EPA-600/2-77-186b
November 1977
Environmental Protection Technology Series
EVALUATION OF LEACHATE TREATMENT
Volume II: Biological and
Physical-Chemical Processes
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-186b
November 1977
EVALUATION OF LEACHATE TREATMENT
Volume II
Biological and Physical-Chemical Processes
by
Edward S. K. Chian
Foppe B. DeWalle
Environmental Engineering
Department of Civil Engineering
University of Illinois at Urbana-Champaign
Urbana, Illinois 61801
Contract No. 68-03-0162
Project Officers
Dirk Brunner
James A. Heidman
Richard A. Carnes
Solid and Hazardous Waste Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
11
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solu-
tion and it involves defining the problem, measuring its impact, and
searching for solutions. The Municipal Environmental Research Laboratory
develops new and improved technology and systems for the prevention, treat-
ment, and management of wastewater and solid and hazardous waste pollutant
discharges from municipal and community sources, for the preservation and
treatment of public drinking water supplies, and to minimize the adverse
economic, social, health, and aesthetic effects of pollution. This publi-
cation is one of the products of that research; a most vital communications
link between the researcher and the user community.
This study involved extensive analysis of different organics and in-
organics present in leachate samples from landfills located in different
regions of the United States, These analysis were then used to predict the
effectiveness of different biological and physical chemical treatment methods
for contaminant removal.
Francis T. Mayo, Director
Municipal Environmental
Research Laboratory
m
-------
ABSTRACT
The efficiencies of three different types of biological treatment units
in removing organic matter from solid waste leachate were extensively
evaluated. The units studied were the anaerobic filter, the aerated
lagoon, and an activated sludge unit treating combined leachate and
municipal sewage. The effectiveness of physical-chemical treatment
steps in further removing organic matter from the biological unit
effluents was also examined.
A completely mixed anaerobic filter, in which the influent is diluted
with recirculated effluent, was studied for a 52-day period and found
to effectively remove organic matter concentrations in high-strength,
solid waste leachate over a range of organic loadings and shockloads.
Recirculation increases the acidic pH of the feed, eliminating the need
to add the costly buffer solutions required in plugflow anaerobic
filters. Ninety-nine percent of the COD removed in the unit was accounted
for in the methane produced. This valuable energy source can be used
for combustion to heat the unit. A microbial solids balance indicated
that only 0.012 grams of VSS was produced per gram of COD removed.
Because of the low solids production and initial seeding of the unit with
digested sludges, it was not necessary to add nutrients. Although a
possible heavy metal toxicity was observed, it was eliminated by adding
sulfide. A high organic matter removal percentage was observed at
hydraulic detention times greater than 7 days, but the percentage was
considerably lower at shorter detention times. Increases in organic
loading had a substantial effect on the relative organic matter composition
of the effluent. Testing of a fixed film biological reactor model showed
that the substrate removal rate is primarily affected by substrate concen-
tration, specific surface area, flow rate, and temperature of the unit.
Evaluation of various physical-chemical treatment methods such as
chemical precipitation, activated carbon adsorption, and chemical
oxidation showed that these methods were not very effective in removing
organic matter from high strength leachate. Although free volatile
iv
-------
fatty acids, present in large quantities in leachate, have a relatively
low activated carbon adsorption capacity, even lower capacities were
noted for the non-fatty acid fraction using batch isotherms. Although
relatively high organic matter removals were obtained using activated
carbon columns, rapid breakthrough occurred after passage of the initial
bed volume of effluent. Activated carbon treatment of the high strengh
leachate was found to be unfeasible because of rapid headless buildup
resulting from the formation of iron precipitates in the carbon column.
These solids are difficult to remove in subsequent backwash operations.
The organic matter removal rate achieved using lime precipitation was
as low as 26% and was realized only at very high dosages.
Substantially higher adsorptive capacities were observed for biologically
pretreated leachate. Removal of biodegradable organics with an anaerobic
filter increased the adsorption capacity by 50%, while aerated lagoon
treatment of the anaerobic filter effluent resulted in an adsorptive
capacity 2.5 times higher than that of untreated leachate. Membrane
fractionation of the anaerobic filter effluent followed by activated
carbon column treatment of each fraction produced relatively low removal
rates for both the high-molecular-weight organics collected in the
18,000 MW UF retentate and the low-molecular-weight organics present in
the 150 MW RO permeate. The highest removals were observed for the
intermediate-molecular-weight fulvie-like organics which were also
characterized by a high aromatic hydroxyl group content. Aerated
lagoon treatment of the anaerobic filter effluent resulted in organic
matter removals as high as 79% using activated carbon columns. This
increased removal rate was shown to result from the higher adsorption
characteristics of the low molecular weight organics present in the 150
MW RO permeate. Anaerobic filter treatment of the leachate resulted in
lower organic matter removal rates than were obtained with untreated
leachate. Aerated lagoon treatment of the anaerobic filter effluent
resulted in a slightly higher TOC removal of efficiency. Both removal
rates, however, were obtained only at very high lime dosages, precluding
the use of this treatment method as a feasible alternative to activated
carbon. It was therefore concluded that physical-chemical treatment
-------
methods are only effective after extensive biological pretreatment Of
all physical-chemical methods tested, activated carbon treatment was
the most effective in removing organic matter.
Studies of the biological aerated lagoon or extended aeration process were
conducted in six completely mixed reactors with no recycle fed with
undiluted leachate and maintained at detention times of 85.7, 60, 30, 15,
and 7 days. These units were operated for periods from 70 to 150 days
The resulting average TOC removal efficiencies varied between 99% and '
96.8% at organic loadings ranging from 0.224 kg/m3.day (14 Ib TOC/1000
cu ft-day) to 1.65 kg/m3.day (105 Ibs TOC/1000 cu ffday). The phosphate
requirements of the aerobic biomass were extensively evaluated The
30 day aerated lagoon did not show any deterioration when the COD:P
ratio in the influent was increased from 165:1 to 300:1. A significant
increase in effluent TOC values was observed, however, when the ratio
was increased to 800:1. Such low ratios did not affect the 60 day and
the 87.5 day units, which were able to function at a ratio as high as
1540:1; the sludge settling characteristics, however, deteriorated
substantially at this ratio. Cessation of nutrient addition to the
units having relatively low detention times caused an immediate increase
in effluent organic matter concentrations and a decrease in MLVSS.
Calculations using the effluent data and MLVSS values obtained under
optimum conditions indicated that the bacteria yield was 0.42 mg VSS/
mg COD, the respiration rate content was 0.025 day'1, and the overall
first-order substrate removal rate constant was 4.9-10'4 liter/mg VSS-day.
All units showed high heavy metal removal rates, especially for iron
( 99.9%), while low rates were observed for calcium (99.3%) and magnesium
(75.9%). The lowest removal rates were obtained with sodium (24.1%) and
potassium (17.0%).
The settling characteristics of the sludge from the aerated lagoons were
found to be comparable to those of lime softening sludge and primary
sludge, which have higher settling rates than observed for secondary
sludge. The interface settling rates generally decreased during condi-
tions of nutrient limitations. The dewatering characteristics of the
sludge were greatly improved by adding cationic polymers and inorganic
vi
-------
coagulants. An increase of approximately 20 times in specific resistance
was obtained at a polymer dosage of 1.5% (Nalco 73C32) and 2.3% (Primafloc Cy)
and at coagulant dosages of 14.8% (FeCl3-6H20) and 23.7% (Ca[OH]2). These
increases generally corresponded to increases of 6 times in vacuum filler
yields. Analysis of the effluent of the 30 day aerated lagoon showed
that 97.6% of the organics had molecular weights greater than 100, with 33%
in the range from 100 to 500, 41% in the range from 500 to 5000, 20% in
the range from 5000 to 50,000 and 4% greater than 50,000, indicating
that these organics are amenable to reverse osmosis and activated carbon
treatment but not to chemical precipitation.
It was also found that physical-chemical treatment methods are not
effective in removing large quantities of organics from the leachate
and that biological pretreatment is required. Several physical-chemical
methods were therefore tested using the aerated lagoon effluents. While
ozonation removed only 48% of the lagoon effluent TOC after a 3 hour
treatment period, activated carbon columns were able to remove 86% of the
organic matter using an empty bed detention time of 3.7 minutes. A
maximum initial COD removal of 53% was realized with a weak base anion
exchange resin, while 82% and 85% of the COD was removed using strong
base anion exchange resins. Membrane reverse osmosis was the only
process capable of removing 91 to 96% of the salts initially present at
a IDS concentration of 6200 mg/1. The organic matter removal rates using
reverse osmosis ranged from 86% to 97%. These rates were not enhanced
by ion exchange or activated carbon pretreatment. Removal of suspended
solids by sand filtration or chemical precipitation is likely required
for reverse osmosis, activated carbon, or anion exchange resin treatment
of the aerated lagoon effluent.
The combined treatment of leachate and municipal sludge was evaluated
in a conventional plugflow activated sludge unit. It was found that the
test unit could effectively treat the high strength leachate. Directly
after the addition of leachate the effluent BOD and COD concentration
deteriorated somewhat, but after sufficient adaptations the effluent
qualities were generally comparable to that of the control units. While
BOD values were not greatly affected, COD concentrations showed a gradual
vii
-------
increase as more leachate was added. The test unit was not able to treat
the high strength leachate at 4% of the municipal sewage flow rate,
as evidenced by high effluent BOD values and deteriorating sludge condi-
tions. This failure was attributed to limiting phosphate concentrations
in the influent and the relative composition of the soluble high molecular
weight organics which tend to affect the flocculation of the sludge.
This report was submitted in fulfillment of Contract No. 68-03-0162 by
the University of Illinois under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period June 30, 1972 to
November 30, 1974, and work was completed as of September 1976.
vm
-------
TABLE OF CONTENTS
Abstract iv
List of Figures xii
List of Tables ' xviii
Acknowledgments xx
Recomnendations for Further Research 1
I. Treatment of a High Strength Solid Waste Leachate with
a Completely Mixed Anaerobic Filter
Conclusions 2
Introduction 4
Observed substrate removal in fixed film reactors 7
Simulation of substrate removal in biofilms 10
Materials and methods 15
Results and discussion 18
Start-up of the anaerobic filter 18
Recirculation ratio 21
Shockloading 23
Gas production 28
Heavy metal toxicity 28
Nutrient requirements 32
Factors affecting effluent organic matter 34
Effect of removal rate on organic matter 35
Organic matter composition 39
Sludge production 44
Effect of pH on gas production 46
Gas composition 49
Effluent buffer capacity 51
Kinetics of substrate removal 51
Specific surface area 55
Temperature 57
Mass transfer 57
References 58
II. Physical-Chemical Treatment of Leachate and Anaerobic
Filter Effluent
Conclusions 62
Introduction 64
Materials and methods 67
Results and discussion 69
References 97
III. Treatment of a High Strength Solid Waste Leachate with
the Aerated Lagoon
Conclusions 98
Introduction 99
Aerobic biological treatment of leachate 101
Kinetic considerations in aerobic biological treatment 103
IX
-------
Materials and methods 107
Lysimeter and solid waste leachate 107
Aerated lagoon 107
Sludge settling and dewatering methods 112
Analytical procedures 115
Results and discussion 117
Aerated lagoon treatment of leachate 117
Sludge settling and dewatering characteristics 132
Effluent organic matter characteristics 137
References 146
IV. Physical-Chemical Treatment of Leachate and Aerated
Lagoon Effluent
Conclusions 147
Introduction 148
Physical-chemical treatment of leachate and
biological effluents 149
Removal of organics by ozone 151
Removal of organics by activated carbon 153
Removal of organics by anion exchange resins 154
Removal of organics by reverse osmosis 156
Materials and methods 157
Results and discussion 163
Removal of organics in leachate by reverse osmosis 163
Organic removal in aerated lagoon effluent by
ozonation 165
Organic removal in aerated lagoon effluent by
activated carbon 168
Organic removal in aerated lagoon effluent by
ion exchange 168
Organic removal in aerated lagoon effluent by
reverse osmosis 179
References 185
V. Combined Treatment of Leachate and Municipal Sewage in
an Activated Sludge Unit
Conclusions "87
Introduction 188
Activated sludge processes 189
Combined treatment of high strength waste and
municipal sewage 192
Materials and methods 197
Results and discussion 206
Sewage analysis 206
Hydraulic flow regimen of the aeration unit 206
Evaluation of leachate additions to the
activated sludge unit 209
Evaluation of leachate sludge additions to the
anaerobic digester 226
References 231
-------
VI. Estimated Costs for Leachate Treatment
Conclusions 233
Introduction 234
Methods and procedures 236
Activated sludge 236
Aerated lagoon 237
Anaerobic filter 238
Slow sand filtration 239
Activated carbon 239
Reverse osmosis 240
Results and discussion 241
References 244
XI
-------
LIST OF FIGURES
Number
1. The Completely Mixed Anaerobic Filter 16
2. The Startup of the Anaerobic Filter 20
3. Determination of the Minimum Required Recirculation Ratio 22
4. Stability of the Anaerobic System Under Different Shockloads 24
5. Effluent Quality During Phase II as Measured by COD, Fatty
Acids and Gas Production 25
6. Effluent Quality During Phase II as Measured by Carbo-
hydrates, Aromatic Hydroxyls, Color, ORP, Conductivity
and Inorganic Carbon 26
7. Effluent Quality During Phase II as Measured by Heavy
Metals, Suspended Solids, Phosphorus, pH and Alkalinity 27
8. Effect of Increasing Hydraulic Loadings on Percentage
Removal of Different Heavy Metals 31
9. Effect of Increasing Hydraulic Loadings on the Metal
Concentrations in the Sludge Collected from the Bottom
of the Anaerobic Filter 33
10. Effluent Quality During Phase III as Measured by COD, Fatty
Acids, Gas Production, Carbohydrates, Aromatic Hydroxyls
and Alkalinity 37
11. Effluent Quality During Phase III as Measured by Conduc-
tivity, pH, Alkalinity, Inorganic Carbon and Phosphorus 38
12. Percentage Organic Matter Removal at Different Hydraulic
Detention Times 40
13. Dilute-Out Curves for Color, Total Phosphorus and
Suspended Solids 41
14. Effect of Rate of Substrate Removal on Relative Organic
Matter Composition of the Effluent 42
15. Effect of Effluent pH on Corresponding Rate of Gas
Production 47
16. Effect of Effluent pH on Corresponding Relative Rate of
Gas Production as Observed by Several Investigators 48
xn
-------
Number Paqe
17. Effect of Rate of Substrate Removal on Gas Composition 50
18. Relation Between pH of Inflection Point and Fatty Acid
Concentration 52
19. Effect of Substrate Removal Rate on Effluent Concentration
as Measured in Different Studies Using Various Wastes 53
20. Effect of Specific Surface Area, on Substrate Removal
Coefficient 56
21. Activated Carbon Adsorption Isotherms with Different
Leachate Samples 70
22. Activated Carbon Adsorption Isotherms with a Diluted
Leachate Sample and a Fatty Acid Mixture 71
23. Activated Carbon Adsorption Isotherms with Fatty Acid
Mixture 73
24. Breakthrough of Effluent TOC in Diluted Leachate and Acid
Mixture Passed Through Activated Carbon Columns 75
25. Headless Build-up During Passage of Diluted Leachate and
Pretreated Anaerobic Filter Effluent Through Activated
Carbon Columns 77
26. Removal of Accumulated Iron and Organic Matter During
Backwashing of the Activated Carbon Column 79
27. Activated Carbon Adsorptive Capacities of Leachate After
Biological Pretreatment with the Anaerobic Filter and the
Anaerobic Filter Followed by Aerated Lagoon 81
28. Ratio of Absorbance to COD and Ratio of COD to TOC at
Decreasing COD Concentrations, Corresponding to Increasing
Activated Carbon Dosages Added to Anaerobic Filter
Effluent 33
29. Ratio of Aromatic Hydroxyls to COD and Ratio of Carbo-
hydrates to COD at Decreasing COD Concentrations,
corresponding to Increasing Activated Carbon Dosages
Added to the Anaerobic Filter Effluent 84
30. Changes of Different Parameters in the Filtered Mixed
Liquor Occurring During Aeration of Anaerobic Filter
Effluent 37
31. Changes of Different Parameters in the Filtered Mixed
Liquor Occurring During Aeration of Anaerobic Filter Effluent 88
xiii
-------
Number Page
32. Sephadex Eluate of the Organic Matter in the Anaerobic
Filter Effluent Before and After Aeration 90
33. Breakthrough of COD, Color and Turbidity in Activated
Carbon Effluent During Passage of A) Unfractionated
Anaerobic Filter Effluent B) 18,000 MW UF Retentate of A
C) UF Permeate and RO Retentate of A D) UF and RO
Permeate of A. 92
34. Breakthrough of COD, Color, Turbidity in Activated Carbon
Effluent During Passage of A) Unfractionated Aerated
Anaerobic Filter Effluent B) 18,000 MW UF Retentate of A
C) UF Permeate and RO Retentate of A D) UF and RO
Permeate of A. 93
35. Effect of Lime Precipitation Treatment of Anaerobic Filter
Effluent on Different Parameters Measured in the
Supernatant 95
36. Effect of Lime Precipitation Treatment of Aerated Anaerobic
Filter Effluent as Different Parameters Measured in the
Supernatant 96
37. Mixed Liquor Volatile Suspended Solids Concentration in
Aerated Lagoons 1, 2 and 3 Treating Leachate 118
38. Total Organic Carbon in Effluent of Aerated Lagoons
1, 2 and 3 Treating Leachate 119
39. Effect of Reduction of Daily Phosphate on Total-P and
COD/TOC Ratio in Effluent of Aerated Lagoons 1, 2 and
3 Treating Leachate 120
40. Mixed Liquor Volatile Suspended Solids Concentration in
Aerated Lagoons 4, 5 and 6 Treating Leachate 122
41. Total Organic Carbon and COD/TOC Ratio in Effluent of
Aerated Lagoons 4, 5 and 6 Treating Leachate 123
42. The Calculation of Substrate Removal Rate Constant Based
on TOC Data 128
43. The Calculation of Substrate Removal Rate Constant Based
on COD Data 129
44. The Calculation of Growth-Yield and Microorganism-Decay
Coefficients Based on TOC Data 130
45. The Calculation of Growth-Yield and Microorganism-Decay
Coefficients Based on COD Data 131
xiv
-------
Number Page
46. Comparison of Settling Velocities of Sludges from Aerated
Lagoons Treating Leachate with Other Sludges 133
47. Effect of Omission of Nutrient Addition of Settling
Velocities of Sludges from Extended Aeration Units 4,
5 and 6 Treating Leachate 134
48. Effect of Chemical Doses on Specific Resistance of
Sludge from Aerated Lagoon 4 Treating Leachate 135
49. Effect of Chemical Doses on Filter Yield of Slduge from
Aerated Lagoon 4 Treating Leachate 139
50. Elution Profile of the NS-100 Membrane Retentate on a
G-75 Sephadex Column as Characterized by Total Organic
Carbon 140
51. Elution Profile of the NS-100 Membrane Retentate on a
G-25 Sephadex Column as Characterized by Total Organic
Carbon 141
52. Visible Spectrum of Effluent from Aerated Lagoon Unit
4 Treating Leachate 143
53. Ultraviolet Spectrum of Effluent from Aerated Lagoon
Unit 4 Treating Leachate 144
54. Absorbance at 400 NM of a Serial Dilution of Effluent
From Aerated Lagoon Unit 4 Treating Leachate 145
55. TOC Decrease of Ozonated Effluent from Aerated Lagoon
3 Treating Leachate 166
56. Results of Aerobic Biological Polish of Ozonated Effluent
from Aerated Lagoon 3 Treating Leachate 169
57. Activated Carbon Breakthrough Curve for Effluent from
Aerated Lagoon 4 Treating Leachate, at a Flowrate of
1.76 cm/min 170
58. Activated Carbon Breakthrough Curve for Effluent from
Aerated Lagoon 4 Treating Leachate, at a Flowrate of
0.35 cm/min 171
59. Duolite A-7 Breakthrough Curve for Effluent from Aerated
Lagoon 4 Treating Leachate, at a Flowrate of 1.76 cm/min 172
60. Duolite A-7 Breakthrough Curve for Acidified Effluent
from Aerated Lagoon 4 Treating Leachate at a Flowrate of
1.76 cm/min 173
xv
-------
Number Pa9e
61. Duolite A-7 Breakthrough Curve for Acidified Effluent
from Aerated Lagoon 4 Treating Leachate, at a Flowrate
of 0.35 cm/min 176
62. Amberlite IRA-938 Breakthrough Curve for Effluent from
Aerated Lagoons 4 Treating Leachate at a flowrate of
0.35 cm/min 177
63. Amberlite XE-279HP Breakthrough Curve for Effluent from
Aerated Lagoon 4 Treating Leachate at a Flowrate of
0.35 cm/min 178
64. Activated Carbon Breakthrough Curve for RO NS-100 Membrane
Permeate from Effluent of Aerated Lagoon 4 Treating
Leachate at a flowrate of 0.35 cm/min 183
65. Sanitary Landfill Leachate Treatment Schematic Diagram 184
66. The Hydraulic Flow System of the Plugflow Activated-
Sludge Pilot Plant 198
67. Airflow System for the Activated-Sludge Pilot Plant 199
68. Leachate Feeding and Automatic Sampling Systems 201
69. Electrical Systems for the Activated-Sludge Units 202
70. Schematic of the Anaerobic Digesters 203
71. The Distribution of BODs Concentration in the Daily
Composite Sewage Samples 207
72. The Flow Pattern of the Designed Aeration Tank 208
73. Influent and Effluent COD and BOD Concentration of the
Control and Test Unit Receiving 0.5 Percent by Volume
of Leachate 211
74. The Effect of 1% Leachate Addition on Effluent Quality
of the Activated-Sludge Process 213
75. The Effect of 2% Leachate Addition on Effluent Quality
of the Activated-Sludge Process 214
76. The Effect of 3% Leachate Addition on Effluent Quality
of the Activated-Sludge Process 215
77. The Effect of 4% Leachate Addition on Effluent Quality
of the Activated-Sludge Process 216
xvi
-------
Number Page
78. The Effect of Leachate Additions on the Concentrations
of P04-3-P and N(Kj) in the Effluent of the test Unit 217
79. The Effect of Increased Leachate Additions on the
Influent BOD/P Ratio and the Effluent COD and 8005
Concentration of the Test Unit 219
80. Effluent Quality During 2% Leachate Addition at the
0.6 F/M Ratio 220
81. Effluent Quality During 2% Leachate Addition at the
1.0 F/M Ratio 221
82. The Effect of Leachate Addition on the Sludge Settling
Characteristics 222
83. Elution Profile of the 500 MW UF Retentate of the Control
Unit on a G75 Sephadex Column as Characterized by TOC,
Carbohydrates and Carbonyl Groups 224
84. Elution Profile of the 500 MW UF Retentate of the Test
Unit Receiving 0.5% Leachate on a G-75 Sephadex Column
as Characterized by TOC Carbohydrates and Carbonyl Groups 225
85. Adsorptive Character of the Soluble Organic Matter in the
Effluent of the Test and the Control Unit Both Operated
at F/M Ratio's of 0.3 and 0.6 227
86. Daily Gas Production from the Control- and Test Anaerobic
Digester Treating Waste-Activated Sludge 228
xvi i
-------
LIST OF TABLES
Number Page
1 Different Phases and Experimental Conditions During the
Anaerobic Filter Study 19
2 Measured Adsorptive Capacities for Leachate and Biologically
Pretreated Leachate Using Adsorption Isotherms 86
3 Physical Characteristics of Solid Waste Placed in the
Lysimeter Used for Leachate Generation 108
4 Analysis of Leachate Collected from Solid Waste Lysimeter 109
5 Operating Parameters of the Aerated Lagoons 111
6 Concentration of Heavy Metals in Effluent of Aerated
Lagoons 125
7 Characteristics of Effluent from Aerated Lagoons with
Sufficient Nutrient Addition 126
8 Kinetic Constants of Aerated Lagoon Treatment of Leachate
with Sufficient Nutrient Addition 127
9 Results of Buchner-Funner Test Using Mixed Liquor From
The Aerated Lagoon 4 136
10 Results of Filter-Leaf Test Using Mixed Liquor From The
Aerated Lagoon 4 138
11 Characteristics of Ion Exchange Resins Used to Treat the
Aerated Lagoon Effluent 159
12 Treatment of Leachate by KP-98 and NS-100 Membranes at
50% Product Water Recovery 164
13 Determination of Various Coefficients for Ozonating Aerated
Lagoon Effluent of Unit 3 167
14 Removal of Organics by Resins and Activated Carbon 180
15 Removal of Organics and Salts with the NS-100 Membrane
from Effluent of the Aerated Lagoon 4, and from Effluent
of the Activated Carbon and Ion Exchange Columns 181
16 Required Reactor Volume for Completely Mixed Reactor as
Compared with a Plugflow Reactor at Various Removal
Efficiencies 191
1\7. Composition of Leachate Samples Collected from Eleven
Different Sources 195
xviii
-------
Number
Page
18 The Amount of Air Supplied to the Laboratory Activated
Sludge Units as Compared with the Amount Supplied to
Actual Sewage Treatment Plants 210
19 Anaerobic Digestion of the Waste-Activated Sludge from
the Leachate Treatment 229
20. A Summary of Cost Estimates for Leachate Treatment 242
xix
-------
ACKNOWLEDGEMENTS
The chemical analysis and treatment studies were performed by Richard
L. Davison, Laboratory Chemist, and the Research Assistants J. T. Y.
Chu, Y. Chang, G. Velioglu, and F. M. Saunders. Invaluable help was
also provided by Laboratory Assistants B. MacPherson, J. Hansen,
C. Stroupe, T. Brozozowski, M. Sweeny, P. J. Strange, B. Clar, G. S.
M. Yi, J. Young, J. Hunsicker, S. Schoemaker and D. Ellis
xx
-------
RECOMMENDATIONS FOR FURTHER RESEARCH
Biological methods of high-strength leachate treatment which were not evaluated
because of budget constraints included the fixed media unit (trickling filter
and rotating disk), the anaerobic lagoon, and modified soil treatment system
currently being tested at the University of Illinois. It is therefore recom-
mended that these systems be studied,on a laboratory scale before being
applied at the demonstration level.
Much research is needed to optimize the design and operation of anaerobic
filters. The type of solid media to which the bacteria are attached should
first be studied in a series of parallel tests using various soft materials
and shapes. The substrate removal efficiency over a range of loadings should
be evaluated and the amount of solids leaving the unit through the effluent
should be determined. The second phase should evaluate different feed con-
figurations, such as conventional plugflow feeding, step feeding, and tapered
step feeding and the location of the recirculation outflow. Different filter
shapes (heights and widths), all having the same volume, should be tested for
a given volume, since a low but long unit is less expensive than a low but
wide unit, and the amount of substrate removed is likely to be higher in the
latter unit. The third phase of the study should focus on operational modes
of the filter, and the effect of different recirculation flowrates and oper-
ating temperatures should be tested in parallel.
An extensive study should be conducted to determine when leachates are suf-
ficiently biologically stabilized to be effectively treated by physical-
chemical methods. The study should combine an in-depth organic chemical analy-
sis of different leachate samples with a limited treatability study.
Finally, research should be undertaken to evaluate combined physical-chemical
and biological methods, such as biological aerated lagoons to which powdered
carbon is added or activated carbon upflow columns in which a biomass is
maintained. The application of such treatment methods could enhance the life
and flexibility of leachate treatment units already constructed and in operation.
-------
I
TREATMENT OF A HIGH STRENGTH SOLID WASTE LEACHATE WITH
A COMPLETELY MIXED ANAEROBIC FILTER
CONCLUSIONS
A completely mixed anaerobic filter, in which the influence organic matter
concentration is diluted with recirculated effluent, was found to effec-
tively remove organic matter concentrations in high strength municipal
solid waste leachate at a range of organic loadings and shock!oads.
Recirculation can effectively increase the acidic pH of the feed to a
pH value close to the optimum, thus .eliminating the addition of costly
buffer solutions as is required for the plugflow anaerobic filter. In
the anaerobic filter, complex organics in the feed are hydrolyzed, first
by acid fermenting bacteria to free volatile fatty acids, primarily
acetic-and butyric acid, which in turn are removed by methane fermenting
bacteria and converted to methane (CH4) and carbon dioxide (COJ. The
methane in the generated gas accounted for 93 percent of the COD removal
in the unit, while a solids balance indicated that only 0.012 g VSS was
produced per g COD removed. Due to the low solids production and the
initial seeding of the unit was digested sludge no nutrient additions
were required during the 518 day period even though the COD:P and COD:N
ratios in the feed were as high as 4360:1 and 39:1, respectively.
Although a possible heavy metal toxicity, likely the result of high
copper concentrations, was observed in the unit, this was eliminated
after sulfide addition.
A high percentage of organic matter removal was observed when the hydraulic
detention time was maintained above 7 days, but, showed considerably lower
removal percentages below this detention time. Increases in organic loading
had a substantial effect on the relative organic matter composition of the
effluent and the magnitude of the fatty acid fraction showed a pattern
inverse to that of the nitrogenous organics. A fixed film model was formu-
-------
lated which indicated that at high substrate concentration the substrate
removal rate is proportional to the square root of the substrate concen-
tration and the specific area of the filter medium. A comparison of the
biofilm model and the measured effluent concentrations tended to indicate
that the substrate removal rate is primarily affected by substrate concen-
tration, specific surface area, flow rate and temperature of the unit.
-------
INTRODUCTION
It has been recognized that biological removal of organics in high strength
wastewaters is most economically realized in anaerobic biological systems,
since these systems do not have the high energy needs associated with
aeration as required in aerobic biological systems (Metcalf and Eddy,
1972). As the cell yield in anaerobic systems is about ten to twenty
times lower than that in aerobic systems, the costs of sludge disposal
are greatly reduced. During anaerobic fermentation the acid-fermenting
bacteria degrade the complex waste to free volatile fatty acids, primarily
acetic and propionic acid, which are subsequently converted by the methane
fermenting bacteria into methane and carbon dioxide. The generated gas
can be combusted to yield valuable energy.
A major disadvantage of the anaerobic fermentation is the sensitivity of
the anaerobic bacteria which are inhibited at acidic pH values. In addition,
the acid- and methane-fermenting bacteria can experience heavy metal
toxicities. Such inhibitions will reduce the biological growth rate and
result in their subsequent wash-out. An ideal process is therefore one
which is able to retain biological solids independent of the waste flow
and maintain sufficient high solids concentration for long periods of time.
Such an objective is realized by employing a solid media in the unit to
which bacteria can attach. Coulter zt at. (1954) for example, employed
an anaerobic filter filled with rock media to retain solids from an
anaerobic digester effluent. Although the majority of the BOD was removed
in the preceeding digester, additional removal was observed in the filter.
Young and McCarty (1967, 1968, 1969) developed the upflow anaerobic filter
in which the anaerobic bacteria are present in a film attached to a rock
medium to remove organics in the waste flowing upward through the column.
Attachment to the medium results in sludge ages of more than 600 days
(Young and McCarty, 1968, 1969) and 150 days (Plummer
-------
since 1876 to purify sewage but the organic removal was mainly thought to
be due to adsorption (Truesdale ut at., 1961). The removal of organics in
septic tank effluent by bacteria attached to soil particles is based on a
similar principle.
The removal efficiencies of the anaerobic filter can be higher than those
obtained with anaerobic digesters operated at the same volumetric loading.
A comparison of the filter with a digester both maintained at a loading of
1.28 kg COD/m3 day (80 Ib COD/1000 cu ft day).showed that the effluent
COD of the anaerobic filter was only one fifth of that of the digester
effluent (Foree and Reid, 1973). Tadman (1973) noted that an aerobic
filter produced an effluent concentration one tenth of that of an anaerobic
digester maintained at the same hydraulic retention time of 2 days.
Although Chian and DeWalle (1976) showed that the COD removal in the
filter was slightly less than observed in aerated lagoons treating the
same waste, the color removal and suspended solids removals, however, are
comparable. Moreover, the installation and operating costs of the
anaerobic filter are about half of that of an aerated lagoon (Pailthorp
et at., 1971).
In a plug flow reactor, the pH decreases initially as a result of the
acid fermentation, and then increases in the direction of the flow, due
to the biological removal of the generated fatty acids, formation of
ammonia and reduction of sulfates. Since the acidic pH in the bottom
section of the filter can potentially inhibit the methane fermenting
bacteria, substantial amounts of buffer solutions are added to the
influent waste stream to prevent such pH decrease. El-Shafie and
Bloodgood (1973), for example, added 8,000 mg/£ NaHC03 to the influent
having a COD of 11,800 mg/l in order to maintain a pH above 6.5. Young
and McCarty (1968, 1969) added NaHC03 at a concentration of about half
of the COD content of the waste, whle Taylor (1972) added 1,000 mg/£
NaHC03 at an influent concentration of 8,800 mg/l COD.
-------
A completely mixed anaerobic filter would not experience the pH decrease
observed in plug flow units, since the mixing maintains a fairly uniform
pH throughout the depth of the filter. This in turn would eliminate the
need for adding costly buffer solutions. In a completely mixed filter,
a major part of the effluent is returned and mixed with the influent stream.
If the effluent has a sufficient bicarbonate buffer capacity, it is even
able to neutralize feed solutions with an acidic pH. Since all of the
previous studies with anaerobic filters used once through plug-flow
reactors, the present study was conducted to evaluate a completely mixed
anaerobic filter.
The high strength acidic wastewater evaluated in this study was a solid
waste leachate containing free volatile fatty acids and complex high-
molecular-weight carbohydrate-like organics (DeWalle and Chian, 1974) to
give a COD of 54,000 mg/l and a pH of 5.4. The fatty acids represented
49 percent of the total COD, while 1610 mg/t carbohydrate, 605 mg/£
tannins and 1270 mg/t proteins were also detected. Such leachate is.
generated in solid waste landfills when infiltrating rainwater allows
degradation of primarily cellulosic materials. Collection and treatment
of such leachate alleviates the potential problem for groundwater
pollution. Due to leaching of metal objects, relatively large amounts
of potentially toxic heavy metals were also detected in the leachate
used in the present study. The observed concentrations were 2,200 mg/t
Fe, 104 mg/t Zn, 18 mg/£ Cr, 13 mg/t Ni and 0.5 mg/t Cu.
-------
OBSERVED SUBSTRATE REMOVAL IN FIXED FILM REACTORS
An engineered design of the anaerobic filter requires a knowledge of the
substrate removal kinetics in fixed film biological reactors. Such infor-
mation is obtained when the influent substrate concentration and the
hydraulic detention time are varied independently of each other while
the system response is studied. Schulze (1960) noted that the percentage
of BOD removal did not change with the increasing influent concentration
when an aerobic fixed film reactor was maintained at a constant hydraulic
loading. However, when the hydraulic loading was increased and the
influent concentration maintained, the percentage removal decreased,
indicating that the percentage removal was solely determined by the
detention time in the filter. The detention time is a function of hydaulic
loading and depth of the filter. The observed results thus follow the
first order substrate removal kinetics.
Rinke and Wolters (1970) also varied influent concentration and hydraulic
loading independently of each other in aerobic trickling filters and noted
that, at a constant hydraulic loading, an increase in influent substrate
concentration corresponded to a constant percentage of removal at low
organic loadings but showed a decreasing percentage of removal at high
loadings. Maintaining a constant organic loading by increasing the
hydraulic loading and simultaneously decreasing the influent concentration
resulted in a constant amount of substrate removed and a constant percentage
substrate removal. Cook and Kincannon (1970) showed, in contrast to
Schulze (1960), that at a constant hydraulic loading an increase in
influent concentration resulted in a lower percentage substrate removal.
Similar to Rinke and Wolters (1970) they noted that at constant organic
loading the percentage of removal did not change when the hydraulic
loading was increased and the influent concentration was decreased
simultaneously. Based on these data one would conclude that recircula-
tion of the effluent to reduce the influent concentration would be
-------
beneficial since this would result in a same percentage COD removal with
respect to the diluted influent. When the substrate concentration of the
effluent is small as compared to the influent concentration, recircula-
tion would then result in a lower actual effluent concentration.
Less variable results have been obtained in evaluating anaerobic filters.
Young and McCarty (1968, 1969) showed that increasing the influent concen-
tration at constant but low hydraulic loading resulted in a constant
percentage removal but showed a decreasing percentage at high hydraulic
loadings. Maintaining the organic loading while decreasing both the
influent concentration and the hydraulic detention time resulted in a
constant percentage removal at low organic loadings and a decreasing per-
centage at high organic loadings. The actual effluent concentration,
however, was only determined by the organic loading and was independent
of influent concentration. Caudill (1968) similarly noted that the COD
removal efficiency in an anaerobic filter treating potato starch waste
remained at 77 percent when the influent concentration increased from
500 to 1000 mg/COD at constant hydraulic loading. El-Shafie and Bloodgood
(1973) also observed a constant percentage removal independent of the
influent concentration at a given hydraulic loading. Jennett and Dennis
(1975) found that at a constant organic loading of 3.5 kg COD/m3 day
(220 Ib COD/1000 cu ft day) the percentage removal decreased from 98
percent to 95 percent and 94 percent respectively when the influent
concentration decreased from 16,000 mg/l to 8,000 mg/l and 4,000 mq/Jt.
The observation that the actual effluent concentration also decreased,
would indicate that recirculation could result in slightly lower effluent
concentrations.
The above considerations are important in evaluating the effect of recir-
culation on substrate removal efficiency as practices in the present
study. Based on fundamental considerations, both Schulze (1960) and
Germain (1964) concluded that recirculation would have no effect on
organic matter removal. However, both Eckenfelder (1961) and Gallar
and Gotaas (1964), using a large number of actual aerobic trickling
8
-------
filter plant data, concluded that recirculation resulted in a larger
percentage substrate removal possibly due to the inoculation of the
incoming sewage with bacteria present in the effluent. Several studies
compared treatment efficiency before and after recirculation was prac-
ticed and generally noted an improvement in effluent quality. Moore
&t at. (1950) noted that a recirculation ratio of one improved the BOD
removal from 85 percent to 93 percent at an influent BOD of 240 mg/£.
Using a similar recirculation ratio Hanumanula (1969) noted that the
BOD removal increased from 62 percent to 91 percent at an influent BOD
of 330 mg/£. Oleszkiewics (1974) showed that recirculation is especially
beneficial at high influent substrate concentrations due to better
aeration of the liquid film and the elimination of substrate inhibition.
He observed an improvement of the removal from 30 percent to 85 percent
at a recirculation ratio of 5 using an influent COD of 3,000 mg/l.
-------
SIMULATION OF SUBSTRATE REMOVAL IN BIOFILMS
The substrate removal kinetics in fixed film biological systems can either
be described in terms of substrate concentrations existing in the bulk
liquid phase or simulated substrate concentrations in each layer of the
biofilms, the sum of which predicts the response of the entire film.
Kornegay and Andrews (1969) and Young and McCarty (1968, 1969) used the
former approach; the latter authors introduced a substrate gradient
factor defined as the ratio of the concentration at the liquid-biofilm
interface to the effective concentration in the biofilm. As a result,
the calculated apparent half velocity concentrations are several times
larger than would be obtained if all bacteria were complete dispersed.
Recent studies however, generally use the second approach in which the
process of substrate diffusion and bacterial uptake are simulated for
each of the successive layers of the biofilm; the present study used
a similar approach.
In a biological film reactor at steady state, a mass balance within the
film can be made by equating:
input - output = uptake (1)
According to Pick's law of molecular diffusion the mass transfer rate dF/dt
through a surface area A is proportional to the concentration gradient of
the substrate S at the interface:
f - - A Ds if (z)
j\C
TT- = rate of mass transfer at interface (mass/time)
ot
2
D = diffusion coefficient (length /time)
gC A
|^-= substrate concentration gradient (mass/length )
aZ
z = depth of biofilm starting from liquid-film interface (length)
2
A = surface area of biofilm (length )
10
-------
The biological uptake within the biological layer can be expressed as:
dSz
~dt
USZX
dSz 3
-jfi = rate of substrate uptake (mass/length time)
U = maximum substrate removal rate ( mass of substrate }
vmass of biomass -time'
Sz = substrate concentration at depth z (mass/length3)
X = biomass concentration within the biolayer (mass/length )
o
K$ = half velocity concentration (mass/length )
By making a material balance between a differential distance within the
biolayer using equations (1), (2) and (3), it results in the following
equation for a unit cross sectional area:
6\_ USZX
which states that the second derivative of the substrate concentration with
respect to z is dependent on both the substrate and biomass concentration.
j\C
When axial dispersion becomes important the term v-^-will have to be added
oZ
to the right hand side of Equation (4). However, since the filter media
in the present study consists of long curved channels, this term can be
omitted.
Equation (4) is a second order non-linear differential equation. As such
it does not have a simple solution. However the equation can be solved
when Sz is either much larger or much smaller than K$ (Hang and McCarty,
1971). When S2 is much larger than K Equation (4) reduces to a zero
order kinetic equation;
.
.20
dz s
11
-------
which can be integrated for a unit cross sectional area, and the mass
transfer rate at the interface becomes:
(6)
h = thickness of biofilm (length)
Since for a given waste, U and X are not expected to vary greatly, equation
(6) would indicate that the rate is independent of substrate concentration
but directly proportional to surface area and thickness of the biolayer.
In such a situation the decrease of substrate within the biofilm is small
as compared to the concentration at the interface so that the rate of
substrate utilization per unit of biomass is approximately the same as if
the solids were evenly dispersed throughout the liquid. According to Pirt
(1973) and Saunders and Bazin (1973) the thickness of the biolayer can be
approximated by
2D -S.
h= -or1 (7)
o
$b = substrate concentration in bulk liquid (mass/length )
However, equation (7) was derived based on the boundary conditions of zero
substrate concentration at the interface of the biolayer and the solid
support; it therefore violates the constraint that S is much larger than
K and is therefore only an approximation. Since D , U, and X are not
J O
expected to vary to a great extent for a given substrate, Equation (6)
can therefore be further modified to indicate that the rate of substrate
removal per unit reactor volume is related to the specific surface area,
A/V, and the square root of the bulk substrate concentration:
dF - k, 4' SK (8)
Vdt n V °b
= coefficient based on zero order kinetics (mass*/length*-time)
o
V = reactor volume (length )
12
-------
An equation similar to equation (8) was also used by DeWalle and Chian (1976)
to describe the rate of biological regeneration of activated carbon particles
due to the presence of activated sludge solids. It is realized that at very
high substrate concentrations the rate of substrate removal will be indepen-
dent of the substrate concentrations due to the zero order kinetics of the
reaction. In that case the rate of substrate removal will only be dependent
on the specific surface area in the reactor.
When Sz is much smaller than KS> equations (4) reduces to a first order
kinetic equation:
(9)
Which after integration yields the mass transfer rate equation at the
interface:
£•»> °f
Since DS, U, X and K$ are not expected to vary greatly with a given sub-
strate, Equation (10) reduces to
dF A
= k''S
VdT 27b
k2 = coefficient based on first order kinetics (length/time)
The equation indicates that the rate of substrate removal per unit reactor
volume is related to the specific surface area of the biofilm and the bulk
substrate concentration, but is independent of film thickness. A similar
equation has been given by Atkinson and Daoud (1968) for large film thick-
ness. Equations (8) and (11) should be used in small successive sections
in plug flow anaerobic filter columns. However, since such detailed data
are generally not available, both equations can be used with respect to
the whole column while $b is substituted by the effluent concentration of
the column.
13
-------
Additional resistance to mass transfer from the bulk liquid to the liquid-
biofilm interface is caused by the liquid film adjacent to the biofilm.
The thickness of this layer can be described by the Nusselt (1916) equation:
•? n V3
5 - Crrft (12)
Q = flow rate (Iength3/time)
v = viscosity (mass/time'length)
p = density of liquid (mass/length3)
g = gravitational constant (length/time2)
W = width of film (length)
6 = thickness of film (length)
Since the mass transfer rate decreases with increasing liquid film thickness
in the laminar sublayer, the flux of substrate would also be proportional
to the flow rate to the power 1/3. This agrees with the analysis by Takeshi
at at. (1972) that the mass transfer coefficient KL for laminar flow can be
expressed as:
KL = kQ1/3 for Re < 10 and (13)
kL = kQ1/2 for Re > 10 (14)
Both equations (8) and (11) illustrate the importance of the specific
surface area, A/V, of the fixed film reactor. Based on similar considera-
tions Ames et at. (1962) stated that the substrate removal was proportional
to the specific surface area of the media. Lamb and Owen (1970 came to
the identical conclusion. For the above reasons, a filter medium with a
relatively high specific surface area of 206 m2/m3 (63 ft2/ft3) was
selected for the present study.
14
-------
MATERIALS AND METHODS
The laboratory scale anaerobic filter column was constructed of Plexiglas
with an overall height of 246 cm and 20.2 cm OD (Figure 1). The height
of the filter medium in the column was 199 cm and comprises a volume of
54.6 liter (1.93 cu ft). A head space of 1.4 liter is provided at the
top of the medium while the inlet section is 2.8 liter. A solids collec-
tion device, installed at the end of the second period in Phase II, is
located at the bottom of the column and has a volume of 3.8 liter. A
recirculation surge vessel with a volume of 5.2 liter provides the separa-
tion of the effluent stream and the recirculation stream. The latter is
subsequently used to dilute the incoming leachate and raise its pH. The
total volume of the column is therefore 67.8 liter of which 54.6 liter
or 81 percent contains the filter medium to which the bacteria are
attached.
The medium in the column consists of plastic "Surpac" slabs (Dow Chemical
Midland, MI) while additional plastic strips were placed between each
sheet. The plastic material has a specific density of 1.45 g/cm3. As
the average thickness of the slab is 0.57 mm, on the basis of 100 measure-
ments, the specific surface of the plastic material itself is therefore
3490 m /m3 (1067 ft2/ft3). The specific surface area of the media per
unit column volume is 206 m2/m3 column volume (63 ft2/ft3 column volume);
the total surface area in the column is therefore 11.3 m2 (121.6 ft2)
and only 6 percent of the column volume is taken up by the plastic
material resulting in a porosity of 94 percent. A comparable study by
Young and McCarty (1968, 1969) used smooth quartzite stones with a
porosity of only 42 percent and a specific surface area of 114 m2/m3
2 3
column volume (37.5 ft /m column volume) as estimated from data provided
by Truesdale vt at. (1961).
The leachate employed in this study was obtained from a lysimeter filled
with solid waste to which simulated rainwater was added (Chian and DeWalle
1976). After collection from the lysimeter, the leachate was refrigerated
15
-------
V= 1.4
15.1cm
E
o
E
u
CVI
Recirculation
Pump
20.2 cm t
_ 18.7cm
Gas Collection Line
Wet Gas Meter
/-as o
Inflow-
\
J1\
Outflow
Recirculation
Vessel
-Filter Media
H i i u u -=r
Sludge T IM^3-^ I
**~ c I ,—11—=-J \ IL,
10.2 cm
28cm
Solids Collection
Figure 1. The Completely Mixed Anaerobic Filter
16
-------
and fed from a 10 liter container into the anaerobic filter using an
electrolytic gas cell, in which the 02 and HZ generated were used to
expand a bladder located in the leachate container to displace its con-
tent. Due to the formation of iron hydroxide solids, the leachate
could not be fed with regular positive displacement pumps without
excessive wearing.
17
-------
RESULTS AND DISCUSSION
Studies on the effectiveness of the anaerobic filter in treating high
strength leachate from the lysimeter were conducted for a period of 518
days starting July 3, 1973. During Phase I of the study, which lasted
218 days the different start-up procedures and pH stabilities of the
unit were tested, while in Phase II which lasted 250 days, the various
operational difficulties of the completely mixed unit were evaluated.
Phase III, which lasted 50 days, was used to study the effluent charac-
teristics as affected by organic loading. A summary of the different
steps is given in Table 1.
Start-up of the Anaerobic Filter
The start-up of the anaerobic filter was tested during the early stage
of the Phase I study. It was noted that when the undiluted leachate
was passed through the unit without seeding no significant biological
degradation would occur for a 33 day period when the unit was operated
at a detention time of 42 days. Also, no biomass would develop on the
filter media, even when the pH of the incoming leachate was adjusted with
sodium hydroxide from 5.8 to 7.0, i.e., the minimum required pH for
methane fermentation. The filter was therefore emptied and the bottom
section was seeded with 1 liter of digested sludge containing a total
25 g solids while the column was filled with the leachate diluted to
one tenth of its strength with distilled water. Thereafter the unit was
fed 1.6 liter of the diluted leachate to give a detention time of 42
days. A high initial gas production was observed which gradually decreased
(Figure 2a). Increasing the influent concentration by using a lower
dilution ratio resulted in increasing amounts of gas being generated in
the unit as a result of the microbial methane fermentation. However,
when the undiluted leachate was fed, inhibition was observed as the gas
production ceased completely. A larger quantity of 2 liter of digested
sludge, containing 50 g solids, and a less rapid increase in the influent
concentration, however, resulted in a satisfactory gas production,
approaching the theoretical amount as calculated from the COD removal
18
-------
Table 1. Different Phases and Experimental Conditions During the Anaerobic Filter Study
Duration Duration
Phase (days) Period (days)
I 218 la
Ib
2a
2b
3a
3b
3c
II 250 la
Ib
Ic
Id
2
3
4
III 50 1
2
3
1-33
33-64(31)
64-103(39)
103-146(43)
146-188(42)
188-200(12)
200-218(18)
218-291(73)
291-293(2)
293-298(5)
298-362(64)
362-398(36)
398-432(35)
432-467(35)
467-497(30)
497-512(15)
512-517(5)
Hydraulic
Detention
Time
(days)
42
42
42
42
42
42/7
42/4.25
42
42/3.0
42
42
42
42
74
74
17.5
7.5
Leachate
Feeding
Rate
1.6
1.6
1.6
1.6
1.6
1.6/9.7
1.6/16
1.6
1.6/22.7
1.6
1.6
1.6
1.6
0.9
0.9
4.0
9.1
Combi ned
Influent
Flow Rate
32.4
32.4
32.4
32.4
32.4
32.4/40.5
32.4/46.8
32.4
32.4/53.5
32.4
32.4
32.4
32.4
31.7
31.7
34.8
39.9
Leachate
Influent
Recircula- COD
tion Ratio (mg/t)
1:20
1:20
1:20
1:20
1:20
1:20/1:4.2
1:20/1:2.9
1:20
1:20/1:29
1:20
1:20
1:20
1:20
1:35
1:35
1:8.7
1:4.4
62,000
62,000
6,200-62,000
6,200-31,000
27 ,000
27 ,000
27,000
27,000
27,000
27,000
19,500
19,500
19,500
39,000
32,000
32,000
32,000
Experimental
Conditions
no pH adjustment;
no seeding
pH adjustment to
7; no seeding
seeding 1£ sludge;
decreasing dilution
seeding 2JL sludge;
decreasing dilution
1:10 to 1 :2
feed 1:2 diluted
1 eac ha te
first shockload
second shockload
start of extensive
monitoring
Third shockload
extensive monitoring
gradual toxicity,
Na2S
solids collection
device installed
between period 2
and 3
doubling of leachate
concentration
kinetic analysis
kinetic analysis
kinetic analysis
-------
50
40
x>
$30
C
o
T3
I 20
10
C
o
40
30
o
•o
S 20
w
o
o
No Dilution
•Added 1.0
Digested
Sludge
B
Added 2jO J?
Digested
Sludge
1
1
0 10 20 30 40 50
Time (Days)
Figure 2. The Startup of the Anaerobic Filter
20
-------
(Figure 2b), after 43 days. In order to prevent possible further inibi-
tion, a new batch of leachate was fed to the unit at a 1:2 dilution
(1 part of feed to 1 part of dilution water) resulting in an influent
COD of 27,000 mg/l. Similarly to the present study Young and McCarty
(1968, 1969) noted that about 40 days of operation were required to
reach equilibrium conditions when two additions of 30 g volatile solids
from a sludge digester were used to start the filter. Adding 30 g of
digested sludge to the bottom section of an aerobic filter, Jennett and
Dennis (1975) noted that only 20 days were required to reach equilibrium
conditions.
Recirculation Ratio
After a 42 day operating period, equivalent to one volume turnover, the
required recirculation ratio was studied. Approximately 63 percent of
the constituents in the initial leachate batch would be displaced after
one volume turn over for a completely mixed reactor. Initially, a 1:20
recirculation ratio was arbitrarily selected at the beginning of the
study. A recirculation ratio of 1:20 was defined as 1 part of feed to
19 parts of effluent to results in a total volume of 20 parts. A
titration curve of the effluent as shown in Figure 3a, indicates that the
highest buffering capacity of the effluent is observed at a pH of 6.2,
shile less is available at the actual effluent pH of 7.2. The pH of the
inflection point at 3.9 reflects the presence of both bicarbonates and
fatty acids. If only bicarbonate had been present, the pH of the inflec-
tion point would have been close to 4.3. Because of this observed shape
of the effluent titration curve, five parts of effluent are required to
increase the pH of the influent to 7.0, thus, requiring a minimum ratio
of 1:6 (Figure 3b). In order to operate the unit safely and well-mixed
the ratio of 1:20 was maintained by using a recirculation flow rate of
30.8 £/d, resulting in a complete mixing of the unit every 1.8 days as
calculated for the column volume subject to mixing. Since the time
required for mixing was short as compared to the hydraulic detention
time, i.e., 1.8 days versus 42 days, the unit can be considered completely
mixed. As a result of the recirculation, the vertical upflow rate in the
unit was 1.1 m/d.
21
-------
8
6
pH 5
4
Amount Of Meq Of H2 S04 Added To
50ml Of Anaerobic Filter Effluent For
The Alkalinity Determination
7.2
7.0
6.8
pH 6.6
6.4
6.2
6.0
5.8
B
2 3
Meq. Of Acid Added
Amount Of Effluent Of Anaerobic
Filter Necessary To Increase The pH
Of 20ml Of Leachate
L I I I I I I
20 40 60 80 100 120
Amount Of Effluent (ml)
140 160 180 200
Figure 3. Determination of the Minimum Required Recirculation Ratio
22
-------
Shock!oading
Before a more extensive monitoring was initiated, the ability of the
anaerobic filter to withstand shockloads was studied. When the detention
time, based on feed stream flow alone, was reduced from 42 days to 7 days
for a one-day period, only a small change in the pH was observed (Figure
4a and b). Since the recirculation flowrate was kept constant, the
recirculation ratio experienced a decrease. The pH decreased to 6.9
when the detention time was reduced to 4.25 days for a one-day period
(Figure 4c and d). A larger shockload was conducted in Phase II, 73
days after the more intensive monitoring of the unit had started. The
detention time was reduced to 3 days for a two day duration which also
caused a similar lowering of pH values (Figure 4e and f). The gas produc-
tion did not show a corresponding increase, as high concentration of
organics were present in the effluent of the unit (Figure 5a) which
reduced the organic removal efficiency to 54 percent. The high COD
values corresponded with high concentrations of fatty acids (Figure 5b),
aromatic hydroxyls and carbohydrates (Figure 6a). The large differences
in the values for the filtered and unfiltered COD may indicate that
substantial quantities of solids were resuspended at the higher flow
rates. This was also indicated by the suspended solids (SS) analysis
(Figure 7b). As most of the phosphates in the unit are present as
suspended solids, a similar pattern as the SS was observed (Figure 7c).
Since fatty acids contribute to the alkalinity, an increase parallel to
the fatty acids was observed (Figure 7e). After the detention time was
restored back to 42 days after the two days shockload, the effluent COD
and SS concentrations returned to values only slightly higher than those
observed before the shockload. It is realized that at short detention
times the flow regime of the unit tends to depart from a completely mixed
pattern and approaches that of a plugflow, since at a three day deten-
tion time, the content is only mixed 2.7 times during one volume turnover.
Based on these tests it was concluded that the buffer capacity of the unit
is sufficient to prevent large pH depressions at relatively short deten-
tion times. However, at detention times as short as three days a large
portion of the organics leave the unit in the effluent stream. Further-
23
-------
150
— 100
TJ
e
Q.
at
o
O
X
Q.
50
7.2
7.1
7.0
6.9
6.8
T
o>
I*
-------
•*" "• ^0.45,1 Filtered
COD
100
Time, Days
200
Figure 5. Effluent Quality During Phase II as Measured by
COD, Fatty Acids and Gas Production
25
-------
200
Time, days
Figure 6. Effluent Quality During Phase II as Measured by
Carbohydrates, Aromatic Hydroxyls, Color,
ORP, Conductivity and Inorganic Carbon
26
-------
100
Time, days
200
Figure 7. Effluent Quality During Phase II as Measured
by Heavy Metals, Suspended Solids, Phosphorus,
pH and Alkalinity
27
-------
more high concentration of suspended solids are detected in the effluent
at such short detention times.
Gas Production
High organic matter removals were observed during the first 60 days of
monitoring in the Phase II period and 97 percent of the COD was removed
at a rate of 0.62 kg COD/m3 day (39 Ib COD/1000 cu ft day) (Figure 5a).
As the oxygen equivalent of 1 mole of CH4 is 202, 22.4 liter of methane
are therefore produced at 0°C and standard pressure for each 64.4 g COD
removed. Thus, 1 g of COD removed produces 0.38 liter of methane gas at
25°C. The average gas production during that period was 23.3 III influent,
and the gas contained 78 percent methane. This corresponds to 89 percent
of the theoretical gas production as calculated from total COD removals.
The difference of 11 percent was attributed to the small amount of
biological solids produced in the unit and organics removed by physical
processes. A theoretical carbon balance calculation showed that 1.75
liter of gas is produced per gram of carbon removed from solution. The
leachate and effluent obtained in the first phase of the study had COD/TOC
ratios of 3.5 and 2.8, respectivley, while the inorganic carbon concentra-
tion present in the influent and effluent was 10 mg/£ and 750 mg/£,
respectively. This amount of carbon removed from solution would theoret-
ically correspond to a gas production of 23.2 Jilt of influent which was
99 percent of the amount actually measured.
Heavy Metal Toxicity
Continuing observation of the filter after a 100 day period showed a
gradual deterioration of the effluent COD (Figure 5a). This was sub-
sequently reflected in a higher fatty acid (Figure 5b), carbohydrate and
aromatic hydroxyl concentrations (Figure 6a). After 144 days, the COD
removal decreased to as low as 64 percent, while the pH decreased below
6.8 (Figure 7d) which effectively reduced the methane fermentation
(Figure 5c). Analysis of the effluent showed a gradual increase in
soluble (0.45 y filtered) heavy metals with maximum Fe, Cu and Zn con-
centrations of 2.8, 0.9 and 0.2 mg/£, respectively (Figure 7a), indicating
28
-------
that heavy metal toxicity may have inhibited the anaerobic fermentation.
The metals were mostly present in the suspended form and the dissolved
iron, for example, represented only 9 percent of the total iron content.
Lawrence and McCarty (1965) and Mosey zt at. (1971) showed that the con-
centration of soluble heavy metals was determined by the amount of sulfide
present in the unit to form insoluble sulfide precipitates. Since iron
is the metal generally present in highest concentrations, the ferrous
sulfides in the unit constitute a sulfide reservoir to precipitate the
more toxic and less soluble heavy metals such as copper, zinc and nickel.
Based on the measurement of soluble sulfides in digesters, Mosey and
Hughes (1975) concluded that a 50 percent decrease in gas production would
occur at concentrations of 10"4 mg/l Zn++, 10"7 mg/l Cd++, 10"12 mg/£ Cu+
and 10~ mg/l Cu . Inhibition during shockloads of short duration occurs
at much higher soluble heavy metal concentrations of 0.5 mg/£ Zn++, 1 mg/l
Cd++, 10"7 mg/l Cu+ and 10"8 mg/l Cu++. Kirsch and Sykes (1971) reported
that at a total zinc concentration of 100 mg/l and a soluble concentra-
tion of 1.5 mg/l, gas production was inhibited. However, failure did
not occur when the zinc was fed as zinc sulfate as a result of which the
soluble zinc concentration never exceed 0.4 mg/£. None of the studies
noted a toxicity due to iron.
Since the present study noted significant concentration of copper, one of
the most toxic elements, the observed inhibition could well be due to this
element. Similar to Lawrence and McCarty (1965) it was noted that the
initial deterioration of the effluent COD was not caused by a higher fatty
acid concentration, indicating that the acid fermenting bacteria were more
affected by the inhibiting substance than the methane bacteria. A
limited number of analyses indicated that approximately half of the
sulfate concentration of 550 mg/l, in the 1:2 diluted leachate was
removed in the unit by reduction and precipitation. Stoichometric
amounts of sulfide would only precipitate 9 percent of the iron removed
in the unit, indicating that the majority of the iron at neutral pH
values would be precipitated as carbonates and hydroxides; a conclusion
similar to that reached by Mosey and Hughes (1975). Unlike the metal-
29
-------
sulfide precipitates, the carbonate precipitates are very pH sensitive
and dissolve at low pH values. This may explain the increase in Zn and
Cu concentrations between 120 and 144 days (Figure 7a). These results
may indicate that initially the acid fermenting bacteria were partially
inhibited by metals resulting in higher effluent COD (Figure 6a). This
was subsequently followed by inhibition of the methane bacteria, as a
result of which the fatty acid concentration gradually increased (Figure
5b). Since less bicarbonate alkalinity was formed, the pH in the unit
started to decrease, which in turn dissolved additional heavy metal
precipitates; the solubilized metals when resulted in further inhibition.
The results further show that the values of pH and conductivity reflect
increases in the fatty acid concentration. The earliest warning of
possible metal toxicity, however, is given by a gradual decrease in
total gas production and an increase of the effluent COD. To reduce the
observed toxicity half of the column volume was drained and filled with
tap water to reduce the inhibiting substances. Since metal toxicity was
suspended, their concentration was decreased by addition of 75 mg/£ Na?S
to the column content, while a similar concentration was added to the
influent for the next 18 days. After these operational changes the
observed effluent quality during the second monitoring period of Phase
II improved greatly while the gas production, the organic matter removal,
and the bicarbonate alkalinity increased gradually. The sulfide addi-
tions caused a decrease in ORP values while reducing the heavy metal con-
centrations (Figure 6c). Furthermore, no H2S odor was detected also
indicating that the toxicity was caused by metals.
In subsequent studies with the same anaerobic filter using a different
feed, experiments were made to study the effect of hydraulic loading on
the removal of heavy metals. Four different loadings were evaluated in
duplicate during which the metal concentrations in samples withdrawn
from the different sampling ports along the filter were measured. Using
the effluent and influent concentrations, the removal percentages can be
calculated. The analytical results in Figure 8 show that the removal of
most metals is 68 percent or higher, except for copper and cadmium. The
30
-------
too
90
80
70
o 60
o
0)
a 50
-------
percentage removal generally decreases with increasing flowrates with the
exception of chromium, which element is likely present as the chromate
anion and not as the cation. Iron showed the highest removals while it
also had the highest initial concentrations. Cadmium had the lowest
removals while it also had the lowest initial concentration. The most
rapid decrease in percentage removal with increasing feed flowrates was
observed for copper. Thus at high loadings copper is removed to a lesser
extent than the other metals, again indicating that this element may
cause a metal toxicity in the anaerobic filter. Although the percentage
metal removal decreases at increasing flowrates, the absolute quantity
of metals increases with increasing flowrates. Since a large portion
of the metals form insoluble sulfide, carbonate and hydroxide precipitates,
which coagulate and settle out in the unit, increased metal concentrations
were observed in the sludge drawn from the bottom portion of the filter
(Figure 9).
Nutrient Requirements
Comparison of effluent and influent total phosphorus (P) showed that 55
percent of the concentration was removed in the unit. The COD:P ratio of
the influent, however, was as high as 4360:1 indicating that a phosphorus
limitation could occur. Young and McCarty (1968, 1969) selected a 470:1
ratio using a protein-carbonate waste and a fatty acid waste. McCarty
and Speece (1963) measured that a feed COD:P ratio of as high as 2200:1
was feasible in an anaerobic digester fed with fatty acids. Force and
Lovan (1971) maintained a 500:1 ratio in their unit treating press
brewery liquor. The latter investigators noted that about half of the
applied phosphorus was utilized in the unit. In a subsequent investiga-
tion using the same unit they maintained a COD:P ratio of 1000:1 using a
wastestream similar to the one used in the present study (Foree and Reid,
1973). They noted that 96 percent of the applied phosphorus was removed,
indicating that only slightly higher ratios are feasible. Although the
waste in the present study was fed to the units without nutrient suppli-
ments, a significant amount of phosphorus of 1328 mg was initially intro-
duced in the unit through the anaerobic sludge used to feed the unit.
32
-------
10,000
1000
o»
o»
•o
V)
_«
£ o
c
o
•^ o>
P o
0)
o
o
"5
100
10
1.0
O.I
^'
.**"
t*' *' -.--'"*'
^s^^r^"^^
S^ l^*
•°'oL
8
10 12 14 16
Flow Rate, Vday
Figure 9. Effect of Increasing Hydraulic Loadings on the Metal
Concentration in the Sludge Collected from the Bottom of
the Anaerobic Filter
33
-------
The COD:P ratio of 4360:1 should therefore be considered as an upper limit.
The COD:N ratio of the feed was 39:1 indicating that no nitrogen limita-
tion would occur. When the phosphates were distinguished in total, soluble
and orthophosphates, it was "noted that about three fourths was present in
the suspended solids form (Figure 7c). The total phosphate concentration,
therefore parallels the suspended solids measurements, as observed during
the evaluation of the third shockload. During the toxicity period a
significant decrease was observed in total phosphate concentration,
probably as a result of the formation of metal phosphate precipitates.
After addition of 75 mg/l Na2S, higher soluble phosphate concentrations
were observed, possibly because less metals were available to form
precipitates with the phosphates; the total effluent P concentration,
however, did not change substantially.
Factors Affecting Effluent Organic Matter Concentrations
It was noted that after addition of the sulfides, the differences between
total COD and the COD in filtration of 0.45 pm Millipore filter became
more substantial. As it was felt that the higher organic solids content
could be due to insufficient settling within the unit, a solids collec-
tion compartment of 3.8 liter was added to the column below the influent
inlet compartment (Figure 1). Visual observations showed that solids
were collected in this compartment to gradually build up a sludge slurry.
Comparison of the effluent data showed both a lower total COD and soluble
COD in period 3 as compared to period 2 probably as a result of the solids
collection device (Figure 5a). The differences in suspended solids con-
centration between the two periods, however, were not noticeable.
In order to determine whether the effluent concentration is solely deter-
mined by organic loading, and not by influent concentration and hydraulic
detention time, the organic loading was maintained constant, while the
influent concentration was doubled, by feeding the unit with undiluted
leachate instead of the 1:2 diluted leachate. Simultaneously the deten-
tion time was doubled to maintain the constant loading. The results in
Figure 5, 6 and 7 indicate a temporary deteriorating effluent quality of
34
-------
COD, fatty acids and carbohydrates. Both pH and bicarbonate alkalinity
decreased during this period but returned to initial values after a 15-
day period. After this initial adaption, effluent qualities were gener-
ally similar for period 3 and 4, indicating that the influent concentra-
tion does not affect the effluent quality under the conditions studied;
however, it does affect the percentage of COD removal. Similar conclu-
sion was reached by Young and McCarty (1968, 1969). They noted that when
the organic loading was maintained while both the influent concentration
and the hydraulic detention time were decreased, the unit gave a constant
percentage COD removal at low loadings but a decreasing percentage at
high loadings. Jennett and Dennis (1975) found that at a constant removal
rate of 3.5 kg COD/m3 day (220 Ib COD/1000 cu ft day) the percentage
removal decreased from 98 percent to 95 percent and 94 percent respectively,
when the influent concentration decreased from 16,000 mg/£, 8000 mg/£ and
4000 mg/£. Contrary to the results in the present study, they also noted
that the actual effluent concentration decreased. Based on these results
one would conclude that recirculation does not affect the effluent quality
as the latter is only determined by the substrate removal rate. Recircula-
tion, however, is necessary to maintain sufficient high pH values to treat
the acidic feed. A completely mixed regime would allow utilization of the
entire length of the column for substrate removal; in plug flow units,
however, only the lower one meter (3.3 ft) is effective (Young and McCarty
1968, 1969; Jennett and Dennis, 1975). As such, increasing the height of
the completely mixed unit with respect to the area at the base will reduce
the total column requirement for the same degree of treatment and, there-
fore, lower construction costs (Germain, 1964).
Effect of Removal Rate on Organic Matter
Since the effluent organic matter concentration is mainly determined by the
substrate removal rate of the unit and not by influent concentration or
hydraulic detention time, the third phase of the study was conducted to
establish such relationships. During the initial 30 days of the third
phase, the detention time was similar to the fourth period of the second
phase, i.e. 74 days. Although slightly less than one volume turn-over was
35
-------
realized, analysis of the effluent quality indicated that steady state
operations were realized (Figure 10 and 11).
When the detention time of the unit was reduced to 17.5 days, higher
effluent concentrations were generally observed (Figure 10 and 11). An
increase was observed for both the biodegradable fatty acids and for the
less readily degradable carbohydrates and aromatic hydroxyls. Reduction
of the detention time to 7.5 days caused a further increase in organic
matter concentration in the effluent. As the quantity of available
leachate was limited, only a 0.7 volume turnover was realized at this
loading. The effluent parameters, with exception of the carbohydrates,
however, indicated that steady-state operations were obtained. Both pH
and inorganic carbon concentrations showed that the unit was effectively
operating at this detention time. The phosphate analyses also indicate
that this nutritional element is not limiting the biological growth at
higher rates of substrate removal. Although a larger amount of organic
matter is removed at the higher loading, the resulting higher organic
matter content in the effluent is a major disadvantage. A material
balance during Phase III indicated that 93 percent of the COD could be
accounted for by the methane gas leaving the unit. The effluent concen-
tration generally reached new equilibrium levels after about 2 days after
a change in the loading. Young and McCarty (1968, 1969) similarly noted
that when the loading to the anaerobic filter was increased, the gas
production rate increased within two days. Jennett and Dennis (1975)
found that immediately after a loading change the effluent COD concentra-
tion increased temporarily but returned to new steady-state levels after
about 10 days. They also noted that the temporary increase in effluent
COD was more noticeable when the loading increase resulted from a higher
influent concentration than from a decrease in detention time. This
tends to agree with the results obtained in this study.
Although the actual effluent concentration is determined by the rate of
substrate removal, the percentage of removal, however, is determined by
the hydraulic detention time. The removal percentage for the completely >
36
-------
3000
Detention Time
74 days
20 30
Time, days
Figure 10. Effluent Quality During Phase III as Measured by COD,
Fatty Acids, Gas Production, Carbohydrates, Aromatic
Hydroxyls and Alkalinity
37
-------
o
u
S
Figure 11
20 30
Time , days
Effluent Quality During Phase III as Measured by Conductivity,
pH, Alkalinity, Inorganic Carbon and Phosphorus
38
-------
mixed unit was therefore calculated during the three detention times of
74, 17.5 and 7.4 days in Phase III and the 3 day detention time during
the shockloading in Phase II (Figure 12). In addition, the percentage
removal as predicted for plugflow units by Young and McCarty (1968,
1969) is shown for comparison. They formulate that the efficiency was
inversely related to the hydraulic detention time. The results of the
two systems indicate that at long and short detention times reasonable
agreement is observed. At intermediate ranges, however, the observed
results showed higher percentages of removal for the unit in the present
study. The data also indicate that the unit should be operated above
a detention time of 7 days in order to obtain more than 95 percent COD
removal.
At the end of the third phase, tap water was introduced in the unit to
study whether the mixing regime as a result of the recirculation agreed
with the completely mixed model. The results of the color dilute-out
clearly indicate that this is still true after accumulation of apprec-
iable amount of biomass, since 63.2 percent is removed after one volume
turnover (Figure 13a). The initial section of the total P curve (Figure
13b) shows an identical result, but shows higher removals after about
one volume turnover, possibly as a result of precipitation under the
more aerobic conditions resulting from the tap water additions. The
suspended solids show the highest decrease since their removal is also
aided by sedimentation. A lower dilute-out is observed after about 70
percent of the suspended solids are removed indicating that the remaining
fraction consists of smaller particles.
Effluent Organic Matter Composition
An evaluation of the relative compositions of the effluent organic matter
at the three loadings is shown in Figure 14. The fourth data point was
obtained during Phase I when the unit was subject to a shock!oad in which
the retention time was reduced to three days for a two day duration. The
results in Figure 14 show that with increasing loadings a larger percentage
of the organic matter consisted of free volatile fatty acids (Figure 14)
39
-------
100
90
^ 80
o
I 70
-------
Total-P Dilute-Out
Suspended
Solids Dilute-Out
I 2
Volume Turnovers
Figure 13. Dilute-Out Curves for Color, Total Phosphorus and Suspended Solids
41
-------
I
50
40
30
0)
81
1 20
o
£
10
0
o
"c
0)
CL
Percentage of Effluent Organic
Matter Consisting of
O Fatty Acids
D Proteins
A COD/TOC Ratio
H 1 1 1
Percentage of Effluent Organic
Matter Consisting of
O Carbohydrates
A Aromatic Hydroxyls
_L
I 2 3 4 5 6
Rate of Substrate Removal (kg/m3 day)
o
o
o
o
o
Figure 14. Effect of Rate of Substrate Removal on Relative Organic
Matter Composition of the Effluent
42
-------
assuming a COD/weight ratio of 1.07. This would indicate that the methane
fermenting bacteria are determining the overall substrate removal rates.
Young and McCarty (1968, 1969) also observed that at low loadings the
percentage fatty acids decreased with decreasing loadings. However, at
removal rates about 1.5 kg COD/m3 day (94 Ib COD/1000 cu ft day) the per-
centage fatty acids decreased with increasing removal rates, possibly
indicating that at this rate the acid fermenting bacteria also become
limiting. A similar optimum at 1 kg COD/m3-day (63 Ib COD/1000 cu ft day)
was obtained with data from Jennett and Dennis (1975); Plummer at at.
(1968) also observed a decreasing percentage fatty acids with increasing
removal rates above a removal rate of 1.5 kg/m3-day. A lower optimum was
of 0.2 kg COD/m day was calculated from data of Arora at at. (1975). This
lower optimum may have been due to the less degradable nature of the sub-
strate as reflected by the low BOD/COD of 0.52. The data, however,
indicate that the above mentioned optimum corresponds to a substrate
removal rate between 1 and 1.5 kg COD/m3-day. The absence of such optimum
in the present study is due to the fact that fatty acids are already
present in the influent stream. The decrease of the percentage fatty
acids at decreasing removal rates is reflected by a decrease of the
COD/TOC ratio's in the effluent, indicating an increase of the oxidation
state of the organic (Figure 14a).
The trend observed for the fatty acids is opposite that of the nitrogenous
organics assuming a COD/N ratio of 10. The highest actual concentration
was obtained at the high loading, which decreased at the intermediate
loading and showed a slight increase at low loadings; the soluble COD
showed a continuous decrease resulting in a gradual increasing percentage
nitrogenous organics. This trend is in agreement with studies by Chian
and DeWalle (1975) who noted an increase in amino acids and nitrogenous
organics after removal of the free volatile fatty acids using an aerobic
system. The data in Figure 14a would indicate that a similar trend is
occurring in anaerobic systems. DeWalle and Chian (1974) noted a sub-
stantial increase of high molecular weight carbohydrates after removal of
both fatty acids and amino acids. The absence of such increase in the
43
-------
present study (Figure 14b) is probably due to the low bacterial growth
rates which are characteristic of anaerobic systems. This was confirmed
in further tests in which the effluent of the anaerobic filter was sub-
sequently treated in an aerated lagoon; after the fatty acids were removed
a substantial increase in high molecular weight carbohydrates was observed.
Sludge Production
A solids balance was made during the third phase of the study. For this
purpose any solids accummulated in the bottom section were removed prior
to Phase III. At the end of the fifty-day period, sufficient solids were
accummulated in the bottom section. A total of 2.5 liters of sludge were
3
removed. The sludge had a density of 1.026 g/cm and contained 3.86
percent solids with a volatile solids/total solids ratio of 0.377 to
give a total of 37.3 g volatile solids. The solids density in the present
study is less than observed by Arora &t at. (1975) who obtained an 8
percent slurry with a volatile solids/total solids ratio of 0.34, but
more than the 2.3 percent observed by Jennett and Dennis (1975). The
calculated average concentration of volatile suspended solids in the
effluent samples was 106 mg/£. As 20 mg/£ was present in the influent
the net accummulation of suspended solids during the fifty-day period
was 11.2 g. Thus, about 67 percent of the recoverable solids generated
in the unit will accumulate in the bottom section and can be removed by
subsequent discharge to sludge drying beds; 33 percent of the recoverable
solids were present in the effluent of the unit. During the fifty-day
period a total of 3992 g of COD was removed, resulting in an apparent net
synthesis of 0.012 g VSS formed/g COD removed. A higher yield would have
been obtained if the accumulation of biological solids onto the plastic
media slabs had been measured. The present value compares favorably with
data calculated by Young and McCarty (1968, 1969) who observed a net
synthesis of 0.015 g VSS/g COD for fatty acid waste and 0.118 g VSS/g
COD for a protein-carbohydrate mixture. Jennett and Dennis (1975) using
a pharmaceutical waste consisting mainly of methanol, observed a yield
of 0.027 g VSS/g COD. The former included in their calculation the solids
that were attached to the filter media, while the latter measured the
44
-------
solids that were loosely attached to the media and could be removed
through the sampling ports. Assuming a VSS/P in the synthesized sludge
of 60:1 (McCarty and Speece, 1963), the maximum COD"P ratio in the feed
should be 5000:1 based on an apparent net synthesis of 0.012 g VSS
formed/g COD removed. This ratio is comparable to that actually
measured in the influent stream.
Several factors have been recognized to determine the amount of solids
leaving the anaerobic filter. Young and McCarty (1968, 1969), observed
a gradual accumulation of solids in the anaerobic filter during which
period the effluent suspended solids remained low. Only after the filter
reached its maximum storage capacity would the effluent solids show an
increase. More than 70 percent of the solids synthesized within the
filter can be retained in the filter when treating a fatty acid waste.
However, with substrates that result in higher yields such as protein-
carbohydrate wastes only 30 percent may be retained (Young and McCarty,
1968, 1969). When the influent waste contains solids, no net removal
may be observed (Jennett and Dennis, 1975). Both Plummer oJt at. (1968)
and Jennett and Dennis (1975) observed that the solids concentration was
mainly determined by the hydraulic detention times. The present data
are in agreement with both studies (Figure 7b). In addition, it is noted
that the porosity of the filter may have a large effect on the solids
concentration which is to be expected since a lower void ratio will
increase the collision frequency between the solids and the filter media.
The suspended solids, for example, were only 10 mg/£ at a void ratio of
0.42 (Young and McCarty, 1968), 34 mg/l at a void ratio of 0-47 (Jennett
and Dennis, 1975), 40 mg/£ at a void ratio of 0.46 (Foree and Reid, 1973),
450 mg/l at a void ratio of 0.85 (Mueller and Mancini, 1975), and 1200
mg/l at a void ratio of 0.68 (Plummer vt at. , 1968). The average sus-
pended solids concentration of 190 mg/l at a void ratio of 0.94 as observed
in this study compares favorably with other data and is probably due to
the detention times which were higher than most other studies.
45
-------
Effect of pH on Gas Production
After termination of the Phase I, II and III studies, all the analytical
data were correlated with each other. When the daily gas production rates
were related to the effluent pH values, the results in Figure 15 indicate
that the highest rates are observed at a pH of 7.1. The relatively low
gas production rates at a pH of less than 6.9 were observed during the
period of heavy metal toxicity. The rate of optimum pH values as observed
in the present study is considerably smaller than other researchers have
found (Figure 16). Speece and Clark (1969), for example, still obtained
half of the optimum gas production rates at a pH value of as low as 5.4,
using anaerobic filters fed with acetate. However, they related gas pro-
duction to the average pH at different depths in the plug-flow unit and a
more narrow band would have been obtained if the pH in the effluent had
been used for such comparison.
Clark and Speece (1970) also summarized existing literature data relating
gas production of anaerobic digesters to corresponding pH values, indicating
that at a pH of 6.0 the gas production was reduced by 50 percent. Using
laboratory digesters, Thiel
-------
150
6.0 6.2 6.4 6,6 6.8 7.0 7.2 7.4 7.6 7.8
PH
Figure 15. Effect of Effluent pH on Corresponding Rate of Gas Production
47
-------
100
i 50
O>
O
s
u
£
0
Speece and Clark
(1969)
Clark and Speece
(1970)
Thiel et al
(1968) ^
I
\ X Heinemann, (1939)
t. / \
8
PH
10
II
Figure 16. Effect of Effluent pH on Corresponding Relative Rate
of Gas Production as Observed by Several Investigators
48
-------
This would only correspond to about 10 mg/£ free ammonia at the pH of the
effluent which is well below the stated toxic level.
Gas Composition
The gas composition measured in the present study was within the range of
values reported by other investigators studying anaerobic filters (Figure
17). The methane content was generally 75 percent but decreased to 63
percent at the highest loading. Using a similar type of wastewater, Foree
and Reid (1973) reported a 70 percent methane content. The decrease of
the methane content at higher rates of substrate removal was also observed
by Thiel
-------
en
O
-------
Effluent Buffer Capacity
It was noted in the previous section that several factors influenced the pH
and buffer capacity of the system. Figure 10 and 11 indicated that the
alkalinity as determined by titration was relatively insensitive to changes
in the bicarbonate concentration as measured with the inorganic carbon
channel of the TOC analyzer. This is a result of the Ionized fatty acids
which register in the alkalinity determination, but are not very effective
in maintaining the pH of the solution due to their low pK values. It was
therefore not surprising to note a general decrease in the inflection point
of the effluent titration curve and the fatty acid concentration in the
effluent (Figure 18). This then shows that the alkalinity as determined
by titration is not a useful monitoring tool to measure the buffer capacity
of the system, and more useful information is obtained when both the pH,
inorganic carbon and fatty acid concentrations are monitored.
Kinetics of Substrate Removal
The kinetic consideration discussed earlier indicate that at relatively low
substrate concentrations the removal rate increases linearly with effluent
concentration. A compilation of the different studies evaluating anaerobic
filters indicate that this is indeed the case (Figure 19). The rate of
substrate removal was calculated as kilograms of COD removed per m3 empty
column volume per day; while the effluent concentration was used as the
corresponding substrate concentration.
Considerable spread in the results of the different studies is apparent
(Figure 19), indicating that other factors such as specific surface area,
biodegradability, temperatures and mixing regime also exert a significant
influence. Figure 19 also indicates that several effluent concentrations
have a finite value when the removal rate approaches zero. The highest
such residual concentration was noted in the present study and extrapola-
tion would indicate a valuesof about 850 mg/l. Four other studies showed
such residual concentration. Further evaluation indicated that its magni-
tude generally increased with increasing influent concentrations. Residual
effluent and related influent concentrations reported by these studies
51
-------
o>
L_
O
g
E
c
£
4.0
.5 3.5
o
a.
3'°50
100 500 1000
Fatty Acid Concentration (mgA)
5,000
Figure 18. Relation Between pH of Inflection Point and Fatty
Acid Concentration
-------
en
CO
o
•o
10
\
o>
o
O)
cr
0>
.0
3
in
rr
7.0
6.0-
5.0 -
4.0-
3.0-
2.0-
1.0
"I^ai f
/
8000
1 1 1
Author Waste
VArora etal(!975) ( I)
BCaudill (1968) (2)
^Speece and Clark(l969) (3)
OEI-Shafie and Bloodgoodll973) (4)
AForee and Reid (1973) (5)
a Foree and Horsley (1972) (6)
OHovious etal(l972) (7)
D Jennett and Dennis (1975) (8)
O Mueller and Mancini (1975) (9)
OPailthorp et al (1971) (2)
' APIummer et al (1968) (10)
A Taylor (1972) (II)
O Young and Me Carty(l968) (9)
A Young and Me Carty (1968) (9)
• Young and Me Carty (1968) (3)
A Young and Me Carty (1968) (3)
D Young and Me Carty (1968) (3)
O Young and Me Carty (1968) (3)
*• Young and Me Carty (1968) (3)
VThis Study
'Strength
(mg/Jj COD)
300-5610
400-1000
6400 -
11,800
12,900
6000-27,000
2000
1000-16,000-
4800
3000
8500
8800
1500 ~
3000
375
750
1500
3000
6000
1
0
500
1000 1500 2000 2500 3000
Soluble Effluent Concentration (mg/j? COD)
3500
4000
4500
Figure 19. Effect of Substrate Removal Rate on Effluent Concentration as
Measured in Different Studies using Various Wastes
-------
were 850 mg/l and 32,000 mg/l (present study), 240 mg/£ and 13,000 mg/£
(Foree and Horsley, 1972), 230 mg/l and 4,800 mg/l (Mueller and Mancini,
1973), 100 mg/l and 500 mg/l (Arora
-------
of Plummer vt at, (1968) gave values of 3.5 kg COD/m3 day and 850 mg/£
respectively using three of the four reported data points. These maximum
rates also tended to increase with biodegradability of the substrate and
q
were 1.8 kg COD/m day with a corresponding BOD/COD ratio of 0.52 (Arora
o£ al., 1975), 3.5 kg COD/m3 day with a BOD/COD ratio of 0.61 (Plummer
vt al., 1968) and 7 kg COD/m3 day with a BOD/COD ratio of 0.65 (El-Shafie
and Bloodgood, 1973). Of all studies, the highest rate was reported by
Q
Mueller and Manzini (1975) who observed a removal rate of 15.6 kg COD/m
day using a biodegradable dextrose-peptone mixture.
Specific Surface Area
Both Equation (8) and (11) indicate that the specific surface is the next
most important factor determining the rate of substrate removal. When the
k2'u-values (in Equation 11) obtained by dividing the rate of substrate
removal by the substrate concentration obtained at low substrate concentra-
tions were related to the specific surface area in each reactor, no
definitive correlation was observed (Figure 20). Excluded from Figure 20
were the data from those studies in which an increase in removal rate did
not correspond to an increased substrate concentration as a result of which
the kp values approached infinity. This was mainly due to the fact that
none of the studies evaluated specific surface areas using the same sub-
strate. Five studies, that evaluated the effect of specific surface areas
using aerobic trickling filter, are included in Figure 20 and all show an
increasing kp value as a result of the increasing surface area as predicted
by Equation (11). The relation tends to be linear at low surface areas but
becomes less than linear at high surface areas. This can be attributed to
the bacterial growth which tends to fill up the void spaces as result of
which the specific surface area at the liquid-biofilm interface becomes
smaller than the specific surface area of the filter medium. Thus, although
insufficient data are available from the different anaerobic filter studies,
all results shown in Figure 20 tend to indicate that the value of k2 would
increase with increasing surface area. As synthetic media having large
surface areas are several times more expensive than the rock media generally
used, selection of the required specific surface area should be based on
economic considerations.
55
-------
o
TJ
tn
O Bruce etal
Aerobic IA Cook and Fleming (1974)
Trickling
-------
Temperature
The effect of temperature on rate of substrate removal was evaluated in only
two studies. Tadman (1973) using an industrial waste and column filled with
Intalox saddles observed a 10 percent increase in removal efficiency when
the temperature was increased from 25°C to 35°C. Caudill (1968) noted a
27 percent improvement in effluent quality when the temperature of an
anaerobic filter unit treating a dilute fatty acid solution was increased
from 26°C to 37°C. A preliminary comparison of heating costs versus the
cost associated with increasing the specific area in the column by installing
more expensive high surface area synthetic filter medium would indicate
that the latter alternative is less costly in order to reach a desired
treatment efficiency.
Mass Transfer
The effect of the flow velocity on the mass transfer in anaerobic filters
as illustrated in Equations (11) (12) and (13) is difficult to evaluate.
The most accurate data to illustrate this effect were obtained from the
study by Young and McCarty (1968, 1969) when for a given flow rate in the
column the substrate removal rate in each 30 cm segment was calculated
and related to the concentration leaving that segment. Such calculations
resulted to the concentration leaving that segment. Such calculations
resulted in relations similar to those shown in Figure 19. The calcula-
A
the k2 77 values showed a linear increase with actual flow rates within the
column when plotted on log-log paper. From the limited data an exponent
of 0.46 was calculated at flow velocities ranging from 1 to 10 m/d, which
approaches the value of 0.5 predicted by Takeshi it at. (1972). Since
the value of k2 is related to K, , this in turn indicates that it is
beneficial for a given column volume to minimize the diameter whil
maximizing the height of the column in order to obtain higher velocities
and thus higher Reynolds numbers within the column.
57
-------
REFERENCES
Ames, W. F. e£ at. , "Transient Operation of the Trickling Filter," JOUA.
Sanit. Engnfi.. ftcv. ASCE, ««, SA3, 21 (1962).
Arora, H. C. &t at., "Treatment of Vegetable Tanning Effluent by the
Anaerobic Contact Filter Process," Wote/L ?oJUtu£. Con&iot (GB) 74,
584 (1975).
Atkinson, B. and Daoud, I. S., "The Analogy Between Microbiological Reactions
and Heterogeneous Catalysis," T>UWA. lvu>t. Cham. Eng*. 46, 19 (1968).
Bruce, A. M. at at., "Research Developments in High Rate Organic Reactions,"
J. IMA*. Pub£. Uaatth Eng*. , 69, 178 (1970).
Caudill, H. , "Application of the Anaerobic Trickling Filter to Domestic
Sewage and Potato Wastes," M.Sc. Thesis, Dept. of Civil Engr. University
of Washington, Seattle, p. 62 (1968).
Chian, E. S. K. and DeWalle, F. B., "Sanitary Landfill Leachates and their
Treatment," J. Env. Engng. V
-------
DeWalle, F. B. and Chian, E. S. K. , "Biological Regeneration of Powdered
Activated Carbon Added to Activated Sludge," (tiateA Ruzasich, in print
DeWalle, F. B. and Chian, E. S. K. , "The Kinetics of Formation of Humic
Substances in Activated Sludge Systems and their Effect on Flocculation ,"
&Lote.c.hnol. and frionngng., 14, 739 (1974).
Eckenfelder, W. W. , "Trickling Filter Design and Performance," JOUA. Santt.
i. Viv. ASCE, 87, SA4 , 33 (1961).
El-Shafie, A. T. and Bloodgood, D. E. , "Anaerobic Treatment in a Multiple
Upflow System," J. Wot. PoUut. ContAol Fed., 45, 2345 (1973).
Foree, E. G. and Lovan, C. R. , "The Anaerobic Filter for the Treatment of
Brewery Press Liquid Waste," Proceed. 26th Induct*. Wa&te. ConteA&nze.,
Purdue University, Engr. Ext. Series 140, 1074 (1971).
Foree, E. G. and Horsley, E. E., "Advanced Studies of the Submerged Anaerobic
Filter from Brewery Press Liquor Stabilization," Dept. of Civil Engr.,
University of Kentucky, Technical Report UKY 48-72-CE, 41 p. (1972).
Foree, E. G. and Reid, V. M. , "Anaerobic Biological Stabilization of
Sanitary Landfill Leachate," Dept. Civil Engineering, University of
Kentucky, Technical Report UKY TR 65-73-CE17, 43 p. (1973).
Galler, W. S. and Gotaas, H. B., "Analysis of Biological Filter," JOUA.
Sanit. Engi. tUu. ASCE, 90, SA6, 59 (1964).
Germain, J. E. , "Economic Treatment of Domestic Waste by Plastic Medium
Trickling Filters," 3ouJi. WateA Poliut. Control Fede/u , 38, 192 (1964).
Hanumanula, V., "Effect of Recirculation on Deep Trickling Filter Performance,"
JOUA. WateA PoZJLut. ContAol FedeA. , 41, 1803 (1969).
Haug, R. T. and McCarty, P. L., "Nitrification with the Submerged Filter,"
DePt. of Civil Engr., Stanford University, Technical Report, 149, 206 p.
\ i y/ 1 ; .
Heukelekian, H. and Heinemann, B., "Studies on the Methane Producing Bacteria.
I. Development of a Method for Enumeration," S&oage. Wolfed JouAwl, 11,
426 (1939).
Hovious, J. C. et tt., "Anaerobic Treatment of Synthetic Organic Wastes,"
Water Poliut. Control Research Series 12020 D 15 01/72, 202 p. (1972).
Jennett, J. A. and Dennis, N. D. , "Anaerobic Filter Treatment of Pharmaceutical
Waste," J. Wat. PoUut. Control Fed., 47, 106 (1975).
Joslin, J. R. vt aJL, , "High Rate Biological Filtration: A Comparative
Assessment," WcuteA ?oU,at. ContAol, 70, 383 (1971).
59
-------
Kirsch, E. J. and Sykes, R. M. , "Anaerobic Digestion in Biological Waste
Treatment," PiogAe44 tnd. M£cAobJ.ol. , 9, 155 (1971).
Kornegay, B. H. and Andrews, J. F. , "Application of the Continuous Culture
Theory to the Trickling Filter Process," Proceed. 24th InduA&t, Wabte.
Ccwjjetenee, Purdue University, Engnr. Extension Series, 735, 1398
(1969).
Lamb, R. and Owen, S. G. G., "A Suggested Formula for the Process of Biological
Filtration," Watvi PoUwt. Control, 69, 209 (1970).
Lawrence, A. W. and McCarty, P. L. , "The Role of Sulfide in Preventing Heavy
Metal Toxicity in Anaerobic Treatment," J. Wat. PoMLut. Con&iol Fed.,
37, 392 (1965).
Levine, M. it at. , "Observation on Ceramic Filter Media and High Rates of
Filtration," Sewage. WoAfci JOUA. S, 701 (1936).
McCarty, P. L. and Speece, R. E. , "Nutrient Requirements in Anaerobic
Digestion," Dept. of Civil Engr., Stanford University, Technical
Report, 25, 115 (1963).
McCarty, P. L. and McKinney, R. E., "Volatile Acid Toxicity and Anaerobic
Digestion," J. Wat. PoUut. Control fed., 33, 223 (1961).
Moore, W. A. et at. , "Efficiency Study of a Recirculating Sewage Filter at
Centralia, Mo," Sewage and lnduu>t>ual Waa-tei, 22, 184 (1950).
Mosey, F. E. &t oJL. , "Factors Affecting the Availability of Heavy Metals
to Inhibit Anaerobic Digestion," J. Wat. Poltut. Control Fed., 74,
18 (1971).
Mosey, F. E. and Hughes, D. A., "The Toxicity of Heavy Metal Ions to
Anaerobic Digestion," J. Wat. Poilut. CoyvUioJi Fed., 74, 18 (1975).
Mueller, J. A. and Mancini, J. L., "Anaerobic Filter Kinetics and Application,"
Proceed. 30th InduAtA.. Wcu>t^ Con^eA-ence, Purdue University, West
Lafayette (1975).
Nusselt, W. , ZCA*. VeAh. Pen*. IYIQ. , 60, 569 (1916).
Oleszkiewics, J. A., "Efficiency of Plastic Media Trickling Filters Operating
Under Extreme Organic Loadings," Proceed. 7tk Int&in. Con^etemie dloutoA
Potiut. Re4eoA.eJt, Paris, Paper 3B (1974).
Pailthorp, R. E. it aJL. , "Anaerobic Secondary Treatment of Potato Process
Waste Water," Paper presented at the 44th Water Pollut. Control Fed.
Annual Conf . , San Francisco (1971).
Pirt, S. J., "A Quantitative Theory of the Action of Microbes Attached to a
Packed Column: Relevant to Trickling Filter Effluent Purification and
to Microbial Action in Soil," J. Appl. Chm. Rto-tecfw. 23, 389 (1973).
60
-------
Plummer, A. H. e£ at. , "Stabilization of a Low Solids Carbohydrate Waste by
an Anaerobic Filter," Proceed 23td InduAtA. Watte. Contf&tenc.e, Purdue
University, Engr. Ext. Series J32, 462 (1968).
Rinke, G. and Wolters, "Technology of Plastic Trickling Filter Media,"
Ptoceed. 5th Int&in. WateA. Potiut, Raaeo/ich Con^etence, San Francisco,
Paper 11-15 (1970).
Saunders, P. T. and Bazin, M. J., "Attachment of Micro Organisms in a Packed
Column: Metabolic Diffusion Through the Microbial Film as a Limiting
Factor," J. App£. Ckw. &U>ti.cA, 2, 393 (1968).
Truesdale, G. A. vt at., "A Comparison of the Behavior of Various Media in
Percolating Filters," J. lvu>t. Pabi. HeotC/i Engu., 60, 273 (1961).
Young, J. C. and McCarty, P. L., "The Anaerobic Filter for Waste Treatment,"
Pn.oc.n2d. 22nd InduAtn. WaAte. Con^e/ience, Purdue University, Engr. Extensio
Series, 729, 559 (1967).
Young, J. C. and McCarty, P. L., "The Anaerobic Filter for Waste Treatment,"
Dept. of Civil Engr., Stanford University, Technical Report 87, 235
(1968).
/oung, J. C. and McCarty, P. L., "The Anaerobic Filter for Waste Treatment,"
J. Wat. PoUat. Contiiot Fed., 41, R160 (1969).
61
-------
II
PHYSICAL CHEMICAL TREATMENT OF LEACHATE AND
ANAEROBIC FILTER EFFLUENT
CONCLUSIONS
It was concluded that high-strength leachates cannot be effectively
treated by physical-chemical treatment methods such as activated carbon
adsorption, chemical precipitation and chemical oxidation. Activated
carbon was observed to have a relatively low sorptive capacity when used
to treat high-strength leachates. Although free volatile fatty acids,
major constituents of such leachates, have a relatively low sorptive
capacity, even lower capacities were noted for the non-fatty-acid frac-
tion in leachate using activated carbon adsorption isotherms. Although
an initial removal rate as high as 72% was obtained in activated carbon
columns through which the diluted leachate was passed, much lower
removal percentages were observed after passage of several bed volumes,
and almost complete breakthrough occurred after 200 bed volumes. The
column studies also illustrated the lower removal rate for the non-
fatty-acid fraction in leachate. Activated carbon treatment of leachate
is also unfeasible because headless builds up rapidly in the column
with the formation of iron precipitates. Although most of these precip-
itates were removed in the first backwash, difficulty was again encountered
in subsequent runs.
Substantially higher adsorptive capacities of activated carbon were found
for biologically pretreated leachate. Removal of biodegradable organics
with an anaerobic filter increased the adsorptive capacity by 50%.
Aerated lagoon treatment of the anaerobic filter effluent further removed
low-molecular-weight organics, resulting in an adsorption capacity of
0.174 mg TOC/mg AC, a value approximately 2.5 times that observed for
untreated leachate. Batch sorption tests showed that color and aromatic
hydroxyls are removed at lower carbon dosages than are high-molecular-
62
-------
weight carbohydrate-like materials. Membrane fractionation of the
anaerobic filter effluent followed by activated carbon column treatment
of each fraction showed relatively low removal rates for both the high-
molecular-weight organics collected in the 18,000 MW UF retentate and
the low-molecular-weight organics present in the 150 MW RO permeate.
The highest removal rates were observed for fulvie-like organics of
intermediate molecular weight. The increase in removal rates of organics
in anaerobic filter effluent treated by aerated lagoon as mentioned
previously was attributed to the higher adsorption characteristic of
the low-molecular-weight organics.
Organic matter removal by lime precipitation, both before and after
aeration, produced removal rates as low as 20 to 25%, and this removal
was obtained at excessively large dosages. Other physical-chemical
methods tested, such as ozonation and chlorination, resulted in similarly
low organic matter removal rates. These results clearly show that
physical-chemical treatment methods are not feasible for high-strength
leachates and that extensive biological pretreatment is required. Of
all physical-chemical methods tested, activated carbon treatment pro-
duced the highest organic matter removal rates.
63
-------
INTRODUCTION
Solid waste leachate generated as a result of infiltrating rainwater
contains high concentrations of organic matter. These substances can
be removed by physical-chemical or biological treatment processes.
After analyzing leachate samples from landfills of different ages,
Chian and DeWalle (1975) noted a change in organic matter composition
with increasing age of the fill: the percentage of free volatile fatty
acids decreased, while the amount of refractory fulvic-like organics
increased with age. Since free volatile fatty acids, di- and tricarboxylic
acids and alcohols are degradable by aerobic and anaerobic bacteria, it
was recommended that aerated lagoons and anaerobic filters be employed
to remove organics from landfills which have generated leachate only
recently. Physical-chemical methods such as lime coagulation, activated
carbon adsorption and ozonation are generally less effective in removing
such organics. Refractory organics generally formed by bacterial or
chemical processes after removal of the degradable substrate, on the
other hand, are more amenable to removal by physical-chemical processes
then by bacterial processes.
The above recommendations have been borne out by several studies on the
treatability of leachate. While Boyle and Ham (1974) noted a 93% COD
removal in anaerobic digestors, Ho et^ aj_. (1974), using a similar leachate,
obtained only a 34% COD removal rate using activated carbon at a maximum
dosage of 20,000 mg/1. In subsequent column studies with granular
Filtrasorb 400, maximum COD removal rates of up to 59% were observed
at detention times greater than 20 minutes, based on analysis of the
filtered second bed volume of leachate passed through the column. A pH
reduction of the leachate to reduce the ionization of the dissociated
free volatile fatty acids and to enhance their adsorption did not result
in increased COD removal; in fact,the opposite was observed, with high
COD removals corresponding to high pH. A similar comparison showed that
anaerobic digestors were able to remove 99% of the leachate COD, while
64
-------
activated carbon at a dosage of 160,000 mg/1 removed only 60%. The
leachate studied was generated from a relatively recent fill, and 65%
of the 2000 mg/1 of TOC consisted of free volatile fatty acids. The
study further noted that the non-fatty-acid fraction of the organic
matter was adsorbed preferentially to the fatty-acid fraction. After
biological treatment of a similar leachate, Pohland and Kang (1975)
obtained removal rates as high as 91%, indicating that biological treat-
ment followed by carbon adsorption will result in high organic matter
removal rates. Cook and Foree (1974) obtained 68% COD removal in the
first effluent bed volume when leachate was passed through an activated
carbon column. Lime coagulation prior to the activated carbon treatment
increased the overall COD removal to 81%. Van Fleet e_t al_. (1974)
observed a 71% COD removal in the first effluent bed volume when leachate
was passed through an activated carbon column. Aluminum pretreatment
resulted in an overall COD removal rate of 94%. The highest COD removal
percentage, 85%, was observed by Roy Weston Inc. (1974) at a carbon
dosage of 10,000 mg/1. Because the leachate was obtained from an old
abandoned fill which had undergone considerable biological stabilization,
high adsorptive removals were to be expected.
These previous studies show that variable organic matter removal rates can
be obtained in landfill leachate using activated carbon, a fact which may
well reflect the variable nature of the organics. While activated carbon
treatment of concentrated leachate from recently generating fills was not
very effective, higher removal rates were obtained with leachate from
more stabilized fills. High removal rates can perhaps be obtained with
a combination of coagulation and adsorption or of biological pretreatment
and adsorption. Chemical precipitation or chemical oxidation alone were
not found to be very effective (Chian and DeWalle, 1976).
The purpose of the present study was to evaluate in greater detail the
effect of organic matter composition on adsorptive behavior using leachate
samples from different landfills, employing both adsorption isotherm
tests and column studies. The effect of biological pretreatment using
an anaerobic filter, with or without further aerobic effluent stabilization,
65
-------
on activated carbon adsorption was studies more extensively, since
preliminary work showed that this treatment sequence resulted in low
organic matter concentrations.
66
-------
MATERIALS AND METHODS
Batch adsorption tests were conducted with both undiluted and diluted
leachate samples to which increasing dosages of activated carbon were
added. The tests were conducted with Filtrasorb 400 (Calgon, Pittsburgh,
PA) added to rectangular 250 ml precipitation bottles filled with 100 ml
of the sample. Prior to adsorption, the leachate samples were filtered
through a 0.45 ym Millipore membrane filter (Millipore, Bedford, MA) to
remove the small amount of suspended solids initially present, since
only the adsorption of soluble organics was to be evaluated in this study.
After the carbon was added, the samples were shaken for 24 hours at
180 rpm. Distilled deionized water blanks to which activated carbon was
added were included to correct for the small amount of organic compounds
that dissolve from the carbon. The column studies were conducted with
columns 15 cm and 30 cm long having empty bed detention times of 1.5 and
3.0 minutes, respectively, at a flowrate of 10 cm/min (2.5 gpm/ft^)
maintained by means of an accurate FMI positive displacement metering
pump (FMI, Oyster Bay, NY). The columns were backwashed at a rate of
p
24 cm/min (6.0 gpm/ft ) when the headless buildup exceeded 200 cm. The
leachate used for the carbon tests was obtained from a lysimeter filled
with milled solid waste to which simulated rainfall was added (Chian
and DeWalle, 1976).
Anaerobic biological pretreatment of the leachate was accomplished with
the unit described in Chapter I. Further aerobic degradation was conducted
in a 80-liter batch reactor. Before and after aeration the soluble
organic matter was characterized by molecular weight and major classes of
organics using colorimetric tests. The molecular weight distribution was
estimated using Sephadex G-25 and G-75 columns (Pharmacia, Piscataway, NJ).
Larger quantities of the various molecular-weight fractions were obtained
with a preparative ultrafiltration unit HFA-180 (Abcor, Cambridge, MA)
followed by concentration of the UF permeate with a reverse-osmosis B-10
Permeator (DuPont, Wilmington, DE).
67
-------
Each fraction was passed through an activated carbon column. Batch
coagulation studies were conducted using a 1-liter sample to which a
lime slurry was added. Following addition of the slurry, the content
was rapidly mixed at 100 rpm for 1 minutes followed by slow mixing for
20 minutes at 25 rpm. The liquid was then settled for 30 minutes. The
various parameters were measured in the decanted supernatant.
68
-------
RESULTS AND DISCUSSION
Initially, the adsorption of the organic matter in the concentrated leachate
was evaluated. The results of these tests (Figure 21) show a relative low
adsorptive capacity: 0.045 mg TOC/mg AC, corresponding to a maximum COD
removal rate of 73% at the highest dosage. When the leachate obtained
from the same lysimeter two years later was tested, its concentration had
decreased to 29% of the initial concentration. Its adsorptive capacity,
however, was 6.5 times as high, possibly indicating that gradual biological
stabilization of the solid waste and of the interstitial leachate resulted
in organics more amenable to physical adsorption. This explanation is
also indicated by the results in Figure 21 since a higher maximum adsorp-
tive capacity tends to be associated with a lower initial organic matter
concentration. The positive effect of biological stabilization is further
examplified by the results of Karr (1972), who noted a maximum adsorptive
capacity of 0.16 mg TOC/mg AC using leachate without any biological treat-
ment. Using a similar wastestream after aerobic biological treatment,
Pohland and Kay (1975) obtained a capacity of 0.36 mg TOC/mg AC a factor
of 2.3 higher. Since the free volatile fatty acids are dissociated at
the pH of the leachate, higher adsorptive capacities are to be expected
at decreasing pH values when these acids are no longer ionized. The
results in Figure 21 indeed show the highest capacity at a pH of 4. The
lowest adsorptive capacity was observed at a pH of 8. These results are
contrary to those of Ho et ah (1974), who did not note any pH effect,
quite possibly because of the absence of any fatty acids in the analyzed
sample.
Since higher adsorptive capacities tended to be associated with lower
initial concentrations, several adsorption tests were made at the lowest
initial concentration noted in Figure 21; i.e., 120 mg/1 TOC. Using the
diluted leachate dated 6/12/73, a maximum adsorptive capacity of 0 068
nig TOC/mg AC was obtained (Figure 22), 1.48 times higher than the capacity
69
-------
1.0 IT
o»
0.1
u
o
a.
a
O
o
o
(O
•o
0.01
0.001
O Pohtand and Kang (1975)
V Roy Weston (1974)
A Ho et al (1974)
D Karr (1972)
A UI 6/12/73
• UI 8/6/75
x Enf i eld Conn 8/16/75 (pH6)
C0(mg/fTOC)
123
137
1690
2000
I3JBOO
395
5029
(X/M)0(mgTOC/mgAC)
0.36
0.092
0.60
a 16
0.046
0.30
0.155
J—I I I 11
ILL
I I II
I ' I '
I I I I 1111
10
100
1000
IOJOOO
lOOjOOO
TOC Equilibrium Concentration (mg/jf)
Figure 21. Activated Carbon Adsorption Isotherms with Different Leachate Samples
-------
o>
O
O
O
Q.
O
O
O.I
0.05
• Diluted Leachate
O Acid Mixture At
Constant AC Dose
Co
(mg/| TOC)
120
120
0.01 -
i—
(X/M)0
(mgTOC/mgAC)
0.068
0.145
•0 50 100
TOC Equilibrium Concentration (mg/J)
Dilute i'I>;uT*VMIiw" AdsorPtion Isotherms with a
Diluted Leachate Sample and a Fatty Acid Mixture
71
-------
of the undiluted leachate, indicating that lower concentrations do
correspond to higher capacities. The increase, however, is not as sub-
stantial as the increase of 6.5 times found for the biologically stabilized
leachate. Since the free volatile acids comprised 52% of the organic
matter in this leachate sample, an acid mixture was made having the same
relative composition as the leachate sample; i.e., 20 rug/1 formic,
50 mg/1 acetic, 10 mg/1 propionic, 50 mg/1 butyric, 10 mg/1 isobutyric,
10 mg/1 paleric, 10 mg/1 isovaleric, 70 mg/1 hexanoic acid and 100 mg/1
NaCl to give an initial TOC of 120 mg/1. Using this acid mixture, a
maximum adsorptive capacity of 0.145 mg/1 was observed 2.1 times as high
as that obtained with the diluted leachate sample. This result indicates
that the non-fatty-acid portion has a lower adsorption capacity than
the fatty-acid fraction and that simple removal of the free volatile fatty
acids, which occurs during aerobic biological stabilization, will not
result in higher adsorptive capacities. Removal of the non-fatty-acid
fraction and bacterial excretion or formation of refractory fulvic and
humic acids may therefore be necessary before higher capacities are
obtained.
Different methods of arriving at adsorption isotherms were next evaluated.
While the results shown in Figure 22 for the acid mixture were obtained
with a constant carbon dosage and increasing dilution of the acid mixture,
the acid mixture results in Figure 23 were obtained with a constant
organic matter concentration and increasing activated carbon dosages.
Although the maximum adsorptive capacity in both cases was the same, the
slope of the isotherm was lower using the constant carbon dosage; i.e.,
the value of n decreased from 1.0 to 0.79. Since the low adsorptive
capacity at low equilibrium concentrations often results from the presence
of a nonadsorbable fraction, dilution of the organic matter concentration
and therefore reduction of the nonadsorbable fraction at a constant
carbon dosage will result in a higher adsorptive capacity than will the
use of an undiluted sample with increasing carbon dosages, since even
very high carbon dosages will not remove this nonadsorbable fraction.
72
-------
0.
e
a
E
o
<
0.05
X 3E
>»
'o
O
V)
0.01
O Acid Mixture With
Increasing AC Dose
A Acid Mixture With
NaCI Addition
C0
120
120
(X/M)0
0.145
0.088
10 50 100
TOC Equilibrium Concentration (mgA)
Figure 23. Activated Carbon Adsorption Isotherms with a
Fatty Acid Mixture
73
-------
It is known that with increasing leaching of the landfill the cation
composition gradually changes, resulting in relatively higher sodium and
lower calcium concentrations. During biological treatment of leachate,
similar relative changes occur. To study this effect, sodium chloride
was added to the sample to give a final concentration of 2000 mg/1. The
results in Figure 23 show that adding salt decreased the maximum adsorp-
tive capacity to 61% of the initial capacity, while the slope of the
isotherm increased from 1.0 to 1.1, indicating that a high sodium chloride
concentration reduces the adsorption of organics. The low adsorptive
capacity obtained with the undiluted leachate may therefore be caused by
the high salt content, which could interfere with the organic matter
adsorption.
To compare the batch adsorptive capacities with those obtained during
continuous flow conditions, three activated carbon column tests were
conducted using the diluted leachate and the fatty acid mixture. The
undiluted leachate was also tested but resulted in rapid clogging of
the activated carbon column, primarily caused by oxidation of the ferrous
iron and formation of hydroxide precipitates. The effluent of the carbon
columns was monitored for TOC, and the results in Figure 24 show that the
initial bed volume of leachate passing through the column experienced a
TOC removal rate of approximately 59% in the 15-cm column. A 76% removal
rate was observed in the 30-cm column, comparable to the removal percentage
obtained in the batch process. These removal rates, however, are observed
only for the initial bed volumes passing through the column. The TOC
exhibited an immediate, rapid breakthrough, leveling off after about
200 bed volumes. Further passage of the leachate continued to remove
approximately 8% of the TOC. These results clearly show that although
high removals can be obtained with activated carbon, rapid organic matter
breakthrough precludes its consideration as a serious treatment alterna-
tive for raw leachate. Further complications are the rapid buildup of
headless and difficulties in backwashing the column.
74
-------
en
o
O
O Diluted Leachate, Column Length 15 cm
A Diluted Leachate, Column Length 30 cm
D Acid Mixture , Column Length 15 cm
300 400 500 600
Number Of Bed Volumes (-)
700
800
900
Figure 24. Breakthrough of Effluent TOC in Diluted Leachate and Acid Mixture
Passed Through Activated Carbon Columns
-------
At the termination of the column test after slightly more than 900 bed
volumes had passed through the 15-cm column, the adsorptive capacity was
calculated to be 0.037 mg TOC/mg AC, a value equal to 56% of the adsorp-
tive capacity obtained in the batch test. A comparison of the adsorptive
capacities of the 15-cm and 30-cm columns after 208 bed volumes showed
that the former column had reached a capacity of 0.019 mg TOC/mg AC,
while the latter column reached 0.011 mg TOC/mg AC. When the acid mixture
was passed through a 15-cm column, partial breakthrough took place after
250 bed volumes, and a further breakthrough occurred after 600 bed volumes.
The adsorptive capacity observed at the termination of the run was 0.118
mg TOC/mg AC, or 81% of the value obtained from the batch data. Since the
adsorption of organics in carbon columns is often limited by film diffu-
sion, lower capacities are to be expected when using columns rather than
single dosages. The smaller discrepancies between the batch and column
tests using the fatty acid mixture may be attributable to the more rapid
diffusion of these compounds compared to non-fatty-acid organics. To
eliminate some of the effects of film diffusion, the longer 30-cm column
was used for further research.
In the activated carbon column studies, the column was backwashed when
the headless reached 200 cm. The results in Figure 25 show that the
headloss in the 30-cm column increased approximately 2.5 cm for every
bed volume of diluted leachate passed. The suspended solids content of
the undiluted leachate was 94 mg/1. This value decreased only to 8.1
mg/1 when the leachate was diluted at a ratio of 1:115. It was evident
that another factor was contributing to the suspended solids content in
the diluted leachate, since dilution alone would have resulted in a
concentration of only 1 mg/1 or less. The leachate darkened after
dilution, probably a result of partial oxidation of the ferrous iron
content, which amounted to 25 mg/1 in the diluted leachate. When fully
oxidized, this amount of iron could result in a suspended solids concen-
tration of 48 mg/1. It was also observed that a light brown precipitate
accumulated at the top of the column. Entrance of the leachate into the
carbon column caused its color to change from a dark brown-green to light
brown, while a clear effluent was obtained at the bottom of the column.
76
-------
250
200
E
o
CO
T3
O
0)
X
150-
100-
50-
V Diluted Leachote
V id After I Backwash
O Anaerobic Filter Effluent
(8/1 to 2/27)
• id After I Backwash
A Anaerobic Filter Effluent
Lime Treated
D Anaerobic Filter Effluent
Aerated and Lime Treated
50 100
Number of Bedvolumes (-)
150
Pr"; J^dH°SS bu1IduP during passage of diluted leachate and
Pretreated anaerobic filter effluent through activated
carbon columns
77
-------
The data in Figure 25 include for comparison the headless buildup
observed when anaerobic filter effluent was passed through the carbon
column. The anaerobic filters were fed the same leachate as that used
for the carbon adsorption studies. Although the suspended solids con-
tent of the anaerobic filter effluent was as high as 240 mg/1 and this
effluent was not diluted before it entered the carbon column, the buildup
of headloss in the column was comparable to that for diluted leachate.
Visual observation also showed that the anaerobic filter effluent resulted
in a layer of relatively coarse suspended solids on top of the bed. It
did not form a gelatinous-like encapsulation of the carbon particles as
did the diluted leachate.
After the passage of 67 bed volumes, the 30-cm carbon column was backwashed
with distilled water at a rate of 24 cm/min. The initial backwash water
had a maximum iron concentration of 975 mg/1 and a TOC of 108 mg/1, which
after 6.7 bed volumes decreased to 0.2 mg/1 iron and a TOC of 40 mg/1.
No further TOC reduction was realized, since the adsorbed organics were
gradually desorbing. By subtracting these minimum concentrations, the
total quantity of material removed and its rate of removal from the carbon
bed during backwashing could be calculated (Figure 26). Since this amount
of material accumulated throughout the 67 bed volumes, the recovered
quantity can be converted to the equivalent influent concentration
removed in the activated carbon bed. The equivalent iron concentration
removed was calculated to be 22 mg/1, or 88% of the initial content,
while the equivalent TOC content was 1.1 mg/1, or less than 1% of the
initial content. Since analysis of the column effluent showed that 99.8%
of the iron was removed in the column, an amount equivalent to 12% of the
initial content remained in the carbon bed.
Further evaluation of the removal rate during the backwash indicated that
it occurred at a logarithmic rate with respect to time. During the first
backwash, more than 90% of both the iron and the TOC were removed with
the first three bed volumes (Figure 26). Since most of the hydroxide
solids in the carbon column were removed during the backwash, the rate
78
-------
100
o
0)
o.
O
O
O
CD
O
o>
o
E
0>
or
0)
§•
10
O.I
Dilute-Out
Of Completely
Mixed Reactor
0 I
O Iron Removal During First
Back Wash
A TOC Removal During First
Back Wash
• Iron Removal During Second
Back Wash
A TOC Removal During Second
Back Wash
-I I I I I
2
34 5678
Number Of Bed Volumes (-)
10 II 12
Figure 26. Removal of accumulated iron and organic matter during
backwashing of activated carbon columns
79
-------
of headloss buildup during the second part of the test was comparable
to that during the initial period (Figure 25). Complications, however,
were encountered during the backwash after the second period; it was
found more difficult to break up the carbon particles cemented together
by the iron hydroxide, and additional agitation of the top of the bed
was required. This increased cementing of the particles was also
reflected in the slower removal of iron during the second backwash
(Figure 26),
Although 90% iron removal was accomplished in 3 bed volumes during the
first backwash, it required 8 bed volumes in the second backwash. As in
the first backwash, the iron recovered in the second backwash represented
the major portion of the iron removed by the carbon column. The water
from the second backwash contained 7.8 mg/1 of the original leachate TOO,
as compared to 1.1 mg/1 recovered during the first backwash. This
difference may indicate that the second backwash was mroe efficient in
recovering the removed organics as a result of the additional agitation.
Also, the firmer cementing of the carbon particles by iron hydroxide may
have been more effective in coagulating and precipitating certain organics
from the leachate. This finding is further substantiated by the results
of the 15-cm carbon column test (Figure 24), in which a consistent TOC
removal of 18 mg/1, or 8% of the original TOC was observed, even though
the carbon bed was almost saturated.
These results indicate that rapid breakthrough of organic matter occurs
when leachate is treated in activated carbon columns. Rapid headloss
buildup and cementing of carbon particles posed additional operational
difficulties. Since the results shown in Figure 21, however, indicated
that biological stabilization tended to increase the adsorptive capacity
of the carbon, an extensive study was made of the capacities obtainable
after biological pretreatment of the leachate. The biological pretreat-
ment methods evaluated were (1) the anaerobic filter (AF) and (2) the
anaerobic filter followed by an aerated lagoon. Adsorption isotherms
of the 0.45-nm-filtered effluent of the anaerobic filter collected during
five sequential periods are shown in Figure 27. The average adsorptive
80
-------
1.0
Q
O
00
o
o
a.
o
O
o
tn
O.I
0.01
0 5/30-8/4
A 8/4-8/20
o 8/20-9/15
A 9/15-1/10
• 1/10-2/27
• 8/11-8/20
After Aeration
C0
559
759
829
900
1105
630
(X/M)0
0.295
0.108
0.091
0.380
0.175
0.510
10
100 1000 10,000
COD Equilibrium Concentration (mg/J?)
100,000
Figure 27. Activated carbon adsorptive capacities ot leachate after biological pretreatment with the
anaerobic filter followed by aerated lagoon
-------
capacity of the carbon during these five periods was 0.102 mg TOC/mg AC
or 0.261 mg COD/mg AC, a capacity 1.5 times higher than that obtained
for the diluted leachate and 2.2 times higher than obtained for the con-
centrated leachate, indicating that biological pretreatment greatly
enhances the adsorptive capacity. Considerable variation, however, was
noted between the individual results; the capacity ranged from 0.091 to
0.380 mg COD/mg AC.
Increasing carbon dosages were added to aliquots of the AF effluent and
the supernatants were analyzed for absorbance, aromatic hydroxyls, and
carbohydrates in addition to TOC and COD. The ratios of the different
parameters were then calculated for the aliquots of the anaerobic filter.
The results are shown in Figures 28 and 29. Generally, when the ratio
of a certain parameter increases with respect to the COD it indicates
that the COD is removed to a greater extent then that parameter. When
the ratio decreases, however, that parameter is removed to a relatively
greater extent than the COD. Since the ratio of COD to TOC reflects the
oxidation state of organic compounds, its increase with decreasing COD
values indicates that oxygenated compounds having a low COD/TOC ratio
are removed preferentially to the remainder of the COD (Figure 28a). At
increasing dosages, however, those less-oxygenated organics having a high
COD/TOC ratio are removed, resulting in a gradual decrease of the COD/TOC
ratio in the supernatant. At very high carbon dosages, only highly
oxygenated nonadsorbable compounds remain in solution.
Since the maximum of the adsorbance-to-COD ratio (Figure 28a) occurred
at a slightly lower COD value than the maximum of the COD/TOC ratio,
the absorbance or color was removed slightly later than these oxygenated
compounds. The maximum of the aromatic hydroxyls-to-COD ratio occurred
at slightly lower COD values than the absorbance/COD ratio, indicating
that these compounds, in turn, were removed slightly after the color
(Figure 29a). Aromatic hydroxyls are often associated with fulvic-like
materials that exhibit high adsorptive capacities. The maximum of the
carbohydrate-to-COD ratio occurred at the lowest COD values, indicating
that these compounds are only removed after most other compounds are
82
-------
J-I2-K)"
00
CO
O 5/30-8/4
8/4-8/20
a 8/20-9/15
A 9/15-1/10
• 1/10-2/27
500
IOOO
1500
COD (mg/i)
Figure 28. Ratio of absorbance to COD and ratio of COD to TOC at decreasing COD concentration
corresponding to increasing activated carbon dosages added to the anaerobic filter effluent
-------
X
o
"5
ll
oQ
08
a
o
o
V)
a>
TJ
>,
O
JO
O
O
O
tr
0.03
0.02
0.01
0
0.14
0.12
0.10
0.08
0.06
0.04
0.02
o 5/30-8/4
A 8/4 -8/20
a 8/20-9/15
X 9/15- I/10
• I/10 - 2/27
500
1000
1500
COD (mg/Jt)
Figure 29. Ratio of aromatic hydroxyls to COD and ratio of carbohydrates
to COD at decreasing COD concentrations, corresponding to
increasing activated carbon dosages added to the
anaerobic filter effluent
84
-------
already adsorbed onto the carbon. The carbohydrate concentration often
reflects the presence of high-molecular-weight humic-carbohydrate-like
organics. These compounds generally have a low adsorptive capacity
because of their large molecular weight, which hinders their micropore
penetration (DeWalle and Chian, 1974).
Because variable adsorptive capacities were observed (Figure 27), the
magnitude of the maximum adsorptive capacity was related to other
parameters measured in the anaerobic filter effluent to detect any
systematic trend. It was found that the magnitude tended to increase
with the absolute concentration of aromatic hydroxyls; it did not cor-
relate, however, with the ratio of aromatic hydroxyls to COD or the
ratio of aromatic hydroxyls to carbohydrates. These results indicate
that absolute aromatic hydroxyl concentrations may be a valuable para-
meter for predicting adsorptive capacity.
The results in Figure 27 also show that aeration of the anaerobic filter
effluent to which activated sludge was added increased the adsorptive
capacity from 0.108 to 0.51 mg COD/mg AC, a 4.7-fold increase (Table 2).
Since this value represents the highest adsorptive capacity obtained
in the study, a more extensive analysis was made of the changes in
organic matter content that occur during the aeration step.
Sixty liters of anaerobic filter effluent were placed in an 80-liter
batch reactor to which 1000 mg/1 of activated sludge was added. After
the sludge was added and aeration had been started, samples of the mixed
liquor were taken, filtered, and analyzed for various parameters. The
pH, ORP, DO, and conductivity were measured before filtration. The
results in Figure 30 and 31 show that the COD decreased rapidly within
a three-day period. Parallel decreases were noted for free volatile
fatty acid concentrations and for turbidity. Although some fatty acids
remained in solution after the three-day period, their concentration was
close to the detection limit of the test. Furthermore, acids other than
the free volatile fatty acids were also detected by the column
chromatographic method used for fatty acid analysis. As the free
85
-------
Table 2
Measured Adsorptive Capacities for leachate and
biologically pretreated leachate using
adsorption isotherms
Adsorbate
(x/m),
1. diluted leachate isotherm
2. undiluted leachate isotherm
3. acid mixture with constant
AC dose isotherm
4. acid mixture with increasing
AC dose isotherm
5. acid mixture with Nad
addition isotherm
6. diluted leachate with AC
column
7. acid mixture with AC column
8. anaerobic filter effluent
isotherm
9. AF effluent after aeration
isotherm
120 mg/1 TOC 0.068 mg TOC/mg AC
13800 mg/1 TOC 0.046 mg TOC/mg AC
120 mg/1 TOC 0.145^ mg TOC/mg AC
120 mg/1 TOC 0.145 mg TOC/mg AC
120 mg/1 TOC 0.088 mg TOC/mg AC
120 mg/1 TOC 0.038 mg TOC/mg AC
120 mg/1 TOC 0.118 mg TOC/mg AC
317 mg/1 TOC
834 mg/1 COD
225 mg/1 TOC
630 mg/1 COD
0.102 mg TOC/mg AC
0.261 mg COD/mg AC
0.174 mg TOC/mg AC
0.52 mg COD/mg AC
86
-------
00
50
40
H 30
-a
>,
1 20
10
0L
500r
400-
-S300-5!
•o
"5
O Q^onductivity
10 -i 10,000
200-
IOO-
E
u
"o
E
5000
•o
o
o
Time of Aeration (day)
Figure 30. Changes of different parameters in the filtered mixed liquor
occurring during aeration of the anaerobic filter effluent
-------
+200 -
I
o.
8
0.
(T
O
•noo -
7L
D Color (0.45/i
Filtered)
3456
Time of Aeration (day)
30
5!
o>
20
X
S
•o
>s
I
O40
0.39
0.38
0.37 X
0.36 o
035 «
0.34
0.33
-0.32
L3I
-"0.30
Figure 31. Changes of different parameters in the filtered mixed liquor
occurring during aeration of the anaerobic filter'effluent
-------
volatile fatty acids are biodegradable, it is to be expected that they
would be removed relatively rapidly. During the aeration process, small
decreases followed by large increases were observed for the carbohydrate
and aromatic hydroxyl concentrations. The increase in the latter para-
meter may reflect the increasing presence of fulvie-like materials generated
during the biological degradation process. The increase in carbohydrates
occurred after a three-day period when the degradable fatty acids were
removed from the solution. Similar increases in concentrations of high-
molecular-weight humic-carbohydrate-like materials have been observed by
DeWalle and Chian (1974b) during aerobic biological treatment of leachate.
The above results therefore suggest that aeration of the anaerobic filter
effluent in the presence of activated sludge tends to decrease the
concentration of low-molecular-weight free volatile fatty acids while
causing an increase in the concentration of high-molecular-weight humic-
carbohydrate-like materials and of intermediate-molecular-weight fulvic-
like materials.
To obtain further insight into the changes in compounds of various
molecular weights the organics present in the effluent of the anaerobic
filter were analyzed by gel permeation chromatographic analysis before
and after aeration. The TOC data from the Sephadex G-75 and G-25 columns
showed that considerable changes occurred as a result of aeration (Figure
32). Most noticeable was the formation of high-molecular-weight compounds
excluded from the Sephadex G-75 columns as they eluted at 20 ml. Only
organics with a molecular weight larger than 30,000 to 50,000 are excluded
from Sephadex G-75. The results of the G-75 tests further show that
after aeration fewer organics elute beyond the inclusion of the column,
as indicated by the inorganic carbon (1C) peak detected at 51 ml. This
result was further illustrated by the G-25 Sephadex tests, in which the
peak beyond the inclusion of the column almost completely disappeared.
The organics detected around 50 ml in the G-25 eluate decreased to a
lesser extent than did the low-molecular-weight organics. These data
therefore support the analytical data shown in Figures 30 and 31, in
that aeration decreases the concentration of low-molecular-weight organics
corresponding with a concurrent increase in high-molecular-weight organics,
89
-------
5 60
o
o
I 1 1 1 1 1 1
I 1 1 1 1
Sephadex G-75
Before Aeration
Sephadex G-75
After Aeration
40
20
60
£ 40
a
o
20
_ C
Sephadex G-25
Before Aeration
1C
. . D
Sephadex G-25
After Aeration
20 30 40 50 60 20 30 40 50 60
Edition Volume (ml) Elution Volume (ml)
Figure 32. Sephadex eluate of the organic matter in the anaerobic
filter effluent before and after aeration
90
-------
while the intermediate-molecular-weight organics tend to be relative
stable.
The adsorptive behavior of each molecular-weight fraction before and after
aeration was next tested using activated carbon columns. Column break-
through tests represent the actual behavior in full-scale plants better
than do isotherms. The anaerobic filter effluent, after removal of the
suspended solids, was divided into a high-molecular-weight fraction by
an 18,000 MW cutoff ultrafiltration (UF) membrane, an intermediate-
molecular-weight fraction using a 15 MW reverse osmosis (RO) membrane,
and a low-molecular-weight fraction consisting of the RO permeate. The
results in Figure 33 show that while treatment of the unfractionated
anaerobic filter effluent results in a 68% COD removal rate, the rate
is only 52% for the high-molecular-weight fraction present in the 18,000
MW UF retentate. The highest removal percentage, 95%, was observed for
the RO retentate obtained after concentrating the UF permeate with a B-9
Dupont permeator. A relatively low removal rate of 63% was also noted
for the RO permeate. These results agree with the batch adsorption
data in Figure 28 and 29, which show a relatively rapid removal of the
aromatic hydroxyls often associated with intermediate molecular weights
but less rapid adsorption for the carbohydrates often associated with
high-molecular-weight organics.
Different results were obtained after aeration of the anaerobic filter
effluent. The COD removal percentage for the unfractionated sample
increased from 68% to 80% (Figure 34). The initial removal of COD from
the UF retentate increases slightly from 52% to 58%, while the removal
from the RO retentate decreased slightly from 95% to 90%. The RO
permeate containing the low-molecular-weight polar compounds showed
the largest improvement, however. Its COD removal rate increased from
63% to as much as 96%, indicating that aerobic biological treatment
causes the largest changes in concentration and character of low-
molecular-weight organics which permeate the 150 MW RO membrane.
91
-------
0°
o
o
C_>
"N.
O
o
O
x.
O
0°
o
O COD
A Color
D Turbidity
0
Number of Bed Volumes (-)
Figure 33. Breakthrough of COD, color and turbidity in activated carbon
effluent during passage of A) unfractionated anaerobic filter
effluent B) 18,000 MW UF retentate of A C) UF permeate
and RO retentate of A D) UF and RO permeate of A
92
-------
O COD
A Color
D Turbidity
50 100
Number of Bed Volumes (-)
Figure 34. Breakthrough of COD, color, turbidity in activated carbon
effluent during passage of A) unfractionated aerated anaerobic
filter effluent, B) 18,000 MW UF retentate of A, C) UF
permeate and RO retentate of A, D) UF and RO permeate of A
93
-------
While previous studies clearly indicate that activated carbon will result
in the largest organic matter removal rates (Chian and DeHalle, 1976), a
limited study was made of the effectiveness of lime treatment. The TOC
reduction in leachate using lime coagulation was 26%, and it decreased to
20% in the leachate pretreated by the anaerobic filter (Figure 35).
Aeration of the anaerobic filter effluent resulted in a TOC removal
of 23% (Figure 36). These removal efficiencies were obtained at exces-
sively high dosages up to 7000 mg/1, resulting in a sludge volume of
15%. Aeration of the anaerobic filter effluent did decrease the required
lime dosage by about 50%, but even at the lower dosage the sludge volume
was not greatly reduced.
94
-------
I5OO
in
IOOO -
o>
o
500-
1500
1000
o
i
I IO
o
0
o
c
o
500 5
£
i
o
o
O IOOO 20OO 3OOO 4OOO 5OOO 60OO 7OOO 8C
Lime Dosage (mg/£) As C4(OH)2
90OO
-"0
15
Q70
0.60
10 - 0.50 -
-------
I500A
1000-
55
I
o
o
o
10
cr>
500 -
1000 2000 3000
Lime Dosage (mg/£) As C4(OH)2
4000
0 J
1500 -.15 -,1.5 -,13
o>
o
i
a;
o
toooj-
m
•o
c
a
£
o
c
500 |
o
sl-fc
CO
Jo -k
1.0 -
E
c
O
O
12
Figure 36.
Effect of lime precipitation treatment of aerated anaerobic filter effluent
on different parameters measured in the supernatant
-------
REFERENCES
Boyle, W. C. and Ham, R. K. "Treatment of Leachate from Sanitary Landfills,"
Jour. Mater Pollution Control Fed. 46, 860 (1974).
Chian, E. S. K. and DeWalle, F. B. "Sanitary Landfill Leachates and their
Treatment," Amer. Soc. Civil Engrs.. Jour. Env. Engr. Div. 102,
411 (1976).
Cook, E. N. and Foree, E. 6. "Aerobic Biostabilization of Sanitary
Landfill Leachate," Jour. Water Pollution Control Fed., 46, 380
(1974). ~~
DeWalle, F. B. and Chian, E. S. K. "Removal of Organic Matter by Activated
Carbon Columns," Amer. Soc. Civil. Engrs.. Jour. Env. Engr. Div. 100,
1089 (1974a).
DeWalle, F. B. and Chian, E. S. K. "The Kinetics of Formation of Humic
Substances in Activated Sludge and their effect on Flocculation."
Biotechn. Bioengr.. 14., 739 (1974b).
Ho, S. et_ aj_. "Chemical Treatment of Leachates from Sanitary Landfills,"
Jour. Mater Pollution Control Fed.. 46, 1776 (1974).
Karr, P. R. "Treatment of Leachate from Sanitary Landfills "Special
Research Problem, School of Civil Engineering, Georgia Institute
Technology, Atlanta, Georgia, Oct. (1972).
Pohland, F. G. and Kang, S. J. "Sanitary Landfill Stabilization with
Leachate Recycle and Residual Treatment," Amer. Inst. Chem. Engr.
Symp. Series No. 145, Water-1974, vol. 71, 308 (1975).
Roy, Weston "Intern Report, Leachate Treatability Study, New Castle County
Delaware," Roy F. Weston Inc., West Chester, PA (1974).
Van Fleet, S. R. et_ aJL "Discussion, Aerobic Biostabilization of Sanitary
Landfill Leachate," Jour. Water Pollution Control Fed., 46, 2611
(1974)> —'
97
-------
Ill
TREATMENT OF A HIGH STRENGTH SOLID WASTE
LEACHATE WITH THE AERATED LAGOON
CONCLUSIONS
It was concluded that aerated lagoon treatment of high strength
leachate with a COD of 57,900 mg/1 could remove between 93% and 96.8%
of the organic matter without any pretreatment of the influent
leachate at detention times ranging from 85.7 days to as low as 7 days.
An extensive evaluation of phosphate requirements showed that the COD:P
ratio in the influent of the 30 day unit should be at least 300:1. For
units with a retention time of 85.7 and 60 days, they were even able to
be operated with a COD:P ratio of 1540:1 in the feed solution. Cessa-
tion of nutrient addition at a COD:P ratio of 165:1 to the units
operated at relatively low detention times caused an immediate increase
in effluent organic matter, a decrease in biological MLVSS and a
deterioration of the sludge settling rates.
All units showed high removals of heavy metals, especially for iron
(>99.9%), zinc (99.9%), calcium (99.3%) and magnesium 75.9%). Lower
removals were observed for sodium (24.1%) and potassium (17.1
The dewatering characteristics of the sludge of the 30 day unit were
greatly improved by addition of cationic polymers and inorganic coagu-
lants. An approximate 20 times increase in sludge specific resistance
was obtained at polymer dosages varying between 0.15% to 1.5% and
inorganic coagulant dosages of 2.9% to 25.5% on the basis of dry sludge
weight. These increases in specific resistance corresponded to 6 times
increases in vacuum filter yields.
98
-------
INTRODUCTION
The conventional sanitary landfill is considered one of the safest
and least expensive methods currently used for the disposal of munici-
pal solid wastes. In areas where groundwater is polluted by leachate,
however, the sanitary landfill becomes less attractive as a solid waste
disposal alternative. Leachate which is defined as a liquid draining
from the refuse disposal landfill and primarily resulting from infil-
trating rainfall often contains a high concentration of organic matter
and inorganic ions, including heavy metals. The magnitude of leachate
contamination depends largely on the amount, the concentration, and
the extent of migration of polluting species. These factors in turn
depend on the quantity and composition of the refuse, the age of the
landfill, the hydrogeology of the site, and the climate. Hence,
leachate contamination is a matter of real concern to design engineers,
regulatory agencies, and municipalities involved with solid waste
disposal.
To curb the pollution created by leachate, three preliminary measures
are practiced today (Chian and DeWalle, 1976): (1) prevention of
leachate production, (2) recirculation of leachate back onto the
sanitary landfill, and (3) collection and treatment of leachate.
The first alternative seeks to minimize leachate production by prevent-
ing rainfall and penetration of runoff water by installing a low
permeable cover on the fill, diverting surface runoff water generated
upstream from the fill, and locating the fill in areas which minimize
groundwater pollution. These methods have met with limited success,
particularly when the fill was located in areas with fluctuating water
tables, unfavorable geological conditions, and irregular rainfall
patterns. The disadvantage of the above strategy is the reduction in
rate of landfill stabilization, resulting from the lower solid waste
moisture content. The ultimate bearing capacity therefore becomes
lower.
99
-------
The second alternative involves the application of the collected
leachate onto the top of the fill by surface irrigation. As a
resuH the moisture content of the solid waste is increased and
presumably reaches an optimum necessary for anaerobic biological
stabilization. In this alternative, the leachate is stabilized at
a rate equal to that of the refuse itself, with the refuse function-
ing as an anaerobic trickling filter.
The third alternative is the most recent in which the collected
leachate is treated by biological or physical chemical methods.
The amount of leachate to be treated depends primarily on the pre-
cipitation and the effectiveness of the cover material. In this
alternative the landfill is lined with an impervious barrier at the
bottom of the fill which enables the collection of the leachate
which is subsequently transported to the location of the treatment
unit. It has been found that biological treatment methods are
effective when treating leachate generated from recently deposited
solid waste, while physical-chemical methods yield better results
when treating leachate from more stabilized solid waste (Chian and
DeWalle, 1976).
The purpose of this study was to evaluate the treatability of leach-
ate from a recently installed landfill, utilizing the aerated lagoon
or extended aeration process followed by physical-chemical processes
for removal of residual organics and inroganics. Sludge conditioning
and dewatering were also evaluated. To accomplish the treatment
objectives, laboratory scale extended aeration units were operated
under various loadings and levels of nutrient addition. Physical-
chemical processes considered for treatment of the effluent from
biological extended aeration units include ozone, activated carbon,
ion exchange and reverse osmosis.
100
-------
AEROBIC BIOLOGICAL TREATMENT OF LEACHATE
A preliminary study by Schoenberger et al_. (1970) indicated that
although the organic matter in leachate consists of molecules which
are not assimilated easily, aerobic treatment of leachate still
appears feasible. Boyle and Ham (1974) showed that aerobic treatment
of leachate without nutrient addition and with a detention time of 5
days resulted in COD reductions between 80 and 93% at the organic
loadings of 0.545 kg COD/m3-day (0.034 Ib COD/day/cu ft) and 1.04 kg
COD/m -day (0.065 Ib COD/day/cu ft), respectively. An increase in
organic loading from 1.04 to 1.75 kg COD/m3-day resulted in a con-
siderable reduction in process efficiency. An organic loading of
6.1 kg/m -day (0.38 Ib COD/day/cu ft) with a detention time of 1 day
was unsuccessful. The combined treatment of leachate and municipal
sewage in an extended aeration activated sludge unit was feasible as
a level of at least 5% by volume of leachate added to the sewage did
not seriously impair the effluent quality. Only when the unit received
more than 5% leachate at a sludge age of less than 11 days, did the
effluent COD experience a substantial increase. The settling charac-
teristics of the unit decreased concurrently.
Foree and Cook (1974) reported the successful aerobic biological treat-
ment of leachate with a COD stabilization efficiency of greater than 97%
The organic loading was 1.68 kg COD/m3-day (0.0985 Ib COD/day/cu ft),
with a detention time of 10 days. No major beneficial effect resulted
from pH adjustment by lime. The nutrient additions were found to be
unnecessary at such a low organic loading, but did improve the perfor-
mance of aerobic treatment slightly. Nutrient additions were required
at a detention time of 5 days with an organic loading of 3.16 Ibs
COD/m3.day (0.197 Ib COD/day/cu ft), and a detention time of 2 days
with an organic loading of 7.89 kg COD/m3-day (0.492 Ib COD/day/cu ft).
Karr (1972) also found that the aerobic treatment of leachate, produced
101
-------
variable effluent qualities while operating at detention times ranging
from 2 to 15 hours with COD loadings ranging from 4.5 to 54.2 kg COD/
m3.day (0.28 to 3.38 Ib COD/day/cu ft). However these units were able
to generate a sludge with good settling properties as the sludge volume
index (SVI) was approximately 30. In further studies in the same lab-
oratory, Pohland and Kang (1975) operated an aerobic biological leachate
treatment process with detention times varying from 8 to 2.3 hours and
organic loadings varying from 1.5 to 5.3 kg/m -day (0.094 to 0.33 Ib
COD/day/cu ft). The concentration of leachate for this study was very
low as compared with the studies mentioned above. Both anaerobic and
aerobic treatment resulted in large reductions in the calcium, mag-
newium, iron concentration; however, the reduction of Ca and Mg was-
larger in the aerobic units. This was further substantiated by Karr
(1972) and Foree and Cook (1974).
102
-------
KINETIC CONSIDERATIONS IN AEROBIC BIOLOGICAL TREATMENT
The oxidation of organic matter under aerobic conditions in extended
aeration units is analogous to the natural purification process which
occurs in rivers. Aerobic conditions are maintained, provided that the
rate of oxidation does not exceed the rate of reaeration. When the
concentration of organic matter is of such a magnitude that the oxidation
rate exceeds the reaeration rate, anaerobic conditions may result. In
order to maintain an aerobic environment in treatment units, it is
generally necessary to supply additional oxygen by means of mechanical
or diffused aeration systems.
Under conditions of substrate limiting, the rate of biological oxidation
of organic matter can be expressed by a first order reaction as follows:
dS „ c
dt = K1S (1)
where
S = concentration of substrate surrounding the microorganisms,
mass/volume
Kj = overall first order substrate removal rate constant, time"1
It has been shown that at high concentrations of organic matter, the
rate of oxidation and sludge growth is independent of the concentration
of substrate (Eckenfelder and O'Connor, 1954). For this case, the rate
of substrate removal can be approximated by a linear function:
dS „
dt " Ko (2)
where
KQ = rate of substrate removal, mass/time
The oxidation kinetics are complicated in complex wastes, since the
constituents are not oxidized at the same time and at the same rate.
103
-------
In order to define the oxidation rate for these cases, a composite
expression can be employed for concurrent substrate removal
(3)
in which K-| , K2, etc. are the rate constants for the respective concen-
trations of constituents S, , Sp, etc., and S, + Sp + ...+ S = S .
In considering the rates of various biological oxidative systems,
other factors such as the concentration of active solids in suspension,
intensity of fluid turbulence, and the level of dissolved oxygen must
be taken into account. In a large extended aeration lagoon sufficient
turbulence is usually present to create a condition of uniform concen-
tration throughout the reactor. The concentration of substrate in the
effluent is therefore equal to the concentration of substrate in the
lagoon itself. A material balance of the substrate may be developed
as follows:
[Substrate applied] - [Substrate discharged] = [Substrate removed]
SOQ - SQ = (df} v (4)
To simplify the equation, it is assumed that the oxidation of organic
matter in leachate proceeds by a first order reaction. Then Equation
(4) becomes
SQQ = SQ = K^V (5)
where
SQ = concentration of substrate in influent, mass/volume
S = concentration of substrate in effluent, mass/ volume
Q = rate of flow, volume/time
V = volume of the extended aeration unit, volume
104
-------
Equation (5) may be re-expressed in the following manner:
so " 1 + ^(V/Q) " 1 + Kje (6)
e = detention time of the unit, time
Considering the effect of sludge solids, Equation (6) becomes
_-
SQ " 1 + K X6 (7)
where
S = concentration of microorganisms, mass/volume
K1 = Kj-X
Kj = overall first-order substrate removal rate constant per
quantity of microorganisms, volume/mass -time
The efficiency, E, of substrate stabilization is
K1, Xe
(8)
An expression of mass balance for the mass of microorganism in the
aeration unit at steady state condition can be written as:
-'-'d (9)
where
Y = growth yield coefficient, mass of microorganisms/mass of
substrate utilized
kd = microorganism-decay coefficient, time"1
The rate of food utilization, dS/dt, in the aeration unit can be
evaluated on a finite time basis:
At=V(So"S) (10)
105
-------
Utilizing Equation (10), Equation (9) then can be solved for the mass
concentration of microorganisms in the aeration unit, X, to yield
V (S - S)
Equations (7) and (11) were used extensively in this study to determine
the characteristics Kj , Y and k^, based on different hydraulic detention
time or biological solids retention time, e.
106
-------
MATERIALS AND METHODS
Lyslmeter and Solid Waste Leachate
The leachate used in this study was obtained from a large scale, simu-
lated landfill lysimeter located at the Univeristy of Illinois, Urbana,
Illinois. Leachate was collected in batches each week and stored at
4°C before use. Samples were monitored for TOC and COD to ensure
uniform concentrations. The simulated landfill column consisted of
an epoxy-coated steel tank, lined with heavy duty PVC sheeting, having
a diameter of 1.53 m (5 ft) and a height of 3.05 m (10 ft). The solid
waste used to fill the lysimeter was collected in the Hartwell section
in Cincinnati, Ohio. The physical characteristics of the raw segre-
gated solid waste prior to shredding is shown in Table 3 using a 483 kg
sample. A total of 1520 kg (3358 Ib) of solid waste was placed in the
lysimeter, which filled it to a depth of 2.52 m (8.3 ft). Deionized
water was added during the filling operation to simulate a wet period
during solid waste collection. Also, this assisted in bringing the
lysimeter more rapidly to field capacity. Thereafter, an equivalent
of 0.089 cm precipitate per week was added to generate sufficient
leachate for characterization and process evaluation.
Leachate samples collected from the lysimeter were characterized with
respect to certain chemical and physical properties. The results are
presented in Table 4. It may be concluded that these data are within
the range of values reported in the literature for leachate composition
(Salvato et al_., 1972)
Aerated Lagoon
The aerated lagoon or biological extended aeration process studies were
conducted in six completely-mixed vessels fed with leachate. The first
three of these units consisted of 3 liter plastic tanks with detention
times of 30, 60, and 85.7 days, respectively. Each day these units were
fed with 100, 50, and 35 m£ of leachate, respectively, after the equiva-
lent volume of the mixed liquor was withdrawn. The second three units
107
-------
Table 3
Physical Characteristics of Solid Waste Placed in the
Lysimeter Used for Leachate Generation
Food Wastes
Garden Wastes
Paper
Plastics
Wood
Metals
Glass
Rock
Rags
Diapers
Total
Solid Waste Used
in this Study
Percent
11.63
8.34
43.75
4.66
0.61
10.85
15.82
1.20
--
3.13
99.9
Average Values Reported fo
Municipal Solid Wastes
(Salvato, 1972)
Percent
15
5
50
3
2
8
8
7
2
—
100
108
-------
Table 4
Analysis of Leachate Collected from Solid Waste Lysimeter
Leachate Cone.
for Extended
Aeration Unit*
Parameter 1,2,3
PH
Conductivity
ORP
TOC
COD
COD/TOC
Tannins
Protein
Total Solids
Fixed Solids
% Volatile Solids
Calcium
Magnesium
Fatty Acids
FA (C) /TOC
Chloride (Cl)
Sulfate ($04)
Ammonia .Nitrogen (NH3-N)
Nitrate Nitrogen (NOs-N)
Organic Nitrogen (Org-N)
Ortho phosphate 0-P04 (P)
Total P04 (P)
Iron (Fe)
Copper (Cu)
Zinc (Zn)
Cadmium (Cd)
Chromium (Cr)
Lead (Pb)
Nickel (Ni)
Sodium (Na)
Potassium (K)
Ca
Mg
i~ " ~ " L~ • • • - ----- --
5.5
20,000
-50
19,400
57,900
2.99
800
1,480
41,580
20,030
51.8
3,650
525
21,400
0.50
1,800
1,500
540
0.1
760
1.4
4.5
2,125
0.15
72
-
0.52
0.92
1.7
1,350
1,240
3,780
660
Leachate Cone.
for Extended
Aeration Unit
4, 5, 6
5.5
_
_
11,773
35,237
2.99
_
-
_
_
3,000
300
_
_
_
_
_
-
-
-
-
1,020
_
55
-
„.
_
_
800
50
_
-
Range of
Concentrations**
3.5-8.5
_
_
0-89,500
_
mt
1,000-45,000
_
5-4,000
17-15,600
_
34-2,800
1-1,800
0-1,100
0-1,300
150-550
_
0-150
0.2-5,500
0-10
0-1,000
_
_
0-5
0.01-0.8
0-7,700
3-3,800
_
-
**
All concentrations in mg/£ except pH, conductivity (pmho/cm) and ORP (mV)
Salvato et al., 1972
109
-------
were 30, 15, and 7 liter plastic tanks with detention time of 30, 15,
and 7 days, respectively. The same rate of withdrawal and filling was
used daily for all three units (e.g., with every liter of leachate fed
to the 30-day detention unit having a volume of 30 liters, one liter
of mixed liquor was taken out}- The leachate fed to these three units
had a strength of about 60% of the leachate used earlier. This was a
result of the addition of more deionized water to the lysimeter in order
to obtain more leachate to be fed to the last three units having shorter
detention times.
Each unft was equipped; with several porous glass air diffusers at the
bottom to provide complete mixing. Phosphorus (P) and nitrogen (N) were
added prior to the addition of leachate based; on the assumed composition
of GgH^N far bacteria. From this assumption corresponding to a COD:P
ratio of 164:1 and a COD:M ratio of 19vl:l, the predicted nutrient
requirements were 347 mg/Titer P and 1737 mg/Iiter N for leachated added
to units 1, 2, and 3, whereas 211 mg/liter P and 1057 mg/liter N were
added for leachate fed to units 4* 5» and 6.. Nitrogen was supplied in
excess of this predicted! requirement, whereas addedi phosphorus was
decreased stepwise to approach the minimum amount necessary in the first
three units. For the second three units* the nutrients were supplied in
slight excess for the first 35 days and! were eliminated during the last
period of 35 days.. Nutrient solutions were iniitialliy made fron* dibasic
potassium phosphate and ammoitiumi nitrate during, the first 171 days of
operation of the first three uniits. Because- of the resulting high, con-
centration of potassium, ian- ira the effluents* the nutrient solutions
thereafter were changed by/ using; ammonium phosphate and ammonium nitrate
far all six units.
Ihe experimental protocols writto these units, were designed as in Table 5..
Alii units were started with sludgje acclimated in a preliminary experi-
ment. During the operation! of the aerated? lagoons, distillled
dteianized water was added daily to each of the units, to replenish the
loss of water due to evaporation. Because of the relatively low feed
rates and the correspondingly lon§ detention! times, thiis method of daily,
110
-------
Table 5
Operating Parameters of the Aerated Lagoons
Organic Loading Detention
Unit Operation Conditions fcg COD/m3-dayleg TOC/nP-day Time, day
1 Fed leachate with 0.67 0.224 85.7
nutrients
2. Fed leachate with 0.96 0.32 60
nutrients
3. Fed leachate with 1.94 0.64 30
nutrients
4. Fed leachate with, then 1.15 0.60 30
without, nutrients
5. Fed leachate with, then 2.35 0.78 15
without, nutrients
6. Fed leachate with, then 5.02 K68 7
without, nutrients
111
-------
withdrawal and filling can be simulated as a continuous flow condition.
The duration of the operation for these units was 150 days for units 1,
2, and 3, and 70 days for units 4, 5, and 6.
Sludge Settling and Dewaterinq
Sludge from the aerated lagoons receiving leachate is normally suffi-
ciently dense to conduct batch settling tests to determine the inter-
face subsidence velocities at varying initial sludge concentrations.
The settling tests were conducted at room temperature (24°C), using a
one liter graduated cylinder equipped with a stirrer operating at 1 rpm.
The slope of the linear part of the settling curve is the interface
settling velocity, VG> for the sludge at a given initial concentration.
The supernatant in the graduated cylinder resulting from the first
settling experiment was removed and the mixed liquor from the extended
aeration tank was added again. This resulted in a more concentrated
sludge and a resulting slower interface subsiding velocity. By repeating
this procedure the relationship between suspended solids concentration
and interface settling velocity was obtained.
The sludge concentrations of the aerated lagoons were very high, and
approximately half of the mixed liquor suspended solids consisted of
non-volatile suspended solids such as iron hydroxides. After sludge
settling, a subsequent dewatering step of the concentrated sludge is
necessary prior to drying and ultimate disposal. In the dewatering
experiment, the sludge was taken from unit 4 having a detention time of
30 days, and settled for 24 hours. Ferric chloride, lime, and cationic
polymers (Nalco 73C32 and Primofloc C-7) were used as chemical condi-
tioners. The specific resistance of the sludge to dewatering was
examined by using the Buchner-Funnel laboratory method. In view of the
close correlation between laboratory results and those obtained on large
machines, the laboratory testing to result in recommendations for com-
mercial filters is of great importance. To obtain design and perform-
ance data for sludge vacuum filters, the filter leaf test was utilized.
This test consists of a leaf with a predetermined area, so the cake
112
-------
weight per square foot can be determined as in commercial units.
Buchner funnel test procedure:
The apparatus required are a Buchner funnel, filter paper, vacuum supply,
rubber hoses, vacuum bottle, vacuum gauge and general laboratory equip-
ment. Procedures of conducting this test are given as follows:
a) Moisten filter paper (Whatman filter paper No. 1) and place
it in the Buchner funnel. Apply a vacuum to obtain a seal.
Empty water collected in filtrate receiver.
b) Analyze the sludge to be filtered for solids content.
c) Measure a volume of sludge that will yield approximately
4 grams of dry solids in a volumetric flask.
d) Select the conditioning chemicals to be utilized and add a
predetermined amount to the sludge to be conditioned.
e) Agitate the volumetric flask vigorously and allow the sludge
to sit 2 minutes. Always agitate the sludge consistently
during a test series.
f) Add the sludge to the funnel and quickly apply vacuum. As
soon as vacuum is applied, start the stopwatch. A vacuum
reservoir may be needed to hold a constant vacuum.
g) Take filtrate volume reading versus time.
h) Continue the test until the cake cracks, or until no filtrate
is obtained for a one minute interval. Usually five minutes
is a sufficient time. Be sure that the cake edges do not
shrink from the sides of the Buchner funnel. If they do, tamp
the edges of the cake to maintain a seal.
i) Sample cake for total solids.
j) Record filtrate temperature, vacuum level, and cake thickness.
k) Plot a curve of time/volume filtrate vs. volume filtrate and
record the slope of the curve. This slope should include only
the linear portion of the curve.
1) Calculate specific cake resistance from the formula:
r = 2PA2b/yw (15)
113
-------
where
r = specific resistance in sec /gm
p = vacuum level in gm/cm^
A = area of Buchner funnel in crn^
b = slope of t/V vs V curve in sec/cm6
y = viscosity in Poise
w = weight of solids/unit volume of filtrate
m) Repeat step a) through 1) for several dosages of the same
chemical.
Vacuum filter test leaf procedure:
The following apparatus are required: test leaf filter cloth, vacuum
supply, rubber hoses, vacuum bottle, vacuum gauge, scales and graduates,
etc. The time for filtration is primarily based on forming a cake of
sufficient thickness. The experimental procedures are shown in the
following:
a) Condition 2 to 4 liters of sludge according to Buchner-Funnel
test results.
b) Place media selected from media screening test on the filter
leaf and attach leaf hose to filtrate receiver.
c) Crimp the hose connecting the leaf to the vacuum source and
set vacuum to desired level with the bleeder valve.
d) Immerse the leaf in the sludge so that the surface of the
leaf is 2 to 3 inches below the sludge level. Open the stop
in the hose and start the stopwatch simultaneously.
e) Keep the leaf submerged for a predetermined pick up time, as
obtained from preliminary tests. For thin sludges, move the
leaf slowly in a horizontal plane with a circular wrist move-
ment at a rate of 6 rpm. In thick sludges, the leaf should
remain stationary. Keep thin sludges mixed with a small mixer.
Thick sludges should be thoroughly mixed prior to the test.
f) At the end of the pick up time the leaf is rotated out of the
bucket.
g) The leaf is then held with the cake on top for the duration of
the drying cycle. At the end of this time the vacuum is
114
-------
released. Adjust the vacuum, as necessary, during the dry
time to maintain vacuum level. Allow filtrate to drain from
the hose to the filtrate receiver.
h) Remove the cake from the filter leaf by blowing into leaf
hose and dislodging it with a spatula. Analyze the cake for
total solids, Note cake discharge characteristics and
thickness,
i) Analyze filtrate for suspended solids, and record the filtrate
volume.
j) Analyze solids content of remaining sludge,
Analytical Procedures
Conductivity, pH, oxidation reduction potential (ORP), total solids,
volatile solids, total suspended solids, volatile suspended solids,
chemical oxygen demand (GOD), chloride, orthophosphate-, total phosphate,
sulfate, ammonia nitrogen, nitrate, and organic nitrogen, as well as
tannins were all determined according to Standard Methods. Total organic
carbon (TOC) was determined with a Beckman Model 915 Total Organic Carbon
Analyzer (Fullerton, CA), Calcium, magnesium, iron, zinc, copper,
cadmium, chromium, lead, nickel, sodium, and potassium concentrations
were analyzed using a Beckman Model 485 Atomic Absorption Spectrophoto-
meter (Fuller-ton, CA), Total dissolved solids (TDS) were measured with
a Myron DS meter (Encinitas, CA). Volatile acids were determined using
gas-liquid chromatography with a column consisting of 20% neopentyl
glycol succinate and 2% phosphoric acid
-------
room temperature (24°C). The RO retentate so obtained was fractionated
using Sephadex G-75 and Sephadex G-25 (Pharmacia, Piscataway, NJ) columns.
G-75 is useful for separations of molecular weights between 1,000 and
50,000, whereas G-25 is applicable for molecular weights of 100 to 5,000.
The total organic carbon (TOC) analysis was used to establish the dis-
tribution of the organic matter in the different molecular weight fractions,
116
-------
RESULTS AND DISCUSSION
Aerated Lagoon Treatment of Leachate
The aerated lagoons were started with distilled water to which acclimated
seed was added in addition to the daily amounts of leachate. The dis-
tilled water was thereby gradually displaced by added leachate resulting
in a gradual increase in effluent TOC values and mixed liquor solids.
The volatile matter in the mixed liquor suspended solids gradually
increased and stabilized after about one volume turnover in all three
units (Figure 37). The stabilization of the MLVSS coincided with the
leveling-off of the TOC in the effluents (Figure 38). The MLVSS in
unit 3 (having a 30 day detention time) stopped increasing after 30
days while the TOC leveled off after 40 days. Although the predicted
nutrient demand required the addition of 347 mg/£ P, this was decreased
by 45% during the initial period to a concentration of 189 mg/l to be
added to the leachate. This amount resulted in a COD:P ratio of 300:1.
However, after 60 days the TOC started to increase again which coincided
with a which coincided with a 72% reduction of the influent phosphate
added to unit 3 to 67.4 mg/£ corresponding to a COD:P ratio of 806:1.
As expected, the phosphate concentration in the effluents of all three
units decreased after the reduction of the concentration in the feed
(Figure 39). The increase in effluent TOC in unit 3 may therefore be
attributed to the phosphate reduction in the feed for the remainder of
the testing period. The decrease in phosphate addition, however, did
not result in a corresponding increase of the effluent TOC in units 1
and 2, which had lower organic loadings. The TOC in effluents of units
1 and 2 actually approached a steady state level as the nutrient addi-
tions were stepwise reduced. The lowest amount added to the influent
of the 60 and 85.7 day lagoons was 33 mg/£ P corresponding to a COD:P
ratio of 1540:1. Although the biological MLVSS of effluent TOC values
did not show any deterioration, the sludge settling
decreased rapidly. This may indicate that at high detention times
(60 and 85.7 days), microorganisms are able to reuse the added phosphate.
This reuse originates from dead cells or from the decomposition of high
117
-------
00
12000
o
tf)
•glQDOO
~o
0>
t 8000
to
~ 6000
0
,_ 4000
§
.tr
•? 2000
i I r
i i
Unit 3 (0 = 30 days)
Unit 2 (0 =60 days)
°0
-Unit I (0=85.?days)
I I I III
I I
I II
10 20 3D 40 50
60 70 80 90
f ime, days
100 110 120 130 140 150
Figure 37. Mixed liquor volatile suspended solids concentration In aerated lagoons
1, 2 and 3 treating leachate
-------
400
Unit I (9 =85.7 days)
60 70 80 90
time, days
Figure 38. Total organic carbon in effluent of
aerated lagoons 1, 2 and 3 treating leachate
-------
ro
o
Unit I
Unit 2
Unit 3
fUnM
< Unit
lUnit 3
4.05 mg/f PQ;-P per day
5.80
11.60 mg/< P04 -P per day
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Figure 39. Effect of reduction of daily phosphate on total-p and COD/TOC
ratio in effluent of aerated lagoons 1, 2 and 3 treating leachate
-------
molecular weight organic compounds to which phosphates are attached.
The COD/TOC ratio of the effluent from unit 3 showed a significant
drop after the decrease of nutrients which decrease was not observed
for units 1 and 2 (Figure 39). The decrease in phosphate concentration
may have resulted in the decrease in COD/TOC ratio, as a phosphate limi-
tation in the filtered mixed liquor may give rise to lysis of bacterial
cells, which, in turn, release the biologically resistant material,
characterized by a low COD to TOC ratio, into solution.
Units 4, 5, and 6 were started with acclimated sludge obtained from
units 1, 2, and 3, respectively. When sufficient nutrients were added
during the first 35 days, both the MLVSS (Figure 40) and the effluent
TOC reached constant values (Figure 41). This indicates that even at
a low detention time of 7 days, the aerobic biological method can be
used to treat leachate, if enough nutrients are added. When the addi-
tion of nutrients was completely stopped, units 4 and 5 gave high
effluent TOC values, corresponding to gradually decreasing MLVSS, as
shown in Figures 40 and 41. Unit 6 showed a very sharp increase of
effluent TOC values while the sludge obtained showing poor settling
properties. This shows that aerobic biological treatment of the leach-
ate cannot be successful at high organic loading and low detention time
without nutrient addition. This confirms the results of earlier studies
by Boyle and Ham (1974) and Foree and Cook (1974).
The COD to TOC ratios in the filtered effluents of units 4, 5, and 6 are
shown in Figure 41 and indicate that within the initial 35 days, three
units obtained constant COD/TOC ratios. Lower detention times resulted
in higher COD/TOC ratios which are to be expected as aeration increases
the oxidation state of the carbonaceous organic material. After the
nutrient addition was stopped, the COD to TOC ratio in the filtered efflu-
ents of units 4 and 5 decreased indicating the release of oxidized
organics. The increase in COD/TOC ratio in unit 6 would indicate that
such mechanism is not operating at higher detention times.
All six extended aeration units showed high percentages of removal of the
121
-------
1600
Without Nutrients
Urnt5Cff=l5daysl
Unit 4 (8 = 3Odays)
600
o
& 400
200
0
2O
4O
Time, days
50
6O
Figure 40!. Wfxetf Ifquor volatfTe suspended salfdis concentiratfon
in aea-ated; Tagoorts 4» 5 and 6 treatfng! leachate
122
-------
D Unit 4 Cff=3Odays)
UrafS (0 = 15 days)
O Units (ff= 7days*
3D- 4O 50
Time „ days
Figure $1. Total organic carbon and GOD/TOE ratio in. effluent
of aerated lagoons $, 5 and 6 treating; leachate
T23
-------
metals or cations present in leachate (Table 6). The iron concentration
in the feed solution was reduced from 2125 mg/£ for units 1, 2, and 3
or 1020 mg/£ for units 4, 5, and 6 to less than 1.2 mg/£. The high iron
removal was attributed to both high pH values of around 9 and the oxida-
tion of Fe (II) to Fe (III). A similarly large removal was observed for
zinc and to a lesser extent for both calcium and magnesium. Removals
of sodium and potassium were generally less than 50 percent. Calcium
removal will take place as calcium carbonate and calcium phosphate. The
removal of magnesium was less than calcium, since magnesium hydroxide
and magnesium ammonium phosphate (MgNH^PO^) are not precipitated sig-
nificantly at pH values less than 10. Only small amounts of sodium and
potassium were removed by any of the aerated lagoons since sodium and
potassium requirements by biological processes are generally small.
Data from these six units were used to evaluate the kinetic constants
from Equations (7) and (11). Table 7 lists results of TOC, COD and MLVSS
analyses at the plateau region which correspnd to the first volume turn-
over with untis 1, 2, and 3. The average values of units 4, 5, and 6,
with sufficient nutrients added, are also listed.
In determining the constant K], Equation (7) was rearranged so that a
graph of (SQ-S)/(Xe) versus S could be plotted. Straight lines were
obtained from such a graph as shown in Figures 42 and 43; values of K.J
were calculated from the slope of these lines. The value for 1C! was
found to be 4.9 x 10"4 l/(mg/£ VSS)(day). In addition, values for Y
and k^ were determined by rearranging Equation (11) for graphing as
shown in Figures 44 and 45. In determining Y and k., the contact time
or the reciprocal of detention time 1/e was plotted versus (SQ-S)/(Xe),
in which Y represented the slope and k. the negative intercept of the
ordinate. These values were found to be Y = 1.29 mg VSS/mg TOC or
0.42 mg VSS/mg COD, and kd = 0.025 day . A summary of the kinetic
characteristics is shown in Table 8. The values of the yield constant,
Y, can decay rate, kj, are close to those found by Cook and Foree (1974)
as can be seen in Table 8.
124
-------
rvj
en
Table 6
Concentration of Heavy Metals in Effluent of Aerated Lagoons (mg/1)
Ele-
ment
Fe
Ca
Mg
Zn
Na
K
Influent
mg/1
2125
3780
660
72
1350
1240
Unit 1
6 = 85.7 d
Days of
Operation
90 120 150
0.12 ND ND
70 25 20
120 125 165
0.05 0.06 0.06
730 900 1000
1030 900
Unit 2
e = COD
Ave. % Days of Operation Ave. %
Removal 60
99.3 0.3
99.0 15
79.2 58
99.9 0.25
35.0 710
22.2 1060
90 120
ND ND
20 20
85 150
0.07 0.09
970 1070
1080
150 Removal 32
ND 99.9 0.3
40 99.4 8
200 81.9 40
0.07 99.8 0.10
1180 27.2 485
710 23.4 620
Unit 3
6 = BOD
Days of Operation
60 90 120 150
0.5 ND ND
26 40 45
94 80 170
0.15 0.09 0.07
710 1140 1130
1030 - 940
ND
45
240
0.11
1160
790
Ave. %
Removal
99.9
99.1
81.1
99.8
31.5
31.9
Ele-
ment
Fe
Ca
Mg
Zn
Na
K
Influent
mg/1
1020
3010
308
55
812
503
Unit 4
9 = 30 d
Days of
Operation
27
1.2
22
80
0.05
650
450
Ave. %
Removal
99.9
99.3
74.0
99.9
19.9
10.5
Unit 5
9 = 15 d
Days of
Operation
27
ND
12
65
ND
660
465
Ave. %
Removal
99.9
99.6
78.9
> 99.9
18.7
7.6
Unit 6
6 = 7 d
Days of
Operation
27
0.3
12
120
ND
710
470
Ave. %
Removal
> 99.9
99.6
61.0
> 99.9
12.6
6.6
Ave. %
Removal
of all
Units
> 799.9
99.3
75.9
99.9
24.1
17.0
-------
ro
Table 7
Characteristics of Effluent from Aerated Lagoons with Sufficient Nutrient Addition
Characteristics Feed
TOC, mg/1 19,400
Percent TOC Removal
COD, mg/1 57,900
MLVSS, mg/1
6, day
pH 5.4
SVI with sufficient
nutrient addition
Unit 1
160
99.2
415
8000
85.7
8.77
28.4
Unit 2
180
99.1
540
9000
60
8.72
31.4
Unit 3 Feed
240 11,773
98.8
666 35,237
10,000
30
8.7 5.7
24.4
Unit 4
210
98.2
536
9500
30
8.8
14.4
Unit 5
310
97.4
822
11,500
15
8.6
21.2
Unit 6
380
96.8
1034
13,500
7
8.5
27.3
SVI without - - 26.2 47.9 141.6
sufficient
nutrient addition
-------
Table 8
Kinetic Constants of Aerated Lagoon Treatment of Leachate
with Sufficient Nutrient Addition
Constants
Values
Cook and Foree (1974)
4.9 x 10~4 (mg/1 VSS)"1 (day)"1
1.29 mg VSS/tng TOC and 0.42 mg
VSS/mg COD
0.025 day
-1
0.4 mg VSS/mg COD
0.05 day
-1
127
-------
00
QI5
o>
5* o.io
•o
en
to
o>
en
I
i
en
'ox 0.05
D Data From Units I, 2 And 3
O Data From Units 4,5 And 6
100
Slope = K! = 4.9 x lO^mg/J?
200
S, mg/Jf TOC
300
400
Figure 42. The calculation of substrate removal rate constant based on TOC data
-------
ro
0.4
a
O
O
o>
E
">,
I II
of
CO
0.3
Q2
O.I
1 O
D Data From Units 1,2 And 3
O Data From Units 4,5 And 6
a
Slope = K! = 4.9xlO"4(mg/j?VSSr'(dayy
250
1000
500 750
S, mg/S. COD
Figure 43. The calculation of substrate removal rate constant based on COD data
1250
-------
0.15
O.K>
I >»
I °
1*0
-l<
co
o
Q05
-i-Q
a
o
Data From Units I, 2 And 3
Data From Units 4,5 And 6
Slope = Y = 1.29
mgVSS
mgTOC
0 /
kd=0.025 day'1
Q05
QIO
mg/Jt TOC
(mg/l VSS)(day)
0.15
020
Figure 44. The calculation of growth-yield and microorganism-decay coefficients
based on TOC data
-------
Ql!
0.10
-i-
O.05
D Data From Units 1,2 And 3
O Data From Units 4,5 And 6
X
1^=0.025 day
Slope = Y = 0.42
I
mgVSS
mgCOD
0.10 020
-I S-Sp rng/J? COD
X0
030
(mg/l VSSMday)
040
Figure 45. The calculation of growth-yield and microorganism-decay coefficients
based on COD data
-------
Sludge Settling and Dewatering Characteristics
The six aerated lagoon units had a very high concentration of suspended
$olids which resulted in hindered settling characteristics and formation
of a settling zone. Figure 46 shows settling velocities for a range of
concentrations of sludges obtained from the aerated lagoons and other
conventional domestic activated sludges, primary sludge, digested
sludge and lime softening sludge (Weber, 1972). For the same range of
sludge concentrations, sludge from aerated lagoon units with nutrient
addition had better settling characteristics than primary sludge and
digested sludge. In comparing sludge settling characteristics, it was
noted that influent leachate without nutrient addition resulted in poorer
sludge settling properties. Figure 47 shows a rapid decrease of the
settling velocity using sludge from units 4, 5, and 6 at varying initial
sludge concentrations. Table 7 shows the SVI values of each unit before
and after the nutrient additions were stopped. A SVI larger than 100
would indicate poor settling characteristics of the sludge.
The sludge dewatering characteristics were measured with the Buchner-
Funnel Test. The resulting graphs of t/V versus V were constructed
using the experimental results. The volume of filtrate was obtained by
subtracting that portion filtered before the 0 minutes reading from all
the subsequent readings. This initial time period is required for the
cake to form. The slope of the graphs was measured while the specific
cake resistances were calculated from Equation (15). Table 9 and
Figure 48 show the results of these experiments using a vacuum of 37 cm
of mercury. All the chemicals reduced the specific resistance considerably.
The quantity of each chemical used to reach an arbitrary low resistance
is in the order of Nalco 73C32 < Primafloc C-7 < FeCl3 • 6 H20 < Ca(OH)2.
While the addition of chemicals increased the specific resistance of
the sludge, the actual solids content also increased as a result of the
addition. At very high dosages the solids content often decreased.
The filter leaf test was operated with a vacuum of 47 cm of mercury,
and the cycle time was chosen as two minutes for forming and three minutes
for drying. The relation between the chemical dosage and the filter
132
-------
0.05
QOI
£
u
o
o
I
O)
I
Activated
Sludge A
(Dick And
Ewing )
Activated
Sludge C
(Dick And
Ewing)
Data From
-------
0.05
0.01
1 — i — i — i i i
cm/sec
u
J» 0.005
0>
TJ
°35
.O
3
0.001
With
Nutrient
Addition
Without
Nutrient
Addit ion
D, Unit 4 (On The 28 Day Of Operation)
E. Unit 5 (On The 28 Day Of Operation)
.F: Unit 6 (On The 28 Day Of Operation )
Unit J jOn The 59 Day Of Operation )
Unit 5 (On The 59 Day Of Operation)
Unit 6 (On The 59 Day Of Operation)
p;:
I E!
IF';
1 2 4 6 8 10 20 40"
Suspended Solids Concentration, g/jf
Figure 47. Effect of omission of nutrient addi
settling velocities of sludges from
>t b and 6 treating leachate
60 80 100
on
134
-------
I09rr
M
8 io*
'«
&
u
u
0)
o.
CO
0)
o>
CO
Primafloc C7
Nalco 73C32
0.05 QIO 0.15 O20 025 0.30
Chemical Dose Per Quantity Of Dry Sludge, g/g
Figure 48. Effect of chemical doses on specific resistance of
sludge from aerated lagoon 4 treating leachate
135
-------
Table 9
CO
Results of Buchner-Funnel Test using Mixed Liquor from the Aerated Lagoon4
c
Type of ,
Chemical *g
-
Fed 3- 6 H20
Ca(OH)2
Nalco 73C32
Primafloc C7
hemical Dosage
gm dosage
m of
0
0
0
0
0
0
0
0
0
0
0
0
dry sludgi
0
.0287
.0573
.1433
.0510
.1020
.2550
.0015
.0031
.0092
.0153
.0053
.0079
Volume of
x Sludge
*> (ml)
100
100
100
100
100
100
100
100
100
100
100
100
100
Cone, of S.S.
(mg/1)
32673
34892
34892
34892
39215
39215
39215
32673
32673
32673
32673
38128
38128
Volume of Dry Cake
Filtrate Weight
(ml) (gm)
72
73
73
76
73
74
75
73
75
76
80
74
75
3.99421
4.23489
3.87031
3.76318
4.15453
4.37084
4.70723
4.27093
4.15211
4.15211
3.91221
3.39847
3.75441
Solid
Content
W(g/cm3)
0.055476
0.048013
0.053918
0.049516
0.056912
0.059066
0.063612
0.059506
0.055362
0.054633
0.048903
0.045926
0.050059
Slope of
t/V vs. V
b(sec/cm6)
0.35715
0.035715
0.0075
0.001667
0.038
0.015625
0.001667
0.015
0.067813
0.003572
0.0015
0.0082
0.005
Specific
Resistance
r(sec2/g)
1.35 x
1.276
2.933
6.98 x
1.384
5.484
5.433
5.315
2.926
1.355
6.359
3.702
2.071
x
x
X
X
X
X
X
X
X
X
X
108
108
107
106
108
107
106
107
107
107
106
107
107
AP = 37 cm
Area = 45.36 cm2
v = 0.01 poise
-------
yield is shown in Table 10 and Figure 49. These results indicated that
the filter yields give results similar to those obtained with the
specific resistance test. Both tests show that the dosages of lime and
ferric chloride were more than 15 times those of the polymers to reach
an identical filter yield. The actual cost should be evaluated on the
basis of chemical dosages and filter yields. The total cost is reflected
not only by the cost of the chemicals, often the largest single item in
the cost of filter operation, but also by the added hours of operation
that are required.
Effluent Organic Matter Characteristics
In general, most of the organic matter in the effluent of aerated lagoon
aeration units consist of stable refractory materials often having high
molecular weights. Results of the organic analysis of the effluent from
unit 4 indicated that 97.6% of the TOC was retained by the NS-100
membrane which has a molecular weight cut-off of around 100. The
elution patterns of the membrane retentate on Sephadex G-75 and G-25
are shown in Figures 50 and 51, respectively. The high molecular weight
fraction (larger than 50,000 MW) was eluted first from the Sephadex 6-75
column and represented about 3.7% of the TOC in the effluent. This
high molecular weight fraction consists mainly of humic acid-like
materials. The remainder of the eluates from the Sephadex G-75 column
of the membrane retentate, representing 93.9% of the original effluent
TOC, consisted of substances lower than 50,000 MW but higher than 100 MW
due to the MW cut-off of the NS-100 membrane. Elution of the NS-100
membrane retentate on Sephadex G-25 showed that 24% of the effluent TOC
was larger than 5000 MW as it was excluded from that Sephadex column.
Since 3.7% of effluent TOC is larger than 50,000, 20.3% of the effluent
TOC has a molecular weight range between 5000 and 50,000. The organic
fraction in the last peak from the G-25 column represents 33% of the TOC,
and can be assumed to have a molecular weight between 100 and 500. Most
of the organic matter in the extended aeration effluent was present in
the peak eluted in the middle, representing about 40.6% of TOC, and
having a molecular weight range between 500 and 5000. DeWalle and Chian
(1974) defined fulvic acid material as having a molecular weight ranging
137
-------
Table 10
Results of Filter-Leaf Test Using Mixed Liquor from the Aerated Lagoon 4
Type of ,g
Chemical v
FeCl 3-6 H20
a(OH)2
Nalco 73C32
Primafloc C7
Chemical Dosage
m of chemical dosage\
gm of dry sludge '
0
0.0275
0.057
0.143
0.05
0.102
0.255
0.0017
0.0028
0.0085
0.0045
0.0081
Filtrate
Volume
(ml)
88
96
197
370
70
151
380
166
180
400
137
270
Dry Cake
Weight
(gm)
1.64
2.572
5.013
9.039
1.715
3.914
11.108
4.137
4.771
9.244
3.653
5.218
Filter
Yield
(Ib/ft2/hr)*
0.44
0.69
1.345
2.425
0.46
1.05
2.98
1.11
1.28
2.48
0.98
1.40
AP = -47 cm
Filter cloth: Eimco NY-527F
Area = 650 cm2
* Ib/ft2/hr = 1.36 x 10"4 gm/cm2/sec
138
-------
15,
Co (OH)2
N
FeCI3 • 6H2 0
0.01
0.05 0.10 0.15 0.20 0.25
Chemical Dose Per Quantity Of Dry Sludge, g/g
0.30
Figure 49. Effect of chemical doses on filter yield of sludge
from aerated lagoon 4 treating leachate
139
-------
30
40
50 60 70 80 90
Eluate Volume , mJI
100
110
120
Figure 50. Elution profile of the NS-100 membrane retentate on
a G-75 Sephadex column as characterized by total
organic carbon
140
-------
20
40 60 80 100 120 I4O
Eluate Volume, ml?
Figure 51. Elution profile of the NS-100 membrane retentate on
a G-25 Sephadex column as characterized by total
organic carbon
141
-------
from approximately 100 to 10,000. It can thus be concluded that most
of the organic matter in effluents from these extended aeration units
is present as a fulvic acid-like material.
The color of the effluent from the aerated lagoon 4 was yellow, and
equivalent to 375 APHA platinum cobalt units. Figures 52 and 53 show
the absorbance spectra in the visible and ultraviolet range, using a
Beckman ACTA III UV-visible recording spectrophotometer. No absorbance
peaks or maxima were observed. To examine the conformity to the Beer-
Lambert Law, absorbance values at 400 nm were taken on a serial dilution
of effluent from unit 4. Figure 54 shows that the results follow the
linear relationship of the Beer-Lambert Law.
142
-------
1.4
Test Conditions
1.2
Slit! Program
Seam Speed : Inm/sec
Chart Speed ! 50 nm/sec
Span : 2
1.0
8
o
.a
Q8
0.6
Wave Length Used
To Measure Color
intensity
0.4
Q2
§bo
I
400 500
Wave Length, nm
600
Figure 52. Visible spectrum of effluent from aerated lagoon 4
treating leachate
143
-------
o
.0
w
O
J5
41--
0.8
0.6
0.4
0.2
10)
Test Conditions
Sample Diluted ; I To 11
Slit ! Program
Scan Speed ! I nm/sec
Chart Speed ! 50 nm/inch
Span I 2
200 300
Wave Length, nm
400
Figure 53. Ultraviolet spectrum of effluent from aerated lagoon 4
treating leachate
144
-------
in
0.5
0.4
I
E
c
8
Q3
8 0.2
o
x»
o
JS
<
0.1
Q50 0.75
Concentration Factor, C/C0 (-)
1.00
Figure 54. Absorbance at 400 nm of a serial dilution of effluent from aerated lagoon 4
treating leachate
-------
REFERENCES
Boyle, U. C. and Ham, R. K. "Biological Treatability of Landfill
Leachate," J. Water Pollution Control Fed., 46_, 860 (1974).
Chian, E. S. K. and DeWalle, F. B. "Sanitary Landfill Leachates and
Their Treatment," J. Environ. Engr. Div., ASCE, 102, 411 (1976).
DeWalle, F. B. and Chian, E. S. K. "Removal of Organic Matter by
Activated Carbon Columns," J. Environ. Engr. Div., ASCE, 100. 1089 (1974).
Eckenfelder, W. W. and O'Connor, D. J. "Aerobic Biological Treatment
of Organic Wastes," Proceedings 9th Annual Purdue Industrial Waste
Conference, Lafayette, Indiana (1954).
Foree, G. G. and Cook, G. M. "Aerobic Biostabilization of Sanitary
Landfill Leachate," J. Water Pollution Control Fed., 46, 380 (1974).
Karr, P. R. "Treatment of Leachate from Sanitary Landfill," Special
Research Problem, Department of Civil Engineering, Georgia Institute
of Technology, Atlanta, Georgia.
Salvato, J. A., Jr., ejt aJL "Landfill Leaching Prevention and Control,"
J. Water Pollution Control Fed., 43, 2084 (1971).
Salvato, J. A., Jr. "Environmental Engineering and Sanitation," Wiley-
Interscience, New York (1972).
Pohland, F. G. and Kang, S. J. "Sanitary Landfill Stabilization with
Leachate Recycle and Residual Treatment," AIChE Symposium Series Water-1974,
145. 71, 308 (1975).
Schoenberger, R. J., e£ al_. "Treatability of Leachate from Sanitary
Landfills," Proceedings 4th Mid-Atlantic Industrial Waste Conference,
University of Delaware, Newark, Delaware, Nov. 18-20 (1970).
Standard Methods for the Examination of Water and Wastewater. APHA, AWWA,
WPCF, 13th ed., American Public Health Association, New York (1971).
Weber, W. J., Jr. Physicochemical Processes for Water Quality Control,
Wiley-Interscience, New York (1972).
146
-------
IV
PHYSICAL CHEMICAL TREATMENT OF LEACHATE AND AERATED
LAGOON EFFLUENT
CONCLUSIONS
The present study noted that physical chemical treatment methods cannot
remove large quantities of organics from leachate. Such methods, however,
are effective after biological pretreatment of the leachate. Different
physical chemical treatment methods were therefore evaluated using aerated
lagoon effluent. While ozonation only removed 48% of TOC after a 3 hour
period of ozonation, activated carbon columns were able to remove 86% of
the organic matter using an empty bed detention time of 3.7 minutes. A
59% initial COD removal was realized with a weak base anion exchange
resin, while 82% to 85% of the COD were initially removed using strong
base anion exchange resins. Reverse osmosis (RO) was the only process
able to remove 91-96% of the salts initially present at a TDS concentra-
tion of 6200 mg/1. The organic matter removal by RO ranged from 85% to
97%, which removals were not enhanced by ion exchange or activated carbon
pretreatment. The flux through the membranes was relatively high and the
membrane fouling was relatively insignificant if the suspended solids were
removed from the influent. Sand filtration or chemical precipitation are
therefore necessary pretreatment processes prior to the reverse osmosis
units.
147
-------
INTRODUCTION
Formation of leachate from desposited solid waste will cause deteriora-
tion of the environment near the landfill unless proper solutions are
provided to minimize the leachate's impact. One of the possible alterna-
tives consists of treating the generated leachate after its collection
below the landfill with biological and physical chemical treatment methods.
As the generated leachate decreases in concentration and changes in com-
position with increasing age of the fill, different treatment methods are
only applicable during certain time periods of leachate production. In
summarizing different treatment studies conducted elsewhere and at the
University of Illinois, Chian and DeWalle (1976) concluded that biological
treatment methods, such as anaerobic filter, aerated lagoon and combined
treatment, are most applicable for treating leachate from recently
constructed fills while physical chemical treatment methods, such as
activated carbon adsorption, chemical precipitation, chemical oxidation
and reverse osmosis are best applied to biologically stabilized leachate.
Biologically stabilized leachate is generated in old fills in which the
solid waste has been subjected to extensive degradation. Such leachate,
however, is also equivalent to effluent from the biological treatment
units receiving leachate generated from recent fills. The present study
was primarily concerned with the physical chemical treatment of aerated
lagoon effluent.
148
-------
PHYSICAL-CHEMICAL TREATMENT OF LEACHATE AND BIOLOGICAL EFFLUENTS
Several studies have evaluated the effectiveness of chemical precipitation
in leachate treatment. Thornton and Blanc (1973) studied the treatment of
leachate by lime, alum, ferric chloride, ferrous sulfate and polymers.
Although high removal percentages of heavy metals and suspended matter
were observed, only slight reductions of organic matter were found.
Simensen and Odegaard (1971) used aeration in combination with coagula-
tion to enhance floe formation, but no additional organic removals were
noted. Chian and DeWalle (1976) indicated that lime treatment predominantly
removed organic matter with a molecular weight larger than 50,000. This
organic fraction is initially present in relatively low concentration and
increases as a percent of the organic matter in leachate during biological
stabilization of the refuse. As a consequence, the COD removal by lime
treatment will increase as solid waste stabilization proceeds. Eventually,
this high-molecular-weight fraction is slowly degraded and converted into
organic matter in the 500 to 10,000 molecular weight range. Since the
latter fraction is not greatly affected by lime treatment, this method
becomes less effective with leachate from stabilized solid waste.
Activated carbon treatment of leachate generally gives better organic
removal than observed with chemical precipitation. Ho et^al- (1974)
observed a 34 percent COD removal with a powdered activated carbon dosage
up to 16,000 mg/1. A COD removal of 55 percent, after elution of two bed
volumes and a detention time longer than 20 minutes, was achieved in a
column test. A 60 percent TOC removal was observed in the study by Karr
(1972) using a maximum dosage of 160 g/1. Because of the high organic
strength, its low adsorptive capacity, and presence of suspended solids
in raw leachate, fouling of carbon columns was reported by Chian and
DeWalle (1976b). The activated carbon process becomes prohibitively
expensive in practice for raw leachate treatment.
149
-------
Ion exchange processes were considered only for ammonia removal in
leachate with a low organic concentration. Weston (1974) reported that
ion exchange was not a feasible means of treatment for ammonia removal,
due to interfering substances which compete for the resin bed bonding
sites. No study has been conducted for organic removal from leachates
using ion exchange.
Reverse osmosis processes were also studied by Weston (1974) using
leachate with a low organic concentration. The COD removal attained
was approximately 80 percent, while total dissolved solid removal was
81.2 percent using DuPont's Permasep system. The problem of disposing
a highly concentrated, and voluminous RO retentate was reported. Only
if subsequent tests indicated high product water yields, improved
permeate quality, and high treatability for the retentate, would reverse
osmosis processes be feasible for leachate treatment with low organic
concentrations.
Only two studies evaluated the use of an oxidant as a means of removing
organic matter in leachate. Ho e_t al_. (1974) found only a 6 percent COD
reduction in leachate having an initial COD of 7,162 mg/1, using an ozone
dosage rate of 12 mg 0^/100 ml/min. The percentage of removal increased
to 37 percent after 4 hours. COD removals of a similar magnitude were
observed with more stabilized leachate. Weston (1974) reported a COD
removal of 22 percent after a four hour test at a cumulative ozone
dosage of 1 mg 0-/mg COD. Both tests, however, demonstrated that ozona-
0
tion was less effective than activated carbon in removing the organic
matter from leachate.
The studies summarized above show that physical-chemical treatment does
not result in large reductions of organic matter in raw leachate. They
are effective, however, for the removal of heavy metals, turbidity and
color. Since physical-chemical methods are more effective in removing
residual organics, Chian and DeWalle (1976b) indicated that these methods
might be feasible for treating leachate from old and stabilized landfills.
150
-------
The degree of treatability can be correlated to established ratios,
such as the ratios of COD/TOC, BOD/COD, Organic-N/Kjeldahl-N, TVS/TS
or S04/C1 in leachate. Since these ratios gradually change with the
age of the leachate, they can be used in predicting the optimum selec-
tion of treatment methods. Leachate from recently deposited solid waste,
should first be treated by biological methods due to the presence of high
concentrations of degradable volatile acids followed by physical-chemical
methods.
Foree and Cook (1974) found that activated carbon resulted in removals
bf 47-70 percent of the COD of the effluent of extended aeration units,
using a residence time in the carbon column of 15 min. Chlorine bleach
was effective for color removal but ineffective for COD removal in the
effluent of extended aeration units. Pohland and Kang (1975) have studied
powdered activated carbon treatment of effluent from an activated sludge
unit receiving leachate. They found a 70 and 90 percent COD removal at
a dosage of 2,000 mg/1 and 4,000 mg/1 powdered activated carbon respec-
tively. Large dosages of mixed ion exchange resins were necessary to
remove the inorganic ions. Actually, the use of resins as a polishing
step for the effluent of biologically treated leachate is economically
unfeasible if the feed water has a TDS of more than 200 mg/1 (Ahlgnen,
1971).
Removal of Organics by Ozone
A gaseous mixture of oxygen and ozone is bubbled through a vessel contain-
ing an aqueous solution of organics. The ozone dissolves in the liquid
and then reacts with and subsequently oxidizes the organic materials.
The principal reaction of ozone lies in its destruction of an ozonide.
Decomposition of the ozonide gives a mixture of oxygenated products
containing carbonyl compounds and acids (Weber, 1972).
Aromatic or polynuclear aromatic compounds are more susceptible to ozone
oxidation due to the aromatic carbon bonds. Since the reactions produced
with ozone are complex, the composition of the oxidation products becomes
difficult to predict. Some aromatic compounds may be oxidized first to
151
-------
form quinones. Saturated hydrocarbons react slowly with ozone at room
temperature, but, at elevated temperatures, the reaction proceeds quite
rapidly. In the reaction mixture, peroxides, ketones, aldehydes, alcohols,
and acids are found. Ethers are oxidized by ozone at the carbon atom
next to the ether oxygen. Esters are therefore found among the oxidation
products. Lactones are formed from cyclic ethers, and carbonates from
cyclic formals. Organic sulfides are oxidized by ozone through sufoxides,
RSOR', to sulfones, RS02R'. The intermediate sulfoxide may, at times,
be isolated. Primary and secondary amines are only partially degraded
by ozone, but tertiary amines form tertiary amine oxides. Organic
phosphates may be prepared by ozone oxidation of the phosphites, and
phosphine oxides are formed from phosphines.
The ozonolysis and oxidation reaction generally require that stoichio-
metric amounts of ozone are reacted. However, there are catalytic
reactions in which experimental conditions determine the amount of ozone
consumed. For example in the preparation of peroxyacids from aldehydes,
ozone is only used as a catalyst or initiator of the oxidation.
When organic compounds in effluent from extended aeration units are
oxidized by ozone, the reactions become very complicated. Thus the
degradation of organics can be measured only by their gross properties
such as measured by TOC or COD. The rate of organics disappearance
can be expressed as:
^ - k[03f [TOC]n (1)
where m and n are the exponentials to the concentration terms. As such,
the instantaneous concentration of ozone, [0,], in the effluent of
extended aeration units has to be measured. The overall ozonation process
may be either chemical reaction rate controlled or mass transfer rate
controlled, depending on the relative magnitudes of the rates for each
step. The heterogeneous reaction for the transfer of ozone into solution
can be represented by:
152
-------
d[03] -
-gr-= «KLa (6[03r - [03]) - Y (2)
where aKLa (e[03] - [03]) = rate of 03 dissolution due to mass transfer
a,3 = correction factors for K|_a and [03JS, respectively
[03] = saturation concentration of ozone
Y = chemical decomposition rate of 0, in solution.
If the concentration of organic matter is high, the ozone decomposition
rate will be very large due to the large demand of ozone for chemical
reaction. The ozone concentration in solution, [03], will be therefore
approaching zero. From Equation (2) under steady-state transfer of ozone,
d[03]/dt = 0, the rate of ozonation can be represented by:
Y = aKLa e[03]S as [03] -»• 0 (3)
where a and 8 can be determined experimentally based upon oxygen transfer
studies. [03] is calculated from Henry's Law Constant for ozone at the
experimental temperature and ozone partial pressure in the gas stream.
Removal of Orqanics by Activated Carbon
The two main mechanisms involved in organic removal by activated carbon
are transport and adsorption. Bulk solution transport involves transport
of the molecular to the carbon surface. Diffusion or film transport
involves diffusion of the molecule through the hydrated water layer, then
during intraparticle transport, the molecule moves to the adsorption site.
The adsorption process can be either physical or chemical adsorption.
Physical adsorption involves hydrogen bonding, induced dipole interactions,
or van der Waals forces. These adsorptive mechanisms are generally
reversible. Chemical adsorption involves the formation of covalent bonds.
This usually occurs at functional group sites, which are formed either
as a result of additives or impurities in the carbon. Chemisorption is
generally not reversible (Mattson and Mark, 1971).
153
-------
Many factors influence the rate of adsorption on carbon, including temp-
erature, pH, nature and concentration of adsorbate, as well as the nature
of the carbon. Adsorption reactions, being exothermic, normally decrease
with increasing temperature. pH affects the degree of ionization of the
adsorbates, and thus their solubility and efficiency of removal. Neutral
compounds are less soluble in solution than their ionized forms, and are
therefore more likely to be removed. In addition, large molecules may
have difficulty in moving through the pores of the carbon.
The surface area of the carbon, which is dependent upon the size and dis-
tribution of the pores, will have a pronounced influence on the rate of
adsorption. The type and populations of functional groups located on the
surface will likewise influence the amount and mannner of adsorption.
The nature of the carbon also depends upon the amount of inorganic
material, hydrogen, and oxygen present on its surface. Inorganic material
influences pore size and distribution, as well as many adsorptive proper-
ties. Hydrogen, likewise, is involved in many bonding reactions with
the adsorbate. Oxygen, which may make up 2 to 25 percent of the weight
of the carbon, tends to increase the polarity, thus making it more
difficult to adsorb nonpolar compounds from solution (Snoeyink and Weber,
1967). Coughlin and Ezra (1968) found that the increase in percentage
of oxygen on the surface greatly reduces the capacity to adsorb phenol at
a low concentration. Since column operations with granular carbon are
generally more efficient than batch operations with powdered carbon,
this study utilized the former to polish the stabilized organics from
extended aeration effluent.
Removal of Organics by Anion Exchange Resins
Most of the organics in secondary effluent are present in the anionic
form at neutral pH. Consequently, the Donnan exclusion effect of nega-
tively charged organics in the cation exchange resin and the competition
with inorganic anions results in very small organic removals. Therefore,
the cation exchange resin does not seem to have enough sorptive capacity
to warrant its use. In this regard, the anion exchange resins appear to
have better potential in the removal of organics.
154
-------
The kinetics of ion exchange involves transport of solute to the site
of the functional group, the exchange of ion itself, and the transport
of the exchanged ion back into bulk solution. The process of solute
transport is similar to that for activated carbon, as it involves bulk
transport, diffusion or film transport, and intraparticle transport.
There are many possible mechanisms to interpret the results of organic
removal by anion ion exchangers, however, none can be considered the sole
mechanism of importance since they all may operate in conjunction.
Adsorption of organic nonelectrolytes (nonionic species) onto the resin
matrix may be of importance. London-van der Waals forces may act as
the principal adsorption forces in this case (Helfferich, 1962;
Dorfner, 1972). The solubility of the adsorbate is of importance
because molecules which have a low solubility in the solvent will
prefer to reside at the matrix-solution interface rather than in bulk
solution. If the adsorbate structure and the resin matrix structure
are similar in nature, strong forces of adsorption are also expected.
Organic substances may also be removed via interaction with functional
groups. If the substance is an organic ion, for example, a simple
exchange between the organic ion and the counter ion of the functional
groups. If the substance is an organic ion, for example, a simple
exchange between the organic ion and the counter ion of the functional
group may take place. This requires that the functional group be ionized.
Hydrogen bonding also may occur between un-ionized functional groups and
the adsorbates. Chasanov £t al_. (1956) proposed hydrogen bonding of the
-OH group of phenol to the uncharged amine group of the resin as the
principal adsorption force in the removal of phenol by a weak base resin.
The uptake of organics is further affected by resin selectivity, influent
concentration, flowrate, column size, extent of cross-linking, temperature,
and particle size. Weak base and strong base macroreticular resins with
different matrices were chosen in the present study for column operations.
flow rates and pH values were also evaluated.
155
-------
Removal of Organics by Reverse Osmosis
Reverse osmosis separation is the combined result of preferential sorption
of solvent or solute at the membrane-solution interface, and the flow of
the interfacial fluid through the pores on the membrane surface (Sourirajan,
1970). This process is applicable for the separation of both inorganic
or organic substances in aqueous solution, however, it has not been found
to be very promising in separating organic matter of low molecular weight
from aqueous solutions (Chian and Fang, 1974). The research effort in
devising membranes for efficient removal of organic matter from aqueous
solution is continuing.
The mechanism by which the membrane rejects dissolved matter is still a
subject of much controversy. However, any theory of the separation
mechanism is intimately concerned with the structure of the membrane.
Chian and Fang (1974) indicated that polymeric membrane materials which
are less polar, (i.e., those that have small solubility parameters, or
those that are less soluble in water may be applicable for leachate
treatment.
Both cellulose and non-cellulose base membranes were used in the present
study to treat the raw leachate. The membrane giving the highest rejec-
tion of organic was chosen for treatment of the effluent from the
aerated lagoons and the effluent from the activated carbon and ion
exchange columns.
156
-------
MATERIALS AND METHODS
A well-mixed ozone reactor was used to study the oxidation of the organic
matter in the effluent of aerated lagoon No. 3. Since the overall oxida-
tion process may be either chemical reaction rate controlled or mass
transfer diffusion controlled, the reaction vessel had to be built in
such a way as to enhance the mass transfer to such an extent that only
the chemical reaction rate controlled the oxidation process. The
reactor consisted of one-liter graduated Pyrex beaker fitted with three
evenly spaced stainless steel bafflex inside. A Teflon coated magnetic
stirring bar was used to disperse ozone evolving from the fritted glass
diffuser. A stainless steel deflector was installed above the fritted
glass diffuser to prevent coalescence of the ozone gas bubbles, thus
improving the mass transfer rate of ozone to the liquid. This unit was
sealed at the top with a 3/4 inch Plexiglass cover with an 0-ring to
prevent leakage of ozone. Dried cylinder oxygen was used for the
ozonator. The unreacted ozone was trapped with 2 percent potassium
iodide solution. Ozone was determined using the potassium iodide
method under alkaline conditions.
After ozonation of the aerated lagoon effluent, the ozonated effluent
was tested for amenability to biological degradation. The experiment
was conducted in a completely-mixed glass beaker filled with 330 ml of
ozonated effluent. Acclimated sludge from the 15 day aerated lagoon
was added to provide a food to microorganism ratio of 0.1 mg BODs/mg
MLVSS-day. Before sampling, water lost due to evaporation was
replenished with distilled deionized water. Excess nutrients were
added to ensure an adequate growth condition.
The activated carbon adsorption experiments were conducted with Calgon
Filtrasorb 400 activated carbon. This carbon is a coal-based carbon and
was obtained in a size range in which the particles passed a U.S. Standard
Sieve No. 12 and were retained on a No. 40 sieve. The surface area of
157
-------
2
this carbon, as reported by the manufacturer, is 1000-1200 m /g.
Filtrasorb 400 as received was mechanically ground to a smaller size.
It was then sieved to a size range which included particles passing a
No. 40 sieve and retained on a No. 48 sieve. After sieving, the carbon
was washed thoroughly with deionized water to remove dust and fines,
and was then dried for 48 hours at 105°C, before being used in the
column. All of the studies were conducted with carbon from the same
lot to avoid batch-to-batch variations.
The activated carbon column employed in this study consisted of a 25 cm
length and 1.2 cm diameter Plexiglass column. The depth of the carbon
bed was 13 cm. The column diameter to particle size ratio was approx-
imately 30, to minimize entrance and wall effects (Smith, 1970).
Hydraulic loadings with downflows of 4 ml/min or 0.35 cm/min (0.87
2 ?
gal/min/ft ) and 20 ml/min or 1.76 cm/min (4.34 gal/min/ft ), with
empty bed detention times of 3.67 minutes and 0.73 minutes, respectively,
were evaluated. The effluent of the aerated lagoon (unit 4) having a
detention time of 30 days was used for the tests after being prefiltered
with a Whatman filter paper No. 1.
Both weak and strong base anion exchange resins were evaluated in the
column studies. The weak base anion exchange resin was Duolite A-7
(Diamond Shamrock Chemical Company, Redwood City, CA). Two strong
base anion exchange resins, Amber!ite IRA-938 and XE-297HP (Rohm and
Mass Company, Philadlephia, PA) were also used as adsorbents. Duolite
A-7 resin has a phenol-formaldehyde matrix and a secondary amine
functional group. Amber!ite IRA-938 resin has a styrene-divinylbenzene
matrix and a quaternary amine functional gropu. Amber!ite XE-279HP has
an acrylic matrix and a quarternary amine functional group. The charac-
teristics and structure of Duolite and Amberlite resins are shown in
Table 11. The Duolite A-7 resin was received in ionic form (salt form).
Amberlite IRA-938 and XE-279HP were received in chloride form. All
resins were preconditioned before use to remove minute quantities of
soluble impurities prior to experimentation. The preparation and pre-
conditioning procedures for the anion exchange resins are given as follows:
158
-------
Table 11
Characteristics of Ion Exchange Resins Used to Treat
the Aerated Lagoon Effluent
Duolite A-7
Amberlite
IRA-938
Amberlite
XE-279HP
Chemical Classifi-
cation Weak Base
Functional Group Secondary Amine
Matrix Phenol
Formaldehyde
Mesh Range 16-50
Capacity meq/m 2.4
Capacity meq/g(dry) 9.1
Approximate pH Range
Acid Adsorption
Approximate pH Range
Scavenging
Maximum Tempt.
Specific Gravity
0^4
Strong Base
Quarternary Amine
Styrene-DVB
20-50
3.8 (Cl form)
1 * 12
2^8
40°C
1.12 (free
base)
1 * 12
60°C (Cl form)
77°C (OH form)
1.203 (Skeletal)
0.555 (Apparent)
Strong Base
Quarternary Amine
Acrylic
12
12
Porosity
Macroporous Macroreticular
Macroreticular
159
-------
a) Transfer a water slurry of resin to the column. Dry resins
should be soaked in water for one-half hour before transfer
to glass columns.
b) Backwash resin with water to expand bed at least fifty percent;
this removes air bubbles and arranges particles by size.
c) Stop the backwash and allow the resin to settle.
d) Drain water through resin, leaving one inch of water above bed.
e) Flow at least two bed-volumes of 1.5 N NaOH through resin (at
least twenty minutes).
f) Wash out caustic with five bed-volumes of distilled deionized
water (at least thirty minutes), drain water from resin to
within one inch of top of bed.
g) Flow at least two bed-volumes of 2N HC1 through the resin (at
least twenty minutes).
h) Wash out the acid with five bed-volumes of distilled deionized
water (thirty minutes), drain water from resin to within one
inch of top of bed.
i) Repeat the caustic-rinse-acid-rinse cycle outline in step e)-h).
The Amberlite IRA-938 and XE-279HP resins are now in the chloride form,
while the Duolite A-7 resin is in the salt form. To convert the Duolite
A-7 resin to the hydroxide or free-base form, it is necessary to regenerate
as described in step e) and f), while using distilled deionized water for
rinsing the resin. Continue rinsing all resins until the effluent pH is
below 9.0.
Studies using these three resins were carried out in 1.2 cm diameter
Plexiglass columns, each having a bed depth of 13 cm. Duolite A-7 resin
was ground to a size of 40 x 80 mesh, and Amberlite IRA-938 and XE-279HP
were sieved to a size smaller than 40 mesh (U.S. Standard Sieve Size).
Thus the column diameter to particle size ratio for these resins was
brought to approximately 30 in order to minimize entrance and wall
effects. The prefiltered (Whatman filter paper No. 1) effluent from
unit 4 was fed downflow at a flowrate of 20 ml/min or 1.76 cm/min
160
-------
(4.34 gal/min/ft ) and 4 ml/min or 0.35 cm/min (0.87 gal/min/ft2) for
the A-7 column. The flowrate was 4 ml/min or 0.35 cm/min (0.87 gal/min/ft2)
for the IRA-938 and XE-279HP columns. Since the pH might affect the organic
removal efficiency of the Duolite A-7 weak base resin, the original effluent
pH of 8.8 was adjusted to pH of 6.2.
The reverse osmosis process was performed with a small stainless steel
(316) high pressure test cell (Abcor, Cambridge, Mass.). Membrane
performance was tested with respect to flux, rejection, and the percen-
tage of product water recovery. TOC and conductivity measurements were
used to define the membrane rejection of the organics and the inorganics,
respectively. The standard procedure employed in evaluating membrane
processes is to first standardize the membrane with distilled deionized
water. Next, a solution of 5,000 ppm sodium chloride is subjected to
reverse osmosis under standard testing conditions, i.e. 600 psig and
room temperature (^ 24°C). The flux of distilled deionized water was
measured versus time, until a steady-state flux was observed. In
addition to flux measurement, rejection of salt by the membranes was
measured versus time for the 5,000 ppm solution of sodium chloride.
The rejection efficiency and volume recovery were calculated by using
the conventional expression, as given by the following equation:
R - 1 - ^ (4)
\ - \ (5)
where R = rejection (fractional)
Vp = volume recovery (fractional)
Cp = concentration of solute in the permeate
Cp = concentration of solute in the feed
V = volume of permeate solution
V = volume of feed solution
161
-------
In treating raw leachate by membrane processes, two different types of
membranes were evaluated. One is the conventional cellulose diacetate
membrane from Eastman Kodak which is designated as KP-98. The other is
a cross-linked polyethylenimine-tolylene-2,4-diisocyanate active layer
laid on a polysulfone base. This is an ultrathin membrane from North
Star Research Institute (Minneapolis, MM) and is designated as NS-100.
162
-------
RESULTS AND DISCUSSION
Earlier results reported in Chapter II indicated that activated carbon
treatment, and lime treatment did not result in high organic matter
removals. Preliminary results obtained in this study with ozonation of
leachate showed that this also did not give substantial organic matter
reductions. Preliminary tests with leachate using reverse osmosis,
however, showed some promise, and further studies were required to
explore this alternative.
Organic Removal in Leachate by Reverse Osmosis
Both KP-98 and NS-100 membranes showed extremely high rejection with
sodium chloride, KP-98 gave a salt rejection of 96 to 98 percent whereas
NS-100 gave a salt rejection of more than 99 percent. Results of these
tests with both KP-98 and NS-100 membranes on the raw leachate are shown
in Table 6. An increase in the pH from 5.5 to 8.0 (adjusted by adding
concentrated NaOH) increased the TOC rejection from 70 to 91.5 percent
using the KP-98 membranes. A parallel increase was observed with TDS
rejection. Since most of the volatile acids have their pK values
between 4.7 and 4.9, the degree of dissociation of these acids in
leachate at pH of 5.5 is incomplete. The increased amount of fatty
acids dissociated between pH of 5.5 and pH of 8 accounts for the increase
in rejection of TOC. The increase in TDS rejection at a higher pH as
measured by conductivity is also attributed to the increase in the degree
of dissociation of the organic acids at a higher pH. Table 12 also shows
a decrease in both the rejection and the flux with an increase in feed
concentration and a decrease in operating pressure. The rejection of
TOC, total dissolved salt, and flux, however, increased with increasing
pH and pressure. The performance of the NS-100 membrane was in all
cases better than that of the KP-98 membrane.
Fouling of the membranes posed a serious treat to the applicability of
RO to raw leachate. When the batch runs were finished, both the KP-98
and NS-1 membranes were washed. The successive runs were made with the
163
-------
Table 12
The Treatment of Leachate by KP-98 and NS-100
Membranes at 50% Product Water Recovery
Type of
Membrane
KP-98
NS-100
Feed
Solution
5000 mg/1
. Nad
Raw
Leachate
TOC =
12946 mg/1
Raw
Leachate
TOC =
18460 mg/1
5000 mg/1
Nad
Raw
Leachate
TOC =
12946 mg/1
Operation
Pressure
(psig)
600
600
1500
600
1500
600 „
600
1500
PH
5.5
8.0
5.5
8.0
5.5
8.0
5.5
8.0
5.5
8.0
5.5
8.0
%
Rejection
of TOC
70
91.5
75
93
56
89
59
60.3
85
93
88
94
Rejection
of TDS
97
97
98
98.2
99
85
97
87
98.5
99.3
97.5
99
98.5
99
*
Flux ,
(gal/day/ft*)
15.1- - 8.7
5.5
6.1
8.9
10
3.7
3.9
6.2
7.1
21 - 19.6
7
7.3
11
12.5
Flux was measured from fresh membranes within 3 hours operation
164
-------
same membranes, but the flux had decreased by more than 75 percent.
Further runs decreased the flux even more. Therefore, unless suspended
solids, colloidal materials, and iron hydroxides are effectively removed
by other methods prior to the reverse osmosis process, this alternative
is completely impractical. Since the NS-100 membrane demonstrated a
better rejection and flux than the KP-98 membrane, the polishing of
effluent from extended aeration units, activated carbon and iron exchange
columns was conducted with reverse osmosis using the NS-100 membrane only.
Organic Removal in Aerated Lagoon Effluent by Ozonation
A batch sample (700 ml) from unit 3 was ozonated in order to determine
the rate of removal of organic matter. In addition, oxygen mass transfer
coefficients under a gas flowrate of 4 1/min (1.2 to 1.5 percent ozone by
volume) were evaluated. During the test runs, samples were withdrawn at
various times and analyzed for total organic carbon (TOC). Results of
this test are shown in Figure 55 and listed in Table 13. A comparison
was made between the mass transfer rate, obtained from the oxygen trans-
fer experiments, and the experimentally measured rate, determined from
a material balance using the iodiometric titration of the residual ozone
in the effluent gas. Using the effluent from the aerated lagoon, the
mass transfer rate was found to be higher than the chemical reaction
rate, indicating that the latter is the rate controlling step. Therefore,
the magnitude of the first order rate constant k reflects the nature of
the organics in the effluent. The results indicate that direct ozonation
of the extended aeration effluent gives a 48 percent TOC and 90 percent
color removal after 3 hours. Actually, such results make it seem
impractical to apply ozone in such a high dose and with such a long
period of time as the first step to reduce the organic constant of the
extended aeration treatment units.
The ozonated effluent had a TOC of 119 mg/1 and a COD of 300 mg/1.
Further polishing with the aerobic biological process were studied as
a means of removing these residual organics. Acclimated sludge was
taken from unit 4, resulting in a food to microorganism ratio, BOD^/
MLVSS-day = 0.1. Results of the aerobic biological process for ozonated
165
-------
250
30
60
90 120
Time, min
150
180
210
Figure 55. TOC decrease of ozonated effluent from aerated lagoon
3 treating leachate
166
-------
Table 13
Determination of Various Coefficients for Ozonating
Aerated Lagoon Effluent of Unit 3
1st Order TOC TOC Reduction Rate of Ozone
a 3 Initial TOC Rate Constant During Initial Transferred (mg/1, nr)
0 0 (mg/1) k (hr-1) Time Period MTR(1) CR(2) CDR(3)
2 2 —
1.43 0.97 250 0.221 48% in 3 hr 1023 391 408
(1) Rate of ozone transferred into the solution calculated from mass transfer
rate (MTR) assuming [03] is equal to zero and using a, 3 and Kj_a measured
from oxygen transfer equipment.
(2) Rate of ozone transferred into the solution calculated from chemical
reaction rate (CR), as measured by TOC decreases.
(3) Rate of ozone transferred into solution calculated from experimentally
determined rate (CDR) as measured by iodiometric method.
167
-------
effluent are shown in Figure 56. The subsequent aerobic biological
process could only remove 16.7 percent of the TOC after an aeration
time of 6 hours. Beyond 6 hours, the bacterial cells lysed, resulting
in a very high soluble TOC's. Therefore, the biological process is
not a feasible method for polishing the ozonated effluent.
Organic Removal in Aerated Lagoon Effluent by Activated Carbon
Two carbon column tests were conducted with the effluent from the
aerated lagoon unit 4, using two flow rates which gave empty bed
detention times of 0.73 and 3.67 minutes. The activated carbon
breakthrough curves are shown in Figure 57 and 58 in which the C/Co
was calculated for COD, TOC and color. The column effluent COD, TOC,
and color increased gradually as the amount of effluent passed through
the bed increased. The TOC removals generally increased parallel to
the COD values in the column effluent, indicating that adsorption of
organics in the column did not preferentially remove the more oxygenated
compounds. The maximum COD removal in the column with the 0.73 minute
detention time was 67% and decreased to 56% after 50 bed volumes. The
column with the 3.67 minute detention time removed initially 86% of
the COD and decreased to 74% after 50 bed volumes. These results
therefore show that high organic matter removals can be obtained using
relatively low flowrates and longer detention times. Both columns
removed the color bearing organics to a greater extent then the
remainder of the organics.
Organic Removal in Aerated Lagoon Effluent by Ion-Exchange Resins
The ionization of functional groups in weak base resin depends on the pH
of the solution. This also applies to the functional groups in the organic
matter present in the effluent of the aerated lagoons both the influence
of pH of the feed solution and the operating flowrate on organic matter
removal were evaluated using the weak base resin. Sulfuric acid was
used to adjust pH of the influent. Figures 59 and 60 show the COD, TOC
and color breakthrough curves for the Duolite A-7 resin at the identical
operating flowrate of 1.76 cm/min but at pH values of 8.8 and 6.2,
168
-------
700
2 3
Aeration Time, days
Figure 56. Results of aerobic biological polish of ozonated
effluent from aerated lagoon 3 treating leachate
169
-------
PH
12
II
10
9
8
7
6
1.0
0.9
0.8
0.7
0.6
C/C0(-) 0.5
0.4
Q3
Q2
O.I
j TOG '- 214 mq/t
Column Influent < COD = 533 mg/*
IpH =8.8
COD
50 75
Number Of Bed Volumes (-)
100
125
Figure 57. Activated carbon breakthrough curve for effluent from aerated lagoon
unit 4 treating leachate at a flowrate of 1.76 cm/min
-------
pH
C/C0(-)
12
II
10
9
8
7
6
1.0
0.9
OB
O.7
0.6
0.5
0.4
0.3
02
0.1
r»
1 1 1 1 1
fTOC = 214 mg/1
~ Column Influent < COD = 547mg/i
_ IpH =8.8
—
—
-
—
—
—
—
/-TOC
^^0-^0
~ r&^^^^r^ C0^/~ Color at 400 nm
~L_-*— *-r-* — ^ — ^-r^ i
25
50 75
Number Of Bed Volumes (-)
100
125
Figure 58. Activated carbon breakthrough curve for effluent from aerated lagoon
unit 4 treating leachate at a flowrate of 0.35 cm/min
-------
PH
ro
12
II
10
9
8
7
6
1.0
as
OB
0.7
0.6
C/Co(-) 0.5
0.4
Q3
02
(XI
°0
fTOC = 214 mg/f
Column Influent < COD =526mg/Jl
IpH =8.8
Color
25
50 75
Number Of Bed Volumes (-)
100
125
Figure 59. Duolite A-7 breakthrough curve for effluent from aerated lagoon unit
4 treating leachate, at a flowrate of 1.76 cm/min
-------
CO
II
10
9
8
7
6
1.0
0.9
0.8
0.7
0.6
C/C0(-) 0.5
0.4
0.3
02-
0.1-
f TOC = 214 mq/t
Column Influent < COD =544mg/*
i.pH =6.2
TOC
25
50 75
Number Of Bed Volumes (-)
100
125
Figure 60. Duolite A-7 breakthrough curve for acidified effluent from aerated lagoon
unit 4 treating leachate at a flowrate of 1.76 cm/min
-------
respectively. When the pH was adjusted from 8.8 to 6.2, higher removals
for TOC, COD, and color were observed. The initial COD removal increased
from 28% to 37% as a result of the pH decrease. These removals, however,
are less than the 68% removal obtained with activated carbon under similar
flow conditions. Since the pK value of A-7 is approximately 6, the amine
groups of the A-7 resin are mostly un-ionized at pH values much higher
than 6, e.g., a pH of 9. The degree of ionization of the carboxyl
groups of the organic matter in the effluent is lower at pH of 6.2 than
that at a pH of 8.8. The higher removals of TOC, COD and color at pH of
6.2 may therefore result from the adsorption mechanisms of the A-7 resin
in a free base form. A similar pH effect was noted by Schnack and Kaufman
(1970) who used the weak base resin Duolite ES-33, with the same phenol-
formaldehyde matrices as A-7 but with a pK of 4.5. Passage of the
aerated lagoon effluents through the resin resulted in a pH increase
from 8.8 to 9.1 and from 6.2 to 8.2. The initial increase in pH during
adsorption of the organic matter onto the A-7 resin from the aerated
lagoon effluent was also noted by Kim (1974). The pH increase could
result from the following reaction:
H H
A" + H20 + N - RZ -> A H - N - RZ + OH" (6)
H H
where A~ represents the organic anion. The last step in the preparation
of the resin in the free base form involves washing with concentrated
sodium hydroxide followed by rinsing with distilled deionized water
which cause the protonation of the hydroxyl groups of the phenolic
matrix of A-7. If the protonation of the ionized phenolic hydroxyl
groups is not completed during rinsing with distilled deionized water,
the following reaction would account for a pH increase:
R'O" + H20 -> R'OH + OH" (7)
The pH gradually decreases to that of the influent (Figure 60) as the
effluent bed volume increases, and protonation of the ionized phenolic
hydroxyl groups are completed.
174
-------
The breakthrough curve using the Duolite A-7 resin at the same pH of 6.2,
but at a lower operating flowrate of 0.35 cm/min is shown in Figure 61.
The lower flowrate resulted in an approximately 10 percent increase in
color and COD removal but did not greatly enhance TOC removals indicating
that less oxygenated compounds were being removed. However, the initial
COD removal of 59% was still less than the 86% obtained with activated
carbon under similar flow conditions. The initial pH increased more
sharply at a low operating flowrate of 0.35 cm/min than at a high flowrate
of 1.76 cm/min.
As strong base resins do not experience any pH effects this parameter
was not investigated, and only different resin matrices were studied for
this removal of organics. The main mechanisms of organic removal by
strong base resins are adsorption on the resin matrix and ion exchange
by the functional groups. As these mechanisms are applicable for a wide
pH range, no pH adjustments are necessary. The breakthrough curves for
two strong base resins with effluent from the aerated lagoon 4 are shown
in Figure 62 and 63. Both of these strong base resins were operated in
chloride form. Removal of TOC, COD, and color with the IRA-938 styrene-
DVB matrix, was better than with the XE-279HP acrylic matrix. Since
most of the organic matter in the extended aeration effluent is present
as a fulvic acid-like material, containing substantial amounts of
aromatic groups, a high affinity and thus stronger adsorption may be
expected when the chemical configuration of the organic material and the
resin matrix are similar as is the case for the aromatic styrene-DBV
type resins. This affinity can be explained by the interactions of IT
electrons of the aromatic rings. Two clouds of electrons are formed
near the aromatic ring by the TT bonds between carbon atoms of the
aromatic ring. The effects of TT electrons affinity between aromatic
compounds and the non-ionic styrene-DVB type resin obviously aids
adsorption.
The salts in the effluent of these ion exchange columns did not change
appreciable with either type of anion resins used. Slight increases in
TDS were then experienced with the use of the strong base resins which
175
-------
l£
II
10
7
6
1.0
Q9
OB
0.7
0.6
C/C0(-) 0.5
0.4
Q3
Q2
0.1
°C
i I i 1 1
_ f TOC = 214 mg/l
Column Influent < COD =522mg/* "
- IpH =6.2
- ^^"^-^
— _
— _
—
-
-^^^Tr0"^"^^^^^ ™
~*^ u I
—
> 25 50 75 100 125
Number Of Bed Volumes (-)
Figure 61. Duolite A-7 breakthrough curve for acidified effluent from aerated lagoon
unit 4 treating leachate at a flowrate of 0.35 cm/min
-------
PH
C/Co(-)
12
II
10
9
8
7
6
1.0
0.9
03
0.7
0.6
0.5
0.4
0.3
Q2
0.1
n
fTOC = 2l4mg/J
~ Column Influent < COD = 533mg/l
_ tpH =8.8
—
—
— ~
—
— o o o .0— ^ 1U1-
_ / D COD
_(/ ^
25
50
Number Of Bed Volumes (-)
125
Figure 62. Amberlite IRA-938 breakthrough curve for effluent from aerated lagoon
unit 4 treating leachate, at a flowrate of 0.35 cm/min
-------
pH
oo
12
II
10
9
8
7
6
1.0
Q9
OB
0.7
0.6
C/C0(~) 0.5
Q4
0.3
Q2
0.1
fTOC = 2l4mg/jl
Column Influent < COD =527 mg/Jt
LpH =8.8
Color
25
50 75
Number Of Bed Volumes (-)
100
125
Figure 63. Amberlite XE-279HP breakthrough curve for effluent from aerated lagoon
unit 4 treating leachate, at a flowrate of 0.35 cm/min
-------
is due to the exchange of the chloride ion from the resin into the
solution.
The organic removal achieved with both ion exchange and activated
carbon processes are summarized in Table 14. The data collected at
the same flowrate show that activated carbon gives the best removal of
organic matter from aerated lagoon effluent. The next best results
were obtained with the strong base resin, Amberlite IRA-938 having the
styrene-DVB matrix. This resin showed a slightly greater organic
removal than the weak base resin A-7, even though the effluent passed
through the A-7 column was acidified.
Since relatively large and expensive thermal regeneration equipment is
needed for regeneration of the spent carbon in addition to the pumping
costs for transferring the carbon, synthetic resins appear to be an
attractive alternative. These resins can be regenerated by inexpensive
chemicals such as NaOH or NaCl.
Organic Removal in Aerated Lagoon Effluent by Reverse Osmosis
Based on the preliminary results treating raw leachate, the NS-100
reverse osmosis membranes were shown to have a better rejection of
organics than cellulose acetate membranes. This is due to the distinct
differences in the chemical nature of the NS-100 membrane as opposed to
the somewhat more polar nature of the cellulose acetate (CA) membrane.
The NS-100 membrane materials, which are less polar, have less tendency
to act as proton accepters as compared with CA.
The rejection of TDS with these membranes was evaluated simultaneously
with removal of organics from the aerated lagoon effluents. As shown
in Table 15, both organic matter and TDS were removed effectively with
the membrane process. The flux of membranes was also quite high.
Membrane processes are thus an excellent method for polishing effluent
from extended aeration units, activated carbon columns and ion exchange
columns. Since the fouling of membranes by suspended solids is the
179
-------
Table 14
Removal of Organics by Resins and Activated Carbon
Material
Activated
Carbon
Duol ite
A-7
Amberl ite
IRA- 938
Amberl ite
XE-279HP
Flowrate
1.76
0.35
1.76
1.76
0.35
0.35
0.35
Percentage Removal After
50 Bed Volume
TOC
47
71
6
42
43
43
26
COD
56
74
6
37
48
59
41
Color
71
94
31
80
92
95
76
Influent pH
8.8
8.8
8.8
6.2
6.2
8.8
8.8
Effluent pH at
50 Bed Volume
8.8
8.8
8.9
6.8
7.4
8.8
8.8
180
-------
CO
Table 15
Removal of Organics and Salts with the NS-100 Membrane From Effluents of the Aerated Lagoon
4, and From Effluent of the Activated Carbon- and Ion Exchange Columns
Operation Pressure = 600 psig, Temp. = 24°C
Feed TOC IDS
Solution (mg/1) (mg/1)
Extended 214 6200
Aeration
Effluent
Activated 48.2 6200
Carbon
Ion 132.7 6200
Exchange
A-7
Effluent
Ion 118.8 6260
Exchange
IRA- 938
Effluent
Ion 143.4 6250
Exchange
XE-279HP
Effluent
% Volume TOC in %
Recovery Permeate Rejection
(mg/1) of TOC
51.2 10.7 95
66.2 16.6 92.2
50 6.5 86.4
47.6 4.7 96.5
44.4 7.3 93.8
48.4 8.2 94.3
TDS in %
Permeate Rejection
(mg/1) of TDS
390 93.7
550 91.1
270 95.7
267 95.7
294 95.3
310 95
Flux
gal/day/ft2,
(crrr/day/cm^)
9.8 (4.0)
9.0 (3.65)
12.5 (5.1)
12.0 (4.9)
12.4 (5.0)
11.9 (4.8)
The operating conditions for activated carbon and resins columns for treating aerated lagoon
effluent were at a flow rate of 0.35 cm/min and a pH of 8.8; with the exception of the A-7
column which was run at a pH of 5.5
-------
major limitation of this process, pretreatment of effluent from the
extended aeration unit by filtration or chemical precipitation may be
necessary. No significant improvement of reverse osmosis effluent
quality was obtained by pretreatment of the aerated lagoon effluent by
carbon and resin columns (Table 15).
Ahlgren (1971) has studied the costs of different alternatives for
demineralization and concluded that cation and anion exchange methods
are the only applicable for low IDS levels with concentrations less
than 200 ppm. In the transition range between 200 to 2,000 ppm of IDS,
electrodialysis is an attractive treatment method. Reverse osmosis is
the most attractive means of demineralization. With feed water contain-
ing more than 2-00 ppm IDS. Since the IDS level in the effluent of the
aerated lagoon units receiving leachate was higher than 2,000 ppm IDS,
only the reverse osmosis process was considered in this study for
demineralization.
The activated carbon process was further studied to treat the reverse
osmosis permeate in order to obtain effluent of high quality. The
influent to the activated carbon column consisting of the permeate of
NS-100 membrane had an organic concentration of 16.6 mg/1 TOC (33.6
mg/1 COD). The results shown in Figure 64, reveal an 82% COD removal
which is only slightly less than the 86% observed for the unfractionated
effluent. The TOC removal, however, is considerably lower than the COD
removal, indicating that highly oxidized compounds with a low COD/TOC
pass through the activated carbon column.
Based on the results of the above studies, a proposed scheme for treating
leachate with aerated lagoon followed by various physical-chemical methods
is depicted in Figure 65. The detention time of the aerated lagoon could
be varied from 10 to 30 days depending on the nature of landfill leachate
being treated. The NS-100 membrane could be substituted by the duPont's
B-10 Permeator (Wilmington, DE), because both of these membrane materials
have comparable characteristics in terms of organic and TDS removals.
182
-------
PH
00
12
II
10
9
8
7
6
1.0
0.9
0.8
0.7
0.6
C/C0(-) 0.5
0.4
0.3
02
O.I
f TOC - 16.6 mg/l
Column InfluenU COD -33.6mg/*
IpH =8.5
TOC
COD
50 75
Number Of Bed Volumes (-)
100
125
Figure 64. Activated carbon breakthrough curve for RO NS-100 membrane permeate from
effluent of aerated lagoon unit 4 at a flow of 0.35 cm/min
-------
Raw
Leachate
(Nutrient Addition)
1
Classification
Extended Aeration
Treatment
Sludge Cake
Disposal
Sludge
Dewatering
Sludge
Conditioning
Final
Effluent
Reverse Osmosis
NS-ICO
Membrane
Reverse Osmosis
\IS-IOO
/lembrone
pH
'Regenerant 'Adjustment
I.SNNacL
Reverse Osmosis
NS-IOO
Membrane
Strong Base
Ion Exchange
Resin
Amberlite IRA938
' Regenerant
1.5 N Nacl
Reverse Osmosis
NS-IOO
Membrane
Recovery Or
Ultimate Disposal
/ V
{ 1
,RO
Retentate
Ion Exchange)
—[Spent
iRegenerant
H^
Figure 65. Sanitary Landfill Leachate Treatment Schematic Diagram
184
-------
REFERENCES
Ahlgnen, R. M. "Membrane vs Resinous Ion Exchange Demineralization,"
Indust. Water Engineering, 8_, 1, January (1971).
Chassanov, M. G., Kunin, R. and McGarvey, F. "Sorption of Phenol by
Anion Exchange Resins," Ind. Eng. Chem., 48, 305 (1956).
Chian, E. S. K. and DeWalle, F. B. "Characterization and Treatment of
Leachates Generated from Landfills," AIChE Symposium Series, No. 145,
Vol. 71, Water-1974, p. 319 (1975).
Chian, E. S. K. and DeWalle, F. B. "Criteria of Selection of Methods for
Leachate Treatment," Proceed. Second National Conference on Complete
Water Reuse, AIChE (1976a).
Chian, E. S. K. and DeWalle, F. B. "Sanitary Landfill Leachates and
their Treatment," Journal Environ. Engr. Div. ASCE 1_02, 411 (1976b).
Chian, E. S. K. and Fang, H. H. P. "Evaluation of New Reverse Osmosis
Membranes for the Separation of Toxic Compounds from Water," AIChE
Symposium Series 136, 70, 497 (1974).
Cook, C. N. and Foree, E. G. "Aerobic Biostabilization of Sanitary
Landfill Leachate," Journ. Water Pollution Control Fed., 46, 380
(1974). —
Caughlin, R. W. and Ezra, F. A. "Role of Surface Acidity in the
Adsorption of Organic Pollutants on the Surface of Carbon,"
Env. Sci. and Tech., 2, 291 (1968).
Dorfner, K. "Ion Exchange: Properties and Applications," Ann Arbor
Science Publishers, Inc. (1972).
Helfferich, F. "Ion Exchange," McGraw-Hill Book Co., Inc., New York
(1962).
Ho, S.-et al_. "Chemical Treatment of Leachates from Sanitary Landfills,"
Jour. Water Pollution Control Fed.. 46, 1776 (1974).
Karr, P. R. "Treatment of Leachate from Sanitary Landfills," Special
Research Problem School of Civil Engineering, Georgia Institute
of Technology, Atlanta, GA, Oct. (1972).
Kim, B. R. "Adsorption of Organic Compounds by Activated Carbon and
Synthetic Resin," Master Thesis, Department of Civil Engineering,
University of Illinois, Urbana, Illinois (1974).
185
-------
Mattson, J. S. and Mark, H. B. Jr. "Activated Carbon: Surface Chemistry
and Adsorption from Solution," Marcel Dekker, Inc., New York (1971).
Pohland, F. G. and Kang, S. J. "Sanitary Landfill Stabilization with
Leachate Recycle and Residual Treatment," AIChE Symposium Series
No. 145, Water-1974, Vol. 71, 208 (1975).
Schnack, P. G. and Kaufman, W. J. "Removal of Organic Contaminants -
Optimizing Resin Column Operations," SERL Report No. 70-11,
Sanitary Engineering Research Lab., University of California,
Berkeley, California (1970).
Simensen, T. and Odegaard, H. "Pilot Studies for the Chemical Coagulation
of Leachate," Norwegian Institute of Water Pollution Research,
Blindern, Oslo (1971).
Smith, J. M. "Chemical Engineering Kinetics," 2nd ed., McGraw-Hill
Book Co., Inc., New York (1970).
Snoeyink, V. L. and Weber, W. J. Jr. "The Surface Chemistry of Activated
Carbon," Env. Sci. and Tech.. 1, 229 (1967).
Sourirajan, S. "Reverse Osmosis," Academic Press, New York (1970).
Thornton, R. J. and Blanc, F. C. "Leachate Treatment by Coagulation
and Precipitation," J. Environmental Engineering Div., ASCE, 99,
535 (1973). ~~
Weber, W. J. Jr. "Physiochemical Processes for Water Quality Control,"
Wiley-Interscience, New York (1972).
Weston, R. F. "Leachate Treatability Study, New Castle County, Delaware,"
Roy F. Weston Inc., West Chester, Pennsylvania (1974).
186
-------
V
COMBINED TREATMENT OF LEACHATE AND MUNICIPAL SEWAGE
IN AN ACTIVATED SLUDGE UNIT
CONCLUSION
Laboratory study showed that a conventional plugflow activated sludge unit
receiving municipal sewage could effectively treat a high strength leachate
containing high concentrations of free volatile fatty acids such as
acetic- and butyric acid. Directly after the leachate addition some
deterioration in effluent quality of the test unit was observed; after
prolonged periods, however, effluent BOD values of the test unit receiv-
ing low leachate additions were generally comparable to that of the
control unit. While BOD values are not greatly affected, COD concentra-
tions showed a gradual increase with increasing leachate addition
indicating that larger quantities of refractory organics were released
from the test unit. The test unit was not able to treat the high
strength waste at 4% of the influent inflow rate as evidenced by high
BOD effluent concentrations and deteriorating sludge characteristics.
This operating failure was attributed to limiting phosphate concentra-
tions in the influent and the relative composition of the soluble high-
molecular-weight organics that tend to affect the flocculation of the
sludge. A tertiary treatment unit such as activated carbon showed lower
organic matter removals after addition of leachate to the municipal
activated sludge unit. The decreased removal was also thought to be
due to the differences in the low-molecular-weight organic matter
composition. The advantage of combined treatment is that it may be
the least expensive alternative as compared to treating the leachate
in a separate unit at the landfill site. In addition the effluent
phosphate concentration would be reduced due to the low COD:P ratio in
the leachate. A disadvantage of the combined treatment is the required
increase in size of the secondary clarifier while lower organic matter
removals may occur in a tertiary activated carbon unit treating the
effluent from the combined treatment unit.
187
-------
INTRODUCTION
Solid waste landfills can have an adverse impact on the environment when
leachates contaminate underlying groundwater strata or surface streams.
Such impact can be controlled by reducing the quantity of leachate
generated or by treating the leachate already generated. Treatment of
leachate can be realized by recirculation of the leachate back into the
landfill to enhance anaerobic biological removal of the pollutants, by
construction of separate treatment units at the site, or by combined
treatment at a municipal sewage plant. The purpose of this study was
therefore to evaluate the addition of increasing quantities of leachate
to a conventional plugflow activated sludge unit treating municipal
sewage. Results in Chapter III indicate that high strength leachate is
amenable to aerobic biological treatment. Only one earlier study evalu-
ated the addition of leachate to a unit treating municipal sewage.
Boyle and Ham (1974) found that leachate with a COD of 10,000 mg/1
could be added to domestic sewage in an extended aeration activated-
sludge unit at a level of at least 5% by volume without seriously
impairing the effluent quality. At greater than 5% by volume, leachate
additions resulted in substantial solids production, increased oxygen
uptake rates, and poorer mixed liquor separation.
188
-------
ACTIVATED SLUDGE PROCESSES
The activated sludge process is a very flexible process and can be
adapted to almost any type of biological waste-treatment problem. The
purpose of this section is to discuss the characteristics and the
applicability of both the conventional activated-sludge processes as
well as some of the modifications (Metcalf and Eddy, 1972) for
leachate treatment:
(a) The conventional process utilizes a plug-flow pattern with
recycle of the settled sludge. The recommended operating parameters
are an aeration period about 6 hours, an organic loading of about 0.2
to 0.4 mg BOD/mg MLVSS-day. The mixed liquor volatile suspended solids
(MLVSS) ranges from 1,500 to 3,000 mg/1. The ratio of the volume of the
recycle sludge stream to influent stream is often about 0.25 to 0.5.
The sludge age is generally maintained at a value of about five to
fifteen days.
(b) The complete-mix process, in which the influent waste stream
is completely mixed within the total contents of the unit, has a better
resistance to shock loads, and is generally applicable to biological
treatment of industrial wastes (Metcalf and Eddy, 1972).
(c) The step-aeration process in which the settled sewage is
introduced at several points along the aeration tank, has a lower peak
oxygen demand than the conventional plugflow unit as the loading is
equalized over the length of the aeration tank. Higher BOD loadings
per unit aeration tank volume are therefore possible. This process
is also capable of withstanding shock loads better than the conventional
plugflow process.
(d) The contact-stabilization process was developed to take the
advantage of the absorptive properties of activated sludge to remove
colloidal, finely suspended, and dissolved organic matter in the wastes.
Its value in industrial waste treatment is limited largely to wastes in
which the organic matter is predominantly present as colloidal organics.
189
-------
(e) The extended-aeration process, aerated lagoon and oxidation
ditch processes are applicable only to small treatment plants.
(f) The Kraus process is used to treat high-carbohydrate wastes
containing a low amount of N and P nutrients in combination with
domestic sewage. The nutrients are added to the unit by aerating a
portion of the return sludge and the digested sludge containing
large amounts of N and P in a separate reaeration tank. This mixture
is then combined with the remainder of the recycle sludge and returned
to the aeration basin.
All of the above units can be aerated with air or with pure oxygen.
The pure-oxygen process has a number of advantages, such as increased
bacterial activity, decreased sludge volume, reduced aeration tank
volume, and improved sludge settling (Albertson, 1970).
The present study selected the conventional plugflow activated sludge
process for the combined treatment evaluation, as this process is the
most commonly used at the present time to treat municipal sewage.
Although the conventional plugflow activated sludge process would be
subject to upsets more easily than the modifications discussed above,
it is felt that if the conventional process would be able to treat
the combined leachate-municipal sewage flow, the modified processes
would certainly be able to treat this waste stream.
The present research simulated the conventional plugflow system by
employing three completely mixed reactors in series. Theoretically,
a series of an infinitive number of complete-mix reactors becomes a
plug-flow reactor. However, a system with only three complete-mix
reactors in series can be considered as approaching a plug-flow without
much experimental error. Table 16 shows the ratio of reactor volume
between the complete-mix and the plug-flow reactors by assuming
first-order reaction kinetics for BODg removal.
190
-------
Table 16
Required Reactor Volume for a Complete-Mixed Reactor
as Compared with a Plug-Flow Reactor at
Various Removal Efficiencies
Number of
Reactors
in Series
1
2
3
4
6
8
10
The
85%
Removal
Efficiency
2.98
1.67
1.42
1.30
1.17
1.14
1.10
1.0
Ratio of Reactor
90%
Removal
Efficiency
3.91
1.88
1.52
1.32
1.25
1.15
1.13
1.0
*
Volume Vr/Vp
95%
Removal
Efficiency
6.33
2.32
1.73
1.49
1.30
1.20
1.17
1.0
98%
Removal
Efficiency
12.53
3.10
1.99
1.70
1.41
1.29
1.23
1.0
*Vr = Volume of a series of complete-mix reactors
Vp = Volume of the plug-flow reactor
191
-------
COMBINED TREATMENT OF HIGH STRENGTH WASTE AND MUNICIPAL SEWAGE
Several studies have pointed out advantages of combined treatment of an
industrial waste and municipal sewage as compared to separate treatment
of the two waste streams. The significant savings in capital and
operating costs is often mentioned as its largest advantage (City of
Dallas, 1971). Longdon (1969) further pointed out that combined treat-
ment reduced the foaming and bulking of a municipal activated sludge
plant receiving yeast wastes. In fact many industrial wastes such as
those from meat packing plants, cotton mills, metal industries (El-
Gohary, 1971) pharmaceutical industries (Andersen ejt al_., 1971), citrus
processing plants (Eidsness et al_., 1971), drying industry (Hashimoto
et a]_., 1972) and papermills (Clingenpeel and Jones, 1974) are amenable
to combined treatment.
In several instances, however, it was necessary to either modify existing
treatment plants or to adapt the treatment plant operation to the specific
waste. Poon (1970), for example, showed that combined treatment was
successful when not more than one part of a high strength nylon waste
was added to seven parts of municipal sewage. Brosig ejt a_l_. (1971)
indicated that 90% BOD removal was only realized at hydraulic retention
times as high as 9 hours. Often treatment plants have to be converted
to accept the industrial waste. A conventional municipal activated
sludge plant was converted to a contact stabilization plant in order
to treat a nitrogen deficient corn products waste (Niles and Etzel,
1971). Pilot plant investigations of the conventional, Kraus and contact
stabilization process resulted in the selection of the latter process
for combined treatment of municipal wastewater and weak effluents of
four paper mills (Vockel ejt al_., 1974). It was further noted that
nutrient addition and chlorination of the return sludge were requred.
Previous studies at the University of Illinois showed that a plugflow
activated sludge unit treating municipal sewage could effectively
degrade a pretreated industrial wastewater from a chemical industry
192
-------
at a waste to sewage ratio of as high as 2.5 to 1. Although the
effluent COD values showed a substantial increase as compared to the
control unit, neither the effluent BOD nor the settleability of the
sludge showed any change when the F/M ratio of the unit remained con-
stant. The COD removal in activated carbon columns treating the effluent
of the combined activated sludge unit was also not affected by the waste
additions (Chian et_ al_., 1976). The present study used the same activated
sludge units but evaluated a landfill leachate waste stream of a substan-
tially higher strength and of a less biodegradable nature than the
industrial wastes. Using a similar solid waste leachate stream with
a COD of 10,820 mg/1, Boyle and Hqm (1974) added increasing volume
percentages of leachate to a batch operated extended aeration unit with
a 24 hour detention time treating municipal sewage resulting in an
increased organic loading of the units from 0.04 to 0.5 kg BOD/kg MLSS
day. They concluded that the units were not significantly affected
below a 5% addition of leachate, corresponding to an organic loading
of 0.15 kg BOD/kg MLSS day. As they varied two factors at the same
time, i.e., organic loading and percentage of leachate addition, it is
difficult to conclude the impact of each factor. The present study
therefore used a different experimental design in which the loading
to unit was kept constant while the percentage of leachate addition
was gradually increased. The effluent of both the test and the control
unit had to meet Illinois effluent standards of a BOD not to exceed
10 mg/1 and a suspended solids concentration not to exceed 12 mg/1.
Cities with more than 500,000 inhabitants or cities that discharge
into lake Michigan should not exceed 4 mg/1 BOD or 5 mg/1 suspended
solids. Furthermore the effluent ammonia-N concentration should not
exceed 2.5 mg/1 in the summer, and 4 mg/1 the remainder of the year,
while phosphorous should not exceed 1 mg/1 and TDS should not exceed
750 mg/1.
As the combined treatment requires an adequate nutrient balance for the
activated sludge, determination of the N and P content of different
leachates was made. The compositions of leachate samples from different
sanitary landfills reveal a large variation in concentration of differ-
193
-------
ent constituents (Table 17). The characteristics of the leachate pro-
duced from a given landfill depend upon such variables as composition
of the landfill material, environmental conditions, and age of the land-
fill. Morgan e_t aj_. (1973) recommended that the optimum nutrient ratio
for the conventional activated-sludge process should be BOD,-:N:P = 100-5-1
b
These ratios are not fixed but depend upon both the sludge age, and the
operating conditions of the activated sludge system. The average concen-
tration of nutrients in leachate are 10,244 mg/1 of BODg, 215 mg/1 of N
and 24.61 mg/1 of P which would result in average ratio for BOD5:N:P of
100:2.09:0.24. This indicates that the leachate is deficient in both
nitrogen and phosphorus. The municipal sewage on the other hand contains
too much nitrogen and phosphorus relative to BODj- materials, thus sewage
treatment plants have experienced difficulty in removing these nutrients.
Furthermore high nitrate concentrations often cause operating problems
in the secondary clarifier as denitrificstion and formation of nitrogen
gas bubbles may prevent adequate sludge settling. The combined treatment
of leachate with sewage would provide, at least a partial solution to both
of these problems. Since the excess nitrogen and phosphorus present in
the sewage can compensate for the deficiency in nutrients in the leachate,
the potential financial savings resulting from the elimination of require-
ments for nitrogen and phosphorus removal processes may offset some of the
additional costs required for expanding the system to treat leachate.
The analysis of leachate samples from sanitary landfills showed very high
concentrations of metals and total dissolved inorganic solids (TDS)
(Table 17). Many of the metals are expected to precipitate in the
aeration tank as metal hydroxide or calcium carbonate under aerobic
and alkaline conditions. The TDS is not expected to present problems
in the operation of the treatment plant when a high dilution ratio with
the sewage is selected.
The anaerobic-sludge digestion is the most common means of treating the
waste-activated sludge. Metals that are concentrated by the biological
sludge in the aeration tank are often released during anaerobic digestion.
194
-------
Table 17
Composition of Leachate Samples Collected from Eleven
Different Sources (Chian and DeWalle, 1976)
Source
Parameter
(mg/1)
COD
TOC
BOD5
Biocarbonate
Alkalinity (CaCO)
PH
Conductivity
(ymho/cm)
ORP (mV)
Turbidity (JTU)
SS
FSS
TS
FS
Org-N
NH4-N
NCyN
N02-N
Total -P
Ortho-P
SO
Cl
Ca
Mg
Fe
Na
K
UI
Lysimeter
49,300
17,060
24,700
668
5.13
13,700
-60
75
139
92.5
33,989
15,586
544.7
392.6
0.5
Trace
21.5
6.5
1,110
1,480
3,750
650
2,200
1,360
1,140
Range
81.10-71,680
70.00-27,700
3.90-57,000
142.00- 3,520
5.09- 7.25
978.00-16,800
(-220)-(+163)
23.00- 270
8.90- 923
7.50- 901
911.00-55,348
812.00-22,895
3.20- 945
1.40- 1,028
0.40- 10.25
Trace - 0.19
0.50- 98
0.25- 85
7.40- 1,558
60.20- 2,467
76.00- 3,900
35.00- 1,140
0.50- 1,046
37.50- 1,580
35.00- 2,300
Average
15,670
5,567
10,244
914
6.13
6,234
-11.36
91.31
264
193
12,559
5,704
135
211
3.93
0.03
24.61
16.05
284
720
961
334
402
391
527
195
-------
The ferric ion is reduced to ferrous ion, and precipitated with sulfide,
phosphate, or carbonate. Malhotra, et al_. (1971), for example, found
that the pH, alkalinity, volatile acids, and the percentage of reduction
of volatile solids in test digesters used for digesting sludges contain-
ing iron phosphates were not different from those in control digesters.
The addition of ferrous-iron in the feed sludge up to a maximum level
of 5.5% on a dry solids basis did not upset the digestion process.
However, the thickened waste-activated sludge containing the precipitated
iron phosphates resulted in significant releases of phosphate during
digestion. Singer (1972) indicated that the release of phosphate from
anaerobic digesters is limited by the concentration of ferrous iron, which
forms solid ferrous phosphate or vivianite, while retention as solid
calcium phosphate [Ca3(P04)2] or hydroxyapatite [Ca5(P04)3OH] was not
thought to be the major mechanism. Thomas (1972) reported that the
results obtained from eight wastewater treatment plants showed that the
phosphate precipitated with ferric chloride did not appear in the
supernatant during the digestion of activated sludge containing iron
phosphate. The above results, therefore indicate that although addition
of iron will not upset the anaerobic digesters, it may cause significant
release of phosphates during the digestion process.
Based on the above considerations, an evaluation was made of a conventional
plugflow activated sludge unit maintained at a constant loading but
receiving increasing volumetric amounts of leachate. The unit was also
tested with a constant volume addition of leachate but at increasing
loadings. During leachate additions the effluent quality was monitored
for nutrients to ensure adequate growth. The sludge was evaluated for
its settling properties to determine any impairment that might be
introduced as a result of leachate addition. The anaerobic digester
receiving waste sludge from the combined biological unit was evaluated
for potential phosphate recycle.
196
-------
MATERIALS AND METHODS
The continuous flow units employed in the present study consisted of
a primary clarifier, activated sludge units and secondary clarifiers
(Figure 66). A 190 liter (50 gallon) capacity primary clarifier was
operated on a batch basis in which a Moyno pump (Robbin and Meyer,
Arlington Heights, IL) was turned on fifteen minutes every hour to
replace the volume in the clarifier with sewage pumped from the municipal
sewer. The sewage was introduced into the primary sedimentation tank
from the bottom, and the suspended solids that settled in the tank were
flushed out every hour for fifteen minutes through the overflow. The
sand particles that settled on the bottom of the tank were removed
manually every few months. The clarified sewage was drawn from the
middle of the tank and delivered to the aeration tank by the Moyno
feed pumps. A lever switch was installed on the primary clarifier as
a safety device. When the level of sewage in the tank dropped to a
point somewhat above the outlet to the Moyno feed pumps, the feed pumps
were shut off automatically to prevent wearing by running dry.
The aeration basins with a total effective volume of 60 liters consisted
of three completely mixed compartments in sequence. A weir was provided
on each of the partitions to allow overflow to the succeeding compartment.
The design of the air supply system is shown in Figure 67. An air filter
was installed upstream of the air regulator for removing particulate
matter from the air line (the particulates might otherwise clog the air
supplying system). The air line pressure was controlled by a regulator,
and the distribution of air to the diffusers was controlled by six
airflow meters. The diffusers were made of perforated polyethylene
instead of the more conventional porous stone. Plugging of the stone
air diffusers through biological growths had been experienced in early
runs.
197
-------
Sewage Level Switch
22"
igh Water Level
I
Low Wafer Level
Primary
Ctarifier
Bypass,
Scraper
_fl Return Sludge Secondary
Solenoi Valve, <$$$£!
Moyno
Pump
Timer Contro
^J
0
0
o
o
e
o
e
o
o
e
P» /
0
0
e
o
e
Unit B
/-Randolph
/ Pump
j_i_
*
^|,
Figure 66. The Hydraulic Flow System of the Plugflow Activated
Sludge Pilot Plant
198
-------
Air Flow
Meters
Air Dif fusers
Air Filter
Air Pressure
Regulator
CD
Figure 67. Airflow System for the Activated-Sludge Pilot Plant
-------
The secondary clarifiers with a volume of 17.5 liters were designed for
a detention time of 70 minutes at a flowrate of 250 ml/min and a surface
? 2
over flow rate of 2.28 liter/cm -sec (285 gpd/ft ). A rotating scraper
was installed to minimize any wall effect from the sedimentation units.
The thickened sludge was pumped with peristaltic Randolph pumps (Randolph
Company, Houston, Texas). The leachate, that was fed to the test unit,
was stored in a 19 liter (5 gallon) polyethylene bottle which was kept at
5°C. An electrolytic pump, producing hydrogen and oxygen, was connected
to a rubber bladder located in the leachate container. Expansion of the
bladder fed the leachate at a constant rate to the test activated sludge
unit (Figure 68).
The wasting of excess activated sludge, and the sampling of sewage were
performed by Masterflex pumps (Cole Palmer, Chicago, IL) which were con-
trolled by a timer, and operated automatically for five minutes every
hour. A three-way solenoid valve, also controlled by a timer, was
installed to bypass the sewage flow to the test unit for the purpose
of controlling the feedflow rate. A schematic of the overall electrical
system is shown in Figure 69.
The activated-sludge system requires a certain amount of maintenance.
The tubing of the pumps had to be replaced after a certain period of
operation, varying from three days to six months, depending upon the flow
rate. The rubber tubing (Matheson Scientific Company, Elkfron Village,
IL) used in te Randolph pump had a life span of three to six days. The
Tygon tubing used in the Masterflex pump (Cole Palmer, Chicago, IL) had
a life span of one to six months. Packing rings on the Moyno Pumps had
to be replaced every half year, and replacement of the Stator Assembly
was necessary about once a year or more.
The anaerobic sludge digesters consisted of two reactors made of Plexiglas
26.6 cm diameter and 45.8 cm in height having a volume of approximately
24 liters (Figure 70). Each reactor was fitted with a stainless steel
helical worm stirrer, equipped at the bottom with a scraper extending
200
-------
Sewage -
Mixture Of Sewage
And Leachate
Ammete
A,C.
Source—*-
^ |W in • wi ^__j s
y///////////////.
/Refrigerator
•Composite
Sewage
Sample
Power
Control
Figure 68. Leachate Feeding and Automatic Sampling Systems
-------
A.C.
Source A
N Pi lot Box
_L
Sewage
Level
Switch
1:
V
V
• Moyno Pump (Unit A)
Moyno Pump (UnitB)
Electrolytic Pump
Spare
Spare
A.C. Source B
O -
Electrolytic Pump -Leachate
Scraper Of Settling Tank •>
Masterflex Pump For Wasting Sludge J Unit A
Masterf lex Pump For Sampling
Solenoid Valve
Masterflex Pump For Wasting Sludge ,
} UnitB
Scraper Of Settling Tank
A.C. Source C
Timer On-Off Control
Moyno Pump For Pumping Sewage From
Municipal Sewer
Figure 69. Electrical Systems for the Activated-Sludge Units
202
-------
To Gas Meter
Mechanipak Seal
Inlet
18"
Plexiglas
Vessel
Outlet
Figure 70. Schematic of the Anaerobic Digesters
203
-------
almost the whole diameter of the reactor. The stirrer was operated
continuously at 60 revolutions per minute. All reactors were tested
under pressure for gas tightness before operation. The operating volume
containing the anaerobic sludge was 15 liters.
The leachate used in this study was obtained from a laboratory scale,
simulated landfill consisting of an epoxy-coated steel tank lined with
heavy duty PVC sheeting, measuring 1.5 m in diameter and 3 m high to
which simulated rainfall was added. The solid waste used to fill this
column was milled after collection in the Hartwell section of Cincinnati,
Ohio.
Samples were collected at specific intervals from the various sampling
points in the activated sludge unit for specific analysis. A daily
composite sewage sample was collected with a Masterflex pump operated
automatically five minutes per hour at a point immediately before the
sewage was fed to the aeration tank. Grab samples of the effluent
from the secondary sedimentation tank were collected daily, filtered
through the Whatman #4 filter paper and analyzed for COD and BOD5 concen-
trations. A second effluent portion was filtered through a 0.45 ym
Millipore membrane and composited for three or four days and used for
analysis of PO~3-P and k,-N. The excess activated sludge, wasted from
" J
the third compartment of the aeration tank, was collected daily weighed
and used to calculate the sludge age.
Batch settling tests were conducted using the activated sludge collected
from the third compartment to determine the interface subsidence velocities
at varying initial sludge concentrations. The settling tests were con-
ducted at room temperature (24°C), using a one-liter graduated cylinder
equipped with a stirrer operated at 1 rpm. The slope of the linear
portion of the settling curve was used as the interface subsiding
velocity, at a given initial sludge concentration
204
-------
Both the test and control activated sludge unit were initially operated
at a food-to-micro organism (F/M) ratio of 0.3 kg BODg removed/kg MLVSS-day,
while the volume of leachate added was gradually increased from 0.5% to 1%,
2%, 3% and 4%. In order to maintain both a constant F/M ratio and a
constant MLVSS concentration, the hydraulic detention time of the test
unit was gradually increased. In the second phase of the study the volume
percentage of added leachate was kept constant at 2% while the F/M ratio
was increased from 0.3 to 0.6 and 1.0, respectively. The anaerobic
digesters were operated in a semi continuous process with a theoretical
detention time of fifteen days which was maintained by withdrawal of
1 liter of digester content and addition of one liter of fresh sludge
the solids of which consisted of approximately 50% primary sludge and 50%
activated sludge from the 0.3 F/M leachate unit. The primary sludge was
collected from the Sanitary District of Urbana.
205
-------
RESULTS AND DISCUSSION
Sewage Analysis^
During the year- long study which started on September 1973, a total of
74 composite sewage samples were analyzed for their BOD5, COD, P and N
concentrations. The distribution of the analytical results for the BODc
is shown in Figure 71 which indicates that the highest frequency was
observed in the range of 80 to 100 mg/1 ; the maximum was 464 mg/1 , the
minimum 40 mg/1 and the average 214 mg/1. As sewage strength showed
day to day variation, while the activated sludge concentration and waste
was maintained constant, the loading of the unit also experienced some
daily variation. However, when averaged over a longer period of time
the average F/M ratio was indeed 0.3 mg BOD5/mg MLVSS-day.
Hydraulic Flow Regimen of the Aeration Unit
Tracer studies using methylene blue were performed to characterize the
actual flow pattern in the aeration tank while aeration was supplied
for mixing. Figure 72 shows the normalized concentration of
tracer versus time in each of the three completely-mixed compartments
connected in series through the weirs. Results of experimentally-
determined concentration versus time were compared wiht the theoretical
values, as shown by the dotted lines in Figure 72. The theoretical
values were calculated based on the equation given below:
(n
Cn = Concentration of tracer in chamber n at time t (mg/1)
CQ = Initial concentration of tracer at time zero in chamber 1
(mg/1)
Q = Flow rate through the system (ml/min)
V = Volume of the tank (ml)
206
-------
Ill-
ro
o
0
4O
120 160 200
Sewage BOD5, mq/X.
240
280
320
Figure 71. The distribution of BOD5 concentration in the daily composite
sewage samples
-------
ro
o
CO
O
1.0
0.9
0.8
0.7
0.6
0.5
O.4
0.3
0.2
O.I
C/C0 1st Chamber (Observed)
C/C0 2nd Chamber (Observed)
-o- C/C0 3rd Chamber (Observed)
C/C0 Theoretical For The
Corresponding Chamber
Hydraulic Detention Time, V0/Q=80 Minutes
1st Chamber
3rd Chamber
100
200
Time, Minutes
300
400
Figure 72. The flow pattern of the designed aeration tank
-------
The results indicate that the initial wash-out of the first compartment is
less rapid than theoretically predicted, indicating some incompletely mixed
corners in the unit. The maximum concentration in the second compartment
was about 30% higher while that in the third compartment was 55% higher
than predicted from theoretical considerations, indicating partial plug-
flow conditions. However the time preiod that the maxima in each of the
C/CQ curves occurred agreed with that predicted from theoretical
considerations. The hydraulic flow regime is greatly affected by the
degree of mixing of theunit, which in turn depends on the amount of
aeration applied to each of the aeration chambers.
During the operation of the activated sludge unit an airflow rate of
0.5 liter of air per liter of aeration tank volume per minute was supplied
to keep the solids in suspension and maintain aerobic condition. A com-
parison of the air supplied to the laboratory activated-sludge pilot plant
with that supplied to actual sewage treatment plants (Table 18) shows
that the former amount appeared sixteen times too high. However, the air
supplied in the lab pilot plant was often limited by the mixing condition
of the liquid. In an actual sewage treatment plant, a minimum airflow
of approximately 280 1/rnin per meter length of the aeration tank is
required for adequate mixing and to avoid deposition of solids. The
airflow rate in the laboratory plant was only 93 1/min per meter length
of the aeration tank. The width of the laboratory unit, however, is
many times smaller than that of an actual activated-sludge plant.
Evaluation of Leachate Additions to the Activated Sludge Unit
Immediately after introduction of the leachate at 0.5% of the sewage
flow into the test unit on August 31, 1973, the effluent characteristics
showed some deterioration as compared to the control unit (Figure 73).
The average COD and BOD of the test unit were 28 mg/1 and 4.3 mg/1,
respectively while they were 16 mg/1 and 3.1 mg/1 for the control unit,
respectively. However, toward the end of the 0.5% leachate addition
both the effluent BOD and COD values tended to converge, while the BOD
in the effluent of the test unit decreased to 2 mg/1.
209
-------
Table 18
ro
o
The Amount of Air Supplied to the Laboratory Activated Sludge Units as Compared
with that Supplied to Actual Sewage Treatment Plants
Compartment
First
Second
Third
Air Flow Rate
1/min 1/1 -min
11.8
9.4
9.4
0.59
0.47
0.47
Dissolved Oxygen
Concentration,
(mg/1)
Test Unit Control Unit
1.0
3.0
5.0
1.2
4.0
6.0
Ratio of Air
to BOD5*
m3/kg BOD5
average
201.5
Required in the
Sewage Treatment
Riant**
nr/kg DOD5
12.8
Leachate/sewage = 2.0%; F/M = 0.3 mg BOD/mg MLVSS-day
**
The Ten States Standard
-------
s
?
?
o
§
a
o
o
Q)
UJ
400
300-
200 —
COD of the
Influent Sewage
BOD of the
Influent Sewage
Effluent COD of
the Test Unit
_Receiving Q5%
Leachate
Effluent COD of
the Control Unit
Effluent BOD
\ of the Test
Unit
Effluent BOD o1
ihe .Control Unit
Sept. 2
Sept. 9
Sept. 16
Sept. 20
Figure 73. Influent and effluent COD and BOD concentrations of the
test and control unit receiving 0.5 percent by volume of leachate
211
-------
After approximately one week of steady operation at a leachate addition
of 0.5% the amount added to the test unit was increased to 1% leachate
by volume on September 21, 1973. The sludge in the test unit started to
bulk on September 28, and the leachate addition was therefore reduced
back to 0.5%. Recovery to normal operation was obtained after one
additional week of operation. The second increase in leachate addition
from 0.5% to 1% was successful and no abnormal deter-
ioration of effluent quality was observed (Figure 74). When the amount
of leachate added was increased from 1% to 2% and from 2% to 3% no
further deterioration was noted (Figure 75 and 76). However when 4%
of leachate was added a substantial increase in effluent BOD concentra-
tion was noted (Figure 77). At this level of leachate addition the sludge
started to bulk excessively.
The impaired effluent quality and sludge bulking of the activated-sludge
unit receiving leachate may have been due to phoshpate limitations. As
a result of the leachate addition the soluble phosphate concentration
decreased significantly from 4 mg/1 to 1 mg/1 at 0.5% addition and to
0.4 mg/1 at the 1% addition (Figure 78). At the 4% addition the effluent
concentration decreased to as low as 0.03 mg/1. Although the influent
BOD/P ratio only increased from 22 to 50 after the 0.5% leachate addition,
the relative large decrease of effluent P concentration at this addition
is most likely due to the presence of 2200 mg/1 Fe in the leachate to
form precipitates with the phosphorus in the municipal sewage. The
gradual change of the color of the sludge from dark gray to brown would
also indicate the occurrence of such a process. When the effluent BOD
values at the 4% addition started to increase, the BOD/P ratio had
increased to as much as 200; a ratio often recommended for sustained
bacterial growth is 100.
A summary of the effluent BOD and COD concentrations at the different
leachate additions is presented in Figure 79. A comparison of the test
and the control unit shows that increasing leachate additions do not
greatly affect effluent BOD values except at the 4% addition. The COD
212
-------
o>
E
cf
O
CD
*-
Q)
3
UJ
280
240
160
I* 200
cf
o
CD
-------
o>
E
cT
o
OQ
o>
o>
o
0)
CO
o»
E
Cf
O
CD
c
0)
LJ
300
280
240
200
160
120
80
40
0
10
8
6
4
2
jjjj*0.3 day'1
Leachate
= 2%
34567
November, 1973
8 9
Figure 75. The Effect of 2% Leachate Addition on Effluent Quality
of the Activated-Sludge Process
214
-------
280
o»
E
Q"
o
CD
o
o>
i
V
240
200
160
120
80
40
0
o>
E
Q"
O
ffl
0)
3
UJ
8
6
4
2
_T7=0.3day-'
Leochate
Sewage
= 3%
Control Unit
13 14 15 16 17 18 19 20
November, 1973
Figure 76. The Effect of 3% Leachate Addition on Effluent Quality
of the Activated-Sludge Process
215
-------
0>
£
cf
O
CO
0)
o>
O
J
0)
CO
o>
e
cf
O
ffl
Q)
UJ
200
160
120
80
40
0
>7 28 29 30
November
234
December
Figure 77. The Effect of 4% Leachate Addition on Effluent Quality
of the Activated-Sludge Process
216
-------
0»
1 8
6
4
2
0
c
Q)
u
/-
/
Kj-N
>
I 2 3 .4
Leachate / Sewage % (Volume)
Figure 78. The Effect of Leachate Additions on the Concentration
of P04-3-P and N(Kj) in the Effluent of the Best Unit
217
-------
however, shows a gradual increase from 30 mg/1 to 71 mg/1 at the highest
leachate addition. The yellow color of the effluent of the test unit
showed an increase parallel to the COD values.
The second phase of the study was used to evaluate the effect of increasing
loadings on the effluent characteristics while maintaining a leachate
addition of 2%. The results showed that the BOD of the test unit was con-
sistantly higher than that of the control unit at a F/M ratio of 0.6
(Figure 80) and 1.0 (Figure 81). At the highest loading the effluent BOD
of the test unit reached a concentration of as high as 10 mg/1, indicating
that with leachate additions the activated sludge unit is preferrably
operated at low F/M ratio's.
The sludge settling characteristics show that the addition of 0.5%
leachate resulted in a decreased settling rate of the sludge interface;
increasing the leachate addition at the same loading, however, did not
further deteriorate the settling characteristics (Figure 82). A conclu-
sion parallel to the above one was reached with regard to the effluent
phosphate concentration, possibly indicating that the low soluble
phosphate concentrations in the unit interfere with the floe formation
of the activated sludge. The impairment of the sludge settling was even
more noticeable at the higher sludge loading due to a greater phosphorus
limitation.
A further explanation of the observed settling behavior in the test unit
can be found in the differences in effluent molecular weight distribution.
For that purpose a 200 liter amount of effluent was collected from both
units and filtered through a 0.45 m Millipore membrane whereafter the
filtrate was concentrated using an AS-197 reverse osmosis membrane.
The RO retenate was further fractionated using a 500 MW UF membrane as
described by DeWalle and Chian (1974). While 43% of the soluble TOC
of the control unit was retained with the 500 MW UF membrane, only 35%
was retained in the effluent of the test unit. When the 500 MW UF
retentate was applied on a Sephadex G-75 column the TOC distribution
showed that the excluded fraction, eluted first, and having a molecular
218
-------
o
o
CD
c
0)
o»
E
O
o
CD
c
-------
o»
E
M
o
o
CD
Q>
O»
O
o»
E
d"
O
ffl
c
Q>
UJ
200
160
120
80
40
0
Leachote
Sewage
= 2% TT =
F_
M
-I
Control Unit-
j i i i
15 16 17 18 19 20 21 22 23 24
January, 1974
Figure 80. Effluent Quality During 2% Leachate Addition
at the 0.6 F/M Ratio
220
-------
240
200
| 160
Q"
2 120
o>
Q)
80
40
0
o»
e
Q"
o
m
c
-------
I
>
100
80
60
40
20
10
8
6
I I
0.3
0.3
0.3
0.3
0.6
Leachate
Sewage
0%
0.5%
1.0%
3.0%
2.0%
D
X
v
I I I I I
0.1 0.2 0.4 0.6 0.8 10 2.0
Sludge Concentration %
4.0 6.0 8.0 10
Figure 82. The Effect of Leachate Addition on the Sludge
Settling Characteristics
222
-------
weight greater than 30,000, was of equal size in both the effluent of
the test and control unit. The low molecular weight fraction eluted
last from the Sephadex, however, was smaller in the effluent of the
test unit. Considerable differences were further noted for the relative
composition of the different molecular weight fraction. The result of
the aromatic hydroxyl-carbonyl- and carboxyl group analysis are generally
comparable. However, noticeable differences were observed for the
results of the carbohydrate analysis (Figures 83 and 84). While approx-
imately 28% of the high molecular weight fraction in the control unit
consisted of carbohydrates, only 21% of the organics in the high molecular
weight fraction of the test unit consisted of carbohydrates. While the
low molecular weight fraction of the organics in the control unit did
not contain much carbohydrates, much larger concentrations were noted
in the similar fraction of the test unit. Since the carbohydrates in the
high molecular weight fraction tend to affect the sludge flocculation,
the lower content in the high molecular weight organics of the test unit
may be responsible for the decreased sludge settling rates resulting
from smaller sludge floes (DeWalle and Chian, 1974b). The low molecular
weight fraction retained with the 500 MW UF membrane and included in
the Sephadex 6-75 generally shows the highest adsorptive capacities as
compared to other molecular weight fractions. The higher carbohydrate
content in the low molecular weight fraction of the test unit could
therefore result in a lower adsorptive capacity in an activated carbon
column.
The last phase of the activated sludge study therefore evaluated the
adsorptive behavior of the soluble organic matter in the effluent of
both units. The effluent was therefore filtered through a 0.45 urn
Mi Hi pore membrane and passed through a 30 cm long carbon column at a
superficial flow velocity of 10 cm/min (2.5 gpm/ft2) to give an empty
bed detention time of 3 minutes. The results in Figure 85 clearly show
that at the low F/M loading the organics in the control unit adsorb
better than those of the test unit, which may be due to the differences
223
-------
ro
ro
56
48
0> 40
M
O
u
J3
O
< ^
M
J>
2 24
•o
>.
5 l6
r 60
- 52
— 5 20
_<§
— 12
— 4
— 0
— 300
— 260 —
— 20
— 0
Elution Volume, ml
Figure 83. Elution Profile of the 500 MW UF Retentate of the Control Unit on a G-75 Sephadex
Column as Characterized by TOC, Carbohydrates and Carbonyl Groups
-------
ro
en
ISO
.2 100
O 50
- 90
•— 8O
— 70
— . 60 —.
_ O
—-S so
u
->
<
s*
030
5
— 20
— 10
— 0
IOO
Elution Volume, mi
Figure 84. Elution Profile of the 500 MW UF Retentate of the Test Unit Receiving 0.5% Leachate
on a G-75 Sephadex Column as Characterized by TOC, Carbohydrates and Carbonyl Groups
-------
in both the molecular weight distribution and the composition of each
molecular weight fraction. When the F/M ratio was increased to 50% for
both units. Fractionation of the low molecular weight organics that are
retained by the RO membrane but permeate the 500 MW UF membrane showed
that their magnitude increased in the effluent of both units. Further-
more the percentage retained with the RO membrane also decreased at the
higher loading (Chian et^ aj_., 1977). At high loadings bacteria excrete
more low molecular weight organics which generally have a lower adsorptive
capacity towards carbon than larger organics such as fulvic acids.
Evaluation of Leachate Sludge Addition to the Anaerobic Digester
The performance of the anaerobic digesters receiving waste sludges from
both the test and the control-activated sludge units, were monitored
daily for gas production. Figure 86 shows that the daily gas production
rates of the two anaerobic digesters.! is approximately equal. There
appears therefore no deleterious effect on the performance of the
anaerobic digester receiving sludge from the test unit treating the
combined leachate and municipal sewage.
The characteristics of the sludges fed to the digesters, and the digested
sludge are listed in Table 19. In spite of the high concentration of
iron found in the sludge from the test unit, no undesirable effects were
observed with regard to the operation of the unit. The release of the
iron from the solids to the supernatant by the anaerobic digestion process
was negligible and represented only 0.3% of the total iron content.
It is important to establish the fate of iron and phosphate during the
anaerobic digestion of sludges. The iron in the waste-activated sludge
may be present as either ferric hydroxide or ferric phosphate precipitates.
In the anaerobic digester, most of the iron is present in the reduced
Fe form. The iron concentrations in the supernatant of the test and control
digesters were 4.0 mg/1 and 1.6 mg/1 respectively; these values are
higher than predicted from the FeCOg solubility, which in turn depends on
the partial pressure of C02 in the gas phase. Furthermore, because
226
-------
1.0
Effluent of Activated Sludge
Unit Recievirvg Leachate at
Low F/M Load ing ;C0= 44 mg/l COD
-oS
\
Effluent of Activated Sludge
Unit Recieving Leachate at
High F/M Loading;C0 = 54.2 mg/l COD
I
1.0
5,0 100
Number of Bedvolumes
-j- 0.5
co
Effluent of Activated Sludge
Unit af High F/M Loading;
C0=22.9 mg/l COD
Effluent of Activated Sludge
Unit at High F/M Loading;
C0=21.3 mg/l COS
Effluent of Activated
Effluent of Activated
Sludge Unit at low
n.g/. COO
COD
50 100
Number of Bedvolumes
Figure 85. Adsorptive Character of the Soluble Organic Matter in the
Effluent of the Test and the Control Unit Both Operated
at F/M ratios of 0.3 and 0.6
227
-------
ro
ro
C»
O
Q
(A
a
O
Tank A (50°70 Primary Sludge
+50°70 Activated Sludge)
TankB (50% Primary Sludge
+50% Activated Sludge
From Leachate Treat-
ment Unit)
I I I I
19 21 23 25 27 29
June, 1974
13 15 17 19 21 23 25 27
July, 1974
Figure 86. Daily Gas Production from the Control- and Test Anaerobic Digester
Treating Waste-Activated Sludge
-------
Table 19
Anaerobic Digestion of the Waste-Activated
Sludge from the Leachate Treatment
PH
Total Solid (%)
Volatile Solid (%)
Ratio VS/TS
VS Reduction (5)
Total Fe mg/1
Total P mg/1
Gas production, I/kg
VS Reduction
Feed
Control
Unit
7.0
2.54
1.85
0.728
46
480
290
288
Sludge
Test
Unit
7.0
3.10
2.10
0.677
42
1.360
282
278
Digested
Control
Unit
6.9
1.72
1.00
1.6*
14.13*
Sludge
Test
Unit
7.0
2.19
1.21
*
4.0
12.25*
Concentrations in the supernatant
229
-------
different levels of soluble iron were present in these two units, it is
believed that the solubility of the ferrous carbonate is not the limiting
factor for the solubilization of iron in the digester.
The ferrous iron may also react with sulfide or phosphate to form the
relatively insoluble ferrous sulfide (K = 4 x 10 ), or vivianite
on SO
(ferrous phosphate, KSQ = 10 ). However, the amount of sulfide ion
in the anaerobic digester was relatively low due to the low sulfate
concentration in the sewage, as the average concentration of sulfate
is approximately 100 mg/1 as SO^2 or 33 mg/1 as S"2, and the total
phosphorous concentration in the digester is 300 mg/1 only, not enough
is available to precipitate all of the 1,360 mg/1 iron present in the
test unit.
The only remaining mechanism that could explain the retention of iron
in the suspended solids of the digested sludge is an adsorption process
in which the iron hydroxides adsorb onto .the sludge particles. This
possibility is supported by the result of the analysis shown in Table 19,
as a higher concentration of ferrous ion in the supernatant, corresponds
to a larger amount of iron adsorbed onto the sludge. This agrees with
general equilibrium conditions established between the adsorbate iron
and adsorbent sludge. The lower soluble total-P observed in the super-
natant of the digested sludge from the test unit corresponds with the
higher concentration of iron on the supernatant. This correlation was
also noted by Singer (1972).
230
-------
REFERENCES
Albertsson, J. G., et a_l_. "Investigation of the Use of High Purity
Oxygen Aeration in the Conventional Activated Sludge Process,"
Water Pollution Control Research Series 17050 DNW o5/70, 1970.
Andersen, D. R. et al_. "Pharmaceutical Wastewater: Characteristics and
Treatment," Water and Sewage Works, 118, IW/2 (1971).
Boyle, W. C. and Ham, R. K. "Biological Treatability of Landfill Leachate,"
Journal Water Pollution Control Federation, 46, 860 (1974).
Brosig, A. et al_. "Activated Sludge Joint Treatment of Pulp and Paper
Effluent with Municipal Sewage," Tappi, 54, ]86 (1971).
Chian, E. S. K. and DeWalle, F. B. "Sanitary Landfill Leachates and their
Treatment," Jour. Environ. Engr. Div. ASCE. 102, 411 (1976).
Chian, E. S. K., Chang, Y., DeWalle, F. B. and Rose, W. "Combined
Treatment of an Organic Chemical Wastewater by Activated Sludge
Followed by Activated Carbon," Proceedings 30th Annual Purdue
Industrial Waste Conference, 966-572 (1976).
Chian, E. S. K., Cheng, S. S., DeWalle, F. B. and Kuo, P. P. K. "Removal
of Organics in Sewage and Secondary Effluent by Reverse Osmosis,"
Progr. Water Technology. 9^ 761 (1977).
City of Dallas, Oregon "Combined Treatment of Domestic and Industrial
Waste by Activated Sludge," U.S. Environmental Protection Agency,
Water Pollution Control Research Series, 12130 EZR, May (1971).
Clingenpeel, W. H. and Jones, M. K. "Design for Joint Treatment of
Municipal and Paper Mill Waste at Lynchbury Virginia," Water and
Sewage Works. 121, 4, R12 (1974).
DeWalle, F. B. and Chian, E. S. K. "Removal of Organic Matter by Activated
Carbon Columns," Jour. Environ. Engr. Division ASCE. 100, 1089 (1974).
DeWalle, F. B. and Chian, E. S. K. "The Kineti-s of Formation of Humic
Substances in Activated Sludge Systems and their Effects on Floc-
culation," Biotechnology and Bioengineerinq. 1_4, 739 (1974b)
Eidsness, F. A. et a\_. "Biological Treatment of Citrus Processing
Wastewaters," Proceedings 2nd National Symposium Food Processing
Wastes, p. 271 (1971). y
El-Gohary, F. A. "Biological Aspects of Waste and Sewage Combined
Treatment Via Complete Mixed Activated Sludge," Water Research,
i, 103 (1971).
231
-------
Hashimoto, S. ejt al_. "Chemical Flocculation Treatment and Activated
Sludge Treatment of Dying Industry Wastewater and Combined Sewage,"
Mizu Shori Gijutsu. 1_3, 1 (1972).
Longdon, D. "Purification of Wastes from a German Yeast Plant," Proceedings
24 Industrial Waste Conference, Purdue University Extension Series,
135. 770 (1969).
Malhotra, S. K., Parrillo, T. P. and Hartenstein, A. G. "Anaerobic
Digestion of Sludge Containing Iron Phosphates," Journal of the
Sanitary Engineering Division, ASCE, No. SA5, October 1971, pp.
629-645.
Metcalf & Eddy, Inc. "Wastewater Engineering, Collection, Treatment and
Disposal," McGrawHill, New York (1972).
Morgan, W. E. and Fruh, E. G. "Metabolic Mechanisms Not Causes of
Activated Sludge High P Removals," paper presented at the 46th
Annual Water Pollution Control Federation Conference (1973).
Niles, C. F. and Etzel, J. E. "The Lafayette Story," Jour. Water Pollut.
Control Federation. 43, 623 (1971).
Poon, C. P. C. "Biodegradability and Treatability of Combined Nylon and
Municipal Wastes," Journal Water Pollution Control Federation. 42_,
100 (1970).
Singer, P. C. "Anaerobic Control of Phosphate by Ferrous Iron," Journal
Water Pollution Control Federation. Vol. 44, No. 4, April (1972).
Thomas, E. A. "Phosphate Removal by Recirculating Iron Sludge," Journal
Water Pollution Control Federation, Vol. 44, No. 2, February (1972).
Vockel, K. G. ejt al_. "Joint Treatment of Municipal and Pulping Effluents,"
Jour. Water Pollution Control Federation. 4£, 634 (1974).
232
-------
VI
ESTIMATED COSTS FOR LEACHATE TREATMENT
CONCLUSIONS
It was found that aerated lagoons provide the least expensive method of
treating leachates having comparatively low BOD5 values (e.g., 5,000 mg/1)
and relatively high flow rates (e.g., 20 gal/min). As the BOD5 level of
the leachate increases at the same high flow rate, the treatment cost
using anaerobic filters becomes increasingly attractive, and at a BODc
value of 25,000 mg/1 it equals that of the aerated lagoon process if credit
for the methane gas produced is deducted. Taking into consideration the
treatment level and thus the effluent quality, the combined treatment of
leachate and domestic wastewater using the activated sludge process
becomes most desirable because of the high dilution factor, especially
when the leachate BOD5 levels .are low and the flow rates are high.
In minimizing the impact of the treated leachate on the environment, the
treatment of leachate using aerated lagoons and physical-chemical treatment
processes (such as a combination of slow sand filtration, activated carbon
adsorption, and reverse osmosis) is most effective, although such treatment
would cost somewhat more than combined treatment of leachate and domestic
wastewater with activated sludge at high leachate flow rates and low BODg
levels. The difference in treatment costs between the two approaches
narrows, however, as the leachate flow rates decrease and the BOD5 levels
increase.
233
-------
INTRODUCTION
The cost estimates presented in this section are intended to aid solid waste
management planners, decision makers, and design engineers in selecting
alternative on- and off-site leachate treatment methods to achieve a
designated level of treatment for leachates of various strengths and flow
rates.
The leachate flow rates and BOD,, levels examined in this study were 2 and
20 gal/mint and 5,000 and 25,000 mg/1, respectively. These values were
selected so that a broad range of leachates produced by various landfill
sites could be considered. In view of the high BODg levels and the objective
of low treatment cost, biological processes were selected for first-stage
treatment. The costs of physical-chemical treatment processes were
estimated only for cases where such processes would be used to treat the
effluents from aerated lagoons and anaerobic filters.
The cost comparison is based upon the results of this study on leachate
treatment, on the assumption set forth in this chapter, and on the average
estimating data from two recent EPA reports (Black and Veatch, 1971;
Bechtel, 1975). All costs were then updated to August 1977 according to
the Water Quality Office Wastewater Treatment Plant Index (Engineering
News Record. January 1971, January 1975, August 1977).
The user must recognize certain limitations in using the cost estimates
presented. Estimates based upon treatment data obtained from the laboratory
studies and upon average cost estimating data should be considered only of
sufficient accuracy for use in reaching broad, general conclusions. When
used with sound judgment by knowledgeable persons, the data can be useful
in selecting appropriate treatment processes for a specific leachate based
upon known levels of contamination and flow rates.
1 gallon = 3.79 litre
234
-------
It should also be realized that the cost estimates presented in this
section must be revised and updated periodically since waste treatment
costs are continually changing and since new and improved waste treatment
techniques and methodologies are rapidly emerging.
235
-------
METHODS AND PROCEDURES
Activated Sludge
The cost estimates for the combined treatment of leachate and municipal
wastewater were based on the addition of leachate at a level of 1 percent
by volume to the sewage when the nominal daily flow rate of the leachate
is 30,000 gallons (20 gpm). The proportion of leachate added was reduced
to 0.1 percent by volume when its flow rate decreased to 3,000 gallons
per day (2 gpm). If the leachate represents 1 percent of the flow into
the municipal treatment plant, the overall flow rate of the plant must
be 3 MGD.
The BOD5 level of the municipal sewage introduced into the activated sludge
system from the primary sedimentation basin is typically 140 mg/1. When
1 percent of leachate having a BOD5 of 25,000 mg/1 is added, the BOD5 of
the combined wastewater becomes 390 mg/1 (140 + 1% of 25,000), corresponding
to a 279 percent increase in sewage strength. To maintain the same
effluent quality, the municipal treatment plant should therefor have the
capacity to treat the equivalent of 8.4 MGD of sewage having a BODg of
140 mg/1. Also, a 280 percent decrease in sludge settling rate was also
observed in this laboratory when leachate was added at 1 percent. Again,
to counteract this effect the municipal treatment plant facilities should
be expanded to the equivalent of 8.4 MGD to maintain the same effluent
quality. The increase in treatment cost resulting from the need for larger
aeration and sedimentation basins, a greater air supply, larger anaerobic
digesters, and increased-sludge disposal is attributable to treating the
specific leachate under study. The additional costs of transporting
the leachate through pipelines and by trucking were also considered.
The calculations of the aeration basin volume requirements were based upon
an F/M ratio of 0.3 day"1 and a MLVSS of 2,000 mg/1. A yield of 0.65 g
VSS/g BOD5 was used for sewage and 0.80 g VSS/g BOD5 for leachate. A
factor of 0.8 was used to convert MLSS to VSS.
236
-------
Cost figures based upon the total costs of treating 1000 gallons of sewage
influent having a BOD5 of 140 mg/1 and at a specific plant capacity as
given by Bechtel (1975) were used to arrive at the costs for conventional
activated sludge treatment, anaerobic digesters, and sludge drying. A
factor of 1.25 was then used to update the costs to August 1977 based on
the average cost indexes given by Engineering News Record (McGraw Hill)
between January 1975 and August 1977. The costs of transporting leachate by
pipeline were calculated assuming the use of 3-in.-diameter, schedule-40
PVC pipe for a distance of 15 miles at an available pumping head of
75 psig. Data given in Costs of Process Equipment (Chemical Engineering,
March>1964) were used to estimate the costs of installed cross-country
pipeline. A factor of 2.71 was used to update these costs to August 1977.
For smaller quantities of leachate, a trucking cost of 7.5^/100 Ib of
liquid for a distance of 30 miles round trip was used.
Aerated Lagoon
The design criteria for an aerated lagoon to treat leachate were as follows:
BODg removal 99%
Mean Cell Residence 90 days (BOD = 25,000 mg/1)
Time 30 days (BOD = 5,000 mg/1)
MLVSS 8000 mg/1 (BOD = 25,000 mg/1)
6000 mg/1 (BOD = 5,000 mg/1
Y 0.8 g VSS/g BOD
kd 0.025 day"1
Aerator 2 Ib 02/hp-hr under field conditions
BOD5:P 150:1
BOD5:N 20:1
Sludge Production g VSS/0.8
o
Sedimentation Basin 400 gpd/ft
Design equations given in Wastewater Engineering (Metcalf & Eddy, 1972)
were used to determine variables such as the volume of the aeration basins
and the oxygen requirements.
237
-------
The capital (including engineering costs), operation, and maintenance
y
costs were estimated from the Black and Jteatch report (1971). The installed
costs of the basins were increased 15% to cover the piping costs. A factor
of 1.78 was used to update the costs to August 1977 based on the average
cost indexes given in Engineering News Record (McGraw Hill) between
January 1971 and August 1977. The costs of ammonia and phosphoric acid
were taken from the August 1977 issue of Chemical Marketing Reporter
(Schnell Publishing Co.) Costs of $120/ton for ammonia and $3.20/100 Ib
for agricultural-grade phosphoric acid (52-54% available phosphoric acid)
were used. The land cost was estimated at $5,000/acre. For electrical
energy, a cost of 3 cents/KWH was used. An average pay rate of $10.00
per man-hour, including overhead costs, was used in estimating operating
and maintenance costs. Capital equipment was depreciated on a straight-
line basis over a period of 10 years for moving equipment and 20 years for
fixed installations. Interest charges were computed at the rate of 4
percent of the initial capital investment over the entire period.
Anaerobic Filter
The design criteria used for estimating the costs of treating leachate
with an anaerobic filter were as follows:
BOD5 Removal 97%
Loading 0.31 Kg BOD5/M3 day
Yield, VSS 0.024 g VSS/g BODs
Detention time 87 days (leachate BOD = 25,000 mg/1)
16 days (leachate BOD = 5,000 mg/1)
Recirculation Ratio 20:1
Yield, Methane 1.75 liters (78% CH4)/g TOC
The capital investment estimate was based on the use of a rubber-lined
steel tank and B.F. Goodrich vinyl core packings along with the necessary
ancillary equipment such as pumps and piping. The costs of the rubber-
lined vessel and pumps were obtained from equipment cost data (Chemical
Engineering, ENR = 400 for materials; Chemical Engineering, March 16,
1977). Factors of 2.75 and 2.71, respectively, were used to update the
238
-------
costs to August 1977. A verbal quote was obtained from B.F. Goodrich Co.
for the vinyl core packings ($2.75/cu ft installed) on August 19, 1977.
The operating costs were computed in the same way as for the aerated
lagoon; the maintenance costs used, however, were 5 percent of the capital
investment on an annual basis.
Slow Sand Filtration
The cost estimates for the slow sand filtration unit were based upon a
single data point: a unit of comparable capacity installed recently by
Daily and Associates Engineers, Inc. (Champaign, IL 61820). A rather
conservative flow rate of 8 gal Ions/day-ft2 was used in the design of
the filter. The operating and maintenance costs were calculated using the
previously described methods of computing costs such as depreciation,
interest, and maintenance.
Activated Carbon
Based upon both the batch isotherm and the column data, the design criteria
for the activated carbon unit were as follows:
Influent TOC* 200 mg/1 from aerated lagoon (AL)
600 mg/1 from anaerobic filter (AF)
TOC Removal 75%
Color Removal 90% measured at 400 nm
(X/M)Q 0.15 g TOC/g Carbon (AL)
0.10 g TOC/g Carbon (AF)
Activated Carbon Filtrasorb 400
Contact Time 2 hours
Pretreatment Slow sand filter
The capital investment was estimated by extrapolating the cost data provided
by Bechtel (1975). After being updated to August 1977, this cost was
$1.0/1000 gal. The operation and maintenance costs were also obtained from
Bechtel (1975). The cost of activated carbon used was $0.65/lb.
*n5VSnlUen/iT°C f11! b! 1/5 of that Wh11e Usin9 leachate having a BOD5
of 5,000 mg/1 instead of 25,000 mg/1.
239
-------
Reverse Osmosis
The design criteria for the reverse osmosis process were as follows:
Pretreatment Slow Sand Filtration, activated carbon
column, and 5 y prefilter
Product Water Recovery 90%
IDS Removal 95%
TOC Removal 80%
Module DuPont Hollow Fiber
Pressure 600 psig
The cost estimates for reverse osmosis using DuPont's B-10 modules were based
upon average cost data obtained from the manufacturers as of August 1977.
Average costs of $1.67/galIon/day (gpd) and $3.33/gpd were used for the
30,000 gpd and 3,000 gpd plants, respectively. The power requirements are
20 hp for a 30,000 gpd unit and 2 hp for a 3,000 gpd unit. Unlike activated
carbon treatment processes, operating costs for reverse osmosis treatment
are relatively insensitive to the levels of contaminants in the feed. The
depreciation costs are based on 10-year life for the mechanical parts and
a 3-year life for the membrane modules. Other costs for operation and
maintenance are similar to those used in the previous calculations.
240
-------
RESULTS AND DISCUSSION
Using the procedures outlined in the previous section, cost estimates were
obtained for treating leachates at two BODs levels and flow rates. The
results are presented in Table 20. To estimate the cost of treating combined
leachate and activated sludge, fixed costs of $2/1000 gal and $6/1000 gal
of leachate were added for transporting leachate at flow rates of 20 gpm
(30,000 gpd) and 2 gpm (3,000 gpd), respectively. The $2/1000 gal value is
for transporting leachate by pipeline, while the $6/1000 gal figure is for
hauling it by trucks within a radius of 15 miles (i.e., 30 miles round trip).
To transport leachate at a flow rate of 2 gpm by pipeline would cost around
$15/1000 gal even if the pipeline is depreciated over a 20-year period. It
should be realized that the use of a 20-year period to depreciate fixed
installations may not be realistic, since the strength of the leachate
produced would be reduced greatly over such a period.
It can be seen from Table 20 that the aerated lagoon provides the least
expensive method of treating leachate having a comparatively low BODs value
(e.g., 5,000 mg/1) and relatively high flow rate (e.g., 20 gpm). At a
20 gpm flow rate, as the BOD5 level of the leachate increases the treatment
cost using anaerobic filters becomes increasingly attractive, and at a
BOD5 value of 25,000 mg/1 it equals that of the aerated lagoon process if
credit for the methane gas produced (e.g., at $1.50/1000 cu ft) is deducted.
Taking into consideration the treatment level and thus the effluent quality,
the combined treatment of leachate using the activated sludge process becomes
most desirable because of the high dilution factor, especially when the
leachate BOD5 levels are low and the flow rates are high.
It should be noted that at high leachate BOD5 levels (i.e., 25;000 mg/1) and
at both flow rates under study the costs of complete leachate treatment using
aerated lagoons and physical-chemical processes (such as a combination of slow
sand filtration, activated carbon adsorption, and reverse osmosis) are compar-
able to those for combined treatment of leachate and domestic wastewater by
241
-------
TABLE 20
A SUMMARY OF COST ESTIMATES FOR LEACHATE TREATMENT
Typical
Leachate Effluent COD
(gal/min)* (mg/1)
Influent BOD > mg/1
Activated Sludge (AS)
(Combined treatment)
Aerated Lagoon (AL)
Anaerobic Filter (AF)
AL-Sand Filter (SF)
-Activated Carbon (AC)
AL-SF-AC-Reverse
Osmosis (R0)#
AF-SF-AC
AF-SF-AC-RO#
20
2
20
2
20
2
20
2
20
2
20
2
20
2
25,000
30
30
500
500
1500
1500
125
125
25
25
375
375
75
75
5,000
30
30
100
100
300
300
25
25
5
5
75
75
15
15
Costs of Treatment
($/1000 gal leachate)*
25,000
23.6
41.4
17.9
31.6
22.1(17.9)t
43 (38.8)
25.7
39.9
27.6
44.6
32.8(28.6)
54.2(50)
34.7(30.4)
58.9(54.3)
5,000
6.0
11.9
4.1
10.0
6.8(5.9)
17.7(16.8)
7.3
13.7
9.2
18.4
10.6(9.7)
22.0(21.1)
12.5(11.5)
26.7(25.4)
*1 gal = 3.79 litre
tNumbers shown in parenthesis indicate the costs of treatment after deducting
the credit for methane produced @ $1.50/1000 cu ft.
#After RO treatment the total dissolved solids (IDS) decreased to 300 mg/1
and 60 mg/1 for influent leachate #00 concentrations of 25,000 mg/1 and
5,000 mg/1, respectively. c
242
-------
activated sludge. The effluent quality is, however, far better when the
complete treatment scheme discussed above is used. The COD value of the
effluents from the complete treatment scheme and from the combined treatment
process are 5 and 30 mg/1, respectively, whereas the TDS levels are 300 and
750 mg/1, respectively. In addition, the color-bearing materials are
removed completely using the complete process. Therefore, to minimize the
impact of the treated leachate on the environment the complete treatment of
leachate using aerated lagoons and physical-chemical treatment processes
is most desirable. As leachate BODs levels decrease, however, the percentage
difference, in treatment costs would be somewhat greater.
Although the information presented in Table 20 can be of particular value
in establishing the effect of changes in leachate flow rates and BOD5
levels on treatment costs, the specific costs presented should be used with
caution. It should be realized that the circumstances of a particular
situation may alter treatment costs drastically. For example, if the
leachate were allowed to discharge into a municipal wastewater treatment
plant having excess capacity, the cost (in terms of the surcharges paid)
could be substantially lower than the estimates presented here.
243
-------
REFERENCES
Bechtel, Incorporated, A Guide to the Selection of Cost-Effective Wastewater
Treatment Systems, Report No. PB-244-417, National Technical Information
Service, U.S. Department of Commerce (July 1975).
Black and Veatch, Consulting Engineers, Estimating Costs and Manpower
Requirements for Conventional Wastewater Treatment Facilities,
Report No. 17090 DAN 10/77, U.S. Environmental Protection Agency
(October 1971).
"Costs of Process Equipment", Chemical Engineering (March 16, 1964).
Metcalf & Eddy, Inc., Wastewater Engineering: Collection, Treatment,
Disposal. New York: McGraw-Hill (1972).
244
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
i. REPORT NO.
EPA-600/2-77-186b
3. RECIPIENT'S ACCESSION-NO.
4. TITLE ANDSUBTITLE
Evaluation of Leachate Treatment
Volume II
Biological and Physical-Chemical Processes
5. REPORT DATE
November 1977 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Edward S. K. Chian and Foppe B. DeWalle
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Civil Engineering
University of Illinois at Urbana-Champaign
3217 Civil Engineering Building
Urbana, Illinois 61801
10. PROGRAM ELEMENT NO.
1DC618; SOS1; TASK 26
11. CONTRACT/GRANT NO.
68-03-0162
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory—Gin. ,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Volume II of a two-volume final report. See also Volume I. Characterization of
Leachate, EPA-600/2-77-186a, NTIS PB-272 855. P.O.-Dirk Brunner 513/684-7871.
16. ABSTRACT
A completely mixed anaerobic filter was found to effectively remove organic matter
concentrations in high-strength solid waste leachate over a range of organic loadings
and shockloads. Recirculation eliminated the need for buffer solutions. Testing of
a fixed film biological reactor model showed that the substrate removal rate is
primarily affected by substrate concentration, specific surface area, flow rate, and
temperature of the unit. Studies of the biological aerated lagoon or extended aera-
tion process were conducted in six completely mixed reactors (no recycle) fed with
undiluted leachate. Phosphate requirements of the aerobic biomass were extensively
evaluated. .Kinetic constants were calculated for optimum conditions. The settling
and dewatering characteristics of the sludge from the aerated lagoons were studied.
The combined treatment of leachate and municipal sewage in a conventional plugflow
activated sludge unit was found to effectively treat high strength leachate. The
test unit was not able to treat the high strength leachate at >_ 4% of the municipal
sewage flow rate. Physical-chemical treatment methods are not effective in removing
large quantities of organics from the leachate and that biological pretreatment is
required; these methods were therefore tested using aerated lagoon effluents. Ozona-
tion, activated carbon, anion exchange resins, and reverse osmosis were studied.
Treatment costs were estimated for leachate flows of 7.6 and 76 £/min containing 5000
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Waste disposal
Refuse disposal
Waste Treatment
Activated carbon treatment
Ion exchanging
Aerobic processes
Anaerobic processes
b.IDENTIFIERS/OPEN ENDED TERMS
Solid waste
Leachate treatment
Sanitary landfill
c. COSATI Field/Group
13B
18. DISTRIBUTION STATEMEN1
RELEASE UNLIMITED
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
265
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
245
*U.S GOVERNMENT MINTING Of FICE:197«— 757-140/66Z6
------- |