LIBRARY OF THE 
 
 UNIVERSITY OF ILLINOIS 
 
 AT URBANA-CHAMPAIGN 
 
 £28 
 Li65c 
 HO. 57-5S 
 
 ruGlMEEWG L ffiRRR J : 
 
 ENGINEERING 
 
OttttfttW^ 
 
 The person charging thts materia^ re 
 sponsible for its return on or before 
 Latest Date stamped below. 
 
 Theft , mutil o r and -J-J. o^s 
 UNI VH^ O, ,U,NO« — .T^ANA-AMP^ 
 
 ttffiKHtt ROOM 
 
 JUL 17 197 
 
 JUN 1 8 1975 
 
 L161— O-1096 
 
Digitized by the Internet Archive 
 in 2013 
 
 http://archive.org/details/effectofsludgech57wood 
 
- 
 
 CIVIL ENGINEERING STUDIES 
 
 SANITARY ENGINEERING SERIES NO. 57 
 
 THE EFFECT OF SLUDGE 
 CHARACTERISTICS UPON THE FLOTATION 
 
 OF BULKED ACTIVATED SLUDGE 
 
 <i 
 
 fir tz 
 
 2 < 
 
 re £ z 
 tu tr < 
 
 u? as 
 
 o > 
 
 uj 5 
 
 »^SJ* 
 
 At 
 
 <s 
 
 * 
 
 w 
 
 By 
 ROBERT F.WOOD 
 
 ENGINEERING LIBRARY 
 
 UNIVERSITY OF ILLINOIS 
 
 URBANA, ILLINOIS 61801 
 
 ,.^< 
 
 X& 
 
 
 
 Supported By 
 FEDERAL WATER QUALITY 
 
 ADMINISTRATION 
 TRAINING GRANT WP 128 
 
 < 
 tf 
 
 N«& 
 
 DEPARTMENT OF CIVIL ENGINEERING 
 
 UNIVERSITY OF ILLINOIS 
 
 URBANA, ILLINOIS 
 
 SEPTEMBER, 1970 
 
THE EFFECT OF SLUDGE CHARACTERISTICS UPON THE 
 FLOTATION OF BULKED ACTIVATED SLUDGE 
 
 BY 
 
 ROBERT FRANK WOOD 
 B.S. Engr., Walla Walla College, I960 
 M.S., University of Texas at Austin, 1966 
 
 THESIS 
 
 Submitted in partial fulfillment of the requirements 
 
 for the degree of Doctor of Philosophy in Sanitary Engineering 
 
 in the Graduate College of the 
 
 University of Illinois at Urbana-Champaign , 1970 
 
 Urbana, I 1 1 i noi s 
 
Ci £ SV- i||BIIIEEMN6 LtBRAW 
 
 THE EFFECT OF SLUDGE CHARACTERISTICS UPON THE 
 FLOTATION OF BULKED ACTIVATED SLUDGE 
 
 Robert F. Wood, Ph. D. 
 Department of Civil Engineering 
 University of Illinois 
 
 The bulking of activated sludges with its accompanying difficult 
 solids separation is a problem which is commonly faced in waste treatment 
 plants. The purposes of this study were: 
 
 1. to investigate some of the properties of bulked activated sludges 
 and to relate the properties to the flotation behavior of the sludges and 
 
 2. to investigate the possibility of using the batch flux analysis as 
 a method of predicting the results of the operation of continuous flow flo- 
 tation units. 
 
 The effects of the carbon to nitrogen ratio, the nitrogen content, 
 and the morphological characteristics of the sludge as well as the rheological 
 properties, yield strength and plastic viscosity upon the flotation behavior 
 of the sludges were investigated. The float solids concentrations and the 
 interfacial rise rates obtained from a standard batch flotation test proce- 
 dure were used to compare the flotation behaviors of the sludges. Of the 
 sludge characteristics studied, the rheological properties correlated the best 
 with the flotation behavior of the sludges. The rheological properties 
 reflected the morphological characteristics of the microorganisms which made 
 up the sludge floe. 
 
 The diameter of the flotation column, the initial depth of sludge 
 in the column and the column filling time were found to influence the inter- 
 facial rise rates of the batch flotation tests. A retardation factor which 
 
 1 
 
was dependent upon sludge concentration was found to exist in batch flotation 
 tests. The pressure at which the saturation of the recycle water was carried 
 out was found to greatly affect the batch flotation test results even though 
 the air to solids ratio was constant. 
 
 The batch flux method of analysis was used to predict the float 
 solids concentration obtained from a laboratory scale continuous flow flota- 
 tion unit. This method accurately predicted the float concentrations 
 obtained from the continuous unit in some cases but was not always consistent 
 The Talmadge and Fitch method was an unsatisfactory method for extending the 
 flux curve to higher concentrations for the activated sludges used here. 
 
ACKNOWLEDGMENTS 
 
 The author wishes to express his sincere appreciation to Professors 
 R. S. Engelbrecht, J. T. Pfeffer, R. I. Dick, B. B. Ewing and V. T. Chow for 
 serving as members of his doctoral committee. To his thesis advisor, Doctor 
 R. S. Engelbrecht, and to Doctors J. T. Pfeffer and R. I. Dick the author is 
 deeply indebted for their suggestions, guidance and encouragement throughout 
 the course of the research and in the preparation of this thesis. 
 
 The author also wishes to thank the following for helping to make 
 this thesis possible: 
 
 Messrs. G. L. Shell and E. C. Cardwell and the Eimco Corporation 
 for the use of the continuous flow laboratory scale flotation unit. 
 
 The Central Soya Co. for supplying the molasses which was used as 
 the substrate in this study. 
 
 The Dixon Springs Agricultural Experiment Station for loaning the 
 tank used as the aeration tank in the laboratory scale activated sludge unit. 
 
 Miss Chrysilis Rymas and Mr. Mike Smith for their assistance in the 
 laboratory. 
 
 All personnel in the Sanitary Engineering Laboratory who have helped 
 in one way or another toward the completion of this study. Especially thanks 
 are given to Joe Matherly, laboratory manager, for his kind assistance. 
 
 Mmes. Ruth Worner, Mary Grahl, and Betty Wood for their help in the 
 typing of the thesis. 
 
 The author wishes to acknowledge with thanks the financial assistance 
 granted by the Federal Water Quality Administration, Department of the 
 Interior, which made this work possible through Training Grant WP 128. Also 
 
he wishes to thank Loma Linda University for its financial support which was 
 granted in the form of a loan. 
 
 And last but by no means least the author wishes to thank his wife, 
 Betty, for her love, encouragement, and understanding throughout this work as 
 she performed the monumental task of caring for a husband and four children. 
 
TABLE OF CONTENTS 
 
 Page 
 
 ACKNOWLEDGMENTS i i i 
 
 LIST OF TABLES ix 
 
 LIST OF FIGURES x 
 
 I. INTRODUCTION 1 
 
 A. General 1 
 
 B. Purpose and Scope 3 
 
 II. LITERATURE REVIEW 5 
 
 A. Sludge Bulking 5 
 
 1. Effect of Food to Microorganism Ratio on Sludge 
 
 0_ual ity 9 
 
 2. Effect of Substrate Carbon to Nitrogen Ratios on 
 
 Sludge Quality 9 
 
 B. Sludge Quality and Rheological Properties 12 
 
 C. Dissolved Air Flotation 12 
 
 III. EXPERIMENTAL EQUIPMENT AND PROCEDURES 17 
 
 A. Laboratory Activated Sludge System 17 
 
 1. Description of Equipment 17 
 
 a. Constant Head Box 17 
 
 b. Primary Settling Tank 19 
 
 c. Aeration Tank 19 
 
 d. Continuous Flotation Unit 20 
 
 2. Substrate 22 
 
 3- Operation 2k 
 
 a. General 24 
 
 v 
 
Page 
 
 b. Precedure for Changing the Characteristics of the 
 
 Activated Sludge 25 
 
 B. Batch Flotation Tests 26 
 
 1. General Procedure 26 
 
 2. Standard Batch Flotation Test Procedure 30 
 
 3. Comparison of Methods for Performing Batch Flotation 
 
 Tests 31 
 
 k. Reproducibility of Batch Flotation Data 33 
 
 5. Effect of Column Filling Time on Interface Rise Rate . ?>k 
 
 6. Effect of Column Diameter on Rise Rate 36 
 
 7. Effect of Column Depth on Initial Rise Rate hi 
 
 8. Effect of Pressure of Saturation on Batch Test Rise 
 
 Rates kk 
 
 9. Summary h3 
 
 C. Solids Flux Analysis . . . 50 
 
 1. Procedure for Obtaining Data from Continuous Laboratory 
 Flotation Unit 50 
 
 2. Procedure for Determining Quantity of Air Released . . 52 
 
 3. Procedure for Obtaining Data for Batch Flux Curves . . 55 
 
 D. Study of Rheological Properties 58 
 
 1. Description of Viscometer 58 
 
 2. Calibration of Viscometer 61 
 
 3. Experimental Procedure 67 
 
 E. Removal of Extracellular Polysaccharide 68 
 
 F. Analytical Procedures 71 
 
 1. Chemical Oxygen Demand 71 
 
 2. Organic and Ammonia Nitrogen 72 
 
vii 
 
 Page 
 
 3. Suspended Solids Determination 72 
 
 4. Sludge Volume Index 73 
 
 5. Total Carbohydrate 74 
 
 6. Float Solids Concentration 75 
 
 7. Carbon to Nitrogen Ratio of Sludge Solids 75 
 
 8. Photomicrographs 77 
 
 IV. RESULTS AND DISCUSSION 78 
 
 A. Relationships Between Sludge Characteristics and 
 Floatability 78 
 
 1. Shifts in Microbial Population 78 
 
 2. Effects of Percent Nitrogen and Carbon to Nitrogen 
 
 Ratio of the Sludge Upon Flotation 84 
 
 3. Relationship Between the Presence of Extracellular 
 Polysaccharide and Flotation Behavior 88 
 
 4. Rheology of the Activated Sludges 91 
 
 5. Relationship of Rheological Properties to the 
 
 Morphology and Nitrogen Content of the Sludge 100 
 
 6. Relationship Between Rheological Properties and 
 
 Flotation Behavior 104 
 
 7- Continuous Flotation Unit Performance 113 
 
 8. Summary 114 
 
 B. Batch Flux Analysis of Continuous Thickener Operation ... 117 
 
 1. General 117 
 
 2. Comparison of Batch Flux Results with Laboratory 
 Flotation Unit Data 118 
 
 3. Investigation of Possible Methods for Extending the 
 
 Batch Flux Plot 125 
 
 a. Direct Pressur izat ion 125 
 
 b. Talmadge and Fitch Method 126 
 
viii 
 Page 
 
 4. Effect of Anaerobic Conditions Upon Flotation 127 
 
 5. Summary 129 
 
 C. Significance of Results 130 
 
 V. CONCLUSIONS 132 
 
 VI. SUGGESTIONS FOR FUTURE WORK 135 
 
 REFERENCES 136 
 
 APPENDIX 142 
 
 VITA 143 
 
LIST OF TABLES 
 
 TABLE Page 
 
 1 CHARACTERISTICS OF THE MOLASSES USED AS SUBSTRATE FOR THE 
 CONTINUOUS FLOW LABORATORY ACTIVATED SLUDGE SYSTEM .... 23 
 
 2 REPRODUCIBILITY OF BATCH FLOTATION DATA 33 
 
 3 TYPICAL BLENDING SCHEDULE FOR A/S = 0.0046 59 
 
 k C-H-N DATA FOR SEVERAL ACTIVATED SLUDGES 88 
 
 5 SUMMARY BATCH FLOTATION DATA 122 
 
LIST OF FIGURES 
 
 FIGURE Page 
 
 1. SCHEMATIC DIAGRAM OF THE CONTINUOUS FLOW LABORATORY ACTIVATED 
 SLUDGE SYSTEM 18 
 
 2. SCHEMATIC DIAGRAM OF FLOTATION TANK OF LABORATORY CONTINUOUS 
 
 FLOW UNIT 21 
 
 3- SCHEMATIC DIAGRAM OF THE BATCH FLOTATION TEST APPARATUS ... 27 
 
 4. TYPICAL BATCH FLOTATION CURVES 29 
 
 5. EFFECT OF COLUMN FILLING TIME ON RISE RATE 37 
 
 6. EFFECT OF CYLINDER DIAMETER ON INTERFACE VELOCITY 39 
 
 7. EFFECT OF COLUMN DIAMETER ON RISE RATE 41 
 
 8. EFFECT OF COLUMN DEPTH ON RISE RATE 45 
 
 9. DETERMINATION OF RETARDATION FACTOR 46 
 
 10. EFFECT OF PRESSURE OF SATURATION ON RISE RATE 48 
 
 11. SCHEMATIC FLOW DIAGRAM OF THE CONTINUOUS FLOW LABORATORY 
 ACTIVATED SLUDGE SYSTEM 51 
 
 12. APPARATUS USED TO MEASURE VOLUME OF GAS RELEASED FROM WATER 
 
 UPON DEPRESSURIZATION 54 
 
 13- SCHEMATIC DRAWING OF COAXIAL CYLINDER ROTATIONAL VISCOMETER . 62 
 
 14. RELATIONSHIP BETWEEN VISCOSITY AND END EFFECT 63 
 
 15- VISCOMETER CALIBRATION CURVES 65 
 
 16. DETERMINATION OF END EFFECT 65 
 
 17. TYPICAL VISCOMETER DATA 69 
 
 18. STRING-OF-BEADS TYPE OF FILAMENTS 80 
 
 19. CARYOPHANON-LIKE FILAMENTS 80 
 
 20. CARYOPHANON-LIKE FILAMENT WITH CELLS ATTACHED PERPENDICULARLY 
 
 TO THE AXIS OF THE FILAMENT 81 
 
 x 
 
xi 
 
 FIGURE Page 
 
 21. ELONGATED FLOC 81 
 
 22. STIFF FILAMENT 82 
 
 23- RELATIONSHIP BETWEEN NITROGEN CONTENT OF THE SLUDGE AND FLOAT 
 
 CONCENTRATION 85 
 
 2k. RELATIONSHIP BETWEEN NITROGEN CONTENT OF SLUDGE AND RISE 
 
 RATE 87 
 
 25- RELATIONSHIP BETWEEN CARBON TO NITROGEN RATIO AND FLOAT 
 
 CONCENTRATION 89 
 
 26. RELATIONSHIP BETWEEN CARBON TO NITROGEN RATIO OF THE SLUDGE 
 
 AND RISE RATE 90 
 
 27. RELATIONSHIP BETWEEN THE NITROGEN CONTENT OF THE SLUDGE AND 
 EXTRACELLULAR POLYSACCHARIDE 92 
 
 28. RELATIONSHIP BETWEEN EXTRACELLULAR POLYSACCHARIDE AND FLOAT 
 CONCENTRATION 93 
 
 29. RELATIONSHIP BETWEEN EXTRACELLULAR POLYSACCHARIDE AND RISE 
 
 RATE Sk 
 
 30. TYPICAL YIELD STRENGTH VALUES FOR ACTIVATED SLUDGE 95 
 
 31. FLOC CONTAINING LARGE FILAMENTS RESEMBLING CARYOPHANON .... 97 
 
 32. FLOC CONTAINING FLEXIBLE FILAMENTS 97 
 
 33. NORMAL ZOOGLEAL FLOC 98 
 
 3*». TYPICAL PLASTIC VISCOSITY PLOTS FOR ACTIVATED SLUDGES .... 99 
 
 35. RELATIONSHIP BETWEEN THE NITROGEN CONTENT OF THE SLUDGE AND 
 
 YIELD STRENGTH 102 
 
 36. RELATIONSHIP BETWEEN THE CARBON/NITROGEN RATIO OF THE SLUDGE 
 
 AND YIELD STRENGTH 103 
 
 37- RELATIONSHIP BETWEEN THE NITROGEN CONTENT OF THE SLUDGE AND 
 
 PLASTIC VISCOSITY 105 
 
 38. RELATIONSHIP BETWEEN PLASTIC VISCOSITY AND THE CARBON TO 
 NITROGEN RATIO OF THE SLUDGE 106 
 
 39. RELATIONSHIP BETWEEN SLUDGE YIELD STRENGTH AND FLOAT 
 CONCENTRATION FOR BATCH FLOTATION TESTS 107 
 
xii 
 
 FIGURE Page 
 
 40. RELATIONSHIP BETWEEN PLASTIC VISCOSITY AND FLOAT 
 CONCENTRATION 109 
 
 41. RELATIONSHIP BETWEEN YIELD STRENGTH AND RISE RATE 111 
 
 42. RELATIONSHIP BETWEEN RISE RATE AND PLASTIC VISCOSITY 112 
 
 43. RELATIONSHIP BETWEEN SLUDGE YIELD STRENGTH AND FLOAT 
 CONCENTRATION FOR CONTINUOUS FLOTATION UNIT 115 
 
 44. METHOD OF DETERMINING LIMITING CONDITIONS 119 
 
 45. TYPICAL BATCH FLUX PLOT FOR THE LABORATORY CONTINUOUS 
 
 FLOTATION UNIT 121 
 
 46. COMPARISON OF PREDICTED FLOAT CONCENTRATIONS WITH THE ACTUAL 
 FLOAT CONCENTRATIONS OBTAINED FROM THE CONTINUOUS FLOW 
 FLOTATION UNIT 123 
 
 47. TALMADGE AND FITCH EXTENSION OF BATCH FLUX CURVE 128 
 
I. INTRODUCTION 
 
 A. General 
 
 The success of the activated sludge method of waste treatment is 
 dependent upon the ability of the mixed liquor suspended solids to remove 
 soluble organic matter and in turn to be separated from the liquid which is 
 to be discharged from the process as treated effluent. Normally this separa- 
 tion is accomplished by gravity in the secondary settling tanks which follow 
 the aeration process. Bulked sludge is only one of a number of types of 
 sludge which separate poorly. Sludge bulking results from the production of 
 a sludge which is not amenable to gravity settling as commonly practiced. 
 The settling rate and degree of compaction are so low that unless the treat- 
 ment plant has been designed to operate under bulking conditions, solids will 
 be lost in the final effluent and, as a result, the BOD and suspended solids 
 removal efficiencies will be greatly reduced. There is, however, always a 
 definite line of demarcation between the settled solids and the supernatant 
 even with a bulking sludge (Pipes, 1967a). The principal problem is that the 
 rate of separation is too slow. 
 
 Sludge bulking may be caused by many factors, only some of which 
 can be controlled at the waste treatment plant. Factors which have been 
 reported as having caused sludge bulking include the following: the food 
 to microorganism ratio (F/M) , the dissolved oxygen concentration in the aera- 
 tion tank of the waste being treated, and even the temperature of the air 
 being supplied to the aeration basin (Ingols and Heukelekian, 1940). 
 
 Flotation processes have been used for several decades in the mining 
 industry for the purpose of concentrating ores. Dispersed air flotation, a 
 
 1 
 
process which may use a variety of chemicals which act as aids in flotation, 
 commonly is used. While the dispersed air flotation processes used in ore 
 concentration are not generally economically applicable to the solution of 
 sanitary engineering problems, the basic principles of flotation which govern 
 that process are applicable to dissolved air flotation. Gaudin (1957), as 
 well as other authors, has given an extensive treatment of the flotation 
 theory. 
 
 In flotation processes air bubbles become attached to solid par- 
 ticles, oil globules or, when dealing with biological sludges, the sludge 
 floe. The air-solid agglomorate, having a specific gravity less than that 
 of the surrounding liquid, rises to form a foam-like layer at the surface. 
 The floated material is then removed for further processing. Dissolved air 
 flotation, a type of flotation not commonly used in ore processing operations 
 has been used in sanitary engineering practice primarily as a means of solids 
 separation with certain industrial wastes and for thickening biological 
 sludges. Only in a limited number of cases has it been used for the clari- 
 fication of activated sludge mixed liquors. With this process, water is 
 saturated with air at pressures greater than atmospheric and subsequently is 
 passed through a pressure reducing valve thus reducing the pressure to 
 atmospheric. Air in excess of that required for saturation at atmospheric 
 pressure comes out of solution as very small bubbles (50y-100y diam)(Katz 
 and Geinopolos, 1963; Vrablik, 1959). The air-laden water is immediately 
 blended with the solid containing waste water, bubble attachment takes place 
 and the air-solid agglomorates float to the surface and are removed for 
 further treatment. 
 
B. Purpose and Scope 
 
 Sludge bulking is a problem which is frequently encountered in 
 activated sludge treatment plants. Bulked sludges are generally efficient 
 in the removal of the organic material from the waste being treated but, 
 because of their poor settling properties, they are difficult to separate 
 from the liquid flow by gravity sedimentation as commonly practiced. Loss of 
 the sludge particles in the final effluent lowers the efficiency of the 
 treatment process as a whole. If the loss of solids is sufficiently great, 
 the efficiency of substrate removal decreases due to the reduction in mixed 
 liquor suspended solids, (MLSS) . The reduction in MLSS causes the F/M to 
 increase thus increasing the bulking problem. If the solids separation 
 operation could be accomplished efficiently, sludge bulking would no longer 
 be a problem. 
 
 Conceivably dissolved air flotation offers a means of obtaining the 
 necessary solids separation for these difficult to settle sludges. This 
 process has been used in industrial waste treatment practice for separating 
 poorly settling materials as well as for separating substances which have 
 specific gravities which are less than that of water. With this process the 
 separation rate, is greater than for gravity sedimentation. 
 
 The general purpose of this research was to explore the possibility 
 of using dissolved air flotation as a means of separating bulked activated 
 sludges. The specific objectives were: 
 
 1. to investigate the relationships between some of the characteristics 
 of activated sludge and the flotation behavior of sludges, and, 
 
 2. to explore the possibility of using the solids flux method of 
 analysis to predict the operational performance of a continuous 
 flotation unit. 
 
The first objective was accomplished using sludges grown in a lab- 
 oratory continuous flow activated sludge system. Manipulation of the organic 
 loading applied to the aeration tank and the amount of nitrogen supplied with 
 the substrate were used to produce sludges having different characteristics. 
 The sludge characteristics investigated included the carbon to nitrogen ratio 
 (C/N) , the nitrogen content, the amount of extracellular polysaccharide which 
 could be stripped from the floe, the morphology and the rheological properties 
 yield strength and plastic viscosity. A standard batch flotation test proce- 
 dure was used to evaluate the effects of the sludge characteristics upon 
 flotation behavior. 
 
 The second objective was accomplished by comparing the values of 
 float solids concentration predicted by the batch flux analysis with the 
 actual float solids concentrations obtained from a continuous flotation unit. 
 These comparisons were made using a laboratory scale unit. Before the data 
 for the batch flux curves could be obtained, a series of experiments were 
 performed to determine the effects of column diameter, column height and 
 column filling time upon the rise rates of sludge in the batch flotation test. 
 
I I . LITERATURE REVIEW 
 
 A. Sludge Bulking 
 
 The sanitary engineering literature is replete with descriptions 
 of sludge bulking problems associated with the activated sludge process. 
 While many of the reports offer conflicting hypotheses as to the causes of 
 the condition, there is general agreement that there are at least two common 
 types of bulking which can be differentiated microscopically, i.e. zoogleal 
 or non-filamentous bulking and filamentous bulking. 
 
 Non-filamentous bulking is characterized by the production of a 
 sludge which contains abnormally high amounts of bound water and consequently 
 has a specific gravity very nearly equal to that of water. This type of 
 sludge bulking is the least common and the most difficult to control (Sawyer, 
 1966). Bulking due to bound water is normally recognized when microscopic 
 examination of a bulked sludge shows the absence of or the presence of very 
 few filamentous organisms. Microscopic examination of this type of bulked 
 sludge reveals that the particles are ragged and diffuse with relatively large 
 surface areas (Smith and Purdy, 1936). Sludge volume index here is of little 
 value in describing the degree of bulking (Pipes, 1967b). 
 
 Fi lamentously bul ked sludges, as the name indicates, are in the 
 bulked condition due to the presence of significant numbers of filamentous 
 microorganisms. Filamentous bacteria or filamentous fungi may cause this 
 condition to occur (Pipes, 1967a; Lackey and Wattie, 19^0). Members of the 
 following genera have been identified as causing bulking: Sphaerot i 1 us , 
 Bac? 1 lus , Beggiatoa , Nocardia , Thiotrix , Arthrobacter and Geotrichum (Pipes, 
 1967a, 1967b). Pipes (1968) has published an atlas of activated sludge types 
 
which includes photomicrographs of many of the filamentous organisms mentioned 
 above. Sludge volume indexes ranging between 100 and 2,000 ml/g have been 
 observed as a result of this type of bulking (Pipes, 1967b; Kraus , 1 9^*9) • A 
 f i lamentously bulked sludge may appear under the microscope to be a mixture 
 of filamentous growths and bacterial floe. The filaments are believed to 
 provide physical support between sludge floes and to increase the bouyancy of 
 the floes, thus preventing them from settling and compacting well. 
 
 Many factors have been designated as contributing to the bulking of 
 sludge and encouraging the growth of filamentous organisms. Sawyer (1966) has 
 listed the causes of sludge bulking as being related to the characteristics 
 of the waste, to fluctuations in the concentration and volume of the waste, 
 and to limitations placed upon the system by its design and operation. A 
 number of the items listed in these categories will be discussed. 
 
 Waste waters containing considerable amounts of carbohydrates have 
 often been involved in sludge bulking problems (Lackey and Wattie, 19^0; 
 Morgan and Beck, 1939; Smit, 1930). Simple soluble substrates such as sugars, 
 amino acids and carboxylic acids which are readily metabol izable by most 
 microorganisms favor the growth of filamentous organisms while complex insol- 
 uble compounds which have to be hydrolyzed before being metabolized have been 
 reported to favor the growth of organisms which have good settling properties 
 (Ingols and Heukelekian, 19^0; Pipes, 1967b). Often wastes which are high in 
 carbohydrate are deficient in nitrogen and phosphorus. Such deficiencies 
 promote bulking (Greenburg, et^ a]_. , 1955; Jones, 1965). While the nitrogen 
 content of bacteria is about 10 to 12 percent of the dry cell weight, fungi 
 contain only k to 6 percent nitrogen (Jones, 1965; Kaylor et^ aj_. , 1963). This 
 fact gives fungi the competitive advantage for growth on a nitrogen deficient 
 
type of waste. For carbohydrates, carbon to nitrogen ratios of from ^0 to 10 
 are given as promoting bulking (Ingols and Heukelekian, 19^0). Both low and 
 high F/M ratios appear to be frequent causes of sludge bulking even in nutri- 
 tionally balanced wastes (Logan and Budd, 1955)- Bulking at high F/M ratios 
 is a common problem (Genetelli and Heukelekian, 196^; Tischler and Eckenfelder, 
 1968). Ford and Eckenfelder ( 1 966) , reported that in laboratory activated 
 sludge units treating brewery waste, refinery waste and domestic sewage, F/M 
 ratios above about 0.7 produced sludge bulking. A low pH waste will tend to 
 cause filamentous bulking as fungi are able to grow at pH values which are low 
 enough to affect the growth of most bacteria (Jones, I966; Pipes and Jones, 
 1963). Pipes and Jones (1963) reported that Geotrichum candidum grows well 
 in the pH range 3 to 9 while the normal range for the activated sludge process 
 is pH 6 to 9- Sphaerot? lus natans , a filamentous bacterium, commonly asso- 
 ciated with bulking, is less tolerant of low pH values than are most of the 
 other bacteria of the activated sludge biomass. 
 
 Low levels of dissolved oxygen in the mixed liquor due either to the 
 method of operation or to inadequate aeration capacity have frequently been 
 cited as causing bulking (Heukelekian, 19^0- 
 
 The idea that low dissolved oxygen in the mixed liquor promotes the 
 growth of filamentous organisms has recently been challenged by Bhatla (1967) • 
 His work with activated sludge treatment of Kraft paper mill wastes indicated 
 that his plant operated well in a f i lamentously bulked condition with 2.0 to 
 3.0 mg/1 dissolved oxygen in the aeration tanks. The settlability of the sludge 
 improved as the mixed liquor dissolved oxygen concentration was decreased. 
 Stale sewage and septic return sludge have also been reported as causing bulk- 
 ing (Heukelekian, 19^1). Hot weather and the accompanying low flow conditions 
 
which permit sewage to become stale are factors here. On the other hand, low 
 temperatures, particularly below 10 C, have also been cited as causes of 
 bulking (Ludzak et aJL, 1961). 
 
 Sphaerotilus natans has for many years been the organism most com- 
 monly associated with filamentous bulking (Ingols and Heukelekian, 1939; 
 Greeley, 1945; Ruchhoft and Watkins, 1929). Studies conducted in recent years 
 by Jones (1964, 1965) and Pipes and Jones (1963) have revealed that 
 Geotrichum candidum , a member of the biomass of most activated sludge systems, 
 may also cause filamentous bulking. While both organisms are obligate 
 aerobes, they are capable of metabolizing substrates at much lower oxygen 
 tensions than bacteria. Sphaerot i 1 us is a higher bacterium while Geotri chum 
 belongs to the class Fung? lmperfecti . In certain stages of growth the two 
 filamentous organisms look considerably alike; and this, no doubt, is part of 
 the reason why bulking caused by Geotrichum may have been credited to 
 Sphaerotilus at various times. 
 
 Whenever sludge bulking due to filamentous organisms has occurred 
 in waste treatment plants, the usual practice has been to initiate action 
 designed to destroy the filamentous growths and relieve the bulking condition 
 (Heukelekian, 1941; Tapleshay, 1945; Haseltine, 1938). 
 
 The activated sludge process can, however, be successfully operated 
 in a bulking condition provided the plant has secondary clarifiers and sludge 
 recirculation pumps of adequate capacity (Pipes, 1967a; Bloodgood, 1947; 
 Kraus, 1963). When such a system is properly operated, an effluent containing 
 less BOD and suspended solids is produced than is usually obtained wtth normal 
 sludge (Haseltine, 1932; Heukelekian, 1941; Keefer, 1963). Bhatla (1967) 
 reporting on experience with a 10 mgd plant treating Kraft mill waste indicated 
 that, when the plant was operated in a bulking condition, BOD and suspended 
 
solids removals were 5 to 10 percent greater than when operated in a non- 
 bulking state. Tischler and Eckenfelder (1968) reported that in bench scale 
 tests using glucose and phenol as substrates, sludges containing appreciable 
 amounts of filamentous organisms had much higher rates of substrate removal 
 than did the normal sludges. With aniline as substrate, the removal rates 
 were similar for both kinds of sludges. Jones (1966) states that "in high 
 carbohydrate (mainly monosaccharide) or sulfite waste liquor (low pH) , 
 filamentous microorganisms ( Geotr ichum in particular) will thrive and effect- 
 ively remove COD. This system is rendered inefficient only if the final 
 solid-liquid separation fails to provide the desired effect." 
 
 1. Effect of Food to Microorganism Ratio on Sludge Quality 
 
 In addition to the role which F/M plays in inducing bulking as 
 previously mentioned, it also plays a part in determining the percentage of 
 storage products in the microbial cells. Walters (1966) has shown in batch 
 studies on sludges grown on glucose and yeast extract that F/M does influence 
 the percentage of storage products in the MLSS. He found that, up to a F/M 
 of 4.3, the newly synthesized storage products constituted an increasingly 
 larger percentage of the cell weight. Beyond an F/M of 4.3 a greater percent- 
 age of the substrate was used for protein synthesis. 
 
 2. Effect of Substrate Carbon to Nitrogen Ratios on Sludge Quality 
 The carbon to nitrogen ratio (C/N) of the substrate has a marked 
 
 effect upon the quality of the sludge which is produced. Based upon the work 
 of Helmers et ah (1951) a BOD/N ratio of 20 is commonly accepted as being 
 necessary to prevent nitrogen deficiencies. While bulking has been encoun- 
 tered at this ratio, it has also been shown that systems can successfully 
 operate at BOD/N ratios up to 30 or more. Apparently with carbohydrates, a 
 range of BOD/N ratios can be successfully employed. 
 
10 
 
 Symons and McKinney (1957) have shown that in nitrogen deficient 
 conditions the microorganisms in activated sludge produced excess extra- 
 cellular polysaccharide which is apparently not biodegradable by the organisms 
 which produce it. Although the proportion of active mass in the sludge 
 decreases as nitrogen deficiency increases, it is possible to operate a stable 
 system on a continuous basis under nitrogen deficient conditions. The avail- 
 able nitrogen is utilized for the production of amino acids as long as the 
 supply lasts (Bechir and Symons, 1 966) . The remainder of substrate processed 
 by the cell is converted to polysaccharide. The efficiency of COD removal 
 falls off, however, as the percentage of nitrogen in the cell decreases. 
 Bechir and Symons (1966) in a laboratory scale completely-mixed system showed 
 about 97 percent and 92 percent soluble COD removals for sludges containing 
 10 percent and 5. 5 percent nitrogen respectively. Neither sludge age nor 
 sludge loading influenced the efficiency of soluble COD removal. This system 
 did not involve a solids separation step, therefore there were no data pre- 
 sented regarding the solids separation properties of the sludge. Information 
 was not given concerning the types of organisms present. Helmers et a 1 . (1951 ) 
 studied the microbial requirements for nitrogen and phosphorous in the treat- 
 ment of some industrial wastes. They found that the rates of BOD removal were 
 not reduced until the nitrogen content of the activated sludge fell below 7 
 percent of the weight of the volatile solids in the sludge. A sludge nitrogen 
 content of less than 7 percent was indicative of a critical nitrogen deficiency 
 in the substrate. 
 
 Walters (1966) working with activated sludges grown on glucose and 
 yeast extract in batch units, found that for substrate COD/N ratios of 1 6 . 5 
 and 56.0 the sludges contained 9-8 and 36.9 percent carbohydrate respectively. 
 The increase in carbohydrate was due to the formation of excess polysaccharide 
 
11 
 
 In addition to the normal amount of carbohydrate storage products. The excess 
 polysaccharide is relatively nonbiodegradable and does not serve as a storage 
 product for use in endogenous respiration. Hoover and Porges (1949) have given 
 the nitrogen content of activated sludge as being 12 percent as derived from 
 the general formula C £ -H 7 0_N. This value is based on cells in the endogenous 
 phase. McWhorter and Heukelekian (1964) report the nitrogen contents of 
 sludges in the presence of exogenous substrate to be 8 to 9 percent. The 
 difference between these values can be attributed to the presence of storage 
 products in the cells in the growth phase. 
 
 As was stated earlier, bacteria are reported to contain 10 to 12 
 percent nitrogen while fungi contain 4 to 6 percent (Jones, 1965; Kaylor 
 e_t_ a_l_. , 1963)- If the microbial seed contains fungi it would seem that the 
 fungi should tend to predominate in a nitrogen deficient system. Genetel 1 i 
 and Heukelekian (1964b), however, reported that for sludges grown on dif- 
 ferent substrates but having a BOD/N ratio of 20 the nitrogen content ranged 
 from 8.7 to 10.4 percent. On a glucose substrate having insufficient buffer 
 capacity a fungi predominated which had a nitrogen content of 10.4 percent. 
 On the same substrate only with more buffer capacity, Sphraerot? lus containing 
 8.9 percent nitrogen became the predominating organism. Kaylor et_ aj_. (1963) 
 in their work with atmospheric nitrogen fixation employed a filamentous sludge 
 
 j containing 3-8 to 5-4 percent nitrogen to obtain 90 to 93 percent COD removals 
 on a number of simple substrates. 
 
 The carbon to nitrogen ratio of the substrate does influence the 
 
 1 composition of the sludge produced. The proportion of the sludge which is 
 
 j active in substrate removal and the presence of extra amounts of polysac- 
 charide, some of which is in the form of extracellular capsular material, are 
 
 I evidence of this. The types of organisms predominating in the sludge also 
 
12 
 
 may be influenced by the amount of nitrogen available. These factors may 
 well change the surface properties of the sludge and may have an effect upon 
 the flotation characteristics of the sludge. 
 
 B. Sludge Quality and Rheological Properties 
 
 The rheology of activated sludge has been studied by Dick (1965). 
 His data indicate that the physical characteristics of the sludge are reflec- 
 ted in the rheological properties, yield strength and plastic viscosity. 
 Dick and Ewing (1967) have reported that bulked sludges possess much higher 
 yield strengths than do "normal" activated sludges. They have indicated that 
 measurements made using a viscometer are responsive to the physical nature of 
 the activated sludge and that the rheological nature of filamentous sludges 
 are very different from those of nonf i lamentous sludges. Chakrabarti (1967) 
 has shown that as the F/M ratio increased, the yield strength and plastic 
 viscosity of the sludge also increased. For example, for laboratory sludges 
 
 grown on domestic sewage primary effluent he reports, for MLSS of 2,500 mg/1 , 
 
 -2 -2 
 
 yield strengths of about 3-3 x 10 and 7-6 x 10 dynes/sq cm respectively 
 
 for F/M ratios of 0.3 and 0.6 lb BOD/lb MLSS. Yield strength was also found 
 
 to be dependent upon the concentration of solids. 
 
 C. Dissolved Air Flotation 
 
 Dissolved air flotation has seen its greatest use to date as a 
 method of recovering solids and oils from a wide variety of industrial wastes, 
 It has been used since the early part of this century in the pulp and paper 
 industry for the recovery of pulp fibers from white water. Only in the past 
 10 to 15 years has it been used to any extent in the treatment of biological 
 sludges. While this process has been used quite extensively for the thicken- 
 ing of activated sludges (Katz, 1958; Hurwitz and Katz, 1958; Ettelt, 1964; 
 
13 
 
 Katz and Geinopolos, 1 967) , only in a very few instances has it been used as 
 the primary means of solids separation in the activated sludge process. In 
 one instance this process was used to clarify the mixed liquor of an activated 
 sludge process which was treating fine chemical wastes from a pharmaceutical 
 manufacturing plant (Home et_ aj_. , 1962). The flotation unit was fed a mixed 
 liquor having a solids concentration of from 1,000 to 2,500 mg/1 and was pro- 
 ducing a float of 10,000 to 15,000 mg/1 suspended solids. Effluent suspended 
 solids were 600 to 900 mg/1. Reportedly, the high effluent solids and 
 relatively moderate float concentration were due to the nature of the waste 
 and certain operational problems. Data regarding the sludge volume index or 
 other characteristics of the mixed liquor were not published. 
 
 Mulbarger and Huffman (1970) have recently reported the use of 
 dissolved air flotation for mixed liquor solids separation at an activated 
 sludge waste treatment ' plant in Prince William County, Virginia. Domestic 
 sewage was treated at this plant. The activated sludge produced had a SVI of 
 58 ± 8 ml/g. Float solids concentrations between 3 and 5-5 percent were 
 obtained while the effluent suspended solids ranged up to about 200 mg/1, 
 depending upon the operating conditions. There appears to be no published 
 literature on the use of dissolved air flotation for the separation of bulked 
 activated sludges. Most of the data which are available regarding the flota- 
 tion of biological solids have to do with activated sludge thickening. 
 
 Ettelt (1964) has presented data from his thickening work done at 
 Chicago's Southwest Treatment Plant which indicate that as the SVI increases 
 the initial rise rate of the sludge decreases. His data show that as the SVI 
 increased from 70 to 120 ml/g the initial rise rate decreased from 0.78 cm/sec 
 to 0.48 cm/sec. The percentage of sludge which was floated also decreased as 
 SVI increased. It is important to note that the SVI values here were in the 
 
range for "normal" sludge, not nearly as high as would be expected with badly 
 bulked sludges. 
 
 The concentration of solids recovered in the float and the concen- 
 tration of solids in the effluent are functions of the suspended solids of the 
 feed, the retention time of the solids in the flotation unit and the air to 
 solids ratio (Vrablik, 1959)- A solids gradient exists in the vertical direc- 
 tion through the sludge layer with the sludge having the greatest concentration 
 of solids being nearest the top (Katz, 1963). The concentration of solids 
 near the top, increases as the solids retention time in the flotation unit 
 increases . 
 
 The air to solids ratio (A/S) , the weight ratio of air bubbles to 
 suspended solids in the flotation cell, is controlled by the pressure at which 
 aeration is carried out, the percent liquid recycle, the degree of saturation 
 and the concentration of suspended solids in the feed. The rate of rise of 
 the air-solids agglomerate is directly proportional to this ratio. Data 
 showing the effect of recycle rate on rate of rise for recycle rates up to 
 300 percent have been presented by Katz (1958, 1959)- The higher recycle 
 rates have higher initial rise rates because of the increased amount of air 
 available for flotation and the loss of resistance to flotation due to dilu- 
 tion of the sludge by the recycled liquid. Polyelectrolytes are sometimes 
 used in sludge thickening to increase the float solids concentration and to 
 reduce the solids loss to the effluent (Garwood, 1967; Katz and Geinopolos, 
 1967; Jones 1968). The use of polyelectrolytes has permitted flotation units 
 to be operated at loadings more than double those which were possible without 
 them while still obtaining a comparable float solids concentration. 
 
 The limited amount of information found in the literature regarding 
 the design procedures employed in sizing dissolved air flotation units 
 
15 
 
 indicated that the procedures used are largely empirical. Apparently the 
 practice has been to correlate data from batch flotation and pilot plant tests 
 with the performance data obtained from the previous operation of full scale 
 units on sludges of somewhat similar characteristics to determine the size of 
 the flotation unit for a particular application. Katz (1963) has stated that 
 for flotation units which are to be used primarily for sludge thickening, the 
 solids loading rate is the parameter which is used to determine the flotation 
 unit surface area. The surface area is determined by assuming a solids load- 
 ing rate. The area obtained is checked to be sure that a maximum desirable 
 hydraulic loading rate has not been exceeded. Loading rates of from 10 to 20 
 lbs/sq ft/day have been reported when polyelectrolytes were not used (Katz and 
 Geinopolos, 1967). Jones (1968) recommended the use of loading rates of 48 
 lbs/sq ft/day when polymers are used. Typical hydraulic loading rates were 
 1,150 to 1,440 gal/sq ft/day for activated sludges (Jones 1968). The proce- 
 dure used to determine the other dimensions of the flotation unit was not 
 described in the literature. These dimensions are probably determined by 
 empirically applying the rise rate data obtained from batch flotation tests. 
 
 Katz (1963) also pointed out that when the flotation unit is to 
 serve primarily as a clarifier, volumeteric surface loading which is related 
 to the separation rate of the air-solid particle, is the parameter upon which 
 the design is based. No further information as to the procedure for design 
 was found in the literature. 
 
 Mulbarger and Huffman (1970) based upon their work with a full-scale 
 flotation unit serving as a clarifier in the separation of activated sludge 
 solids proposed the use of the following expression in the design of clarifi- 
 cation units: 
 
r 
 
 o exp A/S 
 
 uQ/A h ( 1 ) 
 
 where 
 
 C = effluent suspended solids 
 o 
 
 C. = influent suspended solids 
 
 A/S = air to solids ratio 
 
 u = dynamic viscosity 
 
 Q/A, = hydraulic loading rate 
 h 
 
 A limited amount of information on the use of flotation for the 
 separation of microorganisms from their culture media was found in the micro- 
 biological literature. Froth and dispersed air flotation were the methods 
 generally used. With these processes chemicals are added to alter the surface 
 properties of the cells to be floated. Dobias and Vinter (1966) gave a sum- 
 mary of the literature relating to the flotation of microorganisms. Kalyuzhnyi 
 et_ aj_. (1965) , reported from their work with the flotation of yeasts that the 
 biophysical composition of the flocculated yeast cells were determined by the 
 polysaccharide portion of the cell. Cells containing the higher amounts of 
 polysaccharide showed superior flotability. Likewise the percentage of the 
 yeasts cells which were recovered by flotation was greater for the cells which 
 had high bound water contents than for those yeasts having lesser amounts of 
 bound water. The information found in the microbiological literature appears 
 to have a bearing upon the separation of activated sludge by flotation only 
 in that it provides some information regarding the cell surface characteristics 
 which influence f loatabi 1 i ty . This information, however, does not appear to 
 be directly applicable to the flotation process for activated sludges. 
 
III. EXPERIMENTAL EQUIPMENT AND PROCEDURES 
 
 A. Laboratory Activated Sludge System 
 
 The experimental work of this investigation required a system 
 for culturing activated sludge in sufficient quantities so that sludge 
 samples several liters in size could be removed for use in experiments 
 without greatly upsetting the equilibrium of the system. A continuously 
 fed system was preferred to a batch system because it would more nearly 
 represent the operation of a full scale waste treatment facility. As 
 bulked sludges were of primary interest, a solids separation device capable 
 of providing separation of this type of sludge in a reasonable time was 
 necessary. This requirement together with the fact that the effect of 
 sludge characteristics upon flotation was to be studied resulted in the 
 choice of dissolved air flotation as the means of solids separation to be 
 employed. 
 
 A completely-mixed activated sludge system using dissolved air 
 flotation in lieu of gravity sedimentation as a means of separating the 
 activated sludge solids was chosen as the system to be used in this study. 
 The equipment consisted of a constant head box for maintaining a constant 
 flow of dilution water, a primary settling tank, an aeration tank, a flo- 
 tation unit, a pump reservoir and additional pumps along with the necessary 
 appurtenances. A schematic diagram of the system is shown in Figure 1. 
 1. Description of Equipment 
 
 a. Constant Head Box 
 
 A constant head box was used to control the flow of dilution 
 water into the primary settling tank. This consisted of a small wooden 
 
 17 
 

18 
 
 £2h 
 
 O 
 
 0) 
 0) 
 
 
 
 
 1 
 
 
 
 
 \- 
 
 t> 
 
 ||l' 
 
 
 if 
 
 
 
 
 
 
 Z5 
 
 c o 
 O -J 
 
 iz to 
 o I 
 
19 
 
 box divided into two compartments by a wooden partition. Water or domestic 
 sewage, depending upon the operational conditions, was introduced into one 
 compartment. Inside this compartment the top end of a short length of 
 3.5 in. diam plexiglass column which was positioned vertically and sealed 
 to the bottom of the box acted as a stationary weir. Excess water flowed 
 over the weir and was discharged to the sewer. An adjustable 60 V-notch 
 weir located in the partition between the two compartments controlled the 
 rate of flow of liquid into the second compartment. From this compartment 
 the water entered the primary sedimentation tank. 
 
 b. Primary Settling Tank 
 
 An aluminum tank painted inside with epoxy paint was used as a 
 primary settling tank. The tank had a volume of about 120 gal which pro- 
 vided a detention time of 2 hrs at the nominal flow rate of one gpm. It is 
 important that one point be clarified here. For most of the experimental 
 work the flow through the tank was tap water. It was only during the few 
 occasions when domestic sewage was fed instead of tap water that the pri- 
 mary settling tank served any real purpose. 
 
 c. Aeration Tank 
 
 A cylindrical sheet metal tank 5^ in. diam and 5^ in. high was 
 used as the aeration tank. The tank, when filled to the operating level, 
 held approximately 430 gal. The contents of the tank were mixed by the 
 \ action of an 18 in. diam flat bladed turbine and the turbulence created 
 by the rising of the air which was sparged beneath the turbine. The tur- 
 bine was driven through a gear reduction box by a 0.5 hp electric motor 
 which was mounted on a bridge across the top of the tank. The turbine 
 was run at speeds ranging from 60 to 108 rpm. For the majority of the 
 study, the speed was about 96 rpm. A leiman size 206 rotary oil-less 
 
20 
 
 ft 
 
 pressure pump, supplied air which was sparged under the turbine blade 
 
 through a ring of perforated pipes. The combinations of turbine speed and 
 air flow rate used were capable of maintaining the dissolved oxygen in the 
 tank in excess of 2 mg/1 under all the loading conditions which were used. 
 
 d. Continuous Flotation Unit 
 
 Solids separation was accomplished by a continuous dissolved air 
 flotation unit on loan from the Eimco Corporation, Salt Lake City, Utah. 
 The flotation unit consisted of two main parts, viz; the flotation tank 
 where the actual flotation process took place and the system which 
 saturated "the recycle water. A diagram of the flotation tank is shown in 
 Figure 2. The flotation tank consisted of a plexiglass tank 41.75 in. 
 diam and 2k in. high. A circular plexiglass baffle which extended to 
 within 6 in. of the bottom of the tank separated the flotation and clarifi- 
 cation zones. The areas of the flotation and clarification zones were 3-3 
 and 5^5 sq ft respectively. A 1 in. tube conveyed the sludge and recycle 
 water to the point of discharge which was located near the center of the 
 tank and about 10 in. below the liquid surface. The water level in the 
 tank was controlled by the height of the effluent weir. Four scrapers 
 having adjustable skimming depths were used to skim the concentrated float 
 into a collection hopper. Depths ranging between 0.5 and 3 in. were used 
 at various times during the study. The speed of the scrapers could be 
 varied by changing the pulleys on the driving motor and the gear reduction 
 box. A speed of 0.25 rpm was most often used. When continuous skimming 
 
 Manufactured by Leiman Bros. Inc., East Rutherford, New Jersey 
 
 It should be pointed out that the word saturation as used in relation to 
 the dissolved air flotation process is intended to mean only partial sat- 
 uration of the water with air rather than complete saturation. 
 
Clarified 
 Effluent 
 
 Return Sludge 
 Reservoi r 
 
 ^ 
 
 21 
 
 Sludge to be Floated 
 Recycle Water 
 
 Plan 
 
 
 
 
 LZrH 
 
 !■» r 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Floated #- 
 
 — ' 
 
 _2— 
 
 7 
 
 X 
 
 AZ- 
 
 L 
 
 T 
 
 Sludge 
 Withdrawal 
 
 I 
 
 Sludge and 
 
 
 I* 
 
 — ^= — | 
 
 27.5 Iitt- 
 
 ; 
 
 Recycle Water 
 
 c 
 
 LTV 
 
 
 riii 
 
 III 1 
 
 _i_ 
 
 
 
 , / 
 
 r 41.75 in. — 
 
 
 
 I 
 
 Cros 
 
 y . Settled Sludge 
 
 Withdrawal 
 
 s Section 
 
 FIGURE 2 
 
 SCHEMATIC DIAGRAM OF FLOTATION TANK OF LABORATORY 
 CONTINUOUS FLOW UNIT 
 
72 
 
 was not desired, the frequency and length of the periods of collection 
 could be controlled by a time clock. A small pump reservoir of about 
 3 liter capacity received the skimmed float. The collected sludge was 
 returned to the aeration tank by a small float controlled pump. 
 
 The system used to saturate the recycle water consisted of two 
 centrifugal pumps connected in series, an air compressor, a saturation 
 tank, and a pressure reducing valve, plus the necessary valves and piping. 
 The saturation process was carried out by injecting compressed air into 
 the liquid stream at a point upstream of the second pump. The saturation 
 tank was a kl gal pressure tank which received the discharge of the second 
 pump. The tank was equipped with a float controlled valve which discharged 
 air to the atmosphere when the water level in the tank became too low. A 
 0.5 in. needle valve' was used as a pressure reducing valve. The system 
 was operated at *t0 to 50 psig using flotation unit effluent as a recycle 
 water. 
 
 2. Substrate 
 
 Since f i lamentously bulked sludges were of interest in this 
 study, a substrate which was deficient in nitrogen was desired. Molasses, 
 because of its high carbohydrate and low nitrogen content, was chosen as 
 the substrate to be used in this study. Some of the characteristics of 
 molasses are listed in Table 1. 
 
 Ammonium hydroxide and sodium monobasic phosphate were used to 
 ;supply the nitrogen and phosphorous nutrients. The tap water which was fed 
 continuously to the aeration tank as dilution water for the molasses was 
 
 Cat. No. 628-14, 0.5 in. needle valve, manufactured by Central Brass and 
 Aluminum Company, Portland, Oregon 
 
23 
 assumed to supply the other inorganic nutrients necessary for microbial 
 
 growth . 
 
 TABLE i 
 
 CHARACTERISTICS OF THE MOLASSES USED AS SUBSTRATE 
 
 FOR THE CONTINUOUS FLOW LABORATORY 
 
 ACTIVATED SLUDGE SYSTEM" 
 
 Constituent Content 
 
 COD 1 ,040,000 mg/1** 
 
 B0D 5 830,000 mg/1** 
 
 Crude protein 3 percent 
 
 Ca ++ 0.67 percent 
 
 Fiber nil 
 
 Fat nil 
 
 Ash S.k percent 
 
 1959 Feed Ingredient Analysis Table, Nopco Chemical Co., 
 Newark, New Jersey 
 
 Determined by the author 
 
 Because of its high BOD and viscosity it was necessary that the 
 molasses be diluted before being fed to the aeration tank. The molasses 
 was diluted to the desired concentration with tap water. Fifty-five gallon 
 quantities of diluted molasses had to be made up every 2-3 days. In the 
 early part of the study the ammonium hydroxide and sodium phosphate were 
 mixed with the molasses. The substrate was pumped to the aeration tank at 
 the desired rate using a Model T-8 Sigmamotor pump." During the 2 to 3 
 days required to empty the feed drum some fermentation took place. In 
 order to eliminate this problem, the feed system was revised. The molasses 
 was diluted to a COD of 100,000 mg/1 and the sodium phosphate added. The 
 
 Manufactured by Sigmamotor Inc., Middleport, New York 
 
2k 
 
 mixture was refrigerated and fed directly to the aeration tank. The ammonium 
 hydroxide in concentrations of about 100,000 mg/1 NH -N was pumped to the 
 aeration tank separately by a Beckman Model Ikd solution metering pump. The 
 point of discharge was located well below the liquid level to minimize the 
 possibility of losing ammonia gas to the atmosphere. 
 3- Operation 
 
 a. General 
 
 The laboratory activated sludge system was operated on a continuous 
 flow basis using the substrate previously described. The aeration tank was 
 originally seeded with activated sludge obtained from the Champa ign-Urbana 
 Sanitary District's North-east plant. When the organic loadings to the 
 aeration tanks were being changed in order to promote changes in the charac- 
 teristics of the sludge, domestic sewage primary effluent was substituted 
 for tap water to reseed the aeration tank. With the exception of a couple 
 of brief occasions, the dissolved oxygen level in the aeration tank was 
 maintained above 2 mg/1. The nominal dilution water flow rate was one gpm; 
 however, on many occasions it was not possible to keep the system in opera- 
 tion at that flow rate due to the poor flotation characteristics of some 
 of the sludges. In these instances the flow rate had to be substantially 
 reduced. 
 
 The continuous flotation unit was operated using pressures of 
 saturation between kO and 50 psig. In the early part of the study the 
 unit was viewed primarily as a means of accomplishing the solids separa- 
 tion necessary to keep the continuous system in operation rather than as 
 a part of the experimental work. Changes were made in the operational 
 
 Manufactured by Beckman Instruments Inc., Fullerton, California 
 
 I 
 
25 
 
 settings as needed to maintain the continuous system. In the latter part 
 of the work when comparisons were being made between the continuous unit's 
 operation and the batch flotation data, the operation was closely con- 
 trolled during the times when testing was underway. 
 
 b. Procedure for Changing the Characteristics of the 
 Activated Sludge 
 
 The types and characteristics of the microorganisms which make 
 up an activated sludge biomass can be influenced by a number of factors. 
 Nutrient deficiencies, especially nitrogen deficiencies, and aeration 
 tank organic loading rates are two of the prominent ones. Manipulation 
 of the organic loading rate applied to the aeration tank and the BOD to 
 nitrogen ratio (BOD/N) of the substrate were used in this study to promote 
 changes in the microbial populations. The following combinations of F/M 
 and BOD/N were used: F/M = 0.5, BOD/N = kO ; F/M = 1, BOD/N = 25; 
 F/M = 0.5, BOD/N = 10: and F/M = 0.25, BOD/N = 20. It should be pointed 
 out that in this study F/M was defined as the weight of B0D q applied per 
 unit weight of MLSS per day. The BOD/N is defined as the ratio of the 
 weights of BOD,- and nitrogen in the substrate. Based upon the data in 
 the literature which were cited in the previous chapter the first combi- 
 nation created a nitrogen deficient condition, the second a marginal 
 condition and the third and fourth combinations resulted in conditions 
 where nitrogen was in excess or adequate. When the loading conditions 
 were changed, about one-half of the contents of the aeration tank was 
 wasted, the tank was refilled with dilution water and reseeded by feeding 
 domestic sewage primary effluent for a short period of time. 
 
26 
 
 6. Batch Flotation Tests 
 
 The equipment and procedure used in performing the batch flota- 
 tion tests in this study were different in some respects than those which 
 have commonly been used by other researchers. A diagram of the apparatus 
 used in this study is given in Figure 3 • The apparatus consisted of a 
 saturation cylinder, a sludge discharge cylinder, two flowmeters, a source 
 of compressed air, a plexiglass flotation cylinder and the necessary rubber 
 tubing and fittings. 
 
 1 . General Procedure 
 
 The tests were performed by saturating tap water or effluent from 
 the continuous flow laboratory unit with air at pressures of from kO to 
 120 psig in the saturation cylinder, passing the liquid through a pressure 
 reducing valve, blending the air laden water with the sludge to be floated 
 and introducing this blended material into the bottom of the flotation 
 cell. 
 
 For the experiments requiring volumes of recycle water of 2200 ml 
 or less, the saturation of the water with air was accomplished by placing 
 a given amount of liquid in the saturation tank, and bubbling air through 
 the liquid be means of a diffuser stone located near the bottom of the 
 saturation tank. The pressure in the tank and the desired air flow rate 
 vere maintained by manipulating the pressure regulator on the air supply 
 and by adjusting the purge valve located on top of the tank in order to 
 )ermit continuous venting of air from the tank at the required rate. The 
 )ir flow into the cylinder was measured by a flowmeter. The time of satu- 
 ation was 5 mi n . 
 
 While the recycle water was being saturated, the sludge to be 
 loated was placed in a graduated plexiglass cylinder from which it would 
 

27 
 
 2 
 
28 
 
 be later forced at a controlled rate by air. The sludge was usually placed 
 in the cylinder about one minute before the flotation test was to begin. 
 In instances where it was placed in the cylinder earlier and any settling 
 was noticed, the solids were redistributed by inverting the cylinder 3 or 
 4 times. 
 
 After saturation, the water was released at a constant rate through 
 a pressure reducing valve and was blended in a 0.375 in. diam rubber tubing 
 with the sludge to be floated. The sludge was simultaneously forced from 
 the graduated plexiglass container by a flow of air which was controlled by 
 means of a flowmeter. The blended material was discharged vertically 
 through the bottom of a plexiglass column. Columns having nominal diameters 
 of 0.625, 1.5, 2.5, 3.5 and 7-5 in. were used in various batch experiments 
 which will be explained later. Columns were filled to a depth of 20 in. The 
 volume of water left in the saturation tank and the volume of sludge 
 remaining in the sludge cylinder after the flotation tube had been filled 
 were used to verify the ratio of the volumes of water and sludge blended. 
 The approximate solids concentration of the mixture in the flotation column 
 before flotation was calculated. The level of the sludge-water interface 
 was recorded at different time intervals as the sludge rose toward the top 
 of the column. The rate at which the interface ascended was recorded as the 
 interfacial rise rate. Typical curves illustrating the course of the rising 
 of the interface with time are shown in Figure k. 
 
 For tests which required volumes of recycle water greater than 
 2,200 ml the water was saturated in a different manner. The liquid to be 
 
 No. 431 V-stem, bellows valve or a No. 411 1M 2B V-stem bellows valve, 
 manufactured by Hoke Inc., Cresskill, New Jersey. 
 
TIME, mln 
 
 29 
 
 suspended 
 solids, mg/1 
 
 800 
 
 2935 
 
 4070 
 
 10 
 
 15 
 
 20 
 
 25 
 
 FIGURE h. TYPICAL BATCH FLOTATION CURVES 
 
30 
 
 saturated was placed in a steel tank having a capacity of 10 liters. The 
 tank was sealed and compressed air at the desired pressure was applied to 
 the tank. The tank was placed on a gyratory shaker and shaken for 20 min 
 at a rate of 130 rpm. The water treated using this method usually was 
 saturated to a lesser degree than that saturated by the method described 
 previously. Since the experimental work was carried out using gas release 
 data rather than degree of saturation data, the fact that the two methods 
 of saturation did not give the same degrees of saturation was of little 
 importance. 
 
 2. Standard Batch Flotation Test Procedure 
 
 The standard flotation test described in this section was used 
 to compare the floatability of the various sludges with their rheological 
 properties. These tests were performed using the general procedure de- 
 scribed above. Recycle water saturation was accomplished by bubbling air 
 through 1 ,600 ml of water at 40 psig pressure at a rate of about 7-28 1/min 
 for 5 min. The sludge and recycle water were blended and discharged through 
 the bottom of a 2.5 in. diam plexiglass tube. The interfacial rise rate 
 curve was plotted using data collected as the sludge rose in the column. 
 Twenty min after the columns were filled, duplicate samples for determining 
 the concentration of solids in the float were withdrawn from the top 0.25 in, 
 of the float layer with the aid of a rubber bulb and a short length of 
 glass tubing. The samples were stored in tared, capped, glass weighing 
 bottles until they could be weighed. The samples were weighed, dried 
 overnight at 103 C, cooled and reweighed. The concentration of solids 
 in the float as calculated from these data was expressed as percent 
 
 * 
 Model G-10 gyratory shaker manufactured by New Brunswick Scientific 
 
 Company, New Brunswick, New Jersey 
 
31 
 
 solids in the float. The time of sampling, 20 min, was arbitrarily 
 selected. 
 
 3. Comparison of Methods for Performing Batch Flotation Tests 
 
 As stated earlier, the method of carrying out the batch flota- 
 tion tests in this study differed somewhat from the method more commonly 
 used (Eckenfelder et_ aj_. , 1956) (Hurwi tz and Katz, 1959). With the prac- 
 tice which has commonly been used in batch testing the liquid to be 
 saturated is placed in a pressure vessel, the air above the liquid is 
 pressurized to the desired pressure of saturation, and then the liquid 
 is saturated by manually or mechanically shaking the pressurized vessel 
 for some predetermined length of time. The sludge to be floated is 
 placed in a one liter cylinder which serves as the flotation tube. While 
 maintaining the pressure in the vessel, the air laden water is discharged 
 through the pressure reducing valve and by means of a length of laboratory 
 tubing is conveyed to the one liter cylinder where it is mixed with the 
 sludge. When the cylinder is filled to the level required, the recycle 
 water tube is removed and observations of the interface height are made as 
 the flotation progresses with time. 
 
 The method used in this study differs from the common method in 
 two ways: (l) the saturation of the water was accomplished by the con- 
 tinuous bubbling of air through the water rather than by shaking and (2) 
 the blending of the sludge and the air laden water was done outside the 
 flotation tube by combining the proportioned flows of sludge and aerated 
 water prior to their discharge into the flotation tube. 
 
 In order to determine whether any appreciable difference in rise 
 rate resulted from the use of the two different methods of blending, a 
 
32 
 
 test was conducted in which aliquots of a f i lamentously bulked activated 
 sludge were floated using each scheme. Water was saturated by bubbling air 
 through it for both tests. A portion of activated sludge was placed in a 
 one liter cylinder and aerated water was blended with it in the cylinder 
 until the 15 in. level was reached. A second cylinder was filled to the 
 15 in. level using a tube following the blending of sludge and air laden 
 water in the tube. The concentration of suspended solids in both cylinders 
 was essentially the same. The data necessary for determining the rise 
 rates were recorded. 
 
 The appearance of the contents of the two cylinders differed 
 considerably as flotation progressed. The sludge which had been blended in 
 the cylinder did not appear to have nearly as well a defined interface as 
 did the one which had been blended with the air-laden water prior to being 
 introduced into the cylinder. The former seemed to have quantities of 
 somewhat loose, diffuse, clumps of sludge which remained below as the 
 remainder floated toward the top. The sludge which was blended before 
 being introduced into the column, however, had a much more homogeneous 
 appearance immediately after the tube had been filled and seemed to main- 
 tain this as flotation proceeded. The average rise rates recorded on 
 aliquots of the same sludge were 207 and *t00 ft/day, respectively, for the 
 tests using the common and the in-line methods of blending. 
 
 In order to determine whether the difference in rise rates re- 
 sulting from the different methods of blending would prevail for other 
 sludges, a similar experiment was performed at a later date using a less 
 severely bulked sludge. These experiments which were carried out using one 
 liter cylinders gave average rise rates of 326 and 338 ft/day, respec- 
 tively, for the sludges floated using the common method and the in-line 
 
33 
 
 method of blending. These results compared much more favorably as the 
 difference between the rise rates here are within the range of experi- 
 mental error. 
 
 From these tests made using a bulked activated sludge it appears 
 that in-line blending gives a more homogeneous material in the flotation 
 cell at the beginning of flotation with the result that higher rise rates 
 and more complete flotation of the solids occurs. These improvements are 
 believed to result from more even distribution of air among the solids due 
 to the continuous proportioning and intimate contact which takes place 
 during the in-line blending process. The magnitude in the differences in 
 the rise rates obtained using the two methods of blending appear to change 
 as the sludge characteristics change. 
 
 4. Reproducibility of Batch Flotation Data 
 
 The reproducibility of batch flotation results was evaluated by 
 making 10 replicate determinations on a sludge of essential ly the same 
 concentration. The tests were made using a recycle ratio of 0.77, at 
 hO psig and a stock sludge concentration which gave a final suspended 
 solids concentration in the flotation tube of 2,165 mg/1 when the contents 
 of the tube were uniformly mixed. The results of the series are given in 
 Table 2. 
 
 TABLE 2 
 REPRODUCIBILITY OF BATCH FLOTATION DATA 
 
 ! Measurement No. Det. Mean Std. Dev. Coef. of Variation 
 
 i — — 
 
 | Rise Rate 10 225 ft/day 16.8 ft/day 7-47% 
 
 Float Cone. 10 ] .kZ% 0.061% h.3% 
 
3<< 
 
 These data show that there is considerable variation in the 
 results of batch flotation tests, especially in the rise rate determina- 
 tion. Considering the factors involved in the performance of the tests, 
 these variations are not felt to be excessive. The data indicated the 
 need for making multiple flotation tests. Consequently all batch flota- 
 tion tests performed for the purpose of comparing rise rate and float 
 solids concentrations data with other properties of the sludge were made 
 in tr ipl icate. 
 
 5. Effect of Column Filling Time on Interface Rise Rate 
 
 The problem of attaining a uniform distribution of solids in an 
 experimental flotation column is one which must be considered in the per- 
 formance of laboratory flotation tests. The distribution of the solids 
 in the column is influenced by the rate and manner in which the column is 
 filled. The result of filling a column too slowly may be the redistribu- 
 tion of the solids in the column as they start to float even before the 
 filling of the column has been completed. In extreme cases this might be 
 evidenced by the formation of an interface between rising sludge and the 
 clarified underflow even before the column is completely filled. On the 
 other hand, it is undesirable to fill the column so rapidly that the tur- 
 bulence created will break up the floe to the point that a long lag time 
 for ref locculation is required before interface formation is seen. 
 
 Dick (1965) in his studies on the thickening of activated sludge 
 by gravity sedimentation explored in detail the problem of attaining uni- 
 form solids distribution in 3-5 in. diam laboratory settling columns. The 
 I method he developed for minimizing the uneven distribution of solids in- 
 volved filling the column from the bottom by pumping sludge at a rate 
 which was rapid enough to prevent sedimentation from commencing during 
 
35 
 filling. The filling time also was slow enough to prevent the turbulence 
 from being so great that redistribution would take place during the lag 
 period between the end of the filling procedure and the time when sedi- 
 mentation at a constant velocity began. Farnsworth ( 1 967 ) also investi- 
 gated the effect of column filling time upon settling velocity. The plot 
 showing the relationship between filling time and settling velocity passed 
 through a range of minimum settling velocities. The filling time chosen 
 as giving the most nearly uniform solids distribution also gave a settling 
 velocity which plotted near the minimum value. 
 
 The column filling times chosen for use in this study were those 
 which in a series of tests using various filling times gave the lowest 
 rise rates. The determinations of the column fill times to be used here 
 were made in the following way. Sludge was taken from the continuous 
 activated sludge system in sufficient quantity to complete a series of 
 tests using a column of a particular diameter. The recycle water from 
 the saturation system of the continuous flotation unit was used as the 
 recycle water for the batch tests with the columns. With this exception a 
 system very similar to the one described earlier for performing batch flo- 
 tation tests was used. Sludge was displaced from a 13 liter carboy at a 
 controlled rate by air. The sludge was blended with recycle water and 
 discharged into a flotation column vertically through the bottom. The 
 flow rate of the recycle water was controlled by maintaining a constant 
 differential head on a manometer which was connected across an orifice on 
 a line from the discharge side of the continuous unit saturation system. 
 The times of filling of the columns were varied by continuously bleeding 
 off some of the blended sludge as the column was being filled. The time 
 required to fill a column was thus controlled by the fraction of the feed 
 
36 
 
 which was bled off. In any given series of tests made on a particular column 
 size the air to solids ratio was held constant. 
 
 In order to decrease the turbulence in the columns during filling 
 at the rapid rates required, it was necessary to feed the sludge into the 
 larger columns through more than one tube. Two 0.5 in. diam tubes were used 
 to feed the 7-5 in. diam column. All tests were made with the columns being 
 filled to a height of 20 in. 
 
 The data showing the effect of column fill time on interfacial rise 
 rate are plotted in Figure 5. Based upon these data it was decided that 
 a fill time of one minute or less for the 2.5 in. diam and 3-5 in. diam 
 columns and of about 3 minutes for the 7.5 in. diam column would give reliable 
 results and would be used in future tests. Fill times of less than one min- 
 ute were used for the 0.625 in. diam and 1.5 in. diam tubes. It was noted 
 during the course of the ensuing experiments that when the columns were 
 filled at these rates considerable turbulence was evident during the filling 
 process. For the higher concentrations of sludge the turbulence seemed to 
 be confined primarily to the bottom one third of the column during filling 
 while the sludge above seemed to rise as a unit with the filling of the 
 column. For the lower sludge concentrations there was considerable turbulence 
 throughout the column during and for some time after the column had been 
 filled. 
 
 6. Effect of Column Diameter on Rise Rate 
 
 The importance of considering the effect of column diameter upon 
 settling velocity when making batch laboratory sedimentation tests has 
 been pointed out by recent investigators studying the thickening of acti- 
 vated sludges by gravity sedimentation (Vesilind, 1968; Cole, 1969). 
 
37 
 
 tep/*± '3ivy asiy 
 
38 
 Earlier workers in the same or similar areas of interest had determined 
 that column diameters of about 2 in. were suitable for performing settling 
 tests and that larger columns had no beneficial effect (Mancini, 1962) 
 (Kammermeyer , 19^1). Vesilind ( 1 968) in his work with activated sludges 
 investigated the effects of column diameter on settling rates using col- 
 umns having nominal diameters of h, 8, 18, and 36 in. The sludge concen- 
 trations studied ranged from 2,000 to 10,000 mg/1 . Figure 6 is taken 
 from his data and shows the influence of column diameter on settling rates 
 over a range of solids concentrations. At low concentrations two dif- 
 ferent mechanisms were used to explain the behavior of the sludge in the 
 small diameter columns. The first is the phenomenon of bridging. At high 
 solids concentrations the floe particles of the sludge interact to form a 
 bridging structure across the column resulting in slower settling rates 
 than are attained at the same concentration in larger columns. The other 
 mechanism which probably operates at all sludge concentrations results 
 in faster settling rates in small columns than would be experienced in 
 larger columns containing sludge at the same concentration. In this case 
 water being squeezed out as the sludge compresses takes the path of least 
 resistance and flows along the smooth wall of the column rather than 
 going by the more tortuous route through the pores of the sludge mass. 
 
 One other phenomenon of batch settling tests which is related to 
 column diameter concerns the agglomeration of sludge floe into clumps and 
 the effect of this agglomeration upon settling rates. The agglomeration 
 
 ' or clumping together of sludge floe provides channels and larger pore 
 spaces for the passage of escaping water thus permitting faster settling 
 
 I rates. Vesilind ( 1 968) studied the incidence of agglomeration in columns 
 of several diameters and found that in order to get reproducible laboratory 
 
39 
 
 _ 
 
 
40 
 
 settling data consistently it was necessary that test conditions be con- 
 trolled such that complete agglomeration would take place in all tests. 
 He found agglomeration to be enhanced by low solids concentrations, 
 larger column diameters, and long filling times. Incomplete agglomeration 
 or lack of agglomeration, then, is more likely to occur in tests performed 
 using small diameter columns and sludges having high solids concentrations. 
 
 Since the flotation process is essentially sedimentation in re- 
 verse, it seemed advisable to investigate the effect of column diameter 
 upon the rise rates obtained from batch flotation tests. This was done 
 using columns of 1.5, 2.5, 3-5 and 7-5 in. diam and the sludge blending 
 procedure described in the previous section. The columns were filled to 
 the 20 in. level in all tests as previously stated. The data obtained 
 from several series of tests are presented in Figure 7. It should be 
 pointed out that these data were collected on different dates and not 
 necessarily at the same air to solids ratios. Therefore, the data of one 
 series of tests should not be compared directly with those of another 
 series. The air to solids ratio was held constant for all tests made in 
 any particular series. Thus, comparisons can be made between rise rates 
 in columns of various sizes in a given series of tests but not between 
 columns in different series of tests. 
 
 These data plot reasonably well as straight lines on log paper. 
 This is in general agreement with the data of Vesilind (1968). From this 
 plot it may be possible by extrapolation to obtain rise rate data for 
 columns of other diameters. The data shown here indicate that decreasing 
 column diameters gave higher rise rates for the dilute sludges. For 
 sludges in the 5,000 to 8,000 mg/1 range, column diameter did not have much 
 influence on rise rate. An important point to note, however, is that on 
 
I»1 
 
 1000 
 900 
 800 
 
 700 
 
 600 
 
 500 
 400 
 
 300 
 
 200 
 
 100 
 90 
 80 
 70 
 60 
 
 1 1 — r— ft 
 
 50 
 ko 
 
 30 
 20 
 
 10 
 
 H50 mg/1 
 
 1750 mg/1 
 
 2300 mg/1 
 
 5100 mg/1 
 
 8350 mg/1 
 6800 mg/1 
 
 '8200 mg/1 
 
 *-+ 
 
 I 1 I I 
 
 3 ~** 5 6 7 8 9 10 
 COLUMN DIAMETER, In. 
 
 20 
 
 FIGURE 7- EFFECT OF COLUMN DIAMETER ON RISE RATE 
 

kl 
 
 two occasions while substantial rise rates were being recorded for the 
 7-5 and 3-5 in. diam columns, the sludges in the 2.5 and 1.5 in. diam 
 columns exhibited virtually no rise rate at all. This was probably due 
 to bridging and to lack of agglomeration which was discussed previously. 
 It should be pointed out that the results obtained were depen- 
 dent upon the sludge characteristics and would be expected to vary from 
 sludge to sludge. 
 
 Based upon these data it is apparent that in performing batch 
 flotation tests one should consider the effect of column diameter upon 
 rise rate. Column sizes as large as practicable should be used to mini- 
 mize the possibility of bridging and to encourage agglomeration of the 
 sludge in the column. In this study, columns of 3.5 and 7-5 in. diam 
 were used. The 7-5 in. diam column was used only for tests made using 
 high sludge concentrations. 
 
 7. Effect of Column Depth on Initial Rise Rate 
 
 Until recent years the procedures used in the design of gravity 
 thickeners have not taken into account the effect of the depth of sludge 
 upon the subsidence rate of sludges in laboratory batch test columns. 
 I The Kynch theory which was developed for and has been proven to be correct 
 I for ideal suspensions had been used with modifications in the design of 
 
 I thickeners for thickening activated sludge, a flocculent material. A few 
 investigators had considered the effect of depth on settling rate. 
 ; Gaudin, Fuerstenau and Mitchell (1962) found that for concentrated Kaolin 
 suspensions depth influenced the settling velocity but at dilute concen- 
 trations the velocity was practically independent of depth. Clifford and 
 Windridge (1932) found that the settling velocity of activated sludge was 
 directly proportional to depth, but their work was limited to depths of 
 
43 
 
 less than 1 ft. 
 
 Taking issue with the prevailing concept that the initial set- 
 tling rate of a suspension was a function only of the solids concentration, 
 Dick (1965) investigated the effect of initial sludge depth on the subsi- 
 dence rate of activated sludges in laboratory batch test columns. His 
 experiments with activated sludge showed that the subsidence rate is great- 
 ly influenced by sludge depth. By arranging his data such that the initial 
 depth divided by initial settling rate was plotted against the initial 
 depth he obtained lines which could be described by an equation of the 
 fol lowing form: 
 
 {£ - R + SDo ( 2 ) 
 
 where 
 
 Do = initial depth of sludge 
 Vo = initial subsidence rate 
 S = slope of the plot 
 
 R = w atDo = 
 
 The magnitude of the value of R is a measure of the extent of 
 retardation of settling and hence was called the retardation factor. The 
 retardation factor was interpreted as being a measure of the support which 
 the sludge interface received from the underlying layers of sludge. He 
 found that the retardation factor was dependent upon the solids concentra- 
 tion and the nature of the sludge. Vesilind ( 1 968) found the retardation 
 factor also to be dependent upon the diameter of the column in which the 
 tests are performed. 
 
 To the author's knowledge, the effect of sludge depth upon rise 
 rates in batch flotation tests has never been investigated. In order to 
 
kk 
 
 determine whether or not the depth of the sludge in a flotation column had 
 any effect upon the interfacial rise rate, a series of experiments were 
 performed using laboratory developed sludges and the same batch flotation 
 procedure described previously in the section dealing with the influence 
 of column fill time on rise rates. Experiments were performed using three 
 sludge depths under each set of conditions. For any given series of tests 
 using three depths, the air to solids ratio and the suspended solids con- 
 centration were held constant. The air to solids ratios for different 
 series of tests, however, were not necessarily the same. The data for these 
 tests are plotted in Figure 8. These plots show that the interfacial rise 
 rate is directly proportional to the initial depth of sludge in the column 
 at least for depths between 1.5 ft and k ft. The effects of depth are seen 
 to be greater at high sludge concentrations. When the data are plotted with 
 D /V on one axis against D on the other the data plot as straight lines 
 fitting the equation developed by Dick (1965). Thus, a retardation factor 
 exists in flotation as well as in gravity sedimentation. The data for 
 several sludges are plotted in Figure 3. 
 
 The sludge in gravity sedimentation which provides the structural 
 support for the overlying layers is believed to derive its support from 
 physically contacting the bottom of the column. With flotation there is no 
 fixed surface against which the most concentrated sludge layer can bear for 
 support. The supporting force here may be developed by the weight of that 
 amount of sludge in the topmost layers of float which are lifted above the 
 original liquid level in the column and by the friction of the sludge against 
 the column wal 1 . 
 
 8. Effect of Pressure of Saturation on Batch Test Rise Rates 
 
 In order to obtain batch flotation data at high blended sludge 
 concentrations, recycle water saturated at pressures up to 120 psig were 
 
1*5 
 
 2650 mg/1 
 
 D 1*800 mg/1 
 
 ▼ 5800 mg/1 
 
 • 6400 mg/1 
 
 A 6900 mg/1 
 
 I 
 
 10 
 
 20 30 
 COLUMN DEPTH, In. 
 
 40 
 
 50 
 
 FIGURE 8. EFFECT OF COLUMN DEPTH ON RISE RATE 
 
ke 
 
 160 
 
 \ko 
 
 120 
 
 00 
 
 1 
 
 Suspended Sol Ids 
 
 6900 mg/1 
 
 5 80 
 
 3 60 — 
 
 A0 
 
 20 
 
 FIGURE 9, 
 
 2 3 k 
 
 SLUDGE DEPTH, ft 
 
 DETERMINATION OF RETARDATION FACTOR 
 
used in order to provide the air necessary to maintain a constant air to 
 solids ratio. The air release data were measured for pressures of kO , 60, 
 80, and 120 psig. Two series of tests were performed using recycle water 
 which had been saturated at 40, 60, and 80 psig to determine if the changing 
 of the pressure had any influence on the rise rate. The data obtained for 
 the two series of tests are shown in Figure 10. From this figure it can 
 be seen that even though the air to solids ratio and blended sludge con- 
 centration were held constant in each series the rise rates decreased 
 exponentially as the pressure increased. This may have been due to the 
 increased probability of bubble agglomeration at the higher pressures. 
 Katz (1963) reported that the number of air bubbles produced upon depres- 
 surization will be directly proportional to the product of the absolute 
 pressure and the rate of flow of the pressurized air charged stream. He 
 also stated that increasing the pressure substantially over 80 psig may 
 not result in an increased production of useful bubbles because of the 
 increasing rate of bubble agglomeration. Theoretically, the number of 
 air bubbles necessary to achieve satisfactory flotation is proportional 
 to the number of particles to be floated. The discharge of bubbles having 
 diameters of a millimeter or more was noted here when a saturation pres- 
 sure of 120 psig was used. 
 
 Ettelt ( 1 964) has stated that smaller bubbles have less liquid 
 to displace from the surface of the solids to which they must attach and; 
 therefore, they attach more readily than larger bubbles. Additionally, 
 because their terminal velocities are less than those of larger bubbles, 
 the detention time is also increased, which appreciably enhances the oppor- 
 tunity for contact with the solids. 
 
 Vrablik (1958) reported the bubble size distributions he 
 
A8 
 
 10 
 
 20 TJ0~ 60 ^5b" 100 120 
 PRESSURE OF SATURATION, psig 
 
 HO 
 
 FIGURE 10. EFFECT OF PRESSURE OF SATURATION ON RISE RATE 
 
*»9 
 
 obtained from depressur izing water which had been saturated at pressures 
 of 20 to 50 psig. His data showed that the range of bubble sizes were 
 identical for the two pressures but that the higher pressure gave a more 
 uniform distribution of sizes. Both pressures gave modal values in the 
 70 to 90 micron diam range. Data for higher pressures were unfortunately 
 not available. 
 
 9. Summary 
 
 A number of factors which influence the results obtained from 
 batch flotation tests have been pointed out in the preceeding discussion. 
 The data presented showed that column diameter, the column filling time 
 and column depth, and pressure of saturation do affect the interfacial 
 rise rate. The magnitude of the effects of column diameter and column 
 height would be expected to vary depending upon the characteristics of the 
 sludge. 
 
 A test procedure for performing the batch flotation tests was 
 described. This method which included blending of the sludge and recycle 
 water prior to their discharge into the flotation column was found to give 
 higher rise rates than the more commonly used procedure, especially with 
 the more filamentous sludges. 
 
 Nowhere in the literature was anything found to indicate that the 
 factors mentioned previously have been considered in the performance of the 
 batch flotation tests or in the use of the data obtained from them. That 
 the results obtained from prototype flotation units do not always agree 
 with the results predicted from the data obtained from batch flotation 
 tests has, however, been reported (Ettelt, 1964; Katz, 1968; Shell, 1970; 
 and Home e_t^ aj_. , 1 962 ) . Generally the prototype units give higher float 
 concentrations than are obtained in the batch flotation tests, but not 
 
50 
 always . 
 
 The dynamic and hydraulic factors existing in a continuous full 
 scale unit undoubtedly influence the results obtained. On the other hand, 
 it is suggested that consideration of the factors pointed out earlier 
 which affect the batch flotation test may give data which more nearly 
 approximate the results which may be obtained from a full scale unit. 
 
 C. Sol ids Flux Analysis 
 
 1. Procedure for Obtaining Data from Continuous Laboratory 
 Flotation Unit 
 
 The object of this portion of the study was to investigate the 
 possibility of using the solids flux method of analysis to predict the 
 operation of continuous flotation units. With the continuous system 
 operating at assumed steady state conditons the following operational 
 data were needed in order to make comparisons between it and the batch 
 flux curve: (1) the solids loading to the flotation unit, (2) the vol- 
 umetric and mass rates of sludge withdrawal from the flotation unit and 
 (3) the air to solids ratio being used. The methods used to obtain these 
 data will be explained with the aid of the schematic diagram of the con- 
 tinuous flow system as shown in Figure 11 . Steady state conditions in 
 the flotation unit were assumed to exist after the unit had operated under 
 a given set of conditions for a period of at least 3 hrs. 
 
 With the MLSS having been determined, the hydraulic flow rate, 
 |Qr, to the flotation unit is the only additional information needed to 
 determine the flux being applied to the flotation unit. This information 
 jwas ascertained by measuring the dilution water flow rate, Q; the sub- 
 strate feed rate, q; and the sludge recycle rate, 0- R . The float skimmers 
 

51 
 
 To Sewer 
 
 Q - Dilution Water Flow Rate 
 
 q - Substrate Flow Rate 
 
 Q_ ■ Aeration Tank Discharge Flow Rate 
 
 Sludge Recycle Flow Rate 
 
 Liquid Recycle Flow Rate 
 
 Flotation Unit Effluent Flow Rate 
 
 l SR 
 >LR 
 
 FIGURE 11. 
 
 SCHEMATIC FLOW DIAGRAM OF THE CONTINUOUS 
 FLOW LABORATORY ACTIVATED SLUDGE SYSTEM 
 
52 
 discharged their sludge to a small pump reservoir where it was collected 
 until it made up about 3 to ^ liters and then it was pumped back to the 
 aeration tank. The time interval between pumping, depending upon the 
 operational conditions, ranged from 2 to 6 min. The measured amount of 
 sludge returned at a pumping was divided by the pumping interval to give 
 the average sludge recycle rate. The sum of the flow rates of the dilu- 
 tion water, the substrate and the sludge recycle gave the flow rate to 
 the flotation tank. The product of the weight of the MLSS per unit volume 
 of mixed liquor and the hydraulic flow rate divided by the area of the 
 flotation zone was the applied flux. The solids concentration of the 
 float was determined in the following manner. The float returned to the 
 aeration tank during a single pumping cycle was collected and mixed 
 thoroughly. A 50 ml aliquot was withdrawn and diluted 1:5 or 1:10. 
 Duplicate 10 ml samples were processed according to the procedure used 
 for MLSS determinations. 
 
 The liquid recycle rate and the gas release rate had to be 
 determined so that the air to solids ratio could be calculated. Since 
 the flowmeter on the recycle line was not sufficiently accurate, the 
 liquid recycle rate was determined by measuring the rate of effluent dis- 
 charge from the flotation unit, Q_, and running a flow rate balance on 
 the flotation unit. The air released from the recycle water was measured 
 as described in the following section. An example of a typical calcula- 
 tion is given in the Appendix. 
 
 2. Procedure for Determining Quantity of Air Released 
 
 The air to solids ratio is an important parameter in flotation. 
 In order to match the air to solids ratios of the batch tests to those of 
 the continuous unit, it was necessary to determine the amount of air 
 
53 
 released from the recycle water upon depressurizat ion for each system at 
 the particular operating conditions employed. A method somewhat similar 
 to the one used by Vrablik (1958) was employed to make these measurements. 
 The following procedure was used. 
 
 Two graduated 2-liter gas sampling tubes clamped vertically to a 
 ring stand were connected at the top with a short length of laboratory 
 tubing as shown in Figure 12 . A manometer using colored water as an 
 indicating liquid was positioned such that it could be connected to the 
 tubing joining the two sampling tubes. The bottom end of one of the sam- 
 pling tubes was connected by a short length of tubing to the discharge 
 side of the pressure reducing valve of the water saturation system. Prior 
 to beginning a test the body of tube 1 was empty and the water level was 
 adjusted to level A in the lower tubing connection. The manometer was not 
 connected at this time. Tube 2 was filled through the bottom with dibutyl 
 phthalate to a level of about 1 .5 liters. One leg of the manometer was 
 then connected to the tubing joining the two sampling tubes. The pressure 
 reducing valve was opened to permit air laden water to discharge to tube 1. 
 Simultaneously, dibutyl phthalate was drained from tube 2 at a rate suf- 
 ficient to prevent the loss of manometer fluid and yet slow enough so that 
 a slight positive pressure was maintained on the side of the manometer 
 connected to the apparatus. When about a liter of water had been dis- 
 charged into tube 1 the pressure reducing valve was closed and enough 
 dibutyl phthalate was withdrawn to bring the liquid level in the manometer 
 back to its original level. A minute or two was allowed to make the final 
 adjustment. Once the final adjustment was made the manometer was discon- 
 nected and water was drained from tube 1 until the water level returned to 
 point A. The volumes of water and dibutyl phthalate withdrawn were 
 

Sk 
 
 Tube 
 2 
 
 DBP 
 
 Tube 
 1 
 
 FIGURE 12 
 
 APPARATUS USED TO MEASURE VOLUME OF GAS 
 RELEASED FROM WATER UPON DEPRESSURIZATION 
 
55 
 
 measured and the difference between these two volumes represented the 
 volume of gas released in passage through the pressure reducing valve. 
 The volumes measured were corrected to standard conditions. 
 
 With the batch saturation tank, the gas sampling tube was con- 
 nected to the pressure reducing valve by means of a short length of hose. 
 For the laboratory continuous unit the hose was connected to a valve which 
 was identical to the pressure reducing valve but which had been installed 
 in a tee on the upstream side of the pressure reducing valve which served 
 the continuous unit. 
 
 Data collected from the batch saturation tank indicated that 
 for the AO psig and 80 psig pressures of saturation about 92 percent and 
 96 percent, respectively, of the amount of air which could theoretically 
 be released from solution upon depressur ization , was released. The extent 
 to which the valve was opened did not make a significant difference in the 
 volume of gas released from solution. 
 
 For the continuous flotation unit, when operating at 48 to 50 
 psig, about 70 percent of the amount of air which could theoretically be 
 released upon depressur izat ion was released. 
 
 3. Procedure for Obtaining Data for Batch Flux Curves 
 
 The batch flux curves were drawn from calculations based upon 
 the data collected from a number of rise rate tests made using laboratory 
 columns containing blended sludges at various solids concentrations. The 
 air to solids ratio, which determines the driving force, was kept constant 
 for all tests relating to a particular flux curve. This was accomplished 
 by utilizing, for successive tests, recycle water which had been saturated 
 at different pressures and by employing stock sludges of different concen- 
 trations. The air release data for the various pressures of saturation 
 
56 
 were obtained by the method described in the previous section. As a rule 
 the tests requiring low concentrations of sludge in the flotation tube 
 were made using recycle water which had been saturated at kO psig. Sub- 
 sequent tests, necessitating the use of higher concentrations, were made 
 with more concentrated stock sludges and with recycle water that had been 
 saturated at higher pressures. 
 
 The sludges used in these tests were obtained by collecting the 
 float from the continuous unit and letting it sit quiescently and thicken 
 further. After thickening, the sludge was sampled to determine the con- 
 centration and then held under aeration for the duration of the experiment. 
 The more dilute stock sludges were made by diluting the concentrated stock 
 with continuous unit effluent. The effluent from the continuous unit was 
 saturated under pressure and used as the pressurized recycle water for the 
 batch tests. In the case of the tests performed on the full scale flota- 
 tion unit primary effluent was used as the recycle water. The full scale 
 flotation unit also used primary effluent for recycle water. 
 
 The equipment needed to perform the tests was essentially the 
 same as that described previously in the discussion of the batch flotation 
 testing procedures. The columns used here were 3-5 in. in diam for the 
 lower sludge concentrations and 7-5 in. diam for the highest concentrations. 
 The large column was used in order to try to minimize the opportunity for 
 bridging at high solids concentrations. For all tests performed using 
 the 3.5 in. diam columns the small plexiglass saturation cylinder was 
 used to saturate the pressurized recycle water. Because of its limited 
 i volume it could not be used in the tests performed using the 7-5 in. diam 
 
 column. A steel tank of about 10 liters capacity was used to supply the 
 I recycle water when large volumes were needed. 
 
57 
 
 The blending schedule for a flux curve was determined as follows. 
 
 The air to solids ratio of the continuous unit was determined. Since the 
 
 batch tests had to be performed at the same air to solids ratio the batch 
 
 and continuous air to solids ratios were equal and constant. 
 
 V V P 
 (A/S). = (A/S)_ = -^ — 3 — SL ( 3 ) 
 
 where 
 
 (A/S) = ai r to solids ratio for batch test, g gas/g solids 
 (A/S) = air to solids ratio for continuous unit, g gas/g solids 
 
 V = volume of recycle water, 1 
 w ; ' 
 
 V = volume of stock sludge, 1 
 
 C = concentration of stock sludge, g/1 
 
 V = volume of gas released from recycle water saturated at 
 9 stated pressure, 1/1 at STP 
 
 P = density of gas, g/1 at STP 
 Then at any given pressure of saturation, 
 
 A/S = K = J^_ ( k ) 
 
 V P V c 
 
 g g s s 
 
 S 
 
 Once the K value was determined for each pressure then it was only a 
 matter of multiplying the stock sludge concentration and the K constant 
 to determine the ratio of the volumes of dilution water to sludge which 
 was necessary to maintain the constant air to solids ratio. The result- 
 ing solids concentration, C , in the flotation tube after blending is 
 
 m 
 
58 
 given by the equation: 
 
 C 
 C 
 
 , . _w ( 6 ) 
 
 V 
 s 
 
 If it was more desirable to select values of blended sludge concentration, 
 
 the concentration of the stock sludge required to maintain constant A/S at 
 
 a given pressure of saturation was found by the equation: 
 
 s 1 - K C ( 7 ) 
 
 A typical blending schedule for an air to solids ratio of 0.0046 
 
 is shown in Table 3- The procedure used in this study was to select 
 
 concentrations of stock sludge and by looking at the values of C in the 
 
 blending schedule the most desirable V /V was selected. Because of the 
 3 w s 
 
 limited volumes of the saturation cylinder and stock sludge cylinder it 
 was necessary to keep the ratio V /V between about 0.75 and 1.45. This 
 necessitated the changing of pressures more often than otherwise would 
 have been desired. 
 
 It should be pointed out that in this study, the A/S was kept 
 low in order to permit higher blended sludge concentrations to be obtained. 
 The values used here ranged from 0.0046 to 0.01 calculated on a gas re- 
 leased basis. A common value for the air to solids ratio seen in the 
 literature in connection with full scale units is 0.02 (Ettelt, 1964; Mayo. 
 1965; Jones, 1968). 
 
 D. Study of Rheological Properties 
 1. Description of Viscometer 
 
 The rheological properties of the activated sludges were 
 
TABLE 3 
 TYPICAL BLENDING SCHEDULE FOR A/S = 0.0046 
 
 59 
 
 Pressure 
 of Saturation 
 
 Batch 
 Gas Release 
 
 g/i 
 
 s 
 mg/1 
 
 V /v 
 
 w s 
 
 m 
 mg/1 
 
 5 
 
 ,000 
 
 0.45 
 
 3,470 
 
 10 
 
 ,000 
 
 0.89 
 
 5,300 
 
 15 
 
 ,000 
 
 1.34 
 
 6,430 
 
 10 
 
 ,000 
 
 0.556 
 
 6,430 
 
 15 
 
 ,000 
 
 0.834 
 
 8,180 
 
 20 
 
 ,000 
 
 1.12 
 
 9,440 
 
 25 
 
 ,000 
 
 1.39 
 
 10,450 
 
 15 
 
 ,000 
 
 0.64 
 
 9,150 
 
 20 
 
 ,000 
 
 0.852 
 
 10,800 
 
 25 
 
 ,000 
 
 1.063 
 
 12,100 
 
 15 
 
 ,000 
 
 0.387 
 
 10,800 
 
 20 
 
 ,000 
 
 0.516 
 
 13,200 
 
 25 
 
 ,000 
 
 0.645 
 
 15,200 
 
 40 
 
 60 
 
 80 
 
 120 
 
 0.0517 
 
 0.0827 
 
 0.1079 
 
 0.177 
 
 0.089 
 
 0.0556 
 
 0.0426 
 
 0.0258 
 
 Note: The recycle water for 120 ps i was saturated by mechanical shaking, 
 Water at all other pressures was saturated by diffused air. 
 
60 
 
 measured using a coaxial cylinder rotational viscometer. The viscometer 
 consisted of two plexiglass cylinders aligned concentrically in the vertical 
 position. The inner cylinder, the bob, was supported by a torsion wire, a 
 0.007 in. diam music wire 3 in. long. One end of the torsion wire was con- 
 nected to a fixed external support while the other end was attached to a 
 stiff rod which was connected vertically to the axis of the bob. The bob was 
 weighted with a steel weight in order to keep it immersed when in use. The 
 outer cylinder was mounted vertically on a hollow shaft which served as the 
 means of rotation of the cylinder. The shaft, by means of a set of pillow 
 blocks, was mounted on a rigid frame. The outer cylinder was driven via a 
 belt-pulley system by a variable speed motor which was capable of providing 
 a wide range of angular velocities. The speed of rotation was calculated 
 from the number of impulses recorded by an impulse counter which was activated 
 by a microswitch which closed with each 0.25 revolution of the drive shaft. 
 
 The diameters of the unlined cylinders were 1.24 in. and 1.75 
 in., respectively, for the bob and the cup. The surfaces of the cylin- 
 ders were lined with a rubber matting to prevent slippage between the 
 suspension and the smooth surface of the cylinders. It may be noted that 
 the 0.27 in. gap between the cylinder walls is larger than normally used in 
 commercial viscometers. Dick ( 1 965) found this to be necessary to 
 accommodate the sized particles associated with activated sludge. A 
 short length of wire attached radially to the top of the bob served as 
 a pointer. The deflection of the pointer was read in degrees from an 
 externally supported scale which was aligned concentrically with the 
 
 No. 3070 Neotex Protective Mesh distributed by Research Products Corpora- 
 tion, Madison, Wisconsin 
 
61 
 outer cylinder. A dampening device consisting of paddles immersed in an 
 oil bath was attached to the rod supporting the bob and served to dampen 
 out oscillations in the deflection of the bob thus permitting readings to 
 be made more rapidly. A schematic diagram of the viscometer is shown in 
 Figure 13. 
 
 2. Calibration of Viscometer 
 
 In order for the data collected from the viscometer to be mean- 
 ingful, the viscometer had to be calibrated using a fluid of known vis- 
 cosity. One of the main factors which had to be considered in the cali- 
 bration was the effect of the viscous drag on the end of the bob, i.e. the 
 determination of the end effect. Wang (1967) in his work with this 
 viscometer showed that the end effect, h , was dependent upon the viscos- 
 ity of the fluid being tested. The magnitude of the end effect decreased 
 as the viscosity of the fluid being tested increased. 
 
 In this study several Newtonian fluids of different viscosities 
 were used in the calibration of the viscometer. The fluids used included 
 an aqueous solution containing 10 percent glycerol by weight and 20, hO , 
 and 60 percent aqueous solutions of sucrose. These fluids had viscosities 
 ranging from 1.153 x 10 dyne sec/sq cm to 0.^386 dyne sec/sq cm at 25°C. 
 The end effect varied with the viscosity as shown in Figure ]h . While 
 the diameters of the unlined cylinders were known, it was necessary to 
 determine their effective diameters after having been lined so that the 
 instrument constants which describe the geometric properties of the 
 viscometer could be determined. 
 
 The calibration was performed according to the following pro- 
 cedure. With the bob raised so that the gap between it and the bottom 
 of the outer cylinder was 1.5 in. the viscometer was filled with a 
 
N\\\\\\ 
 
 62 
 
 Torsion Wf re with Of 1 
 Dampening 
 
 Drive Shaft 
 
 Pointer 
 
 Rubber Mat 
 Roughening on 
 Inner and Outer 
 Cyl Inders 
 
 Gap - 0.27 In, 
 
 FIGURE 13 
 
 SCHEMATIC DRAWING OF COAXIAL CYLINDER 
 ROTATIONAL VISCOMETER 
 
63 
 
 1000 
 
 100 — 
 
 10 — 
 
 1 r 
 
 10$ Glycerol 
 
 20$ Sucrose 
 
 Activated Sludge 
 
 40$ Sucrose 
 
 60$ Sucrose 
 
 95$ Glycerol (Wang) 
 
 12 3^! 
 
 END EFFECT, in. 
 FIGURE 14. RELATIONSHIP BETWEEN VISCOSITY AND END EFFECT 
 
ik 
 
 Newtonian liquid until 6 in. of the bob were submerged. As the outer 
 cylinder was rotated, the deflections of the inner cylinder, due to the 
 torque applied by the shearing of the fluid in the annular space between 
 the two cylinders, were recorded for a number of speeds of rotation. 
 This procedure was repeated with the liquid level adjusted so that h and 
 5 in. of the bob were submerged. The data for all depths were then 
 plotted as rotative speed versus deflection. The data for a typical 
 calibration are shown in Figure 15- Both cylinders were lined during 
 this calibration procedure. The reciprocals of the slopes were plotted 
 against the depths of submergence as shown in Figure 16. A line drawn 
 through these points and extrapolated to the h axis gave the value of ho, 
 the end effect. This procedure was followed with each of the Newtonian 
 fluids. The spring constant for the torsion wire was calculated using 
 laboratory experimental data and data provided by the manufacturer of the 
 wire, National Standard Co., Niles, Michigan. The only remaining unknown 
 was the effective thickness of the matting. This was calculated using 
 the following equation (Wang, 1967): 
 
 e/n _ 8 tt y f 1 1 I " 1 ,o\ 
 
 FTTo- ~ L(R. + s) L ~ (R - 6)*J ( 8 } 
 
 where 
 
 fl/o = deflection, degrees 
 rotative speed, rpm 
 
 h + ho = depth of submergence + end effect 
 
 u = viscosity of the Newtonian fluid 
 
 a = torsional spring constant 
 
 R. = radius of inner cylinder, unroughened 
 
 R = radius of outer cylinder, unroughened 
 
 6 = effective thickness of the matting 
 
65 
 
 2 3 4 5 6 7 
 
 DEFLECTION, degrees 
 
 FIGURE 15. VISCOMETER CALIBRATION CURVES 
 
 10 
 
 3 2 10 
 END EFFECT, In. 
 
 12 3 4 5 
 DEPTH OF SUBMERGENCE, in. 
 
 FIGURE 16. DETERMINATION OF END EFFECT 
 
66 
 
 For a given viscometer all of its geometric characteristics 
 may be combined into one constant, K , (Green, 19^9). For a Newtonian 
 fluid the viscosity, u, may be calculated from the expression: 
 
 " " K A I ( 9 ' 
 
 where T is the torque and fi is the angular velocity in rad/sec. Torque 
 may be calculated using the expression 
 
 T = a ( 10) 
 
 where a is the torsional spring constant and 6 is the deflection in 
 degrees. The constant K. can be calculated from the following expression: 
 
 k a = if* (h + ho) ( -p: ~ -p: ] (11) 
 
 i o 
 
 where R'. and R 1 are the effective diameters of the roughened cylinders. 
 
 Activated sludge has been shown to behave as a Bingham plastic, 
 that is, the rate of shear is directly proportional to the shearing 
 stress once the yield value has been exceeded (Dick, 1965) (Dick and 
 Ewing, 1967). Unlike Newtonian liquids, a substantial angular velocity is 
 required in order to obtain complete shear across the gap in the vis- 
 cometer. The yield strength may be calculated using this expression: 
 
 \ = K B e X 0=K B T X ( '2> 
 
 where t is the yield strength and 6 y is the intercept on the deflection 
 axis obtained by extrapolating the straight line portion of the consis- 
 tency curve. K D is an instrument constant. 
 
 D 
 
 (13) 
 
 V B ~ In R' /R'. 
 o | 
 
67 
 The plastic viscosity, the slope of the straight line portion of the con- 
 sistency curve, may be calculated as follows: 
 
 where 0, is in rads/sec. 
 
 Since the end effect was found to be dependent upon the viscos- 
 ity of the material being tested, the instrument constants K. and l< also 
 varied with the viscosity. Using the curve of Figure 1^ which shows the 
 relationship between viscosity and end effect, the instrument constants 
 were calculated using a trial and error procedure. A trial value of end 
 effect was selected and used to calculate a value of K.. The plastic vis- 
 cosity of a sample was calculated using that value of K . The plastic 
 viscosity obtained was applied to Figure 14 to see if the corresponding 
 value of end effect was the same as the selected one. The procedure was 
 repeated until agreement between the end effect values was achieved. When 
 agreement was reached, the K value for that end effect was used to com- 
 pute the value of K g . 
 
 The values of the spring constant and effective thickness of the 
 cylinder linings were 
 
 a = 17-^ dyne cm/degree 
 6 = 0.30 cm 
 Typical values of the instrument constants were 
 K A = 1.182 x 10 cm" 3 
 K B = 6.18 x }Q~ i * cm" 3 
 3. Experimental Procedure 
 
 Activated sludge of known concentration was introduced into the 
 viscometer by pouring it from the top into the annular space between the 
 
68 
 
 two cylinders. A volume of sludge sufficient to submerge the lower 6 in. 
 of the bob was used in all cases. The sludge was redistributed by 
 bubbling air through the hollow drive shaft and into the viscometer at 
 a rate which would generate the turbulence needed to provide good mixing. 
 A mixing period of about 10 sec was used. Dick (1965) determined that a 
 period of about ^5 sec was sufficient to permit ref locculat ion of the 
 sludge. A similar period was used here. Following the period of refloc- 
 culation the motor was started to rotate the outer cylinder. As the 
 inner cylinder began to rotate its motion was restricted manually in an 
 attempt to reduce the magnitude of the initial oscillation of the bob. 
 The deflection as indicated by the pointer after it became somewhat steady 
 was read and recorded, usually within the first 20 to 30 sec after the 
 start of rotation. The speed of rotation was calculated using the counter 
 and a stop watch. The sludge was redistributed again and the process was 
 repeated at another speed. This procedure was repeated at several speeds 
 until sufficient points were available so that the consistency curve could 
 be drawn. The same routine was followed using at least 2 other concentra- 
 tions of the same sludge. The data were plotted and the values for yield 
 strength and plastic viscosity were calculated using the appropriate for- 
 mulas presented in the preceeding section. Plots of typical data are 
 shown in Figure 17- 
 
 E. Removal of Extracellular Polysaccharide 
 
 The extracellular polysaccharides associated with bacteria are 
 present in the form of sheaths or capsules (Wilkinson, 1958). A number 
 of procedures have been used to remove these materials for study. Chemical 
 extractions have frequently been used in work with bacterial 
 
69 
 
 O IA O 
 
 CM «— — 
 
 wdj 'Q33dS IVNOIlVlOy 
 
70 
 
 polysaccharides. Phenol and water, trichloroacetic acid, sodium hydroxide 
 and acetic acid are some of the reagents commonly used in performing these 
 extractions (Luderitz et_ a_l_. , 1 966) . Recently Nishikawa and Kuriyama (1968) 
 and Takiguchi (1968) reported extracting a material which they called 
 mucilage from activated sludge using Na. C0_ and Na OH. The mucilage was 
 found to contain about 35 percent deoxyribonucleic acid with the balance 
 being mostly polysaccharide. 
 
 Methods employing shearing have been used in a number of studies 
 as a means of mechanically stripping the extracellular polysaccharides from 
 the cells. Juni and Heym (1964) removed capsular material by passing a 
 bacterial cell suspension through a glass chromatogram sprayer 3 or k times. 
 Gaudy and Wolfe (1962) recovered the slime from Sphaerotl lus natans by 
 blending a suspension of the washed cells in a laboratory blender and sep- 
 arated the cells from the slime by centr if ugat ion. Szaniszlo et al . , (1968) 
 used high speed blending in a Waring blender to remove the capsular poly- 
 saccharide from a marine filamentous fungi. Conrad (1969) described a 
 method to be used for the isolation of polysaccharides from gram-negative 
 cells. He, too, recommended the use of agitation at top speed in a Waring 
 blender as a means of stripping extracellular polysaccharides from cells 
 harvested while in the late log growth phase. Exposure to ultrasonic sound 
 is another method which has been used to physically shear this material from 
 the cell wall. Freidman et_ a_k (1968) used exposure to ultrasonic sound to 
 remove the matrix surrounding the cells in their pure culture of Zoo g lea 
 ramigera . Wyss et_ a_l_. (1968) used exposure to ultrasonic sound to strip cap- 
 sular material from hydrocarbon utilizing bacteria. In all of the above 
 cases, once the capsular material had been stripped from the cells it was 
 
/I 
 
 eparated from the cells and debris by centri f ugat ion at forces of from 
 1,000 to 20,000 g, usually in refrigerated centrifuges. The supernatant 
 :ontaining the polysaccharide was usually acidified and treated with cold 
 ilcohol or acetone to precipitate out polysaccharides. The precipitate 
 /as usually treated further depending upon the purpose of its recovery. 
 
 After some experimentation with the methods employing blending 
 ind exposure to ultrasonic sound, the blending method of stripping the 
 extracellular material was chosen as the method to be employed in this 
 ,tudy. Activated sludge washed twice with deionized water was resuspended 
 n deionized water. The washed sludge was blended at top speed for 5 min 
 n a Sorvall Omni-Mixer. The sample container was immersed in an ice 
 later bath to keep the temperature of the sample as low as possible. The 
 (lending time was selected after making microscopic examinations of sludges 
 /hich had been blended for various lengths of time. Even after 10 min of 
 )lending, the activated sludge floe, though smaller in diameter, were 
 ►till not completely dispersed. The blended sludge was centrifuged in a 
 ■efrigerated centrifuge at 10,000 rpm for 20 min, the supernatant was 
 iecanted and centrifuged again. Samples of the supernatant were taken for 
 :otal carbohydrate analysis using the anthrone method (Gaudy, 1962) 
 (Snell e^ a]_. , 1 96 1 ) . 
 
 : . Analytical Procedures 
 
 1. Chemical Oxygen Demand 
 
 The chemical oxygen demand (COD) tests performed on the feed, 
 
 Cat. No. 0M-115/X0, manufactured by Ivan Sorvall Inc., Norwalk, Connecticut 
 
72 
 
 the effluent, and the mixed liquor solids were carried out according to the 
 procedure described in the 12th edition of Standard Methods (1965)- Twenty 
 ml samples together with the indicated amounts of potassium dichromate, 
 mercuric sulfate, and sulfuric acid containing AgSO. were used in most cases. 
 For the determinations made on the mixed liquor solids, kO ml samples were 
 used together with the appropriate amounts of reagents. All analyses were 
 made in duplicate and were refluxed for at least 2 hrs. 
 
 2. Organic and Ammonia Nitrogen 
 
 Organic nitrogen analyses were made using the Kjeldahl nitrogen 
 procedure described in the 12th edition of Standard Methods (1965). The 
 sample sizes varied depending upon the expected nitrogen content of the 
 sample. The sizes of samples were selected so that, upon completion of the 
 distillation, the pH of the distillate-boric acid solution would not exceed 
 7- Analyses were made in triplicate using duplicate blanks. One departure 
 from the described method was the use of a pH meter rather than the mixed 
 indicator in the titration of the distillate. The distillate boric acid 
 solution for each sample was titrated back to the pH of the distillate-boric 
 acid solution of the blank using 0.02 N HCL. 
 
 The ammonia nitrogen determinations were made according to the 
 procedure given in Standard Methods (1965) with the exception that 0.02 N HCL 
 was used as the titrant. 
 
 3. Suspended Solids Determination 
 
 The suspended solids determinations were made using the Gooch 
 crucible-glass fiber filter method (Gratteau and Dick, 1968). This method 
 utilized a 2.1 cm diam glass fiber filter mat as the filter media rather 
 
 Glass Fiber Filter: Grade 93 1 tAH glass fiber filter 2.1 cm diam, a product 
 of Reeve Angel and Company, Clifton, New Jersey 
 
73 
 
 than the asbestos mat recommended in Standard Methods (1965)- Gratteau and 
 Dick (1968) have shown the Gooch crucible-glass fiber filter method to be 
 more accurate than, and of comparable precision to, the standard Gooch 
 crucible-asbestos mat and membrane filter techniques. The method for deter- 
 mining suspended solids concentrations using the glass fiber filter is essen- 
 tially the same as the standard method. The filter mat is placed in a clean 
 Gooch crucible, seated by filtering through a few ml of deionized water, dried 
 in an oven at 103 C for at least 1 hr, cooled in a desiccator for kO min and 
 tared. After the sample has been filtered with the aid of a vacuum, the 
 crucible is again cooled in a desiccator for kO min and weighed. The sus- 
 pended solids concentration is then calculated. Individual desiccators 
 consisting of wide mouth bottles partially filled with silica gel and having 
 a piece of plexiglass tubing to support the crucibles above the desiccant 
 were used for cooling and storing the crucibles prior to weighing. For mixed 
 liquor suspended solids determinations 10 ml sample sizes were used. The 
 time required for filtration varied depending upon the properties of the 
 sludge being filtered. For suspensions having lower concentrations of solids 
 than mixed liquors, the sample sizes could be increased several fold. 
 b . Sludge Volume Index 
 
 The sludge volume index (SVl) measurements were made as described 
 in Standard Methods (1965). The sludge volume index is defined as follows: 
 
 ci j w 1 1 j m ' settled sludge , ir x 
 
 Sludge Volume Index = -7— ; rr— (15) 
 
 3 g suspended matter 
 
 One liter cylinders were used in performing the test. For the f i lamentously 
 bulked sludges which settled slowly or not at all, the index was limited by 
 the suspended solids concentration. In some tests the mixed liquor was 
 
Ik 
 
 diluted with effluent from the flotation unit in order to obtain a sludge 
 volume index for a MLSS of 1,000 mg/1 . 
 5. Total Carbohydrate 
 
 The anthrone method was used to determine the carbohydrate content 
 of the mixed liquor and of the extracellular polysaccharide material stripped 
 from the cells (Gaudy, 1 962 ; Snell and Snell, 1961). The anthrone reagent 
 was made by dissolving 0.2 grams of anthrone in 100 ml of 95 percent H SO, 
 and storing in an ice water bath. This reagent was made fresh a few hours 
 before use. The tests were made using the following procedure: samples were 
 appropriately diluted to a volume of 2.5 ml in test tubes with deionized 
 water and were kept cold in an ice water bath. Five ml of cold anthrone 
 reagent were pipetted into each of the test tubes. The anthrone reagent was 
 quickly mixed with the sample in each tube immediately following the addition 
 of the reagent to that tube. This was accomplished using a button ended glass 
 rod or by vigorously striking the side of the tube with a finger several 
 times. The tubes were kept in an ice water bath until all samples had been 
 treated with the anthrone. The samples were then heated in a water bath at 
 100 C for exactly 15 min. The tubes were capped with glass marbles immed- 
 iately upon removal from the water bath and were placed in cold water (not 
 ice water) so as to cool to room temperature. The marbles sealed the tubes 
 preventing water or water vapor from entering the tube during cooling and 
 causing the sample to become milky or turbid. The absorbances of the samples 
 at 620 my were measured on a Beckman DU spectrophotometer using a one cm 
 light path and a slit width of 0.08 mm. A 100 mg/1 glucose solution preserved 
 with 0.6 ml of commerical "Roccal"" per liter of glucose solution was used to 
 
 Roccal (10 percent solution): A product of Winthrop Laboratories, New York. 
 Active Ingredients: alkyld: methyl benzyl ammonium chloride (C 1? , C., , C , 
 and other related alkyl groups from Co to C.g) 10 percent inert. 
 Ingredient: Water, 90 percent. 
 
75 
 
 make the standard curve. Since the anthrone reagent was stable for only a 
 few hours new standard curves were made each time a new volume of anthrone 
 was prepared. 
 
 6. Float Solids Concentration 
 
 The concentration of solids in the float collected from the batch 
 and continuous flotation cells were determined by placing a few ml of the 
 sample in a tared glass weighing bottle, capping the bottle to prevent 
 evaporation of water from the sample, and weighing. The cap was removed 
 from the bottle and the sample was dried at 103 C overnight. After removal 
 from the oven and cooling in a desiccator for one hr the sample was again 
 weighed. The float solids concentration was calculated from these data and 
 expressed as percent solids. The float solids determinations which were 
 made in connection with the solids flux analysis were performed by with- 
 drawing 50 ml of the concentrated sludge, diluting it 1:5 or 1:10 and then 
 proceeding as in making a suspended solids determination. 
 
 7. Carbon to Nitrogen Ratio of Sludge Solids 
 
 The C:N ratio of the microorganisms making up the activated sludge 
 was determined using a Hewlett Packard Model 185 C-H-N Analyzer. This instru- 
 ment was designed to provide a rapid semiautomatic means for measuring the 
 amounts of carbon, hydrogen, and nitrogen in organic materials. The instrument 
 accomplishes this by converting the nitrogen, carbon, and hydrogen in the 
 sample to N , CO , and HO in the presence of an oxidant at high temperature 
 (1,050 C) . The products of combustion are swept into a gas chromatographic 
 system by the flow of helium carrier gas. The gases are separated into dis- 
 crete bands in the column. Upon emerging from the column the component bands 
 enter a thermal conductivity detector which develops an electrical signal 
 
 Manufactured by Hewlett-Packard/ F & M Scientific Division, Avondale, Penn. 
 
76 
 
 proportional to the concentration of the component in the carrier gas. The 
 signal generated is delivered to a potent iometeric recorder where the signal 
 produces a chromatogram. The heights of the N. and C0 ? peaks are directly 
 proportional to the amounts of nitrogen and carbon in the sample while the 
 H_0 peak is related to the amount of hydrogen in the sample. Standard organic 
 compounds having compositions similar to the samples to be analyzed are used 
 to determine the relationships between the peak heights and the amounts of 
 each of the 3 elements in the compound. The instrument has a built-in ratio 
 recording feature which automatically attenuates the peak signals so that 
 they are directly proportional to the sample weight. In other words, the 
 signal is automatically divided by the sample weight so that the resulting 
 peak heights of the chromatogram are directly proportional to the sample 
 composition. 
 
 The instrument is capable of analyzing either liquid or solid 
 samples. The sludge samples analyzed here were prepared for analysis as dried 
 sludge solids. The solids in the mixed liquor from the aeration tank were 
 washed, centrifuged and were dried for several hours at 103 C. The dry 
 sludge was pulverized in a mortar and pestle to make a very fine powder. The 
 powder was stored in an air tight plexiglass vial. 
 
 The size of the sample which could be used for analysis was limited 
 to 0.5 mg - 0.7 mg. The sample was weighed in a tared platinum boat on the 
 instrument's electronic balance. Following the weighing, the boat was posi- 
 tioned in the end of the combustion rod, packed with catalyzed oxidant and 
 inserted into the combustion port of the instrument. A carrier gas bypass 
 cycle of 20 sec was used as the sample was inserted into the combustion tube. 
 The chromatogram was traced out on a Honeywell Electronik 16 Strip Chart 
 
11 
 
 Recorder using the automatic signal attenuation and ratio recording modes. 
 Acetanilide, Cyclohexanane 2, 4-di ni trophenyl hydrazone , and cystine were 
 used as standards. 
 
 8. Photomicrographs 
 
 Routine microscopic observations and photomicrographs were made 
 using a Carl Zeiss Standard RA microscope equipped with a phase contrast 
 condensor. A 35 mm Zeiss Ikon attachment camera was used with the micro- 
 scope for making photomicrographs. The ocular lens for the camera was a 
 standard 10X eyepiece. Kodak Plus-X film was used. Flash illumination of 
 30 watt sec of energy was provided by the camera's microflash device. Com- 
 binations of filters having various degrees of light transmission were used 
 to control the amount of light reaching the film during exposure. The film 
 was developed for 5-5 min at 68 F using a fine grained developer. 
 
 * 
 
 Manufactured by Honeywell Industrial Prod. Group, Philadelphia, Penn. 
 Manufactured by Carl Zeiss, Oberkochen, Wuertt. , West Germany 
 
IV. RESULTS AND DISCUSSION 
 
 A. Relationships Between Sludge Characteristics and Floatability 
 1. Shifts in Microbial Population 
 
 Initially it was planned that the characteristics of the activated 
 sludge developed in the laboratory continuous flow system would be controlled 
 by operating the system at steady state conditions. "Steady state" conditions 
 as used here denotes operation of the system at conditions of constant F/M 
 and BOD/N ratios. Shifts in sludge characteristics would be achieved when 
 desired by changing the BOD/N and F/M ratios. As the work progressed, changes 
 in the types of organisms predominating in the sludge occurred when shifts 
 were not desired as well as when encouraged by the intentional changing of 
 the loading and BOD/N of the substrate. The changes were determined by micro- 
 scopic observation. Cassell et^ aj_. (1966) in their study of population 
 dynamics and selection in continuously-fed mixed cultures observed shifts in 
 populations in laboratory systems which were being operated at steady state 
 conditions using skim milk as a substrate. This natural tendency toward 
 population shift even under steady state conditions together with the opera- 
 tional problems encountered in this study when trying to maintain steady 
 state conditions explain the population shifts which will be discussed in the 
 following paragraphs. 
 
 Sludges were grown under three combinations of F/M and BOD/N ratios 
 with the first being one where nitrogen was limiting, i.e. F/M = 0.5, BOD/N = 
 *»0. Since this was the first set of loading conditions, most of the problems 
 with control of the BOD/N ratio were experienced under these conditions. 
 Several shifts of population took place. The occurence of the shifts could 
 not be correlated directly with any operational changes or problems. 
 
 78 
 
79 
 
 A filamentous type of organism having the appearance of a string of 
 beads was the first to predominate the biomass at this loading. Typical fila- 
 ments of this type are shown in Figure 18. After predominating for about 
 three weeks this type of filament became less prevalent as a filamentous form 
 of another type became predominant. This organism had individual cells 
 arranged within the filament so that the long dimension of the cell was per- 
 pendicular to the axis of the filament. These filaments appeared to be 
 similar to the ones which Pipes (1968) called Caryophanon . Typical filaments 
 of this type are shown in Figure 19- It was interesting to note that often 
 these filaments had cells attached externally at almost right angles to the 
 long dimension of the filament. Filaments with these cells attached are 
 shown in Figure 20. As these filaments diminished, odd shaped floes were 
 often formed as zoogleal growth began to surround the old filaments. The 
 floes were often elongated with filaments serving as a sort of framework 
 around which the floes were organized. Such a floe is pictured in Figure 21. 
 This filamentous growth was replaced by a predominantly zoogleal sludge which 
 predominated for a little more than a week. The Caryophanon- 1 ? ke type of 
 sludge then returned and predominated until the loading rate was changed. 
 
 Following the shift of loading to F/M = 1 and BOD/N = 25 the 
 Caryophanon- 1 ike type of filaments were surrounded by zoogleal growth and 
 diminished in numbers. Accompanying the zoogleal growth were a significant 
 number of long, straight, stiff filaments which showed some branching. This 
 type of filament, whose presence may have been encouraged by a temporary 
 oxygen deficiency, decreased in numbers over a period of 10 days or more. 
 Filaments of this type are shown in Figure 22. The zoogleal growth continued 
 for a brief period before giving way to the string-of-beads type of filament 
 which predominated until the loadings were changed. These filaments appeared 
 
80 
 
 FIGURE 18. STRING-OF-BEADS TYPE OF FILAMENTS (x2160) 
 
 FIGURE 19. CARYOPHANON-LIKE FILAMENTS (x2160) 
 
81 
 
 FIGURE 20. CARYOPHANON-LIKE FILAMENT WITH CELLS ATTACHED 
 
 PERPENDICULARLY TO THE AXIS OF THE FILAMENT (x860) 
 
 FIGURE 21. ELONGATED FLOC (x860) 
 
82 
 
 FIGURE 22. STIFF FILAMENT (x680) 
 
83 
 
 to be the same or similar to the one which had been present in the system 
 earlier under the first set of loading conditions. 
 
 The next combination was F/M = 0.5 and BOD/N ■ 10. With this com- 
 bination nitrification with its accompanying decrease in the pH was a 
 continuing problem. The quality of the sludge became poorer from a solids 
 separation standpoint as the floe diameter decreased until floe practically 
 ceased to exist. With the use of Na.CO it was possible to maintain the pH 
 near 7.0 and finally the system was restored to a more normal condition. A 
 zoogleal type of sludge formed first and then was gradually replaced by the 
 str ing-of-beads type of filament. Generally speaking, the operation of the 
 continuous flow unit under these conditions was difficult, because of the 
 poor separation characteristics of the sludge. 
 
 In addition to the sludges which were clearly filamentous or 
 zoogleal in nature there were many sludges which possessed various combina- 
 tions of zoogleal and filamentous growths. Interestingly enough, on some 
 occasions when it might seem that the sludge was considerably more zoogleal 
 than filamentous on a mass basis, the presence of a few filaments would cause 
 the sludge to exhibit properties more befitting a filamentous sludge. 
 
 The previous discussion has centered primarily around the types of 
 organisms which made up the sludge floe. The presence of free swimming 
 protozoans, stalked ciliates, and rotifers in the mixed liquor at all loadings 
 should be mentioned. The numbers of these higher forms which were present 
 seemed to vary widely with time. At times under a given loading they would 
 be present in large number, yet within the space of a few days they would have 
 almost completely disappeared. 
 
 While the seemingly continual shifting of the microbial population 
 was not always predictable nor desired, it did provide an opportunity to 
 
Sk 
 
 study the flotation of sludges of several different morphological types. 
 
 2. Effects of Percent Nitrogen and Carbon to Nitrogen Ratio of the 
 Sludge Upon Flotation 
 
 Sludges possessing differing characteristics were obtained by 
 changing the F/M and BOD/N ratios of the substrate as described previously. 
 The following properties of the sludges were investigated with respect to 
 their influence upon the flotation process: the amount of strippable extra- 
 cellular polysaccharides, percentage of nitrogen in the sludge, the carbon to 
 nitrogen ratio of the sludge, the morphology of the sludge, and the Theolo- 
 gical properties, plastic viscosity and yield strength. The concentration of 
 solids in the float and the sludge interfacial rise rate were used to 
 evaluate the effects of these properties upon sludge flotation behavior. 
 The batch test procedure described previously was used to perform these tests. 
 
 The first characteristic to be discussed is the percentage of 
 nitrogen in the sludge. The amount of nitrogen in the sludge was determined 
 both by the C-H-N analyzer and by the Kjeldahl method for organic nitrogen. 
 Comparisons of the results obtained using the two methods indicate that the 
 C-H-N analyzer gives values that are about 10 percent higher than the other 
 method for sludges which were grown under nitrogen deficient conditions. The 
 data presented here were obtained using the Kjeldahl method. 
 
 The data relating the percentage of nitrogen in the sludge to the 
 solids concentration in the float after flotation had proceeded for 20 min 
 are shown in Figure 23. These data indicate that in general higher float 
 concentrations can be expected with sludges containing more than 8 to 8.5 
 percent nitrogen. Lower nitrogen concentrations resulted in lower float 
 solids concentrations. The sludges which possessed higher concentrations of 
 nitrogen were more likely to be zoogleal while those of low nitrogen content 
 were more filamentous. There were, as is clearly shown in the figure, some 
 
85 
 
 2.4 
 
 2.2 — 
 
 2.0 
 
 in 
 
 *o 
 
 Z 1.8 
 
 o 
 
 c 
 
 I 1.6 
 
 1.4 
 
 1.0 
 
 0.8 
 
 V /v 
 w s 
 
 C 
 
 s 
 
 - 0.8 
 
 - 4000 mg/1 
 
 C 
 
 m 
 
 - 2220 mg/1 
 
 Time 
 
 - 20 mln 
 
 Q. 
 
 I I I L 
 
 7 8 9 10 
 NITROGEN CONTENT, percent 
 
 II 
 
 FIGURE 23. RELATIONSHIP BETWEEN NITROGEN CONTENT OF THE 
 SLUDGE AND FLOAT CONCENTRATION 
 
86 
 
 notable exceptions to these general statements. The data points plotted as 
 showing a float concentration of about one percent and a nitrogen content of 
 about 10 percent were obtained from a sludge which had been zoogleal but 
 contained some filaments of the type shown previously in Figure 22. The 
 behavior of this sludge in the flotation test was apparently dictated by 
 these stiff filaments even though the sludge had a high nitrogen content 
 indicative of a zoogleal type of growth. The next point above that one was 
 obtained three days later after the stiff filaments had essentially dis- 
 appeared and filaments of the types shown in Figures 18 and 19 were 
 increasing in numbers. The point plotted as showing a sludge containing 
 7-35 percent nitrogen and yielding a float concentration of about 1.95 per- 
 cent was obtained from a sludge which had been predominantly filamentous but 
 was in a transition stage and becoming more zoogleal. In this case the 
 characteristics of the zoogleal sludge apparently influenced the flotation 
 behavior despite the relatively low nitrogen content. 
 
 The results obtained when the amount of nitrogen in the sludge is 
 plotted against the rise rate are shown in Figure 24. Since a relationship 
 exists between the interfacial rise rate and the solids concentration of the 
 float the shape of this plot is somewhat similar to the previous one. The 
 data points are quite scattered. As can be seen, there is not a good rela- 
 tionship between the rise rate and nitrogen content of the sludge. 
 
 The C/N ratios of the sludges were obtained using the data collected 
 with the C-H-N analyzer. Since the values for the percentage of nitrogen in 
 the sludge obtained with the analyzer differed considerably from the Kjeldahl 
 data, it was possible to have two C/N ratios depending upon which value for 
 the percentage of nitrogen was used. The data used here, however, are those 
 
87 
 
 6 - 
 
 5 - 
 
 I 3 
 
 V /v 
 w s 
 
 T" 
 
 0.8 
 
 4000 mg/1 
 
 2220 mg/1 
 
 7 8 9 
 NITROGEN CONTENT, percent 
 
 10 
 
 FIGURE zk. RELATIONSHIP BETWEEN NITROGEN CONTENT OF 
 SLUDGE AND RISE RATE 
 
88 
 
 obtained with the analyzer. The values obtained for a number of different 
 sludges are shown in Table 4. 
 
 TABLE 4 
 C-H-N DATA FOR SEVERAL ACTIVATED SLUDGES 
 
 Date 
 
 %Z 
 
 %W 
 
 %H 
 
 C/N 
 
 
 F/M = 0.5 
 
 BOD/N 
 
 = 40 
 
 
 7-20-69 
 7-29-69 
 8-11-69 
 8-17-69 
 
 42.5 
 41.0 
 39-5 
 40.5 
 
 6.5 
 8.3 
 6.7 
 5.31 
 
 7.6 
 7.1 
 6.75 
 6.9 
 
 6.5 
 5.0 
 5.9 
 7.6 
 
 
 F/M = 1 
 
 BOD/N = 
 
 25 
 
 
 8-29-69 
 9- 4-69 
 9-15-69 
 9-22-69 
 9-19-69 
 
 43.9 
 46.8 
 44.5 
 43.2 
 43.0 
 
 9.25 
 
 9.74 
 
 9-5 
 
 7.72 
 
 8.8 
 
 7.42 
 7.54 
 7.53 
 7.38 
 7.52 
 
 4.75 
 4.81 
 4.68 
 5.58 
 4.89 
 
 
 F/M =0.5 
 
 BOD/N 
 
 = 10 
 
 
 11-11-69 
 11-18-69 
 11-26-69 
 
 46.1 
 45.9 
 43.0 
 
 9.6 
 11.4 
 10.6 
 
 7.0 
 7.2 
 6.8 
 
 4.79 
 4.02 
 4.06 
 
 The C/N ratios of several of these sludges are plotted against their 
 
 respective float concentrations in Figure 25. The data indicate that at best 
 
 there is only a general trend toward higher float concentrations as the C/N 
 
 ratio decreases. As with the nitrogen content data, the scatter in the data 
 
 make the C/N ratio an unreliable indicator of float concentrations. A plot 
 
 of the C/N ratio of the sludges against their rise rates. Figure 26, reveals 
 
 that a good correlation does not exist. 
 
 3. Relationship Between the Presence of Extracellular Polysaccharide 
 and Flotation Behavior 
 
 Extracellular polysaccharide was stripped from the activated sludge 
 floe by the blending procedure described in the previous chapter. Even the 
 
89 
 
 2.4 
 
 2.2 
 
 •D 
 
 2.0 
 
 
 
 
 
 
 
 
 
 v> 
 
 
 
 *J 
 
 
 
 c 
 
 
 
 w . 
 
 
 
 u 
 
 L. 
 
 1.8 
 
 
 V 
 
 
 
 a. 
 
 
 
 * 
 
 
 
 z 
 
 
 
 o 
 
 
 
 5 
 
 1.6 
 
 
 i- 
 
 
 
 z 
 
 
 
 LU 
 
 
 
 O 
 
 
 
 z 
 
 
 
 o 
 
 
 
 <_•> 
 
 
 
 h- 
 
 1.4 
 
 
 < 
 O 
 
 
 
 1.2 
 
 1.0 
 
 0.8 
 
 J L 
 
 n r 
 
 - 0.8 
 C g - 4000 mg/l 
 C m - 2220 mg/l 
 Time - 20 mln 
 
 V /V 
 w s 
 
 14 5 6 7 8 
 
 CARBON/NITROGEN 
 
 FIGURE 25. RELATIONSHIP BETWEEN CARBON TO NITROGEN 
 RATIO AND FLOAT CONCENTRATION 
 
c 
 
 i 4 
 
 2 3 
 
 tn 
 
 1 — r 
 
 i r 
 
 V /v 
 w s 
 
 - 0.8 
 
 - 4000 mg/1 
 
 - 2200 mg/1 
 
 * 5 6 7 
 
 CARBON/NITROGEN 
 
 yo 
 
 FIGURE 26. RELATIONSHIP BETWEEN CARBON TO 
 
 NITROGEN RATIO OF THE SLUDGE AND 
 RISE RATE 
 
91 
 
 shearing of this blending operation was not sufficient to completely disperse 
 the sludge floe. In view of this, the amount of polysaccharide recovered was 
 probably something less than the total amount present extracel 1 ula rl y. The 
 results obtained have been plotted against the appropriate parameters to 
 determine their influence upon the flotation process. The data plotted in 
 Figure 27 indicated that as the nitrogen content of the sludge decreased, 
 the amount of extracellular polysaccharide seemed to increase. The extra- 
 cellular polysaccharide data are plotted against the float concentrations and 
 rise rates obtained as shown in Figures 28 and 29. In either case there does 
 not seem to be a correlation between the results of the batch flotation tests 
 and the amounts of extracellular polysaccharides stripped. 
 k. Rheology of the Activated Sludges 
 
 The rheological behavior of a suspension is dependent upon the 
 characteristics of the materials in suspension as well as those of the sus- 
 pending liquid. Activated sludges behave as Bingham plastics, that is, they 
 possess yield strengths and have shearing stresses which are directly propor- 
 tional to the shear rates once the yield stresses have been exceeded (Dick 
 and Ewing, 1968). The physical characteristics of the microorganisms which 
 make up the sludge are a. major factor in determining the yield strength and 
 plastic viscosity of a sludge. In this work the rheological properties of 
 the sludges were used as a means of quantitatively characterizing the physical 
 properties of the sludges. 
 
 The yield strength data for different types of sludges grown under 
 various conditions in the laboratory have been plotted in Figure 30. As can 
 be seen from this plot, the yield strength is quite dependent on sludge con- 
 centration. This type of data usually plot as straight or slightly curved 
 lines on semilog paper in the 3,000 to 6,000 mg/1 range of solids concentra- 
 tions. As illustrated by curve B, the plot becomes slightly curved over a 
 
V 
 
 Ul V 
 
 o en 
 
 — "V 
 
 CC 3 
 
 o 
 
 O CD 
 
 < v, 
 
 to • 
 
 >- > 
 
 -J — 
 
 O 3 
 
 o. cr 
 
 < v 
 
 —I «A 
 
 =3 o 
 
 -J o 
 
 -I 3 
 
 Ul — 
 
 O CD 
 
 6 7 8 9 10 
 NITROGEN CONTENT OF SLUDGE, percent 
 
 11 
 
 FIGURE 27. RELATIONSHIP BETWEEN THE NITROGEN CONTENT 
 
 OF THE SLUDGE AND EXTRACELLULAR POLYSACCHARIDE 
 
93 
 
 3.0 
 
 2.5 - 
 
 2.0 
 
 .5 - 
 
 .0 
 
 0.5 
 
 - 
 
 1 
 
 
 
 1 1 1 
 
 C - 4000 mg/1 
 
 C - 2220 mg/1 
 
 m 
 
 Time ■ 20 mln 
 
 - 
 
 
 o 
 
 
 
 - 
 
 
 
 
 
 — 
 
 - 
 
 
 
 o o 
 
 - 
 
 — 
 
 1 
 
 1 1 1 
 
 - 
 
 5 10 15 20 25 
 
 EXTRACELLULAR POLYSACCHARIDE, mg glucose equlv/g sludge 
 
 FIGURE 28. RELATIONSHIP BETWEEN EXTRACELLULAR POLYSACCHARIDE 
 AND FLOAT CONCENTRATION 
 
9A 
 
 7 - 
 
 6 - 
 
 i 3 
 
 _ 
 
 1 
 
 1 1 1 
 
 V /V - 0.8 
 w s 
 
 C - i»000 mg/1 
 
 C - 2220 mg/1 
 
 m ^ 
 
 _ 
 
 - 
 
 
 
 
 - 
 
 - 
 
 
 O 
 
 - 
 
 - 
 
 
 
 ° o 
 
 - 
 
 
 o 
 
 
 
 
 1 
 
 1 1 1 
 
 
 5 10 15 20 25 
 EXTRACELLULAR POLYSACCHARIDE, mg glucose equlv/g sludge 
 
 FIGURE 29. RELATIONSHIP BETWEEN EXTRACELLULAR 
 POLYSACCHARIDE AND RISE RATE 
 
% 
 
 O Flexible Filaments 
 
 A Stiff Fi laments 
 □ Zoogleal Sludge 
 
 Stiff Filaments 
 
 ) 2000 i»000 6000 8000 10000 
 
 SUSPENDED SOLIDS, mg/1 
 
 FIGURE 30. TYPICAL YIELD STRENGTH VALUES FOR ACTIVATED SLUDGE 
 
 12000 
 
96 
 
 wider range of suspended solids concentrations. Sludge A was grown at F/M = 
 0.5,B0D/N = kO and was made up almost entirely of the type of organism being 
 referred to here as resembling Caryophanon . A photomicrograph of this type 
 of floe is shown in Figure 31- Sludge C was comprised primarily of those 
 filaments similar in appearance to a string of beads as shown in Figure 32. 
 The third, sludge D, was one which was essentially zoogleal and similar to 
 that shown in Figure 33- It is interesting to note that the yield strengths 
 of sludges A and D at concentrations of 4,000 mg/1 differ by an order of 
 magnitude. Between these two extremes, sludges possessing varying degrees 
 of f i lamentousness provide yield strengths of intermediate magnitudes. The 
 type as well as the quantity of filaments present appear to be reflected in 
 the yield strength of the sludge. Microscopic observations revealed a pos- 
 sible reason why the sludges having filaments of the type represented in 
 sludge C had yield strengths about half as great as those of the Caryophanon - 
 like sludge. The string-of-beads type of filaments were much less rigid as 
 was noted by its folding and bending when struck by floating debris while 
 being observed on a microscope slide. The larger filaments of the Caryophanon - 
 like sludge and the straight, stiff, filaments seen in other sludges possessed 
 more rigidity and, therefore, contributed to higher yield strengths. The 
 influence of the nature of the filaments on the rheological properties is 
 dramatically illustrated by the ability of the presence of a relatively small 
 number of filaments to give a predominantly zoogleal sludge a yield strength 
 equal to that of the essentially filamentous sludge of type A. 
 
 The plastic viscosity data for the sludges described previously are 
 plotted Figure 3^- These data show that plastic viscosity was also dependent 
 upon solids concentration and that sludges having high yield strengths also 
 exhibited high plastic viscosities. For low solids concentrations, and over 
 
97 
 
 FIGURE 31. FLOC CONTAINING LARGE FILAMENTS 
 RESEMBLING CARYOPHANON (xlkO) 
 
 m</ 
 
 FIGURE 32. FLOC CONTAINING FLEXIBLE FILAMENTS (x3^0) 
 
98 
 
 FIGURE 33- NORMAL ZOOGLEAL FLOC (x860) 
 
99 
 
 o 
 
 
 
 o 
 
 
 
 o 
 
 
 
 r^. 
 
 
 to 
 
 LU 
 C3 
 
 Q 
 
 _l 
 to 
 
 o 
 
 •— 
 
 
 o 
 
 ^ 
 
 O 
 
 o 
 
 o> 
 
 LU 
 
 v£ 
 
 E 
 
 1- 
 
 
 
 > 
 
 
 to 
 
 
 
 Q 
 
 h- 
 
 
 
 o 
 
 O 
 
 _l 
 
 < 
 
 o 
 
 O 
 
 
 o 
 
 to 
 
 cc 
 
 LA 
 
 
 o 
 
 
 o 
 
 u_ 
 
 
 LU 
 
 
 
 O 
 
 to 
 
 
 z 
 
 1- 
 
 
 LU 
 
 o 
 
 o 
 
 o. 
 
 _J 
 
 o 
 
 to 
 
 CL 
 
 o 
 
 3 
 
 
 -a- 
 
 to 
 
 >- 
 1- 
 
 to 
 o 
 o 
 to 
 
 o 
 
 
 — 
 
 o 
 
 
 > 
 
 o 
 
 
 
 Ci 
 
 
 o 
 
 \- 
 to 
 
 < 
 
 -J 
 
 o 
 
 
 Ol 
 
 o 
 
 
 
 o 
 
 
 —1 
 
 <M 
 
 
 < 
 
 o 
 
 Q. 
 
 >- 
 
 O 
 
 
 
 O 
 
 
 
 o 
 
 
 • 
 
 lud bs/oas auAp ' Ql * Aiisodsia ojiseid 
 
100 
 
 the relatively short range of concentrations investigated, 2,500 to 5,500 
 mg/1 , the relationship between solids concentration and plastic viscosity may 
 fit Einstein's equation which relates the viscosity of a suspension to solids 
 concentration. His equation is as follows: 
 
 u s = p 1 ( i + 03 c v ) ( 16 ) 
 
 where 
 
 y = viscosity of the suspension 
 
 u. = viscosity of the liquid 
 
 jy] = intrinsic viscosity 
 
 C = volumetric solids concentration 
 v 
 
 The volumetric solids concentrations of the sludges used in this work were 
 
 not determined. From Curve B it can be seen that at higher concentrations 
 
 the relationship between solids concentration and plastic viscosity does not 
 
 follow Einstein's equation. 
 
 5. Relationship of Rheological Properties to the Morphology and 
 Nitrogen Content of the Sludge 
 
 The rheological properties of sludge are influenced to a large degree 
 
 by the physical nature of the microorganisms which make up the sludge. The 
 
 information obtained from the literature indicated that generally filamentous 
 
 organisms, particularly the fungi, possess lower nitrogen contents than do 
 
 zoogleal sludges. Therefore, it seemed important to determine whether or not 
 
 a relationship existed between the nitrogen content of the sludges and their 
 
 rheological properties. It had been hypothesized that the C/N ratio of the 
 
 sludge might be a means of relating other sludge characteristics to the 
 
 flotability of the sludge. 
 
101 
 
 The sludges used in this study had nitrogen contents ranging from 
 about 6 to 10 percent. The sludges which were almost totally made up of the 
 Caryophanon- 1 i ke organisms had nitrogen contents ranging between 6 and 7-5 
 percent. The flexible filamentous and zoogleal sludges had nitrogen contents 
 ranging from 8 to 10 percent with the flexible filamentous sludges generally 
 being lower in nitrogen than the zoogleal sludges. One sludge which was 
 primarily zoogleal but contained some straight stiff filaments had a nitrogen 
 content of about 10 percent. 
 
 The yield strengths of several sludges are plotted against the per- 
 centage of nitrogen and the C/N ratios of the sludges in Figures 35 and 36. 
 These data indicate that at yield strengths of 1.5 dynes/sq cm or less, 
 yield strength generally increased as the nitrogen content increased. Sludges 
 in this range of yield strengths were of the zoogleal and flexible filamentous 
 types of growth. At the higher yield strengths the data were widely scat- 
 tered. The sludges having nitrogen contents of 6.1 and ~l .k percent were 
 comprised almost entirely of the Caryophanon- 1 ike filament. The point repre- 
 senting a sludge containing 10 percent nitrogen has been discussed previously. 
 This was the sludge comprised mainly of zoogleal organisms but containing 
 some long stiff filaments which are believed to have been responsible for the 
 high yield strength of that sludge. 
 
 From the data presented it was concluded that at the lower yield 
 strength values the yield strengths of the sludges are related to the nitrogen 
 content and C/N ratio of the sludge. With the sludges which possess the 
 higher yield strengths there is no apparent relationship between percent 
 nitrogen, C/N and yield strength. Consequently, neither nitrogen content nor 
 C/N by themselves can be considered reliable indicators of yield strength. 
 
102 
 
 3.0 — 
 
 2.5 
 
 2.0 — 
 
 1.5 ~ 
 
 1.0 
 
 
 i r 
 
 \ 
 
 I 
 
 
 I 
 
 
 
 I 
 
 
 
 
 
 
 1 
 
 1 
 
 I 
 
 
 
 
 
 • 
 
 \ 
 
 1 
 
 
 
 • 
 
 
 
 1 
 
 \ 
 \ 
 
 \ 
 
 
 
 
 
 — 
 
 \ 
 \ 
 
 \ 
 
 
 
 
 _ 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 — 
 
 \ 
 
 \ 
 
 • 
 
 
 
 - 
 
 — 
 
 
 Vi 
 
 
 
 - 
 
 — 
 
 1 1 
 
 i 
 
 r^v 
 
 1 
 
 - 
 
 '» 7 8 9 10 
 NITROGEN CONTENT, percent 
 
 FIGURE 35. RELATIONSHIP BETWEEN THE NITROGEN CONTENT OF 
 THE SLUDGE AND YIELD STRENGTH 
 
 11 
 
103 
 
 5.5 
 CARBON/NITROGEN 
 
 FIGURE 36. 
 
 RELATIONSHIP BETWEEN THE CARBON/NITROGEN RATIO OF 
 THE SLUDGE AND YIELD STRENGTH 
 
1 04 
 
 Plots showing the plastic viscosity data plotted against the 
 nitrogen content and carbon to nitrogen ratios of the sludges are presented 
 in Figure 37 and 38. The plastic viscosity varies linearly with the nitrogen 
 content of the sludge while the C/N data plot as a slightly curved line. 
 The fit of the C/N data is not as good, however, as for the nitrogen data. 
 These curves also reflect the morphological nature of the sludges. The 
 sludges made up largely of the Caryophanon- 1 ike filaments contained the least 
 nitrogen, the sludges having the string-of-beads type of filament had a 
 higher nitrogen content and the non-filamentous sludges contained the most 
 nitrogen. The percent nitrogen vs plastic viscosity curve slopes down to the 
 right from the very filamentous sludges to the zoogleal sludges. The C/N 
 ratio vs plastic viscosity curve slopes up to the right from the zoogleal to 
 the filamentous sludges. 
 
 While there does not appear to be a good relationship between the 
 nitrogen content of the sludge and the yield strength, there appears to be a 
 good relationship between the nitrogen content and the plastic viscosity. 
 
 6. Relationship Between Rheological Properties and Flotation Behavior 
 
 The rheological properties of the sludges were found to differ 
 widely during the course of this work. Batch flotation test data have also 
 given various results depending upon the nature of the sludges. The relation- 
 ship between the sludge yield strength and the float solids concentration 
 obtained from batch flotation tests is shown in Figure 39- The yield strengths 
 have been normalized to a suspended solids concentration of 4,000 mg/1 . This 
 concentration was chosen arbitrarily as the concentration to be used for 
 comparing the yield strengths of the various sludges. The batch tests were 
 performed in 2.5 in. columns according to the procedure outlined earlier. Tap 
 water was used as recycle water in these tests. While it has been established 
 
105 
 
 7 8 9 10 
 NITROGEN CONTENT, percent 
 
 11 
 
 FIGURE 37. RELATIONSHIP BETWEEN THE NITROGEN CONTENT OF 
 THE SLUDGE AND PLASTIC VISCOSITY 
 
106 
 
 2 3 *» 5 6 7 
 CARBON/NITROGEN 
 
 FIGURE 38. RELATIONSHIP BETWEEN PLASTIC VISCOSITY AND 
 THE CARBON TO NITROGEN RATIO OF THE SLUDGE 
 
107 
 
 Z.k 
 
 2.2 
 
 2.0 
 
 1.8 
 
 1.6 
 
 2 
 
 ^ 1.2 
 
 1.0 
 
 0.8 
 
 Zoog 1 ea 
 Sludges 
 
 Flexible 
 Fl laments 
 
 V /V - 0.8 
 w s 
 
 C $ - i*000 mg/1 
 
 C - 2220 mg/1 
 
 m J 
 
 t at C - 1*000 mg/1 
 
 Time - 20 min 
 
 Stiff Filaments 
 
 1 L 
 
 1 2 3 
 
 YIELD STRENGTH, dynes/sq cm 
 
 FIGURE 39. RELATIONSHIP BETWEEN SLUDGE YIELD 
 STRENGTH AND FLOAT CONCENTRATION 
 FOR BATCH FLOTATION TESTS 
 
108 
 
 that wall effects are factors to be considered, especially in small columns, 
 with filamentous sludges, checks made using a larger column indicated that 
 apparently at the relatively low solids concentrations in the flotation tubes, 
 the results were not appreciably influenced by wall effects. The data pre- 
 sented in Figure 39 are considered to be a reliable indicator of the effect 
 of yield strength on the float concentration. 
 
 The data show that as the yield strength of the sludge increased 
 the float concentration which was obtained in the batch flotation test 
 decreased. This trend was also observed in the continuous laboratory flota- 
 tion unit. The curve in Figure 39 can be divided into three different 
 sections based upon the types of organisms which determined the yield strengths 
 
 of the sludges. The first section including the range of yield strengths from 
 
 -2 
 about to 0.35 x 10 dynes/sq cm represents the zoogleal sludges. The sec- 
 
 -2 -2 
 
 tion spanning from about 0.36 x 10 to 2.0 x 10 dynes/sq cm represents the 
 
 partially zoogleal sludges and the filamentous growths where the filaments are 
 somewhat flexible, such as the string-of-beads type, for example. The section 
 from about 2 x 10 dynes/sq cm on up is for those sludges comprised of stiff 
 filaments such as the Caryophanon- 1 ike sludges and the straight smooth rigid 
 filaments mentioned earlier. It is indeed interesting to note that the sludge 
 having the long, stiff filaments, which has been somewhat of a "spoiler" in 
 almost every other plot where it has been included, plots very nicely here 
 along with the other high yield strength sludges. Apparently the presence of 
 relatively few of this type of filament markedly influenced the rheological 
 properties and flotation behavior of the sludge. 
 
 When the plastic viscosity data were plotted against the float con- 
 centration data, a curve having the same general shape as the yield strength 
 vs float concentration plot resulted. These data are shown in Figure 40. 
 The reason for the similarity in shapes stems from the fact that, in general, 
 
109 
 
 l.k 
 
 2.2 — 
 
 2.0 — 
 
 1.8 — 
 
 1.6 
 
 ].k 
 
 .2 
 
 1.0 — 
 
 ■■ r i i r ■ 
 
 i ' 
 
 1 1 ! | 1 1 1 I | 1 1 1 
 
 V /V « 0.8 
 w s 
 
 ~r— 
 
 I 
 
 
 C s - *»000 mg/1 
 
 - 
 
 • 
 
 
 C - 2220 mg/1 
 
 m 
 
 
 — 
 
 
 n at C - *»000 mg/1 
 
 — 
 
 - 
 
 
 Time - 20 min 
 
 - 
 
 _ 
 
 • 
 
 
 _ 
 
 - 
 
 
 Zoogleal and Flexible 
 
 - 
 
 Filamentous Sludges 
 
 : 
 
 !• 
 
 Stiff Filamentous 
 Sludges 
 
 : 
 
 • 
 
 \# 
 
 
 - 
 
 
 • > 
 
 
 
 i i i i 
 
 1 I 
 
 • ' 11 
 
 ™JL_ 
 
 1 1 1 1 1 1 1 1 — 1 — 1 — L 
 
 0.8 
 
 5 10 15 20 
 
 PLASTIC VISCOSITY X10 2 , dyne sec/sq cm 
 
 FIGURE 40. RELATIONSHIP BETWEEN PLASTIC VISCOSITY 
 AND FLOAT CONCENTRATION 
 
110 
 
 sludges having high yield strengths have higher plastic viscosities than the 
 
 lower yield strength sludges. The plot may be divided into two sections. The 
 
 -2 
 section from to about 7 x 10 dyne sec/sq cm includes both the zoogleal 
 
 -2 
 and flexible filamentous sludges. The section from 7 x 10 dyne sec/sq cm 
 
 on up represents the more rigid filamentous sludges. 
 
 The effects of the rheological properties of the sludges upon the 
 batch flotation rise rates are shown in Figures k] and kl. The rise rates 
 were inversely related to the yield strengths of the sludges. A fairly good 
 curve can be drawn through the points representing sludges having yield 
 strengths less than 0.5 dynes/sq cm. Figure *t1 can be divided into three 
 sections, to 0.25, about 0.26 to 0.40, and about 0.*t1 to 1.0 dyne/sq cm for 
 the zoogleal, flexible filament and stiff filament containing sludges res- 
 pectively. The curve resulting from the plotting of rise rate vs plastic 
 viscosity had a similar shape but showed more scatter in the data points. 
 
 It should be pointed out that the rheological properties discussed 
 above have been made using activated sludges without the presence of the air 
 bubbles which are attached to the sludge particles during the flotation test. 
 Therefore, the rheological measurements made were intended only to provide a 
 quantitative means of characterizing the sludge. 
 
 Attempts were made to measure the rheological properties of the 
 float using the rotational viscometer. A problem encountered here made reli- 
 able determinations impossible. As the sludge was transfered to the visco- 
 meter, some of the entrained air was stripped out. While being sheared in 
 the viscometer the sludge was continuously having air stripped from it which 
 resulted in decreasing values of torque as the shearing continued. The data 
 which were obtained indicated that float with the air entrained may have a 
 somewhat higher plastic viscosity than the same float with the air stripped 
 
Ill 
 
 7 — 
 
 6 — 
 
 5 — 
 
 S k 
 
 E 
 
 Z 3 
 
 1 — "T 1 1 r— i 1— n 1 
 
 V /V - 0.8 
 w s 
 
 C g - J»000 mg/1 
 
 1 C 2220 mg/1 " 
 1 
 
 olo 
 
 t at C - 2220 mg/1 
 y m 
 
 
 n Zoogleal 
 Sludges 
 
 — \ 
 
 
 - 
 
 
 
 ^Flexible 
 Filaments 
 
 "~"~* \ 
 
 ~ stiff ^ ; 
 
 
 Filaments 
 
 — *\ oo- 
 
 \° 
 
 \ 
 \. 
 
 1 1 1 1 1 1 I 1 1 
 
 0.2 0.4 0.6 0.8 
 YIELD STRENGTH, dyne/sq cm 
 
 1.0 
 
 FIGURE 41. RELATIONSHIP BETWEEN YIELD STRENGTH 
 AND RISE RATE 
 
112 
 
 I I 
 
 I I 
 
 V V s 
 
 - 0.8 
 
 
 C s 
 
 - '*000 
 
 mg/1 
 
 C 
 
 m 
 
 - 2220 
 
 mg/1 
 
 at C 
 
 - 2220 
 
 mg/1 
 
 1 1 I I I I 
 
 3 k 
 
 7 8 
 
 10 
 
 PLASTIC VISCOSITY X10 , dyne-sec/sq cm 
 
 FIGURE l|2. RELATIONSHIP BETWEEN RISE RATE AND 
 PLASTIC VISCOSITY 
 
113 
 
 out of it. This would be expected since the air bubbles adsorbed to the 
 sludge floe result in agglomerates which have properties somewhat different 
 than those of the sludge floe alone. The volumetric concentration of the 
 sludge floe would be increased and , although Einstein's law does not hold 
 for the sludges at high concentrations, it does indicate that plastic visco- 
 sity would increase as the volumetric solids concentration increased. The 
 yield strengths were found to be comparable. 
 
 It should be pointed out at this point that in plotting the float 
 concentration and rise rates as functions of a single parameter such as yield 
 strength or plastic viscosity it was tacitly assumed that the other parameter 
 remained constant. Sufficient data is not available to permit the construc- 
 tion of families of curves such as curves for yield strength vs float con- 
 centration for constant plastic viscosity or vica versa. 
 7- Continuous Flotation Unit Performance 
 
 The continuous laboratory unit was used in the experimental work 
 primarily as a means of separating the activated sludge solids so that they 
 could be recycled to the aeration tank in order to maintain a continuous 
 process. Data were not routinely taken from it as a part of the main experi- 
 mental work. Some float concentration data were collected, however. There 
 were a number of operational variables which influenced the performance of 
 the continuous flotation unit, the effects of which can be noted in the con- 
 centration of float which was obtained. Among these variables were the air 
 to solids ratio, the recycle rate, the solid and liquid loading rates applied 
 to the unit, and the speed, depth of skimming, and operating cycle of the 
 float scrapers. During the course of the months that the experimental work 
 was carried out, these variables were changed a number of times. Often these 
 changes were made out of necessity in order to keep the unit operating so 
 
114 
 
 that continuous solids separation could be maintained. Therefore, it was not 
 possible to maintain any one set of operational conditions from which compar- 
 isons of the floatability of the sludges of various characteristics could be 
 made. Some float concentration data were collected, however. The float 
 concentration data which coincided with the rheological data which were taken 
 have been plotted and are presented in Figure k3. The data obtained from the 
 batch flotation tests has been plotted on the same figure for purposes of 
 comparison. Despite the changes in operational variables which were made in 
 the continuous unit from time to time it is indeed interesting to note that 
 the plots of yield strength vs float solids concentration for the continuous 
 unit and for the batch tests are essentially identical. 
 8. Summary 
 
 The microorganisms which predominated in the activated sludge 
 systems employed in this study shifted unpredictably during the course of the 
 work. The shifts were believed to have resulted from the forces of popula- 
 tion dynamics. The operating problems encountered in trying to maintain 
 steady state conditions may have contributed to these shifts but are not 
 considered to be a major cause. 
 
 The relationship between the nitrogen content of the sludge and the 
 flotation parameters and between the C/N ratio of the sludge and the flota- 
 tion parameters were only general in nature. As the percentage of nitrogen 
 increased and the C/N ratio decreased, the float concentration and rise rates 
 tended to increase. There were valid exceptions to that general rule which 
 made the use of the nitrogen content and the C/N of the sludges unreliable 
 as indicators of flotation behavior. 
 
 There was no apparent relationship between the amount of extracel- 
 lular polysaccharide which could be stripped from the sludge by blending and 
 
115 
 
 2.i» 
 
 2.2 
 
 2.0 
 
 .8 
 
 1.6- 
 
 1.2- 
 
 1.0 
 
 0.8 
 
 1 
 
 • 
 
 1 1 
 
 V /V - 0.8 
 
 w s 
 
 C - ^000 mg/1 
 
 C - 2220 mg/1 
 
 m 
 
 - 
 
 
 Time - 20 min 
 
 -c 
 
 
 
 t at C - 1*000 mg/1 
 
 
 i 
 
 • Batch Test Data 
 
 — 
 
 ^.Zoogleal 
 Sludges 
 
 Continuous Unit 
 Data _ 
 
 - 
 
 Flexible 
 
 - 
 
 — 
 
 Fl laments 
 
 - 
 
 A • 
 
 Stiff Filaments 
 
 •\ 
 
 - 
 
 
 
 — 
 
 <D -*■ — ft - 
 
 1 
 
 
 1 1 
 
 1 2 3 * 
 YIELD STRENGTH, dynes/sq cm 
 
 FIGURE 43. RELATIONSHIP BETWEEN SLUDGE YIELD 
 STRENGTH AND FLOAT CONCENTRATION 
 FOR CONTINUOUS FLOTATION UNIT 
 
116 
 
 the flotation characteristics of the sludge. The amount of polysaccharide 
 stripped varied inversely with the nitrogen content of the sludge. 
 
 The rheological properties of the sludge were found to be the most 
 useful parameters for relating sludge characteristics to flotation behavior. 
 This was as might be expected since the flotation behavior of the sludge is 
 determined by the physical characteristics of the sludge and the rheological 
 properties are measures of the physical and morphological characteristics of 
 the suspension. The rheological properties of the sludges varied with the 
 changes in the morphology of the sludges, with the stiff filamentous sludges 
 giving yield strengths which were an order of magnitude greater than those of 
 the zoogleal sludges at comparable solids concentrations. The plastic vis- 
 cosities varied directly with the yield strengths. The yield strengths of 
 the sludges were found to increase as the nitrogen content of the sludges 
 decreased as long as they did not exceed 1.5 dynes/sq cm. At higher yield 
 strengths there seemed to be no relationship. The plastic viscosity varied 
 inversely with the nitrogen content. 
 
 The rheological data correlated well with the batch flotation data, 
 particularly with the float concentration data. The lower yield strength 
 sludges gave higher float concentrations. Since the yield strength is a mea- 
 sure of the resistance to compaction it is reasonable that a relationship 
 would exist between yield strength and float solids concentration. A similar 
 relationship was found to exist between float concentration and plastic 
 viscosity. The plots where yield strength and plastic viscosity were plotted 
 against rise rate gave curves very similar in shape and again showed that the 
 lower yield strength sludges floated more quickly. Float solids data col- 
 lected from the continuous laboratory flotation unit when plotted against the 
 
117 
 
 yield strength data gave a curve which was essentially identical to the curve 
 obtained from the batch flotation tests. 
 
 B. Batch Flux Analysis of Continuous Thickener Operation 
 1. General 
 
 The batch flux plot provides a convenient method for analyzing and 
 predicting the performance of continuous sedimentation basins and thickeners. 
 This method has been promoted and used in the deisgn and analysis of gravity 
 thickeners (Dick 1970a, 1970b) but, to this author's knowledge, has never been 
 applied to flotation processes. Since flotation is essentially sedimentation 
 in reverse, it would seem that this method of analysis should be applicable 
 to flotation. 
 
 This method is based upon the concept that in a suspension each 
 layer of different concentration has a certain capacity to transmit solids 
 to the next layer of higher concentration. If the rate at which solids are 
 supplied to a layer exceeds the layer's capacity to transfer them, the layer 
 then increases in thickness and limits the operation of the thickener. The 
 limiting layer can be determined by use of the data obtained from a series of 
 batch tests made using a range of concentrations of sludge which would include 
 the concentration of the limiting layer. 
 
 The solids flux is defined as the weight of solids which can pass 
 through a unit area in a unit of time. For a batch process the flux is 
 obtained using the following equation (Dick, 1970): 
 
 where 
 
 G, = C.V. ( 17 ) 
 
 b i i 
 
 G, = batch flux 
 
 C. = initial concentration of solids in 
 terms of wt/unit volume 
 
 V. = interfacial velocity 
 
118 
 
 The flux curve is drawn by plotting the values of batch flux against the con- 
 centrations for which they were determined. A typical flux plot is shown in 
 Figure kk. Straight lines drawn tangent to the flux plot and connecting the 
 two axes represent the operating conditions of a continuous thickener. The 
 slope of the line is the withdrawal velocity of the thickened sludge having 
 
 concentration C , where C is the intercept of the abscissa when the applied 
 u u 
 
 flux is G,, the intercept of the ordinate. The concentration represented by 
 the point of tangency is that of the limiting layer. 
 
 The flux curve can be used both for design of thickeners and in the 
 analysis of the performance of existing ones. When used in design several 
 alternative designs can easily be considered by drawing a number of tangents 
 to the flux curve. Similarly, the flux plot can be used to evaluate the 
 effects of operational changes upon thickener performance. 
 
 2. Comparison of Batch Flux Results With Laboratory Flotation Unit Data 
 In this work the batch flux curve was tested to see i f this method 
 of analysis could be used to predict the operation of continuous flotation 
 units. The batch flux curves were constructed from data obtained from batch 
 rise rate tests which were performed over a range of solids concentrations. 
 The batch testing procedure described in the previous chapter was used. The 
 air to solids ratio was kept constant for all tests made in any particular 
 run. The air to solids ratio used in the batch tests were the same as that 
 of the continuous unit. The tests were run in 3-5 and 7-5 in. diameter 
 columns. The flux applied to the continuous flotation unit and the float 
 solids concentrations were known. The rise rates obtained for the batch tests 
 performed at pressures greater than AO psig were multiplied by the appropriate 
 correction factors to bring them all to the common base of kO psig. The cor- 
 rection factors were based on the data of Figure 10. The factors were 1.35, 
 
119 
 
 L U 
 SUSPENDED SOLIDS CONCENTRATION, mg/1 
 FIGURE kk. METHOD OF DETERMINING LIMITING CONDITIONS 
 
120 
 
 1.80 and 3-26 respectively for pressures of 60, 80, and 120 psig. An opera- 
 tion line drawn from the flux axis tangent to the batch flux curve gave the 
 predicted float solids concentration at the intersection of the suspended 
 solids axis. The predicted float concentration was then compared with the 
 actual values obtained from the continuous flotation unit. A typical batch 
 flux plot constructed from batch flotation test data is shown in Figure k5 . 
 
 A summary of the data obtained from a number of runs is presented 
 in Table 5. In the first several runs which were made the flux curves were 
 not extended far enough so that the limiting solids concentration could be 
 reached. As a result the flux plots did not reach values as low as the 
 applied flux so, of course, it was impossible to determine a predicted float 
 concentration. These runs were made at the air to solids ratios between 
 0.006 and 0.0109 using maximum pressures of 80 psig. The data for these 
 curves are not shown. By increasing the maximum pressure of saturation to 
 120 psig and reducing the air to solids ratio to 0.00^6 it was possible to 
 extend the solids flux curve far enough that a limiting concentration could 
 be reached. A graphical comparison of the actual float concentrations 
 obtained from the continuous flow flotation unit and the values predicted by 
 the batch flux analyses is presented in Figure k(>. These data indicate sur- 
 prisingly good agreement in several cases. This is especially true when the 
 number of factors influencing the results of the batch flotation tests and 
 the magnitude of the correction applied for the differences in pressures of 
 saturation are considered. For some values the agreement was not as good. 
 The lack of agreement in these cases may be due to the effects of the many 
 factors which influence the test results. 
 
 The two cases where the predicted float concentrations greatly 
 exceeded the actual concentrations obtained, that is, batch flux curves 5 and 
 
121 
 
 o 
 
 
 z 
 
 o 
 
 
 Z3 
 
 o 
 
 
 
 vO 
 
 
 Z 
 
 
 
 o 
 
 t- 
 < 
 1- 
 o 
 
 o 
 
 
 _l 
 
 o 
 
 
 L_ 
 
 o 
 
 «— 
 
 
 -3" 
 
 V. 
 
 CO 
 
 *— 
 
 en 
 
 => 
 
 
 E 
 
 o 
 
 3 
 
 
 •> 
 
 z 
 
 
 z 
 
 
 o 
 
 o 
 
 H 
 
 o 
 
 
 z 
 
 o 
 
 H- 
 
 o 
 
 CVJ 
 
 < 
 
 o 
 
 i — 
 
 cc 
 
 
 
 1- 
 
 >- 
 
 
 z 
 
 QC 
 
 
 LU 
 
 o 
 
 
 O 
 
 \- 
 
 o 
 
 z 
 
 < 
 
 o 
 
 o 
 
 a: 
 
 o 
 
 o 
 
 o 
 
 o 
 
 
 CD 
 
 1 — 
 
 CO 
 
 < 
 
 
 Q 
 
 _j 
 
 
 _l 
 
 LU 
 
 
 o 
 
 Z 
 
 
 CO 
 
 y- 
 
 o 
 
 
 
 o 
 
 Q 
 
 cc 
 
 o 
 
 LU 
 
 o 
 
 oo 
 
 O 
 
 Z 
 
 Lu 
 
 
 LU 
 
 1- 
 
 
 Q- 
 
 O 
 
 
 CO 
 
 _J 
 
 
 ZD 
 
 a. 
 
 
 CO 
 
 
 o 
 
 
 X 
 
 o 
 
 
 Z) 
 
 o 
 
 
 _J 
 
 vO 
 
 
 U_ 
 
 z 
 o 
 
 < 
 
 o 
 
 
 03 
 
 o 
 
 
 
 o 
 
 
 _J 
 
 Aep/;j bs/q[ 'xn[j ip}eg 
 
O C X> C 
 
 O L- 
 
 _ ._ <u 
 
 
 
 >~ 
 
 V) 
 
 
 
 
 
 
 
 
 
 
 
 >« 
 
 . — 
 
 ■a 
 
 
 
 
 
 
 
 
 
 
 o 
 
 -Q 
 
 ra 
 
 
 
 
 
 
 
 
 
 
 
 c 
 
 
 c 
 
 i— 
 
 
 
 
 
 
 
 
 
 
 
 
 X) 
 
 <r 
 
 
 
 o 
 
 o 
 
 O 
 
 O 
 
 o 
 
 o 
 
 O 
 
 O 
 
 o 
 
 c_> 
 
 CO 
 
 
 to 
 
 o 
 
 o 
 
 O 
 
 O 
 
 o 
 
 o 
 
 O 
 
 O 
 
 o 
 
 
 
 X 
 
 
 oo 
 
 CM 
 
 -3" 
 
 O 
 
 CM 
 
 CM 
 
 O 
 
 ca 
 
 o 
 
 
 u 
 
 -i 
 
 
 
 
 
 
 
 
 
 
 
 m 
 
 
 — 
 
 c 
 
 p«- 
 
 1^ 
 
 LA 
 
 \0 
 
 -3" 
 
 vO 
 
 vO 
 
 oo 
 
 o 
 
 o 
 
 XI 
 
 u. 
 
 a) 
 
 
 
 
 
 
 
 
 
 rA 
 
 
 U 
 
 
 V 
 
 
 
 
 
 
 
 
 
 
 U_ J- _C I- 
 
 
 o — 
 
 
 c — 
 
 
 O O 
 
 
 O CO 
 
 +j 
 
 4-1 4-1 
 
 .— 
 
 ro c 
 
 c: 
 
 O cu 
 
 => 
 
 — o 
 
 
 U. 1_ 
 
 U) 
 
 oj 
 
 3 
 
 a. 
 
 O 
 
 
 13 
 
 
 C 
 
 
 — 
 
 X >» 
 
 4-J 
 
 3 CO 
 
 c 
 
 — XJ 
 
 o 
 
 u- \ 
 
 o 
 
 
 
 X) <+- 
 
 
 cu 
 
 
 — O" 
 
 
 •— in 
 
 
 Q-^v. 
 
 
 Q.J3 
 
 
 < — 
 
 122 
 
 vO 
 
 CO 
 
 
 
 
 
 en 
 
 OO 
 
 ro 
 
 •— 
 
 -3" 
 
 o 
 
 o 
 o 
 o 
 
 o 
 o 
 
 CM 
 
 o 
 o 
 
 -3" 
 
 O 
 
 o 
 
 -3- 
 
 O 
 O 
 LA 
 
 O 
 O 
 OO 
 
 O 
 O 
 OO 
 
 O 
 O 
 CA 
 
 O 
 O 
 
 -3" 
 
 O 
 O 
 vO 
 
 CM 
 
 r^ 
 
 CT\ 
 
 LA 
 
 LA 
 
 VO 
 
 LA 
 
 CM 
 
 -3" 
 
 OO 
 
 OO 
 
 VO 
 
 LA 
 
 CA 
 
 CM 
 
 O 
 
 LA 
 
 vO 
 
 r-» 
 
 en 
 
 -3" 
 
 -3- 
 
 -3- 
 
 -3" 
 
 LA 
 
 <r 
 
 -3- 
 
 ^r 
 
 LA 
 
 o 
 
 O 
 
 O 
 
 O 
 
 O 
 
 O 
 
 o 
 
 o 
 
 o 
 
 O 
 
 
 o 
 
 O 
 
 o 
 
 o 
 
 O 
 
 o 
 
 o 
 
 o 
 
 O 
 
 o 
 
 vO 
 
 vO 
 
 vO 
 
 vO 
 
 vO 
 
 vO 
 
 vO 
 
 vO 
 
 r^. 
 
 cr> 
 
 -3" 
 
 -3- 
 
 -3" 
 
 -3" 
 
 -3" 
 
 -3" 
 
 -3" 
 
 -3" 
 
 LA 
 
 o 
 
 O 
 
 O 
 
 O 
 
 O 
 
 O 
 
 O 
 
 O 
 
 O 
 
 o 
 
 
 o 
 
 O 
 
 o 
 
 o 
 
 o 
 
 O 
 
 O 
 
 o 
 
 o 
 
 o 
 
123 
 
 </5 
 
 ■z. oo 
 
 O =5 
 
 — O 
 
 I- => 
 
 < 2 
 
 aL — 
 
 O LTv O LT\ 
 
 CM — «- 
 
 spiios luaojad ' uoiie-nuaouoo leou leruoy 
 
\2k 
 
 6, may have resulted from instances when the continuous floatation unit was 
 underloaded. With the flotation unit operating at a given sludge withdrawal 
 rate underloading would result in a float concentration which is less than 
 the maximum obtainable. The cases where the float obtained was more concen- 
 trated than predicted cannot be explained by underloading. Some of the 
 variation in the data may be due to the fact that the flotation unit may not 
 always have been operated at the conditions which would give the maximum 
 float solids concentrations. 
 
 The fact that the laboratory continuous flow flotation unit was not 
 hydraul ical ly nor geometrically a model of a real full-scale unit and is 
 rather a toy in comparison may also help to explain the fact that the solids 
 flux curve did not always accurately predict the float solids obtained from 
 the unit. 
 
 In the batch flotation tests using the most concentrated sludges 
 long lag periods were noted between the time of column filling and the time 
 when an observable interface was formed. It seemed possible that during this 
 long lag period the microoganisms in the sludge might be reducing the flota- 
 tion driving force by consuming the oxygen present in the supplied air. In 
 order to determine if this might be the case the continuous flotation unit 
 was operated on nitrogen gas while the batch tests for the solids flux curve 
 were also performed using nitrogen. The results of that series of tests are 
 listed as Run k in Table 5 and are essentially the same as those run using 
 air as the agent of flotation. 
 
 It was desired that the continuous flotation unit could be operated 
 over a fairly wide range of solids fluxes so that the operating lines drawn 
 between the fluxes and the resulting float concentrations could be treated as 
 tangents and permit the construction of the actual solids flux curve which 
 
125 
 
 s/ou 1 d define the continuous unit's operation. On a number of occasions attempts 
 
 vere made to obtain these data; but, at best, they were only partially suc- 
 
 ;essful . The problem here was to maintain a constant air to solids ratio while 
 
 nanipulating the applied flux. 
 
 With the flotation unit being used as a clarifier, the float solids, 
 
 3fter being collected, were recycled back to the aeration tank. Any adjust- 
 
 nent in the operating parameters of the flotation unit which would result in 
 
 3 change in the volume of return sludge would also result in more feed being 
 
 discharged to the flotation unit. The attaining of a constant air to solids 
 
 ratio was a trial and error procedure. 
 
 3. Investigation of Possible Methods for Extending the Batch 
 Flux Plot 
 
 a. Direct Pressur ization 
 
 The possibility of using direct pressur ization of the sludge itself 
 as a means of extending the flux curve to higher concentrations was explored. 
 With this method the mixed liquor would be aerated and saturated with air 
 under pressure and then released to the flotation chamber. This was tried 
 and compared with the standard method of performing the batch test. A stan- 
 dard flotation test was performed using a sludge which gave a blended suspended 
 solids concentration of about 7,200 mg/1 . The sludge to be directly pressur- 
 ized was diluted to give it a concentration of 7,200 mg/1. The sludge was then 
 pressurized at a pressure which would result in an air to solids ratio similar 
 to that of the standard test. The sludge, upon floating, gave a rise rate that 
 <vas more than twice as fast as that for the sludge which was floated in the 
 standard way. The interface was harder to distinguish and the subnatant was 
 nore turbid than with the standard test. Since the rise rates were so dif- 
 ferent for the two methods and the interface was very poorly defined, direct 
 pressurizat ion could not be used as a means of extending the flux curve. 
 
126 
 
 While the direct pressuri zat ion technique does not represent what 
 is usually done in full scale flotation units which are used to thicken waste 
 activated sludge, it does make one wonder if this could not be done in prac- 
 tice. The subnatant is, of course, more turbid; but if this were recycled 
 through the plant it might not be too much of a problem. 
 
 b. Talmadge and Fitch Method 
 
 Talmadge and Fitch (1955) developed a method based upon the Kynch 
 analysis which made it possible to develop the solids flux plot from one batch 
 settling test rather than from a multiple batch test procedure. Their reason- 
 ing was that since all layers having a concentration greater than the layers 
 above them will ultimately propagate to the liquid interface, the interfacial 
 settling velocity for a layer of any given concentration can be determined 
 from the slope of tangents drawn to the sludge interface-time curve. The 
 sludge concentration associated with any given slope can be determined by 
 extending the tangent back to the axis where time is zero. Since by the time 
 a layer of any concentration has propagated to the surface all the solids in 
 the column have passed through that layer, the concentration at any height 
 can be determined using the following formula: 
 
 r C H 
 C. = o o 
 
 where 
 
 H, ( 18 ) 
 
 C, = concentration at height H, 
 
 C = initial concentration before separation commences 
 o 
 
 H = initial depth of sludge in column 
 
 H, = depth of sludge at the time being considered 
 
127 
 
 This method was tried using laboratory grown sludges with the hope 
 :hat it might provide a means for extending the flux curve. The type of 
 -esults obtained when this method was applied to batch rise rate data is 
 shown in Figure k~J . As can be seen from this plot, this is not a satisfactory 
 nethod for extending the flux curve. If the method were applicable for use 
 vith this kind of sludge the batch flux curves calculated from the different 
 -ise rate curves should be essentially identical. The failure of the flux 
 ;urves to match results from the fact that the characteristics of activated 
 sludge differ from those of the ideal suspension assumed in the Kynch analysis 
 jpon which this procedure was based. The ideal suspension of the Kynch 
 analysis was considered to be one in which the particles were rigid, spherical 
 jniform in size and shape and evenly distributed in any horizontal plane. 
 Activated sludge is made up of non-rigid, flocculent, non-spherical particles 
 tfhich are not necessarily uniform in size nor shape. Additionally, the floes 
 form aggregates and structures which possess yield strengths (Dick, 1965). 
 4. Effect of Anaerobic Conditions Upon Flotation 
 
 When dealing with aerobic activated sludges it is important that the 
 sludge be kept aerobic until its use in the batch flotation test if the rise 
 rates obtained are to represent what happens in flotation of the aerobic 
 sludge. An experiment was performed in order to determine the magnitude of 
 the effect of anaerobic conditions upon rise rate. A sludge sample was col- 
 lected and one float test was performed shortly after collection. The 
 remainder of the sludge was divided with one half being aerated while being 
 held and the other being permitted to enter anoxic conditions. Tests were 
 performed several hours later and the rise rates of the sludges were compared. 
 The rise rate for the unaerated sludge, after 6 hrs, had dropped from the 
 initial value of 15 ft/day to 6.6 ft/day while the aerated sample had a rise 
 
128 
 
 0) 
 
 U 
 
 
 > 
 
 4J 
 
 
 L. 
 
 
 
 3 
 
 U- 
 
 
 c_> 
 
 T5 
 
 
 X 
 
 C 
 
 
 =3 
 
 (TJ 
 
 
 
 
 
 u. 
 
 en 
 
 
 T3 
 
 T) 
 
 
 H) 
 
 (Ti 
 
 
 
 F 
 
 
 o 
 
 
 TJ 
 
 — 
 
 01 
 
 
 
 -o 
 
 H 
 
 -C 
 
 0) 
 
 
 
 L. 
 
 -Q 
 
 J! 
 
 — > 
 
 en 3 
 E o 
 
 Aep/}j bs/qi 'xnu spnos 
 
129 
 
 rate of 17-3 ft/day. The data clearly shows that being held without aeration 
 results in reduced rise rates when the sludge was subsequently floated. This 
 was probably due at least in part to the utilization by the microorganisms of 
 the oxygen in the air attached to sludge particles. Changes may also have 
 taken place in the surface properties of the sludge due to the anoxic condi- 
 tions. 
 
 5. Summary 
 
 In theory the batch flux analysis should be applicable to use with 
 flotation units as well as with gravity thickeners. One serious problem in 
 the development of the flux curve is the difficulty encountered in obtaining 
 blended suspended solids concentrations high enough to equal or exceed the 
 limiting solids concentration. Pressures of saturation higher than those 
 normally employed in flotation were used to extend the flux plot to higher 
 sludge concentrations. The rise rate values obtained at the higher pressures 
 were adjusted to compensate for the loss in rise rate associated with the use 
 of the higher pressures. In view of the number of factors which influence 
 the batch flotation test results and considering the magnitudes of the cor- 
 rection factors which were applied to account for the effect on rise rates of 
 changing pressures of saturation, surprising agreement between the predicted 
 and actual values of float concentration were obtained in several cases. In 
 about an equal number of instances, however, good correlation was not obtained 
 The reasons for the lack of agreement probably stem from the many factors 
 which influence the batch flotation test and the possibility that the contin- 
 uous flotation unit was underloaded in some cases. 
 
 The Talmadge and Fitch method was not a satisfactory means for 
 extending the flux curve for the type of sludges used in this study. Direct 
 pressurization of the sludge gave rise rates in excess of those obtained using 
 
130 
 
 the standard batch flotation procedure. For that reason and also because of 
 the poorly defined interface formed, it did not provide a good means for 
 extending the curve. The batch flux analysis was found to be somewhat limited 
 in its usefulness in predicting the float concentrations obtainable from con- 
 tinuous flotation units. The necessity of applying correction factors to 
 compensate for the effect of column diameter, sludge height, and changes in 
 the pressure of saturation, together with the need for relatively large vol- 
 umes of sludge in order to determine the effects of column diameter, and the 
 difficulty in extending the curve made the use of this method somewhat less 
 attractive than it originally appeared to be. 
 
 C. Significance of Results 
 
 The success of the flotation process as a means of solids separation 
 is dependent upon the properties of the material to be floated. This investi- 
 gation has included the study of some properties of bulked sludges and has 
 related these properties to the flotation behavior. This contribution of 
 information to the meager amount in this area should be helpful to those con- 
 fronted with the problem of separating bulked sludges. The good relationship 
 which was shown to exist between the flotation behavior and the rheological 
 properties suggests their use as an aid in predicting flotation unit perfor- 
 mance. 
 
 Batch flotation tests are usually performed for the purpose of 
 evaluating the flotation characteristics of the sludge or other material to 
 be floated prior to the design and installation of a flotation unit. The rate 
 of separation and the concentration of solids obtained in the float are the 
 two main parameters considered. Some of the significant findings of this 
 research relate to the identification of a number of factors which influence 
 the results of batch flotation tests. The consideration of the effects of 
 
131 
 
 flotation column diameter, column depth and column filling time should lend 
 improvement to results which are obtained from batch flotation test procedures 
 The introduction of a method of performing the batch flotation test which more 
 closely parallels the process which takes place in continuous flow units 
 should improve flotation test results. 
 
 The information provided here regarding the effects of saturation 
 pressure upon rise rates points out the importance of this parameter in the 
 flotation process. 
 
 Consideration of the several factors mentioned here should result in 
 the obtaining of better information upon which flotation units may be 
 designed. 
 
V. CONCLUSIONS 
 
 I. The results of the batch flotation test procedure are influenced by the 
 column filling time, the column diameter and the column depth. The 
 magnitude of these effects upon the interfacial rise rate is dependent 
 upon the solids concentration and characteristics of the sludge. 
 
 I. Data obtained using the method which is commonly in use in practice may 
 give inaccurate results because the factors mentioned previously which 
 influence the test results are not generally taken into consideration. 
 
 J. A technique for performing batch flotation tests which is believed to 
 more closely represent the process in a continuous flotation unit was 
 developed. This method gives better flotation results than the com- 
 monly used method, especially with highly bulked sludges. 
 
 k. A retardation factor exists in batch flotation tests similar to that 
 
 which exists in sedimentation. The magnitude of the retardation factor 
 is dependent upon the solids concentration and the physical properties 
 of the sludge. 
 
 5. Filamentous sludges can be floated to concentrations of at least one 
 percent solids by dissolved air flotation. The maximum applied loading 
 rate must be kept lower in some cases than for the zoogleal sludges. 
 
 6. The carbon to nitrogen ratio and the nitrogen content of the sludge are 
 not reliable indicators of the flotation behavior of the sludge. 
 
 7. The carbon to nitrogen ratio and the nitrogen content of the sludge 
 correlate well with the plastic viscosity of the sludge. The relation- 
 ship between yield strength and the carbon to nitrogen ratio and the 
 nitrogen content of the sludge are not well defined. 
 
 8. The rheological properties of the sludge, yield strength and plastic 
 
 132 
 
133 
 
 viscosity are the best means of characterizing bulked sludges and of 
 relating the sludge properties to its flotation behavior. 
 The yield strengths of the sludges are dependent upon the morphological 
 characteristics of the organisms making up the sludges. 
 The amount of strippable extracellular polysaccharide present in the 
 sludge does not appear to have any influence upon the flotation behavior 
 of the sludge. 
 
 In view of all the factors which influence its results, the batch flux 
 analysis gives surprisingly accurate predictions of the float solids 
 concentrations obtained from the laboratory continuous flow flotation 
 unit on several occasions. The fact that in other instances the pre- 
 dicted and actual do not agree is probably due to the many factors which 
 influence batch flotation test's. 
 
 The pressure at which the recycle water is saturated was found to greatly 
 influence the flotation rise rate of the sludge. The rise rate decreases 
 exponentially as the pressure increases even though the air to solids 
 ratio is held constant and the possibility that in some cases the con- 
 tinuous flotation unit may have been underloaded. 
 
 The use of a number of pressures of saturation as a means of extending 
 the batch flux curve gives good results as long as the rise rates are 
 corrected to account for the loss in rise rates due to the use of the 
 higher pressures. The precision with which the blending operation can 
 be performed together with the necessary but undesirable extrapolation 
 of the data to obtain the correction factors will limit the magnitude of 
 the pressures which will be of practical use. 
 
 The Talmadge and Fitch method of analysis is an unsatisfactory method 
 for extending the batch flux curve, at least for f i lamentously bulked 
 sludges. 
 
13* 
 
 The coaxial cylinder rotational viscometer is not suitable for determining 
 the rheological properties of the floated sludge because the loss of the 
 air bubbles from the float while in the viscometer gives changing values 
 of torque at constant shear rate. 
 
VI. SUGGESTIONS FOR FUTURE WORK 
 
 On the basis of the results of this study and observations made 
 during the course of the investigation, it is suggested that further investi- 
 gation should be pursued in the following areas: 
 
 1. A study of the surface properties of activated sludges and their influence 
 upon the efficiency of air bubble attachment and the subsequent flotation 
 of the sludge should be made. This would involve the devlopment of a 
 method for measuring the efficiency of bubble attachment. 
 
 2. A study of the influence of the dissolved organic material present in the 
 suspending liquid upon the flotation behavior of activated sludge should 
 be made. It was observed in the course of the current study that the 
 addition of primary sewage effluent to the aeration tank for a period of 
 an hour or two markedly improved the flotation characteristics of poorly 
 floating sludges. Information which would explain the reason for this 
 would be enlightening. 
 
 135 
 
REFERENCES 
 
 Bechir, M. H. and Symons , J. M. I966. "The Effect of Nitrogen Deficiency on 
 the Behavior of the Complete-Mixing Activated Sludge Process," Air and 
 Water Pollution International Journal 10:191- 
 
 Bhatla, M. N. 1967- "Relationship of Activated Sludge Bulking to Oxygen 
 Tension," Journal of the Water Pollution Control Federation 39:1978. 
 
 Bloodgood, D. E. 19^7- "Concentration of Activated Sludge," Sewage Works 
 Journal 19:202. 
 
 Cassell, E. A., Sulzer, F. T. , and Lamb, J. C. 1 966 . "Population Dynamics and 
 Selection in Continuously Mixed Cultures," Journal of the Water 
 Pollution Control Federation 38:1398. 
 
 Chakrabarti, S. K. 1 968 . "Changes in Some Physical Properties of Activated 
 Sludge Under Different Biological Conditions, "Master's Thesis," 
 University of Illinois, Urbana, Illinois. 
 
 Clifford, W. and Windredge, M. E. D. 1932. "The Settlement of Activated 
 Sludge." Proceedings of the Institute of Sewage Purification 115- 
 
 Cole, R. F. 1969. "Experimental Evaluation of the Kynch Theory," Doctoral 
 Thesis, University of North Carolina, Dissertation Abstracts 29:4687B. 
 
 Conrad, H. E. I969. Private Communication. 
 
 Dick, R. I. 1965. "The Applicability of Prevailing Gravity Thickening Theories 
 to Activated Sludge," Doctoral Thesis, University of Illinois, Urbana, 
 11 1 inoi s. 
 
 Dick, R. I. 1970a. "Thickening," Jjt_ Gloyna, E. F. and Eckenfelder, Jr., W. W. 
 (ed) Water Quality Improvement by Physical and Chemical Processes , 
 University of Texas Press, Austin, Texas. 
 
 Dick R. I. 1970b. "Role of Activated Sludge Final Settling Tanks," Journal 
 Sanitary Engineering Division, Proceedings of the American Society of 
 Civil Engineers 96:423. 
 
 Dick R. I. and Ewing, B. B. 1967- "The Rheology of Activated Sludge," Journal 
 of the Water Pollution Control Federation 39:5^3- 
 
 Dobias, B. and Vinter V. I966. "Flotation of Microogani sms ," Fol ia Microbiolo- 
 gica 11 :31A. 
 
 Eckenfelder Jr., W. W. , Rooney, T. F. , Burger, T. B. and Gruspier, J. T. 1956. 
 "Studies on Dissolved Air Flotation of Biological Sludges," j_n_ 
 McCabe, B. J. and Eckenfelder, Jr., W. W. , Biological Treatment of Sewage 
 and Industrial Wastes , Vol. II, Reinhold, New York. 
 
 136 
 
137 
 
 Ettelt, G. A. 1 96^4 . "Activated Sludge Thickening by Dissolved Air Flotation," 
 Proceedings 19th Industrial Waste Conference , Purdue University. 
 
 Farnsworth, G. A. 1967- "The Effect of Induced Flocculation on Settling and 
 Thickening Behavior of Activated Sludge," Masters Thesis, University of 
 1 1 1 inois , Urbana , 111 inois . 
 
 Ford, D. L. and Eckenfelder, W. W. I966. "The Effect of Process Variables on 
 Sludge Floe Formation and Settling Characteristics," The University of 
 Texas Center for Research in Water Resources Technical Report to the 
 Public Health Service EHE-11-6604. 
 
 Freidman, B. A., Dugan, P. R. , Pfister, R. M. , Remsen , C. C. 1 968 - "Fine 
 Structure and Composition of the Zoogleal Matrix Surrounding Zooglea 
 rami gera ," Journal of Bacteriology 96 : 2 1 44 . 
 
 Garwood, J. 1967- "Polyelectrolytes in Sewage Treatment," Effluent and Water 
 Treatment Journal 7:380. 
 
 Gaudy, A. F. , Jr. 1962. "Color imeteri c Determination of Protein and Carbo- 
 hydrate," Industrial Water and Wastes 7:17- 
 
 Gaudy, E. and Wolfe, R. S. 1962. "Composition of an Extracellular Polysac- 
 charide Produced by Sphaerotilus natans ," Applied Microbiology 10:200. 
 
 Gaudin, A. 1967- Flotation , Second Edition, McGraw-Hill, New York, N. Y. 
 
 Gaudin, A. M. , Fuerstenau, M. C. and Mitchell, S. R. 1959- "Effect of Pulp 
 Depth and Initial Pulp Density in Batch Thickening," Mining Engineering 
 11:613- 
 
 Genetelli, E. J. and Heukelekian, H. 1964a. "The Effect of Chemical Composition 
 of Substrate and Loading on the Performance and Bulking of Activated 
 Sludge," Journal of the Water Pollution Control Federation 36:643- 
 
 Genetelli, E. J. and Heukelekian, H. 1964b. "Components of the Sludge Loading 
 Ratio and Their Effect on the Bulking of Activated Sludge," 
 Proceedings 19th Industrial Wastes Conference, Purdue University. 
 
 Gratteau, J. C, and Dick, R. I. 1968. "Activated Sludge Suspended Solids 
 Determination," Water and Sewage Works 115:468. 
 
 Greeley, S. A. 1945- "The Development of the Activated Sludge Method of Sewage 
 Treatment," Sewage Works Journal 17:1135- 
 
 Green, H. 1949- Industrial Rheology and Rheological Structures , John Wiley, 
 New York. 
 
 Greenburg, A. E., Klein, G. , and Kaufman, W. J. 1955- "Effect of Phosphorous 
 on the Activated Sludge Process," Sewage and Industrial Wastes 27:277- 
 
 Haseltine, T. R. 1932. "The Activated Sludge Process at Salinas, California, 
 
 with Particular Reference to Causes and Control of Bulking," Sewage Works 
 Journal 4:461 . 
 
138 
 
 iaseltine, T. R. 1938. "The Use of Hydrated Lime in the Activated Sludge 
 Process," Sewage Works Journal 10:1017- 
 
 lelmers, E. N., Frame, J. D. , Greenberg, Sawyer, C. N., 1951- "Nutritional 
 Requirements in the Biological Stabilization of Industrial Wastes, ill, 
 Treatment with Supplemental Nutrients," Sewage and Industrial Wastes, 
 24:496. 
 
 teukelekian, H. 1941. "Activated Sludge Bulking," Sewage Works Journal 19:39. 
 
 teukelekian, H. and Weisberg 1956. "Bound Water and Activated Sludge Bulking," 
 Sewage and Industrial Wastes 28:558. 
 
 loover, S. R. and Porges, P. 1949- "Treatment of Dairy Wastes by Aeration -II: 
 Continuous Aeration Studies," Proceedings of the Fifth Industrial Wastes 
 Conference, Purdue University. 
 
 torne, W. R. , Hartman, H. E., Rinanca, U. S., and St. Clair, R. L. 1962. 
 "Biological Treatment of Fine Chemical Plant Wastes," Journal of the 
 Water Pollution Control Federation 34:833- 
 
 iurwitz, E. and Katz, W. J. 1959- "Concentrating Activated Sludge to a Fuel 
 Value of 4,000 BTU per Gallon," Wastes Engineering 30:730. 
 
 Ingols, R. S. and Heukelekian, H. 1939- "Studies on Activated Sludge Bulking," 
 Sewage Works Journal 1 1 :927- 
 
 Ingols, R. W. and Heukelekian, H. 1940. "Studies on Activated Sludge Bulking 
 II, Bulking Induced by Domestic Sewage," Sewage Works Journal 12:703- 
 
 Jones, P. H. 1964. "The Effect of Temperature and Oxygen Tension on One of the 
 Microorganisms Responsible for Sludge Bulking," Proceedings of the 19th 
 Industrial Waste Conference , Purdue University. 
 
 Jones, P. H. 1965- "The Effect of Nitrogen and Phosphorus Compounds on One 
 of the Microorganisms Responsible for Sludge Bulking," Proceedings 20th 
 Industrial Waste Conference, Purdue University. 
 
 Jones, P. H. 1966. "Experiments with Bulking Sludge and its Ability to Degrade 
 High Carbohydrate Wastes and Sulphite Waste Liquor," Pulp Paper Magazine 
 Canada 67, No. CT-139-T-145 (Convention Issue 1 966) . 
 
 Jones, W. H. 1968. "Dissolved Air Flotation Thickening of Waste Water Sludges, 1 
 Presented at Nebraska Water Pollution Control Association, Great Plains 
 Design Conference, Omaha, Nebraska. 
 
 Juni, E. and Heym, G. A. 1964. "Pathways for Biosynthesis of a Bacterial 
 
 Capsular Polysaccharide, IV. Capsule Resynthesis by Decapsulated Resting- 
 Cell Suspension," Journal of Bacteriology 87:461. 
 
 valyuzhnyi, M. Y. , Petrushko, G. M. and Novikova, G. P. 1965-"Flocculation of 
 Candida uti 1 is and Candida tropical is Yeast Cel Is and its Relation to 
 Flotation," Mikro biologlya 34:800. 
 
139 
 
 Kammermeyer , K. 19^1- "Settling and Thickening of Aqueous Suspensions," 
 Industrial and Engineering Chemistry 32:l / t84. 
 
 Katz, W. J. 1958. "Sewage Sludge Thickening by Flotation," Publ ic Works 89: 1 1^ 
 
 Katz, W. J. 1959- "Adsorption - Secret of Success in Separating Solids by 
 Dissolved Air Flotation," Wastes Engineering 30:37^- 
 
 Katz, W. J. and Geinopolos, A. 1963. "Dissolved Air Flotation as a Method of 
 Thickening Aerobic Biological Solids," Paper presented at the High Level 
 Conference on Sludge Dewatering, University of Michigan, Ann Arbor, 
 Michigan, Jan. 30, 1963- 
 
 Katz, W. J., and Geinopolos, A. 1967- "Sludge Thickening by Dissolved Air 
 Flotation," Journal of the Water Pollution Control Federation 39:9^6. 
 
 Katz, W. J. and Geinopolos, A. 1968. "Concentration of Sewage Treatment Plant 
 Sludges by Thickening," Proceedings of the Tenth Sanitary Engineering 
 Conference , University of Illinois, Urbana, Illinois, Feb. 6 and 7, 1968. 
 
 Kaylor, F. B. , Bechir, M. B. , and Symons, J. M. 1963. "Atmospheric Nitrogen 
 Fixation in Activated Sludge," Proceedings T 8th Industrial Wastes 
 Conference , Purdue University, 1963 • 
 
 Keefer, C. E. and Meisel, J. 1955- "Activated Sludge Studies," Sewage and 
 Industrial Wastes 27:982. 
 
 Keefer, C. E. 1 963 - "Relationship of Sludge Density Index to the Activated 
 Sludge Process," Journal of the Water Pollution Control Federation 
 35:1166. 
 
 Kraus, L. S. 19^9. "Quantitative Relationships in the Activated Sludge 
 Process," Sewage Works Journal 21 : 6 1 3 • 
 
 Lackey, J. B. and Wattie, E. 1 940. "Studies of Sewage Purification, XIII. 
 the Biology of Sphaerotllus natans kutzlng in relation to Bulking of 
 Activated Sludge," Sewage Works Journal 12:669- 
 
 Logan, R. P. and Budd, W. E. 1955. "Effect of BOD loadings on Activated 
 
 Sludge Plant Operation," \n_ McCabe, B. J. and Eckenfelder, Jr. W. W.,(ed) 
 Biological Treatment of Sewage and Industrial Wastes , Pergamon Press, N.Y 
 1955- 
 
 Luderitz, 0., Staub, A. M. , and Westphal , 0. 1966. "Immunochemi stry of and 
 R Antigens of Salmonella and Related Enterobacter iacea ," Bacteriological 
 Review 30:192. 
 
 Ludzack, F. J., Schaffer, R. B. , and Ettinger, M. B. 1961. "Temperature and 
 Feed as Variables in Activated Sludge Performance," Journal of the Water 
 Pollution Control Federation 33 : 1 4 1 . 
 
 Mancini, J. L. 1962. "Gravity Clarifier and Thickener Design," Proceedings of 
 the 17th Industrial Wastes Conference, Purdue University. 
 
140 
 
 Mayo, W. E. 1965- "Industrial Waste Application of Dissolved Air Flotation," 
 Presented at Annual Meeting of Water Pollution Control Federation, 
 Atlantic City, N. J. 
 
 McWhorter, T. R. and Heukelekian, H. 1964. "Growth and Endogenous Phases in 
 the Oxidation of Glucose," Advances in Water Pollution Research, 
 McMi llan Co. 2:419- 
 
 Morgan, E. H. and Beck, A. J. 1939- "Carbohydrate Wastes Stimulate Growth of 
 Undesirable Filamentous Organisms in Activated Sludge," Sewage Works 
 Journal 1 : 46 . 
 
 Mulbarger, M. C, and Huffman, D. D. 1970. "Mixed Liquor Solids Separation by 
 Flotation," Journal of Sanitary Engineering Division, Proceedings of the 
 American Society of Civil Engineers 96:861 . 
 
 Nishikawa, S. and Kuriyoma, M. 1968. "Studies on Mucilage in Activated Sludge, 1 
 J ournal of Fermentation Technology 46:381 . 
 
 Pipes, W. 0. and Jones, P. H. 1963. "Decomposition of Organic Wastes by 
 Sphaerot i lus ," Biotechnology and Bioengineer ing 5:287. 
 
 Pipes, W. 0. 1967a. "Types of Activated Sludge Which Separate Poorly," Paper 
 Presented at the 40th Annual Meeting, Central States Water Pollution 
 Control Association, June 14, 1967- 
 
 Pipes, W. 0. 1967b. "Bulking of Activated Sludge," Advances in Appl ied 
 Microbiology , Academic Press, Vol. 9:185- 
 
 Pipes, W. 0. 1968. An Atlas of Activated Sludge Types , Department of Civil 
 Engineering, Northwestern University. 
 
 Ruchhoft, C. C. and Watkins, J. H. 1929- "Bacteriological Isolation and Study 
 of the Filamentous Organisms in the Activated Sludge of Des Plaines 
 River Sewage Treatment Works," Sewage Works Journal 1:52. 
 
 Ruchhoft, C. C. and Smith, R. S. 1939- "Changes in Characteristics of 
 
 Activated Sludge Induced by Variations in Applied Load," Sewage Works 
 Journal 1 1 :409- 
 
 Sawyer, C. N. I966. "Bulking of Activated Sludge and Related Problems," The 
 Activated Sludge Process in Sewage Treatment Theory and Application , 
 University of Michigan Department of Civil Engineering, Feb. 1966. 
 
 Shell, G. L. 1970. Private Communication. 
 
 Smit, J. 1934. "Bulking of Activated Sludge-Causative Organisms," Sewage 
 Works Journal 6:1041 . 
 
 Smith, R. S. and Purdy, W. C. 1936. "The Use of Chlorine for the Correction 
 of Sludge Bulking in the Activated Sludge Process," Sewage Works Journal 
 8:223. 
 
141 
 
 Snell, F. D. and Snell, C. T. 1 96 1 . Colorimetric Methods of Analysis , Vol. 3, 
 D. Van Nostrand Co. Inc., Princeton, New Jersey. 
 
 Standard Methods for the Examination of Water and Wastewater , 1965- 12th 
 Edition, American Public Health Association. 
 
 Symons, J. M. and McKinney, R. E. 1957- "The Biochemistry of Nitrogen in the 
 Synthesis of Activated Sludge," Sewage and Industrial Wastes 30:87**. 
 
 Szaniszlo, P. J. 1968. "Production of a Capsular Polysaccharide by a Marine 
 Filamentous Fungus," Journal of Bacteriology 95:1474. 
 
 Talmage, W. P. and Fitch, E. B. 1955- "Determining Thickener Unit Areas," 
 Industrial and Engineering Chemistry 47:38. 
 
 Tapleshay, J. A. 1 9^5 - "Control of Sludge Index by Chlorination of Return 
 Sludge," Sewage Works Journal 17:1210. 
 
 Takiguchi, Y. 1968. "Studies on Mucilage in Activated Sludge, (I) On 
 Characteristics of the Mucilage and its Production," Journal of 
 Fermentation Technology 46:488. 
 
 Takiguchi, Y. 1968. "Studies on the Mucilage in Activated Sludge, (II) On 
 
 Correlation Between Viscosity of the Mucilage and Settling Characteristics 
 of Activated Sludge," Journal of Fermentation Technology 46:497. 
 
 Tischler, L. F. and Eckenfelder, W. W. 1968. Linear Removal of Simple Organic 
 Compounds in the Activated Sludge Process , The University of Texas Center 
 for Research in Water Resources, Technical Report to the FWPCA, EHE-12- 
 6702 CRWR-22. 
 
 Vesilind, P. A. 1968. "The Influence of Stirring in the Thickening of Biolo- 
 gical Sludge 1 ,' Doctoral Dissertation, University of North Carolina, 
 Chapel Hill , N. C. 
 
 Vrablik, E. R. 1959- "Fundamental Principles of Dissolved Air Flotation of 
 
 Industrial Wastes," Proceedings of the 14th Industrial Waste Conference , 
 Purdue University. 
 
 Walters, C. F. 1966. "Biochemical Storage Phenomena in Activated Sludge," 
 Doctoral Thesis, University of Illinois, Urbana, Illinois. 
 
 Wang, R. C. T. 1967. "A Viscometer for the Study of the Rheology of Activated 
 Sludge," Unpublished Project Report, University of Illinois. 
 
 Weinshank, D. J. and Garver, J. C. 1967- "Theory and Design of Aerobic 
 
 Fermentations," _h^ Peppier, H. J. (ed) Microbial Technology , Reinhold 
 Publ ishing Co. 
 
 Wilkinson, J. F. 1958. "The Extracellular Polysaccharides of Bacteria," 
 Bacteriological Reviews 22:46. 
 
 jWyss, 0. and Morland, E. J. 1968. "Composition of Obligate Hydrocarbon- 
 Utilizing Bacteria," Appl ied Microbiology 16:185 
 
APPENDIX 
 
 Typical calculations of continuous flow flotation unit operating conditions 
 
 Dilution Water Flow Rate, Q = 3,530 ml/min 
 
 Substrate Feed Rate, g = 20 ml/min 
 
 29 5 1 
 Sludge Recycle Rate, Q = J " . = 8.43 ml/min 
 
 Flow Rate to Flotation Unit, Q = 3,530 + 20 + 843 = 4,393 ml/min 
 
 Clarified Liquid Discharge from Flotation Unit, E = 6,000 ml/min 
 
 Liquid Recycle Rate = E - Q = 6,000 - 3,530 = 2,470 ml/min 
 
 Mixed Liquor Suspended Solids = 4.0 g/1 
 
 Gas Release Rate = 42 ml wet gas/1 recycle 
 
 „ , p. A . / , . \ /Std tempwatm pressure - vapor pressurex 
 
 Vol Dry Air = (vol wet a i r) ( — ■-) ( ^— — -. — c c ) 
 
 7 gas temp std atm pressure 
 
 /,- , • /, i w 273° \ /750 mm Hg - 20.4 mm Hgx 
 = (42 ml air/1 recycle) ( — L=L —) (— -,/Z mm .. -) 
 
 295.5 9 
 
 = 37-4 ml dry gas at STP/1 recycle water 
 
 Wt = (0.0374 0(1.2929 g/D 
 
 = 0.0484 g air/1 recycle water 
 
 .,- _ , 0.0484 g air/1 recycle) (2.47 1/min ) 
 7 ~ K (4.393 l/min)(4.0 g/1) 
 
 = 0.0068 g air/g sol ids 
 
 Load inn RatP - (^ 393 1/m? n) (4 g/1 ) (1 440 mi n/day) 
 Loading Rate (454 g/lb)(3.3 sq ft) 
 
 = 16.9 lb/sq ft/day 
 
 142 
 
VITA 
 
 Robert Frank Wood was born September 1 A , 1937, in Eddington, 
 Maine and was graduated from Brewer High School, Brewer, Maine, in 195**. 
 
 He received his B.S. Engr. degree from Walla Walla College, 
 College Place, Washington, in I960. From I960 to 1 965 he was employed 
 as a sanitary engineer with the Division of Sanitation and Engineering 
 of the Oregon State Board of Health in Eugene and Portland, Oregon. 
 
 He received his M.S. degree in Environmental Health Engineering 
 from the University of Texas at Austin in 1966. He was a Federal Water 
 Pollution Control Trainee from I966 to 1970. He will become Assistant 
 Professor of Environmental Health at Loma Linda University in September, 
 1970. 
 
 He is a member of the American Water Works Association and the 
 Water Pollution Control Federation and is a Registered Professional 
 Engineer in Oregon. 
 
 m