIUIOS GIVIQIOA IB'-l°xL UOVJOÜPGH &‘
A 107
1 0 00‘ f. O
1 1-1 xD 1-1/,' 0 1S
'1-1
/1 0 •' 0
* I> 0 0 +=·$-1 $3 O C6
[ 01 0 1~ 0 Ln 01 :1* 0 · ,0 1
.• 1-11 S C3 S S S S S EE SQ ° °“ °‘
¢*%°®]° -1-1 cu 1-11;; 0 X1 0 0 0 0 0 0 0 0 E + ug
1 1 1 0 0 0 0 0 0 0 0-0 01-1 1-1 1-1 1-1 1-1 1-1 1-1 0 0 - 11,0 E 1
10 0 0 0 0 0 0 -1-1 0 0 -1-1 ,-:1
1 Z Z Z Z Z Z Z +° O Q2 gg E\ 1 0 0 0 0 :0 1-1 01-> 0 E\ 1 -1-1 -1-1 11-1 -1-1 0 0 0 0 0 00 0
1 0 0 0 0 -1-1 -1-1 -1-1 0 N :1*0 1-1 1-1S! S1 0 +> 0 <c•·¤
\ 1 -1-1 -1-1 -1-1 -1-1 0 0 0 ·+¤ 0 1-= 0 ’
\‘ +-> 1-1 1: 0 -1-1 -1-1 --1 0 0 0 0 ,0
\ 0 0 0 0 0 0 0 0 0 0 1-10\ X 0 0 0 z <¤ < << 3+; cg E ·;1-¤
1 1 \, 0 00 :.1 0Q 931 Q Q Q -1: 0 0 0 00
\ ' I >>UJ FWG ¤-1 ’Ö
\ { 1 I 1-1 0 1-> 0 -1-1 0\1 ; '
0 0 01 1 1 11 Q -2111x 1-1 0 0\ · ••
-- 0 011: .0 1-1 $.11 ' _ +¤ CD :3 -1-1 0
1 0 OE -1-1 1-1 gl '· Z 010 1-1 0 111-*13]ä Ä S 1;
_1<1-1 0
‘}‘ , • Qq-{@51 0 1-1 0 1-» 1IO
1-1 -1-11->S
'Ö ""'Q Q)
*\ x 1-1-• 0gg
000 ‘—’ä 11-1 1~
1 0 1 1-10 0 0 0 0 _lf\ :1* NN N
1-1SEVETOSGIl’U?I0A JO UOVUWPGH-1-1
1:1., 1
, 108 4
maximum volatile solids removal of about 50 percent at
- around 4,500 mg/1 alum. While the general trend was thee same for all of the alum-polyelectrolyte systems, some sig-
, nificant differences did appear. The differences were more' easily seen when averages of the three types of polyelectro--
i
Ü ip lytes, anionic, cationic, and nonionic, were depicted, asin Figure 18. Particularly at low and high alum concentra—- _tions a difference in the percent removal of volatile solidwas noted with the anionic-alum systems appearing to bemore efficient. , .
F T
In studying the remaining two independent variables,
'F talum concentration was held constant at 5,000 mg/l, the F
- minimum concentration of maximum removal efficiency, and themore accurate constant concentration. In Figures 19 and 20,
lthe percent reduction in total volatile solids is depictedas a function of polyelectrolyte concentration. The remain-~°ing independent variables, alum concentration and time in- .terval, were held constant at 5,000 mg/1 and 1.2 minutes,
respectively. The polyelectrolytes behaved uniquely as.illustrated in Figure 19. However, there was a comparabletrend between the two best cationic and anionic polymers4and this feature is shown in Figure 20. As seen, increasinganionic polyelectrolyte concentration first decreased, thenincreased substantially the amount of reduction in total .
volatile solids. For the cationic polyelectrolytes, the
109
B4 O CD
O0 Q •‘ I-I go 4-4I or In GS +[
‘ sx I- :> _ ,.::· Co CO $40 C0 4:’ I · 0*0 ·•-II •~ •~ +90 . • 3I · O 4* QO Q4I _ 4-I I~ ·•·I 0 N co .I . Co 0 o 0 zu ·¤ ·¤>I-• 0 0 ·•-•0
•~ •~ EQ} Lt'\ •$;‘, :-404Ö [~ :51 ·•-IN 4-Ib.0 0;>~.O l\ 4: CDE-•
\ ‘ CO CD +9 $-4 cn- I 'ÖCÜ \ G) GJG)I 0 Q bu 0 r"I+-7I ·«-I 0 049 OE ·«-I:>,· I gg ·«-I : Q I-4 4:«-II o Q -0 0 ~ 0 0o\ · · ·•-I 0 <:.>+> 4*:,*4: 4-I qm o q 0+:
0 Q OQ ·«-I 0 >0I\ Q <:: 00 4: E 0O • :6 ·«·-I I-II-I
. \ • • • CD $-4 $4 :60 .\ ¤0 bo :>4··¤:S 4: 0 4::>,\ P P ·-•·-+·-I 0:.<: <t 000 00 >< E-—c>Q ¤•.::> ¤:> om ::4
\ NWQ 9-I\ 0 •~ 05-I\ ••
0 „-I gn-I—\ +9 E .¤ 0\ , G ZS CI3
I-I ‘I'I 'PÖ\ · O<:1 $-4 0·•-I —‘~ 3 S ä*°. _ CU
x "CS·¤ $0
- _ 0 0 ·OO OO
O «·-I OO I QO OE.. E 4-I ·r-IZ
p x ·I··’SZ‘•x\ ßq-I_ x $4::\ 01
CD I><4 x MO
_ O) 1-I •I IDQ v-I
O O O O OLn 4* NN5SPIICIS GIICIBIOA IBIICIIL umE ~«-ICr-«
11_0
O OlQ)Q EI 4 <¤ 0 -:4
I
bo ,-4 an\Drn"“ä
<° Oc:_ [II -4; +> «-I UJMO
. G5 "O\/4 ‘ 0 S-I 00mIE-4-* +> -::,-+4:· -I/I ··-•Q Q
O C)+°<D/\ +> :4-*0 .0EI QS O Q O$-I
cGO• U O>cDQ) ·r·-I Q4
O O O O O O O I-0: 0 -I-> 4-IM
O O O O O O O O*‘O¤·—-I -4-> EI GSE}I--II--II-4,-II-4,-44-4 I QI—Iq$ >„ S-4+->äääääää I S? *3
QÖ @g 40OO O O I E,-IIO C\l+¤ O <•-I-,-I
-1-·I·r-I·•··I·4-I C.) O O I 5cGO O · Q O-49Q Q Q Q ··-4 ··-* ··-* I Q:>0 0 ca ms"‘ "‘-"*
• • • • V +Q
+¤·•¤+> Q-:-I-:-·I••-I I Izu cu 0 O Q Q Q ,-4 —•-> —I=·0OU‘OZ<!l<}<fS _ 1
I _. 00 o-co„ I \r ••" ¤-I $-I Of,)‘ I ‘ X,Bl‘
°mö Ö I, ,5; 030-;,‘ QJ‘I—4v—-·II ’ 0 ¤ 40 ·.;.?:°2 I
I \ IO Q+>O 0-, o0/ ~\ «-4 $0
rg r-I\x Ü?~ [/I
I \ 00 >¤-•I . xO 4
· 9 0 Ef E 9NLOO O l¤ .
~ SPIIOS GTITIBIOA IWIOJ, UOVIÜHPSH
111 *
E ä. Q 7 —° 7 O xo
.. **7
.-
\\1 .
gg S
„ \\ '·\‘ NI--1 pg
0 E .„_;§D Q‘\ 7*277
gw
\” ¤ Q cum
x Q F4 Z ¤7
‘ 1 **® -¢m m pg?\ ¤3:'$_ mi-4 -I-3 I6+O Q ,-4 „
»‘ +9 O ·•-I Q .
\ ""'°Ö O qg G;
· 7 N LO O ¤ :3-I+¤+=·N omÖ _E *O® girl E-IO„ jjg N+° m
Ul‘· Ol Q :° JM
' O*" +·° cu ¤<1>O
•ß
I "“ “° ¤+· $7 #Z16‘*;3‘E.· . 3., O °"" $'¢•'>‘1‘ ‘ man > Sm
.·•·* O 0+: >
I Q Q U) ···* ·¤ mq;[ cu . O cp Q[ ¤ EZ Q., rg Q .
, , _§O OO <•-IÖ ¤D bi) lgI I CS
1 „·„-•g‘
.•• gEZ:
"O $3 ä $62
G} ¤ I I O, #3Ü I
ux C1 I O
~=‘ IG Q)S-ISDIIOS G{t12{0 2 0 u ' ¤
. A I 1 L 0;10np0g. ·«-•Er-,
112 ·
best removals were obtained at low concentrations. Nalco675,.an anionic polymer, clearly appears in Figure 19 tobe the best polyelectrolyte for this application.
l
IThe final independent variable analyzed was the
timeintervalbetween addition of alum and polyelectrolyte.l
Graphs of the time interval versus percent reduction in·total volatile solids are shown in Figures 21 and 22. In
I
IFigure 21, graphs for each polymer are shown. While each _polyelectrolyte system appeared to be unique, there was,in fact, consistency in the noted variation. Increasingcationic polyelectrolyte time intervals brought about adecrease in percent removals. With an anionic polyelectro— l
lyte, there was an increase in removal efficiency as the’ time interval increased. These trends were more clearly
seen when the average of the two anionic and cationic poly-electrolytes were taken as depicted in Figure 22. e
The final graphical presentation in this section,Figure 23, illustrated the removal efficiency with alum
l
concentration for various concentrations of the best poly-electrolyte tested, the anionic Nalco 675. The time interval
was held constant at its maximum effective value of 3.2 .F minutes. From these curves it was observed that acombi-nationof either relatively low alum concentration (3,000
mg/l) and high polyelectrolyte concentration (50 mg/1) or
relatively high alum concentration (5,000 mg/l) and low alum
I
113
1:„ (Nr-4I com1 ¤¤_:> IDA 1 ‘. v··I<D Ez ¤ : .1 ER ä C? :1 gzs—
···‘ • ·¤ ‘I #1E.I > E 2
, I —I \ I Pd •‘ m . Ix ' G) (UE 'Ö .‘ 6-::1 I~: -,-1 ~•O l"\r-I ,-1Ö ¤ Ö? °
«—•o ‘·<¤ ¤\ I l
• •+J Q)1 2 . :1‘ I o :>„I \\ 1 2 :1 53I
‘ 1 ··¤ cn o. \ 1 ct I>zI 1 G3 ‘I-I ° „ .
I OE wwI I 2 22 zgg. ·
°(I
°4 -1-,-• 6-1+-> .c> ·+>C\lOl‘OU'\ c.x14 I" ;:·-1--1 <1-ImIn-1 Ol\l\ l\·l‘ I „ ¤> -:1 cwgw1 oo oo' / IA, A — I +=* ¤+°·,-I-,-1-1-1-,-10 oo / I I ¤> :1 ···-IEE; ggg;--1 -,1--1 I ¤¤ eu pmco cos: :::1 / I ¤> oE·,-I-,-I-,-1-,-Io oo I I 1-I 2 ¤··-•
I 49-P·P¤·1-I ·r·*•r-I __ /_I GS ·P "C$$-Iam asc: 1:::. I :> ca cum¤Q¤z<¤ <<<¤ / /I¤¤I'
° ~1¤ ,-1 gw °Ü C? I „ 1: cßI I QH cz *32I cv =,-1
-„-1 :*,*01-4Ö IL Ö I IE E E/ ~ ,’ — E-• 1-1 ··-1:6. I I +;>es aus-
I · E ·•-1¤>/ /} •r-¢ $-1+-71 / HB cds:. ·/
<(I/ I N 5-01 :>+-1
¤ J 12 @61 E sg qä .• 1-[ '
v-I Noo 1~ xo m 4 m cu ,-1 Ol4 4 4 4 4 4 4 4 4 ESDIIGS GIIIPIGA IWIOJI ¤011CmP9H% ED
I ' -1-I_ 2.
1 114I 1+
1-1-1
1 01~
1
GN: CN@_ fg gg
1\
*1-1 • ¤>
1
E
\
+.,.4
‘‘
0 E-« -
\\
P \OI
1\
•¤
l
\\
¤>_ —1->
1
+Q Q z
1 2 Q cu In _ 1
1 \Ü ° M Q „ '¤
‘
\'
O +5"‘ 1
x*;*1 -1 "*
2 =1:- 2:
•H
x
¤-1 Q +-> .
'Ü
GQ)
1M
1wg S: cc >P‘ 1
UNO 1-Iq)
O N »·=> 2:3 ¤ 2%*
1
\pm
gg O:
I·1€ NI ¤°
fs
'_ CQ
I-\Ü
,-4 G)
1
O1 cu :21-1
·
I•$.I
G
1 1 IY> Sg
0ab
Oo
_I
·1-1QIS‘; *== 2: 1-1 E:. 1
G°‘ > E °°
I0-.-4 &Q
1 1I
<,q OO gz: I 00
1°’ :
|1 cu O1-1
Ol1 E-1 c¤M°FIQ)
1
Ö 1 z <>¤ """I 1 Q :0 @:2 1
1 (D N kg
CD PI>1-1
"‘ ¤‘ <1· IN• 1
Spr4*
"‘
uotlüüpa oHß IÜab•r-I“'*· 1
115 I
. „-Z4 LnORDPOCGO
0 c Q ;;·,~2 ° 4 S E ‘° °’é/ O
/ / / GS • ::20N_ / I! I · PCI 0+; Irn--4 Q; I! GJ GJ GJ GJ GJ ¤E‘ 4-I· / *4 SE
I ·4 4-·4 fü 4-4 4-4 Y"I· • 0 63-4-J ‘0 0 0 0 0 "¤N\ O~ #4-40$-4 $-4 $-4 $-4 $4 4-4 Ln _4—40(O (Ö 4¤ 4-*4 +¤ 4-> -4-> gw 4-4 ::4-40 0 0 0 0 0 ‘ c/10
I 4 0 0 0 0 0 cz} >>I I · 4-4 4-4 4-4 4-4 4-4 4-4GJ E EY"I —
_I I I GJ GJ GJ GJ GJ Q3 50
I I 4 >> >> >> >> {>> I>4·—I '~ CQ-4I I 4-4 4--4 4-4 4-4 4-4 $-40 0:1 -4-4I 0 0 0 0 0 0:> 00 E.0I I CI-4 ¤-4 CI-I CI-4 Q-4 P 3-4-4 :5·4-4• P 4--4;:
I\I \ 4-4 4-4 4-4 4-4 4-4 -5+-> as 430
I ¤-4 $-4 ·4-4II ~ S 2 .2 ä” 2 ä° E S24 -4-4P 0 ·4-4\ I \ N -3*RO CD CJ EICÜ OO 39-4 44-I N 0.cn
°°•• FWG4 I I GJ ·4-ISS
\I ~ I I 'IJ Q) |"IÖ
\ \ Z (I3 CU. .4 4-4sw S2\ \” x mg 4q\
E "°"?E zu wggo4* r—I
_ •
0 Q w+> ·4 l¢\ s O DOG}
s •4·I CDI 4 x\ \\' $-45-445144 .\
¤€·
IFWCJ N
O .0 O O -Ln -3* N\ INSPIIOSGIVIRIOA IP10J4UOVI-OHPGH4
·4-4EL' .
u ,A ,
116concentration(2 mg/l) yielded the best removal efficien-
cies. Nearly 60 percent reduction in total volatile solids
·was projected for either of these system combinations.
Selection of the Best Polyelectrolyte
-
Aluminum System' _ From the data in Tables X and XI it was noted that the
Aanionic polyelectrolytes Nalcolyte 672 and 675 had the high-llest coefficients for the polyelectrolyte variable,·X2 andX22. Strictly on the analysis of these coefficients Nalco-lyte 672 would have been selected as the superior poly-
electrolyte. However, as discussed previously, the variables
were studied from a system viewpoint using the regression
equation, and Nalcolyte 675 had a superior system constant,
bo, to Nalcolyte 672. This superiority is clearly seen in
Figure 19, where the reduction of volatile solids is sub-
stantially better for Nalcolyte 675 than for all other poly-
electrolyte systems. For this reason the extended testing,
including pH variation, was aceomplished using Nalcolyte 675.
Experiment Seven: pH Evaluation in Expanded Test
The purpose of this experiment was to evaluate pH as
an independent parameter for the best performing polyelectro-
lyte and to confirm the earlier results for the other inde-pendent variables. The data and response results for this
experiment are contained in Table XXIII of Appendix A. The I
regression coefficients obtained from this data are presented
117
in Table XII. In comparing these coefficients with data V
presented earlier in Table XI discrepancies were noted.
These Value differences were not due to differing experi-
mental results, but were due rather to a larger coding
v range for the same absolute dimensional parameter ranges.‘ These differences were unimportant when results were com—‘
W
„ pared graphically, but were troublesome in attempting nu-
merical correlation. To overcome these difficulties each X
independent parameter was graphed using the same techniques
employed previously in Experiment Six. Results are shown
in Figures 24 to 27. The results from Experiment Six were
included for comparison where possible.A,
Aluminum and Polyelectrolyte Concentration Variables nX The Variation of per cent reduction of Volatile solids
iwith aluminum and polyelectrolyte concentrations are shown -
in Figures 24 and 25, respectively. “The·results were Very
similar to the previous work, as expected.
Time Interval Variable_‘v·' V
‘ In Experiment Six the results with anionic polymers
indicated that an increased time interval between aluminumisulfate and polyelectrolyte addition resulted in greater
colloidal destabilization. For this reason the time inter- ·
, val tested was increased to six minutes in expectation that S”
an upper limit for time interval could be established. Asl
E 118 V
TABLE XII. REGRESSION COEFFICIENTS FOR EXPERIMENT SEVEN,EXTENDED TEST FOR EH
' Variable Name Coefficient .Coefficient_Qif aEElicable) Symbol Value
_ Equation Constant · bo 44.71Aluminum Conc., X1 bl 7.59lAluminum Conc., X12 b¥1 · -3.57Poly. Conc., X2 ' b2.
I1.35
'Poly. Conc., X22 b22 h 1•77Time Interval, X5 · bj E · 0.51Time Interval, X32 bz; 0.88pH, X4 bk 1.82.pn, X42 1;% 0.26 1Interaction, X1X2 b12 -0.87
EU
Interaction, X1X3 bl; -0.351 Interaction, X1X4 bla ’ · . ~ -0.93
Interaction, XQXB b2; E 0.69 EInteraction, X2X4 _ b24
· ‘ · -0.28Interaction, XBXE __ bzq -0.01
In 119
° :1I 0. ‘ --4x ' P· 0_ . Q 3,-4
Ü _ Q 01 QI • xD + GJI 0 0I ~ :s ¤_ I · :4:,* 4-4 0I ¤„ 46 _ I QI >- I —·0 -40I 4 •4—-IQ}OI- ¤~ Q 00*0 PI 04 :1,:-Io Q ,4 ~ CI-I-4-*· 4-> 0 0 ·uw:><1 . 1~ --4
Q Q Or··<I :5. I 0 - •-4
·4 4 ä ä E2 “ ,¤§."’
-4-4 --4 -\ +>< E ‘‘ $-4 S-4 MEN äß M äI 0 0 4-4\ ¤„ ¤„ 0Q—4¤ Q 4: --4, :4 :4 $.4-4-4:5 -4 ~ ¤0 E4 M M +> Q:'. ··-I 5• · 00-4: :1*0 cn -4\ 05Q --4 0 <:1· x -4,4445 -4-:- Q4 {3 0+>+¤ 0 .0-\ >; CD $-4 • P\ { 4-1*0Q +> ¤-4 -4-4\ ° ¤?4C:äO Qä S.. 0ug G] -40 Q m „QQ *0 *0
\ •• l"’°\O Q) *I'*\ 0 , · Q *0 -4\ „ I -4-> :1 0\\ 0 0 0 cn„ Z —•¤ 4->· · 0 >< 0° I " 3*: "‘ TJ I• . ‘ ° _ 4 OS •~ P„- · 01:/1 0 0 .
\” 4-4 r-IIx E .0 0‘
¤ 0 > •Q ,4-4-4 ux‘ x . ·r-I |$-4 9-Iklg’ E 0 04 2:>\„
Q<:: *0 00. ,4 0 -4-44-4I E x "C5 PES
\•.. . ~. ‘\
· 4 o ¤‘·}‘ .
Q Q Q Q Q ·uw :1* en 01 —
«-ISPIIOSQIVJHIOA {910414 JOUOVIOHPSH_‘‘
_ 120
Q¢ .s
29 O N +:1
\x\ UN
N cx N O
•,N l\ +9
x · J1 >„' C.\] ~ 4-4
Ö V !><! ¤- c\ HN $-4V •~ GJ•~ bl) Q 4-4
Q · x E bu GJO _ \ 4* -1-+ >~,
-4-4 • _ \t*°\•~ CI) 4-4
4* G) s $1*4 G) O:6*:1:3 x o O Q-4$-44-14-1 \ -4-1+->GJcG x 4-J • .¤ _Q-QK> \ 46 ce- 4->GJ \ $-4 JG ·4-4 _cJ,'I4'GJ 4 -4-=» ra 3QQ-4*6 ' QO O Q xOG> O"c rn
NO GJ "¤'Ö$-4 1 U GJ O
td •~GJ 1 ~4-J U)l‘\O ¤1·-44-IN 1 GJ N
· — 1-lm _ 1 ~1¤ E GJ·4-¤ -1-> :$;>+-J . 1 1 4-4
WHG r-1 •~ -4-I. GJ GJ GJ 4 . V coo GJ «1->
5 5 5+*+* / -4$-4 4-1 cßV
~·-• ··-1 5:::: 1 —1-> „¤ 4-1$-4 $-4 Q··-1aS 1 0 cü oGJ GJ -1-1 +7 [ GJ -4--I >
•
¤-4 Q-4 EGJW 1 . 0 4-1 -1$-4 UN ·‘ N N GJ IO 9-IIH
Erl rd -1··-1¤ :>„· > ow<zj—1¤o -1
, 1 OO •'¤ QO1 v-IQ-4 GJ OO
‘
'
•• - / FU •'-(P-4]
V 1 GJ g / o .1I +.> . // . Q gz
Öo / · ·r-4z // - $-4$-1J/’
1 1 G7 ¤°4‘ •
LnN N
4* NN N -4 O ON CD l\ LO UN 4*UN Ln Ln Ih UN 4* 4* 4* 4* 4* 4* SG-JSDIIOS 9'[I1PI¤A [@104], JO UGUOHPGH
121 E 4
seen in Figure 26 this upper limit was still not defined; g _
maximum removal is achieved at the six minute interval.
.pH Variable _lThe relationship between colloid destabilization and
. pH in the presence of coagulants is graphically illustrated
in Figure 27. The best destabilization results were obtained ‘
at pH 7, also near the upper limit of the experimental
design.
Overall Parameter Evaluation
In Figure 28 the per cent reduction of volatile solidsA
was studied as a function of alum concentration for speci-
fied polyelectrolyte concentration in a family of curves.
Time interval and pH were held constant at their maximum
destabilization values of 6 minutes and 7.0, respectively.
I
I122
xo. _ 01° . +
¤ ‘ VN L';:><: :~
. Ln [x. Q_ 0• g—( -. ”°°Q
° +::2, Q 2 E; _I v' ¤ .
\:1* :112 ~rm -,-1 1-1
· m. Q) • +-7
5 P1 **1\x [**7*1-* O Q-)
. X — :-1 s:0 0 .:1*. E 1 > -—= **. \\ 0 id 3· ·‘¤0+¤ x P, \\ l Q]Q •~ U}0000 \ +-1 2 Ex rl)\ · E-4 IS1-11-11-1 0 Q· +-> ~1—= ::o··c·¤¥>gg1 +-1Q 0 0 +-200 1 ~ -\ ’¤E E ao.: 1 ,2 _':·:ä2„2 X ¤¤00·»-10::20 ‘ C9 7>O-1 ¤-•E>> N ‘ ' q_‘N N¤* =-:.2.2; 1 °| O N -¤(9 . 1 1 ~ ' .2| P'B 1 01 ···*roÄ
10._ ~ C;) 1 - >. ‘ · +-1 _ .I •
ux :1* MO» :1- :1* NSIHIOS 9IV}-BIOA I?l¤:L JO uOI1?mP8H% 2. :5 .bß-„-e
· C:-1
I
« 123
. \ ° ·
Ö _ • C\l“ [‘ +
‘ ' J UN{
.N N
· _ ~l‘ C
O° I-!
UN ¤·e G_ Q)
°A
• HM Z+9 ao +£iI —>> - $-1«—I • •~ OOr-IQ) Q 9-I— $-ICUQ bl)05-am rn cz:QJQJP Q)#-*+9 Q S!Q)¤Q) r · +9
. >;·HU• jr-!
I-I o ¤: ;so<1>·o N _ _
. ¤«E C> ::1 cn·«-Io O
•¤ O *:5·‘¤—I—1I—« MOQ-I ·c ·•-II:. cu cu I-IGUN U O
· Q · - ' Q Cf)Q)G+9 Q)
+9 cd +9 Q)· cum . . ~ N ·-I ‘
'~ 'I-I:§+9 „ ::1 ·«-•_ I-IO;} ‘· +9 ·
.'5·•·I¢¤ _ ‘ ’ _ ~ ¢U@+9+9 ' ‘ Q) r-| _
tU@ ‘ •—·I Oas-Is: .¤ Z>:s+¤o ux -::3SSO
°~ vI·r-4 ¢•—I
oEOU - cdSQ- > Q ·1-IOQ) - O<¤c>.¤ ·¤ ·•—•
· Q) +9
IO O .,-4
Q) C,) $-4+9 l Gc · _ _ t>Z „
. r~ N °· _ ' I t~N
UNCN CD
‘l‘ \O UN J FN N Q) .J -3* J J J J J J $5SPVIOS BIICIBIOA {*310:1: JO UGIZIOÜPSH
0ÖÖ 0ox1
l \O + ,,-4
/ *0001 1-1 15/ [ 0 0 0 0 0 ,-1::11: >< 1-,
, 1- 1- 1- 1- 1- 0,-10 1-/ 1 ,: ä,
1/ _ 0 0 0 ,0.0 :3:*, ·- 0 Q$-1 $1 $-1 $-1 $-1 ¤-1EO 1** • S3I 1- 1- 1- 1- 1- :s • O -1 Q1 0 1[ O O 0 O O ’OEl‘ UNNI N O[ 0 0 0 V0 0 *:-1 Cr-1
_ ,-1 ,-1 ,—1·,-1_,—1 Q1-·:s •~ rnQ Q cu - 0 0 0 0 00.,;1
· §§§Z*§‘T1Ä°"’ -„*.E,” J"' \ I O O O -O O >-19 • E' U) U2
| . Q-, $-,0::: O 0 V _‘ G} -1-I ·-·| •~Q,-1,-1 ,-1 ,-1 ,-1 1-1-E 0<: ,-1
EE EFG ° <C· . bß ¤0 bl) bi) bu-,-1c:SO -1-* $31 •
1 EVE E E E 1; SN _· I · cu 4· nc oo 0 E: $-1 --1 • '
I -1 01 rn UN -,-1011-1 1- *0 391\ E-100 :.1 0
00 ·¤ mwQ 9 ¤o Q -10 0:: ·:s
‘ [ •• ®¤ G) ·1·¢•~1
\ I Q) UNO +9 1-I •
\ \ 1 1- O N om __
\ 1- « 00\\ · cu 0 ,-1:;
\ C><1-1 ,-1 -,-11 \\ -1-1 .0
OS 05 1.Ul*\x\ cum‘I ä E >°0_ O
°°¤° ::1 -1 :1-1,-1
1 OLG I' 1 \‘\ 0: ·¤ ::2
· _ \x -+1-1 0 00 ._ N C><! 0 -,-1-,-1-1 -1-<1-1
1 v _
\'·"‘¤
\\x ää”.
1 V. \ .
° O N1-1 [ (D1 N
UNOV
UN OI UN O UN O‘ CDno LO UN UN -:1* 4* rnrnSPIÄIOS
GIIÖIFIOA U1UOVIÜIIPGH1‘ gt,
.f
V. DISCUSSION OF RESULTS ·
The following sections contain a discussion of the
results of this investigation.‘
gt- .Nature of Spent Vegetable Tannin Liquor °
l ” In reviewing the literature on the treatment of spent
Vegetable tannin liquor it was immediately evident thatin—“ ·
dividual investigators were unable to insure the constancy 1
of the quality of the vegetable tannin waste. This situation
led to disclaimers of applicability of the data and results
. from one plant to another and thus to a serious impairment .”’
of the significance of a given investigation. The large ,i
concentration Variations of the tannin waste reported in
g _other works were also observed in this investigation. TheO
tannin solutions at the plant in Pearisburg were preparedl. from weighed mixtures of quebracho barks, and as the per
cent of tannin Varied considerably in concentration forl
‘
bark even from the same tree, it was probable that the tan-
nin concentration of bark mixtures Varied considerably.‘ ° ~
a Tannin concentration measurements at Pearisburg were made
a _with a hydrometer device termed a barkometer, and, since
the extracting liquid was well waters of varying solids
concentrations, some tannin concentration deviation was
expected. The tannin barks were obtained from numerous
· locations and could be expected to Vary in substance.
125 ·
12 126 1
OFinally, operator error in an old, poorly lighted and inade-
quately equipped plant was to be expected, at least occa-
isionally. The remaining factor affecting the tannin liquor
waste was the raw hides. Both the nature of the cowhide andits treatment prior to tanning affected tannin absorption.
Combinations of some or all of these reasons led to about a10 per cent day-to-day Variation in the total solids concen-
tration of tannin waste, as noted in Experiment One. This
level of Variation was prohibitive in attempting to achieve
reproducible results. From a practical, long—term treatment
—Viewpoint, the apparent Variation of average total solids
was critical. The total solids concentration average de-
creased from a level of 30,000 mg/l to 28,000 mg/l over a° four month period. One year earlier, Edwards had reported
levels of 38,000 mg/1. · ·A
A In other characterization tests it was learned that a
given sample of the tannin waste decreased considerably in-‘
volatile solids concentration after a period of two to three2
days at temperatures of 20 and 25°C. This change in Vola-
tile solids concentration was likely due to microbial action,
and was graphically presented in Figure 9. Stability was
achieved by cooling the waste to temperatures near 0°C. At
this temperature the bacterial activity was reduced, pre-
Venting waste deterioration. _
¤
· From both the chemical process and microbial action P
viewpoints a single massive sample and cold storage were
required to obtain a waste of constant parameters. In
other words, the only successful method of consistently
y procuring an unchanging waste was to have a sufficiently
large volume of waste from~which smaller test samples could I
be taken. This technique was followed, but even thenP
stringent mixing procedures were necessary to secure the
desired waste quality. Through this approach an experimen—_
tal design using absolute parameter dimensions rather than
.parameter-waste ratios was possible. This technique made-
possible a methodical laboratory approach, with reproducibleV
results. Previous work was limited by inherent error in
waste condition to rough estimations of gross independent_° variables, with variables of lesser importance then being y
impossible to study.it. l · ‘ - l
lq
Gravimetric Relationship of Treated Waste
Gravimetric work was undertaken early in the experie
mental phase of the investigation, primarily to determine
the best gravimetric index to measure colloidaldestabili—zation.”This work was accomplished. The total volatile
solids was determined to be superior to the other possible
_ gravimetric indices from the standpoint of both accuracy
and experimental ease. In this work three systems of .
treated waste were evaluated for both total and suspendcd
128
solids concentrations. These three systems were the com- ·
pletely mixed waste, the clarified portion and the settled
sludge. Data obtained in Experiment Three showed that much
of the Al2(SO4)3-18H2O added was remaining in the dissolved_ state. Of a 1500 mg/l dosage about 40 per cent was found
in the clarified section of the treated waste; the remaining
A 60 per cent was coagulated with the destabilized colloid.
While the precipitated aluminum had probably taken the
form of a flocculant aluminum oxide prior to precipitation,U
this structure seemed an unlikely species for the portion
remaining in the clarified section. As the fixed suspended“
solids concentration for the clarified section of treatedE
waste did not change from the raw waste value of around ‘
. 450 mg/l, it was evident that the aluminum ion was not being
· trapped on the filter cake, but was passing through the
- porous cake structure. ”This movement would not seem mechanie a
cally reasonable for a flocculant structure as described by
Riddick and Matijevic in the literature. These observations
_ were obtained from experiments designed for other purposes,
but in retrospect were important in providing insight into
the behavior of the aluminum ion in solution. VU
Response Variable Evaluation n
There were two phases of testing the response variables.I
lIn the first phase repeated testing on the same treated waste _
allowed calculation of the coefficient of variation. As
U129
listed in Table IX this index permitted comparison of theaccuracy of the various dependent variables._ As noted
earlier the per cent reduction of total volatile solids was
selected as the primary response, but others would have_ been possible choices with little or no loss of accuracy._ i
During the testing of the 10 aluminum—polyelectrolyte
systems the second phase of the response variable evaluation
y was accomplished. Besides the primary response at least one
additional response was measured.. The additional responsesU
were then compared with the primary response in a series of
graphs, Figures 12 to 15. While the eurves approximating‘ 2 the data were drawn linearly, the values of the slopes were
1 not usually one. i'
Color Removali ”T·n '
i_It·was seen in Figure_14 that the tannin concentration t
removal was more rapid than the reduction of the total
volatile solids. At 50 per cent reduction of total volatile
solids a corresponding 75 per cent reduction in tannin con-
centration was obtained. From a color treatment standpoint
this relationship was critical and fortuitous. As described» in the literature review the purpose of chemical waste
treatment is often directed at color removal; biological
treatment at organic solids reduction. While chemicali
xtreatment in this investigation generally only removed yaround half of the total volatile solids it was much more
130C
successful in color removal and would therefore be appli—4
A° cable to a combined chemical-biological sequential treatmenti
system. . F
~ Chemical Oxygen Demandh
JI
The value of the slope of the chemical oxygen demand-
per cent reduction of volatile solids curve was about one. "
This meant that the volatile solids did not change.as to·
their oxidation requirements as a function of extent of
removal from solution. While this seemed normal at first
. glance, the more rapid removal of the tannin compounds withi
total volatile solid removal meant that the composition ofn
I_ the total volatile solids was changing. Thus, it would not
have been unexpected to have noted a change in the chemicalJ J oxygen demand as the total volatile solids concentration
decreased. I l -.Sett1ed Volume and Filtration Time
l-
There were two curves with negative slopes, those de- ‘
scribing settled volume and filtration time, and they were
illustrated in Figures 13 and 14, respectively. LaMer’s J
J theoretical prediction on filtration time using polyelectro-
lytes was applicable to these tests. Filtration of the.
clarified phase was more rapid as volatile solids removal ‘
increased. As there was less suspended solid material to
be filtered the filter cakes were thinner, resistance less,
1131 L I
and filtration time shorter. The polyelectrolyte molecule —
remaining in the clarified section tended to prevent cake° compression by maintaining a porous structure of stable
channels. The decrease in volume reduction as the volatile
n solids increased was straightforward: as more of the °
volatile solids precipitated a greater coagulated volume
was acquired.’ . 2
l
Turbidity .
__ The transmission turbidity data taken in this investi-~
gation was simply not accurate enough, as was evident in
the data in Table IX and from the graph in Figure 13. In
Figure 12 turbidity as light scattering was presented, and
this data correlation was also very poor. However, there‘· was nne strong advantage to the light scattering data, that
being its facility of operation. No dilution was necessaryS
2and the data was taken immediately. The instrument used
was an expensive Hach nephelometer; Black and Birkner have W _
reported more satisfactory results using the more sophisti-
cated Brice-Phoenix Light Scattering Photometer, Universall
Series 2000. Future turbidity data acquisition is
recommended with an instrument of equivalent caliber. W
2 lIndependent Variable Evaluation
A discussion of four independent variables studied in
this investigation follows.
· l132 ·
Aluminum Sulfate Concentration .
The reduction of total volatile solids showed anexpected relationship with aluminum sulfate concentration
and was illustrated in Figure 17 for the different aluminum
sulfate-polyelectrolyte systems. Figure 18, a graphical·
presentation of the average results of the individual poly-
electrolyte types was easier to follow. Beyond some critical
value of aluminum sulfate concentration, additional aluminum‘ sulfate was searching for smaller and smaller quantities of
_colloidal waste material. The result was, of course, al
decrease in the marginal efficiency of organic matter re-
moval with increasing alum concentration. It was evident in ‘
Figure 18 that the slope of the curve remained relativelyi
constant up to 3,000 mg/1; beyond this value increasingly
.smaller amounts of volatile solids were removed for constant
aluminum sulfate concentration intervals. Thus, the maximum
removal efficiency occurred at an aluminum sulfate concen—
tration of 3,000 mg/l. As the initial total volatile solids
concentration is 21,000 mg/l, the optimum total volatilel
solid—aluminum sulfate ratio was 7 to1.l
The slight dip in
volatile solids reduction beyond an aluminum sulfate con-
centration of 4,800 mg/l may be due to entrance to the re-i
stabilization zone described by Teot and others. In com- .
paring the relative importance of the independent variables
it was obvious from Figures 18, 20, 22, and 27 that the ‘
I
_ .133 ” r
aluminum sulfate concentration was the most important. This
variable affected the per cent reduction in volatile solids
over the maximum 50 per cent range. The maximum effect of
A any of the other independent variables was over a range of
about 10 per cent reduction in volatile solids.
Polyelectrolyte Concentration and Type
The behavior of the system in terms of the polyelectro-
lyte concentration and its effect on colloidal destabilizationis more difficult to explain. The various polyelectrolytes
manifested an individual behavior pattern, as was illustrated
‘Vin Figure 19. In these curves there was no consistency be-
· tween curve shape and polyelectrolyte type. However, in
collation of the two best performers of each polyelectrolyte” type in Figure 20, the anionic type was superior in per-
U
formance to the cationic type. 'Maximum colloidal destabili— F1
zation was attained at highest polyelectrolyte concentrations,
but the decreasing reduction of volatile solids at
thecentralaluminum concentration was not understood. ThisW
performance was later repeated in the extended testing _
sequence as shown in Figure 25. The superior performance of
these anionic polyelectrolytes was surprising in that it hadV
been expected that the cationic polyelectrolytes with their
positive charges would be more effective in destabilizing
the negatively charged colloidal particles. In the design
of the experiment six cationic polyelectrolytes were tested p
134)
while only three anionic polyelectrolytes were evaluated. W° The reasons for this anionic superiority will be considered
U6
later in the mechanism section.
· Time Interval A I_ · The data for the differing aluminum sulfate-
polyelectrolyte systems was consistent according to poly- lelectrolyte type. As was evident in Figures 21 and 22 there
_.
was severe deterioration of the ability of the cationicpolyelectrolyte to destabilize a colloidal system when it
was added after the aluminum sulfate. When the cationic
pl
polyelectrolyte was added prior to aluminum sulfate it wase
ly °ab0ut five per cent more effective in reducing the volatile
solids than when added afterwards. This phenomenon was' qualitatively true for all cationic.polyelectrolytes. The
opposite effect was noted for anionic polyelectrolytes; theU
longer the delay in adding the polyelectrolyte the better -
the.per cent reduction in volatile solids. In fact, con-I
‘ sideration of Figure 22 made it clear that the anionic
polyelectrolyte really had no effect in helping or hindering
colloidal destabilization until after at least a 1.2 minute’time interval. As the polyelectrolyte chosen for the ex- I
tended test sequence was an anionic polyelectrolyte the
final experimental design doubled the time interval to be
investigated to a six minute range (0 to 6 minutes). Un-fortunately, as seen in Figure 2T, this range was still
135;
not sufficient as the maximum volatile solids removal wasattained at the longest time interval.
‘
It was interesting to note that the magnitude of the
removal of volatile solids due to the time interval was ‘
around five per cent, the same value as attained by the
polyelectrolyte concentration. It was equally important,i
_ therefore, to choose the proper time of polyelectrolyte
addition as well as the proper concentration. Previous
investigators who tested various polyelectrolytes in con-
junction with aluminum sulfate at a constant time interval
were certainly not obtaining an accurate estimation of the
true effectiveness of their polyelectrolytes.
_ The results obtained for the pH variable were presented
in Figure 27. It was noted that volatile solids removal Iincreased as pH increased from 5 to 7. These results were
limited to an anionic polyelectro1yte—aluminum sulfatep ‘
system which precluded definite proof as to whether the 0
aluminum or polyelectrolyte substances were most affected
by the pH. However, in the region between pH 5 and 7 the
hydrolyzed metallic state of the aluminum ion is the same,
Al8(0H)204Ü and previous investigators have described de-
creasing effectiveness as the pH increases above the range
·5.6 to 6.0. It is believed that the increased effectiveness
observed in this experiment was partially due to changing
polyelectrolyte structure with increasing hydroxyl ion ·
concentration. With a negatively charged anionic poly-
electrolyte, increased hydroxyl ion migration into the
polyelectrolyte membrane would cause a more strictured
polyelectrolyte conformation. This increased, more open
rigidity would lead to a more effective net for holding
the hydrolyzed aluminum ion flocs in position. —A
Another consideration was the manner of pH control.
As the aluminum sulfate was introduced to the tanninA
waste, hydrolyzation occurred and the pH attempted to'
drop. This drop was prevented and the pH maintained con-
stant at a predetermined value by the addition of a con- IA
centrated solution of potassium hydroxide. The higher
the required value of constant pH, the greater the quantity
-of the potassium hydroxide that was added. The maximum‘ quantity added was around six grams, and, while the potas-
sium ion has a charge of only plus one, this large positive
ion eoncentration undoubtedly helped in destabilizing the _
colloidal waste.A
_A
_
Optimum Values of Independent Variables for ColloidalDestabilization—EconomicAConsiderations
A The final experiment was undertaken to establish the A. best values for treating the colloidal tanning waste. The
A
best values for the time interval and pH were between three _
and six minutes and seven, respectively. The aluminum
A137
sulfate—polyelectrolyte concentration combinations were. ‘ examined in Figure 28 for the best polyelectrolyte, Nalco-
lyte 675. The greatest removal, 65 per cent volatile solids
reduction, occurred at an aluminum sulfate concentration of ’h
_ 5010 mg/l and a polyelectrolyte’concentration of 50 mg/l. ·j · It was interesting to note that 60 per cent removal was 1
obtainable using the lowest polyelectrolyte concentration ‘ .
of 2 mg/l and an aluminum sulfate concentration of 4510
mg/l. The much higher aluminum sulfate concentration made1 this concentration combination unfavorable economically. At
q lower treatment levels, 56 per cent reduction in volatile _~” solids was obtainable using 2 mg/l polyelectrolyte and 5010
img/l aluminum sulfate. At these conditions there was 8.5
F per cent less removal but only four per cent of the poly-6 ”electrolyte used at 50 mg/l levels was required. As the
Ä polyelectrolyte is expensive, $1.00/pound, these latter '
values of 2 mg/l polyelectrolyte and 5010 mg/l aluminum. ~
sulfate probably represented the best compromise of degree ‘
of treatment and expense. It must be remembered thatV·
chemical treatment is generally succeeded by biologicall.
treatment which is capable of less expensively processing
·marginally removable colloidal matter. The major purpose
of chemical treatment is color removal. Extrapolation of g~ the curve in Figure 15 estimated 85 per cent reduction of
color as measured by tannin concentration at 56 per cent
T T138
reduction in volatile solids. The removal of the remainingT
15 per cent color would require the much higher 50 mg/l
polyelectrolyte coneentration, and would, therefore, be
very expensive indeed. TT
TT
q Theoretical Consideration of Polyelectrolyte MechanismT_
T There were three significant facts upon which to base T
a comparison of the mechanisms of anionic and cationic4
Apolyelectrolytes.
*1)The anionic polyelectrolytes gave
rsuperior performance to the cationic polyelectrolytes when
the other independent variables were held constant.
VT
2) Anionic polyelectrolytes were only efficient in destabi-T
T lization when added after the aluminum sulfate. 3) Cationic
polyelectrolytes were only efficient in destabilization‘ when added before the aluminum sulfate{ T
‘ It would definitely appear that entirely differentT mechanisms are functional as a result of the anionic and T T
cationic polymer additions. The negatively charged anionicT
Tpolyelectrolytes largely form molecular bridges between
positively charged hydrolyzed aluminum floc particles bring-
ing about a larger, more dense floc which enhances agglomera-
Vtion of the colloidal material in the waste and results ina floc that is rapidly removed by sedimentation. Bridging
is an effective mechanism in situations where the polymer
is added after the aluminum ion has been allowed to form
insoluble hydrous oxides. The time of addition of anionic
139 3 I
polyelectrolyte is clearly critical in the colloidal de- hAstabilization process. ·Sufficient time must be allowed forthe formation of aluminum hydrous oxide floc and for the
'subsequent colloidal flocculation.‘ Only then should the yanionic polyelectrolyte be added to form the larger, quickerA
W settling particles by bridging. Earlier addition of the
polyelectrolyte results in premature bridging with resulting
loss of destabilizing power of the aluminum ion. Unfortunate-ly, this time interval is probably a function of several
factors, for example, rate of mixing, aluminum concentration,
waste concentration and others, and is therefore difficultV
to pinpoint. A positive fact is that it should not make
much difference if the polyelectrolyte is added later than
the least possible time interval.
· As noted, the cationic polymer gave optimum results
V when added prior to the aluminum sulfate. The mechanismresponsible for this effect is probably electrostatic
neutralization of the negative charge associated with thel
colloidal matter in the wastewater. As both components of
a cationic polymer-aluminum ion system have the same charge,
removal efficiency will be greatest for polyelectrolyteW addition while the negatively charged colloidal concentration
is still high, that is, before the competing aluminum ion
is added. If the cationic polymer is added after the alumi-
num sulfate, it is apparently doing little more than what
140 ‘ T
additional aluminum ion could accomplish. The physical
, appearance of the destabilized colloidal waste for cationic
polyelectrolytes at different time intervals was very
·striking. When the cationic polyelectrolyte was added prior
to the aluminum sulfate, dark brown coagulated masses were1
quickly evident; later addition of aluminum sulfate produced
a different, lighter brown precipitate. When added simul-
. taneously, the two coagulants were obviously in competition
as they produced two different types of precipitates. There
was less evidence of the precipitate produced by cationic
„ polyelectrolyte than in the previous test. When the cationic
polyelectrolyte was added following the aluminum sulfate
there was no indication of any cationic polyelectrolyte
linduced destabilization. In summary, all results suggestedlthat
aluminum ion-cationic polyelectrolyte system was one
4 of rivalry between two c0mponents.' _ ‘
The material presented in the previous two paragraphsin
is the prevailing mechanistic theory applied to this investi-
gation in a manner suggested by the data and results.
Another exposition can be made concerning the anionic poly-l
electrolyte-aluminum ion system. The synergism of the system
is clear. Anionic polyelectrolytes, alone, are ineapable ofA
p destabilizing colloidal wastewater. In conjunction with p
the aluminum ion, colloidal removal is more efficient than
with the aluminum ion alone. A physical picture of the
·141
bridged aluminum floc would be similar to that of a net,·or a three dimensional net with openings offset. At high
polyelectrolyte concentrations the negative charge should 44
become more significant than just from a bridging view-
point. It is known that addition of negatively chargedeparticles to a negatively charged colloidal system further T
stabilizes the system through electrostatic repulsion.
Similarly the highly charged negative polyeleotrolytes yshould repel the negatively charged colloidal particles,
A
even if the polyeleotrolytes are partially neutralized in
the net structure. The polyelectrolyte-colloid electro-
static repulsion will reflect the sol particles closer to°„ the aluminum ion. This has the effect of increasing the
counter ion concentration in the sol particle atmosphere.and reducing the double layer.. The same result may be con-
sidered from another viewpoint. In the reflection of theh
jion away from the polyelectrolyte surface, the effective
volume available to the sol particles is reduoed, and the Veffective counter ion ooncentration thus increases. This
leads to increased colloidal destabilization. Bridging,
provable by electron and microscopy and indicated by results
of the investigation, is of obvious importance in explaining
the mechanism of coagulation in this work. It does seem —
reasonable that the nature of bridging should at high poly-
eleotrolyte concentration lead to related destabilization "
phenomena. _
142
~ Jar Testing Procedure .· Coagulation testing is not always comparable because
—’ there is no standard jar test procedure. Containers, mix-ing speeds and time lengths, coagulant form (solid or
UV solution), type of stirrer, and other factors are not stan-
· dard. There is a need to establish such a procedure. Test-ing of given variables could then be undertaken, introducing.planneddeviations from the standard procedures, and reportedas such. Comparison of similar colloidal experimentation
would then be much more valid.~(
·Limitations of the Investigation l”
.3 · Theoretical considerations made in this investigationwere hampered by the limitations imposed by any applied H
- research problem. The natural waste system was ever vary-
ing and impossible to reproduce synthetically. The tannin
"waste had a high salt concentration, and pH control caused
variance in ion concentration. For these reasons the timeinterval variable should be retested using a synthetic,
·W_
reproducible colloidal system such as the Minusil colloid
described by LaMer and his coworkers.· Massive aluminum ion concentrations·make the study of ‘other independent variables very difficult. Further in-
vestigations of aluminum-polyelectrolyte systems should
incorporate lower aluminum concentrations.
. i VI. CONCLUSIONS
The principal conclusions drawn from this work are as
follows.° l
1. Total volatile solids, tannin, and COD removal _.
l_ are suitable response parameters for spent tan
liquor waste coagulation testing. _ ‘
2. Color removals from spent tan liquor wastes ap- _proaching 90 per cent are possible using an
Caluminum ion—polyelectrolyte chemical coagulationA
system. ~
p:l'
3. A seven to one total volatile solid—a1uminum C·
« sulfate ratio gives maximum removal efficiency
C of total volatile solids. 1‘_C ”
4. Anionic polyelectrolytes give better volatile
solids removal results than cationicpolyelectro—l
- _ lytes when used in conjunction with aluminumW sulfate. -
s 5. The best polyelectrolyte tested was a long chain
C anionic polymer, Nalcolyte 675.
-6. The optimum time interval between addition of _
_” ll
alum and anionic polyelectrolyte is a minimum ofi
three to six minutes, and anionic polyelectrolytes
· · are ineffective when added prior to or simultaneous—
ly with the aluminum ion.
143 V
1 144
n 7. Cationic polyelectrolytes give best performance j
when added prior to the alum.f
8. The optimum pH for the aluminum—anionic polyelectro-
lyte system was 7. ·9. The major mechanism of destabilization of the
anionic polyelectrolyte was bridging of the
aluminum ion.floc. ‘ 1
10. The method developed in this investigation was
j successfully applied to the treatment of spent.
vegetable tannin liquor.'‘ ‘
5 ‘Vi
, · ° VII. BIBLIOGRAPHY °
(1.Adamson, A. W., Physical Chemistry of Surfaces, 2nd Ed.
Interscience Publishers, New York, N.Y., p. 221 (1967)._ 2. Badger,_W. L. and Banchero, J. T., Introduction to Chemi-
cal Engineering, McGraw—Hill Book Co., Inc., New York,
N.Y., p. 561 (1955). ,3.”Barkley, W. A., King, P. H., Myers, R. H., and Randall,
C. w., "Chemical Treatment of Spent Vegetable Tanr
_Liquor", Proceedings of the Twenty—Fifth Annual Purdue(
,
University Industrial Wastes Conference. (1970) ,4. Benoit, R. E. and Starkey, R. L., "Enzyme Inactivation
as a Factor in the Inhibition of Decomposition of
Organic Matter by Tannins", Soil Science. Vol. 105,pr
Np. 4, pp. 203-8 (1968).5. ‘Benoit, R. E., Starkey, R. L., and Basaraba, J., "Effect,
' of Purified Plant Tannin on Decomposition of Some
Organic Compounds and Plant Materials", Soil Science.
Vol. 105, No. 5, PP, 153-8 (1968).
6. Berks, A. A. and Schroeder, W. C., "Determination oflv
_ Tannin Substances in Boiler Waters", Industrial and
5 ,Engineering Chemistry, Analytical Edition. Vol. 14, _,6 9 p. 456 (1942)., V , ·
re
145
(i1467.
Birkner, F. B. and Edzwald, J. K., "Flocculation of° Dilute Clay Suspensions with a Nonionic Polymer", Pre- ‘
lsented at the National Meeting of A.W.W.A., San Diego,
Calif., May 1969. .l„ 8. Birkner, F. B. and Edzwald, J. K., "Coagulation with
l 4Inorganie and Polymeric Coagulants", Fourth Annual
Rgpgrt, Water Resources Research Center, University of' g
Maryland. pp. 57-85 (1968). ·P
9. ibid. p. 77. (
10. Black, A. P., Discussion of Paper "Chemioal Aspects of
_/ ( Coagulation" by Stumm, W. and Morgan, J. J., Journal _·‘ ( A.W.W.A. Vol. 54, No. 8, pp. 992-4 (1962). ·
l11. Black, A. P., Birkner, F. B., and Morgan, J. J., "The
_Effect of Polymer Adsorption on the Electrokineticl VStability of Dilute Clay Suspensions", Journal of Col-
loid Interface Science. Vol. 21, pp. 626-4S (1966).
12. Black, A. P., Birkner, F. B., and Morgan, J. J., "De-
~ stabilization of Dilute Clay Suspensions with Labeled
Polymers", Journal A.W.W.A. Vol. 57, pp. 1547-60 (1965).
13. Black, A. P., Breswell, A. M. and Black, A. L., "Review' of Jar Test", Journal A.W.W.A. Vol. 49, No. 11, p. 1414(1967).14.
Black, A. P. and Chen, C., "Electrokinetic Behavior ofAluminum Species in Dilute Dispersed Kaolinite Systems",
Jgurnal A.W.W.A. Vol. 59, NO- 9- PP- 1173-83 (1967).
‘p I
U Iy 147 ‘
15. ,Black, A. P. and Vilaret, M. R., "Effect of ParticleSize on Turbidity Removal", Journal A.W.W.A. Vol. 61,
No. 4, pp. 209-14 (1969). '6
16. Black, A. P. and Willems, D. G., "Electrophoretic Studies
of Coagulation for Removal of Organic Color", Journal
A.W.W.A., Vol. 55, No- 5- Pp- 589-604 (1961).
17. Burton, E. F., Phil. Magazine. Vol. 11, No. 6, p. 425
(1906); Vol. 17, p. 583(1909).18.
·Chakrabarty, R. N., Khan, A. O., and Chandra, H., "Acti-6
vated Sludge Treatment of Tannery Waste", Journal of
the American Leather Chemists Association. Vol. 62,
„ NO- 4, pp- 733-47 (1967)-ß19.Chapman, D. L., "A Contribution to the Theory of Electro-
(4 3 capillarity", Phil. Magazine. Vol. 25, No. 6, p. 475-81
- (1913)- - g20. Derjaguin, B. V. and Landau, L. D., Acta Physiocochim.
‘ (U.R.S.S.) Vol. 14, pp. 655, 663 (1941); Journal Experi-
mgnjal Theoretical Physics (U.S.S.R.). Vol..11, Ap. 802 (1941). ( ‘
21. Dixon, J. K. and Zielyk, M. W., "Control of Bactorial ~
Content of Water with Synthetic Polymeric Flocculants",
Environmental Science and Technology. Vol. 5, No. 6,
pp- 551-558 (1969)- ~ 3 · —
6 ‘ ” (148
Q
22. Dole, L. R., "Surface Properties of Hydrogels Resulting
from the Treatment of Pulp and Papermill Effluents",
Unpublished Master's Thesis; Department of Chemistry,
Lehigh University, Bethlehem, Pa. (1969).
23. Eckenfelder, W. W., Jr., Industrial Water Pollution‘ Control. McGraw-Hill Book Co., Inc., New York, N.Y.
(1966). 4. 2 p24. ibid. pp. 42-43.
(- '
25. 'ibid.. pp. 87-88.
26. 1b1¤. p. 95. .J I.
9 .27. Edwards, W. H., "Characterization and Treatment of Spent
Vegetable Tan Liquors", Unpublished Master's Thesis, 1
· Department of Civil Engineering, V.P.I., Blacksburg,.
va. (1968). p_28.ibid. p. 1. _
29. 1b1d. p.5-6.30.ibid. pp. 23, 26. · i
31. Eye, J: D., "Characteristics and Treatment of Tannery(
Wastes", Presented at the Industrial Wastes
Conference,Universityof Wisconsin, Madison, Wisconsin, March (1968).
32. Eye, J. D. and Aldous, J. G., UAerobic-Anaerobic Treat-
ment of 8pent Vegetable Tan Liquors from a Sole Leather‘ Tannery", Proceedings of the Twenty-Third Annuai Purdue „ _
(University Industrial Wastes Conference. pp. 4,6 (1968).,
p 149 ( A
33. Eye, J. O. and Graef, S. P., "Literature Survey on Tan-nery Effluents", Journal of the American Leather A
Chemists Association. Vol. 62, p. 194 (1967).
_‘A
34. Flory, P. J., Principles of Polymer Chemistry. Cornell
University Press, Ithaca, New York, p. 585 (1953).35. ibid. pp.629-30.36.
Freundlich, H., Z. Physik. Chem. Vol. 73, P. 385 (1910).
37 p Guoy, G., "Sur la Fontion Electrocapillaire", Journal ·
Physics, Vol. 9, No. 4, p. 457 (1910); Ann. Phys. Vol.
7, No. 9 (1917). ,„38.Gurnham, C. F., ed., Industrial Wastewater Control. A
Academic Press, New York, N. Y., pp. 395 (1965).
39. Gustavson, K. H., The Chemistry of Tanning Process.
_.. Academic Press, New York, New York, pp. 142-3 (1956).
40. riuid. pp. 166-9.’
41.A ibid. p. 192._A
10
42. Hach, C. C., "Introduction to Turbidity Measurement",
Booklet produced by Hach Chemical Company, Ames, Ohio.
43. Hamaker, H. C., "The London-van der·Waals AttractionABetween Spherical Particles", Physica. Vol. 4, pp. 1058-72 (1937). ‘
44. Hannah, J. A., Cohen, J. M., and Robeck, G. G., "ControlTechniques for Coagulation—Filtration", Journal A.W.W;A. °
Vol. 59, No. 9, PP. 1149-63 (1967).
150
45. Hem, J. D., "Complexes of Ferrous Iron with Tannic Acid",(
U.S.G.S. Water Supply Papers, 1459-D (1960).
46. Hommon, H. B., "The Purification of Tannery wastes",
U U.s.P.H.s. Bulletin No. 100 (1919). U· 47. Howes, F. N., Vegetable Tanning Matenlnln. Butterw0rths(
Scientific Publications, London, England, p. 2 (1953). ‘
48. Johnson, S., "Industrial Waste Disposal from the Tanning”
Industry to the Streams of Virginia", Unpublished
Bachelor of Science Thesis, Department of Chemical
_Engineering, V.P.I. (1941). ‘
·49. Kawamura, S., "Removal of Color by Alum Coagulation. ·(( (
Part I", Water and Sewage Works. Vol. 114, No. 7,(
(p. 282 (1967).
U 50. Kawamura, S. and Hanna, G. P., Jr., "Coagulant Dosage
Control", Industrial Water Engineering. (Vol. 4, No. 4,- pp. 21-4 (1967). — 6 6 (
51. Kawamura, S., Hanna, G. P., Jr., and Shumate, K. A.,(
("Application of Colloid Titration Technique to Floccu-
lation Control", Journal A.W.W.A. „Vol. 59, P. 1003
(1967)- ,(
52. Kawamura, S. and Tanaka, Y., "Applying Colloid Titration
to Coagulant Dosage Control", Water and Sewage Works. ·
Vol. 113, No. 9, p. 348 (1966). . —
53. King, P. H., Lecture notes for C.E. 522, Sanitary Engi-
neering Design, V.P.I., Spring Quarter (1968).
_I
Ä 151 '
54. King, P. H., Edwards, W. H., and Randall, C. W., "Char-
acterization and Treatment of Spent Vegetable Tan
' Liquors", Proceedings of the Second Mid-Atlantic Indus-
trial Waste Conference. Drexel Institute of Technology,
(1968).(
55. ibid. p. 9. I
56. King, P. H. and Randall, C. W., "Chemical-Biological
Treatment of Textile Finishing Wastes", Paper presented
at the Nineteenth Southern Water Resources and Pollution
~ Control Conference, Duke University, Durham, North( Carolina, April (1970).57. ibid. ‘pp. 8-9. I
58. Kruyt, H. R., ed., Colloid Science. Elsevier·Publishing(
_ Company, Amsterdam, Netherlands, p. 280 (1952).
59. ibid. p. 283. _
60. ibid. p.291.61.ibid. ‘pp. 302-3. p
62. ibid. p. 306.6
I -
· 63. LaMer, V. K., "Coagulation Symposium Introducti0n“,
Journal of Colloid Science. ·Vol. 19, p. 241 (1964). '
64. LaMer, V. K., and Healy, T. W., "The Role of Filtration
· in Investigating Flocculation and Redispersion of
Colloidal Dispersions", Journal of Physical Chemistry. °
Vol, 67, p. 2417 (1963)._
152 r
65. LaMer, V. K. and Healy, T. W., Review of Pure and Ap-
plied chemistiy. Vol. 15, p. 112 (1965).
66. LaMer, V. K. and Smellie, R. H., Jr., "Theory of Fl0ccu—U
lation, Subsidence and Refiltration Rates of Colloidal
Dispersions Flocculated by Polyelectrolytes", Clay
Mingyal Society, Clays and Clay Minerals. Vol. 9,
pp. 296-7 (1960). _
. 67. ibid. pp. 500-1. J’
68. Levey, M., Chemistry and Chemical Technology in Ancient\
* Mesopotamia. Elsevier Publishing Company, New York,
N.Y., p. 65 (1959).A
69. Matijevic, E., "Coagulation of Hydrophobic Sols in
Statu Nascendi. I", Journal of Physical Chemistry.” Vol. 55, p. 1557 (1951). „
70. Matijevic, E. and Kerker, M., "The Charge of Some Hetero—l
lpoly Anions in Aqueous Solutions as Determined by Coagu—
lation Effects", Journal of Physical Chemistry. Vol.6
2 62, p. 1271 (1958). . ”
g 71. Matijevic, E., Mathai, K. G., Ottewill, R. H., Kerker, M.,
_ "Detection of Metal Ion Hydrolysis by Coagulation. III.
Aluminum", Journal of Physical Chemistry. Vol. 65,
p. 826 (1961).( (
_ 72. Matijevic, E. and Janauer, G. E., "Reversal of Charge of
Lyophobic Colloids by Hydrolyzed Metal lons. I. Aluminum(
Nitrate", Journal Colloid and Interface Science. Vol.
19. p- 333 (196ä)•
A15373.
Matijevic, E. and Stryker, L. J., "Coagulation andReversal of Charge of Lyophobic Colloids byHydrolyzed,Metal Ions. III. Al2(SO4);, Jpprnal of Colloid andIInterface Science. Vol. 22, p. 68 (1966).
74. McLaughlin, G. D. and Theis, E. R., The Chemistry ofLeather Manufacture. Reinhold Publishing Corporation,New York, N. Y., pp. 622-7 (1945).
75. Mysels, K. J., Introduction to Colloid Chemistry. Inter-
science Publishers, Inc., New York, N.Y., pg 87 (1959).76. Nalco Chemical Company, "Nalcolyte 670 Series, Slurry
Process Flocculants and Coagulant Aids", Bulletin A-3,Nalco Chemical Company, Chicago, Illinois.
77. Omelia, C. R. and Stumm, W., "Aggregation of Silica ‘I}
Dispersions by Iron (III)", Journal of Colloid and
· Interface Science. Vol. 23, pp. 437-47 (1967).78. Otts, L. E. and Chicca, W. E., "The Feasibility of Using
a Predictive Equation Based on a Multiple Regression JAnalysis for the Determination of an Acceptable Alum_
_ Dose", Fifth Annual Report, Water Resources ResearchCenter, University of Maryland. pp, 79-100 (1969).
79. Packham, R. F., "The Coagulation Process II. EffectJ of pH on the Precipitation of Al(0H)5", Journal of J
Applied Chemistry. Vol. 12, p. 564 (1964). _
· 154( 1
80. Parker, C. E., "Biological Treatment of Spent Vegetable
- 7 Tannins", Proceedings of the Twenty—Fifth Annual Purdue
7 University Industrial Waste Conference. (1970).
81. Parker, C. E. and Thaker, I. H., "A Study of Kinetic
Parameters Using Spent Vegetable Tannins", Proceedings(of
the Third Mid-Atlantic Industrial Wastes Conference.
7 University of Maryland, (1969). _ ~
82. Rebhun, M., Weinberg, A., and Narkis, M., "Treatment of
Wastewater from Cotton Dyeing and Finishing Works for 7Re—Use", Proceedings of the Twenty-Fifth Annual Purdue
University Industrial Waste Conference. (1970).
83. Rice, C. L. and Whitehead, R., "The Theory of Coagulationof Emulsions", Journal of Colloid Interface Science.voi. 23, N6. 2, pp. 174-181 (1967).
84. ·Rich, Linvil G., Unit Processes of Sanitary Engineering.7 John Wiley and Sons, New_Tork, N.Y. (1963). i
85. Riddick, T. M., "Zeta Potential: New Tool forH20Treatment,Part I.", Chemical Engineering. Vol. 68,~
W 4
7l
No. 13, pp. 121-6 (1961).
86. Riddick, T. M., "Zeta Potential: New Tool for H20l
Treatment, Part II.", Chemical Engineering. Vol. 68,l
No. 14, pp. 141-6 (1961).
87. Riddick, T. M., "Role of Zeta Potential in Coagulation
Involving Hydrous Oxides", T.A.P.P.I., Vol. 47, No. 1,pp. 1714-1794 (1964). 7
j 155 9
88. Riddick, T. M., "Zeta Potential and Polymers", dournal
A.w.w.A., vei. 38, No. 6, pp.·719-22 (1966).89. Riddick, T. M., Control of Colloid Stability Through
° — gera Potential, Vol. I. Livingston Publishing Company,
Wynnewood, Penn. (1968).
90. Ries, H. E. and Myers, B. L., "Floeculation Mechanism:
T Charge Neutralization and Bridging", Science. Vol. 160,1 Ne. 3835, p. 1450 (1968). °
9l.*Rudolfs, W., ed., Industrial Wastes, Their Disposal and)Treatment. Library of Engineering Classics, Valley
stream, New York, N.Y., pp. 141-67 (1961).
T 92. Sawyer, C. N. and McCarty, P. L., Chemistry for Sanitary
1Engineers. McGraw-Hill Publishing Co., New York, N.Y.
9 pp. 290-7 (1966). T . 9 9
993. Sentz, R. E., "Industrial Waste Disposal Investigation:
9 ·The Effect of Tannin Extract Plant Waste on the NewU
T River at Radford, Virginia", Unpublished Bachelor of1
Science Thesis, Department of Chemical Engineering,
v.P.1. (1942). ‘T
C
94. Smellie, R. H. and LaMer, V. K., "Flocculation, Sub-° sidence and Filtration of Phosphate Slimes. VI. A
‘ ·Quantitative Theory of Filtration of Flocculated1 lSuspensions",·eearnal of Colloid Science. Vol. 13,. . .1
4 p. 689 (1968). 1
. 1
· 156
· 95. Smoluchowski, N. V., Bull. Intern. Acad. Polon. Sci.,' Classe Sci. Math. Nat. Vol. 1903, p. 184 (1903);
Adamson, A. W., op. cit. p. 261. " _
96. Standard Methods for the Examination of Water and'(_Wastewater. American Public Health Association, 12th
' · Edition, New York, New York (1966).
97. Stern, 0., "Zur Theorie der Elektrolytischen Doppel- ' -
schicht", Z. Electrochem. Vol. 30, pp. 508-516. " '(
98. Stumm, W. and Morgan, J. J., "Chemical Aspects of
Coagulation", Journal A.W.W.A., Vol. 54, No. 8,
pp- 971—90 (1962)- .-“ 99. ibid. p-979-l
100. Teot, A. S. and Daniels, S. L., "Flocculation of Nega-(
tively Charged Colloids by Inorganic Cations and ·
Anionic Polyelectrolytes", Presented at 156th Meeting,
Division of Colloid and Surface Chemistry, American(
'Chemical Society, Atlantic City, N.J., September (1968). ·
101. Thabaraj, G. J., Bose, S. M., and Nayudamma, Y., "Com-
_ parative Studies on the Treatment of Tannery Effluents- -by Trickling Filter, Activated Sludge, and Oxidation
· _ Pond Systems", Journal of the American LeatherChemists,
Association. Vol. 58, No. 2, p. 130 (1963)..
102. Tomlinson, H. D., Koon, J. H., Krenkel, P. A., and Thack-
ston, E. L., "Removal of Color from Vegetable TanningSolution", Presented at the Annual Conference of the
Water Pollution Control_Federation, Dallas, Texas, IOctober, 1969. l n
- (I
157l
103. Tomlinson, H. D., Thackston, E. L., Krenkel, P. A.,Aand McCoy, V. W., "Complete Treatment of Tannery Waste",VTechnical Report Number 15, Department of Sanitary andWater Resources Engineering, Vanderbilt Univ., Nash-ville, Tennessee (1968).
_) l
104.. ibid. pp. 80-81, 83-86. ·105. Toyoda, H., Yarisawa, T., Futami, A. and Kikkawa, M.,
"Studies of the Treatment of Tannery Wastes III. Ef-‘
fects of Chemical Treatments", Journal of the AmericanLeather Chemists Association. Vol. 58, No. 5, p. 446
(1966)- -.106. United States Department of Interior (FWPCA), "The CostI
— of Clean Water:· Leather Tanning and Finishing", Vol.III, Industrial Waste Profile No. 7, Washington, D.C.
7 pp--6-4 (1967)· 7107.. 1b1d. p.34.108.van Olphen, H.,Clay Colloid Chemistry. Interscience
(
Publishers, New York, N.Y. (1963). _
109. ibid. p. 10. · · _( V
110. ibid. pp. 18-19.J
111. ibid. p. 31. 7 7U
112. ibid. pp. 37-38. -- - "
113. ibid. p. 45. . . _
114. ibid. p. 51. l
115. ibid. pp. .55-56.1
·
4 158 7 87
116. ibid. pp. 169-71. 1
117. ibid. p. 172. 44 118. 1b1a. p. 174.
119. Verwey, E. J. W. and Kruyt, H. R., Z. Physik. Chem.
Vol. A167, p. 312 (1934). (
120. °Verwey, E. J. W. and Overbeek, J. Th. G., Theory of
the Stability_pj Lyophobic Colloids. Elsevier Pub-7 lishing Company, Inc., Amsterdam, Netherlands (1948).
121. ibid. pp. 4-5. (5
122. ibid. p. 19. _I
123. ibid. p. 20. 4(
7 7124. ibid. p. 41. 7 r
125. iibid. p. 49. 6 '7 (
126. Vold, M. J. and Vold, R. D., Colloid Chemistry: The(VScience of Large Molecules, Small Particles and Sur-
8
4 jaggs. Reinhold Publishing Corporation, New York,
N. Y., p. 1 (1964).127. ibid. p. 27. ~ . ”
7__
128. von Helmholtz, H., Ann. Phygik. Vol. 7, p. 337 (1879).
129. von Smoluchowski, M., Bull. Acad. Sci. Cracovie. p. 182
(1906). 7 ’ ’4 130. von Smoluchowski, M., Handbook der Elecktrizitat und
der Magnetismus II. Edited by Graetz, Leipzig, Germany,
p. 366 (1914). ( (
1n
159 '
131. von Smoluchowski, M., Physik Z. Vol. 17, pp. 557, 585
(1916).1 1
_ 132. von Smoluchowski, M., Z. Phy§ik;_Qhgm. Vol. 92, p. 1291(191v)• 1
133. Walles, W. E., "Role of Flocculant Molecular Weight in
Coagulation of Suspensions”, Journal of Colloid Inter-
race Science. Vol. 27, No. 4, pp. 797-803 (1968).
. 134.- Wilson, J. A., The Chemistry of Leather Manufacture.l Chemical Catalog Company, New York, N.Y., pp. 214,
216 (1923). .135. ibid. p. 240. ,
I VIII. APPENDIX
1 . · 160
n Ü 1614 6
e 4 _APPENDIX A Ü
Included in this Appendix are the data from Experimentsl'6A through 6J and 7._ In the data headings response No. 1 isthe per cent reduction in total volatile solids; response
No. 2 is the per cent light transmission for tannin concen-
tration; response No. 3 is the per cent reduction of C.0.D.,response No. 4 is the per cent reduction in volatile suspend—
· ed solids; response No. 5 is the settled volume of the con-
centrated phase after 18 hours; response No. 6 is thefiltra—tion
time for the volatile suspended solids test; response
No. 7'is the per cent light transmission due to turbidity;
. and response No. 8 is the light scattering measured in JTU.
The independent Variable conditions for each test in theExperiment were presented earlier in Tables II, III, IV, ”
and V. » a 'g
162 ·
TABLE XIII. DATA FOR EXPERIMENT 6A! CATIONIC NALCOLYTE 605
Test Resp. Resp. Resp. Resp. Resp. Resp.N0. No. 1 No. 2 No. 5 N0. 5 No. 6 N0. 7
' 1 9.61 55.8 21.1 ~ 210 17.0 9.02 40.75 42.5 55.6 665 21.0 55.55 18.44 40.8 17.2 155 41.0 57.54 21.26 40.0 18.5 170 42.0 54.05 55.50 57.0 52.1 295 85.0 56.06 . 52.07 27.5 50.5 725 60.0 7.07 17.18 45.0 17.2 115 85.0 12.0
. 8 · 57.52 57.0 51.2 560 45.0 49.59 5.12 50.5 4.2 0 5.5 · 12.7
10 59.74 56.5 57.2 450 2.5 57.011 27.40 51.2 28.5 467 65.0 24.812 50.66 45.0 ‘ 24.5 510 ‘5.5 28.015 ‘
57.08 52.5 55.8 770 25.0 25.07 14 56.15 50.5 · '51.7 690 55.0 19.9
15 58.42 55.0 55.2 715 55.0 22.216 51.87 51.5 — 27.5 452 65.0 50.017 27.54 50.9 29.2 455 80.0 50.718 18.28 25.0 16.6 260. 26.0 56.719 46.05 52.0 42.9 ‘ 445 65.0 24.8
· 20 50.28 50.8 25.2 210 65.0 51.5
163
* TABLE XIV. .DATA FOR EXPERIMENT 6Bz CATIONIC NALCOLYTE 634
Test Resp. Resp. Resp. Resp. Resp. ‘Resp.7 No. No. 1 No. 2 N0. 4 . No. 5 N0. 6 N0. 7
1 24.50 37.5 29.8 300 13.0 40.02 38.78 52.7 58.2 440 11.5 42.53 25.62 34.1* 33.2 310 12.0 40.34 24.84 37.5 47.8 285 11.0 35.85 39.45 52.8 47.8 440 12.5 47.2
~6 42.33 37.5 _ 60.4 450 18.0 51.3» 7 24.06 38.3 ' 47.8 305 15.5— 40.38 _ 44.47 58.8 61.3 540 9.0 · 64.29 - 4,92 25,3 21.0 .
1 6.0 12.010— 43.40 53.3 39.8 » 410 14.0 39.2 „11 36.34 46.2 70.8 400* 11.5 41.212 36.92 47.0 70.9 430 18.0 44.213 37.31 46.5 60.4 405 11.0 48.014 36.09 43.5 62.5 375 12.0 37.715 36.60 45.0 . 70.2 385 14.0 40.716 34.00 48.3 42.1 415 14.0 48.3 *
_ 17 .35.6149.3 58.4 385 12.5 50.0
- 18 36.00 44.5 59.7 370 12.0 42.5°.
19 37.12 46.0 60.3 388 10.5 43.520 35.70 45.5 40.0 382 13.5 44.0
164 E
TABLE XV. DATA FOR EXPERIMENT 60, CATIONIC NALCOLYTE 603
Test Resp. Resp.No. No. 1 N0. 5
° . 1 24.1 300· 2 41.9 462
. 5 23.8 5004 24.6 3105 . 42.5 4606 42.0 · 4407 · 25.3 2808 44.2 _ 420
_ 9 —7.0 O .10 45.9 420 ·11 55-5 570
„ 12 34.1 35013 58.7” 44714 . 35.1 37015 55.9 59016 — _ 36.0 375
_ 17 34.1 59518 32.0 59019 33.0 · „ 580
_ 20 32.0 403
« 165_
TABLE XVI. DATA FOR EXPERIMENT 6Dz CATIONIC NALCOLYTE 632
Test Resp. 7Resp.No. No. 1 No. 5
‘
1 33.33 · - 3782 45.39 6153 34-54 4354 _ 37.16 415
- ‘ 7 5 49.41 7 580‘ 6 49-36 5157 „ 30.28 5558 51-63 . 5509 5.11 8 ·
_ 10 49.75 550· 11 44.18- 470 .” ‘ 12 43.16 470
. 13 40.82 49014 43.36 3 490 ‘
715 39.72 48016 42.48 7 490
7 17 — 41.56 485. 18 45.00 7450
7 19 41.61 450 E‘ 10 42.44 - 450
166
TABLE XVII.” DATA FOR EXPERIMENT 6Ez NONIONIC NALCOLYTE 670
Test Resp. Resp. V Resp.N0. No. 1 _N0. 2 No. 7
1 29.1 4840 . E 60.52 48.2 70.8 88.23 30.7 45.8 48.04 _30.6 40.5 61.5‘ 5 48.2 73.0 _ 88.5
- ·6 45.8 ‘. 69.0 76.0
__ 7 27.1_ 27.1 49.3. 8 44.9
” 68.0. 63.09 4.0 30.0 14.5
10 A 51.9 71.2 73.0 · ·· 11 36.2 49.5 56.0 .
12 40.8 62.3 66.513 45.6 Re 61.7 45.614 45.9 · 63.3 45.9°
7 15 44.2 65.0 44.2 116 45.9 63.5 45.917 41.0 62.8 41.0
. 18 45.5 62.4 45.5_
”19 44.6 62.5 44.6 ‘20 44.2 63.0 44.2 .
167
TABLE XVIII.' DATA FOR EXPERIMENT 6Fz CATIONIC NALCOLYTE 607
‘ Test Resp. Resp.No. No. 1 3 No. 7 ·
1 33.90 . . 42.3’ 2 48.34 48.0
’‘5 57-12 58-5
· A 4 . 55.22 44.5A 5 49.85 _79.5
6 52.59 85.5 _7 1 34.15 42.0
· 8 50.73 . 71.09 5.51 9.0 ·
.10 46.98 66.0’ _ 11 43.95 62.0 .
12 59.61 51.013 43.90
’ 59.214 39.02 64.0 ‘
15 ' 42.00 60.016 42.59 59.017 42.24 50.0
· 18 43.41 55.5. 19 V 42.97 63.5 ‘
20 41.12 57.0
168
TABLE XIX. ~DATA FOR EXPERIMENT 6G, ANIONIC NALCOLYTE 675
Test Resp. Resp.No. No. 1 ·N0. 8 .
41 54.7
E265
2 54.2 ' E 2253 37.4 265 3· 4 54.9 217
. 5 · ~ 51-8 575-
' 6 55.5 · 2187 45.2 2588 r 52.0 3 210 .9 0.5 ‘ ‘ 14 _
10 51.8 — 207_ · 11 l 4 48.5 4 462
12 48.6 51515 4 40.8 21214 46.0 215 .15 47.5 21816 44.2 218
4 _ 17 45.1 . 225‘ 18 48.7 218‘ ‘ 19 . 45.5 217 .· 20 45.2 ’ 222
I
169
TABLE XX. DATA FOR EXPERIMENT 6Hz ANIONIC NALCOLYTE 672· Test Resp. Resp.E No. No. 1 No. 4
1 35.4 51.082 _ 51.9 68.97. 3 34.4 55.014 31.4 44.17 '°5 51.2 58.54‘6 48.2 ·° 56.377 -38.4 49.738 51.0 66.53 ' -. 9 -5.9 20.4610 ”
48.1 64.50. 11 44.7. 47.9712 47.4 ' 64.50· 13 39.6 46.48_ _ 14 .47.0 61.6515 A 41.9 49.5916 · 42.4 51.62 .917 40.3 52.17„~ 18 43.3 . 51.63. r 19 43.7 49.50° 20 43.1 43.90
170 °
TABLE XXI. DATA FOR EXPERIMENT 6I. ANIONIC NALCOLYTE 674
Test Resp. Resp.L No. No. 1 No. 8
1 28.7 2302 45.1 ' 3403 28.2 _ 2084 27.0 198‘5 48.7 410 ~ ·6 44-5 35357 25-9 . 135
_ 8 48.7 361.9 4.4 3 75
10 51.5 E 320 E4 11 38.2 225 _
12 ·38.9 1 285 ‘13 . L 42.6 300 M 114 47.4 1 350
· 15 45-6 355 116 42.2 36517 44.1 35518 45-5 · ' 343
L 19 39-9 34020 40.5 355 _
171
TABLE XXII. DATA FOR EXPERIMENT 6Jz CATIONIC NALCOLYTE 610
° Test Resp. Resp.No. No. 1 No. 6
1 27.6 12.5— 2 _ 42.2 4.5~ 5 25.4 > 15.5
_ 4V -25.8 10.5 *_ 5 45.8 _ 5.0
* 6 42.0 10.07 27.5 ‘14.0* 8 40.6 11.0 ° -9 2.4 6.6
10 52.0 5.5 ‘11 - 55.5 12.012 35.5 8.515 ’ 45.8 12.5
_. 14 42.6 2 · 15.015 45.4 15.4
_ 16* 59.7 12.5 _17 44.5 14.1*’ 18 41.5 15.4
A“
19 44.8_ 14.820 59.8 ' 15.5
172
TABLE XXIII. DATA FOR EXPERIMENT 7. EXTENDED PARAMETER
EXPERIMENT USING ANIONIC NALC0 675Test Resp. Resp. Resp. Resp. Resp. Resp. Resp. Resp.No. N0. 1 N0. 2 N0. 5 No. 4 N0. 5- No. 6 N0. 7 No. 8
1 29.9 45.0 25.1 46.6 475 9.3 49.0 2551A 27.7 42.5_ 27.5 45.4 570 10.7 51.0 155· 2 44.8 56.8 45.8 57.4 660 4.5 53.8 4152A 45.5 66.2 45.2' 72.9 ' 605 4.5 81.5 5555 51.0 46.0 24.7 56.5 420 8.8 56.2 2605A 52.2 47.0 51.2 55.0 570 9.5 65.5 1904 28.5 41.1 59.0 47.5 400 9.0 44.9 200 -4A 29.8 48.0 28.5 51.3 395 9.5 60.5 1905‘ 35.4 52.0 55.7 59.2 458 9.8 45.5 182 » 55A 55.7 55.0 55.8 92.8 460 11.5· 48.0 115
_ -6 47.7 64.1 45.2 65.2 655 2.7 61.9 5056A 46.8 65.1 45.5 65.2 690 4.5 67.5 ’ 5507 48.5 66.0 49.1 69.9 565 8.5 69.5 540°7A 46.6. 70.5 48.2 64.8 620 12.4 76.0 5408 47.1 66.5 48.1 57.9 640 10.5 60.5 470
. 8A 49.5 70.0 46.4 58.2 625 14.9 70.7 542.
I941.8 57.3 37.0 58.7 452 6.5 60.5 268
.9A 45.0 64.0 41.5 88.7 490 0.5 89.4 28_ 10 49.5 71.0 48.5 62.4 578 11.0 75.1 480
10A 49.2 71.5 47.9 71.1 620 ·6.5 81.5 540*11 46.5 64.3 44.2 64.3 620 4.5 69.0 475
· 11A 49.7 71.0 48.2 82.9 650 2.5 90.5 210I
12 44.1 64.2 59.8 — 61.8 675 4.5 65.9 450~»12A 45.8 62.0· 57.5 55.1 ~ 720. 6.5 59.5 500 '15 55.2 55.9 52.1 58.5 440 ‘10.8 44.5 154
I
15A 55.6 55.0 55.5 44.1 455 11.3 43.3 8214 58.5 50.0 57.2 69.0 578 4.8 71.0 25014A 52.9* 50.0 24.2 62.6 400 6.2 66.0 17015 47.5 68.2 45.2 57.8 _670 15.5 66.0 440
I
15A 49.4 71.5 58.1 67.6 600 10.7 79.7 520I
16 58.1 55.9 58.4 57.6 422 - 7.8 56.8 22016A 57.2 55.5 39.2 60.5 402 7.9 60.2 21017 3.3 33.5_ 5.2 15.8 10 6.5 11.3 5018 48.8 71.7 48.2 69.7 ‘ 750 4.2 77.0 55019 41.8 59.2 40.1 44.1 · 495 11.5 50.0
273I
20 I42.1 58.9 ‘38.2 61.6 550 6.8 58.5 40421 41.59 67.0 58.2 69.0 740 6.5 82.3 590
I
22 45.21 67.2 I59.9 64.6 605 5.5 71.3 470 ‘ E25 58.15 61.0 39.9 65.2 520- 4.1 71.0 37524 41.69 *71.3 39.0 65.9 600 9.3 78.5 440 ‘25 56.9 65.0 58.2 52.9 540 5.4 56.4 570 _26 57.8 67.0 52.2 60.2 717 9.0 70.5 41027 56.4 78.5 58.4 61.2 665 7.4 68.4 41028 59.1 66.0 58.2 61.5 680 8.5 75.0 410 -29 40.8 64.5 38.2 61.5 655 8.5 66.5 32550 40.2 66.5 58.1 65.9 · 585 8.2 70.0 542
i
EVALUATION OF PARAMETERS AFFECTING THE COLLOIDAL —DESTABILIZATION OF SPENT VEGETABLE TANNIN LIQUOR
r by, Ql _ William A. Barkley V
A
‘ Abstract
The overall objective of this investigation was todevelop a chemical method for treating spent vegetable
tannin liquor. Vegetable tannin wastes are plagned hy a
multitude of undesirable characteristics, principal of
which is a color problem. Histerically, chemieal coagu—
lation processes have been successful in reducing color
associated with industrial wastes. ·‘ .. The coagulants tested were aluminum sulfate and poly-
electrolytes. In addition, the pH and the time interval
between aluminum ion and polyelegtrolyte addition werel
studied. Considerable diffiédltf was encountered in ob-
taining a waste product of unißorm characteristics over an
extended period. This problem was solved by storing a
large quantity of waste of typical properties at tempera— gtures near zero degrees centigrade. '
» The most important variable affecting the treatment
of the tannin.waste was the aluminum sulfate concentration.
Maximum destabilization efficiency was achieved at 3,000 O
mg/l and a waste volatile solids to aluminum sulfate ratio
of seven. Anionic polyelectrolytes, used in conjunction ‘
with the aluminum sulfate, gave superior performance to
aluminum ion-cationic polyelectrolyte systems. The cationic
. polyelectrolytes were most efficient when added prior to the
_ aluminum sulfate, the anionic polyelectrolytes most efficient
when added afterwards. The suggested mechanism for the
anionic polyelectrolyte was chemical bridging between alumi-
num ion flocs followed by_rapid settling. For cationic
polyelectrolytes, charge neutralization of sol particlesV
in competition with the aluminum ion was evident. The
optimum pH was seven with a range of 5toV7
being tested.
For the best polyelectrolyte tested, anionic Nalcolyte 675,V
the use of optimum conditions produced reductions of vola-
tile solids of 65 per cent., At these reduction levelsV
‘upwards of 90 per cent color removal, as measured by tanninV concentration, was achieved. For Nalcolyte 675 the maximumV
V V treatment was observed using 50 mg/l of polyelectrolyte,
but satisfactory performance of 56 per cent volatile solids
reductions was attained through the use of 2 mg/l poly-
electrolyte.V V
_
_ The investigation was successful in defining economi- _
cally feasible conditions for chemical treatment which was
effective in color removal. Subsequent biological treatment ‘
would be required for adequate organic matter reduction. V