EPA/600/R-07/060
September 2006
Landfill Bioreactor Performance
Second Interim Report
Outer Loop Recycling & Disposal Facility
Louisville, Kentucky
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
Page Intentionally Blank
-------
Notice
The U.S. Environmental Protection Agency through its Office of Research and Development partially
funded and collaborated in the research activities described herein under Cooperative Research and
Development Agreement Number 0189-00 with Waste Management Inc. and under contract number
EP06C000146 with GeoSyntec Consultants. This report has been subject to both internal and external
Agency review and has been approved for publication as an U.S. EPA document.
in
-------
Page Intentionally Blank
IV
-------
Acknowledgments
The report presented herein is collaborative in nature. Considerable effort has been devoted to the Outer
Loop Landfill Bioreactor project. Over the past five years, the following individuals and organizations
have assisted in the design, construction, operation, and data collection and analysis aspects of the project:
U.S. EPA
Thabet Tolaymat Ph.D.
Fran Kremer Ph.D.
David Carson
Wendy Davis-Hoover Ph.D.
Waste Management
Gary Hater
Roger Green
John Barbush
Chad Abel
Rick Ban-
Kevin Meiczkowski
Greg Cekander, PE
Chuck Williams, PE
GeoSyntec Consultants
Jon Powell
Mazen Haydar Ph.D.
Robert Bachus Ph.D., PE
Michael Houlihan, PE
Jeremy Morris Ph.D., PE
Alternative Natural Technologies
C. Douglas Goldsmith Ph.D.
North Carolina State University
Morton Barlaz Ph.D., PE
Neptune Inc.
Paul Black Ph.D.
Doug Bronson Ph.D.
David GratsonPh.D.
University of Cincinnati
Robert Grosser Ph.D.
David Feldhake
Brian Morris
-------
Page Intentionally Blank
-------
Foreword
The U.S. Environmental Protection Agency (U.S. EPA) is charged by Congress with protecting the
Nation's land, air, and water resources. Under a mandate of national environmental laws, the Agency
strives to formulate and implement actions leading to a compatible balance between human activities and
the ability of natural systems to support and nurture life. To meet this mandate, U.S. EPA's research
program is providing data and technical support for solving environmental problems today and building a
science knowledge base necessary to manage our ecological resources wisely, understand how pollutants
affect our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for investigation of
technological and management approaches for preventing and reducing risks from pollution that threaten
human health and the environment. The focus of the Laboratory's research program is on methods and
their cost-effectiveness for prevention and control of pollution to air, land, water, and subsurface resources;
protection of water quality in public water systems; remediation of contaminated sites, sediments and
ground water; prevention and control of indoor air pollution; and restoration of ecosystems. NRMRL
collaborates with both public and private sector partners to foster technologies that reduce the cost of
compliance and to anticipate emerging problems. NRMRL's research provides solutions to environmental
problems by: developing and promoting technologies that protect and improve the environment; advancing
scientific and engineering information to support regulatory and policy decisions; and providing the
technical support and information transfer to ensure implementation of environmental regulations and
strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan. It is
published and made available by U.S. EPA's Office of Research and Development to assist the user
community and to link researchers with their clients.
Sally Gutierrez, Director
National Risk Management Research Laboratory
VII
-------
Page Intentionally Blank
Vlll
-------
Executive Summary
A bioreactor landfill is a landfill that is operated in a manner that is expected to increase the rate and extent
of waste decomposition, gas generation, and settlement compared to a traditional landfill. This Second
Interim Report was prepared to provide an interpretation of field data collected as part of a multi-year
Cooperative Research and Development Agreement (CRADA) between the U.S. Environmental Protection
Agency (U.S. EPA) and Waste Management, Inc. (WM). The CRADA was established to evaluate the
performance of landfill bioreactor units at the WM Outer Loop Landfill, located in Louisville, Kentucky.
This report follows the September 2003 U.S. EPA document Landfills as Bioreactors: Research at the
Outer Loop Landfill, Louisville, Kentucky, First Interim Report (i.e., the First Interim Report, EPA/600/R-
03/097), which presented a complete description of the landfill study sites and the data collection
procedures.
The Outer Loop Landfill Bioreactor (OLLB) project considers solid waste decomposition, moisture
balance, landfill gas generation, and leachate quality to evaluate the effect of bioreactor operations on
municipal solid waste (MSW) decomposition. Three types of landfill cells were evaluated in the OLLB
study: (i) a Control cell, in which no liquids were added; (ii) a cell in which liquids were added after the
cell had been completely filled with waste (i.e., the Retrofit cell); and (iii) a cell in which liquids and air
were added as the waste was placed in the landfill (i.e., the As-Built cell). The monitoring data were
sequentially evaluated to identify trends in solid waste decomposition, moisture retention, landfill gas
quality and quantity, and leachate quality. One must recognize the limitation of data presented in this
report in establishing long-term trends in the operation of bioreactor landfills. Below is a brief description
of data evaluation presented in this report.
Operations
• The results of the moisture balance calculations indicate an increase in moisture content of six to seven
percent in the As-Built cells, an increase of approximately one percent in the Retrofit cells and a slight
decrease in the Control cells during the study period;
• Data regarding leachate head in the sump, which was used as an indirect indicator of leachate head on
the liner, indicated that operating a landfill as a bioreactor caused an overall increase in leachate head
in the sump compared to the Control cells. However, in all cases, the average leachate level on the
liner was well below the required 0.3 m (1 ft);
• To date, there is no indication that the bottom liner system of the test cells was compromised while
installing liquid application features, or while applying liquid through those features;
• Although leachate breakouts occurred intermittently in the bioreactor landfill cells, no significant
leachate breakout occurred that would result in any surface-water quality impacts;
• The lack of significant leachate breakouts (breakouts resulting in surface water quality impact) and the
lack of landfill gas emission problems (methane surface emissions were detected, however at
concentrations less than 500 ppm) suggest that, to date, the Outer Loop Bioreactor Landfill has been
operated in a manner that will minimize problems related to excessive pressures in the cells;
• There were indications of "watering-out" of gas collection wells and trenches which was addressed by
using submersible pumps to pump out the free standing liquid;
• During the period of the study, there was an increase in landfill gas collected from the bioreactor areas.
Therefore, it is not possible to conclusively know what affects the watering-out of the collection
features have on the gas collection efficiency;
• During the period of the study, there were no signs that "watering-out" of the gas collection wells
posed a geotechnical instability problem at the site;
• There was no indication of clogging in the leachate collection system during the study period; and
IX
-------
• Overall, the bioreactor landfill cells generally met the criteria of Subtitle D of the Resource
Conservation and Recovery Act for design and operation of MSW landfills. Data, from this site,
suggest that other well-designed and well-operated bioreactor landfills can also be operated in
compliance with the requirements of Subtitle D.
Solids Decomposition
• The solids composition data support conclusions reached by previous research that the operation of an
MSW landfill as a bioreactor results in accelerated solids decomposition. All solid waste monitoring
parameters showed accelerated waste decomposition in one of the As-Built bioreactor landfill cells
relative to the Control cells. However, results were as not as conclusive for the second As-Built
bioreactor landfill cell. The relatively young age of the solid waste made definitive decay rate
estimates difficult.
Landfill Gas
• Results indicate that, although it is variable, the rate of landfill gas generation in the As-Built
bioreactor landfill cells was greater than that of the Control cells, potentially providing a greater rate of
energy production if collection occurred early and consistently;
• The landfill gas decay constant for As-Built bioreactor landfill cells was evaluated to be 0.16 yr^while
the Retrofit cells and the Control cells had a k value of approximately 0.061 yr"1; and
• Although the concentration (ppmv) of non methane organic carbon (NMOC) in the collected landfill
gas did not appear to be higher in the landfill bioreactor cells compared to the Control cells, the overall
production was higher because of the higher gas flow rate.
Leachate Quality
• The leachate quality test results were generally well correlated to the age of the waste for each cell for
the study period. However, no strong correlations were found between leachate quality and
accelerated waste decomposition in the bioreactor landfill cells, as the bottom, most degraded part of
the solid waste mass has the largest impact on the leachate quality;
• Evaluation of the biochemical oxygen demand to chemical oxygen demand ratio (which is generally a
strong indicator of organic solids decomposition) revealed that waste decomposition in the As-Built
bioreactor landfill cells may have been accelerated compared to the Control cells; and
• Liquids that were added to the Retrofit cells were pre-treated to decrease ammonia concentrations.
This was effective since leachate quality test results from this study showed a decrease in ammonia
concentration during the study period in the Retrofit bioreactor landfill cells.
Overall, the analysis of the data collected during the first five years indicate that the addition of liquids
increased the moisture content of waste in the landfill bioreactor cells and accelerated waste degradation.
Leachate quality and solid waste decomposition data indicate that waste degradation was enhanced in the
As-Built landfill bioreactor cells. Landfill gas quantity data indicate that the decay rate was highest in the
As-Built cells and lowest in the Control cells as expected.
Future Activities
To substantiate the effectiveness of landfill bioreactors, research is focused in the following areas:
• Evaluate and adjust sampling frequency of the various monitoring/operational parameters;
• Examine NMOC emissions through various landfill covers;
• Evaluate alternative approaches to assess and engineer controls for fugitive emissions;
• Determine shear strength as a function of age of the wastes and assessment of landfill stability;
• Identify the effectiveness of landfill bioreactors in enhancing carbon sequestration; and
• Identify the effectiveness of bioreactors in treating/containing nanoparticles.
-------
It is anticipated that the data collected throughout the remainder of the CRAD A test period will further
support the conclusions reached in this report and allow better definition of the correlations described.
Finally, it is concluded that, if the trends illustrated in this report for all of the monitored media (i.e., solids,
liquids, and gases) are confirmed at the end of the CRAD A test period, then there will be a significant
increase in our understanding of bioreactors and increased potential to predict their performance.
XI
-------
Page Intentionally Blank
Xll
-------
Table of Contents
Notice iii
Acknowledgments v
Foreword vii
Executive Summary ix
Table of Contents xiii
List of Figures xv
List of Tables xix
List of Acronyms xxi
Chapter 1. Introduction 1
1.1 Landfill Bioreactor Technology 1
1.2 Outer Loop Landfill Bioreactor 1
1.3 Project Objectives 1
1.4 Report Organization 2
Chapter 2. Overview of Bioreactor Landfills 3
2.1 Introduction to Landfill Bioreactors 3
2.2 Regulatory Overview 4
2.3 Anaerobic Decomposition Fundamentals 5
2.3.1 Phase I (Initial Adjustment) 6
2.3.2 Phase II (Transition) 6
2.3.3 Phase III (Acid Formation) 6
2.3.4 Phase IV (Methane Fermentation) 7
2.3.5 Phase V (Final Maturation and Stabilization) 7
2.4 Key Monitoring Parameters for MSW Landfill Bioreactors 7
2.4.1 Physical Monitoring Parameters 8
2.4.2 Analytical Monitoring Parameters 10
Chapters. Site Description and Analytical Methods 19
3.1 Site Description and System Design 19
3.1.1 Control Landfill Cells 19
3.1.2 As-Built Landfill Bioreactor Cells 21
3.1.3 Retrofit Landfill Bioreactor Cells 25
3.2 Sample Procedures and Methods 29
3.2.1 Leachate Sampling 29
3.2.2 Municipal Solid Waste Sampling 30
3.2.4 Landfill Settlement 34
3.2.5 Landfill Gas Sampling 34
3.3 Data Processing and Statistical Analysis 36
3.3.1 Sample Dating and Statistical Analysis 36
3.3.2 Moisture Balance Calculations 37
3.3.3 Statistical Analysis of Leachate Parameters 38
Xlll
-------
Chapter 4. Solid Waste and Moisture Content Analysis 39
4.1 Solid Waste Analysis 39
4.1.1 Results 39
4.1.2 Multiple Linear Regression Models 47
4.1.3 Summary of Solids Decomposition 48
4.1.4 Solid Waste Surface Settlement 54
4.1.5 Landfill Temperature and Oxidation Reduction Potential (ORP) 54
4.1.6 Solid Waste Slope Stability 54
4.2 Moisture Addition 55
4.2.1 Moisture Balance 55
4.2.2 Leachate Head on Liner 59
4.2.3 Measured Waste Moisture Content 65
4.2.4 Evaluation of Calculated and Measured Moisture Content 66
4.2.5 Moisture Content Analysis Summary 69
Chapter 5. Landfill Gas (LFG) 71
5.1 Landfill Gas (LFG) Composition 71
5.1.1 Control cells 71
5.1.2 As-Built cells 71
5.1.3 Retrofit Cells 74
5.1.4 Landfill Gas (LFG) Composition Summary 75
5.2 Measured Methane Production and LandGEM Model Predictions 77
5.3 Comparison of Field Gas Results of Control and Landfill Bioreactor Cells 83
5.3.1 Summary of LFG Generation 86
5.4 None Methane Organic Carbon (NMOC) Concentrations in LFG 87
Where: 91
5.5 Methane Surface Emissions 93
Chapter 6. Leachate Quality 95
6.1 Temperature 95
6.2 pH 97
6.3 Volatile Organic Acids (VOAs) 99
6.4 Total Organic Carbon (TOC) 101
6.5 Biochemical Oxygen Demand (BOD) and Chemical Oxygen Demand (COD) 103
6.6 Nitrogen Content 108
6.7 Metals 112
6.8 Volatile and Semi-Volatile Organic Compounds 113
6.9 Phosphorous Content 115
6.10 Chloride 118
6.11 Leachate Quality Summary 119
Chapter 7. Landfill Bioreactor Performance Analysis 123
7.1 Slope Stability 123
7.2 Liner and Final Cover Integrity 123
7.3 Liquids Addition System Performance 124
7.4 Leachate Collection System Performance 124
7.5 Landfill Gas Production and Emission 125
Chapters. Conclusions and Recommendations 127
8.1 Conclusions 127
8.2 Recommendations 130
8.3 Landfill Bioreactor Monitoring Parameters 131
Chapter 9. References 133
xiv
-------
List of Figures
Figure 2-1 Phases of Solid Waste Decomposition (source: Pohland et al. 1993) 6
Figure 2-2 Changes in VFA in a Single Pass (Conventional) Reactor versus Bioreactor9 and Bioreactor21
over Time (Source: Sponza and Agdag, 2004) 13
Figure 3-1 Outer Loop Landfill Site Map 20
Figure 3-2 Solid Waste Placement in the As-Built Landfill Bioreactor Cells 22
Figure 3-3 Solid Waste Placement in As-Built Landfill Bioreactor Cells 22
Figure 3-4 Cumulative Industrial and Total Added Liquids in As-Built Cell A 24
Figure 3-6 Solid Waste Placement in the Retrofit Landfill Bioreactor Cells 26
Figure 3-7 Bottom Liner Configuration of the Retrofit Unit 27
Figure 3-8 Cross-Sectional Layout of Liquid Introduction Trench 28
Figure 3 -9 Cumulative Liquid Introduction into Retrofit Bioreactor Landfill Unit by Subcell 28
Figure 3-10 Liquid Introduction and Gas Collection System in Retrofit Landfill Bioreactor Cells 30
Figure 4-1 Summary of Waste Age for Retrofit Cells A (5.1) and B (5.2) 40
Figure 4-2 Relationship between CELL and Waste Age in the Retrofit Cells 42
Figure 4-3 Relationship between Moisture Content and Solids Decomposition in the Retrofit Cells 43
Figure 4-4 Waste Age Profile for the Control Cell A (7.3A) and Control Cell B (7.3B) 43
Figure 4-5 Moisture Content Profile for the Control cells at Each Sampling Time 44
Figure 4-6 Relationship between Moisture Content and Solids Decomposition in the Control Cells 44
Figure 4-7 Waste Age Profile for As-Built Cells A and B 46
Figure 4-8 Relationship between Moisture Content and Solids Decomposition in As-Built Cell A 47
Figure 4-9 Relationship between Moisture Content and Solids Decomposition in As-Built Cell B 48
Figure 4-10 Comparison of Moisture Content in the Control, Retrofit, and As-Built Cells 50
Figure 4-11 Trends in CH:L as a Function of Waste Age 51
Figure 4-12 Trends inBMP as a Function of Waste Age 52
Figure 4-13 Trends in Organic Solids as a Function of Waste Age 53
Figure 4-14 Cumulative and Mean Monthly Precipitation 56
xv
-------
Figure 4-15 Moisture Balance and Calculated Waste Moisture Content of the Control Cells 57
Figure 4-16 Moisture Balance and Calculated Waste Moisture Content of Retrofit Landfill Unit 58
Figure 4-17 Moisture Balance and Calculated Waste Moisture Content of As-Built Cell A 58
Figure 4-18 Moisture Balance and Calculated Waste Moisture Content of As-Built Cell B 59
Figure 4-19 Mean Monthly Leachate Head onLinerof the Control Cell 60
Figure 4-20 Mean Monthly Leachate Head and Leachate Volumes Generated in the Control Cell 61
Figure 4-21 Mean Monthly Leachate Head on Liner for the Retrofit Cells 62
Figure 4-22 Mean Monthly Leachate Head and Leachate Generated Volumes in Retrofit Cells 62
Figure 4-23 Mean Monthly Leachate Head on Liner of As-Built Cell A 63
Figure 4-24 Mean Monthly Leachate Head on Liner of As-Built Cell B 63
Figure 4-25 Mean Monthly Leachate Head and Leachate Generated Volumes in As-Built Cell A 64
Figure 4-26 Mean Monthly Leachate Head and Leachate Generated Volumes in As-Built Cell B 64
Figure 4-27 Calculated and Measured Waste Moisture Content in the Control Cells 67
Figure 4-28 Calculated and Measured Waste Moisture Content in the Retrofit Cells 68
Figure 4-29 Calculated and Measured Waste Moisture Content in As-Built Cell A 68
Figure 4-30 Calculated and Measured Waste Moisture Content in As-Built Cell B 69
Figure 5-1 Monthly Average of Gas Composition for Control Cell A 72
Figure 5-2 Monthly Average of Gas Composition for Control CellB 73
Figure 5-3 Monthly Average of Gas Composition for As-Built Cell A 74
Figure 5-4 Monthly Average Gas Composition for As-Built CellB 75
Figure 5-5 Monthly Average of Gas Composition for Retrofit Cell A 76
Figure 5-6 Monthly Average of Gas Composition for Retrofit Cell B 76
Figure 5-7 Cumulative Methane Gas Collection Data versus LandGEM-Predicted Gas Generation for the
Control Cells 79
Figure 5-8 Cumulative Methane Gas Collection Data versus LandGEM-Predicted Gas Generation for As-
Built Cell A 80
Figure 5-9 Cumulative Methane Gas Collection Data versus LandGEM-Predicted Gas Generation for As-
Built CellB 80
Figure 5-10 LFG Generation Rate Prediction for As-Built Cell A and B 81
xvi
-------
Figure 5-11 Cumulative Methane Gas Collection Data versus LandGEM-Predicted Gas Generation for the
Retrofit Cells 82
Figure 5-12 Monthly Average LFG Flow Rate for Retrofit Cells 83
Figure 5-13 Modeled LFG Generation Rate for Waste in the Retrofit Cells 84
Figure 5-14 Normalized Comparison of Gas Collection in Control and Retrofit Cells 85
Figure 5-15 Normalized Comparison of Gas Collection in Control and As-Built Cells 86
Figure 5-16 Normalized Comparison of Gas Collection in Control cells and As-Built Cells 87
Figure 5-17 NMOC Production in the Control Cells 89
Figure 5-18 NMOC Production in As-Built Cell A 90
Figure 5-19 NMOC Production in As-Built Cell B 90
Figure 5-20 NMOC Production in the Retrofit Cells 91
Figure 6-1 Leachate Temperature as a Function of Time in the Control Cells 96
Figure 6-2 Leachate Temperature as a Function of Time in the Retrofit Cells 96
Figure 6-3 Leachate Temperature as a Function of Time in As-Built Cells 97
Figure 6-4 Leachate pH as a Function of Time in the Control Cells 98
Figure 6-5 Leachate pH as a Function of Time in the Retrofit Cells 98
Figure 6-6 Leachate pH as a Function of Time in the As-Built Cells 99
Figure 6-7 Volatile Organic Acids as a Function of Time in Control CellB 100
Figure 6-8 Volatile Organic Acid as a Function of Time in the As-Built Cells 100
Figure 6-9 Total Organic Carbon as a Function of Time in the Control Cells 101
Figure 6-10 Total Organic Carbon as a Function of Time in the Retrofit Cells 102
Figure 6-11 Total Organic Carbon as a Function of Time in the As-Built Cells 102
Figure 6-12 Biochemical Oxygen Demand as a Function of Time in the Control Cells 104
Figure 6-13 Chemical Oxygen Demand as aFunction of Time in the Control Cells 104
Figure 6-14 BOD/COD Ratio as a Function of Time in the Control Cells 105
Figure 6-15 Biochemical Oxygen Demand as a Function of Time in the Retrofit Cells 105
Figure 6-16 Chemical Oxygen Demand as a Function of Time in the Retrofit Cells 106
Figure 6-17 BOD/COD Ratio as a Function of Time in the Retrofit Cells 106
xvii
-------
Figure 6-18 Biochemical Oxygen Demand as a Function of Time in the As-Built Cells 107
Figure 6-19 Chemical Oxygen Demand as a Function of Time in the As-Built Cells 107
Figure 6-20 BOD/COD Ratio as a Function of Time in the As-Built Cells 108
Figure 6-21 Total Kjeldahl Nitrogen as a Function of Time in the Control Cells 109
Figure 6-22 Ammonia as Nitrogen as a Function of Time in the Control Cells 110
Figure 6-23 Total Kjeldahl Nitrogen as a Function of Time in the Retrofit Cells 110
Figure 6-24 Ammonia as Nitrogen as a Function of Time in the Retrofit Cells Ill
Figure 6-25 Total Kjeldahl Nitrogen as a Function of Time in the As-Built Cells Ill
Figure 6-26 Ammonia as Nitrogen as a Function of Time in the As-Built Cells 112
Figure 6-27 Total Iron as a Function of Time in the Control Cells 113
Figure 6-28 Total Iron as a Function of Time in the Retrofit Cells 114
Figure 6-29 Total Iron as a Function of Time in the As-Built Cells 114
Figure 6-30 Ortho-Phosphate as a Function of Time in the Control Cells 116
Figure 6-31 Total Phosphorous as a Function of Time in the Control Cells 116
Figure 6-32 Ortho-Phosphate as a Function of Time in the Retrofit Cells 117
Figure 6-33 Total Phosphorous as a Function of Time in the Retrofit Cells 117
Figure 6-34 Ortho-Phosphate as a Function of Time in the As-Built Cells 118
Figure 6-35 Total Phosphorous as a Function of Time in the As-Built Cells 118
Figure 6-36 Chloride Concentrations as a Function of Time in the Control Cells 120
Figure 6-37 Chloride Concentrations as a Function of Time in the Retrofit Cells 120
Figure 6-38 Chloride Concentrations as a Function of Time in the As-Built Cells 121
xvin
-------
List of Tables
Table 2-1 Mass Balance Monitoring Parameters and Frequencies 9
Table 2-2 Liquid Addition Monitoring Parameters 10
Table 2-3 Tier 1 Landfill Bioreactor Leachate Monitoring Parameters and Frequency 11
Table 2-4 Tier 2 Landfill Bioreactor Leachate Monitoring Parameters and Frequency 11
Table 2-5 Organic Composition of Fresh Residential Refuse (% of Dry Weight) 16
Table 2-6 Chemical Composition of Paper Products Present in Municipal Waste 16
Table 2-7 Landfill Bioreactor Solids Monitoring Parameters 17
Table 3-1 Chemical Properties of Industrial Liquids 23
Table 3-2 Leachate Sampling Parameters and Schedule 31
Table 3-3 Volatile Organic Compounds Examined in Landfill Leachate 31
Table 3-4 Semi Volatile Organic Compounds Examined in Landfill Leachate 32
Table 3-5 Municipal Solid Waste Sampling Schedule 33
Table 3-6 LFG Sampling Schedule 35
Table 3-7 HAPs Analyzed in Quarterly LFG Sampling 35
Table 3-8 Moisture Balance Parameters 38
Table 3-9 Area Distribution of As-Built cells 38
Table 3-10 Area Distribution of Retrofit and Control Units 38
Table 4-1 Summary of Waste Composition Data for the Retrofit cell 40
Table 4-2 Slopes and Their 95% Confidence Intervals for Linear Regressions 42
Table 4-3 Summary of Waste Composition Data for the Control cells and Assessment of Cell Replication45
Table 4-4 Summary of Waste Composition Data for the As-Built cells and Assessment of Cell Replication
49
Table 4-5 Summary of Calculated and Field Measured Waste Moisture Content 67
Table 5-1 Summary Statistics for Gas Composition for Control Cells AandB 72
Table 5-2 Summary Statistics of Gas Composition for As-Built Cells A and B 73
Table 5-3 Notable Air Injection Dates for As-Built Cells AandB 74
xix
-------
Table 5-4 Summary Statistics of Gas Composition for Retrofit Cells 75
Table 5-5 Summary of Waste Mass Inputs for LandGEM, Mg (tons) 78
Table 5-6 Optimized Landfill Decay Constant (k) for Control Cell 79
Table 5-7 Optimized Landfill Decay Constant (k) for As-Built Bioreactor Cells 81
Table 5-8 Summary of Dates Used for Retrofit and Control Cells CH4 Flow Comparison 85
Table 5-9 Summary of Dates Used for As-Built and Control Cells CH4Flow Comparison 85
Table 5-10 NMOC Concentration for Each Sub Cell (ppm as Hexane) 88
Table 5-11 LandGEM-Predicted Mass Production Rate of NMOCs in Different Landfill Units (Mg/yr). 92
Table 5-12 LandGEM Model Input Values for NMOC Side-by-Side Analysis 93
Table 5-13 LandGEM Predicted Mass Production Rate of NMOCs Emissions (Mg/yr) 93
Table 8-1 OLLB Objectives Assessment 127
Table 8-2 Evidence of Accelerated Waste Decomposition Across All Media Analyzed Between Control
and Bioreactor Landfill Cells 130
xx
-------
List of Acronyms
ANOVA Analysis of Variance
BMP Biochemical Methane Potential
BOD Biological (or Biochemical) Oxygen Demand
C&D Construction and Demolition
CAA Clean Air Act
CE Collection Efficiency
CFR Code of Federal Regulations
CH:L Cellulose plus Hemicellulose to Lignin
COD Chemical Oxygen Demand
CQA Construction Quality Assurance
CRADA Cooperative Research And Development Agreement
DOC Dissolved Organic Carbon
FML Flexible Membrane Liner
GEM Gas Emissions Model
GPS Global Positioning System
HAP Hazardous Air Pollutant
HOPE High Density Polyethylene
HELP Hydrologic Evaluation of Landfill Performance
ID Inner Diameter
LCS Leachate Collection System
LFG Landfill Gas
MACT Maximum Achievable Control Technology
MSL Mean Sea Level
MSW Municipal Solid Waste
NCSU North Carolina State University
NESHAP National Emission Standards for Hazardous Air Pollutants
NMOC Non-Methane Organic Compound
NRMRL National Risk Management Research Laboratory
NSPS New Source Performance Standards
OLLB Outer Loop Landfill Bioreactor
OLRDF Outer Loop Recycling and Disposal Facility
OP-FTIR Open Path Fourier Transform Infrared
ORD Office of Research and Development
ORP Oxidation Reduction Potential
PCF Pounds Per Cubic Foot
PSF Pounds Per Square Foot
QAPP Quality Assurance Project Plan
RCRA Resource Conservation and Recovery Act
RD&D Research, Development, and Demonstration
RPD Relative Percent Deviation
SBR Sequential Batch Reactor
SDR Standard Dimension Ratio
SM Standard Methods
SVOC Semi-Volatile Organic Compound
TKN Total Kjeldahl Nitrogen
TOC Total Organic Carbon
xxi
-------
List of Acronyms (continued)
TVA Total Volatile Acid
UCL Upper Confidence Level
U.S. EPA United States Environmental Protection Agency
VFA Volatile Fatty Acid
VOA Volatile Organic Acid
VOC Volatile Organic Compound
VS Volatile Solid
WM Waste Management, Incorporated
XL excellence and Leadership
xxii
-------
Chapter 1. Introduction
1.1 Landfill Bioreactor Technology
In 2005, more than 230 million tons of municipal solid wastes (MSW) were generated in the United States
with 57 percent of that waste managed via disposal in MSW landfills (U.S. EPA 2006a). The majority of
these landfills are permitted under Subtitle D of the Resource Conservation and Recovery Act (RCRA) and
are managed as "dry tombs" where explicit controls are implemented to minimize liquid introduction and
infiltration into the solid waste mass. Since the 1970s, the potential benefits of moisture addition, or
leachate recirculation, into MSW landfills have been examined by the U.S. EPA as well as many other
researchers. These potential benefits include the rapid decomposition of degradable organics, the rapid
generation of landfill gas (LFG), and the stabilization (e.g., low concentration of organics, metals, etc.) of
landfill leachate. Since the 1990s, the coupled effects of controlled introduction of leachate and
groundwater (and sometimes air) as well as LFG extraction have been incorporated into the concept of
what is known as the "landfill bioreactor".
The concept is particularly appealing since under the current Subtitle D regulations landfill owners are
financially responsible for the environmental care and management of their landfills for a minimum of 30
years after closure. If a landfill bioreactor can be demonstrated to consistently enhance/accelerate waste
degradation and reduce to long term risk associated with the site to an acceptable level, then reduced post
closure care may be considered for such sites. Although the technology related to landfill bioreactors has
been investigated since the 1970s, full-scale implementation and the corresponding performance
monitoring results have been limited. Under a Cooperative Research and Development Agreement
(CRADA) between the U.S. EPA and Waste Management, Inc. (WM), long-term performance monitoring
of MSW bioreactor landfill cells was initiated at the Outer Loop Landfill in Louisville, Kentucky. In
addition to confirming/refuting the results of previous laboratory and pilot-scale studies, the Outer Loop
Landfill Bioreactor (OLLB) study also focused on defining the obstacles and limitations to the full-scale
implementation of the landfill bioreactor technology concepts at working MSW landfills.
1.2 Outer Loop Landfill Bioreactor
The study presented in this document was conducted at the Outer Loop Recycling and Disposal Facility,
(OLRDF), which is located in Louisville, Jefferson County, Kentucky. The site has a total property area of
approximately 316 hectares (780 acres) and is located on the north side of Outer Loop Road, immediately
west of Interstate 65. The OLDRF is owned and operated by WM and has been used for waste disposal for
approximately 35 years. The OLRDF comprises eight individual and separate landfill units, designated
Units 1 through 8. Units 1, 2, 3, and 6 are inactive landfill units that are not currently receiving waste.
Unit 4 is permitted as a construction and demolition (C&D) debris landfill that is currently active, and Unit
8 is a newly permitted active bioreactor landfill cell (not included in this report). This study focuses on
portions of Unit 5 and Unit 7, which are both permitted Subtitle D landfill units. During the study, Unit 5
was inactive, while Unit 7 actively received waste.
The bioreactor demonstrations at the OLLB represent large-scale research efforts at a full-scale operational
landfill. The study covers approximately 20.2 hectares (50 acres) in three types of lined landfill units. The
first was a Subtitle D landfill that was retrofitted with moisture addition piping to allow the recirculation of
liquids (Unit 5). The second type of units in this study was a landfill that had a piping network (for liquids
and air addition) installed as waste was being placed (Unit 7.4). The third unit was a Control cell that was
developed and filled as a typical Subtitle D landfill without any intent of supplemental liquid addition (Unit
7.3). Each of the three units was divided into sub cells to provide a quasi-"duplicate" of each test cell for
data quality purposes.
1.3 Project Objectives
The overall objectives for the OLLB project, previously presented in the First Interim Report, are as
follows:
-------
• Extensive monitoring of bioreactor landfill cells to understand the physical, chemical, and
biological activities and changes over time within the landfill bioreactors, with particular emphasis
given to characteristics of in-place solid waste, leachate, LFG, as well as waste settlement;
• Compare and contrast measured information with that of a conventional Subtitle D landfill to
evaluate differences caused by the bioreactor landfill treatments;
• Incorporate statistical techniques to assess the effectiveness and protectiveness of the landfill
bioreactor operational technique;
• Establish best practices and procedures required to operate landfill bioreactors;
• Establish the important and indicative parameters that should be monitored with respect to landfill
bioreactor operations (i.e., critical and non-critical measurements); and
• Obtain sufficient research data to enable improvements that may be applied to future bioreactor
landfills, both in an experimental capacity and as an alternative design and management method
for MSW landfills.
1.4 Report Organization
This report provides a summary and interpretation of the monitoring results from 2001 to the second
quarter of 2006 for the OLLB project. This report is a follow-up to the September 2003 U.S. EPA
document Landfills as Bioreactors: Research at the Outer Loop Landfill, Louisville, Kentucky, First
Interim Report, which presented a complete description of the landfill study sites, the distinction between
the study units, and the data collection procedures. Although this document is a follow-up to the earlier
report, this Second Interim Report is intended to be a stand-alone document. The document is organized as
follows:
• Chapter 1. Provides a brief background related to landfill bioreactor technology and introduction
to the OLLB project,
• Chapter 2. Summarizes previous findings regarding landfill bioreactor technology. The findings
presented were usually obtained at laboratory-or pilot-scale level and served as the springboard for
the OLLB project,
• Chapter 3. Provides a description of the various test cells at the OLLB site and explains the
rationale for selecting the specific cells for study,
• Chapter 4. Focuses on solids decomposition and moisture balance calculations,
• Chapter 5. Focuses on landfill gas generation as well as non-methane organic carbon emissions,
• Chapter 6. Focuses on leachate quality and how they relate to waste decomposition,
• Chapter 7. Provides an evaluation of several performance criteria at the site as well as a discussion
of results compared to regulatory thresholds for MSW landfills,
• Chapter 8. Presents conclusions from the study, as well as recommendations regarding ongoing
and future monitoring activities at the site,
• Chapter 9. Provides a list of the cited references from the report,
• Appendix A. Provides the data validation report generated by Neptune and Company, Inc.,
• Appendix B. Provides supplemental figures related to MSW solids analysis,
• Appendix C. Presents statistical analyses of measured waste moisture content,
• Appendix D. Presents statistical analyses of measured versus calculated waste moisture content,
• Appendix E. Presents summary statistics for the hazardous air pollutant (HAP) analysis; and
• Appendix F. Presents a statistical analysis of the leachate monitoring results.
-------
Chapter 2. Overview of Bioreactor Landfills
This chapter presents a brief overview of previous research and studies related to the bioreactor landfill
concept. Many of the topics covered in this section were presented and evaluated as part of an earlier U.S.
EPA document titled "Monitoring Parameters for Landfill Bioreactors" (U.S. EPA 2004). That document
outlines monitoring at MSW bioreactor landfills based on the OLLB experience. The main purpose of that
document was to provide a regulatory and research-based rationale for the various parameters that were
monitored during the project. The main reason for reintroducing those parameters is to provide a basis for
the re-evaluation of the effectiveness of these parameters that are presented in later chapters.
2.1 Introduction to Landfill Bioreactors
More than 230 million tons of MSW is generated in the United States annually with 57 percent of that
waste disposed of in MSW landfills (U.S. EPA 2006a). Most of the MSW landfills currently in service are
permitted under Subtitle D of the RCRA or Subtitle D regulations (U.S. EPA 2006a). The main purpose of
MSW landfill regulations (just like other environmental regulations), is to minimize risk to human health
and the environment. In the case of MSW landfills, this was accomplished by reducing possible
contamination from the migration of leachate and LFG from landfill sites. Thus, Subtitle D landfill
regulations outline a system that minimizes liquid infiltration into the solid waste mass by controlling the
amount of moisture allowed into these landfills. Because the Subtitle D "dry tomb" landfill design
incorporates features to minimize the potential for the introduction of liquids into the waste, the resulting
waste mass is often maintained at relatively low moisture content. As a result, the conventional Subtitle D
"dry tomb" landfill design does not promote the solid waste decomposition process. Therefore, the risk
associated with environmental emissions from dry tomb MSW landfills may exist longer, justifying the
need for long-term monitoring at MSW landfill sites. Current regulations require leachate and LFG
emissions to be monitored for at least 30 years after closure of a landfill site or as long as environmental
risk are present. Numerous small-scale and large-scale projects have demonstrated that the rate of solid
waste decomposition at MSW landfills can be improved by increasing the moisture content of the waste,
thereby potentially reducing the duration of the required post-closure care requirements. Increasing the
moisture content of MSW by may provide the necessary liquids to accelerate waste decomposition. In
addition, leachate recirculation is an economical means for leachate management at landfill sites (Pohland
1975).
Initially, most of the research examining the effects of moisture addition on solid waste degradation
concentrated on leachate recirculation as a means of economically managing MSW landfill leachate. In a
conventional dry tomb landfill, landfill leachate is mainly generated by rain water percolation through the
solid waste. As a result, the volume of the leachate generated depends largely upon the climate, the type of
waste present in the landfill, the landfill morphology, the landfill surface conditions, and the types of
operations at the facility (Reinhart and Townsend 1998). Research has demonstrated that enhanced
degradation of MSW is possible use leachate recirculation (Pohland et al. 1993; Townsend et al. 1996;
Reinhart and Townsend 1998). Enhanced degradation in a landfill where liquids are recirculated compared
to a conventional landfill, is also characterized by enhanced LFG production. When operated in a fashion
that includes controlled introduction of liquids, the landfill appears to operate like a "biological reactor" or
"bioreactor", as the recirculated liquids increase the moisture content of the waste and enhance the
distribution of nutrients and bacteria, buffer the pH, and dilute inhibitory compounds (Reinhart and
Townsend 1998; Kim and Pohland 2003). Recirculation of leachate also reduces the need and cost
associated with the collection and subsequent removal of leachate from the landfill to some other on-site or
off-site location for treatment through conventional biological and/or physical-chemical processes.
Research has also shown that liquids recirculated in landfill bioreactors tend to act as a medium for
nutrients, provide microbial transport, and enhance the establishment of anaerobes. As a result of the
increase in moisture content, a more rapid rate of solid waste decomposition is achieved when compared to
conventional MSW landfills. The enhanced rate of decomposition, facilitated by the increase in moisture
content, often leads to an increase in the LFG generation rate with a corresponding reduction in
biochemical oxygen demand (BOD), chemical oxygen demand (COD), total organic carbon (TOC), total
-------
volatile acids (TVA), and metals content in the leachate (Pohland et al. 1993; Reinhart and Townsend
1998; Sponza and Agdag 2004). The enhanced rate of landfill decomposition typically results in an
increase in the rate of landfill settlement (El-Fadel 1998; Hossain et al. 2003), which may provide
additional airspace. As a result, the total amount of solid waste placed in landfill bioreactors may be more
than the amount placed in a dry tomb landfill of similar size. Enhancing the rate of LFG production may
also improve the economics associated with gas-to-energy facilities at landfill bioreactor sites (Barlaz et al.
1990; Mehta et al. 2002). As a result of these combined effects, landfill bioreactors may reduce the long-
term environmental impacts associated with the disposal of MSW in landfills.
2.2 Regulatory Overview
Solid waste regulations were established fairly recently relative to other environmental regulations. Most
MSW landfills in the U.S. operate under regulations identified in Subtitle D of RCRA. The specific
regulatory criteria for MSW landfills are presented in 40 CFR Part 258, which was promulgated in 1991.
The main goal of these regulations is to ensure protection of human health and the environment through the
establishment of minimum national criteria for MSW landfills. Subtitle D regulation specifies performance
standards for the location, design, groundwater monitoring, and corrective actions for MSW landfills. In
the preamble to the Subtitle D regulations, the U.S. EPA recognized that landfills are, in effect, biological
systems that require moisture for decomposition to occur. The U.S. EPA further acknowledged that the
increase in moisture content of solid waste landfills may provide specific benefits, which may include
increasing the rate of waste stabilization, improving leachate quality, and increasing LFG production for
potential energy recovery. At the time Subtitle D regulations were promulgated, however, the U.S. EPA
believed that many landfills, particularly those in humid areas, already had sufficient liquid for
decomposition. Therefore, the conventional opinion was that the intentional addition of liquids was
unnecessary. Furthermore, it was hypothesized that intentional liquid introduction may result in
operational problems, including: (i) an increase in leachate production; (ii) clogging of the leachate
collection system; (iii) buildup of hydraulic head within the landfill; (iv) an increase in LFG emissions and
odor problems; and (v) an increase in the potential for the leachate to be released as a pollutant due to
leachate breakouts and/or run-off. These operational problems, should they occur, would likely result in
adverse impacts on human health and the environment (Ikem et al. 2002).
Subtitle D regulations included provisions to prohibit the addition of liquids to MSW landfills. Specifically
40 CFR 258.27 states "bulk or non-containerized bulk liquid waste may not be placed in MSW landfill
units." Subtitle D regulations, however, allow for recirculation of leachate and gas condensate generated
from the gas recovery process as long as the landfill design includes certain liner requirements. Because of
the lack of data on the performance of landfill bottom liners, the U.S. EPA initially required (at a
minimum) that MSW landfills that recirculate leachate and gas condensate use single composite liners to
contain leachate and prevent it from contaminating the underlying soil and groundwater. MSW landfill
liner systems typically consist of high density polyethylene (HDPE) geomembranes, geosynthetic clay
materials, or compacted clay (Foose et al. 2002). In addition to requiring composite liners, the regulations
require that a demonstration be provided to assure that the added volume of liquid will not increase the
liquid (i.e., hydraulic) head on the liner to more than the allowable 30 cm (1 ft). Because of the self-
implementing nature of the regulations, the permitting of leachate recirculation systems for MSW landfills
was delegated to the approved state regulatory agency or tribal community. Thus, regulations regarding
MSW landfills tend to vary between states, which allow various degrees of flexibility within the law.
The U.S. EPA provided limited regulatory flexibility to allow landfill bioreactors because of promising
results from research regarding landfill bioreactors. Most notable was the introduction by U.S. EPA in
1995 of Project XL (excellence and Leadership). For the case of MSW landfills, Project XL allowed for
leachate, as well as other industrial liquids, to be added to Subtitle D landfills that do not meet the
composite liner criterion. In turn, the designers of the Project XL landfill bioreactors hoped that the
leachate recirculation/LFG recovery requirements would enhance groundwater protection and provide for
additional capacity to accommodate more waste at individual landfills, thus extending the life of existing
landfill cells. Unfortunately, because of difficulty in obtaining a Project XL landfill bioreactor permit, only
five landfills were permitted under this effort nationwide before the U.S. EPA ceased accepting Project XL
proposals in 2003.
-------
In 2003, the U.S. EPA defined landfill bioreactors as MSW landfills that utilize liquids other than leachate
and gas condensate to achieve an average moisture content of more than 40 percent on a wet weight basis.
In its final rule on National Emissions Standards for Hazardous Air Pollutants (NESHAP) for landfills and
the Maximum Achievable Control Technology (MACT) regulations, the U.S. EPA required landfill
bioreactors with a total disposal capacity equal to or greater than 2.5 x 106 Mg (or m3) to include a system
to actively collect and control LFG that will commence operation within 180 days after liquids addition or
after the average landfill moisture content reaches 40 percent, whichever occurs later (FR 2003).
Realizing the potential benefits of adding moisture to reduce the long-term risk associated with MSW
landfills, the U.S. EPA promulgated the Research, Development, and Demonstration Permits for Municipal
Solid Waste Landfills (RD&D rule) in 2004 (FR 2004). Under this rule, the U.S. EPA allowed approved
states to permit landfill bioreactors under Subtitle D regulations. The RD&D rule deviates from the
Subtitle D regulations in that it allows approved states to waive some provisions of the Subtitle D landfill
operating criteria (excluding hazardous waste prohibition and explosive gas control), design criteria, and
final cover criteria. RD&D permits, as well as the rule itself, are temporary modifications to the original
Subtitle D regulations. The RD&D permit for a given MSW landfill is initially approved for three years,
with optional three year renewals for a maximum of 12 years. Landfill bioreactors permitted under the
RD&D rule are required to submit annual reports that summarize data obtained during each year and assess
progress towards ultimate solids stabilization.
Approved states may issue landfill permits under the RD&D rule to allow the addition of non-hazardous
industrial liquids to MSW landfills that use conventional composite liners or alternative liner systems. The
landfill owner must also demonstrate to the appropriate State Director that the MSW landfill that is
designed and operated as a landfill bioreactor under the RD&D rule does not pose an additional risk to
human health or the environment beyond what can result from its operation as a dry tomb Subtitle D MSW
landfill. Just like the Subtitle D regulations, the RD&D rule is self-implementing, giving each state or
tribal community the authority to permit landfill bioreactors that may otherwise not have been allowed
previously under Subtitle D regulations.
There are insufficient data on the behavior of landfills that use industrial liquids as a source of moisture.
Since the RD&D rule is temporary (i.e., 12 years), the U.S. EPA Office of Research and Development
(ORD) is actively gathering data that may help in supporting many of the concepts envisioned by this rule.
The data gathered by ORD will not only concentrate on the performance of the containment system (e.g.,
leachate and LFG collection, liners, slope stability, etc.) but also on the microbial stabilization within the
landfill unit. Data are currently being collected from four Project XL landfills and the two MSW landfills
that have CRADAs with the U.S. EPA. More information regarding the other Project XL landfills can be
found on the webpage http://www.epa.gov/projectxl.
2.3 Anaerobic Decomposition Fundamentals
Waste in an MSW landfill does not have a single age because waste is placed incrementally in the various
cells throughout the life of the facility. Rather, waste of different ages is associated with the various cells
within the landfill and their respective "stabilization" stage or phase (Pohland et al. 1993). As can be seen
in Figure 2-1, the different MSW landfill stabilization phases often overlap and can be viewed collectively,
which tends to limit the industry's understanding of the various phases and their interaction. As shown in
Figure 2-1, the initial phase results in aerobic decomposition followed by four stages of anaerobic
decomposition. Thus, the majority of MSW landfill decomposition occurs under anaerobic conditions. It
is noted that virtually all MSW landfills undergo these five stages of stabilization and that operating an
MSW landfill as a landfill bioreactor has an effect only on the rate and not the sequence (and potentially the
duration) of the stabilization phases (Pohland and Al-Yousfi 1994; Reinhart and Townsend 1998; Kim and
Pohland 2003). Thus, it is important to understand each of the
-------
PH
2
T3
O
1,0
0,8
0-6
0,4
0.2
0.0
o
Q.
O
O
-13000
2000
1000
O)
40 -
20
600
200 400
Stabilization Time, Days
Figure 2-1 Phases of Solid Waste Decomposition (source: Pohland et al. 1993)
stabilization phases individually. After a short-duration aerobic phase, it is generally recognized that there
are four more steps involved in the anaerobic solid waste degradation and stabilization process, with each
step involving a set of separate and distinct microbial populations. These steps/phases are shown in Figure
2-1 and further described below. Successful conversion and stabilization of the waste is dependent on
microorganisms performing their respective functions in syntrophic relationships. The application of these
phases to an MSW landfill setting is briefly discussed below (Pohland 1975; Pohland et al. 1993).
2.3.1
Phase I (Initial Adjustment)
This phase is sometimes referred to as the lag phase. As the waste is placed in the landfill, the void spaces
contain oxygen (O2). With compaction, the O2 content of the landfilled solid waste gradually decreases.
As moisture becomes available and the microbial population density increases, biochemical decomposition
under aerobic conditions is initiated.
2.3.2
Phase II (Transition)
The transition phase is relatively short-lived as the O2 is rapidly consumed by the bacteria present, resulting
in a transition from aerobic to anaerobic conditions. During this phase, the primary electron acceptors
become nitrates and sulfates, rather than O2, with the displacement of O2 by carbon dioxide (CO2) in the
effluent gas.
2.3.3
Phase III (Acid Formation)
This phase is marked by the onset of the hydrolysis of the biodegradable fraction of the solid waste, leading
to a rapid increase in the concentration of volatile fatty acids (VFAs) in the leachate. This also corresponds
to a decrease in pH from approximately 7.5 to 5.6 (Pohland et al. 1993). During this phase, the
decomposition intermediates such as VFAs contribute to a high COD and the long-chain volatile organic
acids (VOAs) are converted to acetic acid (C2H4O2), CO2, and hydrogen (H2).
The presence of high levels of VFAs in the leachate will increase both BOD and VOAs, leading to the
onset of H2 production by fermentative bacteria and H2-oxidizing bacteria. The H2 generation phase is
comparatively short-lived as it terminates by the end of this phase. This phase is also marked by an
increase in the biomass of acidogenic bacteria as well as a rapid consumption of substrates and nutrients.
-------
The drop in pH may cause concomitant mobilization and the possible complexation of metal species that
are more soluble at a low pH.
2.3.4 Phase IV (Methane Fermentation)
Intermediary products appearing during the acid formation phase (i.e., mainly acetic, propionic, and butyric
acids) are converted to methane (CH4) and CO2 by methanogens. As a result of the consumption of VFAs
by methanogens, the pH moves back to neutrality. The organic strength of the leachate (as characterized by
high BOD) is dramatically decreased in correspondence with increases in gas (i.e., CH4 and CO2)
production. This phase also signifies the longest overall time duration and represents the period when the
majority of the waste decomposes.
2.3.5 Phase V (Final Maturation and Stabilization)
The final stage of solid waste decomposition is characterized by a lower rate of biological activity due to
the limiting nutrients such as phosphorus. During this stage, landfill CH4 production is almost negligible.
O2 and oxidized species may slowly reappear as O2 permeates from the atmosphere with a corresponding
increase in oxidation-reduction potential (ORP) in the leachate. It is hypothesized that residual organic
materials may slowly be converted to gas in this phase, with the possible production of humic-like
substances. To date, the U.S. EPA has no documentation of an MSW landfill exhibiting such
characteristics.
2.4 Key Monitoring Parameters for MSW Landfill Bioreactors
The landfill bioreactor RD&D rule allows for the controlled introduction of liquids into Subtitle D MSW
landfills to accelerate the decomposition of biodegradable organics. To understand the decomposition
process, it is important to understand the type and role of microorganisms that contribute to this process.
The anaerobic waste degradation process requires at least four different groups of microorganisms (Parkin
and Owen 1986; Pohland et al. 1993). These microorganisms occur naturally in MSW but require different
conditions to achieve optimal performance. The kinetics of microorganisms in landfill bioreactors have not
been widely investigated, likely because these groups of microorganisms (and other anaerobes in general)
are harder to culture compared to aerobes. In fact, it is now recognized that the majority of the
microorganisms in environmental systems cannot be easily cultured (Amman et al. 1995; Hugenholtz et al.
1998; Jjemba 2004). As a result, there is a need to use molecular-based non-culture techniques to study the
microbial populations in MSW landfills as solids decomposition progresses. Therefore, instead of directly
studying the "cause" or source of the decomposition, researchers are often forced to indirectly study waste
stabilization by monitoring the "effect" of the decomposition process. As a demonstration of this, Lay, et
al. (1998), examined the abundance and activity of methanogens in simulated landfill columns. In that
study, both the cumulative CH4 production and moisture content were higher in the columns that included
recirculated leachate. Recirculation of leachate also shortened the initial lag phase. The specific microbes
responsible for these mechanisms (i.e., the cause) were not thoroughly investigated, but the results of the
microbial activity (i.e., the effect) were demonstrative.
In addition to the biological parameters that can be monitored, there are some physical and chemical
parameters that can be monitored that collectively confirm the optimal operation of landfill bioreactors.
The remainder of this section presents results from previous studies that focus on these parameters. It
should be noted that the optimal ranges presented in this review may not apply to all landfills as the
composition of waste varies greatly. Therefore, in response to the heterogeneous nature of MSW, waste
stream variability, and differences in environmental and climate conditions, the use of "Control cells" (i.e.,
conventional dry tomb cells) at operating landfill bioreactor facilities is encouraged to help understand the
site-specific performance of landfill bioreactors compared to the conventional landfill units. Unfortunately,
the literature does not include numerous example projects where Control cells were systematically
constructed and operated; while Control cells were integral to the study at the OLRDF. For research
purposes, it is recommended that landfill cells used as a control be comparable to the landfill bioreactor
cells in age, depth, and composition of waste so that specific differences in monitoring parameters may be
evaluated more effectively. They should be monitored separately whenever possible to demonstrate
specific impacts of landfill bioreactor operations on the volume and quality of leachate, the volume and
-------
composition of LFG generated, changes in the waste mass, and effects on the leachate collection system.
While it is believed that the use of Control cells associated with landfill bioreactor are beneficial and that
use of the Control cells will likely help in the assessment of site-specific data, it is understood that the use
of Control cells may not be feasible in all cases.
2.4.1 Physical Monitoring Parameters
2.4.1.1 Geotechnical Considerations
Slope stability is an important parameter in MSW landfill design, particularly with regard to landfill
bioreactors. Often landfill slope stability focuses on the stability of the final landfill configuration.
Operating landfills as bioreactors increases the importance of a slope stability assessment as part of the
overall design, not only for the final configuration but also for when the landfill is at interim grades. Large
changes in liquid levels in the landfill impact the development of pore pressures, which can influence slope
stability. Pore pressures may increase in landfill bioreactors because of the addition of liquids and the
concurrent increase in the rate of LFG production. However, there are little well documented field data
available to substantiate this possibility.
Several approaches for predicting and periodically assessing stability in landfills have been proposed
(Bachus et al. 2004). Stability calculations can be conducted using the strength properties of the waste and
foundation soils, the geometry of the waste mass, and the pore pressures within the waste and foundation
soils. Unit weight, shear strength and frictional characteristics of MSW vary widely because of differences
in waste characteristics and compaction techniques. The unit weight of MSW was reported by Landva and
Clark (1990) to range from 320 to 1580 kg/m3 (20 to 99 pcf). As expected, the shear strength of the waste
and the calculated stability of the landfill depend on the composition of the waste. MSW may have an
internal friction angle of 1 degree with cohesion as large as 2200 psf to a friction angle as high as 36
degrees with no cohesion (Singh and Murphy 1990). Kavazanjian et al. (1995) reviewed the technical
literature and recommended design shear strength for waste that is represented as a bilinear Mohr-Coulomb
failure envelope exhibiting a friction angle of 33 degrees with minimum shear strength of 550 psf. Large-
scale laboratory direct shear tests conducted on "conventional" and bioreactor landfill wastes indicate
similar strength parameters. Leachate levels in the waste and liquid in the subsurface also affect slope
stability.
In conjunction with a slope stability study, it is recommended that operators follow simple guidelines to
promote landfill bioreactor slope integrity. Operators should avoid any excavation at the toe of the slope
that can create local zones of high stress that may potentially lead to instability. Operators should also
avoid filling waste in cells at steep grades [i.e., greater than 3 horizontal to 1 vertical (3H:1V)]. The
placement of fill for on-site roads and a component of the cover may also lead to a translational or veneer
instability, and should be considered during design (Stark et al. 2000).
As required under Subtitle D regulations, leachate head on a landfill bottom liner should not exceed 30 cm
(1 ft). The addition of moisture into the landfill may cause excess amounts of leachate to build up on the
bottom liner if the liquids are not effectively removed. Before the addition of moisture into a landfill
bioreactor, therefore, the ability of the leachate collection system (LCS) to effectively minimize the
potential impacts of increased leachate flow must be explicitly considered. Performance of the LCS and
resistance to clogging should also be examined during design and monitored during operation. Potential
clogging of the LCS may lead to a buildup of leachate within the drainage layer, causing the head on the
liner to exceed 30 cm (1 ft). As described previously, this may increase the potential for slope instability.
Alkalinity, hardness, iron, and manganese compounds, total organic carbon, COD, and BOD are all
involved in reactions that can result in buildup of precipitates that could potentially lead to LCS clogging
(Fleming et al. 1999; Rowe et al. 2000; Cook et al. 2001; Rittmann et al. 2003). Monitoring the
concentration of these parameters may provide an indication as to whether leachate concentrations
approach saturation levels for calcium carbonate and other compounds that contribute to clogging and poor
LCS efficiency. LCS clogging may also be caused by the settling of suspended particles from the leachate
and biological growth on or in the LCS (Koerner and Koerner 1995).
-------
2.4.1.2
Mass Balance
It is important that the mass of the landfilled solid waste within each landfill bioreactor cell be identified
and recorded. Conducting regular surveys of in-place volume and total disposed mass can be helpful in
estimating the density of solid waste placed and may provide insight into waste decomposition. As the
solid waste decomposes, the density of the landfill tends to increase (Tiquia et al. 2002). Parameters that
may assist with mass loading calculations are presented in Table 2-1. With regard to the total disposed
mass, it is recognized that soil used as daily cover needs to be considered in any mass calculation. It has
been recommended to avoid daily cover materials with a low permeability to minimize perched leachate
zones within the landfill (U.S. EPA 2002). Efforts may be made during site operations to remove
temporary cover, temporary roadways, and piles of soils used for daily cover prior to the placement of
waste in a particular part of the cell. Additionally, using materials other than soil (e.g., temporary tarps,
foams, etc.) as daily cover may be an effective alternative in preventing perched leachate zones.
Table 2-1 Mass Balance Monitoring Parameters and Frequencies
Parameter
Visual Landfill Inspection
Mass of Landfilled MSW
Mass of Landfilled Construction and
Demolition Waste
Mass of Soil (other than daily cover)
Type of Daily Cover
Mass or Volume of Daily Cover
Landfill volume
Settlement
Frequency
Daily
Daily
Daily
Daily
Daily
Daily
Quarterly
Quarterly
Units
-
Mg (tons)
Mg (tons)
Mg (tons)
Mg (tons) or m3 (yd3)
m3 (yd3)
m(ft)
2.4.1.3
Moisture Balance
A key component in operating a landfill as a bioreactor is the introduction of moisture from internal (i.e.,
leachate) and external (e.g., precipitation, stormwater, groundwater, and industrial liquid waste streams)
sources into the landfill. In general, the decomposition and select stabilization rate of biodegradable solid
wastes increases with increasing moisture content of the waste (El-Fadel 1998; Olayinka 2003). Research
has shown that the optimum moisture content for biological degradation is greater than 40 percent (Pohland
et al. 1993; Reinhart and Townsend 1998). Deliberate moisture addition to the landfill should be applied
uniformly to the extent practicable to evenly wet the waste and to reduce differential settlement. Typical
moisture addition techniques include applying liquids to the surface and to the subsurface. Surface
applications include spray irrigation and surface ponds. Subsurface techniques include horizontal trenches
and vertical wells, in addition to other methods such as permeable blankets, which have been described
elsewhere (Haydar and Khire 2006). Limited research has been performed at full-scale landfills regarding
design guidelines for subsurface application techniques; however, hydraulic properties of waste were
estimated using leachate recirculation via vertical wells (Jain et al. 2006). Similarly, hydraulic properties
while using horizontal injection have been reported (Townsend and Miller 1998, Haydar and Khire 2005).
As described previously, an increase in landfill moisture content that results in a build-up of pore pressure
may decrease slope stability. Assuming the volume of water consumed to be negligible during waste
hydrolysis, moisture balance can be calculated as presented in the following equation:
AS = Moisturem - Leachateout
Equation 2-1
Where:
AS = net moisture storage;
Leachateout = leachate generated by the landfill; and
Moisture,n = liquids added into the landfill, including precipitation.
As mentioned previously, liquids addition occurs in various forms including, but not limited to, the
introduction of leachate, stormwater, liquid waste, and municipal and industrial wastewater. Parameters
that assist in performing a water balance are presented in Table 2-2. These should be monitored daily when
-------
liquids are being introduced into the waste. It is also important to take into account the moisture content of
the incoming waste when using the moisture content of the waste to quantify or to assess moisture balance.
Table 2-2 Liquid Addition Monitoring Parameters
Parameter
Volume of Leachate Added
Rainfall
Volume Outside Liquid Added (e.g.,
Groundwater, Industrial Wastewater)
Volume of Leachate Generated
Mass of Sludge Addition
Wet Basis Moisture Content of Sludge
Added
Frequency
Daily
Daily
Daily
Daily
Daily
Daily
Units
L (gal)
mm (inch)
L (gal)
L (gal) of leachate generated by the
landfill bioreactor cells only
Mg (tons)
Percent (M/M)
2.4.2 Analytical Monitoring Parameters
2.4.2.1
Leachate Monitoring
Leachate monitoring parameters presented in this section may be used to enhance the operational control of
MSW landfill bioreactors. These parameters are categorized as either "Tier 1" or "Tier 2", shown in Tables
2-3 and 2-4, respectively. Tier 1 parameters are relatively inexpensive to obtain. Tier 2 parameters are
usually more time intensive to assess and, as a result, the testing is incrementally more expensive than
testing for Tier 1 parameters. The extent and frequency of leachate monitoring at each site will ultimately
be dependent on local, state, and federal regulations. It is, however, recognized that the value of some of
the individual parameters can change significantly over the life of the landfill bioreactor cell. It is noted
that the constituents listed in Table 2-3 and 2-4 are simply suggested monitoring parameters, and the
specific leachate monitoring needs and requirements of a bioreactor landfill should be evaluated on a site-
specific basis. Because of their importance in understanding the decomposition process and therefore
attaining a properly functioning landfill bioreactor, some of these monitoring parameters are discussed in
detail below.
2.4.2.1.1 Leachate Temperature
Research suggests that anaerobic processes are optimized when the waste is within either the mesophilic
(30 to 38 °C (86 to 100 °F)) or thermophilic (50 to 60 °C (122 to 140 °F)) temperature range (Parkin and
Owen 1986). The higher thermophilic temperatures enhance the rate at which organic matter is converted
to VOAs but lead to a lower yield of CH4 compared to the lower mesophilic temperatures (Pohland et al.
1993). This trend is possibly attributed to an increase in the activity of acetogens and a decrease in the
activity of methanogens. Optimum CH4 generation from solid wastes, however, was found to occur at
41°C (105.8 °F) (Hartz et al. 1982). It is recognized that operating temperatures in the landfill at the
thermophillic range may present concerns regarding fire, health and safety issues, and sustained vitality of
the microorganisms. The maintenance of a uniform leachate temperature is believed to be a fundamental
monitoring parameter that is indicative of an efficient anaerobic stabilization process.
Since landfill temperature is not typically controlled by the operator, the temperature ranges stated above
reflect a combination of effects due to ambient temperature conditions, microbial activity, and the extent
and effectiveness of insulation provided by the specific landfill configuration. In a study conducted at the
Outer Loop Landfill in Louisville, Kentucky, the temperatures of the leachate were initially approximately
7° C (45 °F) but steadily increased to 30 °C (86 °F) or higher within a few months after operating the
landfill as a bioreactor (U.S. EPA 2003). While an increase in leachate temperature may be reflective of
waste degradation in a landfill, it is not solely indicative of biological activity. Leachate temperature may
also be affected by ambient temperature as well, depending on the leachate sampling location.
10
-------
Table 2-3 Tier 1 Landfill Bioreactor Leachate Monitoring Parameters and Frequency
Parameter
Static head on Liner
Temperature
pH
Conductance (laSm/cm)
Total Dissolved Solids (mg/L)
Alkalinity (mg/L as CaCO3)
Anions (mg/L)
Cations (mg/L)
Chemical Oxygen Demand (mg/L)
Biochemical Oxygen Demand (mg/L)
Total Organic Carbon (mg/L)
Total Phosphorous (mg/L)
Ortho Phosphate (mg/L)
Ammonia (mg/L)
Nitrite (mg/L)
Nitrate (mg/L)
Method
Pressure Transducer
Thermometer
U.S. EPA(2) 9045C
Field Electrode
SM(3) 160.1 (C)
SM(3)310.1
SM(3) 300.1
SM(3)410.4
SM(3) 405.1
EPA(2) 9060
SM(3) 365.2 (C)
SM(3) 365.2 (C)
SM(3)350.1(C)
SM(3) 300.1
SM(3) 300.1
Frequency
(i)
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Notes:
1. Head on the liner should be monitored continuously, however, it is suggested that a weekly average is reported.
2. U.S. EPA SW-846 Test Methods for Evaluating Solid Wastes.
3. U.S. EPA Methods for Chemical Analysis of Water and Wastes.
Table 2-4 Tier 2 Landfill Bioreactor Leachate Monitoring Parameters and Frequency
Parameter
Volatile Organic Compounds0 '
(VOCs) (ng/L)
Semi- Volatile Organic
Compounds (SVOCs) (|ag/L)
Volatile Fatty Acids
(mg/L)
Arsenic (mg/L)
Barium (mg/L)
Cadmium (mg/L)
Calcium (mg/L)
Copper (mg/L)
Chromium (mg/L)
Iron (mg/L)
Lead (mg/L)
Magnesium (mg/L)
Mercury (|J.g/L)
Potassium (mg/L)
Sodium
Selenium (mg/L)
Silver (mg/L)
Zinc (mg/L)
Method
SW-846 8260 (B)
SW-846 8270 (B)
GCMS
SW-846 6010(prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 7470 (prepared per SW-846 3005)
SW-846 7470 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
SW-846 6010 (prepared per SW-846 3005)
Frequency
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Note: 1. Constituents listed in 40 CFR 258 Appendix I.
11
-------
2.4.2.1.2 LeachatepH
The optimum pH for anaerobic systems ranges between 6.8 and 7.4 (Parkin and Owen 1986). Initially, the
leachate pH may be neutral; but after the onset of anaerobic conditions there is generally a noticeable drop
in the pH especially during the acid forming phase, as discussed previously. The drop in pH is caused by
the accumulation of VFAs in the leachate. However, the pH will eventually increase to neutral conditions
as methanogens consume these acids. Studies conducted to compare leachate pH in conventional dry tomb
MSW landfills and landfill bioreactors have shown that there may not be a significant difference in pH
between the two systems. As described previously, the pH is expected to vary with time in all landfills,
depending on the initial waste composition and the phase of waste degradation in the areas where the
leachate passes before it is collected in the LCS.
2.4.2.1.3 Volatile Fatty Acids and Volatile Organic Acids
VFAs and VOAs affect microorganisms and the degradation processes in two primary ways. First, they
have a low ionization constant (i.e., low pKa) and can readily dissociate, releasing H+ ions that cause the
pH of the system to decrease and therefore become destabilized. Second, when the acids are non-
dissociated (as is typical at low pH levels), the acids are able to penetrate microbial cell membranes,
establishing a pH gradient by actively transporting protons out of the cell and reducing the internal cell pH
(Zoetemeyer et al. 1982; Aguilar et al. 1995). The decrease in intracellular pH in turn leads to an increased
energy demand by the cell to restore pH levels leaving less energy for growth (Yamaguchi et al. 1989;
Gonzalez et al. 2005). These processes lead to reduction in the rate of solid waste degradation. VOA
concentrations that are in excess of 6,000 mg/L can inhibit microbial processes (Pohland et al. 1993).
However, most research regarding solid waste degradation has not focused on VOAs, but rather has
investigated the effect of VFAs on the menthanogenic population within the landfill.
Acidic leachate typically correlates with a high VFA content and a low CH4 production for prolonged
periods. Most common among these VFAs are acetic acid, propionic acid, and butyric acid (Barlaz et al.
1989; Kim and Pohland 2003). The amount of leachate that is recirculated affects the quantity of VFAs. If
VFAs are high, methanogenesis can be inhibited by the low pH that is induced. Therefore, the volume of
recirculated leachate has to be properly adjusted to minimize a buildup of VFAs. Lab-scale study results
presented in Figure 2-2 demonstrate this. Figure 2-2 shows the effects of leachate recirculation on VFA
accumulation in comparison to VFA generation under conventional landfill management. The leachate in
"Reactor 9" was recirculated at 9 L/day (2.4 gpd) which was 13 percent of the reactor volume, whereas
leachate in "Reactor 21" was recirculated at 21 L/day (5.5 gpd, 30 percent of the reactor volume). It is
interesting to note that VFA buildup in the reactor with a higher leachate recirculation rate of 21 L/day (5.5
gal/day) was nearly as high as the VFAs generated in the single-pass (i.e., conventional) reactor.
Furthermore, at 21 L/day (5.5 gpd), the bioreactor had a spike of almost 30,000 mg VFA/L within 30 days
which can be detrimental to methanogens. This is apparent when comparing the CH4 production in Reactor
21 to Reactor 9 (Figure 2-2). A further demonstration of the affect on degradation is that waste settlement
was greater in Reactor 9 than Reactor 21 (Sponza and Agdag 2004). Settlement may be related to CH4
production in anaerobic landfill systems as it is an indication of mass loss and waste degradation. At high
leachate volumes, saturation, washout of the methanogens and/or ponding may occur in the reactor, thus
contributing to the noted detrimental effects of the acidic condition. At this time, models that adequately
predict the correct amount of leachate to recirculate have not been developed and determinations of what
recirculation volume is adequate (or optimal) are generally made on a case-by-case assessment of
performance.
2.4.2.1.4 Leachate Biochemical and Chemical Oxygen Demand
BOD consists of biologically degradable dissolved organics in the leachate. COD is a measure of
chemically oxidizable components in leachate and reflects the amount of O2 that is required by the bacteria
to metabolize the existing organic substrate as well as the O2 required by other oxidizable chemical
compounds. One of the main consequences of operating a landfill bioreactor is the rapid reduction of BOD
in the leachate. BOD values reported in the literature for conventional landfills ranged between 20 and
152,000 mg/L (Krung and Ham 1991; Chu et al. 1994; Kjeldsen et al. 2002; U.S. EPA 2003), whereas
COD values range between 500 and 60,000 mg/L (Pohland et al. 1993; Reinhart and Townsend 1998; U.S.
EPA 2003). By comparison, BOD values for landfill bioreactors were found to range between 20 and
28,000 mg/L (Pohland et al. 1993; Reinhart and Townsend 1998; U.S. EPA 2003).
12
-------
—;K — Single pass reactor
- R@attof9 —&— Reactor21
.2 20000 -
10000
5000 -
so
100
150
200
250
Operation time (day)
Figure 2-2 Changes in VFA in a Single Pass (Conventional) Reactor versus Bioreactor9 and
Bioreactor21 over Time (Source: Sponza and Agdag, 2004)
While the values of BOD and COD may change individually as a function of decomposition, the ratio of
BOD to COD may be used to assess the relative biodegradability of the leachate substrate. Variations in
BOD and COD may be closely related to the variations observed with VFA production. As a result, the
BOD/COD ratio may act as an indicator of the biodegradability of organics present in MSW. At the time
of initial waste placement within the landfill, the BOD and COD concentrations were relatively low.
The initial low BOD and COD concentration is thought to be caused by the initial aerobic stabilization of
the MSW or by a delay in the hydrolysis of the waste. During the acid formation phase, the majority of the
O2 demand (both BOD and COD) is caused by the presence of high concentrations of VFAs. BOD and
COD concentrations decrease after the onset of the CH4 fermentation phase and the conversion of VFAs.
Landfill bioreactors have a higher BOD/COD ratio during the acid forming phase relative to conventional
landfills (Reinhart and Al-Yousfi 1996; Reinhart and Townsend 1998). However, research suggests that
this ratio decreases during the phase that follows (i.e., during the CH4 fermentation phase (Phase IV)).
After waste stabilization, COD may be influenced by high molecular weight organics such as humics and
fulvics present in the leachate (Pohland et al. 1993). These residuals tend to elevate COD to a higher level
than BOD and reduce the BOD/COD ratio. For instance, leachate BOD/COD ratios are usually higher than
0.5 for acid formation phases of decomposition but may decline to less than 0.1 for well-decomposed
waste. It should be noted that COD is also influenced by an increase in the concentration of ammonia
which has implications that are outlined later in this document.
2.4.2.1.5 Leachate Total Organic Carbon
Similar to the phenomenon observed for COD and BOD, the TOC levels increase after initial waste
placement as a result of microbial solubilization of the organics. During the acid forming phase, TOC
increases rapidly. An increase in TOC may also be observed soon after the introduction of waste
containing high concentrations of organics. Because of the conversion of the VFAs to CH4, TOC
concentration tends to decrease during the CH4 fermentation phase. TOC of conventional landfills has
been reported to range between 30 and 30,000 mg/L (Pohland and Harper 1987; Krung and Ham 1991;
Pohland et al. 1993; Chu et al. 1994; Kjeldsen et al. 2002). As an indication of the similarity in the
decomposition mechanism, the TOC in landfill bioreactor systems ranges between 70 and 28,000 mg/L
(Pohland et al. 1993).
13
-------
2.4.2.1.6Leachate Nitrogen Content
Nitrogen is mainly present in MSW leachate in the following forms: total Kjeldahl nitrogen (TKN),
ammonia nitrogen (NH4-N), and nitrate nitrogen (NO3-N). Ammonia is the most important of the three
forms of nitrogen since at high concentrations (i.e.,1,500 - 2,500 mg/L) it tends to inhibit methanogens
(Hashimoto 1986; Hansen et al. 1998) and therefore reduces the potential for waste degradation. Under
anaerobic conditions, ammonia tends to accumulate in the leachate, especially when the leachate is
recirculated. While the ammonia concentration in conventional landfills ranges between 2 and 2,200 mg/L
(Krung and Ham 1991; U.S. EPA 2003), concentrations in landfill bioreactors range between 6 and 20,000
mg/L (Krung and Ham 1991; Pohland et al. 1993; Chu et al. 1994; Reinhart and Townsend 1998; U.S. EPA
2003). The accumulation of ammonia in the leachate from landfill bioreactors may adversely affect the
methanogenic population. Ammonia concentrations of 1,500-3,000 mg/L are inhibitory to anaerobic
processes at highpH levels, whereas concentrations above 3,000 mg/L can be toxic to most
microorganisms in the waste (Pohland et al. 1993). Increasingly higher concentrations of ammonia in the
leachate may indicate the potential for adverse effects on the methanogenic population, but these elevated
values may also be an indication of an advanced stage of waste decomposition.
2.4.2.1.7 Leachate Metals and Metalloids Content
Heavy metals and metalloids exert toxicity to microorganisms influencing their biochemical activities, cell
morphology, and growth (Hughes and Poole 1989; Gadd 1992; Jjemba 2004). Unlike organic compounds,
metals do not degrade as the waste in the landfill decomposes, but rather they are transformed from one
chemical state to another. Metals may also be precipitated under anaerobic conditions as carbonates,
hydroxides, or sulfides; they may also bind to organic waste ligands. Metals can also be chelated or
subjected to ion exchange within the landfilled waste matrix.
Most of what is known about metal transformations in landfills is based on chemical rather than biological
analyses. Thus, there is limited information about the role of microorganisms on such transformations
under the anaerobic conditions that are typical of landfills with progressively changing redox potential,
moisture content, temperature, and pH. Beyond the adverse effects of metals in the leachate from a
microorganism perspective, metals concentration in the leachate is an important parameter as it can affect
the cost of off-site leachate treatment (e.g., a wastewater treatment plant may reject leachate if metals
concentration (e.g., arsenic) is too high, thus causing the landfill to send the leachate elsewhere for
treatment). The lower pH and higher organic content of the leachate during the initial landfill stabilization
phases may mobilize some metals during the acid forming phase (Pohland et al. 1993). However, after the
onset of the CH4 fermentation phase, metal concentrations in the leachate tend to decrease. The decrease in
these concentrations is a result of metal reduction, formation of metal sulfides, precipitation, and
complexation within the waste matrix. Operators should recognize that the introduction of large
concentrations of heavy metals, through either solid or liquids placed in the landfill, may retard or inhibit
the solid waste degradation process and should be avoided in landfill bioreactors.
2.4.2.1.8 Semi-Volatile and Volatile Organic Compounds
SVOCs and VOCs may represent parameters that are of particular importance for monitoring as there is a
potential for the introduction of complex organic constituents into the landfill, particularly when various
industrial wastes are applied. The ability of microorganisms to assimilate and transform potentially toxic
organic compounds in landfill bioreactors has been documented. For example, in-situ reductive
dehalogenation of organic compounds (e.g., trichloroethylene and hexachlorobenzene) has been
demonstrated in bench-scale landfill bioreactor studies (Kim and Pohland 2003). However, additional data
are needed to compare and contrast the volatilization potential for specific VOCs and SVOCs that are most
likely to appear in leachate for both conventional and bioreactor landfills.
2.4.2.1.9 Leachate Phosphate Concentration
The carbon to phosphate (C:P) ratio is important as the presence of low phosphorus quantities will slow
down microbial growth and decomposition. A C:P ratio of 60:1 is deemed optimal for microorganisms to
actively assimilate substrate carbon. Reducing the C:P ratio to this range in landfills may require adding
commercial phosphorous-rich fertilizers together with the recirculation leachate. The beneficial effects of
adding phosphates to laboratory-scale landfill bioreactor cells have been documented (Sheridan 2002). To
the authors' knowledge, this practice has not been further examined in the field.
14
-------
2.4.2.2 Solids Monitoring
2.4.2.2.1 Refuse Composition
Cellulose and hemicellulose are the principal biodegradable components of MSW. While the anaerobic
biodegradation of refuse in landfills requires the coordinated activity of several trophic groups of
microorganisms, the conversion of cellulose and hemicellulose to CH4 can be described by the following
reactions:
(C6H10O5)n + nH2O -> 3nCO2 + 3nCH4 Equation 2-2
(C5H8O4)n + nH2O -> 2.5nCO2 + 2.5nCH4 Equation 2-3
The other major organic compound in refuse is lignin, which is, at best, only slowly degradable under the
anaerobic conditions that prevail in landfills (Colberg 1988). In addition to its recalcitrance, lignin
interferes with the decomposition of cellulose and hemicellulose by physically impeding microbial access
to the more easily degradable carbohydrates. Thus, the complete conversion of cellulose and hemicellulose
to LFG is not expected. Research to date from both field-scale and laboratory-scale systems has not
established a lower level of cellulose and hemicellulose biodegradation as many components of waste
contain cellulose, hemicellulose, and lignin and each waste component has unique biodegradation
characteristics (Barlaz 2006).
The cellulose, hemicellulose, and lignin concentrations for a series of waste samples collected over the past
14 years are summarized in Table 2-5. Some of the samples represent fresh residential refuse while others
were collected at a landfill. The samples collected at a landfill may not be comparable to the residential
waste samples that do not include the large quantities of office paper or wood waste that are typically
generated in the commercial and construction sectors, respectively. The relatively large range in cellulose,
hemicellulose, and lignin concentrations is indicative of their variability in refuse composition.
Unfortunately there is not a single factor to which this variability can be attributed. The last row in Table
2-5 represents values for the aggregate composition of waste entering a Canadian landfill. These values
were calculated from measurement of the cellulose, hemicellulose, and lignin content of individual waste
components and a waste composition survey. The composition of the aggregated waste stream includes
some inert wastes that dilute the cellulose, hemicellulose, and lignin concentrations.
The precise composition of waste that enters a landfill is never known and can vary over time and with the
landfill location (e.g., proximity to urban, residential, or rural waste streams). Nonetheless, it is generally
agreed that the major sources of organic matter in landfills are paper, food waste, and yard waste. Of
course, the disposal of yard waste in landfills is banned in many states, thus decreasing its importance as a
source of CH4 production. The U.S. EPA published an estimate of the composition of MSW based on the
materials flow methodology (U.S. EPA 2006a). Using these data, coupled with data on the fraction of each
MSW component that is recycled, it is estimated that paper makes up approximately 26.3 percent of the
mass of MSW entering landfills. In that other waste streams (i.e., those that do not contain food waste,
paper, and yard waste) are typically buried in landfills, the actual paper content in a landfill is lower than
26.3 percent. This is consistent with the lower cellulose content of waste entering the Canadian landfill as
presented in Table 2-5.
Available data on the cellulose, hemicellulose, and lignin concentrations in fresh paper and wood waste are
summarized in Table 2-6. It is noted that a relatively large difference in the cellulose content exists for the
two office papers that were tested. The office paper described in Wu et al. (2001) had an ash content of
11.6 percent compared to a 1.4 percent ash content in the sample obtained by Eleazer et al. (1997). This
difference is likely due to the presence of inorganic coatings in the sample obtained by Wu et al. (2001).
15
-------
Table 2-5 Organic Composition of Fresh Residential Refuse (% of Dry Weight)
Reference
Ham and Bookter, 1982
Jones etal, 1983(1)
Bookter and Ham,
1982(2)
Barlaz et al, 1989(3)
Eleazer et al. 1997
Rhew and Barlaz, 1995
Ressetal., 1998
Price et al., 2003
Barlaz unpublished(4)
Cellulose
(C)
42.4
25.6
63.4
51.2
28.8
38.5
48.2
43.9
22.4
Hemicellulose
(H)
nm
6.6
nm
11.9
9
8.7
10.6
10
5.8
Lignin
(L)
10.9
7.2
15.7
15.2
23.1
28
14.5
25.1
11
(C+H)/
L
3.89
4.5
4.04
4.15
1.64
1.68
4.06
2.15
2.57
VS.
nm(5)
59.6
nm
78.6
75.2
nm
71.4
nm
nm
Source(6)
L
L
L
R
R
R
R
R
L
1. Refuse recovered from a landfill in the UK. The type of refuse and sampling strategy were not
specified. The samples also contained 2.3% starch and 5% protein. Analyses were conducted using
detergent fiber techniques.
2. Average of ten samples from the working face of a NY landfill.
3. The following additional analyses were performed on this sample: protein - 4.2%, soluble sugars -
0.35%, starch - 0.6% and pectin - <3%.
4. These values represent the composition of waste entering a Canadian landfill as described in the text.
5. nm = not measured.
6. (L) = landfill and (R) = residential refuse.
Table 2-6 Chemical Composition of Paper Products Present in Municipal Waste
Source
Reference
Cellulose
Hemicellulose
Lignin
Volatile solids
Newsprint
Wu et al.
2001
48.3
18.1
22.1
98.0
Eleazer et
al. 1997
48.5
9
23.9
98.5
Office Paper
Wu et al.
2001
64.7
13.0
0.93
88.4
Eleazer et
al. 1997
87.4
8.4
2.3
98.6
Corrugated
Cardboard
Eleazer et al.
1997
57.3
9.9
20.8
92.2
Coated
Paper
Eleazer
etal.
1997
42.3
9.4
15.0
74.3
Branches
Eleazer et
al. 1997
35.4
18.4
32.6
96.6
In summary, although the exact composition of waste that is buried in landfills varies, it is well established
that the waste contains large concentrations of cellulose-containing materials. Paper is the major
contributor to the cellulose and hemicellulose concentrations in landfills, followed by contributions from
wood, yard debris and food waste. Therefore, regardless of initial concentrations of cellulose,
hemicellulose, and lignin, it may be a useful measure of decomposition to track the relative changes in
cellulose, hemicellulose, and lignin during the life of the facility.
2.4.2.2.2 Solids Monitoring Parameters
In the previous sections, discussions were presented regarding the physical and analytical measurements
that can be made to demonstrate waste decomposition. Ultimately, it is the degradation of the waste solids
themselves that is the true measure of decomposition. The values of the parameters presented in this
section can vary widely due to heterogeneity of the waste, variations in the degree of decomposition, and
moisture content. Unlike leachate and LFG sampling (where "weighted average values" may be obtained),
it is physically difficult and expensive to obtain representative individual samples of waste for analysis.
Table 2-7 presents suggested solids monitoring parameters for landfill bioreactors. More data from other
bioreactor landfills regarding solids monitoring are needed; monitoring of the parameters shown in Table 2-
7 (with the exception of moisture content) may not be necessary during routine operation.
16
-------
Table 2-7 Landfill Bio reactor Solids Monitoring Parameters
Parameter
Average Temperature
Average pH (2)
Average Volatile Solids
(percent M/M)(3)
Average Wet-Based Moisture
Content (percent M/M)
Method
Thermometer
U.S. EPA
9045C
U.S. EPA 1684
-
Frequency
(i)
(i)
(i)
(i)
Optimum Range
35 - 55 °C
6.5-7.6
Decreasing Trend
< 45 %
Notes:
1. Frequency of solids monitoring should be determined on a site- and project-specific basis.
2. U.S. EPA, 2003.
3. Mehta etal., 2002.
2.4.2.2.3 Temperature
Temperature of the waste mass may be determined by thermocouples or thermistors within the waste mass
and monitored electronically (via a data logging station) or by using portable meters. Temperature
monitoring in bioreactor landfills operating aerobically is essential, as the regulation of temperature is
critical to preventing elevated waste temperatures resulting from aerobic waste decomposition.
2.4.2.2.4 Volatile Solids
Adding moisture to the MSW stimulates biological activity in landfill bioreactors. This increase in
biological activity may directly translate into an increase in the degradation of cellulose and hemicellulose,
which may translate into a measurable increase in the rate of waste settlement. A three-year study at the
Yolo County Landfill in California demonstrated that the abundance of cellulose, hemicellulose, and lignin
are strongly correlated to the volatile solids (VS) content of MSW (Mehta et al. 2002). VS content in
MSW is expected to decrease as the refuse decomposes because of cellulose and hemicellulose content loss
from the waste. The main disadvantage of using VS as an indicator, however, is that, unlike cellulose,
hemicellulose, and lignin, the analysis of VS offers a lower level of accuracy and is affected by daily cover
application (Mehta et al. 2002).
2.4.2.2.5 Moisture Content
Moisture content of MSW should be examined to ensure relatively uniform distribution of the liquids that
are added to the landfill bioreactor. The moisture content of the "fresh" incoming solid waste needs to be
evaluated to calculate moisture addition requirements. Moisture addition requirements (i.e., volumes) will
likely dictate the liquids addition rate required for landfill bioreactors.
2.4.2.3 Landfill Gas Monitoring Parameters
During the process of anaerobic solid waste decomposition, landfills generate significant quantities of CH4
and CO2. Both of these gases are undesirable greenhouse gases (Conrad 1995; Jjemba 2004), with CH4
having a global warming potential approximately 20 times that of CO2 (IPCC 2001). Controlling and
monitoring the emissions of these gases is an essential element of any controlled landfill operation.
Estimates of gas production rates in bioreactor landfills relative to conventional (i.e., Subtitle D) landfills
vary, but previous investigations have indicated rates of landfill gas production at landfill bioreactors
between two (Reinhart and Townsend 1998) and seven (U.S. EPA 2005b) times higher than that at
conventional landfills. As a result, it is critical to design gas collection systems at landfills operating as
bioreactors to capture the expected additional LFG. The rate and quantity of LFG generated is dependent
on the mass and composition of waste as well as the moisture conditions within the landfill. Equation 2-4
describes the relationship of the predicted CH4 generation over time with regard to the composition of the
waste, disposed waste mass, and time.
17
-------
QCH4 = L0Il(e~kc - e~kt) Equation 2-4
Where:
QcH4 = methane generation rate at time t (m3/yr);
L0 = methane generation potential of the waste (m3 CH^/Mg waste);
R = average annual refuse acceptance rate during the active life of the landfill (Mg/yr);
k = methane generation rate constant (yr"1);
c = time since landfill closure (yr) and
t = time of initial waste acceptance (yr), respectively.
As will be discussed later in chapter 5, LFG generally consists of approximately 50 percent CH4 and 50
percent CO2by volume, so to calculate the total LFG generated at a specific time, Equation 2-4 should be
multiplied by 2.
Of particular interest when assessing LFG generation with regard to landfill bioreactors are the parameters
L0 and k. The capacity to generate CH4 generally depends on the organic content and moisture content of
the waste (U.S. EPA 1998). As described previously, operation as a landfill bioreactorby increasing
moisture content of the waste may accelerate the decomposition of the waste and therefore increase the
value of k. These parameters can be estimated on a site-specific basis and implemented in the Landfill Gas
Emissions Model Version 3.02 (LandGEM), developed by U.S. EPA, to predict LFG generation based on
waste placement data and kinetic parameters. Since landfills operated as bioreactors are expected to
generate LFG at a greater rate than dry tomb landfills, gas recovery systems at landfill bioreactors need to
be designed to handle larger flow rates than those at conventional landfills.
The discussion thus far has mainly been on the anaerobic reactions that occur in landfill bioreactors. It is
recognized that landfill bioreactors can be operated as aerobic units by the controlled introduction of air
into the waste after placement. Potential issues regarding air addition in bioreactor landfills include
hydraulic limitations as well as subsurface fires. Jain et al. (2005) found that leachate recirculation had a
significant impact on the air permeability of MSW. One of the regulatory concerns associated with aerobic
landfill bioreactors has been the potential for subsurface fires from spontaneous combustion. Literature
suggests that carbon monoxide (CO) production may be an indicator of subsurface fires (U.S. EPA 1998).
The U.S. EPA's Compilation of Air Pollutant Emission Factors (AP-42) cites a default CO concentration of
142 ppm for MSW landfills; however, it is noted that this value was based on data collected from a limited
number of sites, and the data were classified with a qualifier of "poor" (U.S. EPA 1998). A study at a
landfill undergoing leachate recirculation and air addition in Florida found CO concentrations between 0-30
ppm in areas operating under anaerobic conditions, and CO concentrations up to 1,200 ppm in areas
operating under aerobic conditions (Powell et al. 2006). However, areas showing higher CO concentrations
did not have elevated temperatures indicative of a subsurface fire. A combination of CO and temperature
monitoring is likely the most effective approach in evaluating the presence of subsurface fire.
MSW landfills with an actual or design capacity equal to or greater than 2.5 million Mg of waste (or 2.5
million m3) are subject to New Source Performance Standards (NSPS) for air emissions, which require the
collection of LFG and 95% destruction (by weight) of non-methane organic compounds (NMOCs), a class
of compounds that depletes stratospheric ozone, contributes to global warming, and includes hazardous air
pollutants (HAPs). Currently, specific regulations regarding anaerobic landfill bioreactor LFG collection
requirements are found in 40 CFR Part 63, Subpart AAAA. As a result of these requirements, landfill
bioreactor facilities must comply with LFG collection requirements sooner than dry tomb landfills.
Because of limited data available for LFG emissions at aerobic landfill bioreactors, these systems are
regulated in a similar manner to dry tomb landfills under NSPS. The compounds included in the
monitoring program and the monitoring frequency may be determined on a site-specific basis and guided
by local, state, and federal regulations pertaining to the site or by the goals of the landfill bioreactor
owner/operator.
18
-------
Chapter 3. Site Description and Analytical Methods
3.1 Site Description and System Design
The OLRDF is located in Louisville, Jefferson County, Kentucky. The site, shown in Figure 3-1, is located
on the north side of Outer Loop Road, immediately west of Interstate 65 with an approximate area of 316
hectares (780 acres). The site comprises eight landfill units, designated Units 1 through 8. Units 1, 2, 3,
and 6 are inactive. Unit 4 is an active, permitted C&D debris landfill. Unit 8 is an active bioreactor landfill
unit that is located to the north of Unit 7 and overlays parts of Units 7 and 4 (not shown in Figure 3-1).
Waste placement in Unit 8 just started and will not be discussed in this report. The focus of this study is on
portions of Unit 5 and Unit 7, which are both Subtitle D landfill cells; Unit 5 did not receive any waste
during the study, while portions of Unit 7 received waste until 2005, as described later. The landfill
bioreactor permit approval for this site was received from the Commonwealth of Kentucky, Kentucky
Natural Resources and Environmental Protection Cabinet, Department for Environmental Protection in
2001 under Permit No. 056-00028.
The OLDRF is owned and operated by WM and has been used for waste disposal for approximately 35
years. The site is situated within the alluvial valley of the Ohio River; approximately nine miles southwest
of River Mile 614. The area is generally flat with elevations averaging 139 m (456 ft) above mean sea
level (MSL). The region is effectively enclosed by topographically elevated areas on the west, east, and
south. Elevations range up to 229 m (750 ft) above MSL in areas surrounding the site. Topography and
stream development in the area have been modified by construction and development activities of the
region. Because of the flat topography, the clayey nature of the soil, and the relatively low elevation, the
area is naturally poorly drained. To enhance surface drainage for the development of the region, several
engineered drainage channels have been constructed in the vicinity of the landfill. The channels drain
toward the southwest, eventually discharging into the Ohio River. Some of the cells on the site are unlined
and the entire site utilizes pumps to provide an inward gradient to groundwater flow into the site.
The average yearly regional temperature is 13° C (55° F), ranging from -30 to 40° C (-21 to 103° F).
Average annual precipitation is approximately 113 cm (44 in.) of rainfall, plus approximately 38 cm (15
in.) of snow (U.S. Department of Commerce 2004). The average number of days with precipitation is 125
annually, with 47 days being associated with thunderstorms. The prevailing wind in the area is generally
from the south. Details regarding the meteorological conditions at the site, with particular focus on
precipitation as it relates to the moisture balance analysis at the site, will be presented in later sections.
In contrast to some of the landfill bioreactor research described in Chapter 2, the demonstrations at the
OLLB are large-scale research efforts at a full-scale operational landfill. The study covers a total of
approximately 20.2 hectares (50 acres) in lined landfill units. In this study, three types of landfill units
were studied. The first unit was a control that was developed and filled as a conventional Subtitle D
landfill without any intent of supplemental liquid addition (Control cells). The second type of unit in this
study was a landfill that had a piping network (for liquids and air addition) installed as waste was being
placed (As-Built landfill bioreactor cells). The third was a Subtitle D landfill that was retrofitted with a
moisture addition piping network to allow the recirculation of liquids (Retrofit landfill bioreactor cells).
Each of the three units was divided into subcells to provide a quasi-"duplicate" of the test cell, which will
be discussed in more detail in subsequent sections.
3.1.1 Control Landfill Cells
3.1.1.1 General
Unit 7 is located in the western portion of the OLDRF as shown on Figure 3 -1. Portions of that Unit,
designated 7.3 A and 7.3B, have been designated as the Control cells for the study. These cells have been
operated as conventional Subtitle D (i.e., dry tomb) landfills since initial waste placement began in 1998.
19
-------
Easement Lines
Bioreactor Research & Development Project(s)
Sedimentation Basins
ITS
Contained
reactor Landfill
COMPOST
AREA
UNIT 4
Permitted Construction and
Demolition Debris Landfill
Permitted Contained Landfill
Figure 3-1 Outer Loop Landfill Site Map
The cells were monitored and sampled in a similar manner to the landfill bioreactor units. The Control
cells consist of two paired landfill cells (i.e., Control cell A and Control cell B) that are not hydraulically
separated. However, a barrier layer was installed between the Control and As-Built units. The layer
consists of low permeability clay along with an additional layer of permeable tire chips.
The Control cells were selected to represent waste decomposition in a conventional Subtitle D landfill
without liquids addition and with standard vertical LFG extraction wells. Waste placement in the Control
cells commenced in 1998 and ceased in 2004. At the start of the landfill bioreactor project in 2001, solid
waste in the Control cells was approximately three years old. Prior to the start of the project in 2001,
approximately 342,000 Mg (380,000 tons) of solid waste was present in the Control cells. Additional waste
was placed in Control cell A and Control cell B during the project to bring the cells to final grade.
3.1.1.2
Leachate Collection System (LCS)
The Control cells were lined with a smooth 60 mil HDPE geomembrane on the base areas while textured
60 mil HDPE geomembrane was used on the perimeter slopes (2.5H: IV). A 10 oz. non-woven geotextile
cushion layer was placed over the bottom liner and a double-sided geocomposite was placed over the side
slope. A 0.3 m (1 ft) granular leachate drainage blanket covered the base areas, and was constructed with
non-carbonate coarse gravel. In swales in the valleys in the Control cells, an 8 in. standard dimension ratio
(SDR) 11 perforated HDPE pipe was laid on a gravel bed then covered with gravel. This coarse % in. to 1
!/2 in. gravel material had the same basic characteristics as the remainder of the leachate drainage blanket,
except it was a larger grade. Each leachate collection line terminated on the west side of the Control cells
into a sump underlain by the composite liner system. A pumping system was installed at each sump, which
20
-------
provides a means to pump leachate that accumulates. These pumping systems discharge into a force main
that terminates at the site's leachate treatment plant.
3.1.1.2 Gas Collection
As with any "dry tomb" landfill cell, each Control cell contained two vertical gas collection wells installed
in the center of the solid waste mass. LFG collection wells were connected to header lines that connect to
the flare station. The LFG collection headers maintained separate gas collection fields in the Control cell.
The first gas collection header collects landfill gas mainly generated in Control cell A, while the second
collects the gas generated in Control cell B. Before merging together, gas flow in each header line is
measured via an orifice plate installed within the pipe. A probe was also installed in each gas collection
well to allow for LFG sample collection, as will be discussed in Section 3.3.
3.1.2 As-Built Landfill Bioreactor Cells
3.1.2.1 General
The As-Built cells are also located within Unit 7, designated 7.4A and 7.4B, as shown in Figure 3-1. The
cells were designed as a sequential aerobic-anaerobic landfill bioreactor (also referred to as "hybrid"). The
rationale behind the sequential approach was to promote the rapid decomposition of easily degradable
organic matter in the aerobic stage of treatment with the intent of reducing the amount of fermentable
organic matter entering the anaerobic stage. This strategy may shorten the acid-generating phase of
anaerobic waste decomposition and result in a more rapid onset of methanogenesis. Thus, the objective of
these cells was to examine the impact of increased moisture content and waste aeration on the degradation
of recently placed MS W. It is noted that the aim of this research is not to examine the performance of
WM's patented design (U.S. Patent No. 6,283,676 Bl) but rather the effect of liquid addition and sequential
aerobic/anaerobic operation on the solid waste degradation.
The As-Built cells were divided to provide quasi-"duplicates" that are hydraulically separated with a clay
barrier [0.3 m (1 ft)]. A layer of shredded tires was placed onto the clay barrier to act as a conduit for
leachate to reach the bottom liner. The placement of shredded tires was discontinued in the first quarter of
2004 since the layer acted as a conduit for LFG and rainwater migration. A schematic of the configuration
discussed here is presented in Figure 3-2.
The initial waste placement in As-Built cell A and As-Built cell B occurred during June and October of
2001, respectively as presented in Figure 3-3. As of June 2005, approximately 1,461,000 Mg (1,610,000
tons) of solid waste had been placed in the two As-Built cells [i.e., 590,000 Mg (650,000 tons) in As-Built
cell A and 871,000 Mg (960,000 tons) in As-Built cell B]. The cells were constructed in 4.6 m (15 ft)
vertical lifts for a total of seven different lifts. Based on mass, solid waste comprised approximately 71 %
while biosolids from a municipal wastewater treatment plant, C&D debris, and soil accounted for
approximately the remaining 13%, 9.2% and 6.8%, respectively.
3.1.2.2 Leachate Collection System
Since the As-Built cells are a part of Unit 7, the LCS is similar to that of the Control cells as described in
section 3.1.1.2. Each leachate collection line drains into a separate sump underlain by the composite liner
system. A pumping system was installed at each sump, which provides a means to pump leachate from
each sump as it accumulates to a designated depth. As with the Control cells, these pumping systems
discharge into a force main that terminates at the site's leachate treatment plant.
3.1.2.3 Piping Network Installation
The design of the As-Built cells liquid addition, air injection, and LFG extraction piping network utilized
one piping network for delivering liquids and a second piping network for distributing air and extracting
LFG. The sequence and method of placing the pipes is described below.
The first 4.6 m (15 ft) thick lift of waste was placed on top of the leachate collection system, followed by
placement of the first pipe layer (comprised of 10.2 cm (4 in.) ID perforated HDPE) on the top surface of
the first lift. The pipes were placed at approximately 18.5 m (60 ft) wide intervals across the top surface of
21
-------
JUN;OS(2nd-0<«>TOPO
1140 l!« 1)40
EAST-WEST PROFILE
o in so no *o SB so m m m ion n» ran
NORTH-SOUTH PROFILE
UNIT 7
N.T.S.
GRAPHIC SCALE
GRAPHIC SCALE
JNIT 7
• TOPD rBW JUNE 2005
TOPfl FIBW WRLL SOB
• - TQPO FROH DCCENBCff EO&l
TCPO rH» SEPTEMBCB eDD*
TQPO FHM JLNE 20CM
TOPE* FROM MARCH 2KM
—- TDPfl FIRM nCCEHKIt K»3
TQPQ FROM KTDKR £DQ3
TCPg FHW *Uq«HT a»3
TDPfl FRON *WIL 2003
TOPD FROM JANUAIW ^00^
TOPa FROM SCPTCHBER £002
IDPO FROM JWE 3002
Figure 3-2 Solid Waste Placement in the As-Built Landfill Bioreactor Cells
1e+6
8e+5 -
6e+5H
I
4e+5 -
g
O
2e+5 -
As-Built Cell A
As-Built Cell B
3/1/01 9/1/01 3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 3-3 Solid Waste Placement in As-Built Landfill Bioreactor Cells
22
-------
the waste. The length of these pipes was dependent on the length of the individual lift within the cell. The
pipes were perforated with three 1 cm (0.4 in.) diameter holes at 120° intervals around the pipe, with holes
spaced at 0.3 m (1 ft) intervals along the length of the pipe. The pipes were covered with a permeable
media generally consisting of tire chips; however, an alternative design was introduced in the second
quarter of 2004 that involved placing a thin horizontal layer of permeable material on top of each lift to
facilitate liquid recirculation, air addition, and LFG collection. Each end of the perforated piping has a
section of solid pipe of the same diameter that is subsequently connected to a common 10.2 cm (4 in.) ID
manifold pipe. After placement and installation of the first lift and pipe layer, a second lift was placed,
followed by a second piping layer. In this design, the uppermost lift of waste is aerated, while the lift
immediately below this lift receives liquid, and LFG is extracted from all deeper lifts.
3.1.2.4
Air Introduction
Air addition began once a completed lift was placed on top of the piping layer (e.g., the second lift was
aerated by injecting air through the first layer of piping). Each 4.6 m (15 ft) lift of waste received injected
air beginning approximately 30 days after lift placement. This was accomplished by connecting the header
pipe for a particular lift to a high-volume air compressor. Compressed air was continuously pumped into
each new lift at a maximum rate of 57 m3/min (2,000 ft3/min) for a period ranging from 30 to 90 days. The
following three set points were designated as the threshold points for the cessation of air introduction: (i)
waste mass temperature of 71 °C (160 °F); (ii) net change in waste mass temperature of 6.7 °C (12 °F)
within 24 hours; and/or (iii) after 90 days of air introduction. Set points (i) and (ii) were established to
prevent elevated waste temperatures and potential subsurface fire caused by aerobic decomposition. In all
cases, air introduction stopped whenever one of these set points was reached. Until January of 2006, 9.6 x
107 m3 (3.4 x 109 ft3) of air was injected into As-Built cell A while 4.8 x 107 m3 (1.7 x 109 ft3) of air was
injected into As-Built cell B. Based on the total mass of solid waste placed in the cells, As-Built cell A
received almost five times more air per unit mass of waste [162 m3/Mg (6,300 ft3/ton)] compared to As-
Built cell B [32 m3/Mg (1,260 ft3/ton)].
3.1.2.5
Liquids Addition
Liquids addition into the As-Built cells was accomplished through the piping network previously described.
Liquids were gravity-fed through one of four on-site tanks to the lift of waste directly below the lift being
aerated. The addition of liquids was controlled and included intermittent dosing depending on several daily
and seasonal factors including the apparent moisture content of the in-place waste, forecasted precipitation
events, and recent moisture additions.
Liquids added into the As-Built cells included a combination of industrial liquids as well as recirculated
leachate obtained from other lined units at the OLDRF site. The industrial liquids added to the As-Built
cells consisted mainly of beverage waste (75 percent); oily wastewater (10 percent); paint waste (9
percent); ink water (2 percent); and other (4 percent). Volumes and chemical properties of the industrial
liquids are presented in Table 3-1. A summary of the liquids injection history as related to moisture
balance in the As-Built cells is provided in Section 4.2. Figures 3-4 and 3-5 present the volume of
industrial liquids added to As-Built cell A and As-Built cell B, respectively. Figures 3-4 and 3-5 show that
the majority of industrial liquids were added to the As-Built cells after March 2004.
Table 3-1 Chemical Properties of Industrial Liquids
Liquid Type
Beverage waste
Oily wastewater
Paint waste
Ink water
Food waste
Other (septic, municipal,
food, beverage, cleaning)
State
Liquid
Liquid
Liquid & Sludge
Liquid
Liquid & Sludge
Liquid & Sludge
Percent
of Total1
75
10
9
2
1
3
pH
NA
5.7-8.1
5.0-9.1
9.9
4
3.6-7
BOD
(mg/L)
NA
NA
NA
NA
1,800
NA
COD
(g/L)
NA
19-134
2.4-16
38
33
NA
NH3-H
(mg/L)
NA
1-95
1-550
163
NA
NA
1: This represents the approximate percentage of the total industrial liquids added to the As-Built cells.
23
-------
60000 1
50000 -
o
I 40000
3
I
3 30000
3
CT"
'-§ 10000 -
3
1.6e+7
'^
g 1.4e+7 H
13
oo
^ 1.2e+7 -
I
5 l.Oe+7 -
CT- 8.0e+6 -
T3
•o 6.0e+6 -
T3
3
| 2.0e+6 -
O
^^^— Total Added Liquids
— — - Industrial Liquids
60000 1
50000 -
40000 -
o
3 30000 -
Z
I
20000 -
10000 -
3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 3-4 Cumulative Industrial and Total Added Liquids in As-Built Cell A
1.6e+7
Total Added Liquids
Industrial Liquids
3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05
Date
9/1/05
3.1.2.6
Figure 3-5 Cumulative Industrial and Total Added Liquids in As-Built Cell B
Gas Collection
LFG was collected in deeper waste lifts through the pipes previously utilized for air addition. In general, a
series of LFG collection pipes were connected to the active gas collection system once air addition and
liquids addition were no longer occurring in a particular lift. The LFG collection started in April of 2003
for the As-Built cells.
24
-------
3.7.3 Retrofit Landfill Bioreactor Cells
As part of this research, an existing "dry tomb" MSW landfill cell was retrofitted with a piping network to
operate as a bioreactor landfill cell. Apart from the main objective of examining the effect of liquids on
solid waste degradation, research in the cell aimed to assess the effects of nitrate-enriched leachate on
landfill bioreactor performance. Previous research concluded that leachate recirculation was shown to
increase the ammonia concentration within landfill bioreactor cells. Ammonia rich leachate from the
OLRDF was treated in a sequencing batch reactor aerobically to convert ammonia to nitrate. This leachate
containing nitrate was then recirculated in the Retrofit area. As nitrate-containing liquid moves through the
upper sections of the landfill bioreactor cells, denitrifying bacteria convert nitrate to N2 gas, resulting in a
net loss of nitrogen from the landfill.
The Retrofit bioreactor landfill Unit (Unit 5) is located in the northern portion of the OLRDF complex, as
shown in Figure 3-1. The Unit consists of four separate landfill subcells (designated 5.1 A, 5.2A, 5. IB and
5.2B) and is permitted as a Subtitle D cell with a single composite clay liner. Each subcell is equipped with
its own leachate collection line, which allows sampling at each distinct leachate sump. To provide for
hydraulic separation between the test areas and because of geometric similarities, Unit 5.1 A (the most
southern cell, referred to as Retrofit cell A) and Unit 5.2B (the most northern cell, referred to as Retrofit
cell B) were selected for this study. Liquid introduction also occurred in the two middle cells of the
Retrofit unit (5. IB and 5.2A) as will be discussed later. Leachate analysis results presented in Section 4.4
is for the leachate collected from the sump of cell 5.1A and 5.2B (referred in Section 4.4 as Retrofit cell A
and Retrofit cell B, respectively).
Waste placement was initiated in July 1995. Approximately 1,752,000 Mg (1,931,000 tons) of solid waste
were in place by October 1997, of which 32 percent was special waste consisting mainly of contaminated
soils, while the remaining 68 percent was MSW. Between October 2000 and March 2001, an additional
136,000 Mg (150,000 tons) of MSW was placed in the cell to adjust the final top deck elevations as
presented in Figure 3-6. During 2004, daily cover material was stockpiled on the western side of the cell,
as presented in Figure 3-6. Since the cell did not receive new waste during the study, aim thick long-term
clay cover was placed on the Unit. A 5 cm thick layer of cover soil was then placed on top of the clay
cover and the cell was seeded with grass for erosion control.
3.1.3.1 Leachate Collection System
The LCS consists of a collection layer (Figure 3-7), separator geotextile, collection pipe, and cleanout riser.
The collection layer consists of a 0.3 m (1 ft) thick layer of sand on the 10H: IV cell floor and 2H: IV
intercell berms, and geocomposite on the 3H: IV perimeter berms. Perforated pipes used to collect the
leachate were surrounded with coarse aggregate and wrapped by a nonwoven geotextile. The collection
pipes discharge via perforations in the pipe into the Retrofit cell's sumps. Also, collection pipe cleanout
risers connect to the collection pipes in the cell sumps. The pumped discharge from the sumps is connected
into the site's leachate management system via a force main. Underlying the LCS in the Retrofit unit is a
geomembrane that was constructed of 60 mil HOPE that was smooth on both sides.
3.1.3.2 Liquids Addition
Construction of the Retrofit bioreactor cells took place between March and May 2001 and included
installation trenches, moisture distribution and gas collection piping, thermocouples, and ORP probes. To
increase the moisture content of the Retrofit cells, 26 horizontal infiltration galleries were constructed. The
trenches were 4.6 to 6 m (15 to 20 ft) deep, 1 m (3.3 ft) wide and were constructed just below the surface of
the landfill (Figure 3-8). The trenches were spaced approximately 18.3 m (60 ft) apart as presented in
Figure 3-8. Each trench contains two 7.6 cm (3 in.) ID HOPE pipes, one for liquid introduction and the
other for gas collection. Each pipe was perforated with 0.95 cm (0.4 in.) holes every 0.3 m (1 ft) along the
length of the pipe. The trenches were bedded with tire chips and backfilled with permeable material like
shredded tires to allow for liquid and gas flow as illustrated in Figure 3-8. The liquid introduction lines
were then connected through a solid 7.6 cm (3 in.) ID riser pipe to a valve and then to an HOPE liquid
distribution header. Temperature and ORP probes were also installed at the end of each horizontal trench.
25
-------
560
550
540
5301
520;
510|
500°
490 £
480
470
460
450
440
430
0.00 1*00
2001 TOPO
2003 TOPO I\
2003 TOPO :T\
1998 TOPO-, \^i/4
1999TOPO^y££
2000 TOPO
3-00 4-W S-00 MS 7-00 1*00 9-00 10+W 11*00 12*00 13*00 14-00 15*0) 16t« 17-00 18*00 19*00 20-IX
UNIT 5 - EAST-WEST PROFILE
-900 -800 -700 -600 -500 -400 -300 -200 -100 0 100 200 300 400 500 600 700
UNIT 5 NORTH-SOUTH CROSS SECTION AT 6+00
GRAPHIC SCALE
GRAPHIC SCALE
UNITS
SCALE: r = 1000'
Figure 3-6 Solid Waste Placement in the Retrofit Landfill Bioreactor Cells
26
-------
GEOTEXTILE FILTER (TOP OF LEACHATE DRAINAGE LAYER)
GRANULAR DRAINAGE LAYER
60 MIL HOPE GEOMEMBRANE (BASE GRADED
COMPACTED COHESIVE
SOL LINER
TOP OF SUBCRADE
GEOCOMPOSITE DRAINAGE STRIPS
SITU
SUBGRADE
V_ IN-
SUBl
Figure 3-7 Bottom Liner Configuration of the Retrofit Unit
Nitrified leachate was added intermittently depending on daily and seasonal factors, apparent moisture
content of the in-place waste, forecasted precipitation events, recent moisture additions as well as operator
judgment. The ultimate goal was to achieve uniform infiltration while avoiding leachate outbreaks, seeps,
and the reduction of performance of the LFG collection. Initial liquid introduction commenced in March
2002 and additions occurred steadily until October 2002 when liquid introduction stopped for a period of
about 10 months. During that time, no liquids were added to the cells except for limited amounts of
infiltration from precipitation through the long-term cover system. After this 10-month period, the
injection episodes continued until spring 2006 (with the exception of a 4-month period between January
2005 and April 2005). The cumulative liquids injection is presented in Figure 3-9, and is discussed in more
detail in Section 4.2.
The main source of moisture added to the Retrofit cells was nitrified leachate which was treated ex-situ by
chemolithotrophic bacteria that converts NH4+ to NO3" as presented in Equation 3-1.
; +2O2
; +2H+ +H2O
Equation 3-1
Laboratory research suggests that denitrifying bacteria that are already present in the waste mass will utilize
the NO3" as a terminal electron acceptor to form both N2 and small amounts of N2O gases. Liquid sources
other than leachate, including water from the landfill underdrain, sedimentation pond, and other liquid
waste streams as allowed by the permit, were used to augment the supply of leachate. These liquid sources
were pumped from the sequential batch reactor (SBR) pretreatment plant to holding tanks that were then
used to distribute leachate to the trenches via a force main and manifold system.
27
-------
t 3e+6 -
T3
'1
3
T3
^ 2e+6 -
T3
<
y
_£
1 1e+6 -
O
n
_.-•
Retrofit Cell 5.1 A £•''
Retrofit Cell 5.1 B .•'x'''
Retrofit Cell 5.2A _..''.'
^ - Retrofit Cell 5.2B f-!~"^ .s '~f
*°jj^ S
tff *
/ — ••••*? r' — "
» f" ,..' f— *""*
V'*™ .•'-•'
•.' ;.••'. — /
/^ '
v
»ff
P
3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 3-9 Cumulative Liquid Introduction into Retrofit Bioreactor Landfill Unit by Subcell
28
-------
3.1.3.3 Gas Collection
In addition to the horizontal gas collection trenches discussed previously, there are 17 vertical gas
collection wells installed in the Retrofit cells, as shown in Figure 3-10. These vertical LFG collection wells
served the dual purpose of collecting LFG and penetrating layers of daily soil cover. Probes for measuring
temperature and ORP were installed during vertical gas well installation in 2000. Additional
thermocouples and ORP probes were installed during 2001 when the Retrofit cells were retrofitted with the
additional gas collection and the liquid distribution piping network.
LFG collection wells and trenches were connected to two header lines that connect to the flare station. The
LFG collection headers maintained separate gas collection fields in the Retrofit unit. The first gas
collection header collects landfill gas mainly generated in cells 5.1 A and 5. IB, while the second collects
the gas generated in cells 5.2A and 5.2B. Before merging together, gas flow in each header line can be
measured via an orifice plate installed within the pipe. A probe was also installed in each header pipe to
allow for LFG sample collection, as will be discussed in Section 4.3.
3.2 Sample Procedures and Methods
The primary objective of this research was to assess the effect of liquids addition on waste decomposition.
The type and frequency of each analysis follows most of the guidelines presented in Chapter 2, as well as in
the First Interim Report. A discussion is presented herein of the sampling locations and sample handling
and analysis protocols. Methods used in the project meet specifications for U.S. EPA-approved
methodologies and are appropriate for the parameter and matrix of interest. The methods are generally
either from SW 846 (U.S. EPA 1996) or Standard Methods for the Examination Water and Wastewater
(APHA 1999). Equipment used for field sampling is calibrated and maintained according to manufacturer's
guidelines. Redundant solid waste probes (i.e., temperature and ORP) were employed to circumvent any
premature instrument failure. Collected leachate samples were placed in a cooler and maintained at 4 °C
(40 °F) using crushed ice and were shipped overnight to the appropriate laboratory for analysis. During the
study, inclement weather and component equipment failure affected sampling frequency and the sampling
periods were altered to accommodate such occurrences.
3.2.1 Leachate Sampling
Leachate was collected monthly from each of the landfill bioreactor and Control cells. The design of the
landfill units (i.e., paired cells) was such that, with the exception of the Control cells, each cell was
hydraulically separated at the base from the surrounding cells. The Retrofit cells were separated by
approximately 305 m (1,000 ft) laterally. The As-Built cells were constructed with a clay barrier to
hydraulically separate As-Built cell A and As-Built cell B.
Leachate samples were collected directly from the tap on the leachate riser line for each subcell. Switching
the riser pump from automatic mode to manual mode (i.e., turning the pump on manually) prior to purging
and sampling was shown to be an effective method for obtaining an adequate volume of leachate. Leachate
sample bottles were collected in the following sequence: COD, BOD, VFAs, pH, temperature, VOCs,
SVOCs, TKN, ammonia-N, nitrate-N, nitrite-N, metals, calcium, sodium, ortho phosphate, total phosphate,
chloride, sulfate, TOC, DOC, TDS and specific conductance. Methods used to analyze for these
parameters are presented in Table 3-2, while the list of VOCs and SVOCs examined are presented in
Tables 3-3 and 3-4, respectively. To obtain a representative sample, effluent was purged prior to actual
sample collection in accordance with the approved sampling protocol in the QAPP.
29
-------
UNIT 5 LANDFILL GAS
COLLECTION SYSTEM
i Ffer^
I j SOUTH I ~ M]
I I METERING If"
I j .STATION" ~
i f
, 114", ,
I j I ' (' «M.1A
« 'liJ^r i
I 1 %-524 , 1
| W < ! ,
I K-
Zr---J \ W-539T
W.5OT ABANDONED ^
«12 ~ *• T~ - • ^50
J _ _ a i /"-ABANDONED- - +
6~ ~ • - - '
~ . I*V-.
. J_ ._*W-526 II _
,- -
1 W-
¥-5/6 II _ i
i-
—£
ABANDONED
LgACHATE COLLECTION RISER
GAS WELL
CONDENSATION DRIP LEG
CONTROL VALVE
REDUCER
PIPE DIAMETER
METERING STATION
_ m m m^m COtSSTRUCTED UNIT LIMITS
— — — CELL LIMITS
ABANDONED GAS LINE
— - ^^— GAS HEADER
QAS LATERALS
GAS COLLECTION PIPES
Figure 3-10 Liquid Introduction and Gas Collection System in Retrofit Landfill Bioreactor Cells
3.2.2
Municipal Solid Waste Sampling
MSW samples were collected biennially (every two years) through discrete borings advanced into the
landfill units. The sampling location protocol required dividing the cell into six sections and then dividing
each section into approximately 3 m x 3 m (10 ft x 10 ft) grids and randomly choosing a square within the
grid. This boring location protocol was used for assigning the sampling locations in the remaining sections.
Within each unit, the same 3 m x 3 m (10 ft x 10 ft) grid was used for subsequent sampling events.
Sampling at the edges of the cell was avoided. In addition, if drilling at a selected location could not be
initiated (e.g., known location of asbestos placement) or if the boring could not be completed (e.g., an
impenetrable object was encountered) at a selected location, a randomly selected square adjacent to the
original location was selected.
The following protocol was used for MSW sampling during the study: (i) the surface elevation of the
sampling location was established; (ii) a drill rig equipped with an approximately 0.9 m (3 ft) diameter
bucket auger was used to drill into the solid waste mass; (iii) each location was sampled in approximately 3
m (10 ft) long vertical sections; (iv) a composite sample was obtained at each 3 m (10 ft) long depth
interval at each boring location; (v) the initial 3 m (10 ft) of material at each location was generally
discarded as it predominantly contained cover soil (or at least a disproportionate amount of soil); (vi) as the
boring advanced, each 3 m (10 ft) long sample was extracted from the auger and the appearance of the
waste was observed and recorded; and (vii) at least five composite samples were collected from each of six
sampling locations in each sampling time period. As such, a minimum of 30 MSW samples were collected
for each cell on a biennial basis. This sampling schedule is summarized in Table 3-5.
Temperature and ORP of the in-place MSW were measured by Type T-thermocouple probes connected to a
PC-driven data collection system. The data communications system for the probes was designed to record
the temperature and OPJ3 for each probe once every 30 minutes. These data were used to construct a
control chart for each probe. Probes installed in the Retrofit cells were mostly temporary, and as such were
removed, inspected, reconnected, and replaced (as necessary).
30
-------
Table 3-2 Leachate Sampling Parameters and Schedule
Parameter
Head on liner
Leachate production
COD
BOD
Ammonia-nitrogen (NH3-N)
Ortho Phosphate / Total Phosphate
Nitrate-nitrogen (NO3-N)
Nitrite-nitrogen (NO2-N)
Total volatile fatty acids
Temperature
TOCandDOC
pH
VOC
SVOC
TKN
Total dissolved solids
Sulfate
Chloride
Potassium
Conductance
Metals (As, Ba, Cd, Ca, Cu, Cr, Fe,
Pb, Mg, Hg, K, Na, Se, Ag, Zn)
Frequency
Continuous
Continuous
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Monthly
Quarterly
Method
Pressure Transducer
Flow Meter
SM410.4(1)
SM 405.1
SM 350.1
SM 365.2 (c)
SM 300.1
SM 300.1
GCMS
Thermometer
U.S. EPA 9060
U.S. EPA 9045
SW-846 8260(B)
SW-846 8270(B)
SM160.1(C)
SM 300.1
SM 300.1
Electrode
SW-846 60 10
Notes: Standard Methods (APHA 1999).
Table 3-3 Volatile Organic Compounds Examined in Landfill Leachate
Chemical Compound
Name
Ethylbenzene
Styrene
cis- 1 ,3 -Dichloropropene
trans- 1 , 3 -Dichloropropene
1 ,4-Dichlorobenzene
1,2-Dibromoethane (EDB)
Acrolein
1 ,2-Dichloroethane
Acrylonitrile
Vinyl acetate
Methyl Isobutyl Ketone
Toluene
Chlorobenzene
trans- 1 ,4-Dichloro-2-butene
2-Chloroethylvinyl ether
Dibromochloromethane
Tetrachloroethene
Chemical Compound Name
Carbon Tetrachloride
2-Hexanone
1,1, 1 ,2-Tetrachloroethane
Acetone
Chloroform
Benzene
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dichlorobenzene
1 ,2-Dibromo-3 -Chloropropane
DBCP
1,2,3 -Trichloropropane
Ethyl methacrylate
Methyl Ethyl Ketone
1 , 1 ,2-Trichloroethane
Total Xylenes
cis- 1 ,2-Dichloroethene
Dichlorodifluoromethane
1,2-Dichloropropane
Chemical Compound Name
1,1,1 -Trichloroethane
Bromomethane
Chloromethane
lodomethane
Dibromomethane
Bromochloromethane
Chloroethane
Vinyl chloride
Methylene chloride
Carbon Disulfide
Bromoform
Dichlorobromomethane
1 , 1 -Dichloroethane
1 , 1 -Dichloroethene
Trichlorofluoromethane
Trichloroethene
trans- 1 ,2-Dichloroethene
31
-------
Table 3-4 Semi Volatile Organic Compounds Examined in Landfill Leachate
Chemical Compound Name
4-Nitroaniline
4-Nitrophenol
Benzyl alcohol
N-Nitrosopiperidine
4-Bromophenyl phenyl ether
2,4-Dimethylphenol
N-Nitrosomethylethylamine
Cresol, p-
1 ,4-Dichlorobenzene
4-Chloroaniline
p-Phenylenediamine
Cresol, m-
2,2'-Oxybis(l-Chloropropane)
Phenol
Bis(2-chloroethyl) ether
Bis(2-chloroethoxy) methane
Bis(2-ethylhexyl) phthalate
Di-n-octyl phthalate
Hexachlorobenzene
3 ,3 '-Dimethylbenzidine
Anthracene
Isosafrole
1 ,2,4-Trichlorobenzene
2,4-Dichlorophenol
2,4-Dinitrotoluene
Diphenylamine
1,4-Dioxane
0,0,0-Triethyrphosphorothioate
Pyrene
1 ,4-Naphthoquinone
Dimethyl phthalate
Dibenzofuran
1-Naphthylamine
Kepone
Hexachloropropene
Benzo(ghi)perylene
Indeno( 1,2,3 -cd)pyrene
Benzo(b)fluoranthene
Chemical Compound Name
Thionazin
Methyl parathion
Phorate
Disulfoton
Isodrin
Benzo(a)pyrene
2,4-Dinitrophenol
Chlorobenzilate
Famphur
Dibenzo(a,h)anthracene
2-Acetylaminofluorene
Cresol, 4,6-Dinitro-O-
1 ,3 -Dichlorobenzene
N-Nitrosodiethylamine
Parathion
3 -Methy Icholanthrene
Benzo(a)anthracene
7,12-
Dimethylbenz(a)anthracene
2,3 ,4,6-Tetrachlorophenol
Cresol, p-Chloro-m-
p-Dimethylaminoazobenzene
Dimethoate
2,6-Dinitrotoluene
Pentachlorobenzene
Phenacetin
Ethyl methane sulfonate
N-Nitrosodimethylamine
N-Nitroso-Di-n-propylamine
Methyl methanesulfonate
Hexachloroethane
4-Chlorophenyl phenyl ether
Hexachlorocyclopentadiene
Isophorone
Pentachloronitrobenzene
Acenaphthene
Diethyl phthalate
Di-n-butyl phthalate
Phenanthrene
Chemical Compound Name
Pentachlorophenol
2,4,6-Trichlorophenol
2-Nitroaniline
2-Nitrophenol
2-sec-Butyl-4,6-dinitrophenol
Naphthalene
2-Methylnaphthalene
2-Chloronaphthalene
2-Naphthylamine
Methapyrilene
3 ,3 '-Dichlorobenzidine
4-Aminobiphenyl
N-Nitrosodi-n-butylamine
N-Nitrosopyrrolidine
Safrole
Cresol, o-
1 ,2-Dichlorobenzene
o-Toluidine
2-Chlorophenol
1,2,4,5-Tetrachlorobenzene
2,4,5-Trichlorophenol
Acetophenone
Nitrobenzene
3-Nitroaniline
sym-Trinitrobenzene
5 -Nitro -o -toluidine
m-Dinitrobenzene
N-nitrosodiphenylamine
Fluorene
2,6-Dichlorophenol
Hexachlorobutadiene
Benzo(k)fluoranthene
Acenaphthylene
Chrysene
Diallate
Pronamide
Butyl benzyl phthalate
Fluoranthene
32
-------
Table 3-5 Municipal Solid Waste Sampling Schedule
Parameter
ORP
Temperature
Solid waste settlement
Waste moisture
PH
Cellulose and lignin content
Organic solids
BMP
Collection Frequency
Daily
Daily
Quarterly
Biennially
Biennially
Biennially
Biennially
Biennially
3.2.2.1
Analytical Methods for Solids Analysis
The procedures that were used to process solid samples have been fully described in the QAPP presented in
the First Interim Report, and are summarized here. Samples were excavated as described above. After
excavation, a composite sample was mixed and then grab samples were placed in a 19 liter (5 gallon)
bucket for shipment for analysis. Upon receipt, samples were stored at 4 °C (40 °F) to preclude additional
biodegradation. Samples were initially shredded to a width of about 20 mm (0.8 in.) and a length of 50 to
100 mm (2 to 4 in.). After shredding, samples were again stored at 4 °C (40 °F) until they could be dried.
Approximately 1 to 2 kg (2.2 to 4.4 Ib) of each sample was dried for analysis of moisture content and solids
composition. After drying, samples were ground using a Wiley Mill to pass a 1 mm (0.04 in.) screen and
then re-dried prior to analysis. When pieces of metals and/or textiles were encountered in the sample, the
materials were cut into small pieces prior to grinding. Alternately, small pieces of metal were substituted
for larger pieces to maintain a similar metal concentration.
The concentrations of cellulose and hemicellulose were analyzed by subjecting a sample to a two-stage acid
hydrolysis (Petterson and Schwandt 1991). After washing a sample with a 2:1 mixture of toluene and 95
percent ethanol to remove lipids, a solid sample was subjected to a 72 percent H2SO4 primary hydrolysis
followed by a 28 fold dilution for the secondary hydrolysis. These hydrolyse convert the cellulose and
hemicellulose to their component simple sugars. Cellulose is a polymer of glucose, while hemicellulose is
a polymer of arabinose, galactose, mannose, and xylose. The solubilized sugars were analyzed by using a
high performance liquid chromatograph equipped with a pulsed amperometric detector. Sugars were
separated with a Carbo-Pac PA1 column manufactured by Dionex Corp., Sunnyvale, CA. The mobile
phase was 50 percent NaOH with 40.4 mL of 1.5 M sodium acetate added per liter of 50 percent NaOH
eluent. Water was circulated around the column to dampen temperature variations. Fucose was used as an
internal standard to correct for losses during the hydrolysis procedure (Davis 1998).
A new high performance liquid chromatograph system was used for the samples analyzed in 2005. The
same procedure was used although the detector was a pulsed electrochemical detector. In addition, it was
no longer necessary to circulate water to maintain column temperature with the new liquid chromatograph.
The solids that remained after the acid hydrolysis consisted of organic and inorganic materials. These
solids were then combusted for 2 hrs at 550 °C (1,022 °F) with the weight loss on combustion considered
lignin (Petterson and Schwandt 1991). In all likelihood, this procedure slightly over predicted the amount
of lignin as the material "counted" as lignin also included some plastic, rubber, and leather that were not
removed during the acid hydrolyses.
Organic solids concentrations were determined by the weight loss of the dried samples after combustion for
2 hrs at 550° C (1,022 °F). All solids analyses were performed in duplicate and average data were utilized
for the regression analyses described below. Samples were reanalyzed in all cases where the relative
percent deviation (RPD), as defined in equation 3-2, exceeded 20 percent.
Rpp=
standard deviation
mean
Equation 3-2
33
-------
3.2.2.2 Biochemical Methane Potential
Samples were placed in 160 mL capacity serum bottles after which a biochemical methane potential (BMP)
medium was added (Wang et al. 1994). Bottles were then inoculated using 15 mL of a methanogenic
consortium that were maintained on freshly ground MS W. Serum bottles were filled and maintained under
anaerobic conditions with an 80/20 mixture of N2 and CO2 gas and incubated at 37 °C (99 °F). After 60
days of incubation, the majority of the gas volume was measured by using a 60 mL capacity plastic syringe.
A 5 mL capacity wetted glass syringe was then used to measure the remaining overpressure. BMP results
were corrected for CH4 production attributable to the inoculum and to standard temperature and pressure.
The concentration of CH4 was measured by using a gas chromatograph equipped with a thermal
conductivity detector and a CTR1 column manufactured by Alltech, Deerfield, IL. BMP assays were
conducted in triplicate and five replicates were conducted to measure CH4 production associated with the
inoculum. If the RPD exceeded 20 percent, then samples were reanalyzed unless two replicates were
consistent, in which case the outlier was discarded.
3.2.2.3 Use of Cellulose Plus Hemicellulose to Lignin Ratios
Historically, daily and intermediate covers were applied to each landfill cell and the amount of soil was
estimated to range from 10 to 25 percent of the waste volume. Thus, there is the potential to include soil in
refuse samples that are excavated from landfills. The soil component of the sample will dilute the
cellulose, hemicellulose, and lignin concentrations in the sample. The most effective way to eliminate the
effect of soil dilution is to analyze data based on the cellulose plus hemicellulose to lignin (CH:L) ratio
rather than the concentration of either the cellulose or hemicellulose. While advantageous from the
perspective of eliminating the influence of the soil, this method has the disadvantage that the initial CH:L
ratio of the buried refuse is often unknown or only known within a range. The variation in the CH:L ratio
for fresh refuse was presented in Table 2-5. This ratio may be different when all of the different waste
streams entering a landfill are considered.
3.2.4 Landfill Settlement
Settlement of the landfill was monitored on a quarterly basis as a secondary indication of decomposition
and stability using a GPS measurement of surface elevation. GPS surveying was performed using a
Trimble Model 4800 for the settlement plates within each cell using the following protocol: (i) every
sampling event was initialized from a known point that was controlled at + 5 cm (2 in.) for horizontal and
vertical control (if sampling within a cell was interrupted, the system was reinitialized from the known
point before sampling resumed); (ii) sampling was initiated if the root mean square reading from the system
was < 10 cm (4 in.); and (iii) the positional dilution of precision as measured on the device was < 6 before
the GPS system could be used. In addition, one of every 20 settlement plates measured by the GPS unit
was randomly selected and re-tested. These results were compared to the limits established in the QAPP
presented in the First Interim Report. If the three aforementioned conditions were met, the positional
accuracy of the GPS readings was reported to be sufficient to meet the analytical needs of the study. A
detailed landfill settlement analysis is not included in Chapter 4, as waste filling occurred continuously in
the Control cells until 2004, and waste filling occurred in the As-Built cells until 2005. Since waste filling
occurred in these cells throughout the monitoring period, an assessment of total settlement and a correlation
between waste decomposition and settlement could not be made. Monitoring of the settlement points will
continue at the site and a correlation between waste settlement and waste degradation will be investigated
as part of the Final Report.
3.2.5 Landfill Gas Sampling
3.2.5.1 Gas Collection System Sampling
The LFG sampling schedule is presented in Table 3-6. Primary LFG constituents (i.e., CH4, CO2 and O2)
were measured weekly, whereas trace gases (i.e., NMOCs, including individual HAPs) were measured
quarterly. Surface emissions monitoring of CH4 was conducted quarterly, if required. Table 3-7 indicates
the specific HAPs (which are a subset of NMOCs) that were analyzed for in each sample. Similar to the
leachate sampling protocol, LFG sampling occurred at one point per cell. The LFG extraction wells are
34
-------
located systematically across the cell, approximately equidistant from each other, to provide relatively
uniform extraction coverage.
Table 3-6 LFG Sampling Schedule
Parameter
LFG flow
LFG percent composition (CH4, CO2, O2,)
HAPs
Surface emission monitoring (CH4)
Collection Frequency
Weekly
Weekly
Quarterly
Quarterly
LFG analysis was also performed for CH4, CO2, and O2 using a CES Landtec GEM 2000 as outlined in the
QAPP. This instrument is a portable field gas analyzer and uses a self-compensating infrared detector.
After calibration, the instrument was connected to a gas sampling port on the selected gas header using
flexible plastic tubing. Gas was drawn into the instrument by an internal pump. Results are date and time
stamped and logged by the instrument. Gas standards for CH4, CO2 and O2 were also analyzed twice daily
on the day of sampling to evaluate accuracy objectives as outlined in the QAPP. Concentration readings
for CO2 and CH4were to be within 15 percent of the actual concentration or of the sample duplicate; the
tolerance for O2) was to be ± 30 percent. LFG analyses were made by electronically logging three
consecutive measurements at a frequency of one measurement per minute of LFG composition and flow.
Flow rate, differential pressure, static pressure, and temperature were recorded on the field instrument at
each sample location (i.e., well, lateral, header, and flare). The mean value for each of these measurements
was recorded and selected as the reported value for each parameter of interest.
LFG samples were also collected for laboratory analysis of CH4, CO2, N2 and O2 by U.S. EPA Method 3C,
NMOCs were sampled and analyzed by U.S. EPA Method 25C, and HAPs were analyzed as identified in
U.S. EPA Method TO-14. For the gas samples obtained for external laboratory analysis, samples were
collected in 6 liter capacity SUMMA® passivated stainless steel canisters.
Table 3-7 HAPs Analyzed in Quarterly LFG Sampling
Chemical Compound
Dichlorodifluoromethane
Chloromethane
1,2-Dichloro-l, 1,2,2-
tetrafluoroethane
Vinyl chloride
Bromomethane
Chloroethane
Trichlorofluoromethane
1, 1-Dichloroethene
Carbon disulfide
l,l,2-Trichloro-l,2,2-
trifluoroethane
Acetone
Methylene chloride
trans-l,2-Dichloroethene
1 , 1 -Dichloroethane
1,2,4-Trichlorobenzene
Hexachlorobutadiene
cis- 1 ,3 -Dichloropropene
Chemical Compound
Toluene
trans- 1 , 3 -Dichloropropene
1 , 1 ,2-Trichloroethane
Tetrachloroethene
2-Hexanone
Dibromochloromethane
1,2-Dibromoethane (EDB)
Chlorobenzene
Ethylbenzene
Xylenes (total)
Styrene
Bromoform
1 , 1 ,2,2-Tetrachloroethane
Benzyl chloride
4-Ethyltoluene
4-Methyl-2-pentanone
(MIBK)
Chemical Compound
Vinyl acetate
cis- 1 ,2-Dichloroethene
2-Butanone (MEK)
Chloroform
1,1,1 -Trichloroethane
Carbon tetrachloride
Benzene
1 ,2-Dichloroethane
Trichloroethene
1 ,2-Dichloropropane
Bromodichloromethane
1 ,3 ,5-Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 , 3 -Dichlorobenzene
1 ,4-Dichlorobenzene
1 ,2-Dichlorobenzene
35
-------
3.3 Data Processing and Statistical Analysis
3.3.1 Sample Dating and Statistical Analysis
The primary goal of the study at the OLLB was to evaluate the extent to which the operation of a landfill as
a bioreactor accelerated refuse decomposition. The CH:L will be used as a metric to compare samples
excavated from different landfill cells. However, waste age must also be considered as refuse decomposes
over time, resulting in a change in this ratio (Barlaz 2006). Thus, six year old refuse from a Control cell
may (or may not) be more decomposed than six month old refuse from a bioreactor landfill cell.
To evaluate the effect of decomposition over time, it was necessary to establish the age of the solid waste at
the time of sample collection. The age of the waste represented by each sample was estimated from site
survey data that were specific to the location of each boring. The survey data provided historical waste
placement information that allowed an estimate of the age of each sampling interval for each boring.
Typically, survey data were available to document the date of waste filling through the entire depth of the
landfill. In some cases, it was not possible to estimate the age of waste from a specific boring. In such
cases, that data were excluded from the analyses presented in this section. Over the entire data set, 3.4
percent of the samples were excluded because of a lack of information regarding waste age.
To evaluate reproducibility between cells, a given characteristic of the refuse (e.g., moisture content) was
initially compared between cells by using a t-test in Microsoft Excel®. All of these analyses were
conducted using a two-sided t-test, assuming unequal variance in the data. In general, data sets were only
considered significantly different if the results of the t-test indicate a probability (p-value) of less than 0.05,
meaning that there is a 95% probability that a difference is relevant in consideration of sample variability.
To evaluate decomposition as a function of time, the CH:L, BMP and organic solids data for a specific cell
were plotted as a function of waste age, and a linear regression was conducted. The slope of the best fit
line from the linear regression was then evaluated to determine whether the slope was significantly
different from zero by calculating the 95 percent confidence intervals for each slope using the regression
analysis tool in Microsoft Excel. If the 95 percent confidence interval included a slope of zero, then the
slope was judged to be statistically similar to zero. A slope of zero means that there was not a significant
decrease in the measured parameter (e.g., BMP) with respect to waste age. Slopes for all linear regressions
and the corresponding 95 percent confidence intervals are presented in Section 4.1. The same procedure
has been used in previous research (Barlaz et al. 2004).
Diagnostic plots were analyzed to ensure linear regression model assumptions were met and to investigate
the influence of individual points upon the fit. A plot of residuals versus fitted values was used to assess
the assumption of constant error variance. A normal quantile-quantile plot was used to assess the
assumption that errors were distributed normally. A residuals-versus-leverage plot was used to investigate
the influence of individual points on the regression fit.
As illustrated by Equations 2-2 and 2-3, the biodegradation of cellulose and hemicellulose is linked to CH4
production. As described in Section 4.3, the predictive CH4 generation model is exponential. Thus, the
CH:L and BMP are more likely to decrease exponentially as the remaining substrate (i.e., cellulose and
hemicellulose) becomes less bioavailable. In this respect, a first order decay model may be a more
appropriate mathematical representation of solids decomposition as a function of waste age. A linear
regression model was used here only as a method to rapidly assess a large amount of data and to assess
statistically whether the calculated slopes were significantly less than zero. The calculated model is not
intended for use in a predictive mode based on the analyses presented in this section. It should also be
noted that a linear model is not ideal when working with a ratio (CH:L). However, of the three solids
decomposition parameters evaluated (CH:L, BMP, and organic solids), CH:L is the only one that eliminates
the influence of soil dilution. Thus, it was judged important to use this ratio in a linear regression.
A multiple linear regression analysis was also conducted. Five variables were considered (waste age,
sampling date, sample elevation, moisture content, and temperature). First, the correlation between the five
variables was investigated. Next, multiple parameters were evaluated, including both individual parameters
(e.g., moisture content), and parameter interactions (e.g., sampling date x elevation). Best-fit models were
developed using R, a statistical modeling software (www.r-project.org). and the results were inspected
36
-------
qualitatively using both the adjusted R2 and the significance of the parameter modeled. The adjusted R2 is
comparable to R2 for the linear regression analysis in that it is corrected for the used multiple parameters.
3.3.2 Moisture Balance Calculations
As a part of this study a moisture balance was conducted to provide confirmation of the "apparent" or
"anticipated" increase in the bulk moisture content of the waste. Waste moisture content was also directly
measured during the biennial solid waste sampling events. An evaluation was made to compare the
calculations to the measured values. The Control and Retrofit cells were each considered as a single unit
due to the lack of any hydraulic separation between their subcells. As-Built cells A and B, however, were
considered as individual units because of the clay barrier separating them. The moisture balance
calculations were performed for the all the cells between March 2002 and December 2005. Parameters
used in preparing the moisture balance included: (i)volumes of recirculated leachate; (ii) supplemental
liquid addition; (iii) estimated infiltration; (iv) estimated surface run-on; and (v) measured leachate
generation. Moisture generated as a result of biological or chemical interactions within the landfill cells
was neglected as a source of additional liquids in the moisture balance calculations. Moisture leaving the
landfill through the gas collection system as condensate was also not included in the moisture balance
calculations. These two sources were not expected to significantly influence the results of the analysis (less
than approximately 1500 cubic meters per cell) expressed in %.
March 2002 was chosen as the start date because solid waste samples were collected from each cell during
that month and analyzed for (among other parameters) moisture content. Since the sampling was
conducted prior to liquids addition in the bioreactor cells, the mean moisture content from the March 2002
solid samples was used as the initial moisture content in the moisture balance calculations. Thus, the
estimated initial moisture content (on a wet weight basis) was 32.5 percent for the Control cells, 37 percent
for the Retrofit cells, 42 percent for As-Built cell A, and 43 percent for As-Built cell B. Typical moisture
content on a wet weight basis (i.e., mass of water divided by mass of wet waste) of incoming waste can
range from 15 to 40 percent (Tchobanoglous 1998). It is acknowledged that these moisture content values
are on the high end of the "typical" range of waste moisture content reported by Tchobanoglous (1998).
Given the relatively wet climate of Louisville and the fact that the data utilized are from a single sampling
event, the actual moisture content of as-received waste may be less than what was measured in the March
2002 samples. However, using these data provided the best starting point for estimating an initial moisture
content amount for use in the moisture balance calculations.
With regard to the parameters used in performing the moisture balance, the volume of recirculated leachate,
the volume of supplemental liquids added, and the volume of the generated leachate were recorded daily.
Precipitation at the site was recorded daily at an on-site weather station. To account for infiltration, runoff
and evapotranspiration, parameters were estimated using the Hydrologic Evaluation of Landfill
Performance (HELP) model (Schroeder et al. 1994). The type of cover material, thickness of the cover
material, slope of the waste surface, and slope length were selected as input parameters to the HELP model.
Only the surface of each landfill cell was modeled in HELP to calculate runoff, evapotranspiration, and
infiltration. Details regarding selection of the specific input parameters were obtained using available
survey maps and by communication with the OLRDF's site manager. For the Control cells, Retrofit cells,
and sloped surfaces of the As-Built cells, the cover material was assumed to be clay with a hydraulic
conductivity of 1 x 10"6 cm/s (4 x 10"7 in/s). The thickness of the clay cover was assumed to be
approximately 1 m (3 ft). The flat areas of the As-Built cells consisted of a cover material made of an
approximately 0.3 m (1 ft) thick layer of compost with a hydraulic conductivity of about 4 x 10"2 cm/s. The
compost cover material was replaced by a 1 m (3 ft) thick layer of clay in August 2004. The slope of the
landfill units was 4 horizontal to 1 vertical (4H:1V) and the length of the slope was assumed to be 76 m
(250 ft). Additional details on the moisture balance parameters for the different units are presented in
Table 3-8. The topography of the landfill units and the ratio of the sloped versus flat areas of the landfill
units was estimated using contour intervals from the survey maps that were available from March 2002 to
December 2005. These estimates are presented in Table 3-9 (As-Built cells) and Table 3-10 (Retrofit and
Control cells). It is noted that because the day-to-day details of waste filling and covering were not
available, the inputs to the HELP model are best estimates, and the results should be treated as such.
Run-on was the only parameter that was not recorded or calculated directly; rather, it was estimated based
on runoff values output by the HELP model and the geometry of each cell. Run-on from Unit 7.2 (located
37
-------
to the north of As-Built cell A) to As-Built cell A, run-on from Control cell A to Control cell B, and run-on
from the Control cell to As-Built cell B was included in the estimated/calculated infiltration volumes. Run-
on percentage was estimated using the parameters presented in Table 3-8. Run-on from unit 7.2 to Control
cell A was assumed to be 25 percent of rainfall. This was based on an average value of runoff in the sloped
and flat surfaces of unit 7.2. It is noted that prior to late 2003, most of the run-on volume did not infiltrate
to the As-Built cells, but rather was channeled directly to the LCS, as presented in survey maps and
reported by WM. This was due to the inclusion of a "column" of rubber tire chips that were placed
between: (i) unit 7.2 and As-Built cell A; (ii) As-Built cell A and As-Built cell B; and (iii) As-Built cell B
and Control cell B. The channeling was assumed to cease after the first quarter of 2004 when tire chips
were no longer placed at the edges of the cells.
Table 3-8 Moisture Balance Parameters
Parameter
Runoff (percent)
Evapotranspiration
(percent)
Infiltration (percent)
Control &
Retrofit
Sloped Area
51
47
2
Control &
Retrofit Flat
Area
49
48.5
2.5
As-Built
Sloped
Area
51
47
2
As-Built
Flat Area
3.5
60
36
Table 3-9 Area Distribution of As-Built cells
Date\Units
Mar 02-Oct 02
Nov 02-Feb 03
Mar 03- May 03
Jun03-Nov03
Nov03-Oct04
Nov 04 - Apr 05
June 05-Dec 05
As-Built cell A
Sloped Area, m2
(acres)
12141(3.0)
10117(2.5)
12950 (3.2)
12141 (3.0)
12141 (3.0)
12141 (3.0)
566 (1.4)
As-Built cell A
Flat Area, m2
(acres)
39659 (9.8)
18211(4.5)
16187 (4.0)
18211(4.5)
18211(4.5)
18211(4.5)
13759 (3.4)
As-Built cell B
Sloped Area, m2
(acres)
0
3237 (0.8)
8094 (2.0)
14164(3.5)
14164(3.5)
40469 (10)
40469 (10)
As-Built cell B
Flat Area, m2
(acres)
36017 (8.9)
38850 (9.6)
32375 (8.0)
32375 (8.0)
46539(11.5)
22662 (5.6)
22662 (5.6)
Table 3-10 Area Distribution of Retrofit and Control Units
Date\Units
Mar 02-Dec 05
Retrofit Sloped
Area, m2 (acres)
99553 (24.6)
Retrofit Flat
Area, m2 (acres)
19425 (4.8)
Control Sloped
Area, m2 (acres)
33994 (8.4)
Control Flat
Area, m2 (acres)
0
3.3.3 Statistical Analysis ofLeachate Parameters
Multiple linear regressions were utilized to quantify trends in the leachate parameters of interest. Two
explanatory variables were used in the regression fits, sampling date and a phase-shifted sine function of
the sampling date (to capture a possible seasonal component). The sine variable was phase-shifted to have
a maximum on September 1st and minimum on March 1st. For the regression fits, the parameter Sampling
Date is the number of days between the sampling date and 1/1/1970.
While not presented, several diagnostic plots were analyzed to ensure linear regression model assumptions
were met and to investigate the influence of individual points upon the fit. A plot of residuals versus fitted
values was used to assess the assumption of constant error variance. A normal quantile-quantile plot was
used to assess the assumption that errors are distributed normally. Residual versus leverage plot was used
to investigate the influence of individual points on the regression fit.
38
-------
Chapter 4. Solid Waste and Moisture Content Analysis
A data validation report, which confirms that data presented herein were collected according to the
specifications outlined in the QAPP in the First Interim Report and provides validation of data presented in
this section, is included as Appendix A.
4.1 Solid Waste Analysis
The objective of Section 4.1 is to present and analyze data on the composition of solids excavated from the
Control, As-Built, and Retrofit cells of the OLLB between 2000 and 2005. Major sampling events were
conducted in 2000, 2002, 2003, and 2005. In addition, fresh waste entering the Control and As-Built cells
was characterized in 2001 and 2004. The solids composition data from the OLLB are presented and
discussed in Section 4.1.1. To the authors' knowledge, the data presented in this section represent the
largest data set of refuse samples from a single site available to date.
4.1.1 Results
The primary focus of this section is to present and analyze data on the composition of the solids excavated
from the Control, As-Built, and Retrofit areas. The discussion focuses on the following two questions:
1. Do the solids composition data support the assumption that replicate cells actually behaved as
replicates? This applies to a comparison of results from replicate areas of the Retrofit,
Control, and As-Built cells. Data from cells 5.1A and 5. IB are referred to as Retrofit cell A,
and data from cells 5.2A and 5.2B are referred to as Retrofit cell B.
2. Do the solids composition data support the hypothesis that the operation of a landfill cell as a
bioreactor accelerates the biodegradation of organic matter as described by equations 2-2 and
2-3. It is recognized that there are inherent differences in the Retrofit and As-Built landfill
bioreactors. As such, some caution is warranted in analysis of the results.
The discussion in this section addresses these two questions from the perspective of the solids composition
data only. Later in this report, the solids, gas, and leachate data are considered together to evaluate whether
there is evidence that operation as a bioreactor landfill resulted in accelerated refuse decomposition.
4.1.1.1 Solids Decomposition in the Retrofit cell
The first assessment of "data reproducibility" is to compare the calculated waste age from Retrofit cells A
andB (i.e., 5.1 and 5.2). If there is a significant difference in the waste age between cells that cannot be
explained by the cells' fill histories, it suggests that the procedure used to date the waste is not entirely
appropriate. In comparing the intercell variability, the waste age in Retrofit cell A was older than that in
Retrofit cell B by 222, 89 and 89 days in 2000, 2002 and 2005, respectively (Table 4-1). This difference
was only significant (p < 0.05) in 2000 when the waste was youngest and small differences would be more
significant. The actual difference may be important for waste less than five years old, but becomes less
important over time. The waste age data are encouraging as they are reasonably consistent. This suggests
that the solids decomposition data can be assessed with respect to the calculated waste age. Another way to
assess differences in waste age between cells would be to evaluate the distribution of the sample age data as
presented in Figure 4-1.
As described in Chapter 3, the Retrofit cells are reported to be composed of approximately 32 percent soil.
While this will not affect the CH:L, it will reduce the BMP, which is expressed on a "per dry mass" basis.
As presented in Table 4-1, there are significant differences between Retrofit cells A and B in 2005 for
CH:L, BMP, and organic solids. The 2005 trends are consistent in that all three measures of solids
decomposition suggest that the solids from Retrofit cell A are more decomposed relative to samples from
Retrofit cell B. However, this observation is not consistent with the measured moisture contents, which are
statistically similar between Retrofit cells A and B in 2005. The decomposition rate data, which are
described below, do not support the hypothesis that the rate of decomposition is different between Retrofit
cells A andB.
39
-------
Table 4-1 Summary of Waste Composition Data for the Retrofit cell
Sampling
Year
0
o
0
-. O
<
to O
c3 O _
O
O
O
n=34 n=42 n=45 n=44 n=36 n=27
— •-
'
1
f^H "^ 8
— ' — ' 9
]
8
i i i i i i
O O (N (N in iri
O O O O O O
o o o o o o
(N (N (N (N (N (N
^H (S| ^H (S| ^H (S|
>ri in >n in in
-------
A graphical depiction of trends in the relationship between waste age and CH:L, BMP, and organic solids
are presented in several figures, included in Appendix B. Figure 4-2 depicts the combined Retrofit cell
data, with no distinction between Retrofit cells A and B. This was done to utilize the largest data set
possible, minimizing the confounding effects of other variables that may affect refuse decomposition.
Additional analysis of the CH:L and BMP results in which the data were separated by area are presented in
Appendix B. The statistics associated with the linear regression analysis are presented in Table 4-2.
For analysis purposes, regression modeling was chosen. There are several limitations to the use of linear
regression models including whether a linear model is most appropriate, whether waste age is the most
important independent variable, and whether multiple variables should be considered concurrently. As
discussed in Chapter 3, a first order decay model may be a more appropriate mathematical representation of
solids decomposition. However, as illustrated in Figure 4-2, the calculated exponential function is similar
to the linear model for CH:L. Further assessment of exponential models is beyond the scope of this interim
report, as the results suggest that additional data collection and analysis over the next several years is more
critical.
When all of the CH:L data are combined for the Retrofit Unit, the slope of the regression is significantly
less than zero at the 95% confidence level despite the high degree of scatter. This is also the case for the
combined data for BMP and organic solids. A negative slope means that there is a statistically significant
decrease in the value of a solids decomposition parameter with waste age. The BMP is a more sensitive
measure of decomposition as it tracks changes in biodegradable solids only, while the organic solids
includes both biodegradable organics (e.g., cellulose) and non-degradable organic solids (e.g., HDPE).
With respect to the trends in decomposition as presented in Figure 4-2, the low correlation coefficients
indicate that a single variable linear regression model does not completely characterize the measured
trends. This is reasonable as there are many other variables affecting refuse decomposition. Multiple
parameter regressions are discussed below.
When Retrofit cells A and B are analyzed separately (as presented in Table 4-2) for CH:L and BMP, the
slopes are significantly less than zero for both measures in Retrofit cell A but only for BMP in Retrofit cell
B. The results do not provide sufficient evidence to pursue further analyses of the data in which Retrofit
cells A and B are treated differently as the larger data set associated with the combined area is more robust.
Many factors other than waste age are expected to be correlated with refuse decomposition including, for
example, moisture content, pH, substrate quality and temperature. To evaluate whether moisture content
alone was a better indicator of decomposition, the BMP and CH:L were plotted as a function of moisture
content in Figure 4-3. As illustrated, moisture content was not a good predictor of the extent of
decomposition. Two explanations for this result are that (i) the samples that contained elevated moisture
were not wetter long enough for moisture to influence decomposition at the time of sampling; and (ii) other
confounding variables precluded a clear trend. Ultimately, when all of the data from the Retrofit cells are
considered, the increase in the average moisture content between 2000 and 2005, from 34.7% to 38.6%,
was statistically significant (p<0.05) but not so dramatic as to be the only variable controlling either the
CH:L or BMP. The use of multiple parameter regression models is discussed in Section 4.1.2.
4.1.1.2 Solids Decomposition in the Control cell
As presented in Table 4-3, there was not a significant difference in waste age between samples obtained
from Control cells A and B in 2000, 2003, and 2005. Both the median waste age and the range were quite
similar (Figure 4-4). The 225-day difference associated with the 2002 sampling, as well as the different
ranges for Control cells A and B, was surprising and likely reflects some imprecision in the sample dating
process (Figure 4-4). There were no significant differences in any of the waste monitoring parameters at
any of the sampling periods, with the exception of the moisture content in 2005, when the average moisture
content of samples from Control cell B was about four percent higher than samples from Control cell A
(Figure 4-5). The moisture data for Control cell A in 2000 shows some particularly wet samples relative to
Control cell B. The two wettest samples have moisture contents of 49.6 and 56.4%, respectively.
However, the corresponding CH:L (2.84 and 1.49) and BMP (114.9 and 75.8 ml CHydry gm) do not
suggest that these two samples were well decomposed relative to the bulk of the samples (Figures presented
in Appendix B). As the characteristics of Control cells A and B were quite similar, all subsequent analyses
41
-------
Table 4-2 Slopes and Their 95% Confidence Intervals for Linear Regressions
Between Waste Age and CH:L, BMP, and Organic Solids in the Retrofit, Control, and As-Built Cells
Cell
Retrofit
cells (A
&B)
Retrofit
cell A
Retrofit
cellB
Control
cells
As-Built
cells (A
&B)
As-Built
cell A
As-Built
cellB
CH:L
Lower
Level
-3.3E-4
-3.9E-4
-3.4E-4
-2.6E-4
-8.3E-4
-1.2E-3
-7.5E-4
Slope
-2.3E-4
-2.6E-4
-1.5E-4
-9.5E-4
-5.1E-4
-7.2E-4
-3.0E-4
Upper
Limit
-l.E-4
-1.2E-4
-3.0E-5
-7.0E-5
-1.9E-4
-2.5E-4
-1.5E-4
BMP
Lower
Level
-1.4E-2
-1.3E-2
-1.7E-2
-1.8E-2
-1.6E-2
-3.5E-2
-2.3E-2
Slope
-9.7E-2
-8.8E-3
-9.7E-3
-1.1E-2
-3.2E-2
-1.9E-2
-4.3E-3
Upper
Limit
-5.8E-3
-4.5E-3
-2.5E-3
-4.5E-3
9.2E-3
-2.6E-3
1.4E-3
Organic Solids
Lower
Level
-8.3E-3
Slope
-6.2E-3
Upper
Limit
-4.1E-3
Not Analyzed
-5.3E-3
-1.4E-2
-1.7E-2
-1.5E-2
-2.1E-3
-8.1E-3
-l.OE-2
-6.0E-4
1.2E-3
-2.4E-3
-3.3E-3
3.3E-3
o
4 -
3 -
2 -
1 -
1000
Linear
Exponential
y = -0.0002x +1.5695
R2 = 0.0738
y=1.3524e~l
R2 = 0.0601
2000 3000
Waste Age (days)
4000
Figure 4-2 Relationship between CH:L and Waste Age in the Retrofit Cells
42
-------
200
BMP
y = 0.6568x +3.2994
R2 = 0.0373
CH:L
y = 0.0143x +0.5663
R2 = 0.0243
- 3
ffi
O
40 60
Moisture Content (%)
80
• BMP
O CH:L
Linear - CH:L
Linear-BMP
Figure 4-3 Relationship between Moisture Content and Solids Decomposition in the Retrofit Cells
O
o
o
in
-------
o
oo
o _
o _
o
m
o
(N
n=59 n=30 n=39 n=31 n=40 n=31 n=30 n=28
oO(N(Nmminin
oooooooo
oooooooo
< PQ
Figure 4-5 Moisture Content Profile for the Control cells at Each Sampling Time
ffi
O
OH
S
m
200
180 -
160 -
140 -
120 -
100 -
80 -
60 -
40 -
20 -
0 -
BMP O
y= 1.927x + 3.9506
R2 = 0.
o o CH:L
y = 0.0292x+1.1404
R2 = 0.0366
- 5
- 3
ffi
O
- 1
10 20 30 40 50
Moisture Content (%)
60
70
• BMP
O CH:L
Linear - CH:L
Linear - BMP
Figure 4-6 Relationship between Moisture Content and Solids Decomposition in the Control Cells
44
-------
Table 4-3 Summary of Waste Composition Data for the Control cells and Assessment of Cell
Replication
Sampling
Year
0
o
0
O
0
-------
difference in the moisture content in 2004 between the refuse added to As-Built cells A and B. Although
both sets of samples were collected in May, 2004, their age is different and presumably more moisture was
added to As-Built cell B over the 76-day residence time of these samples in the landfill.
The waste age increased in As-Built cell A between 2003 and 2005, whereas the average waste age was
statistically similar between 2003 and 2005 in As-Built cell B. This was consistent with the addition of
fresh waste to As-Built cell B between 2003 and 2005. A comparison of the waste age profiles for As-Built
cells A and B is presented in Figure 4-7.
Trends in the relationship between waste age and CH:L, BMP, and organic solids are presented in
Appendix B, and the statistics associated with the linear regression analysis for these figures are presented
in Table 4-2. For As-Built cell B, the linear regression slopes were not significantly different from zero for
any of the monitoring parameters (i.e., CH:L, BMP, and organic solids). As the calculated average waste
age in As-Built cell B was only 545 days, thus a clear trend was not apparent. This is a very short time
interval over which to observe any refuse decomposition.
For As-Built cell A, the slopes of all measures of decomposition, CH:L, BMP and organic solids were
significantly less than zero (Table 4-2) which indicates that decreases in CH:L, BMP and organic solids
with waste age were all statistically significant even though the average waste age was relatively young
(730 days). The steeper slopes for the solids data for As-Built cell A relative to As-Built cell B were
consistent with the fact that As-Built cell A received more air. The presence of air would have stimulated
aerobic decomposition which is faster than anaerobic decomposition. Thus, the data suggest that the
additional air added to As-Built cell A accelerated decomposition. As was the case for the Retrofit and
Control cells, the moisture content was not a useful predictor of waste decomposition (Figures 4-8 and 4-9).
This is not to suggest that moisture content is not a significant variable, but rather that the relatively young
waste age coupled with other confounding variables mask a simple statistical relationship. As for the other
test areas, these results suggest that additional monitoring in the coming years is warranted to be able to
assess results and trends over a longer time period.
0
Ti
3
60
O
Ti
0 -
n=24 n=27 n=42 n=52 n=32 n=31
ts
ts
ts
m
•*t
r--'
ts ts
Ti
o
o
ts
m
Figure 4-7 Waste Age Profile for As-Built Cells A and B
46
-------
4.1.2 Multiple Linear Regression Models
Linear regression analyses using either waste age or moisture content as the independent variable indicated
that neither of these parameters was able to explain the observed variation in CH:L, BMP or organic solids.
This is reflected in the correlation coefficients presented with each linear regression. These coefficients
were typically below 0.1 and never exceeded 0.13. In conducting a methodical search of potential single
parameter relationships using moisture content, waste age, sampling date, sample elevation, and
temperature, it was found that all variables except temperature were about the same in their ability, or
inability in this case, to explain the observed trends in CH:L, BMP or organic solids. As a result, multiple
linear regression analyses were conducted in an attempt to better explain the observed variation in solids
decomposition monitoring parameters.
Subsequent work with temperature as a third variable did not improve the two variable (age, moisture)
models, while the use of elevation offered slight improvements. Specifically, when elevation was used as
the third variable for the same twenty four models, six had an adjusted R2 below 0.1, twelve had an
adjusted R2 of 0.1 to 0.2 and six had an adjusted R2 of 0.27 to 0.50. Finally, work was conducted to
explore models that included a sample date and an elevation term. The logic for this interaction was that it
might account for waste settlement. While this model did offer an improved adjusted R2, the physical
meaning of this term is not clear and the extent of settlement would actually reflect decomposition of waste
under the sample, and not the sample itselfs. Thus, no further discussion of this model is presented.
The best multiple linear regression model was determined to be:
Decomposition Parameter = f (Moisture Percent + Waste Age + Elevation) Equation 4-1
In summary, the use of multiple linear regression analysis validates the significance of moisture content as
a parameter that controls waste decomposition. However, given the number of other factors that affect
decomposition, even a multiple parameter model could not account for most of the variability in the solids
decomposition parameters.
200
150 -
oo
e?
T3
nf
o 100
50 -
BMP
y = -0.4149x +68.86
R2 = 0.017
o
CH:L
y = -0.0221x + 2.5792
R2 = 0.057
- 4
- 3
- 2
- 1
ffi
U
15
35 45
Moisture Content (%)
55
65
• BMP
O CH:L
Linear - CH:L
Linear - BMP
Figure 4-8 Relationship between Moisture Content and Solids Decomposition in As-Built Cell A
47
-------
oo
o
£
2
m
200
150 -
100 -
50 -
BMP
y = 0.0527x + 49.567
R2 = 0.0006
CH:L
y = 0.0009x+ 1.6349
R2 = 0.0003
o o
o
- 1
40 60
Moisture Content (%)
80
• BMP
O CH:L
Linear - CH:L
Linear - BMP
Figure 4-9 Relationship between Moisture Content and Solids Decomposition in As-Built Cell B
4.1.3 Summary of Solids Decomposition
The objectives of this section were to: (i) assess reproducibility between the cells and (ii) evaluate whether
the solids composition data support the concept that operation of a bioreactor landfill can accelerate solids
decomposition relative to a traditional landfill. The waste age is significantly older in the Retrofit cell
relative to the As-Built and Control cells. As such, trends that include the influence of waste age should be
most apparent in the Retrofit cells. However, as refuse decomposes the nature of the remaining degradable
organic matter changes. Thus, the refuse in the As-Built and Retrofit bioreactor landfills cannot be
assumed to be the same.
Analysis of cell reproducibility was important as it showed significant differences between As-Built cells A
and B. Given the impracticalities associated with the operation of a full-scale landfill as a research site, it is
not surprising that the cells were not perfect replicates however, this project will continue utilizing during
the lifetime of this CRADA. As all solids data could be analyzed as a function of waste age, the need for
replication was less important than cases where waste age is difficult to assess. In the case of the OLLB, it
was possible to assign a waste age to individual samples on the basis of landfill fill records, a capability not
afforded at many other study sites.
The available solids composition data document significant decreases in degradable solids as a function of
waste age in the Retrofit cell. The average waste age in Retrofit cells A and B in 2005 was 3,044 and 2,955
days, respectively. In contrast, the average waste age for Control cells A and B in 2005 was 1,657 and
1,655 days, respectively. Thus, there is less time for trends to become apparent in the Control cell relative
to the Retrofit cells. In 2005, the average moisture contents were similar in the Retrofit and the Control
cells, although the recirculation of leachate in the Retrofit cells would be expected to provide some
stimulation even if the average moisture contents are similar.
The slope of the CH:L and organic solids linear regression lines for the Retrofit cells were greater (i.e.,
more negative) than that for the Control cells as would be expected given that the Control cells were not
operated to accelerate decomposition. However, the slopes of the BMP linear regression lines for these two
48
-------
Table 4-4 Summary of Waste Composition Data for the As-Built cells and Assessment of Cell
Replication
Sampling
Year
>•>
"i
W r>
0
O
ts
Subcell
A
B
Statistical
Parameter
average
stnd. dev.
average
stnd. dev.
p-value
A
B
average
stnd. dev.
average
stnd. dev.
p-value
A
B
average
stnd. dev.
average
stnd. dev.
p-value
A
B
average
stnd. dev.
average
stnd. dev.
p-value
A
B
average
stnd. dev.
average
stnd. dev.
p-value
Waste Age
at Time of
Sampling
(days)
36
0)
72
0)
0)
290.3
155.01
207.5
115.3
0.05
418
175.48
562
123.4
1.05E-05
22
0)
76
0)
0)
730
321.3
545
405.3
0.048
Moisture
(%)
40.0
4.5
45.8
7.0
0.12
41.9
9.19
41.5
9.48
0.90
41.1
8.3
48.4
16.7
0.0114
28.1
4.7
39.5
6.38
0.0055
39.9
7.6
39.0
9.6
0.708
CH:L
1.54
0.77
3.10
0.66
0.00
0.96
0.39
1.00
0.43
0.76
2.12
0.55
1.62
0.51
1.53E-05
2.99
0.80
2.40
0.62
0.186
1.46
0.63
1.79
0.63
0.045
BMP
(mL
CH4/dry
gm)
57.7
18.5
84.2
22.3
0.05
26.6
12.20
20.3
11.89
0.08
68.8
25.0
49.0
21.7
9.05E-05
75.9
15.7
97.8
22.56
0.080
44.2
19.7
62.9
26.5
0.002
Organic
Solids
(%)
62.4
12.1
82.6
4.2
0.00
41.9
5.96
37.5
9.03
0.054
49.9
11.0
42.5
10.1
0.001
56.4
11.3
58.9
16.13
0.769
40.4
9.8
51.9
12.5
0.00014
1. The 2001/early 2002 and 2004 sample sets were only six samples per cell. All samples were collected on the same day so the
standard deviation is zero and a t-test is not appropriate. These samples were collected to characterize fresh refuse used to fill the
cells.
test cells were similar. As will be discussed in Chapter 5, the LFG collected in the Retrofit cells appears to
be greater than that collected in the Control cells, indicating that decomposition in the Retrofit cells may
have been accelerated relative to the Control cells.
With regard to the behavior of samples from the As-Built cells, the calculated average waste age in As-
Built cell A (730 days) and As-Built cell B (545 days) were significantly different and much lower than the
average age in the Control cell (1,656 days). The young waste age in the As-Built cells makes it difficult to
compare decomposition as trends would be expected to emerge over a period of five years or more.
Nonetheless, As-Built cell A, which received more injected air than Cell B, exhibited more rapid
decomposition (steeper slopes) than both As-Built cell B and the Control cell for all three measures of
decomposition (CH:L, BMP, organic solids). This was also consistent with the higher moisture content in
As-Built cell A (Figure 4-10). With respect to As-Built cell B, the linear regression slopes were not
significantly different from zero which means that there was not a statistically significant decrease in CH:L,
BMP or organic solids with waste age. Given the young age in As-Built cell B, it is difficult to document
accelerated solids decomposition on the basis of the solids data alone. It is expected that if additional
49
-------
sampling and testing is performed in subsequent years, consistent trends with regard to the solids data will
become apparent.
Finally, while the emphasis of this section has been to examine statistical evidence for accelerated
biodegradation in the bioreactor landfill cells, it is also important to consider visible trends. Trends over
time, in which the data were grouped by year, are presented in Figures 4-11 through 4-13. These figures
illustrate trends in CH:L, BMP, and organic solids for both the As-Built A and Retrofit cells that are
encouraging and support the supposition that the operation of a landfill as a bioreactor results in accelerated
solids decomposition. Thus, while the statistical analyses were not overwhelming, this is likely a result of
limited data, time period of sample collection, and waste heterogeneity, not the absence of accelerated
decomposition. This observation is made in full recognition that the data set from the OLLB is perhaps the
most comprehensive available to date. These data suggest the need for waste samples that are at least five
years old to document decomposition. Going forward, it may be cost-effective to collect a smaller set of
targeted samples of known waste age to provide information for a comparison of the effect of bioreactor
landfill operation on solids decomposition.
i-
0
1 o
o >j->
o _
o _
-------
n=14 n=104 n=13 n=67 n=23 n=62 n=0
O
cl 1-2 2-3 3-4 4-5 5-8 8-11
Waste Age (yr.)
Control cells
ffi en -
O
n=0 n=7 n=8 n=47 n=32 n=86 n=43
cl 1-2 2-3 3-4 4-5 5-8 8-11
Waste Age (yr.)
Retrofit cells
n=47 n=43 n=15 n=5 n=0 n=0 n=0
ffi
O
1-2 2-3 3-4 4-5 5-8 8-11
Waste Age (yr.)
As-Built cell A
ffi
O
n=49 n=54 n=8 n=4 n=0 n=0 n=0
1-2 2-3 3-4 4-5 5-8 8-11
Waste Age (yr.)
As-Built cell B
Figure 4-11 Trends in CH:L as a Function of Waste Age
51
-------
ra
n=14 n=103 n=13 n=67 n=23 n=62 n=0
cl 1-2 2-3 3-4 4-5 5-8 8-11
Waste Age (yr.)
s
1-1 rt
dn
ra
n=0 n=7 n=8 n=47 n=32 n=85 n=41
<1 1-2 2-3 3-4 4-5 5-8 8-11
Waste Age (yr.)
Control cells
Retrofit cells
gg
n=47 n=43 n=15 n=5 n=0 n=0 n=0
cl 1-2 2-3 3-4 4-5 5-8 8-11
Waste Age (yr.)
As-Built cell A
gg
1-1 rt
a
n=8 n=4 n=0 n=0 n=0
cl 1-2 2-3 3-4 4-5 5-8 8-11
Waste Age (yr.)
As-Built cell B
Figure 4-12 Trends in BMP as a Function of Waste Age
52
-------
•O
n=14 n=104 n=13 n=67 n=23 n=62 n=0
<1 1-2 2-3 3-4 4-5 5-8 8-11
Waste Age (yr.)
Control cells
00
6
n=0 n=7 n=8 n=47 n=32 n=86 n=43
cl 1-2 2-3 3-4 4-5 5-8 8-11
Waste Age (yr.)
Retrofit cells
8-
o _
fc. -
'O
xo" \O ~
-a
0
C/3 O
o ^
3
00
o 0
-------
4.1.4 Solid Waste Surface Settlement
Accelerating waste settlement in a landfill unit is one of the key objectives of operating a landfill as a
bioreactor. Moisture addition to the waste is intended to increase the rate of waste decomposition, resulting
in accelerated waste settlement (and therefore a gain in airspace). Surface elevation data was collected for
the different landfill units using GPS measurements. As discussed in Section 3.2.4, settlement in the
Control and As-Built cells could not be estimated, since these units received waste for most of the study
period, making an assessment of settlement as a function of waste decomposition alone difficult.
Additional waste placement is not planned for the test areas. Periodic monitoring of settlement will
continue at the site, and a discussion of settlement will be presented in the final report.
4.1.5 Landfill Temperature and Oxidation Reduction Potential (ORP)
As discussed in Chapter 3, instrumentation for measuring in-situ temperature of the waste and ORP was
installed in the Retrofit and As-Built cells. Because of an installation defect, the majority of the
thermocouples installed to measure temperature were heavily influenced by ambient temperature. As a
result, the data collected for the Retrofit and As-Built cells were not representative of the waste
temperature, and these data are not presented. Additionally, the majority of ORP probes returned data that
indicated the probes did not function correctly; as a result, an analysis of ORP data is not presented in this
report.
4.1.6 Solid Waste Slope Stability
A slope stability study for the Retrofit and As-Built cells was completed in March 2000. Results were
presented in a report prepared by Vector Engineering, Inc. (Vector) titled: Stability Analyses for Unit
7.4A/7.4B and Area 5 at the Outer Loop RDF, Louisville, Kentucky. For the stability analyses, both the
intermediate and final waste filling conditions were analyzed. Two potential stability failures were
considered. The first failure condition assumes the failure surface to be based within the leachate collection
system/geomembrane liner/compacted clay liner system interfaces. The second failure condition assumes
failure to occur in the subgrade soils beneath the cells.
For the intermediate waste filling conditions, waste grades of 3H:1V and 3.5H:1V (horizontal to vertical)
were analyzed for both static and pseudo-static conditions. For these analyses, a unit weight of 1041 kg/m3
(65 pcf) was assumed and a friction angle of 33° for the waste was used. For the final waste filling
conditions, the final design waste grade was assumed to be 4H: IV for both static and pseudo-static
conditions. For these final conditions, a unit weight of 1361 kg/m3 (85 pcf) and a friction angle of 25° for
the waste were assumed.
For the stability analyses in the Retrofit cells, both existing (i.e., before stockpiling) and proposed final
conditions (i.e., after stockpiling) were evaluated. The western slope of the stockpiled material intersected
the existing eastern slope of the Retrofit cells. Stability analyses were evaluated for both the eastern and
western side slopes under static conditions.
Based on the assumptions made in the stability study, it was concluded that selected waste grades for
intermediate and final conditions in As-Built cell A and As-Built cell B are stable under static and pseudo-
static conditions. Similarly, the existing and proposed conditions for waste grades in the Retrofit cell were
stable under drained and undrained static conditions.
Liquid addition to the Retrofit and As-Built cells was initiated early in 2002 and is ongoing. For the
monitoring period between 2002 and 2005, approximately 65,000 m3 (17.4 million gallons) and 105,000 m3
(27.8 million gallons) of liquid were introduced to the Retrofit and As-Built cells, respectively.
Approximately 20 L (5.3 gallons) of nitrified leachate per Mg of in-place waste was applied in the Retrofit
cells. Approximately 206 L (55 gal) of liquid per Mg of in-place waste was added to the As-Built cells.
No stability problems have been reported for any of the landfill units at the OLLB to date. A slope stability
analysis was not performed as part of this report. However, a follow-on study that utilizes data in this
report may be beneficial and provide valuable information as to the potential for stability problems. A
general discussion on slope stability is presented in Chapter 7
54
-------
4.2 Moisture Addition
The purpose of this section is to provide a summary of the moisture balance assessments and a discussion
of the calculated waste moisture content, liquid head on the liner, and measured moisture content within the
study cells. In this section, "calculated moisture content" refers to the moisture content as determined
using methods described in the U.S. EPA document Example Moisture Mass Balance Calculations for
Bioreactor Landfills (U.S. EPA 2005c). The "measured moisture content" is the percent moisture content
measured in the solid waste samples collected from the site throughout the study period (presented in
Section 4.1). The discussion includes an assessment of the moisture balance calculations in the cells
followed by a discussion regarding the leachate head in the LCS sump, which was used as an indicator of
head on the landfill liner.
4.2.1 Moisture Balance
A moisture balance for each of the landfill cells was performed. Moisture balance calculations were
performed for the Control, Retrofit, and As-built cells for the time period ranging from March 2002 to
December 2005. The bulk waste moisture content was calculated for each of the landfill units.
Furthermore, this calculated moisture content was compared to the one measured from field-collected
samples of waste. The addition of leachate and liquids to the Retrofit and As-Built cells was expected to
increase the moisture content of the waste. Field-measured moisture content was expected to exhibit large
variability because of the spatial variation in waste hydraulic properties (i.e., waste heterogeneity) and the
non-uniform wetting of the waste. It is noted that the method of solid waste sampling described in Chapter
3 may have impacted the liquid distribution in the bioreactor landfill cells. The advancement of the 0.9-m
boreholes may have drilled through perched liquid zones, possibly leading to an improved liquid
distribution. The degree, to which this occurred, if at all, is unknown.
For the moisture balance presented in this section, the "Liquid In" value is the summation of the estimated
infiltration from rainfall, recirculated leachate, and added liquids. The procedure for incorporating
infiltrated rainfall into the moisture balance was described in Chapter 3. The "Liquid Out" value is the
leachate collected at the leachate collection sump. The difference between Liquid In and Liquid Out, AS,
represents the change in moisture content due to liquids introduction. The bulk waste moisture content is
calculated as follows:
AS
Moisture content = M H xlOO Equation 4-2
1 Mw
Where:
M! = Initial moisture content (%);
AS = Change in storage (tons of water); and
Mw = Wet weight of waste (tons).
Equation 4-2 was used to assess the moisture balance of the landfill units, and calculate moisture content of
the waste for each landfill unit.
4.2.1.1 Precipitation
Rainwater infiltration into the solid waste is typically governed by the type of cover material and vegetation
on the surface of the waste. At the OLLB, a vegetative cover was not constructed with the exception of the
Retrofit cells. Precipitation data was collected by an on-site weather station. Figure 4-14 presents the
mean monthly precipitation and the cumulative precipitation that was recorded at the site during the period
from March 2002 to December 2005. The annual average precipitation at the OLLB is about 100 cm (40
in.). However, it is important to note that evapotranspiration is believed to be relatively high, ranging from
40 to 70 percent (based on HELP input parameters) depending on surface soil conditions.
55
-------
&
o
2
a
150 -
100 -
50 -
10 -
Cumulative Precipitation
Mean Monthly Precipitation
180
- 160
- 140 .S
- 120 •&
- 40 Q
5000
- 2500 g
2000 >
I
- 1500 1
3
1000 °
- 500
3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 4-14 Cumulative and Mean Monthly Precipitation
4.2.1.2
Control Cells
The Control cells did not receive any supplemental liquids addition during the study. Control cell A and
Control cell B were combined in this analysis because there was no hydraulic separation between the cells.
However, since the cells were filled until 2004, infiltration from precipitation occurred during the
monitoring period. The moisture balance and the calculated bulk moisture content of the waste in the
Control cells are presented in Figure 4-15. The cumulative infiltrated volume reached about 3,000 m3
(800,000 gallons) by December 2005 in this area. Infiltration to the Control cells was relatively low after
waste acceptance ceased. The leachate generation in the Control cells was also relatively low and steady
compared to the other landfill units during the monitoring period. Figure 4-15 shows no general seasonal
patterns and indicates no general trends during the monitoring period. Prior to the middle of 2003, the
calculated value of Leachate Out was higher than Leachate In, resulting in a net decrease in the calculated
bulk waste moisture content from 32 to about 29.5 percent. However, after mid-2003, the rate of Leachate
Out decreased, resulting in a relatively constant calculated bulk moisture content of about 29.5 percent.
With infiltration being almost the sole source for Liquid In for the Control cells, the calculated moisture
content showed a slight decrease over the study period. This decrease may be attributed to the small
amount of infiltration that occurred following the completion of waste filling, as well as the increase in
vertical stress from added waste, causing compression and release of the pore liquids. During the study
period, the calculated mean and standard deviation of Liquid In per Mg of in-place waste for the Control
cells were 2.8 liters/Mg (0.7 gallons/ton) and 1.5 liters/Mg (0.39 gallons/ton), respectively. The measured
mean and standard deviation of Liquid Out per ton of in-place waste were 22.7 liters/Mg (6.0 gallons/ton)
and 7.8 liters/Mg (2.1 gallons/ton), respectively.
56
-------
50
-
o
o 40
O 20
O Moisture Content
^^— Liquid Out
~ ~ • Liquid In
- 4e+7
§
2e+7
O
0
- 1.6e+5
- 1.4e+5 "S
o
- 1.2e+5 £
- l.Oe+5
- 8.0e+4 >
- 4.0e+4 ts
3
- 2.0e+4
O
3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 4-15 Moisture Balance and Calculated Waste Moisture Content of the Control Cells
4.2.1.3 Retrofit Cells
The Retrofit cells received nitrified leachate as well as the infiltrated liquids. There was no addition of
waste into the Retrofit cells during the monitoring period. The moisture balance and the calculated bulk
moisture content of the waste in the Retrofit cells are presented in Figure 4-16. Cumulative Liquid In
reached about 76,000 m3 (20.2 million gallons, approximately 15% infiltration) by December 2005, while
cumulative Liquid Out was steady during the monitoring period and reached about 62,000 m3 (16.2 million
gallons) by December 2005. Infiltration volume for the Retrofit cells was relatively low compared to the
As-Built cells because of the clay cover. Figure 4-16 shows no general seasonal patterns and indicates no
cyclical trends during the monitoring period. Since Liquid In and Liquid Out were nearly identical, there
was a small change in the calculated moisture content. The calculated bulk waste moisture content
increased from approximately 37 to 38 percent as of December 2005. During the monitoring period, the
calculated mean and standard deviation for Liquid In per Mg of in-place waste for the Retrofit cells were
20 liters/Mg (5.3 gallons/ton) and 9.5 liters/Mg (2.5 gallons/ton), respectively. The mean and standard
deviation of the measured Liquid Out per ton of in-place waste were 13.5 liters/Mg (3.6 gallons/ton) and
8.8 liters/Mg (2.3 gallons/ton), respectively.
4.2.1.4
As-Built Cells
The As-Built cells received moisture through the addition of liquids and infiltration from precipitation.
Liquids addition and infiltration occurred as waste was being placed as presented in Figure 4-17. The
moisture balance and the calculated bulk moisture content of the waste in As-Built cell A and B are
presented in Figure 4-17 and Figure 4-18, respectively. Infiltration volumes in the As-Built cells were the
highest compared to the other landfill units. The primary reasons for the high infiltration volume were that
waste placement occurred until 2005 and relatively high permeability compost was used as daily cover, in
addition to the large volume of run-on because of the location of As-Built cells relative to the other units.
The location of the As-Built cells resulted in a high percentage of run-on volumes to the As-Built cells,
particularly from Unit 7.2 toward As-Built cell A and from the Control cells to As-Built cell B. The
infiltration volumes comprised about 65 percent of the total liquid added to each of the As-Built cells.
Cumulative Liquid In was greater than that of Liquid Out in the As-Built cells resulting in an increase in
the calculated bulk moisture content of the waste.
57
-------
60
§
O
o
40 -
O Moisture Content
^^™ Liquid Out
— — - Liquid In
- 4e+7
- 3e+7
2e+7
- 1.6e+5
- 1.4e+5
- 1.2e+5
- l.Oe+5
- 8.0e+4
- 6.0e+4
- 4.0e+4
- 2.0e+4
- 0.0
V3
U
"S
S
o
£
3
o
fD
3
*O
|>
T3
'3
.^
(D
>
•£2
ed
3
g
a
Date
Figure 4-16 Moisture Balance and Calculated Waste Moisture Content of Retrofit Landfill Unit
For As-Built cell A, cumulative Liquid In reached about 152,000 m3 (40.2 million gallons, approximately
62% infiltration) by December 2005, while cumulative Liquid Out reached about 80,000 m3 (20 million
gallons). Figure 4-17 shows no general seasonal patterns and indicates no cyclical trends during the
monitoring period. A fluctuation in the calculated waste moisture content was observed in As-Built cell A.
A rapid increase in the moisture content was noted around the middle of 2002 when the total waste mass
was generally unchanged while the liquids introduction rate remained high. During the monitoring period,
the mean and standard deviation of the calculated Liquid In per Mg of in-place waste for As-Built cell A
were 247 liters/Mg (65.3 gallons/ton) and 47.2 liters/Mg (12.5 gallons/ton), respectively. The mean and
standard deviation of the measured Liquid Out per Mg of in-place waste were 175.8 liters/Mg (46.5
gallons/ton) and 46.1 liters/Mg (12.2 gallons/ton), respectively.
60
S
c
o
O
2 40
n
'o
S
^ 30 -
U 20
- 4e+7
9?^ cfcP^
O Moisture Content
— Liquid Out
' — • Liquid In
le+7 g
O
- 1.6e+5
^
• L4e+5 £
e
o
- 1.2e+5 'S
o
- l.Oe+5 |
"o
- 8.0e+4 >
T3
'3
a*
- 6.0e+4 j
- 4.0e+4 5
3
2.0e+4 (3
3/1/02
9/1/02
3/1/03
9/1/03
3/1/04
Date
9/1/04 3/1/05
Figure 4-17 Moisture Balance and Calculated Waste Moisture Content of As-Built Cell A
58
-------
For As-Built cell B, cumulative Liquid Out was relatively constant during the monitoring period.
Cumulative Liquid In reached about 147,000 m3 (39.0 million gallons, approximately 68% infiltration) by
December 2005, while cumulative Liquid Out reached about 80,000 m3. Figure 4-18 shows no general
seasonal patterns and indicates no general trends during the monitoring period. The calculated bulk waste
moisture did not show large fluctuations which increased from 43 percent to about 50 percent by December
2005. During the monitoring period, the calculated mean and standard deviation of Liquid In per Mg of in-
place waste for the As-Built cell B were 164.0 L/ton (43.4 gallons/ton) and 52.9 L/ton (14 gallons/ton),
respectively. The measured Liquid Out per ton of in-place waste mean and standard deviation were 106.2
L/ton (28.1 gallons/ton) and 34.0 L/ton (9.0 gallons/ton), respectively. The mean and standard deviation of
the calculated bulk moisture content for As-Built cell B were 47.8 percent and 2.3 percent, respectively.
It is important to note that the calculated net increase in bulk moisture content in the As-Built cells above
was conservative, based on the high liquid addition volumes. As described in Chapter 3, a layer of
shredded tires was used between As-Built cell A and As-Built cell B during the construction of these cells
until the first quarter of 2004, when it was noted that the shredded tires were acting as a conduit for
leachate to drain quickly into the leachate collection system thus the use of shredded tires was discontinued.
As shown on Figures 4-17 and 4-18, approximately 66% of the total leachate generation during the study
period occurred prior to the first quarter of 2004 in the As-Built cells. Furthermore, the difference between
the rates of increase for Liquid In versus Liquid Out is much more pronounced after the first quarter of
2004. Therefore, a portion of the Liquid Out used in the moisture balance calculation was likely
overestimated, resulting in an underestimation of the AS.
cS
o
"
o
o
O
I
-8
cS
O 20
50 -
40 -
30 -
- 4e+7
O Moisture Content
~~ Liquid Out
• — • Liquid In
le+7 g
3
O
- 1.6e+5
- 1.4e+5
- 1.2e+5
- l.Oe+5
- 8.0e+4 I
- 6.0e+4 ;
- 4.0e+4 ;
- 2.0e+4 (
- 0.0
a
o
IS
o
3/1/02
9/1/02
3/1/03
3/1/04
Date
9/1/04
3/1/05
9/1/05
Figure 4-18 Moisture Balance and Calculated Waste Moisture Content of As-Built Cell B
4.2.2 Leachate Head on Liner
Under RCRA regulations, the head on the bottom liner of an MSW landfill should not exceed 0.3 m (1 ft).
The main purpose of this regulation was to reduce risks associated with leakage of leachate from a landfill
unit that could cause groundwater contamination. The potential for leachate head to build up on the bottom
liner typically correlates with leachate generation, given that greater leachate generation rates would
logically increase the potential for leachate accumulation of liquids on the liner. As mentioned previously,
the leachate head presented in this section is the leachate head measured in the sump and not on the bottom
liner itself. Caution must be taken when examining head on the liner data in this section since it is
recognized that leachate head in the sump can be influenced by the size of the pump used to convey
leachate from the leachate collection system, as well as the pump control system. Furthermore, the leachate
59
-------
head level in the sump may be affected by precipitation events (e.g., runoff from the landfill into either the
sump or the vaults connected to the sump). The OLLB was permitted to have the leachate head in the sump
temporarily exceed the regulatory limit of 0.3 m on days following heavy rainfall events, per the site's
permit issued by Kentucky. Recognizing that the Control cells were considered as a whole in the moisture
balance, leachate head in the sump of Control cells A and B were averaged for this assessment. Similarly,
leachate head measured in the sump in the Retrofit cells was also averaged for these calculations.
4.2.2.1
Control Cells
The Control cells received liquids from infiltration only (i.e., the cells did not receive any supplemental
liquids). Thus, leachate in the collection system is expected to be generated from infiltration and trends of
leachate head are expected to correlate well with precipitation. The mean monthly leachate head in the
Control cells is presented in Figure 4-19. The mean monthly leachate head was mostly below the
regulatory limit. The mean and standard deviation of leachate head in the sump were 14 and 7.4 cm (5.5
and 2.9 in.), respectively. The leachate head and generation were generally higher prior to mid-2003, while
waste was still being placed in the Control cells. Once waste placement ceased in 2003, both leachate head
measurements and leachate generation decreased, as shown in Figure 4-20. The higher leachate head and
generation rates prior to mid-2003 were likely caused by a portion of the runoff from the landfill being
routed to the sump.
o
03
ffi
O
03
40 -
30 -
20 -
10 -
>1
53
o
14 -
o
.a 12
03
ffi 10
•S
03
03
4-
2-
Leachate Head
Regulatory Limit
3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 4-19 Mean Monthly Leachate Head on Liner of the Control Cell
60
-------
40 H
«•> 20 -
o
10 -
16
"S
o
o
- 50000
o
40000 o
o
30000 J
"o
- 20000
O
- 10000 =
31/02 91/02 31.03 91/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 4-20 Mean Monthly Leachate Head and Leachate Volumes Generated in the Control Cell
4.2.2.2
Retrofit Cells
The Retrofit cells received moisture through the addition of nitrified leachate (beginning in March 2002)
and infiltration from precipitation. The clay cap that was placed on the Retrofit cells resulted in a low
infiltration volume relative to the As-Built cells. Since there was no hydraulic separation between the
Retrofit cells with respect to liquid addition, the leachate head values as recorded from Retrofit cells A and
B were averaged. Since the pressure transducers are located in the sumps of all cells, it is necessary to
account for the elevation head in the sump. In the case of the Retrofit cells, the leachate sump was located
at an elevation that added a pressure head of approximately 97 cm (40 in.). Thus, an "equivalent regulatory
limit" is noted on Figure 4-21. The equivalent regulatory limit accounts for the additional 97 cm (40 in.)
added to the transducer's measurement caused by the higher elevation of the sump. The mean monthly
leachate head in the Retrofit cells was generally below the regulatory limit. The mean and standard
deviation of the leachate head were 95 cm and 25 cm (37.4 and 9.8 in.), respectively. The leachate head in
the Retrofit cells was considerably higher than in the Control cells. The leachate head in the Retrofit cells
is expected to be greater than in the Control cells since the Retrofit cells received additional liquids. The
mean monthly leachate head correlates well with volumes of generated leachate in the Retrofit cells as
presented in Figure 4-22. A temporary increase in leachate head correlated well with an increase in
leachate generation towards the end of 2002 and at the beginning of 2005, as shown in Figure 4-22.
4.2.2.3
As-Built Cells
The As-Built cells received moisture through the addition of industrial liquids and infiltration from
precipitation. Infiltration volumes for the As-Built cells were higher than those in the Control and Retrofit
cells. This was the result of high infiltration allowed by the compost cover used for the As-Built cells
during waste filling, as well as the higher volume of run-on from surrounding cells. Figure 4-23 and Figure
4-24 present the mean monthly leachate head in As-Built cells A and B, respectively. The mean monthly
leachate head in the As-Built cells was below the 30 cm (1 ft) leachate head regulatory limit and similar to
values measured in the Control cells. For As-Built cell A, the leachate head mean and standard deviation
were 16 cm (6.4 in.) and 6.9 cm (2.7 in.), respectively. For As-Built cell B, the leachate head mean and
standard deviation were 14.5 cm (5.7 in.) and 6 cm (2.35 in.), respectively. When compared to the Control
cells, the addition of industrial liquids and the relatively large infiltration volumes did not have a significant
effect on the head measured in the sump - Figures 4-19, 4-23, 4-24 indicate the number of times the 0.3 m
(1 ft) threshold was exceeded (four, four, and one in the Control and As-Built cells A and B, respectively)
during the study period. This is likely due to effective drainage of leachate in the cells'
61
-------
200 -
IT
^a 15° -
K
"eS
^
%
jo 100 -
>->
£
"3
o
s
§ 50 -
o>
S
n -
ou
^
-S
o
^ 60 -
ITH
h-H
"eS
1 40-
hJ
-^
J^
"^
O
^ 20 -
fl
n -
1 1 Leachate Head
T-1 • 1 J. T~> 1 J. T '
PI i-i
n
-
nfl
PI n nnnfln
mil ri
n
i-i
1 H nn
n
3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 4-21 Mean Monthly Leachate Head on Liner for the Retrofit Cells
40 -
30 -
20 -
10 -
1.4e+6
- 50000
o
40000 o
(D
30000 |
I
-o
- 20000
a
O
- 10000 ,
0
3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 4-22 Mean Monthly Leachate Head and Leachate Generated Volumes in Retrofit Cells
62
-------
40 -
a
-a" 30 "
hSn
1)
ta
o
C3
-------
40 H
Q
^ 30
rt
u
o
a
^
Q
20 -
§ 10-
0 J
16
14 -
• Leachate Head
I Leachate Generated
- 50000
o
40000 o
30000 |
I
o
c
-------
4.2.3 Measured Waste Moisture Content
This section describes statistical analyses that were conducted to evaluate moisture content trends with
respect to waste age and sampling date. The moisture content of the waste was assessed on samples
collected from each of the landfill units during the study period, as described in Section 4.1. The frequency
of solid waste sample collection varied between the different landfill units. Samples collected from the
Control and Retrofit cells in 2000 were included in this analysis. Waste age and moisture content were
assessed using time plots and running linear regression fits of the data. Moisture content and sampling data
were evaluated using box plots and performing multiple comparisons using Tukey's procedure. Details of
the statistical analysis of these two methods are provided in Appendix C.
4.2.3.1 Measured Moisture Content versus Waste Age
Linear regression fits and time plots were used to evaluate the measured moisture content versus age and
are presented in Appendix D. Inspection of the time plots indicates large variability in moisture content in
all cells. This is also evidenced in the low adjusted-R2 values, even though waste age is a significant
predictor for all cells. The linear regression fits indicate that all cells had an increasing trend in moisture
content with waste age. However, only the Retrofit cells' time plots indicate a definitive linear trend.
Linear regression assumptions may not be met in the Control and As-Built cells.
The time plots for the As-Built cells appear to show moisture content levels increasing and then decreasing
as waste age increases. It appears that the collection of additional older waste samples is needed to
determine if a trend exists. The diagnostic plots show the residuals were not random in the As-Built cells.
The diagnostic plots also show the residuals may not have been normally distributed in As-Built cell B.
These characteristics make it difficult to definitively conclude if there was an increasing trend in moisture
content with waste age in the Control and As-built cells.
4.2.3.2 Measured Moisture Content versus Sampling Date
For the analysis of measured moisture content versus sampling date, it was assumed that all sampled waste
is similar for a specific sampling date and that the age difference in sampled waste between sampling dates
is equal to the difference in sampling dates. One violation of this assumption would be if waste sampled in
2002 is actually older than waste sampled in 2003. Analysis of variance (ANOVA) summary tables and
box plots were used in this analysis and are presented in Appendix C.
The ANOVA summary tables present the F-test of the null hypothesis that all means are equal versus the
alternative hypothesis that there are at least two significantly different means. The other part of the
summary tables provides Tukey multiple comparisons (i.e., comparisons of mean moisture content between
each pair of sampling dates). A multiple comparison method is necessary to control the overall
significance level since more than one significance test is being performed. Box plots are also provided to
allow a qualitative assessment of each trend with sampling date.
For the Control cells, the F-test p-value indicates that at least two mean moisture content levels were
significantly different. The Tukey multiple comparisons analysis indicates that mean moisture content
levels were relatively constant across sampling dates except for the drop in 2002. In addition, the box plots
did not provide evidence of a trend. Therefore, there is no evidence of a trend in the measured mean
moisture content with sampling date in the Control cells.
For the Retrofit cells, the F-test p-value indicates that at least two mean moisture content levels were
significantly different. The Tukey multiple comparisons indicated that the 2002 and 2005 mean moisture
content levels were equal and both are significantly greater than the 2000 mean moisture content. The box
plots did not provide any evidence of a trend. These results indicate there is no trend in the mean measured
moisture content with sampling date in the Retrofit unit.
For As-Built cell A, the F-test p-value indicates there were no differences between mean moisture content
levels. The box plots also indicated no differences present. Therefore, there is no evidence of a trend in the
mean measured moisture content with sampling date in the As-Built cell A.
For the As-Built cell B, the F-test p-value indicates that at least two mean moisture content levels were
significantly different. The Tukey multiple comparisons and the box plots indicate the 2003 mean was
65
-------
higher than the 2002 mean and the 2005 mean. Therefore, there is no evidence of a trend in the mean
measured moisture content with sampling date in the As-Built cell B.
4.2.4 Evaluation of Calculated and Measured Moisture Content
Because solid waste samples analyzed for moisture content were measured for a given year with collection
obtained in a single month, it was decided to compare these measured values to the calculated values for
the entire given year. The calculated moisture content showed no patterns within a given year; therefore,
the entire year is considered comparable to the measured values collected in a single month of that year.
A qualitative comparison of the distributions by year consisted of summary statistics and box plots. A
quantitative comparison of the distributions by year consisted of a two-sample t-test and a nonparametric
Wilcoxon Rank Sum test. The qualitative comparison is of the whole distribution while the quantitative
comparison is of the mean of the distributions. The qualitative and quantitative comparison of the
distributions is included in Appendix D.
Both the qualitative and quantitative comparisons indicate that measured moisture content was higher than
the calculated moisture content for all years in the Control cells. All comparisons indicate that the
measured moisture content and calculated moisture content were equivalent in both 2002 and 2005 in the
Retrofit cells. The box plots and hypothesis tests indicate that the average calculated moisture content was
higher than measured moisture content in the As-Built cells in all years except for As-Built cell B in 2002.
While the box plots indicate the measured distributions span was smaller than the calculated distributions,
the general conclusion must be that calculated moisture content was higher than measured moisture content
in the As-Built cells.
A statistical summary including: (i) number of samples; (ii) mean moisture content; (iii) standard deviation;
and (iv) coefficient of variation is presented in Table 4-5. The relatively large standard deviation in the
measured waste moisture content is believed to be caused by a combination of the heterogeneous nature of
the waste, the spatial variation in waste composition, and the varying macro-pore structure and resulting
hydraulic properties within a given cell.
As shown in Figure 4-27, there was an overall increase in the mean measured moisture content in the
Control cells of about five percent during the monitoring period, while the calculated moisture content
showed a slight decrease. The amount of measured increase is unlikely to have occurred since the Control
cells did not receive any addition of liquids. This may indicate that calculated values of the bulk waste
moisture content are more accurate than the measured moisture content of discrete solid waste samples, or
may be indicative of the inherent variability in the waste moisture content in different locations within the
area. The mean measured waste moisture content showed an increase of about 2 percent for the Retrofit
cells during the monitoring period. Figure 4-28 indicates that there was good correlation between
calculated and measured waste moisture content in the Retrofit cells. This may have been due to the well-
defined geometry of the Retrofit cells that resulted in relatively more accurate estimates of rainwater
infiltrated volumes. However, it is recognized that the sampling frequency of solid waste samples in the
Retrofit cells was relatively low as compared to other units.
In the As-Built cells, the measured moisture content revealed large spatial variation. The distribution of
injected liquids will likely result in relatively wetter areas where the injected liquids follow preferential
flow paths and accumulate, and the addition of air may have formed dry areas within the waste and
removed some moisture via aerobic decomposition of the waste. The difference in calculated and
measured moisture content of the As-Built cells may be caused by the assumptions taken for calculating the
bulk waste moisture content. Infiltration and run-on volumes (estimated using HELP) to estimate the
moisture balance of these cells in specific portions of the cells were difficult to estimate due to cell
geometry. The measured waste moisture content showed large fluctuations with an average decrease of
about 2 percent in the mean waste moisture content, as shown in Figure 4-29 (As-Built cell A) and Figure
4-30 (As-Built cell B). A decrease in the moisture content of the As-Built cells is doubtful because of the
addition of relatively large liquid volumes into these cells. The calculated bulk waste moisture content
appears to be a more realistic representation of the As-Built moisture content than the measured moisture
content from waste samples taken from those cells.
66
-------
As expected, there was lower variability in the measured moisture content of the Control cells as compared
to the Retrofit cells and As-Built cells (refer to Table 4-1) since no additional liquids were added. Leachate
generation and liquid addition to the Retrofit and As-Built cells resulted in large spatial variation of the
moisture content, likely due to the non-uniform wetting of the waste. Thus, the variability in the measured
moisture content in the bioreactor landfill cells was greater than that in the Control cells.
Table 4-5 Summary of Calculated and Field Measured Waste Moisture Content
Is
a o
a°
^H W
2^
2°
|<:
pp'S
^0
i«
pp'S
^0
ParametersVDate
Number of Samples
Mean Measured Waste MC + Stdev
Coefficient of Variation (percent)
Mean Calculated Bulk Waste MC +
Stadev Coefficient of Variation (percent)
Number of Samples
Mean Measured Waste MC + Stdev
Coefficient of Variation (percent)
Mean Calculated Bulk Waste MC +
Stadev Coefficient of Variation (percent)
Number of Samples
Mean Measured Waste MC + Stdev
Coefficient of Variation (percent)
Mean Calculated Bulk Waste MC +
Stadev Coefficient of Variation (percent)
Number of Samples
Mean Measured Waste MC + Stdev
Coefficient of Variation (percent)
Mean Calculated Bulk Waste MC +
Stadev Coefficient of Variation (percent)
2002
70
32.5 + 5;
15.3
31.5+0.4;
1.2
84
37.0+7.5;
20.1
37.7+0.2;
0.5
24
41.9 + 9.2;
21.9
49.3+3.4;
6.9
31
41.6 + 9.0;
21.6
44.2+1.1;
2.5
2003
73
36.1+5.9;
16.3
29.7 + 0.4;
1.3
NA
NA
37.6 + 0.1;
0.2
42
41.1+8.3;
20.2
46.1+2.3;
5.0
56
48.6+16;
33.5
47.6 + 1.7;
3.6
2005
60
38 + 5.3;
13.9
29.6 + 0.1;
0.3
64
38.8 + 6.6;
17.0
37.7 + 0.1;
0.2
32
39.9 + 7.6;
19.0
50.3 + 1.2;
2.4
31
39 + 9.6;
24.6
49.1+0.5;
1.0
1. NA = Not Applicable.
60
50 -
g
&
I 40
o
O
20
O Calculated Bulk Moisture Content
• Mean Measured Moisture Content
3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 4-27 Calculated and Measured Waste Moisture Content in the Control Cells
67
-------
60
50 -
40 -
U
22
S 30 "
-------
§ 40 H
I
'o
^ 30 -
O Calculated Bulk Moisture Content
• Mean Measured Moisture Content
3/1/02 9/1/02 3/1/03 9/1/03 3/1/04 9/1/04 3/1/05 9/1/05
Date
Figure 4-30 Calculated and Measured Waste Moisture Content in As-Built Cell B
4.2.5 Moisture Content Analysis Summary
A moisture balance of the different landfill units indicated an increase in the calculated bulk waste moisture
content in the Retrofit and As-Built cells. The time plots for the calculated bulk waste moisture content
showed no general seasonal patterns and indicated no general trends during the monitoring period. A
decrease in the moisture content of about two percent was calculated for the Control cells, as a result of a
larger Liquid Out volume versus Liquid In. By the end of December 2005, the calculated waste moisture
content increased by about one percent in the Retrofit cells. The rate of increase in moisture content was
relatively consistent over time. There was a relatively large variation in the calculated waste moisture
content of the As-Built cells, as a result of the various assumptions considered in the estimate of infiltrated
volumes into these cells, the ongoing waste placement in the cells, and the effects of air addition to the
waste. By the end of December 2005, the calculated bulk waste moisture content had increased by six to
seven percent in As-Built cell A and B, respectively. In making moisture balance computations, several
estimates had to be made regarding the factors that influence infiltration. While it is believed that a net
increase in moisture content occurred in the Retrofit and As-Built cells, it is difficult to verify the
magnitude of the phenomenon, even moisture content data from solid waste samples.
Leachate head in the sump of the different landfill units was consistent with trends of leachate generation.
Leachate head measurements were generally below regulatory limits and correlated well with trends of
leachate generation. Overall, the mean leachate head measurements for all cells were below the regulatory
limit of 30 cm (1ft).
Trends in measured moisture content were assessed in two ways - moisture content versus waste age and
moisture content versus sampling date. The first method provided some evidence of a trend in the
measured moisture content with waste age, but the second method did not provide evidence of a trend in
the mean measured moisture content with sampling date. The measured waste moisture content revealed a
large variability because of the spatial variation in the waste. This variability was greater in the Retrofit
cells and the As-Built cells compared to the Control cells.
The calculated bulk moisture content appeared to provide a more realistic estimation of the actual moisture
content increase compared to the measured moisture content data. It appeared that on-going waste filling
and the construction of tire chip layers at the edges of the As-Built cells reduced the magnitude of the
expected moisture content increase, based on the estimated infiltration volume and the liquid addition
volume. Overall, the results emphasize the large liquid volumes required to increase the moisture content
of a landfill to the level cited in the literature as ideal for anaerobic waste decomposition.
69
-------
This page is intentionally left blank
70
-------
Chapter 5. Landfill Gas (LFG)
LFG quality and quantity were measured on an approximately weekly basis in accordance with procedures
identified in Section 3 and the QAPP. This section consists of the following: (i) presentation of the LFG
composition and flow data; (ii) comparison of field data and predicted LFG generation based on U. S.
EPA's LandGEM at various waste decay rates; (iii) comparison of LFG flow data between the Control cells
and the landfill bioreactor cells; (iv) presentation of non-methane organic carbon (NMOC) concentration
data; (v) calculation of NMOC production based on LFG flow data; and (vi) calculation of NMOC
production based on in-place waste data using LandGEM.
5.1 Landfill Gas (LFG) Composition
The quantity and composition of LFG is anticipated to vary with the stage of solid waste decomposition.
However, since solid waste decomposition was believed to be occurring predominantly under anaerobic
conditions at the time the gas measurements were obtained, LFG at the OLLB was expected to consist of
approximately 55 percent by volume CH4, 40 percent CO2, 5 percent N2, and trace amounts of NMOCs
(U.S. EPA, 1998). Often however, LFG is assumed to consist of only methane and carbon dioxide at 50%
ratio. LFG was monitored on site using a LandGEM 2000 landfill gas analyzer. Sampling was performed
approximately once per week during active gas collection for each cell. In accordance with the QAPP,
triplicate gas readings were made for each sampling event at each sampling point. In this analysis, the
mean of the triplicate readings is presented, and samples in which the gas flow was zero were eliminated
from the analysis.
5.1.1 Control cells
As presented earlier in Chapter 3, LFG collection sampling at the Control cells began in January 2002. The
data and analysis presented here show gas samples collected through February 2006. Gas samples were
analyzed at two gas collection wells (GW) for each cell (GW A-l, GW A-2, GW B-l, and GW B-2). The
number of triplicate sets of sample data evaluated varies between 116 (Control cells A-2 and B-2) and 147
(Control cell A-l). Summary statistics for the field gas samples taken from the Control cells are shown in
Table 5-1. Statistically, the LFG composition was similar in each of the Control cells (GW A-l and 2, B-l
and 2) and across duplicate cells (Control cells A and B).
The data presented in Table 5-1 shows LFG concentrations indicative of anaerobic decomposition;
furthermore, the data are generally consistent (i.e., there was minimal variation in gas composition
throughout the sampling period). To show the variation of gas composition over time within the Control
cells, data collected between 2002 and 2006 were aggregated for each month. The results are shown in
Figures 5-1 and 5-2 for Control cells A and B, respectively. The figures indicate that the gas composition
of the primary LFG constituents was generally consistent throughout the sampling period. Additionally,
the concentrations of O2 and balance gas (i.e., N2) were generally low for both cells throughout the
sampling period, indicating minimal air intrusion and proper balancing of the gas collection system. It is
also noted that the sampling location (well heads) may have played a role in minimizing the O2 and N2
concentrations by minimizing the effects of breaks and leaks in the header piping system.
5.1.2 As-Built cells
In As-Built cells A and B, LFG was sampled from the LFG extraction header pipe of each subcell starting
in April 2003. Summary of gas composition statistics for As-Built cells A and B are presented in Table 5-
2. Figures 5-3 and 5-4 summarize the temporal variation of the data presented in Table 5-2, and indicate
substantial variability, particularly in comparison to the relatively consistent results from the Control cells.
This variability is unusual given air injections into the cell. Potential sources of the erratic behavior include:
(i) effects of air injection activities in the cells; (ii) leakage and air intrusion around the header pipes; (iii)
air intrusion through the cover; and (iv) "over- pulling" of the well field.
71
-------
Table 5-1 Summary Statistics for Gas Composition for Control Cells A and B
Sub Cell
i~*\\r A i
LrW A-l
fN — 1 4.7s!
LrW A-z
fN — 1 1 6s!
LrW B-l
/•1SJ — llf^
f~^\\7 T3 1
LrW J3-Z
/•TXJ — l^rn
Parameter
Mean ± Stdev
Median
Range
Mean ± Stdev
Median
Range
Mean ± Stdev
Median
Range
Mean ± Stdev
Median
Range
CH4
(%)
58±3
57
45-65
58±34
57
48-68
58±3
58
46-66
57±2
57
43-62
C02
(%)
41±3
42
31-44
42±3
42
33-48
41±2
41
34-43
41±1
41
38-44
02 (%)
0.6±1
0.1
0-4.6
0.2±0.3
0
0-2.4
0.4±0.7
0.1
0-3.7
0.2±0.4
0.1
0-1.9
Balance
(%)
0.5±3
0.2
0-18
0±3
0.2
0-11
1.3±3
0.2
0-15
0.8±2
0.3
0-10
Note: Each sample represents the mean of triplicate readings collected at gas collection well
70
60 -
50 -
o
40
O
S 30
o
20 -
10 -
Figure 5-1 Monthly Average of Gas Composition for Control Cell A
72
-------
70
60 -
50 -
o
•-S 40 H
O
O
O
O
30 -
20 -
10 -
CH4
C02
°2
Balance
Figure 5-2 Monthly Average of Gas Composition for Control Cell B
The cause of this erratic behavior is likely a combination of the effects of air injection and sampling
location. LFG sampling in the As-Built cells occurred in a header pipe, compared to the well heads in the
Control cells, which allowed for leaks to play a role in LFG composition. The second reason is the
aerobic/anaerobic operation of these cells. While air injection occurred at various times throughout the
study period, it is likely that the months where more air was injected had an effect on the overall LFG
composition. For example, in As-Built cell A, large-scale air injection activities occurred at various
occasions as presented in Table 5-3. Figure 5-3 shows reduced CH4 and CO2 and elevated O2 and balance
gas during the same month where large air injection occurred. Large-scale air injection activities occurred
in As-Built cell B. Again, Figure 5-4 shows that the CH4 and CO2 percentages were reduced, while the
balance gas and O2 concentrations were elevated at these times. It is interesting to point out that the O2 to
N2 ratio during these times (assuming balance gas consists of only N2) is (1 to 4) similar to that in ambient
air.
Table 5-2 Summary Statistics of Gas Com
osition for As-Built Cells A and B
Sub Cell
Cell A
fN — 197s!
/~^11 TD
L.C11 D
fN" — 1?^
Parameter
Mean ± Stdev
Median
Range
Mean ± Stdev
Median
Range
CH4
(%)
51±11
55
21-63
52±9
54
5-82
CO2
(%)
37±6.4
39
18-45
37±6
39
4.2-45
02
(%)
2.4±3
0.9
0-12
2.1±3
1.1
0-18
Balance
(%)
9.7±13
2.9
0.1-48
9±11
5.5
0.1-73
Note: Each sample represents the mean of triplicate readings collected at gas collection header
73
-------
o
1
+J
(L>
O
C
O
O
O 20-
10 -
Figure 5-3 Monthly Average of Gas Composition for As-Built Cell A
Table 5-3 Notable Air In ection Dates for As-Built Cells A and B
Cell
As-Built cell A
As-Built cell B
Month
11/2003
6/2004
1/2005
8/2003
7/2004
8/2005*
Air Injected, m3 (ft3)
2.8xl05 (9.9xl06)
2.1xl06(7.5xl07)
I.lxl06(3.9xl07)
5.7xl05 (2.0xl06)
7.7xl06 (2.7xl07)
9.2xl06(3.2108)
* Air injection occurred over a period of 6 month ending with the date indicated. Air
injection occurred in equal volumes over that time period.
5.1.3
Retrofit Cells
In the Retrofit cells A and B, there was a single sample collection point in each cell. LFG sampling from
the Retrofit cells began in November 2001 and summary statistics are presented in Table 5-4. Figures 5-5
and 5-6 show significant variability and inconsistency, although both cells seem to be exhibiting similar
behavior. Potential sources of the erratic behavior include: (i) leakage and air intrusion around LFG header
pipes; (ii) "over-pulling" of the well field and subsequent air intrusion through the cover; and (iii) watering
out of gas collection wells. The most likely explanation is a combination of these factors, since watering
out of gas collection wells can cause an imbalance in the vacuum applied to the landfill, resulting in over-
pulling of the well field and subsequent air intrusion. These effects will be discussed in more details in
Section 5.2.3. As with the As-Built cells, the O2 to N2 (assuming balance gas consists of only N2) ratio is
(1 to 4) is similar to ambient air throughout the monitoring period.
74
-------
Figure 5-4 Monthly Average Gas Composition for As-Built Cell B
Table 5-4 Summary Statistics of Gas Composition for Retrofit Cells
Sub Cell
Cell A
fi\r — 1 77s!
Cell r>
(XT - 1 A()\
Parameter
Mean ± Stdev
Median
Range
Mean ± Stdev
Median
Range
CH4
(%)
49±9
51
4-63
43±13
46
4-62
C02
(%)
35±6
36
3.2-46
32±9.3
34
3.2-46
02
(%)
3.3±3
2.7
0-19
4.9±4.3
4
0-19
Balance
(%)
11±12.3
7.2
0-74
19±19
13
0-74
Note: Each sample represents the mean of triplicate readings collected at the header pipe of
each cell
5.1.4 Landfill Gas (LFG) Composition Summary
In general, duplicate cells had statistically similar LFG composition, with CH4 and CO2 constituting the
bulk of the LFG stream. Oxygen concentrations in LFG of the As-Built bioreactor landfill cells may not
necessarily indicate a problem with the gas collection system. While some oxygen can be consumed in the
process of aerobic degradation, the rate of air injection is much larger than what is needed for
stoichiometric conversion; thus the majority of O2 will pass through the system. However, the ratio of
concentration of O2 to N2 in LFG of the Retrofit system at 1:4 (ratio of O2 to N2) is evidence of LFG
collection problems. It is most likely that leaks in the header piping system played a large role in air
intrusion as well as watering-out of LFG wells.
75
-------
\ \ \<^ \<^ \
Figure 5-5 Monthly Average of Gas Composition for Retrofit Cell A
70
o
60 -
50 -
40 -
o
§ 30-|
1
O 20H
10 -
CH4
Ul
/./Vl u-J
vx\ 'v/-x\ v ^-v\ /•• ix\ /
^^ —\x--/ V ^ L..^-C-
>--r-r
Figure 5-6 Monthly Average of Gas Composition for Retrofit Cell B
76
-------
5. 2 Measured Methane Production and LandGEM Model Predictions
For this section, the actual LFG collection rates (i.e., field data) were compared to predicted gas generation
rates modeled in the U.S. EPA's LFG production model, LandGEM. LandGEM models landfill gas
generation using a first-order decay equation. Detailed information regarding the parameters and
assumptions used in LandGEM can be found on the U.S. EPA's website
(http://www.epa.gov/ttn/catc/dirl/landgem-v302-guide.pdf ). The equation used by the model to calculate
LFG generation is similar in mathematical form to Equation 5-1 :
''j
Equation 5-1
Where:
Q = LFG generated (nrVyr);
k = first order waste decay rate (yr"1);
L0 = CH4 generation potential (m3/Mg waste);
M; = waste mass placement in year i (Mg); and
ty = time (yr).
LandGEM allows the user to input various parameters, including waste mass placement (MO, waste decay
rate (k), and the CH4 generation potential (L0). One cited benefit of landfill bioreactors is the acceleration
of the waste decomposition rate (i.e., higher k) (Reinhart and Townsend 1998). The objectives of the
analysis presented in this section are to: (i) estimate an approximate k for each cell based on gas collection
data; (ii) compare the actual landfill gas generation rates to those predicted using LandGEM; and (iii)
compare the k values based on field data to the expected k values for landfill bioreactors as reported in the
literature.
For the purposes of this analysis, waste mass inputs used in LandGEM for each cell are presented in Table
5-5. These values were calculated using waste receipts at the site. Model default values of L0 or site-
specific estimates of L0 may be used in LandGEM. The U.S. EPA's compilation of pollutant emission
factors, AP-42, states "Estimation of the potential CH4 generation capacity of refuse (L0) is generally
treated as a function of the moisture and organic content of the refuse. " Moisture content and BMP (i.e., a
measure of the CH4 yield of an organic material during anaerobic decomposition) data from fresh waste
samples taken at the site in 2004 were available and were used to calculate a site-specific L0.
Since the waste placed in each cell at OLLB was primarily MSW, it was assumed that the BMP and
moisture content results of the fresh waste were indicative of waste placed in all cells. It is noted that the
site specific L0 (59 m3/Mg-wet) is lower than the LandGEM default L0 values for wet and conventional
landfills (96 and 100 m3/Mg (3,000 and 3,200 ft3/ton), respectively).
The LFG flow rate and volume collected at OLLB, like any other landfill, represents the fraction of the gas
captured by the LFG collection system. Thus, the collected LFG is generally less than the LFG generated
since collection systems do not operate at 100 percent collection efficiency (CE). Rather, AP-42 states that
reported CEs typically range from 60 to 85 percent, with a 75 percent CE being the value that is most
commonly assumed. To this end, each of the subsequent sections will compare the following to the
LandGEM-predicted gas generation rate: (i) the raw field data collected; and (ii) transformed field data,
using the assumption that the gas samples represent a collection system operating at 75% CE. To
accomplish this latter comparison, the collected LFG was divided by 0.75 to calculate the landfill gas
generation rate. It is noted that this procedure does not follow U.S. EPA Method 2E for estimating a site-
specific gas generation rate. The sites actual CE can be higher or lower than 75%, but for the purposes of
this analysis 75 % will be used.
To carry out the comparison using field LFG data, the cumulative CH4 generation for each cell was
calculated by (i) averaging the triplicate total gas flow readings for each date of collection; (ii) multiplying
the mean of the daily readings by the mean of the CH4 percentage for that day; (iii) taking the average of
77
-------
the weekly readings for each month; (iv) multiplying the monthly average flow rate by the number of days
in the month; and (v) summing the total monthly gas collection amounts. Dates with a zero gas flow
reading were eliminated. It is also noted that LandGEM typically provides gas generation estimates on an
annual basis, so the equations were entered manually to provide monthly results as necessary.
In order to optimize the CH4 production equation from LandGEM (as presented in Equation 5-1) for the
waste decay parameter k using annual waste mass placement data and monthly CH4 generation volume, the
sum of squared errors was used. Since only two to four complete years of CH4 flow data exists,
optimization had to be performed on a deciyear basis (decimal year). This required stating Equation (1) in
terms of decimal year production (as presented in Equation 5-2).
kLn
Equation 5-2
Where d is the number of deciyears since initial waste placement, M; is the waste mass placed in deciyear i,
and Qd is the methane production for deciyear d. Estimation of the waste decay rate parameter, k, for a
particular cell proceeded by minimizing the sum of squared errors of Equation 2 (as presented in Equation
5-3).
s
d
kT d
K1^o -y
10 t;
v
Equation 5-3
Where d varies over the deciyear methane data for the cell. The actual data for the minimization consisted
of monthly methane observations and annual waste mass values. Therefore, the Qd values were
interpolated and the M; values were the annual waste mass observations divided by 10, so it was assumed
that waste was placed uniformly throughout a given year. A quasi-Newton method was used to minimize
this function.
Table 5-5 Summary of Waste Mass Inputs for LandGEM, Mg (tons)
Year
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Control Cells
193,771 (213,148)
148,131(162,944)
35,801 (39,381)
49,398 (54,338)
58,456 (64,302)
27,139 (29,853)
Retrofit Cells
424,810 (467,287)
745,369 (819,898)
585,133 (643,640)
112,263 (123,489)
28,577(31,435)
As-Built
Cell A
183,680 (202,048)
100,574(110,631)
178,424 (196,267)
56,759 (62,435)
CellB
256,247 (281,872)
2,188(2,407)
289,747(318,722)
250,641 (275,705)
5.2.1
Control Cells
Waste placement began in the Control cells in 1999; however, gas collection and measurement did not
commence until 2003. As a result, only LandGEM results from 2003 onward were used for comparison to
field data to estimate k. Since no hydraulic barrier exists between the subcells of the Control cells, the data
from each subcell were aggregated to compare to the LandGEM prediction. In general, the cumulative
methane generation for Control cells appears to follow the expected landfill gas generation rate as predicted
for a typical "dry tomb" landfill scenario in LandGEM. Without accounting for gas collection
inefficiencies, the optimized k value (0.043) matches up very well with the curve representative of the dry
78
-------
tomb case of k = 0.04 yr"1 as presented in Table 5-6. When assuming a gas CE of 75 percent, the
transformed field data indicate a waste decay rate of 0.061 yr"1, greater than the AP-42 default k value of
0.04 yr"1. It is noted that the actual CE of the Control cells is not known, so the actual waste decay rate may
be more or less than 0.061 yr"1. It is also recognized that the time period of data collection may not
necessarily be indicative of long-term LFG generation trends.
Table 5-6 Optimized Landfill Decay Constant (k) for Control Cell
£0 = 59
Collected
Methane
0.043
Generated
Methane1
0.061
L0 = 96
Collected
Methane
0.025
Generated
Methane1
0.035
1.Assumed a collection efficiency of 75%
fi
o
I
u
u
16
14
12
10
8 -
6 -
4 -
2 -
•O
LandGEM k=0.04
LandGEM k=0.05
--T- LandGEM k=0.15
A Collected LFG
O Generated LFG Assuming CE=75%
Figure 5-7 Cumulative Methane Gas Collection Data versus LandGEM-Predicted Gas Generation
for the Control Cells
5.2.2
As-Built Cells
Data from As-Built cells A and B were analyzed separately since detailed waste placement data were
available for each cell (see Table 5-5) and because of the hydraulic separation of these cells. LFG
collection data for the As-Built cells began in April 2003. Since LandGEM provides an annual output, the
output for 2003 was scaled such that it represented nine months of predicted generation (i.e., representing
April 2003 to December 2003 rather than the entire year).
Cumulative methane generation predictions using LandGEM with mass input based on the available waste
acceptance data (Table 5-5), beginning with a k value of 0.04 yr"1 (the AP-42 default for "dry tomb"
landfills), and incrementally increasing up to a k value of 0.25 yr"1 are presented in Figures 5-8 and 5-9. A
decay rate (k value) of 0.25 yr"1 represents the expected decay rate for wet landfills according to the U.S.
EPA report "First-Order Kinetic Gas Generation Model Parameters for Wet Landfills" (U.S. EPA 2005).
As demonstrated by the figures, the cumulative collected methane for both As-Built bioreactor landfill cells
(as presented in Table 5-7) indicate a k value of between 0.15 yr"1 and 0.20 yr"1 in As-Built cells. The
optimization of the decay constant k indicated that the k value for both bioreactor cells are statistically
similar at 0.11 yr"1. Moreover, if it is assumed that the collected gas represents 75% of the methane being
79
-------
25
m 20
•-v—
o
O
LandGEM k= 0.04
LandGEMk = 0.15
LandGEM k =0.20
LandGEM k = 0.25
Collected LEG
Generated LEG Assuming CE = 75%
/V
*
..O
Figure 5-8 Cumulative Methane Gas Collection Data versus LandGEM-Predicted Gas Generation
for As-Built Cell A
35
30 -
25 -
•••O
-v —
o
o
LandGEM k= 0.04
LandGEM k= 0.15
LandGEM k= 0.20
LandGEM k= 0.25
LandGEM k= 0.28
Collected LEG
Generated LEG Assuming CE = 75%
Figure 5-9 Cumulative Methane Gas Collection Data versus LandGEM-Predicted Gas Generation
for As-Built Cell B
80
-------
Table 5-7 Optimized Landfill Decay Constant (k) for As-Built Bioreactor Cells
Landfill
Cell
As-built A
As-built B
L0 = 59
Collected
Methane
0.11
0.11
Generated
Methane
0.16
0.15
L0 = 96
Collected
Methane
0.064
0.062
Generated
Methane
0.089
0.086
generated (i.e., a CE of 75 percent), the k value for As-Built cells is optimized at a value of 0.16 yr"1 which
falls within the range that was visually estimated before and that is statistically higher than that observed
for the control cells.
These results suggest that based on gas collected at the site between 2003 and 2006, the As-Built cells
appeared to generate LFG at a rate that is clearly higher than would be expected at traditional "dry tomb"
landfills. As described earlier, the waste decay rate in the Control cells is approximately 0.06 yr: which is
much lower than that of the As-Built cells. As illustrated in Figure 5-10, LFG generated in the early stages
of the landfill's life was significantly greater in the As-Built bioreactor cells relative to the Control and dry
tomb cells.
Modeling of LFG generation rates for 50 years based on current loading rates and k of 0.16 (yr:) indicates
the effectiveness of bioreactor landfills at generating methane at a faster rate relative to a "dry tomb" cells
as presented in Figure 5-10. The higher LFG generation rate, especially in the early stages of landfill
development, has implications for potential beneficial reuse of LFG. Furthermore, the As-Built LFG
curves, shown in Figure 5-10, indicate that LFG generation declines to very low rates after approximately
25 years; this has the potential to reduce long-term concerns regarding LFG management during post-
closure. These results should be viewed with caution since the developed k values are based on only 3
years worth of landfill gas collection data.
12
10 -
8 -
CD 6 -
o
E 4H
2 -
/\
/\
/ \
As-Built Cell A k = 0.04
As-Built Cell Bk = 0.04
As-Built Cell Ak = 0.16
As-Built Cell Bk = 0.16
0
2000 2010 2020 2030 2040 2050
Figure 5-10 LFG Generation Rate Prediction for As-Built Cell A and B
81
-------
5.2.3
Retrofit cells
Unlike the As-Built bioreactor landfill cells, the Retrofit cells were not hydraulically separated. As a result,
the LFG collected from the two collection points was aggregated to calculate methane collection for further
use in modeling the Retrofit unit. LFG collection started in November 2001. For LandGEM modeling of
the Retrofit cells, a k for a "dry tomb" landfill case (i.e., k = 0.04 yr"1) in addition to higher k values were
explored. Results presented in Figure 5-11 show that direct field measurements of LFG generally fall just
above (k = 0.041 yr"1) the expected gas generation in "dry tomb" landfill. However, when a CE of 75%
was applied, methane generation approximately matched the curve corresponding to a k = 0.061 yr"1 which
is slightly higher than "dry tomb" landfill but not higher than that of the Control cell. The U.S. EPA
document regarding k values for wet landfills indicated that modeled landfill sites had a k value that ranged
between 0.11 and 0.3 yr"1. Even with the assumption of 75 percent CE, the data for the Retrofit unit fall
well below the lower bound of that range.
As discussed previously, LFG collection in the Retrofit cells exhibited erratic behavior with regard to CH4,
CO2, and O2 concentrations during the sampling period. It is believed that this trend can be attributed to a
combination of watering out of gas wells, air leakage, air intrusion, and operator error. Upon review of
Figures 5-5 and 5-6, it is noted that each figure shows LFG concentration for Retrofit cells A and B as
expected for the first four months of operation (i.e., November 2001 through February 2002). Figure 5-12
depicts the average LFG collection flow rate for the Retrofit cells during the study. This figure shows that
the LFG generation was highest during the first four months before starting a precipitous decline in flow for
the first few months of 2002, when the installation of new vertical gas wells was stopped. As a result, it
appears that the erratic behavior in terms of LFG composition began as the LFG collection rate decreased,
indicating that the existing vertical gas collection wells may have been affected. It is likely that the wells
filled with either the liquids that were being applied during this time period and/or with gas condensate.
The magnitude of the impact on LFG CE is not known, since the determination of specific gas collection
efficiency was beyond the scope of this project. Furthermore, the fluctuation of LFG flow rate beginning in
50
rO 40 -
-•— LandGEM k = 0.04
•O" LandGEM k = 0.07
-•T- LandGEMk = 0.15
O Collected LFG
O Generated LFG Assuming CE = 75%
^
.O
Figure 5-11 Cumulative Methane Gas Collection Data versus LandGEM-Predicted Gas Generation
for the Retrofit Cells
82
-------
70
^ 50 -
I
I 4° "
£ 3(H
o
§? 20 -
>> 10 -
I OH
cv^
c^
Monthly Average LFG Flow Rate
Date
Figure 5-12 Monthly Average LFG Flow Rate for Retrofit Cells
late 2004 (see Figure 5-12) may be the result of an increase in the vacuum on the Retrofit cells. The
comparison of Figures 5-5 and 5-6 to Figure 5-12 indicates that the increase in gas flow rate generally
coincides with a decrease in CH4 and an increase in O2, indicating over-pulling on the LFG well field.
Furthermore, operators for LFG well monitoring were replaced at the same time and this may have
contributed to the erratic readings after that time.
Figure 5-13 compares the modeled LFG generation rate for the Retrofit unit (k = 0.061 yr"1) to the
decomposition rate in the AP-42 default. The results indicate that the decomposition rate in the Retrofit
unit was greater than the modeled LFG production based on the decomposition rate of the regulatory
default value. The difference in LFG generation is not as dramatic in the Retrofit cells compared to the As-
Built cells (assuming similar loading rate). It is important to note that the Retrofit Unit was not operated as
a bioreactor landfill until 2001, approximately six years after initial waste placement. Based on the first-
order decay LFG production model for MSW, fresher (i.e., newly placed) waste generates LFG at a greater
rate than older waste. Therefore, the timing of landfill bioreactor operations and the time period of
collected LFG field data (2001 to 2006) is not perfectly comparable to the Control cell. In addition, the
older refuse in the Retrofit cell may have contributed to a lower decay rate.
5.3 Comparison of Field Gas Results of Control and Landfill Bioreactor Cells
A comparison of the CH4 flow rates between the Control (i.e., dry tomb) cell and the experimental (i.e.,
landfill bioreactor) cells is critical in understanding the effectiveness of landfill bioreactor operations. The
hypothesis is that with the greater degradation rate in bioreactor landfill cells, compared to dry tomb cells,
landfill gas will be generated at a greater rate. Since waste placement and mass in the test cells varied, a
direct side-by-side comparison over the entire monitoring period was not possible. Comparisons were only
made in months for which gas collection data for waste of similar age was available as summarized in
Table 5-8 and 5-9. The mass of waste in each cell was normalized to the mass in the Control cell as the
rate and amount of CH4 generated, over time, is proportional to the mass in the cell. The waste age, shown
in Tables 5-8 and 5-9, is relative to the year of initial waste placement for each cell (e.g., a waste age of six
83
-------
12
10 -
fi
'a
6 -
4 -
2 -
k = 0.061yr
k = 0.04 yr"1
1990
2000
2010
2020
2030
2040
2050
Figure 5-13 Modeled LFG Generation Rate for Waste in the Retrofit Cells
is compared in Table 5-8 since waste filling began in 1999 and 1995 for the Control and Retrofit cells,
respectively). To carry out the comparison, the ratio of CH4 generated in a bioreactor landfill cell to that
generated in the Control cell on these dates was calculated and plotted as presented in Figures 5-14 and 5-
15. The observation number listed in the tables corresponds to the observation number located on the x-
axis of each of the figures. The dashed horizontal line at a ratio of one indicates no difference in gas
generation (based on collected gas data) between the Control cells and the landfill bioreactor cells.
Figure 5-14 shows that the data for the comparison of the Control and Retrofit Unit fell below the ratio of
1.0 for the first two observations and above the line for the last two observations. As presented earlier in
Chapter 3, liquid introduction into the Retrofit landfill Unit did not started until May 2001. As a result,
there was not adequate time for good moisture distribution, between May and November, to see an
enhancement in the LFG generation rate in the Retrofit Unit. The increasing trend in the ratio over time,
observed in Figure 5-14 suggest that, with time, the Retrofit Unit was able to produce more gas. This
observation is inline with the solids data results (presented in Chapter 4) that suggested that the Retrofit
cells are potentially decomposing at a rate higher than the Control cells. However, as with the solids data,
more data are needed to better document trends.
Because of the filling schedule, more gas observations were available for the comparison of the bioreactor
landfill cells and Control cells as presented in Table 5-9. As shown in Figure 5-15, the ratios of the landfill
gas generated in the bioreactor landfill cells to that generated in the Control cells were often higher than 1,
suggesting that the rate of degradation was greater in the As-Built bioreactor landfill. This conclusion is
further supported by the solids data (see Chapter 4) that show more rapid solids decomposition in the As-
Built cells.
84
-------
TJ
ible 5-8 Summary of Dates Used for Retrofit and Control Cells CH4 Flow Comparison
Observation #
1
2
3
4
Waste Age (Yr)
6
7
Control Month
11/2005
12/2005
1/2006
2/2006
Retrofit Month
11/2001
12/2001
1/2002
2/2002
able 5-9 Summary of Dates Used for As-Built and Control Cells CH4 Flow Compariso
Observation #
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Waste Age (yr)
3
4
Control Month
1/2002
2/2002
3/2002
4/2002
5/2002
6/2002
7/2002
8/2002
9/2002
10/2002
11/2002
12/2002
1/2003
2/2003
As-Built Month
1/2005
2/2005
3/2005
4/2005
5/2005
6/2005
7/2005
8/2005
9/2005
10/2005
11/2005
12/2005
1/2006
2/2006
2.5 -
2.0 -
§ 1.5 H
IS
1.0
0.5 -
2 3
Observation Number
CH4 Flow Rate Ratio
Figure 5-14 Normalized Comparison of Gas Collection in Control and Retrofit Cells
85
-------
7 -
Pi
"a
"3
o
£J
m
tH
O
u
m
3 -
6 8 10
Observation Number
12 14
16
• CH4 Flow Ratio - As-Built Cell A
O CH4 Flow Ratio - As-Built Cell B
Figure 5-15 Normalized Comparison of Gas Collection in Control and As-Built Cells
5.3.7 Summary ofLFG Generation
The measured LFG generation rates, the decay constant (k), and the solids data presented in the previous
sections indicate that there is an increase in solid waste degradation in bioreactor landfills compared to
Control cells. The enhanced LFG generation rate indicates that the majority of LFG production will occur
in the early stages of landfill development (in the case of the As-Built cells in this study, within the first ten
years following initial waste placement). Figure 5-16 shows the percent of LFG generated as a function of
years after waste placement for the various k values estimated earlier in this section. Based on this
analysis, 50% of LFG is generated within 7 years of waste placement at a k of 0.16 yr"1. This trend has
significant implications for potential beneficial use options for LFG at MSW landfills operated as a
bioreactor during and after waste filling since the rapid LFG generation would mean more recoverable
energy is available sooner. The trend further underscores the importance of having a LFG collection
system in place in the early stages of landfill development for: (i) ambient emission control if a landfill is to
be operated as a bioreactor and (ii) an increase in LFG collected if a beneficial use of the LFG (e.g., LFG-
to-energy) is considered for the site. If a LFG collection system is not in place during the early stages of
landfill development, the most LFG-productive years will be missed, resulting in greater emissions and
lower cumulative LFG utilization.
These trends also have an implication for post closure care. Based on LandGEM model data presented in
Figure 5-16, more than 90% of LFG is produced within 12 years after waste placement in bioreactor
landfill cells as compared to less than 50% in the Control Units. Thus, if the potential for LFG generation
is low (e.g., 90% of LFG has already been produced) in MSW landfills operated as bioreactors, then an
argument could be made that the duration of post closure care monitoring for LFG may be reduced.
86
-------
k = 0.04 /yr
k= 0.061/yr
k=0.16/yr
k = 0.7 /yr
0.1
1
10
100
Year After Waste Placement
Figure 5-16 Normalized Comparison of Gas Collection in Control cells and As-Built Cells
5.4 None Methane Organic Carbon (NMOC) Concentrations in LFG
Samples for analysis of total NMOCs and a number of speciated HAPs that are a subset of NMOCs were
collected for all cells on a quarterly basis and sent under chain-of-custody protocol to Severn Trent
Laboratories as outlined in the QAPP presented in the First Interim Report. Summary statistics for the
HAP data are provided as Appendix E. Non-detected constituents were not included in the statistical
analysis represented in the tables. In evaluating the air pollutant concentrations, a correction was used to
account for air infiltration. Furthermore, only constituents detected in more than 50 percent of the samples
were included; these constituents were then corrected for air infiltration using the following equation from
U.S. EPA's AP-42 guidance document:
_
=
0^(1x10")
Equation 5-4
--CH,
Where:
Cp
CpjU
IxlO6
CcO2
Corrected concentration of pollutant P in LFG, ppmv;
Uncorrected concentration of pollutant P in LFG, ppmv;
Constant used to correct concentration of P to units of ppmv;
CH4 concentration in collected gas, (ppmv); and
CO2 concentration in collected gas, (ppmv).
Currently, little is known about the kinetics of NMOC emissions from MSW landfills, specifically as the
NMOC emissions relate to waste type, waste age, and other factors. NMOCs are generated in MSW
landfills from volatilization of organic compounds and, to a limited degree, by biological and chemical
87
-------
processes (U.S. EPA, 1998; Staley et al., 2006). The Clean Air Act (CAA) default concentration for
NMOCs in MSW LFG is 4,000 ppm as hexane. This value was developed for compliance purposes, but
AP-42 indicates that NMOC concentrations greater than this default concentration can exist at MSW
landfills. AP-42 defaults for NMOCs are 600 ppm for MSW landfills with minimal co-disposal occurred at
the site.
When considering a comparative analysis of NMOC production, the use of only NMOC concentration data
from a landfill cell may be misleading. NMOC concentrations must be tied in with the LFG flow rate to
evaluate production. Additionally, the mass of waste and the composition of waste would likely affect the
magnitude of NMOCs produced. In this section, the analytical data (i.e., NMOC concentrations) will be
tied into available LFG flow rate data to calculate the mass flow rate of NMOCs from the different cells.
Only total NMOC data are analyzed in this section, since the individual compounds listed in the QAPP and
analyzed in the laboratory (i.e., HAPs) are a subset of NMOCs. The data will be analyzed to present the
analytical results on a unit (i.e., mass) basis. Summary statistics for total NMOCs are presented in Table 5-
10.
The mean NMOC concentration (presented in Table 5-10) varied between cells. Differences between
concentrations ranged from a factor of approximately 4.5 (e.g., Retrofit cells versus Control cell A) to as
little as approximately 1 (e.g., As-Built cell B versus Retrofit cells). It is noted that higher NMOC
concentration in a particular cell does not necessary mean higher emissions since NMOC production rate
depends on the volume of LFG generated by each cell. Furthermore, the calculated NMOC production
averages do not directly account for the differences in the mass of waste in the cell or waste age. Thus, a
direct between-cell comparison of NMOC production is inaccurate. The aforementioned lack of accurate
predictive methods of NMOC production does not allow for any meaningful conclusions to be drawn based
on NMOC concentration alone. Furthermore, the differences between the cells with regard to the time of
waste placement, coupled with the NMOC sampling frequency outlined in the QAPP, limits the pool of
data that can be compared.
Table 5-10 NMOC Concentration for Each Sub Cell (ppm as Hexane)
Subcell
Control cell A
Control cell B
As-Built cell A
As-Built cell B
Retrofit cell A
Retrofit cell B
Number of
Detects
14
14
10
10
15
14
Range (min-
max)
383-1,608
450-867
97-633
85-467
63-350
143-383
Median
1,146
633
458
233
250
250
Mean ± Standard
Deviation
1,060±368
611±117
393±177
242±100
246±73
237±71
5.4.1
Landfill Gas (LFG) Flow-Based NMOC Production Analysis
As noted in Chapter 3, the Control and Retrofit cells were sampled at more than one location; however,
there was no hydraulic separation between the cells, and therefore the LFG flow data was combined for
comparison with LandGEM. In this section, a similar approach will be used in evaluating the mass flow
rate of NMOCs in the Control and Retrofit cells as was used for methane production. Since samples were
taken from more than one point for NMOC analysis, the mean of the concentration at each selected date
will be used to calculate the NMOC mass emission rate. Equation 5-5 depicts the approach for calculating
the NMOC mass emission rate:
-------
m = C x QLFG x
1
385.4
- x 86.17x0.454 xl(T
Equation 5-5
Where:
m
CP
QLFG
385.4
86.17
0.454
io-6
mass flow rate of NMOC, kg/min;
concentration of NMOC corrected for air intrusion, ppmv as hexane;
flow rate of LFG, scfm;
conversion of ft3 per Ib-mol of gas at STP;
molecular weight of hexane, Ib/lb-mol;
conversion of Ib into kg; and
conversion factor from ppm to volume fraction.
Since NMOC samples were collected quarterly, the monthly average flow rate corresponding to the month
that the NMOC sample was collected was used in calculating the mass flow rate. These calculated NMOC
production rates were plotted over time to evaluate potential temporal trends in NMOC production.
Figures 5-17 through 5-20 depict the results for the Control, Retrofit, and As-Built cells A and B and
demonstrate somewhat erratic behavior in NMOC production rate.
The NMOCs from the Control cells appear to increase and decrease randomly while showing an overall
decreasing trend during the sample collection period. The As-Built cells also appear to fluctuate during the
sample collection period. Overall, it does not appear to be a strong temporal trend for each cell. In general,
the production rate for each cell falls between 0.002 and 0.035 kg/min (0.004 and 0.07 Ib/min). The mean
NMOC production rate that takes into account the cell's LFG flow rate for the Control, Retrofit, and As-
Built A and B cells is 0.012, 0.012, 0.010, and 0.008 kg/min (0.026, 0.026, 0.022, and 0.018 Ib/min),
respectively.
e
w
Date
Figure 5-17 NMOC Production in the Control Cells
89
-------
0.03 -
O 0.01
S
0.03 -
OB
O
°-02 '
§
I
Date
Figure 5-18 NMOC Production in As-Built Cell A
Date
Figure 5-19 NMOC Production in As-Built Cell B
90
-------
0.03 -
g
"So
(3 0.02
o
0.01 -
5.4.2
Date
Figure 5-20 NMOC Production in the Retrofit Cells
Waste Mass-Based NMOC Production Analysis
The regulatory requirements for installing a LFG collection and control system are based on the mass flow
rate of NMOCs, which is based on a tiered system. If a landfill's design capacity is 2.5 million Mg or 2.5
million m3, a LFG collection and control system must be put in place. However, this is based on a
regulatory default NMOC concentration of 4,000 ppm. If it can be demonstrated that a landfill's NMOC
concentration is less than 4,000 ppm, the NMOC production rate may be modeled using the site-specific
NMOC concentration and Equation 5-6.
= j2kLoM,(e-k'0(CNMOC)(3.6xlO-9)
Equation 5-6
Where:
MNMOC
k
Lo
M,
ti
CNMOC
3.6xlO"£
= Mass emission rate of NMOC (Mg/yr);
= CH4 generation rate constant (yr"1);
=CH4 generation potential (m3/Mg);
= Mass of MSW placed in year i (Mg);
= age of the 1th section of waste (yr);
= Concentration of NMOC (ppm); and
= Conversion factor.
After calculating the annual mass of NMOC production, the result is compared to a regulatory threshold of
50 Mg/yr. If the calculated NMOC production exceeds this amount, then a LFG collection system must be
installed in accordance with 40 CFR Subpart WWW. The site-specific NMOC concentration (i.e., CNMOC)
is determined by performing Tier 2 testing, which is described in Subpart WWW.
The calculation of NMOC production depicted in Equation 5-6 differs from Equation 5-5 in that it is based
on the mass of waste in place as well as the parameters k and L0 of the landfill cell instead of actual LFG
collection data. Equation 5-6 can be modeled in LandGEM using site-specific waste placement data and a
91
-------
representative NMOC concentration (based on the sampling guidelines stated in Subpart WWW). The
parameters k and L0 can either be the Clean Air Act regulatory default values of 0.05 yf: and 170 m3/Mg or
determined on a site-specific basis using U.S. EPA Method 2E.
The NMOC generation rate was modeled for each cell based on available waste placement data, the
regulatory default k and L0, and the mean NMOC concentration during the study period. The mean NMOC
concentration for each cell was presented in Table 5-10. Table 5-11 depicts the LandGEM-predicted
NMOC production for each cell using the waste placement data presented in Table 5-5, k=0.05 yr"1 and
L0=170 m3/Mg.
Typically, a demonstration of whether a landfill is above or below the 50 Mg/yr threshold is performed by
summing the NMOC production rate for each landfill unit at a site for a particular year. Since this report
focuses on a portion of the OLRDF, a comparison to the 50 Mg/yr threshold would be incomplete, based on
available NMOC concentration data (i.e., no NMOC data were available for the cells that were not part of
this study). It is not feasible to make a strong comparison of NMOC production between the Control cells
and the landfill bioreactor cells using the method prescribed by the regulations (i.e., using Equation 5-6).
The data presented in Table5-l 1 are a reflection of the waste placement data (time of waste placement and
total in-place waste mass) and the NMOC concentration based on several sampling events. Since all of
these data are incorporated in the LandGEM model, it is difficult to establish a cause-and-effect
relationship between operation as a bioreactor landfill and NMOC production. Section 5.3, which
investigated relationships between waste placement, NMOC concentration, and LFG collection data,
indicated that there was no clear distinction between the NMOC productions in the Control cells versus the
landfill bioreactor cells. However, it is noted that the increased waste decomposition rates discussed earlier
may necessitate the early collection of LFG (regardless of predicted NMOC production) for bioreactor
landfills, especially if the landfill owner/operator seeks to capture LFG for beneficial reuse (e.g., gas-to-
energy).
Table 5-11 LandGEM-Predicted Mass Production Rate of NMOCs in Different Landfill Units
(Mg/yr)
Year
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Control
Cell
-
-
-
-
0.0
9.7
16.6
17.5
19.1
21.1
21.4
20.4
19.4
18.5
17.6
16.7
Retrofit
Cell
0.0
6.1
16.6
24.2
23.0
21.9
22.5
21.8
20.7
19.7
18.7
17.8
17.0
16.1
15.3
14.6
As-Built
Cell A
-
-
-
-
-
-
-
0.0
4.3
6.4
10.3
11.1
10.6
10.1
9.6
9.1
As-Built
CellB
-
-
-
-
-
-
-
0.0
3.7
3.5
7.6
10.8
10.3
9.8
9.3
8.8
Note: A dash (-) indicates NMOC production was not modeled for that year since no waste was in place.
To directly compare cells, the NMOC generation rate for each cell was based on normalized mass
placement and site specific L0, k and the 95 percentile upper confidence level on the mean (as a
conservative estimate of the mean) of the NMOC concentration for each cell during the study period as
presented in Table 5-12. Results of the analysis are presented in Table 5-13. Although the NMOC UCL95
for the bioreactor cells was much lower than the Control cell (less than half), bioreactor cells produced
more NMOC per year when compared to the Control cell. This observation is a direct result of the increase
92
-------
in LFG generation rate (higher k values) of bioreactor landfill cells. However, the amount of NMOC
generated (Mg/year) drops quickly to levels below that of the Control cell within a few years of operations
(7 years in the case of OLLB).
These results shed light on the importance of a LFG management system for bioreactor landfills.
Bioreactor landfills are expected to emit NMOCs more rapidly than dry tomb cells. However, bioreactor
landfill NMOC emissions tend to drop much quicker than dry tomb cells, thus reducing long-term
concerns.
Table 5-12 LandGEM Model Input Values for NMOC Side-by-Side Analysis
Parameter
k (year1)
L0 (m3/Mg)
NMOC
UCL95 (ppm)
Control
cell
0.061
59
993
Retrofit
cell
0.061
59
271
As-Built
cell A
0.16
59
495
As-Built
cellB
0.16
59
300
Table 5-13 LandGEM Predicted Mass Production Rate of NMOCs Emissions (Mg/yr)
Year After
Waste Placement
1
2
o
J
4
5
6
7
8
9
10
20
30
60
Control
Cells
0
4.83
8.24
8.64
9.36
10.27
10.34
9.72
9.15
8.61
4.68
2.54
0.41
Retrofit
Cells
0
1.32
2.25
2.36
2.56
2.81
2.82
2.66
2.50
2.35
1.28
0.69
0.11
As Built
Cell A
0
6.05
9.79
9.46
9.60
10.01
9.38
7.99
6.81
5.80
1.17
0.24
0.00
As Built
CellB
0
3.66
5.92
5.72
5.81
6.06
5.67
4.84
4.12
3.51
0.71
0.14
0.00
5.5 Methane Surface Emissions
Methane emissions were measured on using a CEC-Landtec SEM-500 field instrument. Surface
concentrations were monitored around the perimeter of the collection area and along a pattern that traversed
the landfill at 30 m intervals and where visual observations indicated elevated concentrations of landfill
gas. Emissions were monitored and recorded separately for Unit 5 and 7. The climatic conditions and the
background methane concentration up and downwind were recorded for each sampling event. Background
concentrations averaged 5.82 ppm upwind and 9.97 ppm downwind for Unit 5, and 2.98 ppm upwind and
21.73 ppm downwind for Unit 7, for the period December 2001 to June 2006. Permit requirements
necessitate a methane concentration greater than 500 ppm above the measured background level to be
marked, adjustments made to reduce the surface emissions at that location, and the location to be
reanalyzed within 10 days. If an exceedance exists on reanalysis, additional adjustments and/or cover
maintenance must be performed and the location reanalyzed within 10 days. On a third exceedance, the Air
Pollution Control District (APCD) must be notified, and either a new well installed within 120 days of the
initial exceedance, or an alternative remedy submitted for approval to the APCD.
93
-------
During the period from December 2001 to June 2006, Unit 5 recorded the following permit response
actions:
• Reported twenty occasions of exceedances which were resolved within 10 days via adjustment of
the gas collection system;
• . Seven exceedances where additional soil cover was added; and
• Installation of one new gas collection well.
During the period from December 2001 to June 2006, Unit 7.3 recorded the following permit response
actions:
• • Reported fifteen exceedances which were resolved within 10 days via adjustment to the gas
collection system;
• Fourteen exceedances where additional soil cover was added; and
• Five instances that required maintenance of leachate risers or changes to the gas collection header
to resolve the issue.
During this same period, Unit 7.4 recorded the following permit response actions:
• Reported four exceedances which were resolved within 10 days via adjustment to the gas
collection system;
• Four exceedances where additional cover soil was added; and
• One instance that required modification of the gas collection header to resolve the issue.
94
-------
Chapter 6. Leachate Quality
This section considers the following parameters for the leachate quality analysis and the interpretation of
the treatment performance in the different landfill units: temperature, pH, volatile organic acids (as acetic
acid), BOD, COD, BOD/COD ratio, TOC, TKN, ammonia nitrogen (NH3/NH4-N), total iron, total
phosphorous, total metals, VOCs, SVOCs, and chloride.
Leachate quality changes over time as a landfill matures and the waste characteristics change due to
changes from waste decomposition. However, often that quality is governed by the bottom, most degraded,
portion of the landfill (Kjeldsen et al., 2002). In this section, leachate quality over time (not waste age) is
evaluated. As the statistical analysis of the solids samples demonstrated in Section 4.1, both waste age and
sample date were found to have an equivalent impact in demonstrating trends or changes in waste
characteristics (and therefore leachate characteristics). As such, an examination of leachate quality over
time was deemed to be an appropriate context in which to view the data.
Trends of leachate parameters versus time were plotted for leachate parameters with detection frequencies
exceeding 50 percent. Descriptive statistical summaries are presented in the analysis for each of the
leachate parameters discussed below: (i) number of samples; (ii) percent detected; (iii) minimum number
of samples detected; (iv) maximum number of samples detected; (v) mean and standard deviation of the
detected samples; (vi) 95 percent UCL; (vii) 75th percentile; and (viii) 95th percentile. When the percentage
detected was less than 100 percent, detection limits were noted in the summary statistics. Multiple linear
regressions were used to quantify trends of the leachate parameters. When linear regression assumptions
were not met, a qualitative assessment was utilized. Details on the quantitative and qualitative assessment
of the leachate parameters are included in Appendix F.
The 95 percent confidence interval for the mean is the range of values that will contain the true mean (i.e.,
the average of the full statistical population of all possible data) 95 percent of the time. The 75th and 95th
percentiles represent values that are less than or equal to the selected value, 75 percent and 95 percent of
the values respectively.
6.1 Temperature
Leachate temperature may be an indicator of microbial activity in a landfill. However, temperature alone
cannot predict the stages of waste decomposition. For example, temperature may also reflect seasonal
variation and, to a lesser extent, the effectiveness of insulation provided by the landfill configuration.
Anaerobic processes are favorable at mesophilic (30 to 38 °C) and thermophilic (50 to 60 °C) temperature
ranges.
Leachate temperatures as a function of time in the Control, Retrofit, and As-Built cells are presented in
Figures 6-1 through Figure 6-3, respectively. Based on the multiple regression analysis conducted, leachate
temperature exhibited a statistically significant positive trend in the Control and As-Built cells. The same
analysis, however, indicated that the leachate temperature in the Retrofit cell A had a statistically
significant decreasing trend while Retrofit cell B did not display any trends. The detailed statistical
analysis is presented in Appendix F. These trends can graphically be observed in Figures 6-1 through 6-3.
Overall leachate temperature in the As-Built cells (average 33 °C) was higher than in Retrofit cells (average
27 °C), and Control cells (average 22 °C), potentially indicating a higher waste decomposition rate in the
As-Built cells relative to the other treatments as a result of air and liquids injection that likely had
stimulated and increased microbial activity. It is noted that a decrease in temperature of about 7 °C was
observed in Control cell A compared to Control cell B following the end of 2004. Reasons for such a
change are currently under investigation and will be further discussed in later reports.
95
-------
50
40 -
O
^ 30
22
1
8,
S 20
£
10 -
• Control Cell A
O Control Cell B
80
•
O
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-1 Leachate Temperature as a Function of Time in the Control Cells
50
40 -
O
^ 30 H
3
8 ,
O
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-2 Leachate Temperature as a Function of Time in the Retrofit Cells
96
-------
O
^
22
50
40 -
30 -
S 20 -
H
10 -
• As-Built Cell A
O As-Built Cell B
o
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-3 Leachate Temperature as a Function of Time in As-Built Cells
6.2 pH
The leachate pH is typically measured in landfill leachate to assist in identifying the stabilizing phase (see
Chapter 2). For optimal anaerobic activity, the leachate pH is expected to range from 6.8 to 7.6 (Parkin and
Owen 1986). Although the initial pH may drop, because of the production and accumulation of VOAs, it
tends to rise back to the neutral range in the CH4 fermentation phase when methanogens start consuming
the VOAs. The air addition, in the As-Built cells, is expected to shorten the acid-forming phase and cause
a rapid progression towards the CH4 fermentation phase.
As with temperature, there are significant increasing pH trends (Figures 6-4 through 6-6) in the Control
cells and As-Built cells, and significant decreasing trends in the Retrofit cells. The model fits are only fair
as seen in the adjusted-R2 values (See Appendix F). The pH time plots indicate that a seasonal component
may be present, but when this component was added to the model, the regression fits did not improve.
Although in both the Control and As-Built cells, the pH reached neutral values by the end of 2003, a clearly
defined acid forming phase was not observed. The acid forming phase occurs at the early stages of a
landfill's life and may last for a relatively short period. In this study, the introduction of air may have also
shortened the period of acid forming phase in the As-Built cells. Thus, in future study, measurement of
leachate pH may be needed at a greater sampling frequency to identify small or sudden changes resulting
from air addition. During the monitoring period, the mostly likely reason for the neutral pH values was that
the characteristics of the bottom portion of well degraded solid waste influenced the overall pH of the
leachate, as described earlier in this document.
A good fraction of the industrial liquid introduced into the As-Built bioreactor Cells was carbonated
beverage waste which tends to be acidic in nature. Based on the leachate pH values, the addition of these
types of liquids did not result in a decrease of the measured leachate pH in the As-Built bioreactor landfill
cells. A sudden drop in pH to about 5.2 was observed in As-Built cell B during September 2004. This,
however, was not observed in As-Built cell A, and so the limited and short-period drop in leachate pH
could not be attributed to the addition of beverage waste. Once again it is noted that a decrease in pH of
about 0.4 was observed in Control cell A relative to Control cell B following the end of 2004. This
decrease was consistent with trends of other parameters.
97
-------
10
9 -
ffi
6 -
• Control Cell A
O Control Cell B
00
o
•
o
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-4 Leachate pH as a Function of Time in the Control Cells
10
ffi
M
o
9 -
6 -
• Retrofit Cell A
O Retrofit Cell B
5*
o
8° 8
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-5 Leachate pH as a Function of Time in the Retrofit Cells
98
-------
10
60
O
9 -
+ 8 -
7 -
6 -
• As-Built Cell A
O As-Built Cell B
00° »CL
° 88 .c
O
o
8° 8o0*
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-6 Leachate pH as a Function of Time in the As-Built Cells
6.3 Volatile Organic Acids (VOAs)
The production and accumulation of VOAs is an indicator of the acid formation phase of a landfill. The
concentration of VOAs decreases over time as these acids are consumed by methanogens in the CH4
fermentation phase. At this phase, pH is expected to increase and to stabilize around neutral values. VOAs
measured in the leachate included: acetic acid, butyric acid, formic acid, lactic acid, propionic acid, and
pyruvic acid. For the purposes of this report, rather than reporting the concentration of each VOA by itself,
the total VOAs concentrations were calculated as acetic acid using the following equation:
[Total VOAs] = [Acetic] + 60
[ButynCJ
90
+ [PymVlc] + [ProPlomc]1
110 74
Equation6-l
where:
numerals indicate the molecular weight of each compound in grams;
brackets indicate concentration in mg/L; and
total VOAs are expressed in mg/L as acetic acid.
Of the VOAs examined in the leachate, only acetic and propionic acids were detected above their
respective detection limit at sufficient frequencies. Because of the large percentage of non-detects,
statistical analysis and comparisons were rather difficult to conduct. Leachate VOA concentration as a
function of time in the Control and As-Built cells are presented in Figure 6-7 and Figure 6-8, respectively.
VOAs in the Retrofit cells were detected at a relatively low frequency, thus will not be graphically
presented in this section. VOAs concentrations exhibited a spike in the leachate of the Control cells in late
2003 . Since then, with a few exceptions, concentrations of VOAs in these cells remained below 100 mg/L
during the monitoring period suggesting that the CH4 fermentation phase in the Control cells had already
started. This observation is further supported by the leachate pH measured in the Control cells,
99
-------
1a>
Jf
rS
'o
5000
4000 -
o 3000
VI
c<3
rS
'o
2000 -
o
g 1000
td
"o
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-7 Volatile Organic Acids as a Function of Time in Control Cell B
uuuu -
M)
-— 4000 -
o
o
8 3000 -
t«
CS
o
^ 2000 -
o
I)
6
^ 1000 -
ji
"o
0 -
• As-Built Cell A
O O As-Built Cell B
•
0 0
0 •
• °
o
o
•
o 0° « •<$>-™0. A
, , » , • , , .•CV-OPO^OD-r^b-rO^-rO^-Q* PO, ".TVOO-^POCX-
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-8 Volatile Organic Acid as a Function of Time in the As-Built Cells
100
-------
which is consistent with on-going waste decomposition in the CH4 fermentation phase. For the As-Built
cells, concentrations of VOAs were relatively high with a large degree of variability from approximately
4,000 mg/L prior to mid 2004 before stabilizing below 100 mg/L at a later time. This correlated well with
leachate pH increasing to reach a neutral value following the end of 2003 in the As-Built cells. It is
particularly interesting to recognize that the As-Built cells generated high volumes of LFG, as was
presented in Chapter 5, even during times with high VOA concentration in the leachate. The existence of
this phase "accelerated methane phase" suggests that for bioreactor landfills, because of the increase in the
degradation rate, it is possible to have high VOA concentration as the methane generation starts.
Compared to Control cells, higher concentrations of VOAs (about five-fold) were observed in leachate
from As-Built cells until mid 2004.
6.4 Total Organic Carbon (TOC)
TOC includes a variety of organic compounds, including humic and fulvic acids, VOAs, and
carbohydrates. TOC measurement was not included as a leachate monitoring parameter until early 2004.
A statistical summary of the TOC measurements in the leachate of the landfill units is presented in
Appendix F. Leachate TOC concentrations as a function of time in the Control, Retrofit, and As-Built cells
are presented in Figures 6-9 through 6-11, respectively.
Statistically, the TOC concentrations in the leachate showed no apparent trends in any of the treatment
cells. The addition of beverage waste to the As-Built cells was likely responsible for the higher TOC
concentrations (about three-fold) observed in this unit as compared to the Control and Retrofit cells. TOC
measurements increased from about 800 to 1,500 mg/L in the As-Built cells while it ranged from 500 to
800 mg/L in the Control cells. Higher TOC concentrations could also have been caused by an increase in
the rate of waste decomposition in the As-Built cells, which may have resulted in higher concentrations of
VOAs, as discussed earlier. TOC concentrations in the Retrofit cells were lower compared to the other two
treatment units and appeared to be stable ranging from 200 to 500 mg/L, which is again consistent with a
stable phase of CH4 production. A decrease in TOC concentration of about 600 mg/L was observed in
Control cell A following the end of 2004.
zuuu -
5
If ^500 -
o
£4
0
•| 1 ooo -
o
0
'a
6
13 500-
S
0 -
• Control Cell A
0 Control Cell B o
•
O
• • ° °^
.°<> ^°
° ^ 0°
•-.. «.•
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-9 Total Organic Carbon as a Function of Time in the Control Cells
101
-------
2000
e 1soo -
o
o
H
a
o
"I
U
re
60
1000 -
13 500 -
o
H
• Retrofit Cell A
O Retrofit Cell B
0
0 •
>°8
d»
o .
°°
00
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-10 Total Organic Carbon as a Function of Time in the Retrofit Cells
i
O
o
H
o
"i 100°
o
o
O
-g 500
£
o
• As-Built Cell A
O As-Built Cell B
• O
•8
o o o
o a
o »o •
' 88° cPo*
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-11 Total Organic Carbon as a Function of Time in the As-Built Cells
102
-------
6.5 Biochemical Oxygen Demand (BOD) and Chemical Oxygen Demand (COD)
BOD is generally higher in younger leachate and it decreases more rapidly than COD over time. COD
includes the recalcitrant organic compounds (e.g., high molecular weight compounds as well as synthetics).
Consequently, as waste ages, the ratio between BOD and COD decreases (i.e., more COD in relation to
BOD). Leachate generated from waste with relatively low biodegradability was reported to have a
BOD/COD ratio lower than 0.5 (Christensen et al. 2001, Reinhart and Alyousfi 1996).
A statistical summary of the BOD and COD measurements in the leachate of the different landfill units is
presented in appendix F. Leachate BOD, COD, and BOD/COD ratios as a function of time in the Control,
Retrofit, and As-Built cells are presented in Figures 6-12 through Figure 6-20, respectively.
BOD concentrations showed an increasing trend in the Control cells and a decreasing trend in the Retrofit
and As-Built cells. Leachate from the As-Built cells exhibited BOD concentrations that were substantially
higher (about five-fold) than in the Control and Retrofit cells. This could be a consequence of the high
BOD in the beverage waste added to the As-Built cells. Retrofit cells displayed lower concentrations of
BOD than both Control and As-Built cells, which is indicative of the more mature waste in the Retrofit
cells. BOD concentrations in the Control cells ranged from 20 to 1,000 mg/L. The BOD concentration in
the Retrofit cells was typically below 200 mg/L. BOD concentrations in the As-Built cells ranged from
300 to 10,000 mg/L prior to early 2004 before stabilizing at about 250 mg/L.
COD concentrations showed an increasing trend in the Control cells, a decreasing trend in the Retrofit cells
and no apparent trend in the As-Built cells. Similar to BOD concentrations, the highest COD
concentrations were found in the leachate from As-Built cells, followed by the Control cells and the
Retrofit cells. COD concentration in the Control cells was typically below 1,000 mg/L prior to September
2003 before increasing to about 2,000 mg/L. COD concentration in the Retrofit cells decreased from about
2,000 to 1,000 mg/L during the monitoring period. COD concentration in the As-Built cells increased from
about 2,000 to 5,000 mg/L during the monitoring period.
The BOD/COD ratios are comparable between the As-Built and Control cells as they both appear to show a
decreasing trend early on with a more stable ratio as the waste ages. The BOD/COD ratio in leachate from
the As-Built cells dropped below 0.5 at the end of 2003, almost two years after initiating landfill bioreactor
activities. Similarly, the BOD/COD ratio in the Control cells dropped below 0.5 after June 2002, almost
four years after initial waste placement. These results indicate that the As-Built cells exhibited a more
rapid rate of decomposition relative to the Control cells. The BOD/COD ratio in both the As-Built and
Control cells reached about 0.1 in December 2005. The BOD/COD ratio in Retrofit cells generally
remained constant throughout the monitoring period. The ratio ranged from 0.05 to 0.08, which is
indicative of mature leachate generated from less degradable (i.e., more mature) waste. Similar to the other
parameters previously discussed, there is a clear difference between Control cell A and Control cell B
following the end of 2004 with respect to BOD and COD. Control cell A has distinctly lower
concentrations of BOD and COD compared to Control cell B.
103
-------
^
|p 10000 -
Q
O
m
•d"
§ 1000 -
g
Q
s
60
o 100 -
o
's
O
o0° o * 'o0*^**0*0 a°°o °
• o* • V> °° ° o o
• ° 0 0 o 0
% • n _ •
• «0 «
CS ^ ^ 0 90
0
o
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-12 Biochemical Oxygen Demand as a Function of Time in the Control Cells
10000 -
Q
O
O
1000 -
13 100
• •
•
o
o •
o
•••
• Control Cell A
O Control Cell B
10
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-13 Chemical Oxygen Demand as a Function of Time in the Control Cells
104
-------
^o
Q
O
O
Q
0
m
1 .U
1.4 -
1.2 -
1.0 -
0.8 -
0.6 -
0.4 -
0.2 -
00 -
• Control Cell A
O Control Cell B
0 °
0
•
0
8 •'» 8*o.o«/ •' . oSd, O.o
o • 2 8« •* o c^S^oC^O* ^•••8 »» •
• un
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-14 BOD/COD Ratio as a Function of Time in the Control Cells
10000 -
Q
O
m
•a
•
- o .
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-15 Biochemical Oxygen Demand as a Function of Time in the Retrofit Cells
105
-------
10000 -
Q
O
O
I 1000
Q
a
So
&
O
13 100 H
o
O
• Retrofit Cell A
O Retrofit Cell B
Vs
«;«"„.„"
Q.
. v.
10
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-16 Chemical Oxygen Demand as a Function of Time in the Retrofit Cells
1.6
1.4 -
1.2 -
O 0.8 H
o
Q
O 0.6 ^
0.4 -
0.2 -
• Retrofit Cell A
O Retrofit Cell B
T*
0.0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-17 BOD/COD Ratio as a Function of Time in the Retrofit Cells
106
-------
10000 -
Q
O
m
I 1000
Q
100
o
o
m
10
r8
cP
• As-Built Cell A
O As-Built Cell B
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-18 Biochemical Oxygen Demand as a Function of Time in the As-Built Cells
Sb 10000 -
£
Q
0
O
"S
| 1000 -
Q
a
60
X
0
•a 100-
^0
-------
^o
rt
BOD/COD
1.4 -
1.2 -
1.0 -
0.8 -
0.6 -
0.4 -
0.2 -
00 -
• As-Built Cell A
0 As-Built Cell B
•
O
•
O
•
• °
8
•o . . o o
0 •
• o 8 •
o OcP° ^ 0 •OO»«OTJO 0«o g3g£o
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-20 BOD/COD Ratio as a Function of Time in the As-Built Cells
6.6 Nitrogen Content
TKN is the sum of organic nitrogen plus ammonia. Ammonia concentrations should not be higher than
TKN concentrations, since they are a subset of TKN. Ammonia is the reduced, inorganic form of nitrogen
and is formed as a result of the release of organic nitrogen in waste. The conversion of organic nitrogen to
ammonia is called ammonification. As waste degrades, ammonia concentrations in leachate generally
increase as the nitrogen-bearing compounds in waste (e.g., proteins, amino acids, etc.) mineralize (i.e.,
become ammonified). As leachate matures, most of the organic nitrogen is mineralized and TKN is
(almost) equivalent to ammonia. The conversion of ammonia to nitrate (nitrification) under aerobic
conditions produces nitrite as a short-lived intermediary. Due to the anaerobic nature of leachate, nitrite
and nitrate are generally not detected. However, it is noted that the Retrofit cells were injected with
nitrified leachate as an additional source of liquids, which could lead to the detection of these compounds
in the leachate. Furthermore, the As-Built cells were injected with air, which would be expected to
facilitate nitrification and therefore increase the detection of nitrate in the leachate.
A statistical summary of the nitrogen content in the leachate of the different landfill units is presented in
appendix F. It is noted that nitrate was not detected at a frequency greater than 50 percent in any of the
landfill units. Nitrite was detected at greater than 50 percent frequency in the As-Built cells, but not in the
Retrofit or Control cells. This was likely caused by the introduction of air into the waste in the As-Built
cells. However, while nitrate also briefly spiked in the As-Built cells, it was detected at lower frequencies
than nitrite, potentially indicating incomplete nitrification. Furthermore, this effect occurred over a brief
period that quickly disappeared as conditions returned to anaerobic. Leachate nitrogen content (TKN and
ammonia) as a function of time in the Control, Retrofit, and As-Built cells, is presented in Figures 6-21
through 6-26.
Ammonia concentration trends often reflect those of TKN. TKN concentration trends in the As-Built cells
and Control cell B were similar and showed an increasing trend. TKN concentrations increased from about
250 mg/L to 1,500 mg/L and from about 100 mg/L to 1,500 mg/L in the Control cell B and As-Built cells
during the study period, respectively. An increase in TKN concentrations may be an indicator of
additional waste degradation in the Control and As-Built cells. TKN concentrations showed a decreasing
trend in Retrofit cell A and no trend in Retrofit cell B. Ammonia concentrations showed a significantly
108
-------
decreasing trend in the Retrofit cells. A decrease in TKN concentrations of about 1,200 mg/L was
observed in Control cell A following the end of 2004. The As-Built cells and Control cell B exhibited an
increasing trend of ammonia with a similar magnitude as TKN. Ammonia concentrations increased from
about 250 mg/L to 1,800 mg/L and from about 500 mg/L to 1,800 mg/L in Control cell B and the As-Built
cells, respectively. A decrease in ammonia concentrations of about 1,200 mg/L was observed in Control
cell A following the end of 2004, which is consistent with the decrease in TKN concentrations.
The Retrofit cells were injected with nitrified leachate and exhibited decreasing trends of TKN and
ammonia concentrations. The decrease in ammonia concentrations was likely related to the nitrification
process where ammonia in leachate was converted into nitrate, which counteracted the build-up of
ammonia concentrations in leachate from the Retrofit cells. However, no spike in nitrate (or nitrite)
concentrations was observed in these cells as a consequence of the additional liquids injection. This could
be expected given the strongly anaerobic nature of leachate in these cells and the relatively low volume of
nitrified leachate introduced as compared to generated leachate. It was expected that denitrification
(therefore an increase in N2 gas) of the nitrified leachate upon injection would occur; however changes in
N2 gas concentration in the Retrofit cells appeared to be intermittent and short-lived, which may be
indicative of changes in N2 concentration that were caused by operational issues (e.g., over-pulling on the
gas well field and watering out of gas wells) rather than a result of denitrification. The results discussed
above are again consistent with the more stable and mature waste in the Retrofit cells. TKN concentrations
in the Retrofit cells gradually decreased from about 1,000 mg/L to stabilize around 500 mg/L. It is
expected that the TKN concentrations in the Retrofit cells will remain at this level for decades, since
research using leaching tests on MSW has shown that the nitrogen content in leachate will be substantial
for centuries (Christensen et al. 2001). The Retrofit cells display a decreasing trend in ammonia
concentrations from about 1,500 mg/L to 500 mg/L; these concentrations are generally lower compared to
the Control and As-Built cells, indicating more stable concentrations, which is indicative of a more mature
and stable waste.
3000
%
2500 -
2000 -
H
d
60
8 1500 -
iz
2
3 1000 -
500-
• Control Cell A
O Control Cell B
O
o o
o 8
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-21 Total Kjeldahl Nitrogen as a Function of Time in the Control Cells
109
-------
i
bfj
8
3000
2500 -
2000 -
1500 -
C3
o 1000 -
500 -
0
• Control Cell A
O Control Cell B
°°8.
• o»°
0 08
oo
0°
0 •
°8 c9 *
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-22 Ammonia as Nitrogen as a Function of Time in the Control Cells
3000
2500 -
2000 -
bfj
8 1500 -
iz
3
2 1000
500 -
0
• Retrofit Cell A
O Retrofit Cell B
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-23 Total Kjeldahl Nitrogen as a Function of Time in the Retrofit Cells
110
-------
ouuu
2500 -
i
& 2000 -
bfj
O
g 1500 -
•^H i
C5
Q 1000 ^
S 1
a
}-H
.^^
500 -
0 -
• Retrofit Cell A
0 0 Retrofit Cell B
*•
0 •
o • cP
8° ° o
) • •
o ^^ 0 o •
0 O A ^ TJ O ^
° °0c*0 ;° .0°°°^ ° o X0
° *"** »° ° 0 °*
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-24 Ammonia as Nitrogen as a Function of Time in the Retrofit Cells
3000
2500 -
2000 -
bO
8 1500 -
iz
2
3 1000 -I
500 -
• As-Built Cell A
O As-Built Cell B
O O
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-25 Total Kjeldahl Nitrogen as a Function of Time in the As-Built Cells
111
-------
i
3000
2500 -
2000 -
1500 -
I 1000 -I
S
500 -
00*0
o o
0 •
• O 08
•* A
o • 0-
*• ^
000*
•
•
o •
o
• As-Built Cell A
O As-Built Cell B
6.7
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-26 Ammonia as Nitrogen as a Function of Time in the As-Built Cells
Metals
Addition of industrial liquids into the As-Built cells may have caused an increase in the heavy metals
content of the leachate. Paint and ink waste made up about 11 percent of the total liquids added to the As-
Built cells. Metals that are typically constituents of paint and ink waste include arsenic, barium, cadmium,
chromium, mercury, selenium and silver. Elevated concentrations of trace heavy metals in wastewater
streams have been reported to retard or inhibit biological processes (U.S. EPA 1987). In general, aerobic
processes are more sensitive to elevated trace metal concentrations compared to anaerobic processes, and
inhibition of biological processes (and therefore waste degradation) in MSW landfills, as a consequence of
elevated trace metal concentrations, is very unlikely since metals concentrations are well below the ranges
reported to inhibit biological processes (U.S. EPA 1987). Trace metal concentrations are generally low in
leachate from MSW landfills and are not very useful leachate parameters for a geochemical evaluation.
Iron concentrations, on the other hand, are generally substantial in leachate and may be more useful as a
monitoring parameter. Under anaerobic conditions, it is expected that more iron gets mobilized since the
reduced form of iron (Fe2+) is more soluble than the oxidized form (Fe3+), which precipitates out as oxides
and hydroxides.
A statistical summary of leachate metals, as presented in Appendix F, show that trace metal concentrations
generally do not appear to exhibit a trend. However, it is noted that aluminum, cadmium, copper, and lead
were observed at greater frequency and concentration in the As-Built cells compared to the other landfill
units. This may have been caused by the injection of industrial liquids containing paint waste and ink
waste into those cells. Arsenic and chromium were detected at 100 percent frequency in both the Retrofit
and As-Built cells. Arsenic concentrations appeared to be slightly higher in leachate from the Retrofit
cells. Arsenic is more mobile under anaerobic conditions since it is present in its more mobile (and more
toxic) trivalent state (As3+) compared to its pentavalent oxidation state (As5+). Conversely, chromium
concentrations appeared to be higher in leachate from the As-Built cells. This is consistent with the fact
that chromium is more mobile and soluble in its oxidized state (i.e., Cr6+) as compared to its reduced
trivalent oxidation state (i.e., Cr34). This difference could also be the consequence of higher concentrations
of dissolved organic carbon (which is captured in the TOC measurements), which has been shown to
increase metal mobility through the formation of DOC-metal complexes.
112
-------
Total iron concentrations in leachate as a function of time in the Control, Retrofit, and As-Built cells are
presented in Figures 6-27 through 6-29, respectively. Iron concentrations showed a decreasing trend with
concentrations gradually decreasing from about 15 mg/L to 5 mg/L and from about 30 mg/L to 5 mg/L in
the Control and As-Built cells, respectively. In the Retrofit cells, iron concentrations showed an increasing
trend in Retrofit cell A and no significant trend in Retrofit cell B and ranged from approximately 10 to 20
mg/L. Iron concentrations were higher in the Retrofit cells as compared to the Control and As-Built units.
This is consistent with more reducing conditions in the Retrofit cells compared to the Control and As-Built
cells. This is also consistent with the interpretation of arsenic concentrations above. Injection of air into
As-Built cells appears to have had a fairly long-lasting effect as iron concentrations are low. The iron
levels are likely to rebound as the waste matures and the potential effect of the aeration dissipates.
However, a longer monitoring period is required to support more conclusive interpretations with respect to
iron concentrations.
6.8 Volatile and Semi-Volatile Organic Compounds
The occurrence of VOCs and SVOCs in leachate may be of use in evaluating the landfill bioreactor's
capacity for microbial assimilation and transformation of organic and potentially toxic compounds. A list
of VOCs and SVOCs that were measured in the leachate of the different landfill units is presented in
Chapter 3. A statistical summary of the VOCs and SVOCs detected at greater than 50 percent frequency in
the different landfill units is presented in Appendix F. Similar to the discussion regarding trace metals,
VOCs and SVOCs were generally only detected at fairly low concentrations in leachate from MSW
landfills and meaningful trends are generally hard to discern. Overall, the concentrations of VOCs detected
in leachate of the different treatment units were consistent with similar landfill settings. Acetone, MEK,
and toluene were detected at a greater frequency and concentrations in the As-Built cells compared to other
units. However, these compounds are generally quickly degraded under aerobic conditions and do not pose
challenges for leachate management.
50
8
40 -
30 -
20 -
10 -
• Control Cell A
O Control Cell B
O
o •
00*
o o
• o
8 8
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-27 Total Iron as a Function of Time in the Control Cells
113
-------
o
o
H
50
40 -
30 -
20 -
10 -
• Retrofit Cell A
O Retrofit Cell B
O O
O
o
o e
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-28 Total Iron as a Function of Time in the Retrofit Cells
50
p
40 -
30 -
20 -
10 -
• As-Built Cell A
O As-Built Cell B
O O
o o
• 8
o
o
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-29 Total Iron as a Function of Time in the As-Built Cells
114
-------
Similarly, leachate from MSW landfills may contain some SVOCs including phenolic compounds such as
phenols and cresols. Low concentrations of cresols were detected in the As-Built cells at higher
frequencies and concentrations compared to the other cells. This may be a consequence of the addition of
paint waste. SVOCs and VOCs may have a small effect on COD measurements. In general, the VOC and
SVOC concentrations were relatively low and there were no observed trends of VOC and SVOC
concentrations in the leachate samples.
6.9 Phosphorous Content
Phosphorous may be one of the rate-controlling macro nutrients in landfills. Since leachate samples were
not filtered prior to analysis, a high phosphorous concentration may be an indicator of microbial growth as
it was incorporated into the microbial cell mass, which was analyzed as total phosphorous. Mineralization
of organic phosphorous in the waste generates inorganic phosphorous. In its inorganic form, phosphorous
is typically measured as ortho-phosphate. Leachate ortho-phosphate and phosphorous concentrations as a
function of time in the Control, Retrofit, and As-Built cells, are presented in Figures 6-30 through Figure 6-
35.
Phosphate concentration trends in the As-Built cells and Control cell B are similar; both show an increase
over time with concentrations in the As-Built cells being slightly higher. The Retrofit cells exhibited a
slightly increasing trend with lower concentrations when compared to the As-Built and Control cells.
Phosphate concentrations were below 3 mg/L before they gradually increased up to 15 mg/L after the end
of 2003 in the Control cells. The increase in phosphate may be an indicator of increased decomposition
following the end of 2003 as organic phosphorous in waste became mineralized. On the other hand,
microbial growth may account for an increase in total phosphorous concentrations, which would also be an
indirect measurement of increased decomposition as an increase in microbial cell mass, suggesting an
increase in waste decomposition.
Total phosphorous concentration trends are generally a reflection of phosphate concentrations. Similar to
phosphate, phosphorous concentrations in the Control cells increased from below 5 mg/L to about 15 mg/L
following the end of 2003. Similar to other parameters discussed above, a decrease in phosphate and
phosphorous concentrations of about 5 mg/L was observed in Control cell A following the end of 2004.
The Retrofit cells displayed a slightly increasing trend with lower concentrations of phosphate and
phosphorous compared to Control cell B and the As-Built cells. Phosphate and phosphorous concentrations
were at or slightly below 5 mg/L in the Retrofit cells.
The As-Built cells displayed an increasing trend of phosphate and phosphorous concentrations, similar to
that in the Control cells. Phosphate concentrations were mostly below 5 mg/L before they increased up to
18 mg/L following mid 2004. Similarly, total phosphorous concentrations were below 10 mg/L before
increasing up to 25 mg/L following mid 2004. The increase in phosphate and phosphorous may be an
indicator of increased waste decomposition and microbial growth following mid 2004. However, it cannot
be excluded that the addition of the beverage waste to the As-Built cells increased the phosphorous
concentrations in these cells, since many beverages contain phosphoric acid; however, this cannot be
verified with the available analytical data for the beverage waste
Leachate carbon to phosphate ratio (C: P) was determined for the different treatment units. For the Control
cells, the C:P ratio mean and standard deviation were 124.1 + 69.5 (Control cell A) and 124.2 + 53.3
(Control cell B). For the Retrofit cells, the C:P ratio mean and standard deviation were 105.3 + 40.4
(Retrofit cell A) and 121.2 + 53.4 (Retrofit cell B). For the As-built cells, the C:P ratio mean and standard
deviation were 160.1 + 77.7 (As-Built cell A) and 181.0 + 58.5 (As-Built cell B). A C: P ratio of 60:1 is
deemed optimal for microorganisms to actively assimilate substrate carbon.
115
-------
30
25 -
bb 20 -
f 15 H
o
°r
o
5 -
• Control Cell A
O Control Cell B
8
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-30 Ortho-Phosphate as a Function of Time in the Control Cells
25 -
t 20-
I
£ 15 -
ft
o
PH
O
H
5 -
0 -
• Control Cell A
0 Control Cell B
O
O
00
o
0
•••
0 »8°0 00 n 0°0
o 0 ° ** 0°
O .£) •
O Q^3 .O O O
85 ft* 0 8 °r«D^3» •%.•*«•*
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-31 Total Phosphorous as a Function of Time in the Control Cells
116
-------
30
25 -
20 -
•a 15
I
PH
10-
5 -
• Retrofit Cell A
O Retrofit Cell B
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-32 Ortho-Phosphate as a Function of Time in the Retrofit Cells
30
25 -
20 -
43 K
a
Jj
PH
13 10 -I
o
H
5 -
• Retrofit Cell A
O Retrofit Cell B
#
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-33 Total Phosphorous as a Function of Time in the Retrofit Cells
117
-------
30
25 -
20 -
•3 «
a 15
I
5 -
• As-Built Cell A
O As-Built Cell A
«
O O
•
o . o • ° ° •
-------
the nature of the landfilled waste changes). Furthermore, chloride is not redox sensitive and should not
change much as a result of the different treatment approaches.
Leachate chloride concentrations as a function of time in the Control, Retrofit, and As-Built cells are
presented in Figures 6-36 through Figure 6-38. Interestingly, chloride concentrations appeared to show an
increasing trend over time in both the Control and the As-Built cells, although it is observed that the initial
chloride concentration in the Control cells was lower than the other cells. Changes in chloride
concentrations are generally more a reflection of dilution rather than a result of waste decomposition.
Chloride concentrations in the Control cells were below 1,000 mg/L prior to mid 2003 and increased to a
peak at about 5,000 mg/L before decreasing again at the end of 2004. Consistent with most parameters
presented in this section, chloride concentrations in Control cell A decreased compared to Control cell B
following mid 2004. In the Retrofit cells, chloride concentrations showed no trend and ranged between
1,000 mg/L to 2,000 mg/L. Chloride concentrations in the As-Built cells showed an increasing trend with
concentrations increasing from about 1,000 mg/L to 3,000 mg/L. The increase in chloride concentrations
in the Control and As-Built cells may have been caused by the ongoing waste placement in those cells until
2004 and 2005, respectively. It is expected that these concentrations will stabilize over time.
6.11 Leachate Quality Summary
In summary, all parameters measured provide a consistent geochemical picture for each cell with respect to
each cell's waste age and waste decomposition phase. The As-Built cells, which received air and liquids
injection, have the youngest waste age and appear to display the most active phase of decomposition.
Trends are generally not much different in the bioreactor landfill cells compared to the Control cells, even
though concentrations of some parameters appear to be higher. The substantial difference in temperature,
however, may indicate more active decomposition in As-Built cells, possibly reflective of the exothermic
nature of aerobic MSW decomposition. Although the overall waste age in the As-Built cells is less than
that of the Control cells, the geochemical analysis of the leachate indicates that the cells are approximately
at the same stage of waste decomposition. Judging from the reviewed data, it can be hypothesized that the
combined treatment of air and liquids appears to have accelerated waste decomposition in this
investigation. The Retrofit cells, which received nitrified leachate, did not show any significant signs of
accelerated waste decomposition based on the leachate chemistry. This is likely the result of waste age (as
much as six years old when treatment began), which is significantly older than both the Control and As-
Built units. The waste in the Retrofit cell may have already reached a more mature stage prior to the
addition of supplemental liquids, which is reflected in stable and lower concentrations of many leachate
parameters reviewed. From a geochemical perspective, it appears the waste may have already gone
through most of its decomposition; thus, the injection of liquids did not increase waste decomposition in the
Retrofit cells. The addition of nitrified leachate to the Retrofit cells resulted in a significant decrease of
ammonia concentrations in the leachate of the Retrofit cells.
119
-------
I
.
o
6000
5000 -
4000 -
3000 -
2000 -
1000 -
• Control Cell A
O Control Cell B
o • o
8 o °
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-36 Chloride Concentrations as a Function of Time in the Control Cells
6000
5000 -
4000 -
3000 -
2000
1000 -
o -(o
• Retrofit Cell A
O Retrofit Cell B
O O
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-37 Chloride Concentrations as a Function of Time in the Retrofit Cells
120
-------
I
.
o
6000
5000 -
4000 -
3000 -
2000 -
1000 -
• As-Built Cell A
O As-Built Cell B
•
O
0 O
•
• o
•
o
o
0
6/1/01 12/1/01 6/1/02 12/1/02 6/1/03 12/1/03 6/1/04 12/1/04 6/1/05 12/1/05
Date
Figure 6-38 Chloride Concentrations as a Function of Time in the As-Built Cells
121
-------
This page is intentionally left blank.
122
-------
Chapter 7. Landfill Bioreactor Performance Analysis
As described in Chapter 1 of this report, one of the primary objectives of the OLLB study was to evaluate
whether the benefits of landfill bioreactors as reported in laboratory- and pilot-scale projects could be
realized at a full-scale and operating MS W landfill. The previous chapters of this interim report focused on
describing the tests that were performed to evaluate the performance of the OLLB and presenting an
interpretation of the test results. This chapter provides a general discussion of the overall performance of
the OLLB with respect to certain permitting issues that are relevant to landfill bioreactors. The issues
addressed in this chapter are slope stability, liner and cover integrity, liquids addition system performance,
LCS performance, and LFG production and emissions. Relevant sections of the "Solid Waste Disposal
Facility Criteria Technical Manna?" (U.S. EPA, 1993) are referenced, where appropriate.
7.1 Slope Stability
MSW Landfills must be designed and operated in a manner that maintains the stability of natural and waste
slopes. Based on evaluations of past failures of both natural slopes and waste slopes, it is widely known
that excessive amounts of liquids within soils or waste can contribute to failure and/or movement of slopes.
The failure to control liquids within a slope or a landfill could result in problems related to slope stability.
Because bioreactor landfill operations involve adding liquid to the wastes, it is appropriate to pay
particularly close attention to the potential destabilizing effects of liquids at such facilities. Excess liquids
can result in saturation and a decrease in the effective stress within the waste, with a corresponding
decrease in effective shear strength. Control of excess liquid (i.e., "pore") pressures in a bioreactor landfill
is provided by minimizing the potential for large portions of the landfilled waste to become saturated. This
can be provided by: (i) controlling liquid application rates; (ii) maintaining the LCS operation and
preventing the LCS from being overloaded; and (iii) controlling exit gradients (i.e., seeps), usually
achieved by providing a minimum of 15 - 30 m (50 to 100 ft) between landfill slopes and the end of an
injection pipe and/or trench.
As described in Section 4.2 (i.e., Moisture Addition) of this report, the data from the OLLB collected under
the CRAD A to date show that: (i) liquid application rates resulted in increased moisture content but not
saturation of the waste as some of the solid waste samples examined had low moisture content; (ii) the LCS
remained functional and showed no signs of excessive head buildup; and (iii) there was no evidence of
excessive seeps occurring. Based on these findings, it is concluded that operation of the OLLB has not
caused conditions that would decrease the stability of natural or waste slopes resulting from the
development of excess pore pressures. Note that adding liquid increases the weight of the waste and,
therefore, could reduce the overall stability of the waste. Although evaluation of this aspect of slope
stability is beyond the scope of this report, it is likely that the increased weight resulting from moisture
content increases (which are reported in Section 4.2 to be about, on average, six to seven percent in the As-
Built cells and one percent in the Retrofit cells) would not have a significant adverse effect on stability if
the slopes were originally designed having a factor of safety meeting the recommendations provided in
U.S. EPA (1993).
7.2 Liner and Final Cover Integrity
The integrity of the base liner is of particular concern for bioreactor landfills because of the need to contain
the additional amount of liquid that is collected and removed from atop the liner of a bioreactor landfill. In
addition, the integrity of the final cover is of concern for a bioreactor landfill because it may be exposed to
more seeps and differential settlement than a non-bioreactor landfill. The study reported in this document
was not designed to evaluate liner and final cover integrity, and so this study does not directly provide
information for evaluating effects of bioreactor operations on liner or cover integrity. However, it is
important to note that a recent U.S. EPA study on the performance of Subtitle D landfill liner systems of
landfills that are younger than 15 years old (U.S. EPA 2002) confirmed that the integrity of well-
constructed liner systems that are properly monitored during construction is not jeopardized over the
operating life of the facility. Leakage in the base liner is usually associated with poor construction,
123
-------
inadequate construction quality assurance (CQA) monitoring, and construction of features near the liner.
The best assurance that the liner system will perform well for operating landfill bioreactors is to provide a
high-quality construction and CQA monitoring (paying particular attention to sealing around all
penetrations) and installing liquid application features in a manner that does not damage the liner (e.g.,not
installing liquid application wells to a depth near the liner). Based on a review of the installation of the
liquid application features for the OLLB project, there is no indication that the base liner system at the
OLLB was compromised while installing liquid application features or while applying liquid through those
features at this point of the project.
The integrity of the final cover could be compromised if excessive differential settlement occurs or if slope
instability occurs. These problems could be worsened if landfill bioreactor operations increase the amount
of differential settlement or increase the pressure on the bottom of the cover system (e.g., liquid pressure or
gas pressure acting on the bottom of the geomembrane or clay barrier layer), which could lead to
instability. Because final cover has not yet been installed over the cells that were part of the OLLB study,
an evaluation of these potential effects. However, the lack of leachate breakout problems and the lack of
landfill gas emission problems suggest that the OLLB has been operated in a manner that will minimize
problems related to excessive pressures under the liner. Although no similar information is available to
evaluate the potential for significant differential settlements, such settlements (if they were to occur) could
be easily repaired. Based on these considerations, impacts on the integrity of the OLLB final cover are not
expected to be a problem.
7.3 Liquids Addition System Performance
The performance of the liquid addition system can have a significant effect on other systems present at a
landfill. Systems that could be affected include the surface-water management system (which could be
affected if seeps from the landfill impinge on drainage features), LFG management system (which could be
affected if gas extraction wells become filled with liquid), or the LCS (which is discussed separately in
Section 7.4). Based on observations at the site, no significant leachate breakout problems were reported
that would have resulted in surface-water quality effects. Therefore, the operational practices used at the
OLLB appear to have been successful in preventing surface-water quality problems. Also, although there
were indications of "watering-out" of gas collection wells and trenches during some periods of the study,
there was also an increase in landfill gas collected from the bioreactor areas and so it is not possible to
conclude whether the watering-out of the collection features would be expected to have an impact on future
stability of the landfill cover system; see Section 8 for recommendations on future monitoring and
performance evaluation of the LFG management system. Finally, as discussed in the body of the report,
wastes having a low pH and/or high sugar levels are expected to have an adverse effect on waste
degradation and should be avoided unless lab- or bench-scale testing is performed to assess the potential
effects of the candidate liquid waste stream on the microbial populations and resulting performance of the
landfill environmental protection systems.
7.4 Leachate Collection System Performance
If liquid application activities promote microbial activity within LCS features, then it is conceivable that
those features could become clogged. In that case, the LCS features might not function as originally
intended and cease to transmit leachate from the landfill as designed. Upon achieving field capacity or if
preferential flow paths develop in the solid waste, there is a concern that the amount of liquids that pass
into and through the LCS will increase, thus raising the concern regarding the integrity of the LCS. This
concern could extend from an increase in flow volume or flow rate in the drainage materials or the
premature "fouling" of the pores in either the protective geotextile or the hydraulically transmissive
granular drainage component of the LCS. Although it is certainly recognized that this can be a potential
problem, the authors are not aware of any case where the concern has been suspected, studied, and
confirmed. It is encouraging to note that the effect of the landfill bioreactor performance is to merely
accelerate the waste degradation processes, not introduce new mechanisms or processes. Therefore,
components or construction techniques that are effective for the LCS systems for conventional landfills are
anticipated to be effective for landfill bioreactor landfills. Because of the likelihood for additional
microbial activity and higher flows rates in landfill bioreactors compared to conventional landfills, it may
124
-------
be advisable to consider increasing LCS hydraulic conductivity and transmissivity for landfills bioreactor
projects. In addition to the construction and materials selection controls, it is recommended that operating
practices that result in the potential for clogging or blinding of the LCS components be avoided. For
example, if waste streams are identified that may react and introduce precipitates (e.g., aluminum dross),
these waste streams should be avoided. Similarly, the use of fine-grained soils near the bottom of the waste
mass should not be used, as they have the potential for migration into the LCS.
Signs of clogging of the LCS can include reduced leachate collection rates, watered-out landfill gas
collection wells, or breakouts at the sides of the landfill. At the OLLB, none of these conditions was
observed to the degree that would imply clogging of the LCS features. However, this finding should be
confirmed again in future studies.
7.5 Landfill Gas Production and Emission
A final concern regarding landfill bioreactors is the emissions from the facility that result from enhanced
LFG generation. If LFG emissions at a bioreactor landfill are greater before capping than for a non-
bioreactor landfill, then it could be expected that a greater quantity of landfill gases would be emitted to the
atmosphere from the bioreactor landfill. Chapter 5 indicates that, at the OLLB, LFG production in the As-
Built cells was significantly higher than would be expected for a conventional dry tomb landfill and slightly
higher in the Retrofit cell. However, despite the higher gas production rates during bioreactor operations,
there does not appear to have been excessive surface (surface scans with greater than 500 ppm CH4)
emissions from the facility.
125
-------
This page is intentional left blank.
126
-------
Chapter 8. Conclusions and Recommendations
8.1 Conclusions
The data set from the five-year study at the OLLB represents one of the most comprehensive compilations
of information regarding the performance of landfill bioreactors. The study was conducted at a full-scale
operating MSW landfill and, as such, was faced with the challenges of performing leading-edge research at
an operating landfill site. The objectives of the OLLB study were presented in Chapter 1; a summary of
these objectives and their status (i.e., "met", "not met", or "on-going") as of the date of this Second Interim
Report is presented in Table 8-1.
Table 8-1 OLLB Objectives Assessment
Objective
Design and implement two
alternative large-scale bioreactor
operations.
Monitor sufficient parameters to
understand the physical, chemical,
and biological activities within the
landfill bioreactors.
Compare and contrast monitoring
results with the requirements for a
Subtitle D Landfill.
Incorporate statistical techniques to
assess effectiveness of the landfill
bioreactor operation.
Establish best practices and
procedures required to operate
landfill bioreactors.
Establish important and indicative
parameters that should be monitored
with respect to landfill bioreactor
operations.
Obtain sufficient data to enable
improvements that may be applied
to future bioreactor landfills, in both
an experimental and practical
capacity.
Met / Not Met /
On-going
Met
On-going
On-going
On-going
On-going
On-going
On-going
Comment
The Retrofit and As-Built cells represent the
two alternative bioreactor operations.
Several conclusions are presented in this report;
however, research and performance monitoring
continues at the site.
Many comparisons between the bioreactor and
Control cells are provided in this report;
however, research at the site is on-going.
Statistics were utilized to evaluate data for all
three media (solid, liquid, and gas). More data
are being collected at the site for future
statistical assessment.
Additional data collection and analysis will
further refine conclusions reached in this report,
leading to a complete evaluation of the
operational techniques used.
Monitoring parameter recommendations for
bioreactor landfills in general are provided
below in Section 8.2. Ongoing work at the site
will expand on these recommendations.
Recommendations for additional data collection
parameters are provided in Section 8.3.
8.1.1
Solids Analysis
The intent of the solids decomposition analysis was to assess the reproducibility of the data between the
cells and their replicates (i.e., Control cells A and B, Retrofit cells A and B, and As-Built cells A and B)
and to evaluate whether the solids decomposition data support the concept that operation as a bioreactor
landfill accelerates waste decomposition relative to operation as a conventional landfill. The assessment of
cell replicates indicated that the waste in the Retrofit and Control cell replicates were similar, but As-Built
cells A and B were found to be significantly different. The differences between As-Built cell A and As-
Built cell B were attributed to the fact that the waste in the two cells was relatively young.
The three primary parameters used to assess the extent of decomposition were BMP, CH:L, and organic
solids content. Generally, a decrease in these parameters indicates an increase in the degree of waste
decomposition. CELL and organic solids content were found to be significantly different in the Retrofit
127
-------
cells compared to the Control cells; however, linear regressions of the data indicated that BMP was similar
in the Retrofit and Control cells. All three decomposition parameters were less in As-Built cell A relative
to both the Control cells and As-Built cell B, indicating more rapid waste decomposition. The accelerated
waste decomposition in As-Built cell A relative to As-Built cell B was attributed to the much larger volume
of air injected into As-Built cell A (approximately five times greater on a mass basis). However,
regressions of solids data for As-Built cell B did not indicate a difference when compared to the Control
cells. One limitation of the analysis of the As-Built cells was the young age of the waste, which precluded
significant trends in waste decomposition to develop. An additional limitation in developing strong trends
in decomposition was the inherent heterogeneity of the MSW. Despite nearly five years of operation, the
solids analysis results still indicate that, although there are statistical differences in waste decomposition
between the three distinct cells, the waste has not undergone significant solids decomposition. It is
expected that data from sampling and testing of solid waste from these cells in subsequent years will show
more consistent solids decomposition trends.
8.1.2 Liquids Analysis
The liquids analysis involved assessing the moisture balance of the different landfill cells, evaluating
leachate head on the liner, and evaluating leachate quality data. Data showed an inconsistent trend in
moisture content with respect to waste age and the type of landfill cell. For instance, the mean measured
moisture content for the Control cells showed an overall increase of approximately six percent between
2002 and 2005, which is a large increase considering that no moisture was added to the Control cells during
the study period. Overall, measured waste moisture content data as obtained from moisture content
measurements on discrete waste samples should be interpreted with caution; variability in waste
composition, hydraulic properties, compaction rates, and other factors likely contributed to the inconsistent
behavior of the measured moisture content data.
A moisture balance was conducted using a combination of information related to rainfall, liquids addition,
and leachate generation quantities. The initial moisture content percentage used in the calculation was
based on the measured moisture content of samples collected in 2002, prior to initiation of liquids addition
in the bioreactor landfills cells. This assumption has obvious limitations in that it represents data based on
discrete samples from a single sampling event. As described earlier, the measured moisture content data
showed inconsistent trends in all cells during the study period; however, this was assumed to be a more
appropriate starting point compared to literature-reported moisture content values. Liquid addition rates of
5.3, 65.3, and 43.4 gallons/ton in the Retrofit cells, As-Built cell A, and As-Built cell B resulted in a
calculated moisture content increase (on a wet weight basis) of approximately one, six, and seven percent,
respectively. The slight increase observed in the Retrofit cells is attributed to a low liquid injection volume
at those cells relative to the overall mass of the cells, as well as the cover system that minimized infiltration
of rainfall. The six and seven percent moisture content increase in As-Built cell A and As-Built cell B is
believed to be conservative relative to the "Liquid In" volumes that likely infiltrated the waste mass in
these cells. Prior to the first quarter of 2004, a0.3m(l ft) thick layer of tire chips was installed between the
As-Built cells, resulting in short-circuiting of injected liquids to the LCS. This led to an overall increase in
the "Liquid Out" value used in the moisture balance calculation prior to the first quarter of 2004. When the
practice of installing the tire chip layer between the As-Built cells ceased following the first quarter of
2004, the rate of leachate generation slowed significantly, indicating that a greater increase in moisture
content may have occurred in the As-Built cells (based on the Liquid In volumes) if the tire chip layer had
not been installed prior to 2004. Data following the first quarter of 2004 indicated a significantly lower
leachate generation rate in the As-Built cells. Nevertheless, the moisture balance calculation results for the
bioreactor landfill cells underscore the fact that very large volumes of liquid are required to significantly
increase the moisture content of a landfill. Despite efforts to significantly increase the moisture content, it
appears that additional liquids could have been assimilated by the waste if the operation procedures had
been different.
Leachate head in the sump for each cell was measured to provide an indication of the leachate head on the
liner. The mean leachate head in the sump for each cell was below the regulatory threshold of 0.3 m (1 ft).
However, exceedances of head on the liner did occur occasionally in all cells; the exceedances generally
corresponded with the active filling phase of a particular landfill cell, and the frequency of exceedances
128
-------
decreased after the final lift of waste had been placed. The exceedances during waste filling were primarily
attributed to inadvertent routing of runoff from the cell to the sump.
The leachate quality analysis results were generally consistent with the corresponding calculated age of the
waste for each cell throughout the study period. However, the results of the leachate quality analysis for
the majority of the parameters did not reveal a strong trend that would indicate significantly accelerated
waste decomposition in the bioreactor landfill cells. Still, the results of some of the parameters were
encouraging. For example, the BOD/COD ratio in the As-Built cells decreased to below 0.5 within two
years of initial waste placement, which is indicative of waste decomposition. In comparison, the
BOD/COD ratio for the Control cells did not decrease to below 0.5 until approximately four years after
initial waste placement, supporting the premise of accelerated waste decomposition in the As-Built cells.
Additionally, Chapter 2 described the ex-situ nitrification process that was used to treat the leachate prior to
injection into the Retrofit cells; this process, implemented with the intent of decreasing the ammonia
concentration, was expected to have an increasing benefit over time as a result of increased waste
decomposition. The mean ammonia concentration in the Retrofit cells showed a strongly decreasing trend
during the study period, which indicates that treating the leachate ex-situ was effective at decreasing the
ammonia concentration in the leachate.
8.1.3 LFG Analysis
The goals of the LFG analysis were to evaluate the LFG kinetics in each cell, draw comparisons between
the cells, and assess the NMOC concentration data for each cell. Using the field-collected LFG data and
in-place waste mass as inputs, the U.S. EPA model LandGEM was used to estimate waste decay rates for
each cell. Assuming a 75 percent LFG collection system efficiency, waste decay rates of approximately
0.16, 0.061, and 0.06 yr"1 were estimated for As-Built cells, the Retrofit cells, and the Control cells,
respectively. It is noted that there are some limitations to the estimation of these decay rates. First, the
LFG collection efficiency was assumed based on the AP-42 recommended value of 75 percent. Second,
the estimation of the decay rates was based on only three years (As-Built cells) or four and a half years
(Retrofit cells) of field-collected data. Third, the CH4 generation potential, L0, used in LandGEM was
assumed to be constant and was selected based on the analysis of fresh waste, although this value likely
does not remain constant over time as waste degrades. Despite these limitations, the results strongly
indicate that the LFG generation rate was accelerated in the As-Built cells relative to the Control cells. The
results also indicate that the LFG generation rate in the Retrofit cells appeared to be somewhat greater than
in the Control cells; however, the LFG flow and composition data from the Retrofit cells indicated a degree
of variability that was attributed to the watering out of LFG collection wells. As a result, the LFG
collection efficiency in the Retrofit cells may be less than the assumed 75 percent, which would indicate
that the waste decay rate in the Retrofit cells may be greater than 0.061 yr:.
The NMOC concentration data were evaluated in two ways: (i) NMOC production based on the measured
NMOC concentration and LFG flow data from each cell; and (ii) NMOC production based on the NMOC
concentration data and mass of in-place waste for each cell. The first method was developed to tie in the
years of field gas data to the NMOC concentration and to investigate any trends. The second method was
similar to the method prescribed in a typical Tier 2 NSPS analysis, in which a site-specific NMOC
concentration is used in LandGEM to predict NMOC production based on waste placement data. Analysis
of NMOC production using the first method indicated little difference between the cells, resulting in a
mean NMOC production rate of 0.012, 0.012, 0.010, and 0.008 kg/min for the Control cells, Retrofit cells,
As-Built cell A, and As-Built cell B, respectively. Analysis of NMOC production using waste placement
data and the mean NMOC concentration for each cell indicated NMOC production rates of less than 50
Mg/yr for each cell; it is noted that the predicted NMOC production using LandGEM was performed
primarily to observe whether each cell exceeded the regulatory threshold of 50 Mg/yr of NMOCs.
8.1.4 Summary
Overall, the performance monitoring results are very encouraging. The most significant conclusion is that,
despite the five-year duration of the study, two of the three media analyzed (i.e., solids and LFG) indicated
that waste decomposition was accelerated in the As-Built and Retrofit cells relative to the Control cell.
Table 8-2 summarizes a comparison of the different waste decomposition indicators for each media (solid,
liquid, or gas) between the Control cells and the bioreactor landfill cells.
129
-------
For the Retrofit cells, two of the three solids monitoring parameters indicated accelerated decomposition
relative to the Control cells; therefore the results were believed to be inconclusive. Furthermore, for the
Retrofit cells, the generally stable leachate data did not indicate that waste decomposition was accelerated
as a result of liquids addition. The LFG results indicated a waste decay rate that was similar to that of the
Control cells; the authors believe that if the actual LFG collection efficiency in the Retrofit cells could be
established, it would likely reveal an efficiency of less than 75 percent, thereby resulting in a greater
estimated waste decay rate.
For As-Built cell A, all three solids monitoring parameters indicated a trend of accelerated waste
decomposition relative to the Control cells. The BOD/COD ratio of the leachate indicated that perhaps
waste decomposition was accelerated, but that ratio alone is not enough to conclude that the leachate
indicated accelerated waste decomposition. At the other extreme, the waste decay rate as estimated by
using the LFG collection data in the As-Built cells was several times greater than the Control cells.
Not all of the solids monitoring parameters indicated accelerated waste decomposition in As-Built cell B
relative to the Control cells. The authors believe that future solid waste sampling will likely indicate
accelerated decomposition of waste in this cell. Similar to As-Built cell A, As-Built cell B exhibited a
decrease in BOD/COD ratio, indicating that waste decomposition was accelerated relative to the Control
cells, but as stated above this is not enough to conclude that the leachate results (by themselves) indicated
accelerated waste decomposition. LFG data from As-Built cell B indicated a waste decay rate several times
that of the Control cells.
Table 8-2 Evidence of Accelerated Waste Decomposition Across All Media Analyzed Between
Control and Bioreactor Landfill Cells
Bioreactor
Landfill Cell
Retrofit cells
As-Built cell A
As-Built cell B
Solid
I
Yes
TBD
Liquid
No
I
I
Gas
TBD
Yes
Yes
Notes:
1. TBD - To be determined; a conclusion will likely be reached after additional data is collected.
2.1- The results of the waste decomposition data were positive, but inconclusive.
8.2 Recommendations
As Table 8-2 shows, and as described in Sections 8.1.1 through 8.1.3, ongoing performance monitoring is
anticipated (and is recommended) to continue at the OLLB. The parameters reported in Chapter 2 will
continue to be monitored for the next several years. The authors recommend that additional analyses be
considered. The results of these additional analyses could support the conclusions presented in this report.
Additional recommendations are provided below:
• Collect data that can be used to estimate LFG collection efficiency for each cell. This could be
accomplished by using U.S. EPA Method 2E to estimate the LFG generation rate and indirectly
estimating the CE by comparing the LFG generation rate with the LFG collection rate.
Alternatively, the fugitive emissions for each cell may be estimated using a technology such as
open path Fourier-transform infrared spectrometry. Fugitive emissions at the OLLB were
previously reported using this technique (U.S. EPA 2005), but the measurements were taken in
2002 and 2003, prior to the end of waste placement in the As-Built and Control cells.
• Continue collecting topographic survey data of the Control, Retrofit, and As-Built cells. The
authors note that it may take several more years before a quantitative correlation between waste
decomposition and settlement can be made for the landfill bioreactor cells.
• Provide more detailed sample identification information during the collection of solid waste
samples, especially with regard to the sample proximity of a liquid addition pipe or trench. The
authors believe this information may provide insight related to the variability of the measured
moisture content data. For example, during solid waste sample collection, note the distance of the
borehole to the nearest injection pipe/trench, and identify the volume of liquids added to the
corresponding trench/pipe, if that information is available.
130
-------
It may only be necessary to collect one more set of solid waste samples over the next approximately five
years to analyze the various decomposition parameters. As noted in Section 4.1, a high degree of scatter
seen in the data may be attributable to the relatively short time window during which the samples were
collected, resulting in relatively young waste.
8.3 Landfill Bioreactor Monitoring Parameters
The OLLB will continue to monitor the parameters (presented in Chapter 2) at the same frequency for the
next several years under its RD&D permit. As a result, the trends of solids decomposition, moisture
balance, LFG composition and flow, and leachate quality will be further developed. It is expected that the
variability seen in some trends (e.g., solids decomposition) will tend to decrease during this period, as the
study cells are no longer active (i.e., the study cells no longer receive new waste). Based on the results
from the first five years of monitoring at the OLLB, the following observations can be made regarding the
monitoring parameters and frequency described in Chapter 2.
The mass balance and liquid addition monitoring parameters listed in Tables 2-1 and 2-2 are considered to
be a fundamental part of the OLLB program and any landfill bioreactor project. The liquids added and the
nature and amount of the deposited material in the landfill cell provide the basic information for conducting
moisture balancing and predicting LFG generation using LandGEM. The monitoring frequency listed in
Tables 2-1 and 2-2 was appropriate for the first five years of the OLLB project.
LFG monitoring of CH4, CO2, O2, and balance gas once per week appeared to be an appropriate monitoring
frequency for the OLLB study. Monitoring at this frequency minimized significant variability that may
have occurred with less frequent monitoring; however, even with weekly sampling, there was a relatively
high degree of variability in the As-Built and Retrofit bioreactor landfill cells. NMOC monitoring,
however, appeared to be more frequent than needed. The comparison of NMOC production between cells
did not indicate a significant difference between the bioreactor landfill cells data and the Control cell data.
The authors believe an NMOC sampling frequency of once per year is adequate to further develop the
trends (or lack thereof) seen in the first five years of the OLLB project. Furthermore, the authors believe
that an NMOC sampling frequency of once per year at other landfill bioreactor sites may be appropriate.
Solid waste sampling and analysis parameters used for quantifying solid waste decomposition were
appropriate; however, the period of the study (i.e., only five years) appears to have been too short to see
significant trends, even in cells where large decomposition rates were anticipated. There was significant
variability in BMP, CH:L, and organic solids between the three sampling periods; the authors believe that
the scatter observed in these parameters may have been caused by the relatively short time window
analyzed in this study. For bioreactor landfills in general, the authors believe that it is critical to obtain a
sufficient number of solid waste samples for characterization at the outset of a bioreactor landfill project
and less frequent monitoring (e.g., every five years) be conducted following the initial sampling period.
Analyses on samples of solid waste are important as they provide the most direct evidence of waste
decomposition. However, the solids analyses are performed on discrete samples and the inherent variability
in waste composition may "mask" the indicators of decomposition during the short monitoring time period
that has elapsed so far for the OLLB.
131
-------
This page is intentionally left black.
132
-------
Chapter 9. References
Aguilar, A.; Casas, C.;and J. Lema (1995). "Degradation of Volatile Fatty Acids by Differently Enriched
Mehtanogenic Cultures: Kinetics and Inhibition." Water Research 29: 505-509.
American Public Health Association (1999). Standard Methods for the Examination of Water and
Wastewater, 21st Ed., 1999.
Amman, R., Ludwig W., and K-H. Schleifer. (1995). "Phylogenetic identification and in situ detection of
individual microbial cells without cultivation" Microbiological Reviews, 59: 143-169.
Bachus, R., Houlihan, M., Kavazanjian, E., Isenberg, R., and J. Beech (2004). "Landfill Bioreactor
Stability: Key Issues", Municipal Solid Waste Management, Sept/Oct.
Barlaz, M., Ham, R., and D. Schafer (1989). "Mass-Balance Analysis of Anaerobically Decomposed
Refuse." Journal of Environmental Engineering, 115(6): 1088-1102.
Barlaz, M., Ham, R. and D. Schaefer, (1990). "Methane Production from Municipal Refuse: A Review of
Enhancement Techniques and Microbial Dynamics." CRC Critical Reviews in Environmental
Control, 19(6): 557 - 584.
Barlaz, M., Green, R., Chanton, J., Goldsmith, C., and G. Hater (2004). "Evaluation of a Biologically
Active Cover for Mitigation of Landfill Gas Emissions." Environmental Science & Technology,
38(18): 4891-4899.
Barlaz, M. (2006). "Forest Products Decomposition in Municipal Solid Waste Landfills." Waste
Management, 26(4): 321 - 333.
Bookter, T., and R. Ham (1982). "Stabilization of Solid Waste in Landfills." Journal of Environmental
Engineering, ASCE, 108: 1089-1100.
Christensen, T., Kjeldsen, P., Bjerg, P., Jensen, D., Christensen, J., Baun, A., Albrechtsen, H-J., and G.
Heron (2001). "Biogeochemistry of Landfill Leachate Plumes." Applied Geochemistry, 16: 659-
718.
Chu, L., Cheung, K., and M. Wong (1994). "Variations in the Chemical Properties of Landfill Leachate."
Environmental Management, 18: 105-114.
Clescerl, C., Greenberg, A., and Eaton, A. (eds.) (1999). "Standard Methods for Examination of Water and
Wastewater." American Public Health Association, 20th Edition.
Colberg, P.J. (1988). "Anaerobic Microbial Degradation of Cellulose, Lignin, Oligolignols, and
Monoaromatic Lignin Derivatives." 333-372. In A. J. B. Zehnder (ed), Biology of Anaerobic
Microorganisms, Wiley-Liss, New York.
Conrad, R. (1995). "Soil Microbial Processes Involved in Production and Consumption of Trace Gases."
Advances in Microbial Ecology, 14: 207-250.
Cook, A., Rowe, R., Rittmann, B., and J. VanGulck(2001). "Biofilm Growth and Mineral Precipitation in
Synthetic Leachate Columns," Journal of Geotechnical and Geoenvironmental Engineering,
10(127): 849-856.
133
-------
Davis, M. (1998). "A Rapid Modified Method for Compositional Carbohydrate Analysis of
Lignocellulosics by High pH Anion-Exchange Chromatography with Pulsed Amperometric
Detection (HPAEC/PAD)," Journal of. Wood Chemical Technology, 18(2): 235-252.
Eklund, B., Anderson, E., Walker, B., and D. Burrows (1998). "Characterization of Landfill Gas at the
Fresh Kills Municipal Solid Waste Landfill." Environmental Science and Technology, 32: 2233-
2237.
Eleazer, W., Odle, W., Wang, Y. and M. Barlaz, (1997), "Biodegradability of Municipal Solid Waste
Components in Laboratory-Scale Landfills," Environmental Science and Technology, 31(3): 911 -
197.
El-Fadel, M. (1998). "Leachate Recirculation Effects on Settlement and Biodegradation Rates in MSW
Landfills," Environmental Technology, 20: 121-134.
Haydar, M.and M. Khire (2006). "Leachate Recirculation Using Horizontal Trenches in Bioreactor
Landfills." Journal of Geotechnical & Geoenvironmental Engineering, ASCE, 131(7): 837-847.
Haydar, M.M. and M. Khire (2006). "Leachate Recirculation Using Permeable Blankets in Engineered
Landfills." Journal of Geotechnical & Geoenvironmental Engineering, ASCE, 133(2): 166-174.
Fleming, I., Rowe, R., and D. Cullimore (1999). "Field Observations of Clogging in a Landfill Leachate
Collection System," Canadian Geotechnical Journal: 36, 685-707.
Foose, G., Benson, C., and T. Edil (2002). "Analytical Equations for Predicting Concentrations and Mass
Flux from Composite Liners" Geosynthetics International, 8(6): 551-575.
Gadd, G. (1992). "Metals and Microorganisms: A Problem of Definitions." FEMS Microbiology Letters,
100: 197-203.
Gonzalez, G.; Urrutia, H.; Roeckel, M.; and E. Aspe (2005). "Protein Hydrolysis under Anaerobic, Saline
Conditions in Presence of Acetic Acid." Journal of Chemical Technology and Biotechnology,
80(2): 151-157.
Federal Register (2003). National Emission Standards for Hazardous Air Pollutants: Municipal Solid
Waste Landfills; 68(11): 2227-2242.
Federal Register (2004). Research, Development, and Demonstration Permits for Municipal Solids Waste
Landfills; 69(55): 13242-13256.
Halvadakis, C., Findikakis, A., Papelis, C., and J. Leckie (1988). "The Mountain View Controlled Landfill
Project Field Experiment." Waste Management and Research, 6: 103-114.
Ham, R. and T. Bookter (1982). "Decomposition of Solid Waste in Test Lysimeters," ASCE, J. of
Environmental Engineering, 108(EE6): 1147 -70.
Hansen, K., Angelidaki, I., and B. Ahring (1998). "Anaerobic Digestion of Swine Manure Inhibition by
Ammonia." Water Research, 32(1): 5-12.
Hartz, K., Klink, R., and R. Ham (1982). "Temperature Effects: Methane Generation from Landfill
Samples." Journal of Environmental Engineering, 108(4): 629-638.
Hashimoto, A. (1986). "Ammonia Inhibition of Methanogenesis from Cattle Wastes." Agricultural Wastes,
17: 241-261.
134
-------
Hossain, M, Gabr, A., and M. Barlaz, M. (2003). "Effect of Decomposition and Stress on Creep
Compression Index of Municipal Solid Waste in Landfill bioreactors." The 12th Pan-American
Conference for Soil Mechanics and Geotechnical Engineering and the 39th US Rock Mechanics
Symposium, Soil and Rock America 2003, June 22-26, Cambridge, Massachusetts, USA: 1435-
1440.
Hugenholtz, P.; Goebel, B.; and N. Pace. (1998). "Impact of Culture-Independent Studies on the Emerging
Phylogenetic View of Bacterial Diversity." Journal of Bacteriology, 180(18): 4765-4774.
Hughes, M., and K. Poole (1989). Metals and Micro-organisms. Chapman and Hall, New York
Ikem, A.; Osibanjo, O.; Sridhar, M.; and A. Sobande (2002). "Evaluation of Groundwater Quality
Characteristics near Two Waste Sites in Ibadan and Lagos, Nigeria." Water Air and Soil Pollution,
140 (1-4): 307-333.
Intergovernmental Panel on Climate Change (2001). Climate Change 2001: The Scientific Basis,
Cambridge University Press, United Kingdom.
Jain, P., Powell, J., Townsend, T., andD. Reinhart (2005). "Air Permeability of Waste in a Municipal Solid
Waste Landfill." Journal of Environmental Engineering, ASCE, 131(11): 1565-1573.
Jain, P., Powell, J., Townsend, T., and D. Reinhart (2006). "Estimating the Hydraulic Conductivity of
Landfilled Municipal Solid Waste Using the Borehole Permeameter Test." Journal of
Environmental Engineering, ASCE. 132(6):645-652.
Jones, K., Rees, J., and J. Grainger, (1983). "The Application of Enzyme Activity Measurements to a Study
of Factors Affecting Protein, Starch and Cellulose Fermentation in Domestic Refuse," Eur. J.
Appl. Microbiol. Biotechnol., 18: 181-185.
Jjemba, P. (2004). Environmental Microbiology: Principles and Applications. Science Publishers: Enfield,
NH.
Kim, J., and F. Pohland (2003). "Process Enhancement in Anaerobic Landfill bioreactors." Water Science
and Technology, 48(4): 29-36.
Kjeldsen, P., Barlaz, M., Rooker, A., Baun, A., Ledin, A., and T. Christensen (2002). "Present and Long-
Term Composition of MSW Landfill Leachate: A Review." Critical Reviews in Environmental
Science and Technology, 32(4): 297-336.
Koerner, G., and R. Koerner (1995). "Long-Term Permeability of Granular Drainage Material."
Proceedings Seminar Publication Landfill Bioreactor Design and Operation, Wilmington, DE.
Krung, M., and R. Ham (1991). "Analysis of Long-Term Leachate Characteristics." Sardinia International
Landfill Symposium, Cagliari, Italy: 9-12.
Landva, A. and J. Clark (1990). "Geotechnics of waste fills." ASTM STP1070 Geotechnics of
Waste Fills: Theory and Practise: 86-103.
Lay, J., Li Y., and Tatsuya, N. (1998). "Mathematical Model for Methane Production from Landfill
Bioreactor." Journal of Environmental Engineering. 19:730-736.
McCarty, P. (1964). "Anaerobic Waste Treatment Fundamentals." Public Works, 9-12.
Mehta, R., Barlaz, M. A., Yazdani, R., Augenstein, D., Bryars, M. and L. Sinderson (2002). "Refuse
Decomposition in the Presence and Absence of Leachate Recirculation." J. Environmental
Engineering, 128(3): 228-236.
135
-------
Olayinka, A. (2003). "Microbial Respiration and Nitrate Production in Cowdung Amended Alfisol as
Affected by Temperature and Moisture Content." Discov. Innov., 15(1-2): 75-82.
Parkin, G., and F. Owen (1986). "Fundamentals of Anaerobic Digestion of Wastewater Sludges." Journal
of Environmental Engineering, 112(5): 867-920.
Petterson, R., and V. Schwandt (1991). "Wood Sugar Analysis by Anion Chromatography." Journal of
Wood Chemical Technology, 11: 495-501.
Pohland, F. (1975). "Sanitary Landfill Stabilization with Leachate Recycle and Residual Treatment." EPA-
600/2-75-043, U.S. Environmental Protection Agency, Cincinnati, Ohio.
Pohland, F., and S. Harper (1987). "Retrospective Evaluation of the Effects of Selected Industrial Wastes
on Municipal Solid Waste Stabilization in Simulated Landfills." Georgia Institute of Technology,
School of Civil Engineering, Atlanta, Georgia, 251 pp.
Pohland, F., Cross, W., Gloud, J., and D. Reinhart (1993). "Behavior and Assimilation of Organic and
Inorganic Priority Pollutants Co-disposed with Municipal Refuse." EPA/600/R-93/137a, Risk
Reduction Engineering Laboratory Office of Research and Development, Cincinnati, Ohio.
Pohland, F., and B. Al-Yousfi (1994). "Design and Operation of Landfills for Optimum Stabilization and
Biogas Production." Water Science and Technology, 30(12): 117-124.
Powell, J., Jain, P., Kim, H., Townsend, T., and D. Reinhart (2006). "Changes in Landfill Gas Quality as a
Result of Controlled Air Injection." Environmental Science and Technology, 40: 1029-1034.
Price, G., Barlaz, M. and G. Hater, (2003), "Nitrogen Management in Bioreactor Landfills." Waste
Management, 23(7): 675 - 88.
Reinhart, D., and B. Al-Yousfi (1996). "The Impact of Leachate Recirculation on Municipal Solid Waste
Landfill Operating Characteristics." Waste Management and Research, 14: 337-346.
Reinhart, D., and T. Townsend (1998). Landfill Bioreactor Design and Operation, Lewis Publishing, New
York.
Rittmann, B., Banaszak, J., Cooke, A., and R. Rowe (2003). "Biogeochemical Evaluation of Mechanisms
Controlling CaCO3(s) Precipitation in Landfill Leachate Collection Systems." Journal of
Environmental Engineering, 129(8), 723-730.
Ress, B., Calvert, P., Pettigrew, C. and M. Barlaz (1998). "Testing Anaerobic Biodegrability of Polymers in
a Laboratory-Scale Simulated Landfill," Environmental Science & Technology, 32(6): 821 - 827.
Rhew, R. and M. Barlaz, (1995), "The Effect of Lime Stabilized Sludge as a Cover Material on Anaerobic
Refuse Decomposition," Journal of Environmental Engineering, 121(7): 499 - 506.
Rowe, K., Armstrong, M., and D. Cullimore (2000). "Particle Size and Clogging of Granular Media
Permeated with Leachate." Journal of Geotechnical and Geoenvironmental Engineering, 126(9):
775-786.
Schroeder, P., Dozier, T., Zappi, P., McEnroe, B., Sjostrom, J. and L. Peton (1994). "The Hydrologic
Evaluation of Landfill Performance (HELP) Model: Engineering Documentation for Version 3."
EPA/600/R-94/168b, US. Environmental Protection Agency, Risk Reduction Engineering
Laboratory, Cincinnati, OH.
Sheridan, S (2002). Modeling Solid Waste Settlement as a Function of Mass Loss, Master Thesis,
University of Florida. Gainesville FL.
136
-------
Singh, S. and B. Murphy (1990). "Evaluation of the Stability of Sanitary Landfills." Geotechnics of Waste
Landfills - Theory and Practice" Landva, A. and Knowles, G.D., Editors, ASTM Special
Publication 1070, proceedings of a symposium held in Pittsburgh, Pennsylvania, USA: 240-258.
Sponza, D.; and O. Agdag (2004). "Impact of Leachate Recirculation and Recirculation Volume on
Stabilization of Municipal Solid Wastes in Simulated Anaerobic Bioreactors." Process Biochem.,
39(12): 2157-2165.
Staley, B., Xu, F., Cowie, S., Barlaz., M., and Hater, G. (2006) "Release of Trace Organic Compounds
During the Decomposition of Municipal Solid Waste Components". 40:5984-5991
Tchobanoglous, G., Theisen, H., Vigil, S. (1998). Integrated Solid Waste Engineering Principles and
Management Issues. New York: McGraw-Hill.
Tiquia, S. M.; Wan, J. H. C.; Tarn, N. F. Y. (2002). "Dynamics of Yard Trimmings Composting as
Determined by Dehydrogenase Activity, Atp Content, Arginine Ammonification, and Nitrification
Potential." Process Biochem., 37(10): 1057-1065.
Townsend, T., Miller, W., Lee, H., and J. Earle (1996). "Acceleration of Landfill Stabilization Using
Leachate Recycle." Journal of Environmental Engineering, 122(4): 263-268
Townsend, T., and L. Miller (1998). "Leachate Recycling Using Horizontal Injection." Advances in
Environmental Research. 2(2): 129-138.
U.S. Department of Commerce, National Climatic Datacenter (2004). Climatology of the United States No.
20, 1971-2000, Louisville Standiford AP, KY Station.
U.S. EPA (1987). Guidance Manual for Preventing Interference at POTWs, September 1987.
U.S. EPA (1993). "Solid Waste Disposal Facility Criteria." U.S. EPA 530/R-93/017; U.S. EPA: Cincinnati,
Ohio.
U.S. EPA (1996). Test Methods for Evaluating Solid Waste: Physical/Chemical Methods, EPA Office of
Solid Waste.
U.S. EPA (1998). Compilation of Air Pollutant Emission Factors, 5th Edition: EPA AP-42; U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards, Supplement E,
Section 2.4.
U.S. EPA (2002a). State of the Practice for Bioreactor Landfills. U.S. EPA 625/R-01/012; U.S. EPA:
Cincinnati, Ohio.
U.S. EPA (2002b). Assessment and Recommendations for Improving the Performance of Waste
Containment Systems. EPA/600/R-02/099, USEPA ORD, Cincinnati OH, USA, December 2002.
U.S. EPA (2003a). Municipal Solid Waste, Generation, Recycling, and Disposal in the United States; Facts
and Figures for 2003, U.S. EPA 530/F-05/003; U.S. EPA: Washington, DC.
U.S. EPA (2003b). Landfills as Bioreactors: Research at the Outer Loop Landfill, Louisville, Kentucky.
First Interim Report, U.S. EPA 600/R-03/097; U.S. EPA: Cincinnati, Ohio.
U.S. EPA (2004). Monitoring Approaches for Landfill Bioreactors, U.S. EPA 600/R-04/301; U.S. EPA:
Cincinnati, Ohio.
U.S. EPA (2005). Measurement of Fugitive Emissions at a Bioreactor Landfill, U.S. EPA 600/R-05/096;
U.S. EPA: Washington, DC.
137
-------
U.S. EPA (2005). First-Order Kinetic Gas Generation Model Parameters for Wet Landfills, U.S. EPA
600/R-05/072; U.S. EPA: Washington, DC.
U.S. EPA (2005b). Example Moisture Mass Balance Calculations for Bioreactor Landfills, U.S. EPA
456/R-05/004; U.S. EPA: Research Triangle Park, North Carolina.
U.S. EPA (2006a). www.epa.gov/reg3wcmd/solidwastesummarv.htm. Accessed 6/30/2006.
U.S. EPA (2006b). http://www.epa.gov/eogaptil/module2/review2.htm. Accessed 6/28/2006.
Wang, Y., Byrd, C., and M. Barlaz (1994). "Anaerobic Biodegradability of Cellulose and Hemicellulose in
Excavated Refuse Samples using a Biochemical Methane Potential Assay." Journal of Industrial
Microbiology, 13: 147-153.
Wu, B., Taylor, C., Knappe, D., Nanny, M., and M. Barlaz (2001). "Factors Controlling Alkylbenzene
Sorptionto Municipal Solid Waste," Environmental Science & Technology 35(22): 4569-4576.
Yamaguchi, M., Minami, K., Tanimoto, Y., K. Okamura (1989). "Effects of Volatile Fatty Acids on
Methanogenesis of Methanol and of Pregrowth with Methanol on Acetate Utilization by
Methanogens." Journal of Fermentation Bioengineering: 68: 428-432.
Zoetemeyer, R.; Matthijsen, A.; Cohen, A.; and C. Boelhouwer (1982). "Product Inhibition in the Acid
Forming Stage of the Anaerobic Digestion Process." Water Research, 16: 633-639.
138
------- |