Comparative studies of aerobic and anaerobic landfills using simulated landfill lysimeters

MISSING IMAGE

Material Information

Title:
Comparative studies of aerobic and anaerobic landfills using simulated landfill lysimeters
Physical Description:
xv, 231 leaves : ill. ; 29 cm.
Language:
English
Creator:
Kim, Hwidong
Publication Date:

Subjects

Subjects / Keywords:
Environmental Engineering Sciences thesis, Ph. D
Dissertations, Academic -- Environmental Engineering Sciences -- UF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2005.
Bibliography:
Includes bibliographical references.
Additional Physical Form:
Also available online.
Statement of Responsibility:
by Hwidong Kim.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 028423298
oclc - 847496256
System ID:
AA00025781:00001

Full Text











COMPARATIVE STUDIES OF AEROBIC AND ANAEROBIC LANDFILLS
USING SIMULATED LANDFILL LYSIMETERS















By

HWIDONG KIM














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005































Copyright 2005

by

Hwidong Kim































This document is dedicated to my parents and loving wife














ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. Timothy G. Townsend, for showing such great patience as a mentor. He gave me this great opportunity to study on the field of solid waste. He also showed me the way of living as an engineer, professor, and a family man. I cannot forget his tears when Mr. Townsend passed away. I would also like to thank my committee members, Dr. Angela Lindner, Dr. Frank Townsend and Dr. Roger Nordstedt, and my other spectacular faculty members, Dr. David Chynoweth, Dr. Gabriel Bitton and Dr. Matthew Booth, who gave me great help.

I wish to thank my colleagues in the Solid and Hazardous Waste Research group, in particular, Brajesh Dubey, Qiyong Xu, Kim Cochran, Steve Musson, Aaron Jordan, Pradeep Jain, Jaehak Ko, Murat Semiz, Judd Larson, and Yong-Chul Jang, a faculty memeber of Chung-Nam University in South Korea. I also thanks goes to my first mentor and graduate advisor, professor, Byung-Ki Hur, a faculty of Inha University in South Korea.

A special thanks goes to my mother as well as my father, who is fighting against disease. Finally, greatest thanks go to my wife, Eunkyoung Choi, for her patience, encouragement, and love.










iv














TABLE OF CONTENTS

pne

A CKN OW LED G M EN TS ................................................................................................. iv

LIST O F TA BLES ........................................................................................................... viii

LIST O F FIG U RES ............................................................................................................. x

CHAPTERS

1. IN TROD UCTIO N .......................................................................................................... I

1. 1 Problem Statem ent ................................................................................................. 1
1.2 Objectives .............................................................................................................. 2
1.3 Research Approach ................................................................................................ 3
1.4 O utline of D issertation ........................................................................................... 5

2. COMPARATIVE STUDIES OF LEACHATE AND GAS QUALITY OF
AEROBIC AND ANAEROBIC SIMULATED LANDFILL BIOREACTORS ......... 6

2.1 Introduction ............................................................................................................. 6
2.2 M aterial and M ethods ............................................................................................ 7
2.2.1 G eneral D escription of the Lysim eter ......................................................... 7
2.2.2 Tem perature Control ................................................................................... 8
2.2.3 Fabricated W aste Stream ............................................................................. 9
2.2.4 A ir Injection ............................................................................................... 10
2.2.5 Leachate and G as Analysis ........................................................................ 10
2.2.6 Recovery of the A naerobic Lysim eters ..................................................... 11
2.2.7 Prediction of W aste M ass Loss ................................................................. 12
2.3 Results and D iscussion ........................................................................................ 12
2.3.1 pH .............................................................................................................. 13
2.3.2 O rganic Carbon Concentration .................................................................. 14
2.3.3 N itrogen ..................................................................................................... 16
2.3.4 D issolved Solids Content .......................................................................... 17
2.3.5 O xidation Reduction Conditions ............................................................ 18
2.3.7 G as Q uality ................................................................................................ 19
2.4 D iscussion ............................................................................................................ 20
2.4.1 Differences between Aerobic and Anaerobic Lysimeters ......................... 20
2.4.2 The Comparison of Leachate Parameters with Other Studies ................... 21
2.4.3 Im plications for Full-scale Application ..................................................... 22


v








2.4.4 Lim itations ................................................................................................. 23
2.5 Conclusions .......................................................................................................... 24

3. THE FATE OF HEAVY METALS IN SIMULATED LANDFILL
BIOREACTORS, UNDER AEROBIC AND ANAEROBIC CONDITIONS ............ 47

3.1 Introduction .......................................................................................................... 47
3.2 M aterials and M ethods ........................................................................................ 48
3.2.1 H eavy M etal Sources in Synthetic W aste ................................................. 48
3.2.2 Sam pling M ethods ..................................................................................... 49
3.2.3 A nalytical M ethods ................................................................................... 49
3.3 Results and D iscussions ....................................................................................... 50
3.3.1 Changes in Metal Concentrations versus Time and the Percentage of
M ass Loss ......................................................................................................... 50
3.3.1.1 A lum inum ........................................................................................ 50
3.3.1.2 A rsenic ............................................................................................ 51
3.3.1.3 Chrom ium ........................................................................................ 53
3.3.1.4 Copper ............................................................................................. 54
3.3.1.5 Lead ................................................................................................. 56
3.3.1.6 Iron .................................................................................................. 57
3.3.1.7 M anganese and Zinc ........................................................................ 58
3.3.2 Organic Wastes as Absorbents of Heavy Metals ...................................... 59
3.4 D iscussion ............................................................................................................ 61
3.4.1 O verall Com parison of M etal Behavior .................................................... 61
3.4.2 Com parison to O ther Studies ..................................................................... 63
3.4.3 Im plication for D isposal of H eavy M etals ................................................ 65
3.4.4 The Im pact of A ir on M etal M obility ........................................................ 66
3.5 Conclusions .......................................................................................................... 67

4. THE EVALUATION OF LIGNOCELLULOSIC WASTE DECOMPOSITION OF
AEROBIC AND ANAEROBIC SIMULATED LANDFILLS .................................. 95

4.1 Introduction ........................................................................................................... 95
4.2 M aterials and M ethods ........................................................................................ 97
4.2.1 Com position of Fabricated W aste ............................................................. 97
4.2.2 Excavation and Processing of Decomposed Solid Waste ......................... 97
4.2.3 M ethane Y ield D eterm ination ................................................................... 98
4.2.4 Cellulose and Lignin D eterm ination ....................................................... 100
4.2.5 D ata A nalysis ........................................................................................... 101
4.3 Results ................................................................................................................ 101
4.3.1 M ethane Y ield of Raw W aste .................................................................. 101
4.3.2 Solid W aste Excavation ........................................................................... 102
4.3.2 M ass Loss for Individual Com ponents .................................................... 103
4.3.3 Biodegradability of Excavated W astes .................................................... 104
4.3.4 B iodegradability of W ood W aste ............................................................ 105
4.4 D iscussion .......................................................................................................... 106
4.5 Conclusions ........................................................................................................ 108


vi








5. LANDFILL SETTLEMENT BEHAVIOR WITH WASTE DECOMPOSITION ..... 119

5.1 Introduction ........................................................................................................ 119
5.2 M aterials and M ethods ...................................................................................... 120
5.2.1 Lysimeters ............................................................................................... 120
5.2.2 Application of Overburden Pressure ....................................................... 121
5.2.3 Compression Index and Phase Separate M ethod ..................................... 122
5.2.4 Estimation of M ass Loss ......................................................................... 123
5.2.5 Volume Loss versus M ass Loss .............................................................. 124
5.3 Results ................................................................................................................ 125
5.3.1 Settlement Behavior over Time ............................................................... 125
5.3.2 The Relationship between The Settlement and Mass Loss ..................... 126
5.3.3 Ultimate Settlement ................................................................................. 127
5.4 Discussion .......................................................................................................... 127
5.4.1 Compression Index .................................................................................. 127
5.4.2 Correlation of M ass Loss and Volume Loss ........................................... 128
5.4.3 Application .............................................................................................. 129
5.5 Conclusions ........................................................................................................ 131

6. SUM M ARY AND CONCLUSIONS ......................................................................... 141

6.1 Summ ary ............................................................................................................. 141
6.2 The Implication of This Research ...................................................................... 143
6.3 Conclusions ........................................................................................................ 145
6.4 Future W ork ....................................................................................................... 147

APPENDIX

A. ADDITIONAL PROCEDURES AND CONCEPTS ................................................. 149

A. I Prediction of M ass Loss by Gas and Leachate ................................................. 149
A.2 Estimation of Biodegradable Volatile Solids (BVS) ........................................ 152
A.3 Lysimeter Dismantlement ................................................................................. 153

B. SUPPLEM ENTAL FIGURES ................................................................................... 159

C. LYSIMETER EXPERIMENT RAW DATA AND GRAPHS ............................... 169

C. I Graphs ............................................................................................................... 169
C.2 Raw Data ........................................................................................................... 189

LIST OF REFERENCES ................................................................................................. 219

BIOGRAPHICAL SKETCH ........................................................................................... 231






vii















LIST OF TABLES

Table page

2-1. MSW components .......................................................................... 25

2-2. Parameters and methods for analysis ..................................................... 26

2-3. Comparison of initial and final characteristics of the aerobic lysimeters............. 27

2-4. Comparison of initial and final characteristics of the anaerobic lysimeters.......... 28

2-5. Comparison of leachate parameters with other aerobic landfill studies .............. 29

2-6. Comparison of leachate parameters with other anaerobic landfill studies ........... 29

3-1. Heavy metal sources in fabricated waste stream ......................................... 69

3-2. Results of statistical analysis of metal leached between aerobic and anaerobic ....69 3-3. The amount of leachate produced and used for analysis................................ 69

3-4. Leachability of As, Cr, and Cu ............................................................ 70

3-5. Comparison of cumulative mass of metal dissolved in leachate and adsorbed on
lignocellulosic materials.................................................................. 70
3-6. Comparison of average metal concentrations of the aerobic and anaerobic
lysimeters with MSW leachate and regulatory levels ................................. 71

3-7. Comparison of characteristics of CCA-treated wood used for Jambeck (2004) and
this study................................................................................. 71

4- 1. Methane yields, VS and mass fraction of the lignocellulosic materials in raw
waste .................................................................................... 109

4-2. Comparison of methane yields of MSW with other studies.......................... 109

4-3. Biodegradable volatile solid (BVS) of organic fraction of the raw waste........... 109

4-4. The physical characteristics of excavated waste ....................................... 110

4-5. Overall methane yields of waste layers of the lysimeters 2 and 4 .................111II


viii








4-6. Summary of cellulose and lignin content of the wood samples ............................... 112

5-1. (Ca)min and (Ca)max values of lys I through 4 ........................................................... 132

5-2. k values of aerobic and anaerobic lysimeters ........................................................ 132

5-3. Comparison of compress indices between current study and other studies ............. 133

A- 1. Actual mass loss and predicted values of the aerobic and anaerobic lysimeter ...... 155 A-2. Mass and density of wastes excavated by depth ................................................. 155

C-1. pH of the aerobic and anaerobic lysimeters ...................................................... 189

C-2. Conductivity of the aerobic and anaerobic lysimeters ......................................... 192

C-3. Alkalinity of aerobic and anaerobic lysimeters ...................................................... 194

C-4. Total dissolved solids (TDS) of aerobic and anaerobic lysimeters .......................... 196

C-5. Total organic contents (TOC) of aerobic and anaerobic lysimeter ........................ 197

C-6. Chemical oxygen demand (COD) of aerobic and anaerobic lysimeter ................... 199

C-7. NH3 -N concentrations of aerobic and anaerobic lysimeters ............... 201

C-8. Sulfide concentrations of aerobic and anaerobic lysimeters ................ 203

C-9. Volatile fatty acids (VFA) of lysimeter 1 ............................................................205

C- 10. Volatile fatty acids (VFA) of lysimeter 2 ..........................................................206

C-11. Volatile fatty acids (VFA) of lysimeter 3 .................................................. 207

C- 12. Volatile fatty acids (VFA) of lysimeter 4 ...................................................... 209

C- 13. Fabricated waste in lysim eters .......................................................................211

C- 14. M etal concentrations of lysim eter 1 ....................................................................212

C- 15. Metal concentrations of lysimeter 2 .................................................................. 213

C- 16. Metal concentrations of lysimeter 3 .................................................................. 214

C- 17. Metal concentrations of lysimeter 4 ................................................................ 216

C- 18. ANOVA results of metals and organic absorbence ......................................... 218




ix















LIST OF FIGURES

Figure pae

2- 1. Schematic of the lysimeter ................................................................ 30

2-2. The composition of fabricated municipal solid waste for this research............... 31

2-3. Comparison of pH between aerobic and anaerobic lysimeters versus time .......... 32

2-4. Changes in COD of aerobic and anaerobic lysimeters versus time ................... 33

2-5. Changes in BOD of aerobic and anaerobic lysimeters versus time ................... 34

2-6. Changes in VFAs of aerobic and anaerobic lysimeters versus time (A) acetic acid
only and (B) acetic acid, propionic acid and butyric acid ............................ 36

2-7 Changes in the ratio BOD/COD of the aerobic and anaerobic lysimeters over time ..37 2-8. Changes in ammonia concentrations versus time........................................ 38

2-9. Changes in TDS of the aerobic and anaerobic lysimeters versus time................ 39

2-11. Changes in sulfide and pH versus time ................................................. 41

2-12. The changes in sulfate and sulfide versus time in the presence of dissolved
oxygen ................................................................................... 42

2-13. The changes in air injection rate and gas concentrations of aerobic lysimeter ....43 2-14. Changes in gas concentrations of anaerobic lysimeter 4 .............................. 44

2-15. Cumulative biogas vs. days in aerobic and anaerobic lysimeters.................... 45

2-16. Changes in gas concentrations, pH and gas generation rate after air injection into
lysimeter 3 ................................................................................ 46

3-I. Changes of Al concentrations over time.................................................. 72

3-2. Changes of As concentrations over time ................................................. 73

3-3. Changes of Cr concentrations over time.................................................. 74



x








3-4. Changes of Cu concentrations over time ................................................. 75

3-5. Changes of Pb concentrations over time ................................................. 76

3-6. Changes of Fe concentrations over time.................................................. 77

3-7. Changes of Mn concentrations over time................................................. 78

3-8. Changes of Zn concentrations over time ................................................. 79

3-9. Distribution of As over a C-pH diagram ................................................. 80

3- 10. Potential- pH diagram of Cr ............................................................. 81

3-1 1. Distribution of Cu over a C-pH diagram................................................ 82

3-12. Adsorption of metal on solid wastes..................................................... 83

3-13. The comparison of aerobic and anaerobic lysimeters in respect of total mass of
metals adsorbed on lignocellulosic materials........................................... 85

3-14. The comparison of metal concentrations adsorbed on organic (newspaper and
cardboard) and plastic waste............................................................. 87

3-15. Fate of heavy metals thermodynamically occurred in aerobic (oxidizing) and
anaerobic (reducing) conditions ......................................................... 89

3-16. Comparison of concentrations of metal leached between aerobic and anaerobic
lysimeters .................................................................................. 90

3-17. Changes in cumulative mass of meta released over a mass loss, % ................. 92

3-18. Comparison of As, Cu and Cr leaching trend of the lysimeters to other study ....94 4- 1. The dry weight differences between predicted and measured remaining mass...113 4-2. Comparison of dry weights between raw and decomposed lignocellulosic wastes. .114 4-3. The changes in the percentage of waste components after decomposition; (A) raw
waste components and (B) decomposed waste (aerobic) ........................... 115

4-4. Changes in cumulative methane volume of lignocellulosic materials over time..116 4-5. Methane yields and weight differences of lignocellulosic materials among raw
and two lysimeters (A) all lignocellulosic materials; (B) wood only.............. 117

4-6. The comparison of dry masses measured and predicted by gas generated and
BMP assay................................................................................ 118



xi








5- 1. The changes in settlement, cumulative gas (C02) and pH over time................ 134

5-2. The changes in settlement, cumulative gas (CO2 and CH4) and pH over time......135 5-3. Settlement behaviors and compression coefficients of aerobic and anaerobic
lysimeter over a period of time......................................................... 136

5-4. Relationship between settlement and overall mass loss of the aerobic and
anaerobic lysimeters..................................................................... 137

5-5. Relationship between percentage of settlement and mass loss....................... 138

5-6. Correlation of logarithm of mass loss of the aerobic lysimeters over time ......... 139

5-7. Different k values of anaerobic lysimeters at lag and log phases .................... 139

5-8. Settlement prediction of the aerobic lysimeters ........................................ 140

A-i1. Schematic of mass loss by waste decomposition...................................... 156

A-2. Waste mass loss by TOC and gas generation .......................................... 158

B- 1. Schematics of aerobic and anaerobic lysimeters used for this research ............ 159

B-2. The carriage system ...................................................................... 160

B-3. A schematic of the temperature control system........................................ 161

B-4. Schematic of gas volume measuring tool; before gas measurement, fill tap-water
up to the top scale........................................................................ 162

B-5. The nation-wide composition of discarded municipal solid waste in 2003......... 163

B-6. The composition of municipal solid waste in Florida in 2000....................... 163

B-7. (A) 'Blue water phenomenon' observed from gas collection system of aerobic
lysimeters; (B) a hole on copper tube caused by corrosion of Cu.................. 164

B-8. Solid samples excavated from one of the aerobic lysimeter ......................... 165

B-9. Decomposed papers were commingled together (aerobic lysimeter) ............... 166

B-l10. Not well degraded office paper (aerobic lysimeter).................................. 167

B-i 1 Wood blocks excavated from aerobic lysimeter ...................................... 167

C- 1. The change in COD of the lysimeters over the percentage of mass loss ........... 169





xii







C-2. The change in BOD5 of the aerobic and anaerobic lysimeters over the percentage

of m ass loss ............................................................................................................ 170

C-3. The change in ammonia of the aerobic and anaerobic lysimeters over time ........... 171

C-4. The change in fluoride of the aerobic and anaerobic lysimeters over time ............. 172

C-5. The change in chloride (CI) of the aerobic and anaerobic lysimeters over time .... 173 C-6. The change in sulfate of the aerobic and anaerobic lysimeters over time ............... 174

C-7. The change in calcium (Ca) of the aerobic and anaerobic lysimeters over time ..... 175 C-8. The change in sodium (Na) of the aerobic and anaerobic lysimeters over time ...... 176 C-9. The change in biogas produced from the aerobic lysimeters ................................... 177

C-10. The change in biogas produced from the anaerobic lysimeters ............................. 178

C-1 1. Al concentration versus pH in leachate from the lysimeters ................................. 179

C- 12. Cr concentration versus pH in leachate from the lysimeters ................................. 180

C-13. Cu concentration versus pH in leachate from the lysimeters ................................. 181

C-14. Mn concentration versus pH in leachate from the lysimeters ................................ 182

C-15. Pb concentration versus pH in leachate from the lysimeters ................................. 183

C- 16. Zn concentration versus pH in leachate from the lysimeters ................................. 184

C-17. Change in methane yields of the waste layer 2-1 and 2-2 ..................................... 185

C-18. Change in methane yields of the waste layer 2-3 and 2-4 ..................................... 186

C- 19. Change in methane yields of the waste layer 4-1 and 4-2 ..................................... 187

C-20. Change in methane yields of the waste layer 4-3 and 4-4 ..................................... 188













xiii














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

COMPARATIVE STUDIES OF AEROBIC AND ANAEROBIC LANDFILLS
USING SIMULATED LANDFILL LYSIMETERS By

Hwidong Kim

December 2005

Chair: Timothy G. Townsend
Major Department: Department of Environmental Engineering Sciences

Many proposals suggest that air injection into bioreactor landfills enhance waste composition; several potential benefits of air addition have been hypothesized, yet little has been proven about the overall performance of aerobic landfills compared with current anaerobic landfills. Utilizing research conducted with six-foot tall stainless steel simulated landfill lysimeters, complete with fabricated wastes, this Ph.D. dissertation compares aerobic and anaerobic landfills with respect to gas and leachate quality, fate of metals, settlement behavior and biodegradation of lignocellulosic materials.

Through air injection, a large enhancement of waste decomposition was observed. More than 90% of the maximum chemical oxygen demand (COD), biochemical oxygen demand (BOD) and total organic carbon (TOC) concentrations decreased within 100 days. During the methanogenic phase in the anaerobic condition, concentrations of ammonia increased by an amount four times greater than the initial concentrations. A large change of ammonia was not observed from the aerobic lysimeters.



xiv








The fate of metals leached from the various metal sources including cathode ray tube (CRT) monitor glass and ground CCA-treated wood were explored. Metal leaching trends observed varied from anaerobic to aerobic lysimeters; the average concentrations of As, Fe, Mn, and Zn in the anaerobic lysimeters proved significantly greater in concentration than observed in the anaerobic lysimeters. Furthermore, significantly greater concentrations of Al, Cu, Cr, and Pb were detected in the aerobic lysimeters as compared to the anaerobic lysimeters.

Using leachate and gas measurements, mass losses from the aerobic and anaerobic lysimeters were estimated. Mass removed from the wastes was primarily converted into gas; after the water was removed from the lysimeters, the mass of waste excavated from each lysimeter was compared with the estimated loss mass. For wood waste, no great influence on air addition was observed through cellulose/lignin analysis. Methane potential of lignocellulosic materials other than wood waste resulted in great differences of biodegradation between aerobic and anaerobic lysimeters.

The landfill settlement behavior occurring in aerobic and anaerobic simulated

landfills was mathematically analyzed. The logarithm of mass loss was linearly correlated with the percentage of settlement. With this relationship, the secondary settlement of bioreactor landfills could be mathematically modeled using the first-order exponential function.












xv













CHAPTER 1
INTRODUCTION

1.1 Problem Statement

Landfills remain the predominant method for managing municipal solid waste

(MSW) in the U.S. Although modem engineered landfills protect the environment from groundwater contamination and in some cases gas emissions, they are most often operated in a fashion where only a small amount of the disposed waste is permitted to biodegrade to a more stabilized state. This results in large amounts of undegraded waste being stored for many years in the future; their management will continue to demand resources and may pose a long-term environmental risk.

Alternatively, many innovative and more environmental-friendly strategies for operation of MSW landfills have been proposed (Stegmann, 1983; Barlaz et al., 1992; Komilis et al., 1999). Among these techniques, leachate recirculation has been found to be the most practical approach for enhancing waste decomposition and stabilization in landfills (Reinhart et al., 2002). This process stabilizes landfilled waste more rapidly because of the increased moisture content and the more effective distribution of nutrients and microorganisms in the landfill. This result creates a very favorable environment for the existing anaerobic organisms responsible for waste degradation. If controlled, methane produced can be utilized as a resource. This technique has changed the concept of a landfill from a historical garbage dump to a bioreactor, where various biochemical reactions are managed in a controlled fashion.





1





2


Air addition has been suggested as another means, in concert with leachate recirculation, to achieve rapid landfill stabilization. It has been reported that waste decomposes more rapidly in aerobic systems relative to anaerobic systems (Read et al., 2001). Additional reports suggest that air injection may stop the production of methane (one of the most serious greenhouse gases), change the leachate quality for the better, reduce the amount of volatile organic compounds (VOCs), and improve the degradability of anaerobically recalcitrant materials (Grima et al., 2000; Read et al., 2001; Lee et al., 2002; Reinhart et al., 2002). Some of these potential benefits have been investigated at the lab scale (Stessel and Murphy, 1992), and some positive outcomes have been reported from field studies (Read et al., 200 1; Lee et al, 2002). However, in order to apply this new technique successfully to full-scale operating landfills, further investigation is necessary. While anaerobic bioreactors have been heavily simulated in previous studies, there are few cases involving the simulation of aerobic landfills. It is also rare to find side-by-side simulations on the same waste stream under the same field conditions comparing aerobic and anaerobic systems.

1.2 Objectives

The main objective of this research was to compare aerobic and anaerobic landfills using simulated landfill lysimeters. In the early development of anaerobic bioreactors, several fundamental simulated landfill experiments were performed that have provided much of our understanding of such processes to date (Pohland, 1980). This research presents the results of parallel aerobic and anaerobic simulated bioreactors. Several different parameters of concern were investigated: leachate and gas quality, settlement, heavy metal fate, and decomposition of lignocellulosic materials. The following were specific objectives of this research:





3

o To compare leachate and gas quality between aerobic and anaerobic bioreactor

landfills,

o To explore the fate of heavy metals leached from the fabricated wastes in

aerobic and anaerobic bioreactor landfills,

o To explore the decomposition of lignocellulosic wastes in anaerobic and

aerobic bioreactor landfills, and

o To evaluate the loss of mass versus the loss of volume in aerobic and anaerobic

bioreactors for use in future settlement model development.

1.3 Research Approach

Four stainless steel lysimeters were constructed: two were operated aerobically and two were operated anaerobically. These lysimeters were designed and constructed as part of a previous research experiment (Sheridan, 2003). After operating the aerobic and anaerobic lysimeters for 1 and 2 years, respectively, one aerobic and one anaerobic lysimeter were dismantled. Waste samples were collected and characterized. The remaining aerobic and anaerobic lysimeters were kept in operation so that waste stabilization could be completely researched; the results of this extended operation will be presented elsewhere.

To compare leachate and gas quality between the aerobic and anaerobic bioreactors, two pairs of simulated landfill lysimeters containing fabricated wastes were operated as aerobic and anaerobic bioreactors. The fabricated wastes were loaded into the lysimeters, compacted, and mixed with water and seed (either anaerobic sludge or aerobic compost). Leachate generated by the lysimeters was collected and analyzed for leachate quality parameters. A mixture of collected leachate and deionized water was added back to the





4


lysimeters to compensate for the amount of leachate lost by leachate collection. The gas volume and composition were monitored using a gas totalizer and gas chromatography.

To explore the fate of heavy metals leached out of the fabricated wastes under aerobic and anaerobic conditions, heavy metal-containing wastes (e.g., CCA-treated wood, cathode-ray tube (CRT) glass and pieces of sheet metal) were mixed with the other fabricated wastes before loading into the lysimeters. After loading and compacting the fabricated waste, leachate generated by the lysimeter was collected and analyzed for copper, chromium, arsenic, lead, aluminum, zinc, manganese and iron. The change of heavy metal concentrations in the leachate over time was monitored. After the lysimeter work was completed, the wastes excavated from two columns were analyzed for heavy metals in order to compare heavy metal concentrations absorbed on solid waste to those released from the lysimeters through the leachate.

To explore the decomposition of lignocellulosic wastes in aerobic and anaerobic landfill environments, lignocellulosic wastes including paper and wood blocks were prepared. They were included in the fabricated waste and loaded in the lysimeters. After the lysimeter study was completed, lignocellulosic wastes were excavated and separated. Biochemical methane potential (BMP) assays were used to evaluate the degree of biodegradation of each lignocellulosic waste. In order to evaluate the impact of air addition on wood waste decomposition, cellulose, lignin and BMP of raw and excavated wood blocks were compared with respect to cellulose and lignin concentrations and BMP values.

To simulate landfill settlement in aerobic and anaerobic conditions as a function of waste mass loss, overburden pressure was applied to the stainless steel lysimeters using a





5


hydraulic cylinder and hand pump. To correlate mass loss and volume loss, a lab-scale experiment was designed where waste was decomposed in simulated landfills in the laboratory with both mass loss and volume loss being measured. A difficulty with using lab experiments to simulate landfill settlement is that it is hard to simulate true landfill conditions, especially, the large overburden pressure. In this research, the experiments included the application of overburden pressure to make the laboratory condition closer to the field conditions.

1.4 Outline of Dissertation

The dissertation is presented in six chapters. The current chapter presents the

problem statement, objectives and research approach. Chapter 2 presents the comparison of gas and leachate qualities between aerobic and aerobic simulated landfills. Chapter 3 presents the fate of heavy metals in aerobic and anaerobic simulated landfills. The evaluation of biodegradation of lignocellulosic materials is presented in chapter 4. Settlement behavior with waste decomposition is presented in chapter 5. Chapter 6 presents a summary, conclusions and recommendations for failure work. Background and other analytical procedures used for this research are presented in appendix A. Supplemental s are presented in appendix B. All other tables and s pertaining to leachate data and BMP are presented in appendix C.














CHAPTER 2
COMPARATIVE STUDIES OF LEACHATE AND GAS QUALITY OF AEROBIC
AND ANAEROBIC SIMULATED LANDFILL BIOREACTORS

2.1 Introduction

The operation of municipal solid waste (MSW) landfills as bioreactors; for the purpose of rapid landfill stabilization has historically been proposed as an anaerobic process. Conditions within the landfill are controlled to accelerate the activity of the anaerobic microorganisms responsible for waste decomposition. The addition of air has also been proposed as a method to enhance landfill stabilization (Stessel and Murphy, 1992), and recently this technique has gained more attention (Read et al., 200 1; Reinhart et al., 2002). In addition to an enhancement of waste decomposition that is more rapid than anaerobic operation, a major benefit often cited for air addition is the reduction in methane emissions relative to anaerobic landfills (Borglin et al., 2004). These studies also find that the overall strength of leachate (with respect to readily degradable carbon compounds and oxygen demand) is lower in aerobic systems, offering a potential advantage with respect to leachate treatment.

Research examining the relative differences in leachate quality between aerobic and anaerobic systems is very limited. Though the performance of aerobic bioreactor landfills has been simulated in several studies (Agdag and Sponza, 2004; Warith and Takata, 2004), these studies are often limited with respect to their ability to control several key parameters, and their lack of a complementary anaerobic system for comparison purposes. This chapter reports the results of research performed to examine the characteristics of



6





7


leachate and landfill gas that result from aerobic and anaerobic operation of identical MSW streams. The experiments conducted involved a technique long employed in the study of landfills: waste-filled columns constructed and operated to simulate landfill processes, referred to here as lysimeters (Pohland, 1980). The columns were designed and operated to control several parameters not traditionally simulated in such experiments, such as temperature and overburden pressure. The objective was to compare leachate and landfill gas quality between each type of system so that similarities and differences can be better understood and to assist in future decision-making, design and operation efforts. Several complementary objectives were evaluated as part of this experiment and they are described in greater detail in Chapters 3 (fate of metals), 4 (comparison of decomposition) and 5 (comparison of settlement).

2.2 Material and Methods

Four lysimeters were used in this research, and each consisted of a stainless steel column and a carriage system component. The original design and construction of the lysimeters used for this research were described previously by Sheridan (2003). Two were operated aerobically (lysimeter I and 2) and two were operated anaerobically (lysimeter 3 and 4). Three parameters, temperature, air addition, and overburden pressure, were controlled in an effort to simulate actual aerobic or anaerobic bioreactor landfills.

2.2.1 General Description of the Lysimeter

A schematic of each lysimeter type is presented in Figure 2-1 (see Figure 13-1 for additional detail). The 6-ft stainless steel main body contained 5 front ports, 2 back ports and I valve at the bottom for leachate collection. The front ports were used for air addition (in the case of the aerobic lysimeters). The carriage system component was designed to support a hydraulic pressurizing unit installed at the top of each lysimeters





8


for the application of an external load to the fabricated waste. The carriage system consisted of a hydraulic cylinder, carriage, steel shaft, and steel plate. A small port located on the top flange was used as a pathway for liquid addition. Perforations in the steel plate allowed added liquid to percolate into the waste (see Figure B-2 for a detail of the carriage system).

2.2.2 Temperature Control

The temperature at the center of a full-scale landfill usually remains constant

because the garbage and cover soil serve to insulate the system (McBean et al., 1995). In a laboratory environment, however, the heat produced by biologically degrading waste is not sufficient to maintain a temperature close to those normally encountered in a landfill. Thus, a temperature control system was designed and constructed (Figure B-3).

The temperature of each lysimeter was measured using a type T thermocouple wire (SRT2O1-160, Omega) fixed on the outside of each lysimeter. Two temperature controllers (MC240, Electrothermal) were utilized in series to maintain desired temperatures without extreme fluctuations. The lysimeters were insulated with 5-cmthick fiberglass and bubble insulation to minimize heat loss. Prior to operation, the lysimeters were filled with tap water and the temperature controllers were tested by measuring the temperature of the water.

The temperature of the aerobic lysimeters was maintained at a constant 55soC for the entire operating period. The anaerobic lysimeters were started at 350C and at day 400, the temperature was increased from 35'C to 55'C at a rate of 2'C per day. Although 550C is in the optimum range for thermophilic anaerobic waste decomposition (Rittmann and McCarty, 2001) and is often encountered in landfills (Watsoncraik et al., 1994;





9


Townsend et al., 1996), 35C was used as the starting point because the anaerobic seed used was from a mesophilic digester.

2.2.3 Fabricated Waste Stream

The waste stream fabricated for this research was based on typical MSW

composition estimates previously reported for the U. S. and Florida (see Figure B-4 and B-5). For simplification purposes, several minor components, such as textiles and tires, were excluded from the fabricated waste stream. A greater portion of commingled paper was allotted as a substitute for those excluded materials. The relative amount of office paper, cardboard and newsprint in commingled paper (4.6 : 2.6 : 1) was again estimated from previous published data (FDEP, 2003 and USEPA, 2005). Figure 2-2 presents the fabricated waste stream composition used. Table 2-1 presents a description of each component, the source, and the method of sample preparation. Commercial grade dog food (Pedigree, USA) was used as the food waste portion of the fabricated waste stream. To support complementary research on the fate of certain heavy metals in aerobic and anaerobic landfill environment (chapter 3), a part of the wood waste fraction was comprised of CCA and a part of the glass fraction was comprised of leaded cathode ray tube (CRT) glass. Detailed waste components and their sources are presented in Table 21 and Figure 2-2.

Mixed fabricated waste samples were created and loaded into the columns as four distinct fractions to prevent waste component stratification in a particular place in the column, (composition of the fabricated waste fractions and their weight are summarized in appendix Q. Prior to loading, 6 inches (15.3 cm) of river rock was placed at the bottom of each lysimeter, and a geotextile was placed between the rock and waste. Each waste fraction was then loaded and compacted until it occupied 25 % of the depth of the





10


lysimeter. Two liters of DI water were added along with the compaction of each waste fraction. After loading, 11 L of additional water was added from the top of each lysimeter. The goal of adding water was to bring the waste in each lysimeter, at the beginning of the experiment, to field capacity. A capacity of 58% was targeted as this was the field capacity measured for this waste under the initial compaction conditions of the lysimeter. The waste was compacted to a density of 30 lb/ft3 dy (480.6 kg/m3 dry).

2.2.4 Air Injection

Two computer-controlled pump drives (Model No. 7550-10, Cole-Parmer) were

used for air addition. Air was saturated and warmed prior to injection to keep moisture in the waste from evaporating. Air was injected on the ports located at the side of the aerobic lysimeters using a manifold from day 1 to day 164 and changed to the most bottom port from day 164 to the end of a test period. A flow rate of 70 mL/min was found to be suitable for control purposes and to maintain low exit gas oxygen concentrations. The flow rate was adjusted several times during the experiment when oxygen concentrations in the exit gas became less than 1% to maintain aerobic conditions.

2.2.5 Leachate and Gas Analysis

Leachate samples were collected on a weekly basis. Leachate was analyzed for sulfide and dissolved oxygen immediately; analysis for pH, alkalinity, and conductivity was carried out within one hour after collection. After this initial analysis, 15 mL of leachate was preserved with sulfuric acid and placed in acid-rinsed high-density polyethylene (HDPE) bottles for later analysis of chemical oxygen demand (COD), total organic carbon (TOC), volatile fatty acids (VFA) and ammonia. For metal analysis, 50 mL of leachate was preserved with concentrated nitric acid and stored at 4'C. The remaining leachate was recirculated back to the top of the lysimeters. Deionized water





11


was added to make up for the amount of leachate used for analysis. Table 2-2 summarizes the parameters and methods used for each analysis.

Biogas samples generated from both the aerobic and anaerobic lysimeters were collected and analyzed for methane, carbon dioxide, and oxygen. For the aerobic columns, the gas volume was measured using a wet-tip gas meter. For the anaerobic lysimeters, the gas was gathered in 5-L and 10-L air-sampling bags, and the volume contained in the bags was measured using the water-gas replacement method (see Figure B-4). A LANTEC GEM 500 (SAIC, San Diego, CA) gas meter was used for gas analysis for both the aerobic and anaerobic lysimeters. Additionally, gas samples collected from the anaerobic lysimeters were analyzed for CH4 and CO2 using a gas chromatograph equipped with a GS-Carbon plot column (Agilent Technology, Palo Alto, CA) to confirm the measurements analyzed by LANTEC GEM500 gas meter.

2.2.6 Recovery of the Anaerobic Lysimeters

Since both of the anaerobic lysimeters (lys 3 and 4) remained in an acidic condition (pH < 6) for 500 days, 100 g of sodium bicarbonate was added as a buffer to the top of each lysimeter on day 300. The pH of the top part of the lysimeters changed to neutral, but the pH of leachate collected from the bottom port remained low (5 to 5.5). The pH of the leachate from lysimeter 4 increased to pH 7 from day 400. Since only minimum changes in leachate pH of the lysimeter 3 were observed after buffer addition, additional sodium bicarbonate was added to the bottom port rather than to the top of the lysimeters; a total of 1 OOg of sodium bicarbonate was added (20 g each were added on days 420, 453, 469, 532, and 555 again). Only a temporary increase in pH was observed after this addition. As a next step in increasing pH, lab air was injected into lysimeter 3 on day 627. Before air injection, the methane concentration of the lysimeter 3 was 35%, and the pH of





12


leachate was 6.11. Lab air was injected with 70 mL/min from the bottom of the lysimeter for five days. The changes in pH and output gas qualities were monitored on a daily basis. The impact of this addition is discussed in the results section of this chapter.

2.2.7 Prediction of Waste Mass Loss

As described in the following sections in this chapter, the aerobic lysimeters more quickly stabilized the waste in comparison to the anaerobic lysimeters, and thus their period of operation was shorter (379 days vs. 741 days). In an effort to normalize the leachate measurements among the different columns, the biogas data, the leachate data and the initial content of the waste was used to estimate the percentage of waste decomposition for a column at any given time. The detailed procedure for this is presented in appendix A, but, in short, the cumulative volume of biogas measured at any given time (CH4 + C02 for anaerobic columns and CO2 for aerobic columns) was used to calculate the mass of initial waste degraded at that time. This was adjusted to account for the mass of organic carbon solubilized in the leachate. The mass of waste estimated to be degraded at a given time was divided by the estimated total potential mass loss in each column (this total potential mass loss was estimated from measured methane yields of the raw waste; see chapter 4 for details).

2.3 Results and Discussion

The data presented for the aerobic and anaerobic lysimeters in this dissertation represent operation periods of 379 and 741 days, respectively. At the end of each operation period, one each of the aerobic and anaerobic lysimeters was stopped and emptied. The remaining lysimeters were left operational (data are not reported here). Values of all leachate parameters analyzed for this research are presented in Table C- I





13


through C-i 1 in appendix C. These include the raw data and graphs of the leachate parameters.

2.3.1 pH

Figure 2-3 depicts the change in pH over the course of the experiment. Both the

aerobic and anaerobic lysimeters remained in acidic condition during the beginning of the experiment. The period of time required to stabilize the pH for the aerobic and anaerobic lysimeters was 200 and 600 days, respectively. Average pH measurements of approximately 8.9 (aerobic) and 7.1 (anaerobic) were observed at the end of the experiment.

Two phases (acidic and alkaline or methane phase) of the pH of the aerobic and

anaerobic lysimeters were observed during a test period. The low pH occurring during the initial phase of the research was attributed to a build up of organic acid concentrations and the related microbial activities. Once the organic waste decomposition process began, the biodegradable fraction of waste was converted into organic acids by various biological reactions, and the accumulation of the organic acids lowered the pH. For the aerobic lysimeters, air was injected through four front ports of the lysimeter using manifolds. The pH was low (< 6) for the first 150 days, and high VFA and alkalinity concentrations indicated that anaerobic conditions were predominant, suggesting that air was not evenly distributed through the manifolds. An increase in the pH of the aerobic lysimeters was observed after air was injected into the only bottom port. Typically, the pH of the system increases to neutral conditions as the organic acids are consumed by methanogenic bacteria. A large amount Of CO2 production in an unbalanced ecosystem may also contribute to lowering the pH as well. High concentrations of VFA and alkalinity were measured in the anaerobic lysimeter leachate during the initial acid phase.





14


The pH of the aerobic lysimeters measured in the latter half of the experiment (9.0) was more alkaline than the pH measured from the anaerobic lysimeters (7.2). According to other lysimeter studies, higher pH was observed from the aerobic lysimeters in comparison with that of the anaerobic lysimeter. The range of pH of aerobic lysimeters has been reported as 7 9 (Stessel and Murphy, 1992; O'Keefe and Chynoweth, 2000; Agdag and Sponza, 2004). Summerfelt et al. (2003) also observed an increase of pH when air was injected into their aquaculture system. They reported that this increase was because of CO2 stripping by air; a decrease in CO2 leads to a decrease of carbonic acid (H2CO3) and bicarbonate concentrations (HCO3) consuming H ions. These relationships can be described by carbonate systems as follows: CO2 gas <- H2CO3 (1)

H2CO3 --* HCO3 + H+ (2)

HCO3" E--) CO3 + H+ (3)

They additionally concluded that, because the dehydration of carbon acid is rate-limiting, pH may not increase instantaneously.

2.3.2 Organic Carbon Concentration

Figure 2-4 depicts the change of COD concentrations for the lysimeters versus

time. The initial average COD concentrations in the leachate of the aerobic and anaerobic lysimeters were 36,000 mg/L and 66,000 mg/L, respectively. The COD values for the aerobic lysimeters increased up to greater than 84,000 mg/L and decreased rapidly after pH was stabilized. Although one of the aerobic lysimeters showed high COD (70,000 mg/L) at day 50, the overall COD concentrations of the aerobic lysimeters were lower than values in the anaerobic lysimeters. Similar trends occurred for COD as were observed for BOD5 (Figure 2-5). The BOD values of the aerobic lysimeters decreased





15


rapidly down to below 100 mg/L from day 200 while BOD values of the anaerobic lysimeters decreased relatively slowly.

The primary contributor to high COD or BOD concentrations in landfill leachate is volatile fatty acids (McBean et al., 1995). Figure 2-6 (a) depicts the changes in acetic acid, one of the major volatile fatty acids (VFA), as a function of mass loss. Acetic acid is used as a substrate by methanogenic bacteria and contributes to the formation of an acidic environment when they are unbalanced with the growth of methanogenic bacteria. For these reasons, VFA concentrations are used as an indicator to assess landfill conditions (USEPA, 2004). For example, the decrease in acetic acid concentration in lysimeter 3 and

4 corresponds to the point when the pH began to rise. In the aerobic lysimeters, high concentrations of acetic acid were noted during the first phase of the experiment due to improper air distribution as discussed earlier. Acetic acid in leachate from the aerobic lysimeters was degraded to less 1 mg/L by day 200, which corresponds with the time required to deplete COD and BOD.

Among different types of short carbon chain fatty acids, acetic, propionic and

butyric acids are known as major VFAs that are involved in biodegradation processes in anaerobic conditions. Production and degradation of these major VFAs in selected aerobic and anaerobic lysimeters are presented in Figure 2-6 (b). In both aerobic and anaerobic lysimeters, the concentration of VFAs was mainly: acetic acids > butyric acids > propionic acids. These results are similar to those found by Parawira et al (2004). They also explained that high butyric acids were mainly attributed to high carbohydrates in waste. Under the same condition, the degradation of VFAs in anaerobic condition was found to be in the following order: butyric acids > acetic acids > propionic acids. Wang et





16


al. (1999) explained that various enzymatic reactions in microorganisms dictate a greater decreasing rate of butyric acids than that of other VFAs. However, more biosynthetic processes are involved in butyric acid production than acetic acid due to longer carbon chains. In aerobic lysimeters, all three major VFAs were depleted together like other bulk organic carbon.

The ratio of BOD5 to COD is often used to assess the biodegradability of the organic matter in leachate, and thus to assess the degree of landfill stabilization. In old stabilized landfills, the BOD5/COD ratio is below 0.10 (Kjeldsen et al, 2002). A low BOD5/COD suggests that a leachate is low in biodegradable organic carbon and relatively high in hard-to-biodegrade organic compounds such as humic compounds. In this research, low BOD5/COD ratios were observed with the aerobic lysimeters after day 200 (0.04 on average) (Figure 2-7). Relatively high BOD5/COD ratios were exhibited from the anaerobic lysimeters (0.36 on average). These values fall into the range of average BOD5/COD ratios proposed by Kjeldsen et al. (2002) for the acid phase (0.58) and the methanogenic phase (0.06).

2.3.3 Nitrogen

Figure 2-8 shows the changes in ammonia-nitrogen in the lysimeter during the course of experiment. Ammonia concentrations from the aerobic lysimeters remained relatively constant, showing a general increase during the course of the experiment. In a different fashion, ammonia concentrations in the anaerobic leachate increased dramatically at a point corresponding to an increase in system pH. Ammonia concentrations in the anaerobic lysimeters increased to concentration in the range of 1000-1600 mg/L. These values then dropped to 800-1000 mg/L and stabilized. Small increases of ammonia concentrations in leachate of aerobic column were observed after





17


day 180, but they were still approximately 4 times lower than values of anaerobic lysimeters. The trend of increases in ammonia concentrations also can be found in operating bioreactor landfills (Reinhart and AlYousfi, 1996).

Since ammonia is generally produced from the deamination process of amino

acids (a monomer of proteins), elevated ammonia concentrations may be associated with protein decomposition. Cali et al. (2005) reported that an active methanogenic bacteria community increased the ammonia concentration. Several researchers have proposed that the enhancement of waste decomposition and leachate recirculation in anaerobic bioreactor landfills results in increased ammonia concentration (Reinhart and Al-Yousfi, 1996; Berge et al., 2005).

2.3.4 Dissolved Solids Content

Figure 2-9 shows the change in total dissolved solids (TDS) in the aerobic and anaerobic lysimeters through the course of the experiment. For both aerobic and anaerobic lysimeters, like the change in other organic matter, TDS concentrations were lower as the pH was stabilized. TDS of the aerobic lysimeters were rapidly stabilized approximately 8 to 10 g/L after day 200. TDS of the anaerobic lysimeters were still greater than that of the aerobic lysimeters but TDS of lysimeter 4 fell below 20 g/L on day 700. Typical TDS concentration in landfills is within the range of 2 to 60 g/L (Kjeldsen, 2002).

Figure 2-10 depicts the change in alkalinity in the aerobic and anaerobic

lysimeters through the course of the experiment. For the aerobic lysimeters, the alkalinity increased to 16,000 mg/L as CaCO3 in the lysimeter 1, but another lysimeter showed low alkalinity which was below 2,000 mg/L as CaCO3 but it increased 8,000 mg/L as CaCO3 again. The alkalinity was lowered below 2000 mg/L as CaCO3 for both aerobic





18


lysimeters after the pH was stabilized at alkaline condition. For the anaerobic lysimeters, relatively high alkalinity was maintained over the entire test period. Generally, alkalinity could be generated by CO2 accumulation and ammonification (Fannin, 1987). It is also consumed by nitrification (Gujer and Jenkins, 1974).

2.3.5 Oxidation Reduction Conditions

Figure 2-11 depicts the change of sulfide concentrations in the aerobic and

anaerobic lysimeters over a period of time. Little changes of sulfide were observed during the acid phase of both aerobic and anaerobic lysimeters. However, rapid increases in sulfides along with an increase of pH were exhibited from the aerobic lysimeters and lysimeter 4. The sulfide level of lysimeter 1 was lowered on day 2 10 again, but increased up to 2,600 itg/l, during the alkaline phase. The trends of change in sulfide concentration of lysimeter 2 exhibited are similar to that of lysimeter 1. The highest sulfide level found in lysimeter 2 was 1,200 p~g/L. In lysimeter 4, sulfide concentrations increased as the pH increased.

It is notable that great concentrations of sulfide were found in the system where air injection had been taking place. It is hard to understand how sulfide could be presented in an aerobic environment. In comparison with sulfate concentrations, sulfide was formed by sulfate reduction (Figure 2-12). However, Figure 2- 10 shows that high dissolved oxygen concentration was also found in the same condition. Snoeyink and Jenkins (1980) reported that sulfide could be detected under aerobic conditions. They explained that this phenomenon was caused by a non-equilibrium situation for the reaction between oxygen and sulfide. Therefore, it can be concluded that sulfide can be found before it oxidizes for a second time by dissolved oxygen. This result indicates that anaerobic zones were





19


presented in the aerobic lysimeters producing sulfide. Relatively high ammonia concentrations found from the aerobic lysimeters (Figure 2-9) also indicated the presence of anaerobic zones in the aerobic lysimeters.

2.3.7 Gas Quality

Biogas emitted from the aerobic and anaerobic lysimeters was measured for 02, CO2 and CH4. Figures 2-13 and 2-14 depict the changes in gas concentrations of aerobic and anaerobic lysimeters over a period of time. The initial air injection rate of the aerobic lysimeters was 70 mL/min. The air injection rate was regulated by changes of oxygen levels within the range of 70 to 120 mL/min. High CO2 concentrations were observed from aerobic lysimeters during the first 50 days, but decreased to lower than 20%. The concentrations of CH4 and CO2 of the anaerobic lysimeters changed during the acidic phase, but stabilized to approximately 60% CH4 and 40% CO2 during the methane phase.

Overall, a total of 40,100 liters (1,400 ft3) of gas was injected into each aerobic lysimeters for a test period, and 45% and 43% of the oxygen included in the air added was converted into CO2 in lysimeter 1 and 2, respectively. In the anaerobic lysimeters, 500 and 1,600 liters of biogas (CO2 and CH4) were produced from lysimeters 3 and 4. Most of the gas generated was mainly concentrated on the methanogenic phase in anaerobic lysimeters while a relatively steady gas generation was exhibited over time in aerobic lysimeters as summarized in Figure 2-15.

Lab air was added to recover the lysimeter 3, which had remained in acidic

condition (pH <6) for 600 days. Figure 2-16 depicts the change in gas concentrations, gas generation rate and pH during air injection. The pH was adjusted to 7.1 at day 4, and air injection was stopped on day 5. After oxygen was depleted in the lysimeter, methane concentrations substantially increased along with biogas generation rate and reached





20


above 50% at the 9th day. The biogas generation rate of lysimeter 3, after air injection, was 2.3 L/day on average for 10 days. During the rest of test period, the pH of the lysimeter 3 went down to 6.5, but further decrease was not observed. In comparison with the conditions of lysimeter 3 before air addition, the amount of biogas produced was substantially increased and high percentage of methane (> 55%) was maintained (Figure 2-16).

2.4 Discussion

2.4.1 Differences between Aerobic and Anaerobic Lysimeters

The largest differences between the aerobic and anaerobic lysimeters can be found from the enhancement of waste biodegradation. Based on the leachate quality results of this study, a period required for the aerobic lysimeters to decompose 90% of BOD was 160 days in the aerobic lysimeters while more than 700 days were required for the lysimeter 4.

Other differences between the two systems were the methane concentrations

contained in exit gas. Air addition to the aerobic lysimeters lowered CH4 concentrations in the exit gas dramatically. Though a small amount of methane was found in the exit gas of the aerobic lysimeter, it was less than 1% of the CO2 gas generated.

It is noted that pH had a relatively low impact on waste decomposition in the aerobic lysimeter. Though the pH of the aerobic lysimeters was acidic, settlement consistently occurred (see Chapter 5). It was probable that the acidic condition was localized only on the bottom part, where air was not supplied properly.

Tables 2-3 and 2-4 present the initial and final characteristics of the aerobic and anaerobic lysimeters. The data presented in Table 2-4 for the anaerobic lysimeters indicate that these systems were not stabilized yet. Water loss from the aerobic lysimeters





21


was calculated using the water carrying capacity of the exit gas assuming that the gas was 100% saturated with water vapor. The overall performance of the aerobic lysimeters with respect to waste decomposition and leachate quality was substantially greater than those of the anaerobic lysimeters. However, the concentrations of sodium in the aerobic lysimeters were still too high to meet drinking water standards. The final concentration of ammonia in the aerobic lysimeters was also substantially higher than the criteria value of ambient water (0.897 at 30'C and pH 8.0) (USEPA, 1999). Although large quantities of waste were decomposed, leachate of the anaerobic lysimeters still contained high concentrations of organics, ammonia and anions (Table 2-4). Leachate generated would be used for recirculation, but excessive volume of leachate must be treated at an on-site or off-site wastewater treatment plant.

2.4.2 The Comparison of Leachate Parameters with Other Studies

Tables 2-5 and 2-6 summarize the comparison of leachate constituent

concentrations from this study with those from other studies. For aerobic landfill studies, the maximum concentrations of COD, BOD and ammonia in this study appeared to be greater than those of other studies, but they were in a similar range overall. The pH of the aerobic lysimeters of this study was, however, greater than other studies. As previously discussed, this would be because of the relocation of carbonic acids, bicarbonate, and carbonate due to CO2 removal by air stripping (Summerfelt, 2003). The high pH of the aerobic lysimeters implies that great concentrations of carbonate ions were dissolved due to high partial pressure Of C02, and these carbonate ions might consume more H' ions when CO2 was removed. If alkalinity data of other studies were available, it would be





22


clear to describe the differences between this and other studies by comparing the concentrations of carbonate ions.

2.4.3 Implications for Full-scale Application

Unlike the lab-scale simulated landfill, it is extremely difficult to aerate an entire large-scale landfill. Highly compacted wastes make it difficult for an air stream to penetrate into the recesses of a landfill. Moreover, leachate characteristics resulted from air addition may be variable. As the analytical results have shown, leachate characteristics of lysimeters I and 2 were different, despite starting with the same waste stream and the same operational condition. The leachate characteristics of lysimeter 1 were similar to those of the anaerobic lysimeters during the first 180 days showing great concentration of organic matter despite air addition. This was because of the large anaerobic zones formed at the bottom of the lysimeter by improper air addition to the bottom.

Aerobic zones can be formed around air injection wells but anaerobic zones may still be present in the same landfill. However, coexistence of the aerobic and anaerobic zones can be used for recovery of acid-stuck 'sour' landfills. In this research, the air addition was conducted under the hypothesis that environments formed by aeration for a short period can be favorable to anaerobic microorganisms. A great amount of VFAs, which caused acidic conditions, may be rapidly consumed by aerobes living in a relatively wide pH range. Conversion of carbonic acid (H2CO3I) to C02 caused by air stripping may increase the pH. With air addition with low flow rate, the anaerobic zones may be protected from oxygen intrusion because oxygen may be depleted by the respiration of aerobes. An additional technical strategy would be to add buffer such as lime along with air addition. Buffer added may increase the alkalinity concentration.





23


Without high alkalinity, the pH of the landfill may decrease again when air addition is stopped. This could happen when methanogenic bacterial population was not enough to adapt to the new condition.

As Reinhart et al. (2002) pointed out, the reduction of leachate volume due to air stripping could be one of the advantages of the aerobic landfills. In this research, a total of 31m m3 of air was added during a test period (1 year). Comparing the volume of water initially added with final leachate volume, approximately 2 1% of leachate volume was reduced. Reduced volume of leachate implies that the operation of aerobic landfills can be economical in terms of saving the cost for the leachate treatment.

2.4.4. Limitations

Since CI-L gas is one of the gases causing global warming, CH4 reduction can be one of the advantages of the aerobic lysimeters. However, landfill gas released without a' flare system could be adverse to the environment. Berge et al. (2005) and Reinhart et al. (2002) pointed out that various kinds of volatile organic compounds (VOC) and nitrous oxide, a more potent greenhouse gas than methane, can be emitted without the flare system. Future research is required to identify the gas constituents and develop the filtering system as an alternative.

As previously mentioned, an extra monitoring job may be required to check

moisture content and gas contents around the gas injection well. Certain ratios of methane and oxygen can be flammable according to Coward and Jones (1952) and Liao et al. (2005). While air was added, CH4 concentrations were low, but the unpredictable changes in 02 and CH4 concentrations were observed from the aerobic landfill (Read et al, 2001) and high concentrations of CH4 and 02 could coexist when air addition starts (Lee et al, 2002).





24


2.5 Conclusions

In this research, the gas and leachate quality from aerobic and anaerobic

simulated bioreactor landfills were compared. Waste streams referenced from EPA and FDEP were loaded into 4 stainless steel lysimeters with a density of 3500 kPa. All lysimeters were prepared with the same conditions, and two of them were assigned for aerobic and two for anaerobic bioreactor landfill simulations. Leachate and gas generated from the lysimeters were analyzed for chosen parameters to make comparisons between aerobic and anaerobic landfills.

Leachate analysis results indicated that organic compounds as measured by COD, TOC, BOD and VFAs in the aerobic lysimeters were degraded more rapidly than those in anaerobic lysimeters. Except for the acidic phase, the pH of the aerobic lysimeters rapidly increased and was stabilized around pH 9.0 while anaerobic lysimeters had remained in acidic phase for more than 400 days, and stabilized exhibiting pH 7.3. The concentrations of ammonia in anaerobic lysimeters increased along with an increase of pH. Ammonia concentrations in aerobic lysimeters varied little over time, but ammonia levels were significantly lower than those of anaerobic lysimeters. Sulfide results imply that both aerobic and anaerobic zones were coexisting in aerobic lysimeters. This may be caused by the limit of oxygen distribution in the lysimeters because of high density and low hydraulic conductivity of wastes under overburden pressure.






25


Table 2- 1. MSW components
Waste components Sources Processing for size reduction
Office paper Mixed scrap paper purchased at Grind with a paper shredder
office supply store
Cardboard Mixed corrugated boxes Scissors and razor blade
Newspaper Local newspaper Grind with a paper shredder
Plastics PET bottles collected from a Scissors
plastics recycler
Foodwase Comerial og oodGrind with a coffee grinder (less
Foodwase Comerial og oodthan 1/32") Southern yellow pine (SYP) Home improvement store Cut with band-saw (2" x 2")
CCA-treated wood Home improvement store Gather saw dust after drilling

Galvanized steel Home improvement store Cut with metal cutter (1/2" x
__________________________ __________________________1/2")
Aluminum Home improvement store Cut with metal cutter (1/2" x
1/2")
Cathode-Ray Tube(CRT) glass CRT monitors Crush with a hammer (1/4-1/8")
Mixed cullet Mixed container glass Crush with a hammer (1 /4 1/8")






26


Table 2-2. Parameters and methods for analy sis.
Parameters Method
Alkalinity Standard Method 2320B
Ammonia Standard Method 4500-D
BOD Standard Method 5210B
COD HACH 2720
Conductivity Standard Method 2510
pH Standard Method 4500-H'
TOC EPA SW846, Method 9060
Sulfide HACH 8131
Sulfate, Floride and Chloride EPA SW846, Method 9056
Sodium EPA SW846, Method 9060A
VFA VFA analysis method using GC
(Innocente et al., 2000)






27


Table 2-3. Comparison of initial and final characteristics of the aerobic lysimeters Initial Final (1 year)
Lys 1 Lys 2 Lys 1 Lys 2
Water quantity (mL) 19,000 19,000 15,137* 15,582 (15,056*
Dry waste quantity (g) 12,784 12,784 8,389* 8,740 (8,715*)
pH 5.7 5.7 8.5 8.5
COD (mg/L) 20,000 28,000 3,400 4,700
BOD 13,000 16,000 200 30
TOC 6,000 7,000 2,600 2,200
Ammonia 70 40 500 250
Fluoride 80 30 0 0
Chloride 200 130 1,200 1,700
Sodium 80 140 800 900
* predicted






28


Table 2-4. Comparison of initial and final characteristics of the anaerobic lysimeters Initial Final (2 years)
Lys 3 Lys 4 Lys 3 Lys 4
Water quantity (mL) 19,000 19,000 18,844* 18,833 (18,704*)
Dry waste quantity (g) 12,784 12,784 11,290 9,258 (8,997*)
pH 4.5 4.9 6.5 7.4
COD (mg/L) 65,000 67,000 42,000 24,000
BOD 48,000 62,000 14,000 6,500
TOC 26,000 27,000 12,000 5,600
Ammonia 120 100 1,000 800
Fluoride 1,500 1,400 460 200
Chloride 1,450 1,400 670 500
Sodium 2,000 2,000 4,800 3,800
* predicted






29


Table 2-5. Comparison of leachate parameters with other aerobic landfill studies Parameters
(mg/L Compost Lysimeter Lysimeter Lysimeter Lysimeter This tud
except for study study I b study 2c study 3d study 4e This study
pH)
Air flow 20L/min 38L/min for
20L/mmn
rate 30min. at 8.4 70(once per a 12-h 1300L/min 20mL/min 120mL/min
week) intervals
COD 2434-31812 861-22026 130-23000 500-5000 2-1000 3400-47000
BOD5 8-11571 100-10000 10-2000 30-45000
Ammonia 98-558 260-630 7-400 2-100 40-700
TDS 3300-11400 700-7700
pH 7.1-8.2 5.17-7.98 5.24-7.5 7-9 7.8 4.5-9.1
aKrogmann and Woyczechowski, 2000; bAgada and Sponza, 2004; CWarith and Takata, 2004; dStessel and Murphy, 1992; eBorglin et al., 2004





Table 2-6. Comparison of leachate parameters with other anaerobic landfill studies Parameters Bioreactor
(mg/L Conventional Conventional Bioreactor
landfill lab This study
except for landfill 1a landfill 2b landfills scaled
pH) scale
pH)
COD 140-152000 1000-40000 20-17000 100-88000 2000-80000
BOD5 20-57000 50-25000 0-10000 6600-60000
TOC 30-29000 7000-19000
Ammonia 50-2200 50-1500 76-1850 100-1600
TDS 2000-60000 2000-25000 18000-50000
pH 4.5-9 3-7.5 5.4-8.6 4-7.5 4.5-7.5
Ca 10-7200 300-4000 20-4000
a Kjeldsen et al., 2002; bPokhrel, 2004; CReinhart and AI-Yousfi, 1996; dPohland and Kim, 1999






30


Back Front




Leachate Carriage
I e[ injection system


To the 4-gas collection system






Main
body











Air injection Leachate collection Figure 2-1 Schematic of the lysimeter






31




SYP CRT glass Mixed cullet

CCA treated wood5% 1 % 6%
I %
Alumium Paper office paper
4% 27%
Galvanized steel
4%




15%
paper newsprint
6%



Food waste Paper cardboard
15% 16%




Figure 2-2. The composition of fabricated municipal solid waste for this research.






32









10
AEROBIC
9 8 7

6
6 O-0 lys 1
5 000 O 0 0 0-- lys 2

4 I I
0 100 200 300 400
8
ANAEROBIC
7


6


5 lys 3
-0- lys 4


0 100 200 300 400 500 600 700 800

Days Figure 2-3. Comparison of pH between aerobic and anaerobic lysimeters versus time






33





100000
AEROBIC
-0- lysimeter 1
-0-- lysimeter 2 80000




60000



0
S40000




200000
0 100 200 300 400

100000
ANAEROBIC +-- lysimeter 3
-0- lysimeter 4 80000




60000



0
) 40000




20000




0
0 200 400 600 800

Days Figure 2-4. Changes in COD of aerobic and anaerobic lysimeters versus time






34





80000
AEROBIC
----- lys 1
-0- lys2


60000





40000
O
0



20000
O0
600
o
.0

0 a
0 100 200 300 400

80000
ANAEROBIC
-0- lys 3 O-0 lys4





0
60000








20000

O OO


0.
Cbo


0 100 200 300 400 500 600 700 800

Days Figure 2-5. Changes in BOD of aerobic and anaerobic lysimeters versus time






35







30000
AEROBIC
lys 1

25000 -- lys 2



20000 15000 10000




5000



0'
0 100 200 300 400
30000

-0- lys 3 ANAEROBIC
-0- lys 4
25000



S20000



15000



S10000



5000




0 200 400 600 800

Days
(A)






36





30000
lys 1 (aerobic) Acetic acids
0-- Propionic acids 25000 -- V- Butyric acids


20000

-I
15000


S10000
O/
> //

5000

o- ...... 00o..... ...... oo ....... = : ==..
0 1 o o
0 50 100 150 200 250 300 350 400
30000
-0- Acetic acids lys 4 (anaerobic)
0... Propionic acids
25000 -vT- Butyric acids


20000
-o

15000


10000
so o o 0 o .


0



0 200 400 600 800

Days
(B)
Figure 2-6. Changes in VFAs of aerobic and anaerobic lysimeters versus time (A) acetic
acid only and (B) acetic acid, propionic acid and butyric acid





37





1.0
Lys 1
S 0 Lys 2
8 A A v Lys 3
0.8 A Lys 4



0.6 O
0.6 Oiw A m A
0 0
U O


OA
0 V
0.4 V
020 0 0

0
0.2



0.0 8
0 200 400 600 800

Days
Figure 2-7. Changes in the ratio BOD/COD of the aerobic and anaerobic lysimeters over
time






38









600

lys 1 AEROBIC
--0- lys 2
500



400



300



200



100



0
0 100 200 300 400

1800
-- lys 3 ANAEROBIC
1600 lys4

1400 1200

1000

o 800 < 600

400 200

0
0 200 400 600 800
Days Figure 2-8. Changes in ammonia concentrations versus time






39





60
AEROBIC
-0- lys 1 0-- lys2 50



40



30



20



10
0.

0 0 0
0..... d%
0
0 100 200 300 400

ANAEROBIC
60 lys 3
lys4 50


40


30
HA

20


10


0 I I I
0 100 200 300 400 500 600 700 800

Days Figure 2-9. Changes in TDS of the aerobic and anaerobic lysimeters versus time





40




20000
AEROBIC Lys
0 Lys 2

S15000
O
0



b 10000

0


5000


o . 0 0 0 o o o 0 .0 0 0
0 0
.0 -- -0 0 ~ 0~cxc00 OO 00

0 100 200 300 400
20000
ANAEROBIC


S15000 A



-~AA
-" 10000




< 5000

-A- Lys 3
--A Lys4

0
0 200 400 600 800
Days

Figure 2-10. Changes in alkalinity of the aerobic and anaerobic lysimeters versus time






41





4000
Lys 1 AEROBIC
0-O Lys2

3000




a 2000
-U



1000 0 0
1000 0 0 0.000

0O
0

0 0
0 100 200 300 400

4000

-- Lys 3 ANAEROBIC
OLys 4 0.- Lys4

3000


E o

S2000



9
1000ooo


~ ~..........6 --0
0 200 400 600 800

Days Figure 2-11. Changes in sulfide and pH versus time







42






100 4--- 000 10
Lysimeter 1


80 8
3000 x


T 6b
I 60 x 6 E
X2

-2000
EXXX 0001
40 xx / -4

20 I i 2
I I1 1000


I \
000



0. 60 -0
0 100 200 300 400

Days
180 1400 10

160 Lysimeter 2 1200
160 1200

140 -x 8
x 1000
120 x x
x 6





o :" _-'_ 80
100 800




2 4006-*
60 y6
\ -400 t
40 -2 / 200
20 1

0 -0 0
0 100 200 300 400

Days


Figure 2-12. The changes in sulfate and sulfide versus time in the presence of dissolved
oxygen






43





40 140
Lys 1 CH4
C02 120
i 02
0I Air injection
30 ..
"-' 7 I r2 --I100 '
,_. I I Iill "'1i I80

-'.~ ,,
L2 -h - -
60

[..... ... 40

10 .

\ .. '.. .20
k11 1:1I[;/ I I"2
I J /F I I I

0 I 0 I /
0 100 200 300 400

Days



Figure 2-13. The changes in air injection rate and gas concentrations of aerobic lysimeter






44





100
Lys 4
CH4 ........ C02
80 ---02


0


.4.



Ci2
20

20


I\
0

0 200 400 600 800

Days Figure 2-14. Changes in gas concentrations of anaerobic lysimeter 4






45






5000
lys 1 .................. ly s 2
lys 3 S 000 - lys 4

O


3000




0
2000




1000

.. .. . .

0 I I I
0*
0 200 400 600 800

Days Figure 2-15. Cumulative biogas vs. days in aerobic and anaerobic lysimeters






46






60

Air injection started Air injection ended
50



40 1

.6
.30 0
o 0

.0
20 '0.CH

0
10 1
I 0"v
090

3.0




*7.0
2.0

0
0 '0 1.5.o 6.5

00
1.0


6.0 0.5
0 5 10 15 20 25 30

Days
Figure 2-16. Changes in gas concentrations, pH and gas generation rate after air injection
into lysimeter 3














CHAPTER 3
THE FATE OF HEAVY METALS IN SIMULATED LANDFILL BIOREACTORS
UNDER AEROBIC AND ANAEROBIC CONDITIONS

3.1 Introduction

An issue of current debate in the solid waste community is the fate of heavy metals disposed in MSW landfills (SWANA, 2003). Heavy metals may be present as a result of industrial residuals, but more importantly for MSW landfills, they result from manufactured products. Examples include lead from electronic devices and copper, chromate and arsenic from treated wood. This debate has taken on more immediate concern as several US states have banned certain wastes containing heavy metals (e.g., leaded cathode ray tubes) from disposal in landfills (SWANA, 2003). These bans are in part a result of fears regarding the fate of the disposal of metals in landfills.

For the most part, heavy metals have been thought to be relatively well contained in typical anaerobic landfills. According to Kjeldsen et al. (2002), the amount of heavy metals dissolved and contained in leachate is very low relative to those present in the waste. Most metals are thought to be released during the initial stage of landfill decomposition as a result of the lower pH. Once a landfill enters the methanogenic phase, heavy metal concentrations in leachate dramatically decrease, and in many cases, their levels decrease to lower than drinking water standards (Kjeldsen et al., 2002). Bioreactor landfills are becoming a more common method of managing MSW, and the impact of these systems on metal leachability should be evaluated. Since bioreactor landfills involve exposing a much larger percentage of waste to moisture, the total mass



47





48


of metals released might be expected to be high relative to dry landfills. On the other hand, given that bioreactor landfills recirculate leachate to the landfill and that traditional bioreactors promote anaerobic waste decomposition (and thus enhance metal removal by the mechanisms described previously), the impact to the environment may be limited.

An alternative bioreactor landfill technique is to add air in addition to moisture.

Taking into consideration that the leaching behavior of heavy metals is mainly controlled by redox, pH and the presence of ligands (Benjamin, 2002), it may be that the fate of heavy metals in aerobic systems will differ from anaerobic systems. The long term fate of heavy metals in aerated landfills is a question yet to be satisfactorily addressed.

In this research, the fate of heavy metals in simulated aerobic and anaerobic

bioreactor landfills was studied. Four stainless-steel lysimeters were used, two each for aerobic and anaerobic conditions. Heavy metal-containing wastes were included in the waste stream added to each lysimeter. Leachate collected from each lysimeter was analyzed for heavy metals over time. In order to evaluate the heavy metals adsorbed in solid wastes, waste samples were removed from two of the lysimeters at the end of the experimental period and analyzed for heavy metal content.

3.2 Materials and Methods

A detailed description of the lysimeters was presented in chapter 2. The methods as described here focus on the metal-containing components in the fabricated waste and on metal analysis in the leachate and waste samples. 3.2.1 Heavy Metal Sources in Synthetic Waste

Several MSW components were chosen as sources of heavy metals. Each

component, its corresponding heavy metals and the percentage of each component are presented in Table 3 -1. Aluminum and galvanized steel sheets were purchased from a





49


local hardware store and cut into 1.5 cm x 1.5 cm square. Galvanized steel served as a source of both Fe and Zn. CCA-treated wood was used as a source of Cr, Cu and As. Crushed cathode ray tube (CRT) monitor glass was used as a Pb source. Total Cu, Cr, and As concentrations were 2350 50, 2890 56, and 1330 10 mg/kg, respectively. Crushed CRT monitor glass used in this research was a mixture of the funnel sections of 30 CRT color monitors. Jang and Townsend (2003) reported that 413 mg/L of Pb leached from CRT funnel glass using the toxicity characteristics leaching procedure (TCLP).

3.2.2 Sampling Methods

Leachate samples were collected weekly via a sampling port located at the bottom of each lysimeter. A portion of the leachate collected was used for analysis of general water quality parameters and the remainder was injected back into the lysimeters. A 50 mL aliquot was preserved with concentrated nitric acid and used for heavy metal analysis.

Lysimeter studies were conducted for 379 and 741 days for aerobic and anaerobic lysimeters, respectively. After the lysimeter studies were completed, solid wastes were removed from single aerobic and anaerobic lysimeter and analyzed for heavy metals. The samples were divided by depth into 4 fractions. Details about these fractions are summarized in Table A-2 in appendix A. Each fraction was then separated into 5 categories which include office paper, cardboard, newspaper, wood blocks and plastics. The separated samples were ground using an Urschell mill (Fritsch, German).

3.2.3 Analytical Methods

Leachate samples were digested with nitric and hydrochloric acids following EPA method 3050B and 30 1 OA for solid and liquid digestion, respectively (USEPA, 2003). Approximately 2 g of the ground samples were digested using nitric acid and 30%





50


hydrogen peroxide and then analyzed for heavy metals using ICP-AES following US EPA, SW-846 Method 60 1OB (USEPA, 2003). Digested samples were filtered using ashfree cellulose filters and analyzed for heavy metals and cations using Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP-AES) (Thermo Electronics, USA). Leachate samples preserved with concentrated nitric acid were analyzed for a total of 8 metals (As, Cu, Cr, Mn, Zn, Pb, Fe and Al).

3.3 Results and Discussions

3.3.1 Changes in Metal Concentrations versus Time and the Percentage of Mass
Loss

The following section presents the results (for each metal) of the aerobic and

anaerobic lysimeters as separated plots. The experimental time period for the aerobic and anaerobic lysimeters differed. Because of their time scale difference, the cumulatitive mass of metal leaching was plotted for all lysimeters as a function of waste mass loss. The estimation of mass loss is described in appendix A. The total amounts of leachate produced and used for the analysis are summarized in Table 3-3.

3.3.1.1 Aluminum

The changes in Al concentration in aerobic and anaerobic conditions over a period of time are depicted in Figure 3-1. High Al concentrations were observed from both aerobic and anaerobic lysimeters (18 and 20mg/L) for the first 10 to 20 days. Whereas the Al concentrations of anaerobic lysimeters dramatically decreased to below 0.5mg/L within 100 days, great changes in Al concentrations were not observed from the aerobic lysimeters. The changes in Al concentrations in aerobic lysimeters were mainly controlled by pH; high concentrations of Al were observed from both lysimeters 1 and 2 at pH < 6 and pH > 8, and lowest Al concentrations (< 1 mg/L) were observed at 6 < pH





51


< 8. Average Al concentrtaions (7.9 mg/L) of the aerobic lysimeters were significantly higher than those of the anaerobic lysimeters (0.28 mg/L).

Generally, Al leaching is not greatly affected by redox conditions but mainly

controlled by pH. Meima and Comans (1997) reported that Al solubility was low in the pH range 6 to 7, which corresponds to the Al results in the aerobic condition presented in Figure 3-1. Among many ligands forming Al compelxation, hydroxide ion (OH) is known as a major ion to control the solubility of Al in aquatic systems. The equilibrium of Al with gibbsite (AI(OH)3)), an Al-OH complex, is characterized by a U-shaped pHleaching curve (Eary, 1999). However, Al concentrations in anaerobic lysimeters did not appear to follow with the solubility of gibbsite. Besides pH and redox conditions, large differences in leachate characterisitcs between aerobic and anaerobic lysimeters included the high organic content and anions such as floride and sulfate in the leachate of the anaerobic lysimeters. The most likely explanation of low Al solubility in the anaerobic conditions is complexation of Al and organic matter. Tipping (2005) reported that Al solubility is strongly associated with organic content. Skyllberg (2001) also reported that high dissolved organic content made Al solubility significantly decrease. Therefore, low solubilitity of Al during the first phase of the anaerobic lysimeters is the result of complexation of Al with high concentrations of organic matter.

3.3.1.2 Arsenic

Figure 3-2 depicts the change in As over time. High As concentrations were

observed from the anaerobic lysimeter before day 220. The greatest As concentration was

3.2 mg/L on the day 89. The As concentration then lowered below 1.5 mg/L after day 400. For lysimeter 4, As concentrations were continuously low, showing the lowest value,

0.27 mg/L on the last sampling day (day 741). In contrast to the anaerobic lysimeters, As





52


concentrations in the aerobic lysimeters decreased initially and increased with increasing pH. The As leaching pattern of the aerobic lysimeters appears similar to Al leachate trends. The lowest As concentration of lysimeter 1 was 0.12 mg/L on the day 163. For lysimeter 2, extremely low As concentrations were observed, with several samples below the detection limit (0.011 mg/L) despite a pH < 6. Overall, As dissolved in the leachate of the aerobic lysimeters was significantly lower than that of the anaerobic lysimeters (p <

0.05).

Figure 3-9 depicts the distribution of As concentrations observed from the aerobic and anaerobic lysimeters at various pH conditions. It has been reported that As solubility changes with pH and is characterized by a U-shaped curve in oxidizing conditions (Drever, 1988). However, Carbonell-Barrachina et al (1999) reported that in the presence of sulfide, Fe, and Mn, the solubility of As was dramatically lower and did not follow a U-shaped solubility curve. However, As concentrations were not impacted by these constituents in the anaerobic lysimeters. The most likely explantion is that the anaerobic lysimeters had poor-anoxic conditions during the first phase. The low sulfide concentrations at a pH < 6 confirm that the redox potential was not low enough for sulfide to become involved in As precipitation. For the aerobic lysimeters, As concentrations were low under acidic conditions and increased up to I mg/L at a pH of 9.

Masscheleyn et al (1991) found that As solubility decreased substantially as the redox potential increased. The changes in As solubility are also associated with the oxidation state of iron; Fe (III) has a strong affinity for arsenate. Therefore, low arsenic concentrations are likely dictated by the low solubility of arsenate. Under oxidizing conditions, As solubility may increase or decrease by pH changes. This is because of the





53


effect of pH on the total oxyaionic arsenate concentrations by pH. This changes in solubility results in the relatively higher concentrations of As at alkaline conditions observed in the aerobic lysimeters (Figure 3-2).

3.3.1.3 Chromium

The initial Cr concentrations of the anaerobic lysimeters were higher than those of the aerobic lysimeters (Figure 3-3). However, Cr concentrations in the anaerobic lysimeters gradually decreased to below 0.05mg/L by the day 453. As the pH of lysimeter

4 changed to moderately alkali (pH > 7.4) after day 464, minor increases in Cr concentrations were observed. In contrast, clear Cr leaching trends were not exhibited by the aerobic lysimeters before day 100, but an increase in Cr concentration did occur following day 150. This increase in Cr concentration corresponds to an increase in pH. Overall, the average Cr concentrations of the aerobic lysimeters were significantly greater than those of the anaerobic lysimeters. This may be because thermodynamically Cr can be present as an ionic form at alkaline pH under oxidized condition.

The toxicity of Cr is determined by its oxidation state. Among the various Cr oxidation states, only trivalent and hexavalent forms are taken into consideration in natural aquatic systems. Hexavalent Cr is considered more toxic than trivalent Cr due to its high mobility and solubility. Cr (VI) may be reduced to Cr (III) at low ORP potential. Cr (VI) becomes unstable and is reduced to Cr (III) at low pH. In order to maintain the oxidation state of Cr as Cr (VI) at a low pH, it is necessary to keep highly oxidizing conditions (Richard and Bourg, 1991). In contrast to other metals such as As and Cu, Cr (III) is not likely to precipitate with sulfide. Chromium solubility is mainly controlled by Cr(OH)3(s). Generally, Cr(OH)3(s) is formed in a pH range of 6.5 to 7 under moderately oxidizing or reducing conditions.





54


According to the potential-pH diagram of Cr (Figure 3-10), total Cr obtained from both aerobic and anaerobic lysimeters in an acidic environments is likely to be Cr (III) as Cr(OH)2+. In contrast, dissolved Cr from the aerobic lysimeter at a pH 9 could be Cr (VI) as CrO42-. Since all Cr species presented on the potential-pH diagram are based upon assuming thermodynamic equilibrium, all Cr obtained from the aerobic lysimeters at high pH may not be Cr (VI).

It is noted that an increase in Cr was observed from lysimeter 4 around a neutral pH ("A" in Figure 3-3). The most likely explanation for this is the lower Fe concentrations of lysimeter 4 than of those of lysimeter 3 (Figure 3-6). In the presence of Fe, Cr may be precipitated with Fe rather than OH due to rapid kinetics. The complexation of Fe and Cr decreases Cr solubility lower than the complexation of OH and Cr (Eary and Rai, 1987). Therefore, an increase of Cr concentration at the end of lysimeter 4 would be the result of a decrease of Fe concentrations.

3.3.1.4 Copper

Overall copper concentrations of the aerobic lysimeters were one to three orders of magnitude higher than those of the anaerobic lysimeters (Figure 3-4). For the aerobic lysimeters, clear Cu leaching patterns over time were not observed, but relatively large changes in Cu concentrations were observed at lysimeter 1 from the day 140 to 190. This period of time corresponded to a pH change from 5.5 to 9. For anaerobic lysimeters, Cu concentrations gradually decreased for the first 450 days. The initial Cu concentrations of lysimeter 3 and 4 were 0.082 and 0.234 mg/L, respectively. Although the concentrations slightly increased after day 450, final concentrations of Cu remained lower than the initial values.





55


Copper solubility is controlled by several Cu-containing minerals forming

complexation with Fe and sulfide. In addition, Cu sulfides may coexist with the sulfides of other metals such as Zn, Pb and As (Faure, 1991). Representative mineral deposits formed with OH-, Fe and/or sulfide include chalcocite (Cu2S), chalcopyrite (CuFeS2), cuprite (Cu20) and malachite (Cu2(OH)2CO3). These minerals are widely distributed over a pe-pH diagram. Ionic Cu is present only at a pH less than neutral and under highly oxidizing conditions (pe > 2.5). For these reasons, high concentrations of Cu may not be found under landfill conditions where low ORP and neutral pH are predominant. These concepts can be applied to the Cu distribution patterns displayed at a low pH in the pHCu concentration chart shown in Figure 3-11.

However, there is a disparity in the Cu concentrations observed and those

thermodynamically predicted for the alkali conditions of the aerobic lysimeters; most of the Cu precipitated by complexation with various Cu-containing mineral deposits at alkali and oxidizing conditions. Edwards et al (2000) reported high Cu concentrations in drinking water at alkali conditions, calling it 'the blue color phenomenon' since water color changed to blue with high concentrations of Cu. Critchley et al (2004) explained this 'blue water' was caused by microorganism-intermediated-Cu leaching from a part of the water delivery system. In this research, 'blue water' was also observed from condensate passing through a copper tube which connected to a gas collection system of an aerobic lysimeter. Further development of Cu corrosion caused a small hole on the same copper tube and called for a replacement of the copper tube with plastic materials (Figure B-7).





56


Another possibility of copper leaching from the aerobic lysimeters is the binding of Cu with ammonium (NH3+). Since both Cu and ammonium are cations, their complexations are present as an ionic form and can be dissolved in aquatic systems. Arzutug et al (2004) reported that Cu leaching increased with ammonia concentrations. However, complexation of Cu and ammonia may occur in a relatively narrow range of ORP and pH (Hoar and Rothwell, 1970). When plotting Cu and ammonia data obtained from the aerobic lysimeters, no clear evidence to prove the relationship between Cu concentrations and ammonia was found (r2 = 0.021). Furthermore, since Cu complexes with sulfide rather than ammonia in the presence of sulfide (Alymore, 2001), it may be difficult to leach high concentrations of Cu under landfill conditions.

3.3.1.5 Lead

Figure 3-5 depicts the changes in lead concentrations over time. For aerobic

lysimeters, Pb concentrations dramatically increased to 1.7 2 mg/L within 30 days and then gradually decreased. After the pH of the aerobic lysimeters stabilized, Pb concentrations decreased to levels similar to the anaerobic lysimeters. In contrast, little change in Pb concentrations was observed from the anaerobic lysimeters and low concentrations of Pb were maintained over the test period. Generally, Pb concentrations in landfill leachate have been reported to be very low (Charlatchka and Cambier, 2000; Jang and Townsend, 2003). This is because lead may precipitate with various ligands such as carbonate ions (C03-), sulfide, and volatile fatty acids (VFA).

Lead solubility is generally controlled by carbonate, or other Pb hydroxides and phosphate in noncalcareous soils (Bradle, 2005). Charlatchka and Cambier (2000) concluded that Pb solubility increased under oxidizing conditions at a pH of 6.2 However, Pb may precipitate as a form of PbS under reducing conditions in the presence





57


of sulfur. Lead is generally present in an ionic form at a pH < 6 under oxidizing conditions (Drever, 1988).

As shown in Figure 3-5, Pb leached from the aerobic lysimeters significantly

greater than from the anaerobic lysimeters. Most samples with high concentrations were distributed in the acidic phases. This leaching pattern corresponds to the characteristics of Pb previously discussed. For both aerobic and anaerobic lysimeters, Pb concentrations decreased with an increase in pH. Most Pb in alkali conditions may be precipitated as forms of PbCO3 or PbS depending upon redox potentials.

3.3.1.6 Iron

As shown on Figure 3-6, initial Fe concentrations of the aerobic lysimeters

(110mg/L for both aerobic lysimeters) were higher than those of the anaerobic lysimeters (20-22mg/L). Iron concentrations of the aerobic lysimeters increased to 250mg/L on the 30th day and then gradually decreased. During changes in pH of the aerobic lysimeters to alkali conditions, Fe concentrations lowered substantially to below 10mg/L. In contrast to the aerobic lysimeters, Fe concentrations of the anaerobic lysimeters increased from 20 to 600mg/L for the first 450 days and then decreased with increasing pH. Although lysimeters 3 and 4 are both anaerobic lysimeters, the final Fe concentrations were substantially different (1 65mg/L and 5.6 mg/L for lysimeters 3 and 4, respectively).

Generally, free Fe concentration is strongly associated with the redox condition of the system. Iron is present in aquatic systems in two oxidation states; Fe (III) and Fe (II). Ferric (Fe3+) and ferrous (Fe 2) irons can be transformed to each other depending upon the redox conditions. Iron (III) is precipitated as a mineral deposit such as Fe203 or Fe(OH)3 at a pH > 5. Iron (III) is also involved in complexation with metals. Under moderately oxidizing and reducing condition, Fe (II) ions are dominant in the pH range





58


of 5 to 9. In the presence of sulfur, Fe (1I) is likely to be precipitated as pyrite (FeS2) at pH > 5 under reducing conditions (Drever, 1988). Despite the many other reactions Fe is involved with, Fe solubility is strongly controlled by sulfide concentrations. Thus, in order to lower Fe concentration, the redox potential of the system needs to be low enough to reduce sulfate to sulfide. Sulfide concentrations of lysimeter 3 were lower than those of the other lysimeters, resulting in greater concentrations of Fe observed from lysimeter

3.

3.3.1.7 Manganese and Zinc

Leaching patterns of Mn and Zn look very similar for the aerobic and anaerobic lysimeters (Figures 3-7 and 3-8) and are thus discussed together. Relatively high concentrations of Mn and Zn were exhibited from the aerobic lysimeters for the first 150 days. The highest concentrations of Mn and Zn were 11 mg/L and 270 mg/L, respectively. Manganese and zinc concentrations then substantially decreased to below

0.2 mg/L and 10 mg/L, respectively. In contrast, little change in Mn and Zn concentrations was observed from the anaerobic lysimeters for 450 days, but decreased following that period. This corresponds to when the pH of the anaerobic lysimeters increased.

Since Mn and Zn are mainly precipitated by sulfide, differences of Mn and Zn concentrations between lysimeters 3 and 4 are strongly associated with the sulfide concentrations present in each lysimeter. Additionally, the solubility of Zn and Mn is associated with organic matter. A decrease in Zn and Mn solubility accompanies an increase of pH and could be accounted for by the generation of pH-dependent charge sites on organic matter (McBride and Blasiak, 1979; Miyazawa et al, 1993).





59


3.3.2 Organic Wastes as Absorbents of Heavy Metals

Analytical results of the 8 metals absorbed on office paper (OP), cardboard (CB),

newspaper (NP), and wood blocks (WD) are shown on Figure 3-12. Concentrations of Al, As, and Cu adsorbed on the lignocellulosic materials of the aerobic lysimeter appeared greater than those of the anaerobic lysimeter.

Heavy metals adsorbed on solid wastes from the aerobic and anaerobic lysimeter were statistically analyzed using the ANOVA test. Test results are presented in appendix C. All metals, except for Pb, absorbed on CB in the aerobic lysimeter and were significantly higher than those of the anaerobic lysimeter. Other significant differences between the aerobic and anaerobic lysimeters were found with Al, As, Mn and Cu adsorbed on NP, OP and WD.

Figure 3-13 depicts the differences of total mass of metals adsorbed on

lignocellulosic materials between the aerobic and anaerobic lysimeters. These values were calculated by multiplying the metal concentrations by the mass of each waste obtained from the garbage separation. Interestingly, the observed trends of adsorption of some metals did not to correspond with their leaching trends. These trends can be found from adsorption trends of As, Mn and Pb. The amounts of metals leached and adsorbed between aerobic and anaerobic lysimeters are compared in Table 3-5. These results indicate that metal adsorption may be influenced by environmental conditions such as pH, redox, and the presence of other ligands. Ravat et al (2000) reported that the binding of selected metals (Zn, Cu, and Pb) on lignocellulosic materials is strongly pH-dependent in the absence of interference from other ligands. Adsorption of Fe on organic matter is mainly controlled by the oxidation states of Fe; Fe (III) has greater affinity for organic matter than Fe (II) does (Jansen et al., 2003), corresponding to the large differences of Fe





60


adsorbed between the aerobic and anaerobic lysimeters (Table 3-6). Adsorption of As also mainly occurs when As is oxidized. As the pH rises, ionic forms of As changes progressively (H2AsO4, HAsO42 and AsO4 3) with each species showing different adsorption properties (Drever, 1988).

It is noted that there were large differences between metals released through

leachate and metals that remained in the lysimeters by adsorption. These differences can be numerically expressed using the ratios between metal leached (LC) and adsorbed

(AD). The smallest LC/AD ratio can be found from Al; only 0.06% of the amount of Al adsorbed was released from the lysimeters. The LC/AD ratios of most metals fell into around or below 2%. Relatively high LC/AD ratios were exhibited from a few metals in the anaerobic lysimeter; LC/AD ratios of As, Mn and Zn were 13.8%, 16.5% and 8.6%, respectively. However, if the amounts of metals precipitated as particulate forms without adsorption are taken into consideration, the ratio of metals between leached and remained would be much smaller than LC/AD ratios.

Lignocellulosic materials such as paper and wood products occupy as much as 45% of MSW landfills (USEPA, 2003). Cellulose and lignin are reported as the major heavy metal adsorbents (Basso et al, 2004). Lignin especially provides many chemical functional groups such as carboxyl and phenolic groups. Babel and Kurniawan (2003) concluded that lignin was considered as the best low-cost adsorbent for Pb and Zn. Basso et al. (2002) also reported that maximum sorption capacity increases due to lignin contents during Cd sorption research. Cellulose has also been heavily demonstrated to remove heavy metals such as Cd, Cu, Ni, Zn and Pb (Sublet, 2003; Okieimen et al., 2005 and Shukla and Pai, 2005).





61


Figure 3-14 depicts metal concentrations adsorbed on selected lignocellulosic

materials (newspaper and cardboard) and plastic waste. Greater concentrations of most metals except for Pb and Zn adsorbed on lignocellulosic materials were observed in the aerobic lysimeter. Bradle (2005) explained that many metals tend to adsorb the organic matter as the pH increases under oxidizing condition. Zhang and Itoh (2003) reported that carbonized mixture of polyethylene terephthalate (PTE) and waste ash could be used as a metal sorbent. However in this research, metal concentrations adsorbed on plastic waste were substantially low in comparison with metal concentrations adsorbed on the lignocellulosic materials.

3.4 Discussion

Among the various factors affecting heavy metal leaching under landfill conditions, the redox and pH may play the most critical role. Under the given redox condition and pH, the metal oxidation state, ligands, adsorption behavior can be determined. In many cases, metal precipitation can be controlled by Fe (II) and sulfide concentrations in both aerobic and anaerobic condition. Cr, Cu, Pb, Zn and As are reported to adsorb on hydrous ferric oxide at pH > 6, and the precipitation of Cu, Fe, Pb, Mn and Zn is controlled by sulfide in anaerobic condition. In addition to those ligands, hydroxide ion (OH) also can play an important role to precipitate Al and Cr (Drever, 1988). The various chemical interactions are depicted in Figure 3-15.

3.4.1 Overall Comparison of Metal Behavior

Figure 3-16 describes the overall trends of metal leaching in aerobic and anaerobic lysimeters. Among 8 metals under consideration, (Al, As, Cr, Cu, Pb, Mn, Fe and Zn) greater concentrations of Al, Cr, Cu and Pb were observed in the leachate of the aerobic lysimeters, and As, Mn, Fe and Zn were observed in the leachate of the anaerobic





62


lysimeters. For the anaerobic lysimeters, the metal leaching occurred in the acid phase, while occurrence of the metal leaching was relatively well distributed over the pH (5 < pH < 9) for the aerobic lysimeters except for Pb.

Cumulative masses of metals were calculated by the multiplication of the amount of leachate used for analysis by the concentration of the metals in the leachate sample (Figure 3-17). The total amounts of leachate produced and used for the analysis are summarized in Table 3-3.

Among the 8 metals under consideration, As, Fe, Pb, Mn and Zn increased

substantially for the first 10 to 15% of mass loss and reached a plateau. These leaching patterns indicate that metal leaching mainly occurred at the initial phase, an acidic environment, in both aerobic and anaerobic conditions. In contrast to these metals, only minor change in cumulative masses of Al and Cu was observed from the anaerobic lysimeters whereas a consistent increase in these metals was exhibited from the aerobic lysimeters. Relatively high concentrations of Al were exhibited during the initial stage of the anaerobic lysimeters, but no further changes were observed. It is notable that cumulative concentrations of Al and Cr increased more rapidly after 25% mass loss occurred. An increase in the rate of accumulation of these metals corresponds to an increase in pH to 9.

Overall metal leaching behavior is strongly associated with pH and redox

conditions. Since the anaerobic lysimeters remained in the acidic condition (pH < 6) for more than 400 days, great amounts of metals such as As, Mn, Fe, Cr and Zn were released through the leachate. In contrast to the anaerobic lysimeters, greater cumulative concentrations of Al, Cu and Pb were observed from the aerobic lysimeters. Leaching of





63


these metals might be influenced by the different environment of the aerobic lysimeters such as an alkaline pH and oxidizing conditions.

3.4.2 Comparison to Other Studies

Generally, heavy metal concentrations found in anaerobic landfills are reported low (Kjeldson et al., 2002). The presence of sulfide and the low solubility of metals at neutral pH may reduce metal concentrations in leachate. Metal concentrations of the aerobic and anaerobic lysimeters along with MSW leachate summarized from the literature are presented in Table 3-6. Metal concentrations of the aerobic lysimeters listed in Table 3-6 are the average value of the lysimeters 1 and 2. For the anaerobic lysimeters, since lysimeter 3 remained in acidic condition for most of a test period, metal results of lysimeter 3 during the methanogenic phase were not included in Table 3-6. For the anaerobic lysimeters, As and Zn concentrations were substantially higher than those of MSW leachate during acidic phase. They were reduced then during the methanogenic phase and similar to those of MSW leachate despite the presence of CCA-treated wood. Cu concentrations of the anaerobic lysimeters were extremely low, and they were lower than even drinking water standards. Although most metal concentrations of the anaerobic lysimeters were greater than drinking water standards, they were similar or lower than those of general MSW leachate during the methanogenic phase.

For the aerobic lysimeters, As, Cr, Fe, Mn and Zn concentrations were lower than those of MSW leachate and the anaerobic lysimeters during the first acidic phase. Aluminum and copper concentrations were greater than those of the anaerobic lysimeters. Only Pb concentration was greater than that of MSW landfill leachate and the anaerobic lysimeters. However, different aspects of metal leaching were observed from the aerobic





64


lysimeters during the alkaline phase; the concentrations of most metals except for Fe were greater than those of the anaerobic lysimeters.

The leaching behavior of CCA-treated wood of the aerobic and anaerobic

lysimeters was compared to a similar study (Jambeck, 2004). Jambeck (2004) researched the leaching behavior of CCA-treated wood mixed with MSW through the 6.7-m high PVC column tests. A total 2% of CCA-treated wood was included in the column and rain water was used for Jambeck's study while 1% of CCA-treated wood and DI water were used for the present study (Table 3-7). In comparison metal leaching results from Jambeck's study proved similar to the anaerobic lysimeter here (Figure 3-18). Extremely low Cu concentrations were observed in both studies. The range of Cr concentrations of Jambeck's study was higher than that of the anaerobic lysimeters during acid phase, but the median of Cr concentrations of Jambeck's study was bottom of the range. In contrast to anaerobic condition, significantly different leaching trends of As and Cu were exhibited from the aerobic lysimeters; overall As concentrations of the aerobic lysimeter were lower than those of the anaerobic column studies during the acid phase. The 95th percentile of As concentrations of the aerobic lysimeter was in the range of the anaerobic system, but they were detected at the very beginning of the aerobic lysimeter operation. After pH stabilized, the median As concentrations of both the aerobic and anaerobic systems became identical. However, Cu concentrations were two orders of magnitude higher than those of anaerobic lysimeters for both the acid and methane (alkaline) phase.

In comparison with other column study (Jambeck, 2004), it can be concluded that leached As, Cu and Cr concentrations might not always be followed by the initial mass of CCA-treated wood and metal concentrations contained. The differences between As and





65


Cu concentrations between the anaerobic and aerobic columns proved too high to be comparable. Within the same anaerobic systems, As and Cr concentrations between Jambeck and this study were not much different despite different initial As and Cr masses. Similar leaching trends could be observed in the same anaerobic or aerobic system, but overall concentrations of As and Cr per initial mass may depend more on the chemistry of the system.

3.4.3 Implication for Disposal of Heavy Metals

In this research, CCA-treated wood and CRT monitor glass were used as metal

sources to represent treated wood and electronic waste. Though the land-disposal of these wastes was banned in several states in the U.S. (SWANA, 2003), they can still be landdisposed as a form of home appliances and ash. Since the greatest amount of As may be leached during the first acid phase of anaerobic landfills, landfill owners need to monitor the leachate quality during landfill construction. Though thermodynamically As may precipitate with sulfide, a maximum of 70% of As dissolved in solution may combine with sulfide at pH 8 (Carbonell-Barrachina, 1999). This adsorption ratio decreased as the pH decreased. In addition to the sulfide, Fe, organic matters, and carbonate are also known as As adsorbent. However, the adsorption efficiency of those ligands were low relative to sulfide (Carbonell-Barrachina, 1999). Thus, in the presence of an As source, As can be found in landfill leachate for all operation periods. For aerobic landfills, As concentrations may slightly increase during the alkaline phase. Overall As concentrations found in aerobic landfills may be lower than those of anaerobic landfills.

Thermodynamically, Cr concentration in landfill condition during all phases may be low. During the first acid phase, Cr may be combined with high concentrations of Fe

(11), and Cr may precipitate with sulfide during the methane phase. The lower





66


concentration of Fe (11) may increase the Cr solubility, but the change is very small. Cu concentrations in anaerobic landfills are typically extremely low. However, high concentrations of Cu may be found in both acid and alkaline phase of aerobic landfills. Since the microorganisms which contribute Cu leaching may corrode all Cu-made equipment connected to the landfill, it is important to avoid using any Cu-containing equipment for gas and leachate collection systems. Pb has high solubility under oxidizing conditions at pH < 6. Thus it would be recommendable to monitor leachate quality for the first acid phase of aerobic landfills. However, air addition facilities are generally installed after landfill closure, Pb concentrations may not be high over operation period. Pb concentration in anaerobic landfill conditions is generally low.

As previously discussed, air addition into a current anaerobic bioreactor landfill may enhance waste decomposition substantially. However, unlike anaerobic landfills, concerns about high Al and Cu concentration and a risk of Cr (VI) may arise. In order to avoid these potential risks, it would be recommendable to inject air into the shallow well rather than the deep well. Since Al, Cu and Cr are redox-sensitive, great amount of these metals could be reduced after passing through the anaerobic zone. However, since air injection into the shallow wells may reduce the air diffusion efficiency into a landfill, further economical and efficiency of air distribution analysis are needed.

3.4.4 The Imp act of Air on Metal Mobility

Although total amounts of metals leached were not considerably high, it is

necessary to pay great attention to certain metals due to changes of their toxicity by different pH and redox conditions. For example, among Cr species dissolved in leachate, Cr (111) can be dominant in current anaerobic sanitary landfills, however, thermodynamically Cr (VI) becomes a major Cr component in the environment formed





67


upon air intrusion. Hexavalent chromium is a highly toxic metal causing decreased pulmonary function and pneumonia (Bradle, 2005). On the contrary, toxicity of As can be reduced by air injection. In oxidizing conditions, Arsenate can be oxidized to As (VI), which is less toxic. Since As (VI) is less mobile and has a low solubility, the total amount of As leached can be reduced. The influence of air injection on the amount of metals can be more clearly understood by comparing the percentage of As, Cr and Cu (Table 3-4). Under the scenario of high As content in leachate caused by co-disposal of fly ash or ground CCA-treated wood, it would be recommended to inject air in order to reduce As toxicity and the amount leached. However, Cr (VI) in an aerobic landfill can be present as an ionic form under oxidizing conditions, and the cumulative mass may increase over a period of time. For these reasons, it is necessary that landfill personnel develop a strategy to optimize the influence of air injection.

Based on the metal results, the metal leaching behavior of old landfill metals can be predictable. Kjeldsen et al. (2002) proposed that the condition of old landfills would be aerobic due to air intrusion. Once air intruded in anaerobic landfills, metal leaching may occur by the dissociation of the metals that adsorbed on decomposed organic matters. Among 8 metals under consideration, Pb would be the most leachable metal. As Figure 313 shows a large amount of Pb was adsorbed on organic matter. Bozkurt et al. (1999) explained that the pH of the landfill would change to acid again after the long-term process. This was due to the oxidation of sulfate and organic matter. Therefore, the moderate oxidizing condition, and low pH, would accelerate Pb leaching.

3.5 Conclusions

Research on the fate of metals in simulated aerobic and anaerobic landfills was

conducted. The leaching behavior of selected metals was significantly different between





68

aerobic and anaerobic lysimeters. Among the 8 metals evaluated (Al, As, Cr, Cu, Pb, Mn, Fe and Zn), the concentrations of Al, Cu, Cr and Pb in leachate of the aerobic lysimeters were significantly greater than those of the anaerobic lysimeters, and the average concentrations of As, Fe, Mn, and Zn in the anaerobic lysimeters were significantly greater in concentration than observed in the anaerobic lysimeters.

After a test period, one each of the aerobic and anaerobic lysimeters was

dismantled and the metals adsorbed on decomposed lignocellulosic waste were analyzed. Greater concentrations (mg metal/kg waste) of Fe, Mn, As, Al and Cu were found from the aerobic lysimeters. All metals, except for Pb, adsorbed on the cardboard in the aerobic lysimeter and were significantly higher than those of the anaerobic lysimeter. Through this research, aerobic landfills were found to have a greater potential to release several metals through leachate than that of anaerobic landfills; aerobic landfills have greater Al, Cr, Cu and Pb leaching potential due to oxidizing condition. Experimental results confirmed this notion. In the presence of CCA-treated wood and electronic waste (CRT monitor glass), high As and Pb concentrations were observed from the anaerobic lysimeter and the aerobic lysimeter during the acid phase, respectively.





69


Table 3-1 .Heavy metal sources in fabricated waste stream.
Waste components Contained heavy metals % of component in fabricated
____________________waste
CCA treated wood Copper, Chromium and 1%
Arsenic
Cathode-ray Tube (CRT) glass Lead 1%
Aluminum sheet Aluminum 4%
Galvanized steel sheet Zinc, Manganese and Iron 4%



Table 3-2. Results of statistical analysis of metal leached between aerobic and anaerobic Average concentrations (mg/L) F P-value F-crit
Aerobic Anaerobic
Al 7.89 1.28 206.89 8.3E-33 3.89
As 0.40 1.28 66.15 4.1E-14 3.89
Cr 0.19 0.10 40.67 1.19E-09 3.89
Cu 2.87 0.02 81.10 1.58E-16 3.89
Fe 35.06 167.68 53.09 6.92E-12 3.89
Mn 1.91 4.57 35.80 9.75E-09 3.89
Pb 0.22 0.03 31.32 7.07E-08 3.89
Zn 54.36 201.12 66.30 3.87E-14 3.89



Table 3-3. The amount of leachate produced and used for analysis lys 1 lys 2 lys 3 lys 4
leachate produced (mL) 8,717 9,747 18,571 15,979
leachate released (mL) 4,024 4,081 6,135 6,111
(used for analysis) I





70


Table 3-4. Leachability of As, Cr, and Cu

Initial conc. (mg/lys) mg released % released
aerobic anaerobic aerobic anaerobic

As 1279.1 Lys 1 Lys 3 1.53 8.21 0.12% 0.64%
mg Lys 2 Lys 4 1.17 9.30 0.09% 0.73%

Cr 1573.1 Lys 1 Lys 3 0.72 0.50 0.05% 0.03%
mg Lys 2 Lys 4 0.56 1.24 0.04% 0.08%
723.9 Lys 1 Lys 3 9.82 0.10 1.36% 0.01%
CU mg Lys 2 Lys 4 6.37 0.34 0.88% 0.05%



Table 3-5. Comparison of cumulative mass of metal dissolved in leachate and adsorbed
on li nocellulosic materials (unit: mg)
Aerobic Anaerobic LC/AD LC/AD
lys 1 lyis 2 lys 3 lys 4 (lys 2) (lys 4) of lys 2 of lys 4
Al 27.3 25.9 7.8 8.7 41900 15700 0.06% 0.06%
As 1.5 1.2 8.2 9.3 110 67.6 1.06% 13.76%
Cr 0.7 0.6 0.5 1.2 222.2 243 0.25% 0.51%
Cu 9.8 6.4 0.1 0.3 238.1 151.2 2.68% 0.22%
Fe 161.3 64.2 1200 506.4 31100 19200 0.21% 2.63%
Mn 10.5 2 31 29.3 494.2 177.3 0.40% 16.51%
Pb 0.6 0.5 0.2 0.1 47.8 156.7 1.13% 0.08%
Zn 292.9 64.2 1,100 1,100 9,400 12,500 0.69% 8.60%






71


Table 3-6. Comparison of average metal concentrations of the aerobic and anaerobic
lysimeters with MSW leachate and regulatory levels (SAIC, 2000; USEPA,
1996 and 2003; Kjeldsen et al., 2002) (unit: mg/L)

Aerobic lysimeters Anaerobic lysimeters Drinking
MSW TLP TC water
leachate Alkali Methane limits str
Acid phase phase Acid phase phase

Al 15.05 4.56 9.07 0.31 0.17 0.2*
As 0.44 0.19 0.55 1.5 0.43 5 0.01
Cr 0.24 0.09 0.26 0.1 0.14 5 0.1
Cu 0.14 4.01 1.75 0.02 0.07 1.3
Fe 3.00 61.66 3.89 188.26 10.32 0.3*
Pb 0.13 0.37 0.03 0.03 0.01 5 0.015
Mn 6.08 3.35 0.09 5.63 0.07
Zn 5.1 96.32 8.26 250.52 4.58 5*
* secondary drinking water standards



Table 3-7. Comparison of characteristics of CCA-treated wood used for Jambeck (2004)
and this study
Jambeck(2004) This study
The % of CCA-treated included in waste stream 1% 1% 1%
As 1390 20.0 1960 27. 1330 10
Cu 814 52.4 1340 54.0 2350 50
Cr 1450 68.3 2550 48.0 2890 56






72



100
AEROBIC Lys
Lysi1 O Lys 2

o *

10 0*00 0 0 00
00 0 0 00 00 0
0 *0 0
0 0 0


0 0
0







0.1
0 100 200 300 400
100
ANAEROBIC
A Lys 3 Lys4 10 O




-A

A
1&


A AA AA
0.1 A IL



0.01



0.001
0 200 400 600
Days

Figure 3-1. Changes of Al concentrations over time





73





1.0
AEROBIC Lys 1
Lysi1
0.8 00 0 Lys 2
0
0.8 0 0



9 0.6 0
O O O
0.4 O-l 0
0O
40o *o

0.
<0.4 * 0 0



0 0 go
0
0 0
0.2 g
o0

0.0 o

0 100 200 300 400
3.5
A ANAEROBIC A Lys 3
3.0 4~ Lys4
2 5 -,- A
.. 2,0 A A x
2.5 A A
A
4 A A AA
2.0 A
8b A

< 1.5 -A A

A AA AA 4
1.0 A A
Al A A A tt

0.5 A AA
A k AA A 4Q4Z,

0.0
0 200 400 600 800
Days

Figure 3-2. Changes of As concentrations over time





74





0.5

Lys 1 AEROBIC O
0 Lys 2 So
0.4


00 0 0 0
p S

0.3 0
0 0
E o* 00
0
0.2 0 0 O o 0
0 S 0
*o0
o d" o

o 0 0
o 0o
0
0.0 11 1 1 1 1
0 50 100 150 200 250 300 350
0.8

ANAEROBIC A Lys 3
A Lys 4

0.6

A


0.4 A
SAA

A A
0.2 AA A
A AA
,tAAA
A PAA
SAA A

0.0,
0 200 400 600
Days

Figure 3-3 Changes of Cr concentrations over time






75





100
AEROBIC 0 Lys 1 O Lys 2



10 -0 0

e o
0 0 0 0
00


of* 0
U0 0 0
0 0000
0 0 00

0 0

0 0
0 O0

0.1 I I I
0 50 100 150 200 250 300 350
1
ANAEROBIC








0.1 A Ls
;A



-6 A
A A AL A A A



AAA Lys 4 AdUL A A A i&AA A A
UkAA A AAh A j AA
0.01 A&A A A AA,&A A A

Below Detection Limit A A A A


A Lys 3
A Lys 4
0.001
0 200 Days 400 600

DaysFigure 3-4. Changes of Cu concentrations over time Figure 3-4. Changes of Cu concentrations over time






76





10
AEROBIC 0 Lys 1
O Lys 2


1- 000
e*
0 00 0
0 0

0 o
0 0 0
0.1 0
0 0

0 0 0 0 QD
0 00 0


0.01 0





0.001 I I I I I
0 50 100 150 200 250 300 350
1

A Lys 3 ANAEROBIC
A Lys 4



0.1 A A AA A


A AA A
A AA A 4 A
A01 AA A
A:*r6A AA AAAAA AAA A A
A A
0.1 AA AAA A A
0.01 A

A A A
A A

Below detection limit of ICP for Pb

0.001 I I ,
100 200 300 400 500 600 700
Days


Figure 3-5. Changes of Pb concentrations over time





77





1000
AEROBIC Lys
Lys 1
0 0 Lys 2

100 0O O s
00 0 0
10O 0 0 0

8 o4
O0 0
E 1o
6 10

0 0 0
0 *
0.

01

0 .1 I I I I
0 50 100 150 200 250 300 350
1000
A Lys 3 ANAEROBIC AAA A A A 1%
A Lys 4 A AA A


A AA A
100 -A A A A


6A A


10 AA A







0 200 400 600
Days

Figure 3-6. Changes of Fe concentrations over time






78





100
AEROBIC O Lys 1

0 Lys 2


10 @
.0 00 0
0 0 0
0 00 0




11 00 0 0
0 0


0O 0
0.1 eo o 0.0O


.009 o 0 0 0


0.01 ,
0 50 100 150 200 250 300 350
100

ANAEROBIC A Lys 3 A Lys 4


10 -A
AAA


A 1 AA A'

A


A
0.1 A A
A A A



0.01 ,
100 200 300 400 500 600 700

Days


Figure 3-7. Changes of Mn concentrations over time






79





300
AEROBIC Lys 1

250 O Lys 2
250

0 0 200



E 150
N
0 0
100
o *

050
o g0 co 0 0 0 BDL


0 50 100 150 200 250 300 350

Days


500

A ANAEROBIC A Lys 3
A A Lys 4 400
A




200
A A
AA A
P300 A A A

AA
N 200 A A

AA
A A A A
A
100

AA


100 200 300 400 500 600 700

Days


Figure 3-8. Changes of Zn concentrations over time





80






3.5
As 0 lys 1

3.0 A O lys 2
A v lys 3
A A lys 4
2.5 A

V
o 2.0 VV

1.5 A
o 4

1.0 A*
A CkiA

0.5 AA A
.0 ()
0.0
4 5 6 7 8 9 10

pH


Figure 3-9. Distribution of As over a C-pH diagram





81


20 1.2

16- 1.0
CrOT4 -0.8


3+ 0.6

4rO

4-002.

0 r( J ( O ) s I 0 .
I4/ Cr (OH)OV


-82


"0 2 46 8012 14
pH

Figure 3 -10. Potential- pH diagram of Cr (Richard and Bourg, 199 1)






82






100
Cu



10 -0 00 @0
00 *t 000

0 g 0
0 A S 0.1
0V




0.0 0 Iys 2
VS 6 VV Iys 3
V ~A lys 4

0.001 III
4 5 6 7 8 9 10

pH


Figrue 3-11. Distribution of Cu over a C-pH diagram







83






18 0.035

16 Al CB As CB
16__NP 0.030 NP
-OP
'~14 OP WD
Sb WD 0.025
12
0.020
S10

8 0.015
6
0.010


0.005


0 IWLLU 0.000
2-1 2-2 2-3 2-4 4-1 4-2 4-3 4-4 Raw 2-1 2-2 2-3 2-4 4-1 4-2 4-3 4-4 Raw


0.08 0.08

Cr CB Pb CB
NP NP
OP OP
S0.06 WD 0.06 WD



F0.0
0.04 0.04

0
0.0


0.02 0.02
0.00 0.00

2-1 2-2 2-3 2-4 4-1 4-2 4-3 4-4 Raw 2-1 2-2 2-3 2-4 4-1 4-2 4-3 4-4 Raw



Figure 3-12. Adsorption of metal on solid wastes






84





0.16 10
Mn CB Fe CB

0.14 NP NP
OP 8 OP
0.12 WD WD

0.10 6

,F 0.08

0.06 4
0
,, 0.04
2
0.02

0.00 0
2-1 2-2 2-3 2-4 4-1 4-2 4-3 4-4 Raw 2-1 2-2 2-3 2-4 4-1 4-2 4-3 4-4 Raw


0.08 3.5
Cu CB Zn CB
NP 3.0 NP
OP 1OP
( 0.06 WD WD
E 2.5
'o 2.0

0.04
1.5
10
0.02

0.5


0.00 0.0 -2-1 2-2 2-3 2-4 4-1 4-2 4-3 4-4 Raw 2-1 2-2 2-3 2-4 4-1 4-2 4-3 4-4 Raw



Figure 3-12(continued)






85






50000 120
Al As

100
40000

80
30000 60
060
20000 ':



10000 20


0 0
aerobic anaerobic aerobic anaerobic
0
"0 300 300
Cr Cu
0
250 C 250 C

Ca
0 200 200
0

150 -150


100 -100


50 50


0 0
aerobic anaerobic aerobic anaerobic



Figure 3-13. The comparison of aerobic and anaerobic lysimeters in respect of total mass
of metals adsorbed on lignocellulosic materials